Amorey Gethin - Antilinguistics - A Critical Assessment of Modern Linguistic Theory and Practice - Intellect Books (1990)

cover next page >
title: Antilinguistics : A Critical Assessment of Modern Linguistic

Theory and Practice
author: Gethin, Amorey.
publisher: Intellect Books
isbn10 | asin: 1871516005
print isbn13: 9781871516005
ebook isbn13: 9780585209807
language: English
subject Linguistics.
publication date: 1990
lcc: P121.G47 1990eb
ddc: 410
subject: Linguistics.
cover next page >

< previous page cover-0 next page >
Antilinguistics
A Critical Assessment of Modern Linguistic Theory and Practice
Amorey Gethin

First published in Great Britain in 1990 by

Intellect Limited
Suite 2, 108/110 London Road, Oxford OX3 9AW
Copyright © 1990 Intellect Ltd.
Extracts from Linguistics by David Crystal (London: Penguin Books, 1981), copyright © 1971 David Crystal, reproduced by
permission of Penguin Books Ltd.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by
any means, electronic, mechanical, photocopying, recording, or otherwise, without written permission.
Consulting editor: Masoud Yazdani
Cover design: Steven Fleming
British Library Cataloguing in Publication Data
Gethin, Amorey
Antilinguistics: A critical assessment of modern
linguistic theory and practice.
1. Linguistics. Theories
I. Title
410.1
ISBN 1-871516-00-5
Printed on 100% recycled paper by Billings & Sons Ltd, Worcester

< previous page page_i next page >
Page i
Contents
Acknowledgements iii
Introduction 1
Chapter 1
What Language is 5
Chapter 2
The Emptiness of Analysis 11
Chapter 3
Chomskyan Mistakes Made Plain 40
Chapter 4
Unapplied Linguistics 60
Chapter 5
The Fantasy of Structure 93
Chapter 6
More Fantasy of Structure 114
Chapter 7
Ungenerative Grammar 131
Chapter 8
Deep Confusion 151
Chapter 9
Being Able to Use Language 170
Chapter 10
Thinking and Language 194
Chapter 11
Language the Corrupter 220
Chapter 12
Trying to Speak the Truth 242
Chapter 13
Slavery to Authority and the Word 254
Bibliography 260
Index 263
< previous page page_i next page >

< previous page page_ii next page >
intentionally left blank

< previous page page_iii next page >
Page iii
Acknowledgements
I give my warm thanks to the former colleagues, and other friends, with whom over so many years I have spent so many
stimulating hours debating language. Among these I would particularly like to mention Michael McCarthy and David Bond,
who were unfailing in their sympathetic encouragement of my urge to question. David I thank too, very specially, for always
thinking of me whenever he came across something he thought might interest or provoke me.
As much as anybody I thank my son, Terence Gethin, for all the care he devoted to his judgements on the manuscript, and for
his objective criticism, advice, and rationality.
And without the constant support and concern of my wife, Mieko Suzuki-Gethin, I should never have completed the task.
I owe a great debt to John Lennox Cook for providing, over many years, a place of work without fashionable preconceptions
where I could think freely about language.
My gratitude also goes to Erich Müller and Irene Marti, to Agnese, Arianna, Edo, Ida, Luca and Silio Masi, to Paul Mlynek, to
Gerhard Prawda, Ingrid Freidl, Felix, and Rea, and to Giorgia, Metello and Paola Tacconi, for making life so good for us during
what were perhaps the most important stages of writing this book.
I dedicate this book to Pod and to the memory of Tora and Bobi.
AMOREY GETHIN
SASSO PISANO
JUNE 1989
< previous page page_iii next page >

< previous page page_1 next page >
Page 1
Introduction
For thousands of years it was the priests of religion who were the experts not only on the nature of the universe but also on the
nature of man. So it was largely accepted that it was the priests, too, who should tell humans how they should act.
Over the greater part of the world that has changed. The new experts on the nature of humankind and on human needs are the
'social scientists' of psychology, sociology, economics, ethology, linguistics. They may disagree among themselves, just as the
priests did. But just as the priests' authority was almost universally respected whatever they said at different times and in
different places, so the 'social scientists' are widely accepted as the only proper source of understanding, their way of
discovering truth as the only possible way.
From the rule of 'infallible' science there seems to be no appeal. One can reject the old dogmas. They were based on mere
superstition and wishful thinking. But the new, those one must accept, because they are based on evidence. They are scientific.
And everything scientific is true. 'Old' religions normally only claim revelation and faith as authority. Rebels can, at worst,
claim a different revelation, a different faith. They can even appeal to reason. But that which claims to be science is an even
more unbending master than intolerant religion, for it is knowledge, the final truth, and from that there can be no appeal. So the
new religion of Marxism, for example, or the new discovery of linguistics, had to be the truth, because they are based on
scientific analysis.
Today the 'scientific' analysers are the academics. Truth comes from them, practically only from them. They are the
professionals who know what they are talking about. So if there is debate and questioning it is debate and questioning that goes
on almost entirely within that circle of academic experts. Outsiders are untrained and so unfit to have opinions worth listening to
- so practically nobody outside ever tries to interfere, for any who do will almost certainly not have any notice taken of them. In
fact, I think very few outside the universities even dream of speaking up, because they feel they would have to do so on the
academics' own terms, and so would make amateurish fools of themselves.
Yet the power of the half-secret lores of the academic world needs to be questioned. Those lores capture the minds of nearly all
those who come in contact with them, those who are considered the most intelligent members of the community, and so those
who to a large extent are given the power to design how our lives should be, now and in the future.

Page 2
But today there is so much 'literature', there is so much 'research' done, that how does one know what to have faith in? How does
one know who to trust? Respect for the literature and for research only means putting oneself deeper into the hands of the
experts, happily leaving it to others, a special caste, to tell us how things are, and how, in the end, therefore, we need to behave.
We need less reading and more thinking.
If people are not going to surrender completely to authority they must decide for themselves on the basis of what they have the
time and ability to read or hear, and understand, and must be respected for conclusions they make in this way. Experts in any
field that those experts think affects other people's lives have a duty to explain their conclusions and the basis of their
conclusions in such a way that makes them open to judgement by people who are not experts. What matters in the end, or what
should matter, is the reality of what people in general actually believe and think. The arrogance of any other approach appears
even greater when one remembers that the particular dogmas one will be expected to bow to depend on the chance of things like
birthplace. The only thing all the experts everywhere and at all times will agree upon is their superiority as experts.
The abstractions that are 'over the heads' of all but an elite few are probably not very dangerous. A far more serious matter is the
'wisdom' that millions of people think they do understand, things therefore that are widely believed with little or no question
precisely because they come from experts.
Now we need to question that 'wisdom' and give their rightful place to plain truths and rationality - the plain truths almost
anyone can discover with a few minutes' straight thinking based on elementary knowledge. What is really important for us to
look at is the effect of research and the literature on the attitudes and beliefs of people generally. Independent of what has gone
before, free of the academic traditions and the literature and the preconceived notions, in this book I want to bring some of the
plain truths to the surface in their own right.
The particular expert wisdom that I want to question is the 'science' known as linguistics.
We know of no time when humans have not been under the sway of language. But in recent years the hold of language has been
strengthened by the rise to power of linguistics. Language has for so long determined human beings' relationships with each
other and to the rest of the world.

Page 3
Now linguists (properly: linguisticians)1 toll us 'scientifically; that we should indeed respect language, as mankind's first
principle. They hail language as man's unique talent, the key to our personality; and so linguistics is presented as a young
science of great revelation and is allowed to lead much of what passes as modem superior thinking.
To an 'ordinary' person linguistics may not seem very important, partly at least because of its very abstractness and apparent
irrelevance to the real world. But the claims made for linguistics are grand. It is being said that linguistics makes fundamental
contributions to psychology, philosophy, biology and our whole view of ourselves. This, it is said, is now widely recognized. A
journalist reported that two "sober psychologists" told him, independently, that their first encounter with the work of Noam
Chomsky, the most famous living linguist, had had the quality of a mystical experience, a sense of sudden revelation and
insight. Chomsky has been compared to Newton as the initiator of a comparable upheaval in our view of things. It is assumed
unquestioningly that language and thinking are linked together - if indeed they are not the same thing - and that the study of
language will help us to understand the human mind.
It is my purpose in this book to show that modern linguistics presents a false picture of the nature of humans and of language
itself. I shall try to show that language is not a wonderful natural asset; it is an artificial device that constantly misleads us and
does us great harm; and the modern way of studying language is itself harmful because it enhances the reputation of language
and sustains corrupt ways of thought.
I am sad that so much of my criticism should be concentrated on Noam Chomsky. He seems to me a person of conscience and
integrity; I deeply respect his efforts in the cause of good and decency where politics are concerned, efforts that I cannot hope to
equal. And he has attacked the regime of experts and urged people to open their minds and question accepted attitudes,
recognize the many absurdities that pass as learned wisdom.
Yet at the same time he has erected his own expert abstractions, built up a system that is about as expert and learned as one can
get. For all that he claims that he too is an opponent of the cult of the expert, in practice he has for many years now been one of
the world's great champions and supports of that cult, inspiring at least a generation with faith in the cult.
1 Linguisticians have usurped the name "linguist", which used to mean a person good at foreign languages. Today a
person described as a fine linguist is very likely nothing of the sort. However, to avoid possible irritation on any reader's
part I have surrendered to the convention and I shall refer to linguisticians as "linguists" throughout.

Page 4
And the world is sadly such that if he did not hold such a position the influence of his political ideas would be even less than it
is.
It seems the nonsense of one university professor (for example, the behaviourism of Skinner) has to wait for another university
professor (in this case Chomsky) before it can be refuted. And then the nonsense that that professor proclaims has to wait for yet
another professor to refute it. The obvious is clearly not its own justification; it has to be given the certificate of an eminent
authority before it can be accepted. But by that time it is no longer seen to be obvious; it has become the profound insight of a
pioneer thinker. For quite a long time, for instance, philosophers saw language as consisting mainly or even entirely of
utterances that are true or false. Any innocent person knows this is not so, knows that people do other things when they speak
besides making true or false statements. Yet such a sensible opinion did not become worthy of respect in the palaces of learning
until a professional philosopher had questioned the old orthodoxy.
Equally, it is no good expressing a truth in one sentence; it usually has to have at least a whole book written about it before
anyone takes any notice!
A lot of what I say will be obvious; much will have been said before by others. But here too it is the effect that is important, and
if the saying of truths has had little effect, then they must be said again, but without pretensions, until they do have effect.

Page 5
1
What Language Is
It is not my purpose here to propound yet another 'theory' of language. I want only to affirm what I see as the essential principle
of languages. I want to encourage 'ordinary' people, those unconditioned by philosophy or linguistics or psychology, or their
education in general, to trust in their own sense of how languages work. And I think that sense is of something much as I
propose here. So this is no exact and worked out analysis of language, no definition, no classification of parts. It is only an
invitation to recognize the practical reality of languages as humans actually experience them; and an appeal to any wavering
initiates or novices in linguistics to start again from the beginning without preconceptions.
Languages are sets of meanings.1 Those meanings are an attempt to reflect human experience of reality. That reality, of course,
includes human thoughts and feelings and fantasies as well as what 'objectively' exists and happens.
Languages are only an attempt to reflect experience. As I shall try to show later in the book, they cannot do this properly.
But as they do try to reflect life and the world, the various meanings that compose them fit together in accordance with the logic
of life, not in accordance with some abstract, formal logic. Thus the basic principle for the working of language is very simple,
and that is perhaps why almost all humans master it so easily.2 But the detailed relationships of the meanings of a language are
very complicated, because they try to reflect the even more complicated relationships of life. The only true description of a
language is the language itself. Anything else is just a game.
Those human thoughts and feelings expressed in the meanings of a language include varying human reactions to, and the
varying angles from which humans see what exists and what happens. In English, for instance,
1 I have been to London.
and
1 Some of those of a philosophical bent may complain straight away, "What does he mean by meaning?" Yet the
philosophically uninitiated will have little trouble with the term. I think they will know almost exactly what I am talking
about. And it is not by definition that the sticklers for the discipline of definition understand their own word "mean" when
they demand "What do you mean Coy meaning)?" If they did not already know they could not ask the question.
"Meaning" certainly has more than one meaning. But when people start arguing about the meaning of meaning there is
immediately a prime example of the inadequacy of words, of the futile slavery to language itself that I criticize in
Chapters 11 and 12.
2 But see Chapter 9.

Page 6
2 I went to London.
may reflect one and the same event. What actually happens, what the person does, is exactly the same in the two sentences. But
the angles from which the event is seen are quite different, and therefore the sentences mean different things. The failure to
grasp this distinction between meaning and what actually happens is one of the main causes of the confusion linguists have
fallen into.1
The meanings of a language can be divided very roughly into two sorts. Think, as an example, of the words 'asked', 'kissed',
'saw'. They all mean the same in a 'general' sense, in that they all have the same general meaning function, that is, they all refer
to actions. They also mean the same, in another 'general' sense, in that they are all about the past. But they have quite different
meanings in what one could call a 'specific' sense. There are 'general' meanings of actions, things, time, place, relationship,
cause, purpose and so on. 'Cat', 'chair', 'affection', 'milk' all have the same 'general' meaning in that they all refer to things, but
they have different 'specific' meanings.
Some meanings fit together sensibly, others do not. 'Have' and 'done', for example, both have (or can have) 'general' time
meanings and those particular 'general' time meanings can be fitted together sensibly. But 'do', although it too can have a
'general' time meaning, will not fit sensibly after 'have' because it has the wrong sort of 'general' time meaning.
The 'general' meanings in 'I have come' do not clash in any way with the 'specific' meaning of 'on foot', but the 'general' time
meaning of ''I have come" does clash with the 'specific' time meaning of "yesterday". The 'general' meanings of the two words
"can kiss" fit, but the 'general' meanings of each of the two words in "can kissed" or "she kiss" cannot be put together sensibly.
It goes almost without saying that one and the same piece of language can have more than one 'general' meaning. "Cat" and
"cats" not only have the same 'specific' meaning (they both mean the same sort of creature); they also have the same 'general'
meaning in that they will both fit, for example, "I love my..." But they have different 'general' meanings in that one means a
single cat and the other means more than one cat, and so on.
1 See Chapter 8.

Page 7
I have persisted with the use of inverted commas for 'specific' and 'general' in order to emphasize that I am not defining new
linguistic categories. I do not want to defend these particular terms as the accurate or proper ones to describe what I am talking
about. In fact, I know that they could not be so defended, and I should be worried if they were such that they easily seduced
people into a debate about their accuracy and 'rigour'. We are not here playing yet another definition game. I use these terms
here only as an aid to recognizing the existence of meaning in all language, in every piece of language. The particular terms
used to arrive at this insight don't matter. The important thing is: "Do you understand the reality I am referring to?" The irony is
that most children probably would, and many adults will insist they don't.1
It should be obvious too that there is no rigid or clear border between 'specific' and 'general' meanings. What category should we
put words like "by" or "on" in, for instance? Well, it doesn't matter. A language is itself, not a classification system.
The meaning of a sentence is the result of the combination of particular 'specific' meanings with particular 'general' meanings
and other particular 'specific' meanings, either in the same sentence or other sentences; and often with particular things known
but not said by the speaker2. For example:
3 He's made a bed.
4 He's made his bed.
5 He's made a mistake.
The "a" in 3, combined with "bed", gives "made'' the sense "built". In 4 the change to "his", still combined with 'bed", gives
"made" the meaning of "fixed the bedclothes". Combining "mistake" with "made" produces a third sense: "committed".
6 Have you seen the cats?
7 Have you seen the cats' supper?
In the spoken forms of 6 and 7 "cats(')" sounds the same, but the different meanings are made clear immediately by the absence
in 6 and the presence
1 See Chapter 12.
2 See further Chapter 2, p.25 and p.32, and for the contribution to meaning of 'unspoken knowledge' Chapter 5, pp.112-113.

Page 8
in 7 of "supper". In other words, the particular 'general' meaning of "cats(')" is shown by the presence or absence of another
'general' meaning, a 'thing'.
8 I'm working.
9 I'm working tomorrow.
10 "What are you doing tomorrow? I'm working."
8 has present meaning. The addition of the particular 'specific' meaning "tomorrow" gives "I'm working" in 9 future meaning.
And in 10, "I'm working'', although identical with 8, has future meaning as well, from the "tomorrow" of the question.
The orders in which the pieces are arranged in a language are also meanings. But how important this ordering is for meaning
differs from language to language. It is obviously very important in a language like English, and not always so important for a
language with cases, like Latin.
All the features of a language, whether the pieces themselves, or the order the pieces are arranged in, are conventions. As far as
the pieces are concerned, the conventions are the meanings, the meanings are the conventions.
But the conventions for the order of the pieces are not always essential to the meaning in that way. Many of course are. There is
a neat basic example in Swedish, for instance: "bok/en" is not at all the same as "en/bok" ("the book" as opposed to "a book").
"The mouse chased the cat" as opposed to "The cat chased the mouse" is an obvious example of a different sort.
On the other hand, many conventions of order are conventions and nothing else, although one must obey them if one is to be
judged to speak the language 'correctly'. "After dark the cat went out" is perfectly clear in meaning. There is no need, just
because the sentence begins with "After dark...", to say "After dark went the cat out". Yet this is exactly what other Germanic
languages do the equivalent of.
These conventions for the ordering of the pieces of a language are the only things that could remotely deserve the name
'structure', and it is very questionable whether that is a good name even for them. Conventions of order haven't much in
common with what is normally implied by the word "structure", and the sooner we drop this word of fashion the better.

Page 9
People learn all the meanings of a language by observation and imitation.1 There is no other way. But when people fit the
pieces of meaning of their language together, they follow, as I have said earlier, the logic of life. Put it another way: they join
individual meanings together so that they make larger meanings.
There is one detail that is important to understand. Originally all individual pieces of meaning were almost certainly joined
together logically, given the meaning that each one had separately. But in ways that I don't think anyone understands the logic
often becomes forgotten. So there are in English, for example, expressions such as "none/the/less, ex/plain, make do, would
rather (do)" where the 'inside' logic has been lost. In fact there is no longer any 'inside' to have a logic; such expressions have
become lumps of meaning in their own right; they are now single separate pieces of meaning.
There is a good example of this psychology, I think, in the case of "it's" in a context such as "My pen was here. It's gone? What
do most English-speakers think they are really saying when they say this? "It is'' or "It has"? It is no use asking people what
they intend "It's" to represent, because they will immediately start thinking what they ought to be really saying, not what they
are actually saying. And here in fact is the whole point: most people when they say "it's" in such a context are almost certainly
really saying "It's" and nothing more. They do not 'really' mean anything else at all. "It's" is an independent single whole, and
there is no longer any need or room for 'inside' logic in it.
This explanation of language will to many seem ridiculously simple and naive. Yet even if some of the far-fetched constructions
of philosophers -linguistic or otherwise - about words, and their relations to things, are in some abstract sense true, they cannot
give us any effective understanding of language. It seems to me that humans would not be able to use the instrument of
language at all if it was not what the innocent and naive think it is. To say they don't realize what they are using is like saying
that a person who thinks he is using a screwdriver isn't really, he is using a hammer. The successful use of a screwdriver or of a
language depends on recognizing them as what they really are. I think it would be irrelevant to refer to the fact that most of us
have little or no idea of what is happening when, for instance, we pick up a glass, that none of us are actually aware of the
process at the time, and that the scientist has to find out and tell us. One
1 This is nothing to do with the behaviourist absurdities of Skinner etc. I take up the question of an inborn language
ability later (see Chapter 9).

Page 10
surely has to distinguish between on one hand a purely internal process involving functions of which some are not under
conscious control, and on the other a deliberately used instrument which is in a sense 'outside' us. Where language is concerned
the equivalent of the processes of the brain and muscles that control picking up a glass are the processes of brain and muscle
that control the using of language. I have been trying to explain the 'instrument' itself, not the processes by which we control it.1
So I believe it is roughly in the simple and naive way I have described that children, for example, experience language. They do
not, of course, articulate that experience, and they do not need to. Neither they nor we need a 'theory' of language, yet another
grand system of carefully defined and related concepts. It is only later, when the corruption of education sets in, that they
become confused by the abstruse, complicated and tortuous views of their elders.
I now want to examine some of those abstruse, complicated, tortuous and artificial views.
1 Many, if not most, linguists would deny that language is any sort of 'instrument', an invention 'outside' us. See Chapter
9.

Page 11
2
The Emptiness of Analysis
I can see three important purposes in thinking about language.
1 For me by far the most important is that it may help us to understand the relationship between thought and language, how
language influences the way people think and the way they act.
2 Understanding the true nature of language, or at least what language is not, is essential for discovering the processes of the
brain in controlling language.
3 That may in turn help physicians and others to help people who have defects (as a result of birth or accident) that make it
difficult or impossible to use language normally.
On the other hand I do not think it is likely that knowing about the way the brain controls language is going to help in
understanding the effects of language and thought on each other, any more than, say, knowing how hearing works is likely to
help give us better understanding of the psychological effects of music.
There are other purposes in studying language. Language studies can be interesting or entertaining, but that does not make them
important. Some people, for instance, search for the origin of language. This is a very interesting question, but it is not
important if it does not serve any of the three purposes I have mentioned first. Some people are fascinated by quirkish details
and accidents of language, but study of these has no more right to pretensions than the collecting of match box labels. Such
study may be entertaining but it is not important.
Modern linguists, though, are obsessed with describing language and languages.
Description of a language is of course very reasonable, and useful, if it is for people who have no opportunity to go to the
country whose language they want to learn, or who want to learn something of the language before they go; or if there is a risk
of a language dying out before it is recorded in some way; and if by description one only means an account of the characteristics
of the language that are different from other languages, together with a bilingual dictionary.
But the linguists want to show themselves very scientific, and so make their descriptions sound impressive by using phrases
such as "mathematical precision" or the "rules that generate all but only the

Page 12
sentences that constitute the language"; and like most of their kind involved in so-called sciences they take yet one more step
backwards to double the barrenness, and debate the theory of linguistic theory, abstractions about abstractions.
Their methods and ideals must lead, in the complete 'grammars' they aspire to, to descriptions like "always is used with past
tenses, present tenses, future tenses; often is used with past tenses, present tenses, future tenses; yesterday is used with past
tenses, but not with present tenses, future tenses; tomorrow is used with future tenses, but not with past tenses, present tenses; in
the 19th century is used with past tenses, but not with present tenses or future tenses; in the 20th century is used with present
and future tenses, but not with past tenses..." Or "hear, see, watch etc., etc....can be used in Verb + Noun Phrase + Verb
sentences (e.g. I saw her go, She heard me coming), but not bathe, hit, lock, put, save, waste, wind, write etc., etc., etc. Or
"There are no comparative or superlative forms of all, no, each, every,... nor of always, sometimes, never,...".
These are my examples, not theirs. Theirs they will no doubt consider far more 'precise' and 'scientific'.1 All the same, in
principle this sort of stuff is the logical result of their proposals and demands for 'scientific' precision and thoroughness.
Chomskyans and others no doubt defend such description, however, on the grounds that through 'thorough' grammars on these
lines we may be able to discover the principles underlying the 'universal grammar' that many modem linguists believe all human
beings are born in possession of. But even on the assumption of such universality, drawing up exhaustive rules in this way has
both a logical flaw and practical dangers.
The procedure is logically flawed because all normal human beings already know (at whatever age they understand the
experience concerned) which realities in life fit together, and which don't. The language they use, the words they combine and
the ones they don't, are the result of that knowledge; that is, they are aware of the relationships of reality, and so naturally use
the pieces of language that refer to the realities to express those relationships. The linguists' way of doing things is like a
carpenter who knows all about the features of wood, and who, on the basis of that knowledge, devises various tools to work the
wood with, and then turns round and says "Ah, look at my tools! I will examine them closely in order to discover the features of
wood." It is typical of the systematizing mania that pretends to discover new profundity in what everybody knows already.
1 See an example of Chomsky's in Chapter 6, p. 126.

Page 13
One of the practical evils of this exhaustive description is that it may lead to utterly useless lists and other analyses for students
of foreign languages. I hope that such analyses are no worse than useless; that they do not do positive harm by making students
actually learn them, as is presumably their authors' intention. My own experience suggests there are few students who seriously
think of doing such a silly thing, and confusing themselves into frustration and failure.1
A probably far more dangerous effect of the method in practice is the rigidness it leads to. Linguists declare for instance (their
examples this time, not mine) that sentences such as
1 Johns are here.
2 I want me to go.
are not 'well formed', i.e. are not grammatical; and Chomsky calls
3 I'll kill us.
an oddity.
There is nothing wrong linguistically with any of these sentences. The reality they describe may be uncommon, or may be
thought bizarre; that is all. Perhaps in the case of I there is a peculiar competition between teams made up of people with the
same first names:
1 Johns are here today. Tomorrow it's the turn of Henrys. On Friday Bills will be coming.
Again (as regards 2) it is perfectly rational, acceptable, comprehensible, grammatical, recognizable, or what you will, to convey
(perhaps comic) emphasis by declaring, for example:
2 I don't want him to be with her. I want me to be sitting there with her arms round me.
As for 3, it would be a wholly normal thing for an honest motorist to say something like:
3 Whoops! I'll kill us if I'm not careful. I'd better ease up on the speed.
1 But see Chapter 4, p.76ff.

Page 14
There is nothing odd about the language in any of these sentences.
The whole point of language is that we use its pieces, its meanings, to say what we mean, whatever that may be. The
systematizers, the descriptionizers, have become so rule-bound that they often lose the link with reality which is the only excuse
for language, and often mistake situations in life that are 'normal' for rigid rules of language, and appear to think that language
is ungrammatical if it expresses the abnormal in life.1 This does not just lead them to make mistakes like the ones I have shown
above. They are giving the world a completely false picture of the basic nature of language. That dead rule-boundness has
meant that their imagination has died too; that, ironically, they have lost sight of that creativity in connection with language that
Chomskyans, at least, insist on. But that is not surprising, as they are badly muddled about the nature of that creativity.2
The Chomskyan 'transformational grammar' system is only worth criticizing in any of its specific claims on the basis that at least
it aims to discover the 'universal grammar'. In fact it does not manage this at all, but it would like to. But other modern
linguistic theories seem to have even less to do with reality, and are not worth any detailed criticism. They are simply analysis
for the sake of analysis, and indeed are criticized by Chomsky as mere "taxonomy".
Here is a description of part of an early system of modern linguistics (Saussure, early 20th century):
"We thus have another dichotomy, of syntagmatic vs paradigmatic, as illustrated in the following diagram:
1 I wonder if the complete grammars that linguists aspire to would allow Orwell's "some animals are more equal than
others."
2 See further Chapter 7.

Page 15
Figure 1
But as any one sign within a sentence has both a syntagmatic and a paradigmatic role (the function of he, for instance,
being partly a matter of its relationship with the other personal pronouns, and partly its relationships with the other words
which follow it), it is perhaps wise to avoid talk of dichotomies at this point, with all their implications of mutual
elusiveness [sic]. Both kinds of relationship are necessary to carry out the complete analysis of any sentence." (David
Crystal, Linguistics, 1981, pp. 165-66)
Then, for instance, there is 'tagmemics':"
The tagmeme...tries to combine into one conceptual unit two ideas which had previously been quite disassociated, the
ideas of class and function. When you analyse a sentence in Pike's terms, you do not end up with two independent
statements, one telling you what the bits of a sentence are and the classes they belong to, the other telling you the
functions they perform (such as would be illustrated by the analysis of the sentence John kicked the ball as 'Noun phrase +
Verb + Noun phrase' and 'Subject + Predicate + Object' respectively). Rather the sentence is analysed into a sequence of
tagmemes, each of which simultaneously provides information about an item's function in a larger structure, and about the
class to which it belongs that could also fulfil that function. A metaphor that is regularly used here is to talk of a structure
as a series of 'slots' into which various types of 'filler' can go:

Page 16
extending the above example, the Subject slot can be filled by a Noun phrase, a pronoun, a Verb phrase (as in To err is
human), and so on. The working out of a comprehensive tagmemic theory, so as to integrate lexical, grammatical and
phonological kinds of information, is something which has exercised the minds of Pike and his associates for some time -
the idea having been first developed in the early fifties. To do this, Pike makes extensive use of a presentation of
information in terms of matrices - that is, he establishes a network of intersecting dimensions, and plots significant points
within it, thus expressing both structural and functional information simultaneously." (Crystal, 1981, pp.213-14)
Or 'stratificational' grammar:
"The theory is called 'stratificational' because one of its chief features is its model of linguistic structure in terms of several
structural layers, or strata........ Language for Lamb...is best viewed as a system of complex relationships which relates
sounds (or, of course, their written counterparts) to meanings. These relationships are not all of the same kind, however,
but break down into more restricted sub-systems, each of which has its own structural organization; one sub-system
operates at each stratum. There are three main strata hypothesized for all languages, 'semiology', 'grammar' and
'phonology'; and these are subdivided in various ways to produce further systems. The complexity of linguistic
relationships is said to be due to the different structural patterns exhibited by sounds, on the one hand, and meanings, on
the other - there is no clear one for one relationship between the two. The various strata are viewed as in a hierarchy, the
structure of each stratum being represented explicitly in any adjacent stratum by means of a small set of basic
relationships; and Lamb has developed a notation and method of diagrammatic presentation to try to represent these
relationships adequately." (Crystal, 1981, p.216)
Notice the important sounding phrase accounting for (my italics - see p.25) in a description of 'scale and category' grammar:
"The reason for the labels 'scale' and 'category' is that they refer to the basic ideas underlying the theory: it was felt that
the best way of accounting for language's structure was by postulating four major theoretical categories, and relating them
via various scales of

Page 17
abstraction ...... The categories comprised class (covering concepts such as 'verb' and 'noun'), unit (covering concepts such
as 'sentence' and 'clause'), structure (covering concepts such as 'subject' and 'predicate'), and system (covering such
concepts as the set 'personal pronouns' or 'tenses'). Scales were the model's constructs which related these categories, and
the linguistic features subsumed under them, to each other. For example, one scale was the means of relating the 'units'.
The various units recognized (sentence, clause, group, word, and morpheme) were considered to be arranged
hierarchically on a rank scale, and each unit was conceived of as consisting of one or more of the units below it - a
sentence was considered as consisting of one or more clauses, a clause as consisting of one or more groups, a group of
one or more words, and a word as one or more morphemes. Thus the utterance The boy will kick the ball would be
analysed as one sentence, consisting of one clause, which consists of three groups, (the boy, will kick, the ball - Halliday
uses 'group' where most other linguists would talk of 'phrases'), each group consisting of two words, and each word
consisting of one morpheme. Other scales would relate other observations about linguistic structure." (Crystal, 1981,
pp.214-15)
Out of 'scale and category' grammar came 'systemics':
"As its title suggests, it is based on the idea of system, which was a category of the early approach. This idea suggests that
at any given place in a structure, the language allows for a choice among a small, fixed set of possibilities (we can have
the/this/my/a...man, for instance); and it is very similar to the Saussurean concept of paradigmatic relationships ......
Language is viewed as a series of 'system networks', each network representing the choices associated with a given type of
constituent (e.g. clause system network, nominal group (sc noun phrase) system network, and so on); and in this approach,
it is the clause system that is taken as the point of departure in analysis, not (as in most other models) the sentence... this
approach...tries very hard to integrate information about structure with information about classification in a single model.
In the hierarchy of units, for instance, each unit has a particular structure and belongs to a particular class, and thus has a
range of functions. As such, there is an important development here from previous grammatical approaches of the
constituent analysis type." (Crystal, 1981, pp.215-16)

Page 18
'Systems' are defined as "lists of choices that are available in the grammar of a language". In English, for instance, there is a
system of 'number', with the choice between singular and plural; a system of 'person', with the choice between first, second and
third; a 'tense' system with the choice between past, present and future; a system of 'mood', where the choice is between
declarative, interrogative and imperative; and so on.
To be fair to 'systemic' linguists, they do emphasize the importance of meaning in language. They even refer to the structure of
sentences, the forms and patterns of grammar and the way they fit together, as "the more surface aspects of grammar"; and to
"the meanings of grammar" as "the fundamental aspects of grammar".
Figure 2 illustrates some of the 'systems' of 'transitivity' that systemic linguistics make 'applicable' to clauses. The further a
'system' is to the right in the diagram, the more 'delicate' the distinction it makes in meaning. Systemic linguists would say that
the clause
Bill fell into the river.
has made certain 'choices'. Among the many interlocking networks of 'systems' it 'chooses' from, it chooses from various
mutually exclusive 'processes'. Out of the system furthest to the left it has made the choice 'material process'; this term, 'material
process', is subdivided into the two terms of a more 'delicate' system, from which the clause has chosen the term 'action process';
and out of the subdivision of 'action process' the clause has chosen 'supervention process'. The clause
I saw him fall in.
has 'chosen' 'mental process', then 'internalized process', and then 'perception process'. The clause
He said it was very cold.
has chosen 'mental process' (?!) and then 'externalized process'.
Bill is in the river.
has chosen 'relational process'.

Page 19
Figure 2
This is a very clear example of how linguists first invent a purely artificial framework and then fit language into it. They think
up an analytical pattern and impose it on language. It is an activity the opposite of looking at the nature of language and letting it
lead them to the appropriate conclusions. It is just one more exercise in analytical game playing. As the analytical system comes
before what is to be analysed it is, in principle at least, even more barren than old fashioned grammatical parsing.
This trick of inventing the system first and squeezing language into it explains how it has been possible for the linguists to
sprout so many different contraptions out of thin air and keep themselves busy.
''In order to fully understand the meaning of something, we need to know what it does not mean as well as what it does
mean. To take an example from the level of lexis, the lexical item animal has some meaning even if it occurs on its own
out of any particular context. But we do not know exactly what it means until we know whether it means 'not vegetable or
mineral' or 'not human' or 'not bird or reptile or fish'. By naming a meaning and by assigning it to a system, where it is
mutually exclusive with other named meanings, we are able to show both something of what it does include and also what
it does not

Page 20
include; what it does not include, that is, of the area of meaning which the terms in the system have in common,"
(Margaret Berry, Introduction to systemic linguistics. 1. Structures and systems, 1975, p.145)
This is characteristic of the empty formality of linguistics. That "to fully understand the meaning of something, we need to know
what it does not mean as well as what it does mean" (apart from being an illogical statement - if we know what it means we
already fully understand the meaning) is at best only true in an utterly unpractical abstractly formal way. It has nothing to do
with the actual psychology of using words. A child may know exactly what "tiger" means even if she has never heard of
''jaguar". Above all perhaps, notice how, by systemics' own argument, one has to know the exact meaning before one can assign
it to a particular system! Berry's example of "animal" illustrates this perfectly.
So what is it all for? Merely, it seems, a celebration of the purest artificiality. One's conviction of the artificiality is confirmed
by the requirement that 'systems' contain only mutually exclusive meanings. In real language meanings constantly overlap:
"close/shut"; "soft/gentle/mild/tender"; and so on.1
Systemic linguists are particularly obsessed with 'choice'. But they are not the only ones seduced by this idea. Many people now
see a link between language and the fashionable 'information theory'. 'Choice' is the great word. "...the...general principle (which
may be taken as axiomatic in the study of any system of communication) that meaning implies choice." (John Lyons, Chomsky,
1977, p.77) But one does not choose a meaning, select from a range of possibilities. One turns a thought directly into the
meaning associated with it. If one is making an omelette one does not say to oneself: "Now, of the possible ingredients in
cooking I select one." One says simply: "I need an egg." Omelette leads direct to egg. When one uses language the thought
leads even more directly to the word. The thought immediately imposes the word. If the thought is, for instance, that behind "I
have lived in Tokyo", one does not have the choice of "I live in Tokyo", or any other choice. Choice does not enter into it.
This confusion linguists have fallen into is possibly a good example of the corrupting effect of language itself that I try to
explain in Chapter 11.2 'Choice' is one of the key concepts of the mathematically logical world of
1 See also Chapter 5, pp.107-08.
2 The thought unfortunately often becomes, in practice, the victim of the word it ought to be controlling. But that is another
matter.

Page 21
information theory and computers - a world that has notched up much celebrated success in recent decades. Language - or
rather, abstract analysis of language - is very inviting, very susceptible to a superficial treatment that reminds the analysers of
that world. And so, hey presto, everything is 'choice', and the most false associations are established. Things probably work
equally powerfully in the reverse direction. Information theory provides the word "choice" and the word alone puts ideas into
linguists' heads, and in no time at all they are throwing up their hands in delight - we have "choice" in language too!
What is the point of all these linguistic systems? They do not do anything useful for anyone. I have never seen or heard, or
heard of, any evidence that they do. Linguists often write busily of how important it is, for us all, to understand language. But
they never explain how their work is going to do this in any real, practical way. They just assure us vaguely that it will. The
authors of systems such as those I have mentioned above may, I suppose, try to defend themselves by saying that it will all, in
due course, reveal the universal design of language. I cannot see that it will do anything of the sort, that it will reveal anything.
Everybody is already aware, for instance, that there are words that mean actions and others than mean things; everybody is
aware that there are doers and done-tos; that words express various sorts of times, others refer to people; and so on. Everybody
is aware which things in life fit together and which don't - and it is those things alone which determine which meanings (words,
pieces of language) fit together and which don't. Turning all this awareness into some neat self-fitting learned man's system
enlightens nobody, and is actually impossible anyway - which is why the linguists are all so busy, all trying to find "the best
way of accounting for...".
For those who have not seen the sort of thing 'transformational-generative' grammarians go in for, here are some examples. First
a representation of Chomsky's so called 'Aspects-type' system of 'grammar':

Page 22
Figure 3
There are what they call 'phrase-structure rules'. The examples below are from an earlier but not basically different Chomskyan
system. They correspond to the 'Base Component' in the diagram above and they are the basis of the famous Chomskyan 'deep
structure'.1 An arrow indicates that the element to its left is to be rewritten as appears to its right. NP=noun phrase, VP=verb
phrase, sing=singular, pl=plural, T=article, N=noun, Aux=auxiliary verb, V=verb, M=modal verb. The elements in brackets in
rule (10) are 'optional'. The brackets of (3), (7), (9) and (11) mean that any of the elements contained inside them may be used.
1 See Chapter 8.

Page 23
From these 'phrase-structure rules' they get 'underlying strings', e.g:
With such a 'string' they 'associate' a 'phrase marker':
Figure 4
They apply a 'transformational rule' to the 'underlying string', thus:
and claim that they have 'derived'1 a passive 'string':
with its 'associated phrase marker': (The question mark represents a little problem they had previously; they claim to have dealt
with that now.)
1 See Chapter 8.

Page 24
Figure 5
After applying some more 'transformational' rules they finally arrive at a sentence - 'surface structure':
The bird may have been eaten by the cat.
And here is an example of writing which for me sums up and shows up the pointlessness of it all:
"...an important theoretical problem. The derived string produced by one transformational rule may serve as the underlying
string for the operation of a subsequent transformational rule, and will therefore need to have associated with it the
appropriate derived phrase marker. Chomsky and his followers have worked on this problem and tried to establish a set of
conventions according to which a particularkind of formal operation (e.g. deletion, permutation or substitution) is defined
to have a particular effect upon the topology of the phrase marker it transforms; and we followed these conventions when
we decided that the effect of the rule B + D + E -> E + B operating upon the underlying phrase marker shown in Figure
x was the derived phrase marker in Figure y. But this was a very simple example from the point of view of the operations
involved and the shape of the phrase marker to which they applied; and furthermore it was a purely abstract example
unaffected by any empirical considerations. The reader will appreciate that the situation is very different when it comes to
formulating a set of

Page 25
transformational rules for the description of English or some other natural language." (Lyons, 1977, p.72)
For a long time I was very puzzled by that expression much used by linguists, "account for". For example, they talk about
'accounting for' the ambiguity of a phrase like "old men and women", and Chomsky says that his system of transformational
grammar is more powerful as a model for the description of language than phrase structure grammar, as it can, for instance,
more satisfactorily 'account for' certain types of structural ambiguity, such as
5 Flying planes can be dangerous.
(This, it is claimed, can mean either "To fly planes can be dangerous" or "Planes which are flying can be dangerous".)
In my innocence and naivety I imagined that 'accounting for' meant explaining the psychology or logic behind the ambiguity,
some profound and fundamental insight into the nature of language. I indeed rejected the claim that I imagined they were
making, as I did not see any great psychological or logical problem in these ambiguities. The explanation, as nearly always,
seemed simple and obvious. In fact the explanation of an ambiguity lies in the very ambiguity itself. 5 is ambiguous because "-
ing" has more than one meaning and if it is combined with certain words which do not, by their meaning, show which meaning
of "-ing" is intended, one does not know which meaning is intended! But, for instance,
6 Flying planes is dangerous.
or
7 Driving cars can be dangerous.
is not ambiguous, quite simply because extra distinct meanings have been added. In the same way, "old men and women" is
ambiguous but "old men and youth" is not. One finds, as one will always find if one looks straight, that the key to meaning is
meaning.
But the way linguists think is so artificial that I did not realize at first what they were talking about. It is possible, they say, for a
phrase structure grammar to generate a sentence such as 5, but it won't be an intuitively satisfying account of the ambiguity and
it fails to relate the phrase "flying planes" on one hand to "planes which are flying" and on the other to

Page 26
"someone flies planes". However, the transformational analysis, they say, accounts for the ambiguity by relating two different
underlying strings, "plane + s + be + ing + fly" and "someone + fly + plane + s'', to the same derived string; many other
structurally ambiguous sentences can be 'accounted for' satisfactorily by transformational grammar.
It finally dawned on me what they are doing. By 'account for' they mean something completely trivial. All they mean in the case
above, for instance, is that application of 'transformational rules' to two different 'underlying strings' will lead to an "-ing" form.
It is trivial because all they are really doing is finding a sign for something that they hadn't been able to allot a sign for before.
In other words, all these systems they create, whether it is phrase structure grammar, or transformational grammar, or anything
else, are nothing more than codes that they want to translate, or rather force, language into. One code is better than another if it
'accounts for', that is, provides signs for, is able to codify, more aspects of a language. But these codes do not tell us anything or
about anything that we didn't know before; they explain nothing and get us nowhere.
Even if one accepts, for a moment, the transformationalists' whole contraption, it does not 'account for' ambiguity in any true
sense of the expression. Why is the "-ing" form sometimes ambiguous? Because it has two different meanings. Why does it
have two different meanings? Nobody knows. Or, according to the transformationalists, why is the "-ing" form sometimes
ambiguous? Because two different meanings (structures?) are transformed into it. Why are two different meanings transformed
into it? Nobody knows. (It won't do to claim that some universal grammar makes such transformations unavoidable, since it
obviously doesn't. In most other languages in the world an equivalent of the "-ing" form is not the result.)
If it is protested that I have failed to understand that the term 'account for' is used here to refer to Chomsky's discovery of the
process -transformational grammar - by which ambiguities, among other things, arise, I must answer that that is a false claim.
That process of transformation is merely an invention through analysis. In fact, the procedure is doubly fraudulent because the
analysis of underlying strings and transformations is analysis of the imagined, not analysis of actual, existing language. This
imagining analysis is then used to invent a process -a psychological process, that is. But a psychological process is not
demonstrated by formulating such an analysis or code, let alone when that code is based on something purely imagined. What
warrant has Chomsky for inventing these 'underlying strings'? For they are no more than his inventions. There is not the
slightest evidence that they exist in anybody's

Page 27
brain. And the same applies, of course, to the 'transformations', which are simply fantasies used to convert fantasies into reality.
It is very much easier - and more rational - to believe that things are what they look like, namely that some pieces of language
have, for instance, two or more different meanings, that speakers sometimes use such pieces, because they express their intended
meanings, and that in some contexts it is not clear to listeners which meaning is in fact intended. Much easier than to believe
that speakers start with one formulation of language in their heads, then make all sorts of very complicated changes to that
original formulation, in order to produce something quite different that sometimes other people can't understand.
In any case, what is supposed to happen when the ambiguity lies entirely within one piece of language itself, that is, in what I
have called 'specific' meaning? For instance, "Make a bow." What 'accounts for' that ambiguity? In fact, of course, they can
invent as many 'underlying strings' and 'transformations' as they like, but it won't get them anywhere at all, even in their own
terms, in 'accounting for' the ambiguity of "bow".
And even if one assumes that there is indeed something which is converted or 'transformed' into the language actually spoken -
and in a sense there is such a something, that is to say, thought - it is a huge assumption, quite without cause, to take it that that
something is language. I imagine this assumption is made because linguists also assume that thought is language, or that, at the
very least, thought is closely reflected in language. I shall try to show in Chapter 10 that this is not so. What is happening is that
the linguists are determined to analyse, although it is neither necessary nor bears any fruit, and when they run into difficulties
they invent something - in this case transformational grammar - in order to make further analysis possible.
I want to repeat that in a slightly different way, because it is an example of the false principle that is the heart of the whole
irrelevancy of linguistics, of whatever school:
Transformational grammar has been invented, not because it is needed to explain any observed facts of the human experience of
language, but because it is needed to improve an analytical code.
In this respect this sort of analysis is exactly like the rules of the war games some people play on table tops. When they devise
war games they must above all, of course, make sure that their rules are consistent with each other; they may also try to make
rules that are as 'realistic' as possible and take into consideration as many as possible of the factors they believe to be those of
real war. But only the mad would believe that these codes or

Page 28
symbols that they devise for war can tell them anything about the nature of real war. 'Realistic' rules of the game can only arise
out of previous knowledge of the nature of war; they merely simulate a reality known to be of a quite different kind.
Yet it is a precisely equivalent mad belief that linguists appear to have about their various systems of formal grammatical
analysis; they appear to believe that their codes are the discovery of new truth about language, while in fact they are a new
barrier of fog interposed between us and our understanding of it.
The pointless triviality perhaps becomes even clearer when one sees the sort of illustration that the linguists Neil Smith and
Deidre Wilson give of the 'necessity' of transformational grammar (Modern linguistics. The results of Chomsky's revolution,
1979, pp.89-99). They offer the sentences:
8 All the children might have played with their friends.
9 The children all might have played with their friends.
10 The children might all have played with their friends.
11 The children might have all played with their friends.
They say that a phrase structure rule can produce 8, thus:
where the bracket ( ) again means that the element is optional, and S=sentence, Art=article, Adj=adjective, PP=prepositional
phrase, and P=preposition. But they say no modification of this rule can indicate that 9-11 are possible grammatical sentences,
because if we allow for the optional occurrence of the quantifier Q ("all" in this case) before each of the verbs in 8, 9-11 can be
produced, but so can the ungrammatical

Page 29
8* All the children all might have played with their friends.1
9* The children all might all have played with their friends.
10* All the children all might all have played with their friends.
11* All the children all might all have all played with their friends.
So a new sort of rule is needed, one that prevents "all" occurring more than once for each subject NP that occurs. A rule that can
move items around. So 'transformational' rules are constructed. The one Smith and Wilson offer is:
where X is a variable, which can stand for anything at all, or nothing. The effect on the phrase marker is shown by the dotted
arrow. (I have used a simpler example than theirs here, in order to avoid irrelevant complications.)
Figure 6
Here one sees not only how trivial but also how dangerously misleading the urge to describe is. When the Chomskyan linguists
get to the stage I have just reported, surely it is obvious that they have completely forgotten about real language and real
psychology. At this stage they find that they cannot fit all the undoubted outward facts of language into their formal system, and
now all they are aware of is the need to expand their method of expressing these facts on a piece of paper. When they have
1 I follow the linguists' convention of marking ungrammatical sentences with an *

Page 30
expanded their method they declare that they have 'accounted for' more facts. They have indeed, in the sense, again, that they
have found a code, a scheme of representation, for more facts. Actually, they are doing nothing different, basically, from what
all non-Chomskyan linguists are doing as well. What is so wrong is that they make themselves believe, and try to make others
believe, that they are saying something true and new and important about the nature of language and about how humans use it.
In fact they are saying nothing, and appear unable to recognize that there is no reason at all why anyone in the real world would
want to produce sentences like 8*-11*1 and that all that needs to be said about "all" (if anything needs to be said about it at all)
is that one can put it in more than one place. That is the habit of English-speakers. To suggest that speakers start with one word
order in their heads and then transform it to produce what they actually say seems grotesque. Example after example of this sort
of barren analytical circle drawing appears in the linguists' writings. There is a good one in Smith and Wilson's discussion of
transformational rules involving "there" in Modern linguistics (1979) on page 107, but the indulgence in mere codifying is
perhaps seen at its most extreme and exposed in the same book on pages 121-22:
"...it should be noticed that there are certain verbs which enter into structures of the type NP V NP, but which do not have
related passives. 12-17 are examples:
12 That book cost a lot of money.
13* A lot of money was cost by that book.
14 John resembles your mother.
15* Your mother is resembled by John.
16 Mary became President.
17* President was become by Mary.
It would be perfectly straightforward to account for these facts, as well as for the regular passive versions of active
sentences, by simply stating in the lexical entry for each passive forming verb that if a structure of the form 18
18 NP1 - V - NP2
is well formed, so is the corresponding sentence 19
1 And if they say that one cannot assume this, that there may be, at least in principle, an unusual language whose speakers
actually say things the equivalent of, for instance, 11*, one has to reply first that the usage of English can be explained to
would-be speakers of that language in far better ways than by this formal system, and second that such a formal system
tells us nothing about the psychology of language.

Page 31
19 NP2 - is - V-ed by NP1
and where there is no passive form of a verb, such a statement will simply be omitted. If it is objected that such a
treatment misses the obvious generalization that transitive verbs typically passivize, this fact can be easily incorporated
into the grammar by stating once for all in the lexicon that any verb which can occur in a structure of the form 18 will
also be able to occur in a structure of the form 19, unless it is explicitly marked as being unable to. This would have the
effect of leaving the general case, e.g. open, unmarked, and only the exceptions, such as cost, marked in the lexicon. Such
statements are known, rather misleadingly, as redundancy rules, because they enable one to avoid the repetition in each
lexical entry of redundant, or predictable, information."
I am not saying that such entries in a dictionary for students of English as a foreign language would necessarily be useless.
What would be the best way, practically and psychologically, to provide such information, and how much of it is in fact
necessary in the first place, are sensible subjects of debate - and the second question is very debatable. What is so irritating
about the linguists' work is that it is not such things that concern them. They want to create a beautiful, abstract, fitting design
that they can put forward as intellectually profound. So it is beneath them here to recognize that the only thing really at stake is
how it is best to organize a dictionary.
Another bizarre claim is "that it is the structure provided by the syntactic rules which are the inputs to the semantic and
phonological components."1 Such a view of the psychology of language is again the result of their analytical mentality. What
surely happens in fact is that one turns one's thoughts, feelings etc. deliberately into meanings (pieces of language) and at the
same time puts the meanings into a word order to express the broader thought - and the word order is itself a form of meaning.
And if they don't actually mean that their descriptions, and claims such as the above, are psychological reality, then they are
effectively admitting themselves that their activities are indeed analytical uselessness of the most extreme sort. Smith and
Wilson in fact appear to do virtually this in statements such as:
"There are two main types of justification for the level of deep structure: first, having such a level simplifies the statement
of
1 Smith and Wilson, 1979, p. 66. I discuss connected claims in Chapter 8.

Page 32
semantic rules, and second, that having such a level simplifies the statement of syntactic rules. In other words, having a
level of deep structure makes it easier to formulate significant linguistic generalizations on both the syntactic and semantic
levels." (1979, p.100)
They do indeed claim 'significant' generalizations. How are they significant? Significant for what, apart from the contrivance's
own circular self?
Formal description of a language says nothing about how it really works. It is illusion to imagine that bringing in the impressive
sounding word "rule" makes any difference to this truth. Apropos "all", for instance, their 'rules' are just another, complicated
and confusing, way of saying that you will find "all" in certain different places.
To return to the Chomskyan analysis of ambiguity: not only is it pointless; it involves a flaw in the very foundation of the 'deep
structure' idea, which I point out in Chapter 8. However, I think one can see already here that one of the reasons for their
misconception is their assumption that "Flying planes" can mean exactly the same as "To fly planes" (or, to follow their system
of 'underlying strings', "someone + fly + plane + s"); or, alternatively, exactly the same as "Planes which are flying" (or "plane +
s + be + ing + fly''). Again, see Chapter 8. But I must bring up already here the point that in fact no two words or phrases or
sentences or 'general' meanings ('grammar' meanings) ever mean exactly the same.
To make the point clearer, take some different sentences and phrases, similar to sentence 5, but unambiguous simply because
again different 'specific' meanings remove any possibility of ambiguity in 'general' or 'grammatical' meanings. First:
20 Washing dogs can be dangerous.
"-ing" used in this way refers to the action itself, and I believe most English-speakers would interpret1 this sentence to mean
(roughly - only the sentence itself is an exact account of its own meaning) that the action of washing dogs can be dangerous to
the washer because the dogs might, for instance, bite.
1 In circumstances that would produce a spontaneous and uncalculated response. Direct interrogation would be useless.
See Chapter 7.

Page 33
means something else. Here the emphasis is not on the action itself, and I think most English speakers would use such a
sentence to suggest, for instance, that the dogs might die as a result. If that seems to be making too much of a doubtful
distinction, consider that it is a distinction with an unquestionable basis. Take just one of many possible examples, the
distinction between
22 I remembered to post the letter.
and
23 I remembered posting the letter.
Or, perhaps not quite so immediately obvious, but clear cut enough,
24 I like to get up about half past five, as I can really get something done in the day that way; but I can't say I like actually
getting up at that hour!
And from a sentence such as
25 I'd first like defining the phrase "account for".
one can see immediately how real the distinction between infinitive and "-ing" is and how essential it may be. As usual it is
simply the particular realities that decide whether a distinction is going to be crucial in a practical way.
Equally, if one takes the other Chomskyan 'underlying' idea, it is pretty clear, I think, that we would not consider a possible
alternative title for a painting entitled "Laughing children" to be "Children who are laughing".
Another example that shows the nature of the whole 'account for' procedure very clearly is where Lyons (1977, p.76) says that
"repeated application" of the rules of "conjoining and embedding transformations...accounts for the existence of such recursive
structures as This is the....that lived in the house that Jack built or ...a big, black, three foot long,..., wooden box". In other
words, repetition will give you repetition.
It is not only in their structural codes that linguists and many others show their faith in barren analysis. Here is another example.
In a review of Smith and Wilson's Modern linguistics Anthony Burgess (1979) admired the advance he considers we have made
from the

Page 34
old method of merely listing pairs of words like "télegraph - telegráphic, téescope - telescópic, atom - atómic" and showing the
shift in stress, to what he calls explaining it by saying that "stress in the adjective is regularly attracted towards the syllable
immediately preceding the adjective-forming suffix ic". Smith and Wilson themselves go even further (1979, p.238). They give
this as an example of what they call 'descriptive adequacy' in a grammar of a language, and say such a grammar will give an
insight into the minds of those who speak the language.
But such a statement about 'stress attraction' does not explain anything, or give insight into anybody's mind. It is merely
gathering into a formal statement of a 'rule' something that practically all English-speakers are aware of without the benefit of
linguistics, and such awareness is emphasized by the linguists themselves! It is true that such rules, which many teachers 'know'
without being taught by linguists, can be given to students of foreign languages, but they do very little if any good. In practice,
such knowledge should be absorbed by students of a foreign language 'on the way', as it were, without conscious study, in
exactly the same way as native speakers so effectively do. There is something basically wrong with their approach to foreign
languages if students can't do this.
No, such analysis explains nothing. What might justly be called 'explaining', though, would be if someone could discover why
English has the stress in the noun in a different place from the adjective, while many other European languages have it in the
same place.
There are several other sorts of linguistic analysis that academics indulge in. Smith and Wilson's Chapter 8 Pragmatics and
Communication is a survey of a particularly barren form. They say, quite rightly, that people often understand much more from
an utterance than its literal or direct interpretation. What people understand, the conclusions they draw, will depend on factors
like knowledge of the world, previous remarks, the situation, and so on. The linguists want to analyse all these factors and
interplay, as if this was going to make us wiser in some way. But the point is that everyone is basically aware of these factors
and the way they work. If they weren't, these factors wouldn't be factors - their effect depends on people's awareness of them,
and of course there are situations where speakers and listeners don't share the same knowledge or the same powers of drawing
conclusions or the same conventions of implication or, quite simply, the same sensitivity to hidden message. Awareness is the
factor.

Page 35
Some linguists obviously like ordering all the factors into some sort of table. But let them not pretend that the table enlightens
anyone.
Much the same has to be said about the subject of 'sociolinguistics'. There are indeed many various groups of people who use
particular forms of language with particular 'social' meanings or particular 'social' aims or effects. But anything sociolinguists
say about this sort of language that is true is known by the people (London taxi-drivers, say) who use it as well. In this case it
seems pointless for the sociolinguists to say anything about it. The taxi-drivers, and people generally, may not be 'conscious' of
the way their language works 'socially', in the academics' tabulating and organizing sense. But they must be effectively aware of
it. Otherwise they cannot and do not use it. As regards anybody who does not know the 'social' meaning of a particular form of
language (most customers of London taxi-drivers, perhaps) anything sociolinguists say about it is obviously not true. And if taxi
customers want to understand the drivers they will have to learn the language, in principle just as they would any 'foreign'
language. Neither their nor the academics' studies will give new and significant enlightenment, though getting together with the
drivers might.
It is quite true that both whole communities (ethnic minorities, less 'educated' groups, and so on) and individuals can have
'social' problems involving language. But these are basically moral problems. No amount of linguistic data gathering and
analysis will help until that is recognized, and when the moral problem is recognized and tackled the linguistic analysis will be
superfluous. Language should not be used to impress, deceive, dominate, frighten, humiliate, be subservient, or gain approval.
Nor should it be used as the basis of judgements of people's worth. It should only be used to explain or please, to make clear or
give happiness. Certainly language can cause misunderstandings, often with pitiful consequences. Language also distorts thought
and leads emotion astray. These are quite other matters, things I devote the later chapters of this book to. The solution, though,
is not analysis.
In fact, as actually practised, sociolinguistics appears to consist very largely of the collection of vast amounts of detailed
statistics about the use of this or that expression or 'structure' or pronunciation in this or that community. If there are amateurs
excited by such tabulation, long may they enjoy it, just as train spotters derive pleasure from their loggings. And neither
sociolinguistics nor any other sort of linguistics should be allowed the pretensions that might try to justify writing such as:

Page 36
"In developing this idea, Sacks observes that turns [within a group of speakers] consist of one or more sentences, with a
sentence being defined as a unit which has
its completion recognized on its completion, and that it is not completely recognizable by participants; also it can be
monitored, from its beginning, to see from its beginning what it will take for its completion to be produced, in such a way
that, on its completion, its completion may be recognized."
(Coulthard, An introduction to discourse analysis, 1985, p.61)
Great claims are constantly made for linguistics; I can find no basis for them. For example:
"Clearly, the knowledge that speakers have of their own grammars is not conscious knowledge. This is obvious enough in
the case of adults, but even more so in the case of children, who are normally completely unaware of the way in which
they form relative clauses, for example, or the conditions under which they would use the word come rather than go. The
linguistic knowledge that speakers have is unfortunately unconscious knowledge; the job of the linguist is to attempt an
explicit, conscious formulation of the grammatical rules that speakers know. Linguistics conceived of this way is
concerned with one aspect of the human mind and therefore correctly classed as a branch of psychology." (Smith and
Wilson, 1979, p.22)
"Unfortunately", they say, as if a great treasure was lying hidden! I regard myself as a pretty good grammarian of English, that
is to say, an investigator and explainer of the principles of English 'grammar' for those who ask my help in learning it as part of
a foreign language. I have never for a moment believed that I am making the slightest contribution to any sort of psychology. I
am quite clearly doing nothing of the sort. I have, I think, cast useful light on the true meaning1 and use of the 'present perfect
continuous' of English verbs, for instance. That does not add one iota to knowledge or understanding of anybody's psychology,
and I do not see how anybody could ever show that it does:
1 'True' meaning from the non-English-speaker's point of view. Native English-speakers, grammarians, who attempted to
explain it as part of a foreign language mostly got it wrong. Together with all other native speakers they knew the 'true
meaning' all along, of course. But when it came to describing it they failed to observe their own language properly.

Page 37
Linguists' description of a language is far more complicated than the language itself and much more difficult to understand. At
best linguistics is a tortuous coded account of what a language itself tells us much more directly. It does not explain either how
or why a language works as it does. It is the equivalent of describing the passage of the apple to the ground without offering any
reason why it falls. Linguists confuse systems of symbols and formulae with the discovery of what moves or drives language.
They are nowhere near the only important problem - what goes on in individuals' creation of language, how language works
inside us. Chomskyans do pretend to offer some sort of real explanation, but it is fanciful and illogical and still rests on analysis.
Can the linguists point to a single clearly valid and useful piece of information about humans or language that all their analyses
and descriptions have given us? I think not. They cannot even achieve the first step of agreeing on a model of language, which
ought to destroy confidence in the whole enterprise from the start. Berry has written (1975, pp.12-13):"
"Approaches to Language Study Should Be Varied
Present day linguists believe that there is, and should be, more than one way of studying language. Newcomers to
linguistics sometimes consider it a weakness of the subject that there are different schools of linguistics, each considering
language in a different way. It is on the contrary, a strength of linguistics that there are these different approaches to the
subject. Controversy is always a healthy sign. Language is so complex that no one approach can cover all its aspects. The
different approaches are complementary rather than contradictory .... Often one finds oneself faced with a situation in
which two or three different labels could be considered applicable to one bit of language. This is a desirable, not an
undesirable, state of affairs. One gains a greater insight into the bit of language by considering in turn the rival labels than
one would if one could with confidence attach just one label and leave it at that."
This is typical of the solidarity among those in authority. But the defence reveals its own weakness. The statement that it is
better not to be able to "with confidence attach just one label and leave it at that" exposes the whole unreality of linguistics. That
the activity is 'labelling' is alone pointless enough. (And I suspect that there are so many schools of linguistics partly because it
is so easy to invent new ways of labelling. There must be an almost infinite number of ways one could do it.) But the weakness
is much

Page 38
deeper than that. For where in the real world could one allow more than one truth? Could one afford to believe and proclaim
proudly even that one can look at the motion of the earth in at least two valid ways: the sun goes round the earth, but the earth
also goes round the sun? Would an engineer be happy to affirm that a suspension bridge holds up the cables just as much as the
other way round, or that the smoke, the funnel, the wheels and the track can all be usefully seen, when appropriate, as the
driving power of a railway engine, and that confidence in just one answer restricts insight into the technical problems involved?
I suspect that many budding linguists (and students of many other academic 'disciplines' as well) make the mistake of confusing
the understanding of an argument or a system with truth itself. They understand the rules of the game invented by earlier
linguists, and how other rules are developed systematically from those rules, and in their joy and pride at being able to follow
they think the whole self-enclosed structure represents reality. In their joy and pride at following how these learned people have
fitted the jigsaw puzzle together, they may even want to become jigsaw puzzle makers themselves. And they think that being
able to follow an argument means the argument is a good one.
The enterprise of modern linguistics rests on the terrible fallacy that the requirements of analysis prove the existence of real
things we were unaware of before. Belief in the virtues and power of analysis is one of the great illusions from which humans
suffer. I do not mean that all that is meant by the word analysis is useless. It is no doubt essential for chemists to know what a
substance consists of or for cooks to know what is in a soufflé. But the basic mistake that so many humans make, academics,
scholars and philosophers above all, is to believe that the activity of defining, classifying, naming, subdividing, systematizing,
all adds to knowledge, truth, understanding, and gives new insight. It does not. Nor is it admirable, any more than, say, war-
making, just because it is a uniquely human talent.
Yet this kind of analysis seems to be an unconquerable itch. It is done, I think, for its own sake. Collectors usually confess that
they collect things simply for the enjoyment of collecting. Not so the analysers. They claim great purposes.
In The Cambridge Encyclopedia of Language (Cambridge University Press, 1987, p.103) David Crystal describes a method of
"working with semantic space". (The book that apparently launched the idea of 'semantic space' is, he says, a "pioneering
work".) In his example of this method people place mammal names in a diagram where the vertical dimension judges ferocity
and the horizontal dimension measures size. Thus "cat", for

Page 39
instance, is located fairly near one corner (ferocious + small), and "horse" towards the diametrically opposite corner (non-
ferocious + large); words for any other animals can be plotted on the diagram in the appropriate places. The more similar any
animals are in size and ferocity, the nearer to each other they appear in the diagram. Crystal says this is a very simple analysis;
but "the general approach is illuminating, with considerable research potential".
Surely nobody can take such activity seriously, call it illuminating, unless they are so hopelessly taken over by the itch that they
cannot see how absurd it is to try to rationalize it. It is also very disturbing to see how that analytical itch gets enmeshed with
the urge to find subjects for research; research, it seems, for its own sake, but with unlimited possibilities so long as new ways
of analysing anything can be thought up. If you can devise anything to research, research it. Yet if we are not careful it is such
analysing researchers that people are going to believe are the oracle.
Analysis of the kind I am talking about is an empty activity in many fields, but few take it as far as the linguists. The linguists
are rather like the people who once upon a time used to discuss earnestly how many angels could sit on the point of a pin.
Today we cannot understand how they could be so foolish. Yet they were learned men of their time, and to question their
wisdom might be the sign of a fool or worse. If you unquestioningly believe in angels who can make themselves small enough
to sit together on small areas, and if you take it for granted that the amount of space the angels occupy is a very important
question, then the whole discussion appears entirely sensible, and you will find no contradictions in your picture, in your
analysis, of the angels and their pins.
In later chapters I shall try to explain other flaws in the linguists' basic principles. Before that, though, in the next chapter I want
to show examples of how they cannot even get right many of the facts and much of the immediate logic that Chomskyan
linguists in particular claim as the basis of those principles. In the following chapter I try to break the illusion that linguistics
can have any practical value.

Page 40
3
Chomskyan Mistakes Made Plain
The Chomskyan explanations of language - which effectively means explanations of sentences - have several different kinds of
flaw. There is nearly always a refusal to recognize the part played by meaning, and sometimes this is taken to a perverse
extreme. Often the explanations for a given grammatical fact are just wrong, again usually because the authors take no account
of meaning but obsessively follow their usual abstract track. Sometimes even the data on which they base their arguments are
simply false too. Then there are the cases where the confusion arises mainly because the author has muddled up two different
senses of the same word. A basic mistake is where the linguists begin by talking about one meaning, and then suddenly switch
to a quite unconnected meaning, but argue as if they were still talking about the same original meaning. I shall show in more
detail in Chapter 6 how keen Chomsky himself is on this. The approach is sometimes carried even further, so that the author
begins to produce utter nonsense language in support of his theory.
Many of the Chomskyans' accounts of sentences are presented as part of their argument that there is a universal grammar that all
humans are born with a knowledge of.1 Jeremy Campbell writes in Grammatical man (1982, pp.176-77):
"Other kinds of word relations are also forbidden for no apparent reason other than that they would violate principles
which are peculiar to the human mind, and anchored in the human genes. A more recent and sophisticated example of this
phenomenon is the embedded clause in an indirect sentence, where sometimes part of the clause can be missing, and the
sentence still be correct. One can say, for example,
1 It is unclear what to do.
where the subject of the embedded clause, "what to do," is absent. If the subject were present, the sentence would read
2 It is unclear what someone does.
If the subject were simply dropped to make the shortened form, the sentence would become
3 It is unclear what does.
1 See Chapter 9.

Page 41
which does not mean the same thing. When all the possibilities are examined, it turns out that the only part of an
embedded clause that can be missing is the subject of an infinitive. This kind of constraint appears very widely in the
world's languages, much more widely than might be supposed, and the general principles which account for it are
universal, as far as is known."
To begin with, what Campbell says here is untrue. Take Spanish:
4 No está claro qué hace. (It is not clear what he/she/it does.)
The subject of "hace" ("does") is missing. And Japanese:
5 Nani o shita ka shirarete inai
What object did question known is not
= It is not known what he/she/it/they/etc. did.
Campbell has himself unwittingly given the answer to the nonexistent problem. It is, as usual, a matter of meaning. One cannot
say something to mean one thing if it clearly has another, conflicting meaning -compare 2 and 3 (in 3 "what" is the subject
instead of the object). As for 2 and 1, it is very unclear that they have anything to do with each other at all. Further, one
important use of infinitives is precisely to make the subject impersonal, general, unspecific, indefinite (hence presumably the
name). The essential idea (though not the exact meaning) behind I is that it is unclear what anybody ought to do, so of course
there is no specified subject.
Chomsky's obsession with formal abstraction sometimes seems to turn him temporarily barmy. In a radio talk in 1981 he offered
the sentence
6 John bought a dog for Mary to play with
as an example of language properties "that people know, that they have no evidence for - or minute evidence for at the most."
There are two clauses there, he said, one
John bought a dog for Mary

Page 42
and the other
''something about 'play with'. The 'play with' part is missing a subject and an object. We can ask who's the subject and
who's the object in 'John bought the dog for Mary..to play with' - everybody knows that the subject is 'Mary' and the
object is 'dog'. But how do we know that? Why doesn't it mean 'John bought a dog for Mary..for the dog to play with
Mary'? Why doesn't it mean that? Or why doesn't it mean 'John bought a dog for Mary for the dog to play with John'?
Those are all perfectly intelligible meanings and there are other sentences which in fact have those meanings - but this one
doesn't. This one has a very specific meaning, namely that Mary is to play with the dog. Now if you think about it, that's a
very precise kind of knowledge. The difference between Mary playing with the dog and the dog playing with Mary is
infinitesimal, they're virtually synonymous1-nevertheless we know that it means Mary's playing with the dog and not the
dog is playing with Mary and we know for sure that it isn't that the dog is playing with John. Now that kind of knowledge
is absolutely unattainable on the basis of inductive argument... ...these are just things that you know, I mean if you know
the language you know these things...suppose we had a different organism, not us, but some Martian, let's say, who had a
different principle, and the principle of the Martian was - led to the conclusion that the sentence, again, 'John bought the
dog for Mary to play with' means 'John bought the dog for the dog to play with Mary'. That's a perfectly possible
principle; and it's a perfectly sensible meaning - in fact, some sentence does mean that, the one I've just produced. The
question is, why doesn't this sentence mean that? It can't be a semantic reason because the meaning's fine -perfectly
sensible meaning - it must be some structural reason which says 'that's not the way it works'."
Indeed, as he says, if you know the language you know these things. And if you know the language you know what "for" means.
You know that 'this' sentence is quite different from
7 John bought a dog from Mary to play with
and quite different from
1! Is "Mary played with the pencil" virtually synonymous with "the pencil played with Mary"?

Page 43
8 John bought a dog with Mary to play with
and different again from
9 John bought a dog in spite of Mary to play with
and so on.
His structural fixation not only leads Chomsky to ignore the key part played by "for"; it also leads him to miss the whole logic
of the sentence. All he can see is the sort of clauses he is used to seeing and dividing language up into, in his lifeless, rigidly
mechanical analytical way:
John bought a dog for Mary / to play with
He fails to see that this is not the way English-speakers normally experience this sort of "for" language. It completely destroys
the original meaning. In this sentence "for Mary to play with" all belongs together, and in turn belongs as a whole to "dog". If
one has to break the sentence up at all - which is in fact unnatural - it looks like this:
John bought a dog. Why? What sort of dog? A dog-for-Mary-to-play-with.
That is why all English-speakers not afflicted with linguists' tic immediately know what the sentence means. The word "for"
immediately triggers the recognition of a certain kind of situation in life - a "for" situation, as again in 10 and 11.
10 I'll get something-for-you-to-sit-on.
11 "What's that?" "It's something-for-Mary-to-eat."
Part of learning English is learning that "for" expressions mean what they mean - that is all.
In their attempt to argue for universal inborn limits to the patterns of language that humans can use, Smith and Wilson write
(1979, p.27): "The fact that most languages tend to adopt one of...two strategies for forming relative clauses is itself quite
striking: logically speaking, there are thousands of alternative possibilities."

Page 44
First of all, are there really thousands of possibilities? I think not. But even if there are, humans are likely to restrict themselves
because there are only a few ways of expressing relatives that are practical, not because they are genetically programmed. In
any case, if only most languages have these features, it is hard to see how the tendency can be a human genetic trait; if it was,
surely nobody would use any other way of expressing a relative. As examples Smith and Wilson offer, among others, the
sentences
12 I read the book that you read.
and
13 I read the book that you read it.
12 is the way languages like English and French express the relative idea, 13 the way languages like Hebrew express it.
But Japanese, for example, does not express the relative idea in either of these ways - it does not even have relative pronouns.
14 Watashi wa anata no yonda hon wo yonda
I subject you subordinate read book object read.
subject
But their faith in all pervading inborn syntax can lead linguists astray far worse than this.
Smith and Wilson ask us to consider the phenomenon of 'contraction', the process which provides alternative pronunciations of
is. They point out that pairs of sentences like 15/16 and 17/18 appear almost identical, but contraction is possible only in such as
15 and 17, not 16 and 18.
15a I'll ask whether the party is on Thursday.
b I'll ask whether the party's on Thursday.
16a I'll ask where the party is on Thursday.
b* I'll ask where the party's on Thursday.
17a I wonder when the party is happening tonight.
b I wonder when the party's happening tonight.
18a I wonder when the party is tonight.

Page 45
b* I wonder when the party's tonight.
Smith and Wilson say there is a significant generalization to be made about this, and that the correct one is syntactic, as follows:
'Echo' questions related to 15-18 can be formed by placing the wh-word in the middle of the sentence rather than at the front of
its own clause1:
15c You think the party is on Thursday?
16c You think the party is where on Thursday?
17c You think the party is happening when tonight?
18c You think the party is when tonight?
Here contraction of is is possible in all cases.
Syntactic rules, they say, move constituents from one place to another. For instance, wh-words can be moved from sentence-
internal position to the front of a clause. Thus sentences 15-18c may be converted into 15-18a. Then, if a wh-word is so moved
from a position immediately after is (16 and 18) "contraction is blocked"; but if the wh-word is not moved from that position is-
"contraction goes through quite freely" (15 and 17). "This seems to be," they claim, "a clear case where a phonological rule
must have access to syntactic information". (Smith and Wilson, 1979, pp.66-68)
Their explanation looks pretty good; that is, if one is simply looking for something that appears to neatly fit the limited
examples they give.
In fact it is rubbish. It is bound to be rubbish when one considers what they are asking us to believe. This is the idea that
children are born with subconscious knowledge of a purely arbitrary, imposed rule without any foundation in either logic or
practical necessity, and of a very complicated kind. A knowledge involving inner awareness of a potential word order that many
will not come across till they are quite old and will possibly never use themselves, and involving awareness of the effects -
purely arbitrary, remember - of moving in the mind only certain words from these only hypothetical positions to their actual
positions. A knowledge involving, furthermore, the minute detail that if that hypothetical position happens to be immediately,
but only immediately, after "is", then this arbitrary rule is subordinate to another rule, equally arbitrary: the original rule doesn't
apply. It is not just one rule of this kind that children are supposed to know when they are born. They are supposed to know
thousands of other rules of
1 Notice how their so-called 'echo' questions are not actually related to the original sentences at all. They are chosen
arbitrarily to fit their exposition. (And the omission of a wh-word from 15c is Smith and Wilson's own.)

Page 46
the same kind, equally complicated, equally arbitrary, and often equally based on imagining word movements from imaginary
positions.
But this is not the end of it by a long way. We are also asked to believe that all humans are born with knowledge of this
particular rule, as well as all the others, although they are likely to be born into one of the thousands of other languages, just in
case they are born into an English-speaking community. A Japanese unborn child is all ready with this rule and all the other
rules of English (and the rules of all the other human languages) just in case she is born and brought up in Britain or North
America, say; and an English child is ready with all the Japanese rules (as well as all the English and all the other rules) in case
she is born and brought up in Japan.
Perhaps I have misunderstood the Chomskyan claim here. Perhaps they are not really saying that children are born with
knowledge of all these rules. But if not, where do they get them from? Because, as Smith and Wilson themselves declare
(apropos another rule) this sort of rule is never taught or mentioned in schools.
As regards the present instance, add "always" after "is" in 16c:
16d You think the party is always where on Thursday?
The wh-word is no longer immediately after "is". So according to Smith and Wilson "contraction goes through quite freely" if
where is moved to the front.
16e* Where do you think the party's always on Thursday?
But the contraction's as 'wrong' there as without "always".
Equally, there are no end of sentences which according to Smith and Wilson ought to be 'wrong', but which are in fact 'right'.
19 I wonder when the party's on a weekday. cf. You think the party's when on a weekday?
20 When do you think the party's at Bill's place? cf. You think the party's when at Bill's place?
21 Who do you think's the tallest? cf. You think the tallest is who?
22 Where do you think he's on holiday? cf. You think he's where on holiday?
23 When's the party? cf. The party's when?
24 Where's Bill? cf. Bill's where?

Page 47
The basis of the contraction of "is" is in fact lack of emphasis, which is just what one would expect. But one does not get this
lack of emphasis on "is" unless something else is emphasized. So one can always contract "is" to "'s" if one is saying (or
questioning) that something is something that carries the emphasis and takes it away from ''is". But if there is nothing after the
"is" to emphasize, the emphasis naturally falls back on the "is" and one cannot contract it. One can see this principle at work in
all the sentences above. In 15a/b, 17a/b/c, and 19-24, one is saying that the party is something: is on Thursday, is happening
(with the main verb "happening" actually spoken, the mere auxiliary "is" is clearly unemphasized), is on a weekday, is at Bill's
place, is the tallest, is on holiday, is the party?, is Bill? So one can contract "is" in all these.
But in 16(a) and 18(a) one is not saying something is something; one is effectively saying
16ai I'll ask where (always) on Thursday the party IS.
18ai I wonder when tonight the party IS.
The principle can be seen working again in
25a Why do you think the party's formal tonight?
b I think it IS tonight (= Tonight I think it IS)
c because it's somebody's birthday.
'Full' "is" is needed in
26* When's it?* Where's he?
because pronouns are by the very nature of their meaning normally unemphatic. Yet
27 It's me!
where the pronoun is emphatic. And further:
28 When's it formal? Where's he now?
because "formal" and "now" add enough weight or emphasis to the pronoun 'somethings' to make contraction sensible. (Cf. He
is where now? which is supposed to make contraction impossible!)

Page 48
This is a real explanation, because it explains why the usage is as it is.
Once more, everything is basically determined by meaning. The principle for contracting "is" is a further example of the circle
of meaning dependent on meaning. Unfortunately even linguists who might admit they were wrong in detail in this case will
probably turn my explanation into abstract technical terms and then claim that the basis of "is" contraction is still 'syntactic'.
They might argue, for instance, that "on Thursday'' is syntactically different in 15 from 16. But, as always, one can't analyse the
syntax till one knows the meaning. At this point, to continue to insist that meaning or phonology depends on syntactic analysis
is surely perverse.
I have deliberately not used expressions like "complement" or "adverbial phrase", not to try to show I was not talking about
'syntax', but because such analytical abstractions do not tell one what English speakers experience. My own explanation is of
course not entirely 'realistic' either; no explanation can be. But I have tried to get as close as possible, and describe directly what
actually happens, instead of using labels of a kind that often only act as veils that stop people looking properly at what is really
going on.
In a discussion of people's intuitive knowledge of their language Smith and Wilson (1979, pp.39-40) present the sentences
29 I'm leaving, for he makes me nervous.
30 I'm leaving, because he makes me nerevous.
Obviously these are both possible sentences. They then give other similar pairs of sentences; one pair will be enough here:
31a Because he makes me nervous, I'm leaving.
b* For he makes me nervous, I'm leaving.
31(b) is of course not possible in English, and Smith and Wilson say that most native speakers will be able to judge that it is not
possible. "...such judgements" they say, "give us good ground for imputing a particular type of linguistic knowledge to the
speaker: in this case, knowledge of syntactic structure." They say the distinction between "because" and "for" seems to be that
between subordinate and co-ordinate clauses. "For", "and" and "but" link co-ordinate clauses of roughly equal importance, while
"because", "although" and "when" introduce subordinate clauses dependent on the

Page 49
main clause. "It is the fact that co-ordinate clauses have very limited freedom of movement, compared with subordinate clauses,
which explains the differential patterns" in the "because" and "for" sentences above.
Once more they ask us to believe something very improbable: that all humans are born with a knowledge of the syntactic
difference between "for" and "because"; that children have an innate capacity to identify "for" as a coordinating word and
"because" as a subordinating one - that they know how to use and how not to use "for'' because they know about co-ordinating
clauses, and they know that "for" belongs to the co-ordinating category. One would have to suppose once again that children
know the syntactic classification of all words in all languages.
There is only one way children can know about "for". That is by learning from observing what it means.
But this case is also revealing, because in fact their explanation of the distinction between "for" and "because" was relegated to a
footnote, and the quotation above was prefaced by the statement that "For our present purposes it is not the actual rules which
explain the speaker's linguistic judgements that are of interest. What is interesting is that such judgements give us...[etc.]" (see
above). If they assume that humans have special inborn linguistic knowledge then they will of course find, if they can, abstract
distinctions like the one above between co-ordinate and subordinate clauses. It doesn't enlighten us in any way about the human
mind. Once again it is simply the declaration of a lifeless rule, an arbitrary command to the mind. It leads to no useful
conclusion.
It is another of their false problems that they have created. They have confused themselves by fastening onto this particular pair
of words, "for" and "because"; in certain contexts, like the one they have chosen ("I'm leaving, for/because he makes me
nervous.") it happens to make no practical difference to the meaning (see below, pp.107-08) whether one uses "for" or
"because", so they immediately decide there is a problem to be solved, and because they think the meanings are the same (they
aren't) they turn to genetically acquired syntactic behaviour as the answer. But take a pair of sentences like
32 I have a white cat and a black one.
33* I have a white cat but a black one.
Surely nobody is going to find any problem here, let alone start looking for syntactic explanations. Everybody knows that 32
makes sense and 33 does

Page 50
not, simply because of the different meanings of "and" and "but", although both are what grammarians would call co-ordinating
words.
Exactly the same applies to "for" and "because". They have different meanings. If one starts with
34 He forgave them, for they knew not what they did.
one can just as well, for practical purposes, say
35 He forgave them, because they knew not what they did.
although I am certain that many will be uneasy even at this stage, for they will sense the difference in meaning produced by
replacing "for" with "because". But if one starts with
36 Why did he forgive them?
there can only be one way, not two ways, of answering:
35i (He forgave them) because they knew not what they did.
The sentence
34i (He forgave them) for they knew not what they did.
would be nonsense as a reply to 36. So it comes down to the explanation that it nearly always comes down to in the end, a
simple difference in meaning. "For" and "because" do not mean the same. Because of the mere accident that "for" and "because''
are in the same 'area' of meaning - while "and" and "but", for example, are not - Chomskyans get excited about them on
completely false grounds.
Differences in meaning, like those between "for" and "because", may happen, in certain contexts, to result in differences that
can be analysed in terms like co-ordinate and subordinate clauses. But note that even these sorts of abstract term, so far as they
are valid, are really about meaning too.1 In the end it is a matter of reality and unreality, of sense and nonsense; the distinctions
are rooted there, not in abstractions. The linguists, though, are so isolated from reality, so cocooned inside their ideological
conviction that
1 See Chapter 6, p.130.

Page 51
all depends on syntax, that once more they look only for evidence that supports their own thesis.
In Reflections on language (1976, p.90) Chomsky offers the sequences
37 who did John believe that Henry said that Tom saw?
and
38 who did John believe the claim that Tom saw?
to show that general principles of innate language faculty are operating to make 38 impossible. He brings a great machinery to
bear on the problem, a machinery of 'initial phrase markers', 'cyclical derivation', 'NPs', 'COMPlementizers', 'wh-movement' and
- apparently the key- 'subjacency principle'. Many innocent readers may have already seen the simple reason why English-
speakers do not say 38. The meanings of "believe" are quite different in the two sequences. In 37 "believe" has the sense of
"think", while in 38 it has the sense of "accept as true''. So in 37 it is followed by a "that" idea which leads on quite logically to
the "who" at the beginning of the sentence. In 38 the question is already logically complete when one gets to "claim" - there is
no more room for any "who". In other words, if there is a coherent question at all, it is "Did John believe the claim?" and "that
Tom saw" is simply the description of the claim. Add "who" and the sequence, contrary to Chomsky's strange assertion, does
not have definite and unambiguous meaning. It is illogical rubbish, as one can confirm by putting "who" at the end instead:
38a* Did John believe the claim that Tom saw who?
The only difference between this and 38 is that 38 increases the confusion by suggesting that "who" is perhaps the object of
"believe". But notice how one can produce a perfectly logical idea using the same sequence for a start as in 38 if one changes
the meaning of "believe" back to that of 37:
39 Who did John believe (that) the claim that Tom saw him was made by?

Page 52
In fact, though, much of my argument above, although valid, is comparatively superficial and perhaps unnecessary. For the most
stark and final objection to Chomsky's case one goes back to that double meaning of "believe". I have already pointed out that in
38 "that Tom saw" is no more than the description, or 'identification', of "the claim". So what human logic must protest against is
really
Who did John believe the claim?
In a conversation with Bryan Magee (Magee, 1982, p.181) Chomsky says that in English, roughly speaking, one can form a
question from a statement that contains a name by questioning that name. For example:
I saw John. Who did I see?
He thinks that he saw John. Who does he think that he saw?
So a plausible rule for English, he says, would be 'To form a question, take the position in which a name can appear, put in that
position a word like "Who" or "Whom" or "What", and move it to the front of the sentence.' Chomsky says that although this
rule works over a wide range it fails in some interesting cases. By the rule he proposed, the question from
40 He wonders who saw John
would be
40q Who does he wonder who saw?
"Well, we know at once that that's not a sentence. You may say it's not a sentence because it fails to be meaningful, but
that seems quite wrong. In fact the pseudo sentence is perfectly meaningful. If it were a sentence, we'd know exactly what
it would mean. 'Who is such that he wonders who saw him?': that's what it would mean - but we don't say it. It's just not
one of the allowable sentences of English. So there must be some principle, part of English grammar, that prevents us
from saying it. Yet it's extraordinarily unlikely that any such principle was ever taught to anyone."

Page 53
Well, to begin with, I certainly did not, at first, understand Chomsky's 'pseudo sentence' 40q, despite his claim that we would
know what it would mean. And I don't understand, either, what he says it would mean - that really does seem double-Dutch to
me, at any rate in connection with 40.
Chomsky is guilty of a terrible absolutism. For him, it seems, word sequences are sentences or they are not. His formal rule-
bound approach is bound to lead to this. He appears not to see that what is important is not whether a lump of language is a
sentence or not; it is whether it makes sense - and whether it makes sense to the individual concerned. Whether a lump of
language makes sense to someone is often - particularly when the connections are more obscure than usual - completely
subjective. One person may see how it links up with reality, be able to recognize the reality it is connected with, another person
may not. I'm sure, too, that Chomsky's interpretation of 40q - "Who is such that he wonders who saw him?" -means perfect
sense to him. I'm not saying he is wrong to give it that meaning. My point is that in cases like this it is not a matter of being
right or wrong. And one and the same person may at one moment not understand at all, and at another moment find that the
pieces of language all fit together in his or her mind and make sense. For example, is
The man the man the man knew knew knew,
a sentence? What a pointless question. What is important is, does this combination of words make sense - to you, or to me, or to
her? Some people may be able to dig a recognition of reality out of it, but I find it very hard most of the time, even though I
invented it. For a few seconds, if I work terribly hard, I think I've trapped the sense, but then it eludes me again.1 Whether a
person recognizes sense in any 'sentence' that is logically rather 'involved' can depend on several factors. One of the most critical
is whether it is prepared for by the context. And some people are quite simply quicker in the uptake than others.
A basic difficulty is that words have to come one after the other, and there is a limit to the number of places one can put them
in, and so there is a limit to the connections that can be made clear. A sentence may start off in a certain way that leads the
listener to expect a meaning quite different from the one intended. There are restrictions, but it is a purely practical problem, and
nothing to do with unconscious inborn inhibitions.
1 I discuss various implications of this 'sentence' in Chapter 6, p.119, Chapter 7, p.137 and Chapter 8, p.155.

Page 54
Another factor that is often decisive is that the different meanings of different verbs (say) may determine the logic of a sentence,
or, vice versa, the particular logic of a sentence may determine the meaning of a particular verb.
Below are some examples of 'difficult' sentences, some more difficult than others. They are all questions, like Chomsky's
'impossible' sentence. I have put the remarks that might have provoked them at the end of the line instead of at the beginning, so
that readers can test themselves first on the question and then see if the context makes it clearer. In many cases stress is critical
to the sense.
41 Who do you know saw? I know Sue saw.
42 Who do you know saw
Sue? I know she saw Sue.
43 Who do you know she
saw? I know she saw Sue.
44 Who do you know who
saw? I know who saw Sue.
saw? I know a postman who saw.
saw? I know Tom saw Sue.
47 Who do you wonder saw? I wonder whether Sue saw.
48 Who do you wonder
whether she saw? I wonder whether Sue saw.
whether she saw? I wonder whether she saw Tom.
50 Who do you wonder saw? I wonder Tom saw!
51 Who do you wonder saw I wonder Tom saw anything at all, let alone
anything? what Sue was doing!
52 Who do you wonder who I wonder Tom saw Sue at all, let alone what
saw? she was doing.
53 Who do you wonder who
saw? I wonder who saw Sue.
whether who saw? I wonder whether Tom saw Sue.
55 Where do you wonder
who went? I wonder who went to Bath.
56 What do you wonder who
saw? I wonder who saw the flying giraffe.

Page 55
Notice how many confusing things there are about these questions.
The first "Who" is governed by the "you" verb in some questions but by the final verb in others - sometimes even in apparently
identical sentences (44,46/45) or sentences that begin identically (41,42/43 and 50,51/52,53) or sentences identical except that
the "you" verb is ''wonder" instead of "know" (45/52,53).
Chomsky would presumably disqualify 44 as a sentence, as it is the same as 40q, except that "know" replaces "wonder". Yet it
turns out to have two more perfectly logical and comprehensible meanings (for me at least) once the contexts are made clear
(45, 46). And one can even grasp 44 if one can escape the temptation to think it is going to mean 45 - the stress pattern is the
same.
The confusion gets worse when "know" becomes "wonder", because "wonder" has two meanings.1 In 50-52 it has the sense "be
surprised", not the "ask oneself" of the others. It ought to have the "surprised" sense in 47 too. 47 and 48 are strictly illogical as
questions apropos "I wonder whether Sue saw". A really logical question to fit that statement is impossible, but I believe many
native English speakers would ask 47 or 48, and in context they are comprehensible and accepted. There is no absolutism here.2
When there are two "who"s, though, things get difficult. Even so, with the stress and context clear, the question can be perfectly
coherent (52) and turns out to be exactly the same as the 'disqualified' 53/40q, with quite a different meaning! But 53/40q itself
is treacherous because it has exactly the same stress pattern as 45 and one is probably tempted to see "wonder" as working in
the same way as "know".
Yet even these difficulties become less if one adds a guiding "whether" to the two "who"s (54) or changes the first "who" into
something else (55,56). Who is prepared to state absolutely and dogmatically that 56 is not an English sentence? If any or many
of my interpretations above are muddled, that will only thoroughly confirm that this is indeed a matter of subjective uncertainty.
It is interesting that apparently the question is quite simple and unremarkable in Japanese:
dare o dare ga mita no daro ka
who object who subject saw (you)wonder question?
1 So does "know", but the effect is not so important.
2 See p.138 for another example of this accepted illogicality - apparently accepted Chomsky himself!

Page 56
In their discussion of generalizations which leads up to their treatment of the question of a universal theory of language, Smith
and Wilson (1979, pp.240-45) offer the following example, among others:
57 I want to invite that boy to my party.
This, they point out, can be changed into
58 That boy, I want to invite to my party.
But, they say,
59 I want to invite this girl and that boy to my party.
cannot be changed into
60* That boy, I want to invite this girl and to my party.
This is typical of the sort of misleading example that Chomskyan linguists so often use. Smith and Wilson use it, and other
sentences of the same kind, to justify a rule that prevents moving a Noun Phrase out of a coordinate Noun Phrase and Noun
Phrase structure, and go on to suggest that this is a rule that applies to all languages. (They cite examples from French and the
African language Nupe.)
Note that they are using "and" in a very peculiar way that nobody would ever want to use it in in real life, in a way that joins
one thing to something else quite nonsensical: "this girl and to my party". However, there is nothing wrong, except clumsiness,
with sentence 61, 'taken' from 59, where "and'' is used as one would expect.
61 This girl I want to invite, and that boy, to my party. (Better still: This girl I want to invite to my party, and that boy.)
Position is obviously often an essential part of the meaning of a piece of language. "And" goes before the thing it 'adds', so it is
a very silly example that puts it after all the things that are to be joined have already been mentioned. "Too", on the other hand,
goes after the thing it adds.
62 This girl I want to invite, that boy too, to my party.

Page 57
Here, of course, "too" is in its right position, the position that is absurd for "and". So actually there is nothing to stop one
'moving an NP out of a coordinate NP and NP structure', so long as one does it in a way that makes sense. But, once again,
people do not in fact produce sentences by 'moving' words from another more 'basic' sentence. People simply put pieces of
language direct where they want them.
So here the linguists illustrate once more how systems of abstract analysis not only lead to false conclusions as to specific rules
but also strengthen the broader false belief that human language is a special separate faculty, programmed with arbitrary and
narrowly restrictive rules. Their confusion over 60 is one more example of what obsession with abstract structures and failure to
see the simple importance of meaning can lead to.
But they might concede, I suppose, that 61 is all right, but insist that they are concerned with the impossibility of 60. What they
apparently fail to see is that the problem in 60 has nothing to do with co-ordinate NP and NP structures. It is simply that it is
nonsense to say:
I want to invite this girl and
Do linguists formulate universal rules to forbid sentences such as those in 63*?
63* This is the mouse which
That is an instrument for
I know that she
I am not sure of the answer to that question. But their position is indefensible whatever the answer. If their answer is (as I
suspect) "yes", they do, it is a confession of the fruitless pedantry of their concerns, pedantry that ignores the obvious principle
of meaning that prevents anybody wanting to, or needing to, or being tempted, to say such things. If their answer is "no", they
don't, they are illogical and inconsistent, as the 'sentences' of 63 are exactly the same in principle as 60.
Smith and Wilson continue to show their basic misunderstanding of the non-problem when they move on to give an example
from the African language Fe'Fe', one of those "which permit the extraction of NPs from coordinate structures." They are
exercised by this because it is a "violation of the supposedly universal ban on such extractions." Once more they are being
distracted by irrelevant figments of their imagination. I know

Page 58
nothing about Fe'Fe', but one thing is certain: like all languages, it is used in a way that makes sense.
Finally, an example where Chomsky resorts - not for the first time - to utter nonsense in support of his theory; once more he is
trying to demonstrate humans' 'genetic endowment' with abstract grammatical rules. (Magee, 1982, pp.181-82)
64 He told the class that the book was difficult.
64q What class did he tell that the book was difficult?
65 I asked him to tell the class that the book was difficult.
65q What class did I ask him to tell that the book was difficult?
but
66 The book is more difficult than he told the class that it was.
66q What class is the book more difficult than he told that it was?
This is a typical result of the Chomskyan method, in which reality is ignored and artifical word juggling goes on. In the first two
questions (64q, 65q) everything goes all right - naturally, because Chomsky asks the only sensible question about the class that
is possible in the circumstances, i.e. "What class did he tell?" Then suddenly in 66q he arbitrarily abandons the 'telling'
connection and starts, it seems, - to the extent that he does anything coherent at all - making comparison between the class and
the book, and claims it shows there is something about that particular 'embedding' that prevents us questioning "the class". He
insists on defying meaning and practicality and instead engages in mechanical, mathematical word transposing, in order to make
a quite irrelevant and nonsensical comparison. People form questions, or any language, to say what they want to say, not to
follow some rigid geometrical formula. If Chomsky is determined to pursue abstract gibberish he will get gibberish. In real life,
people don't want to play word games or follow regular procedures when they use language; they use it to produce meaning.
They use whatever words and sequences are necessary to that meaning. It isn't demonstrated that their freedom in using
language is limited by genetic or any other sort of restriction just because these artificial abstract procedures don't always
produce sense. (But as I shall point out later, there are restrictions of a practical kind.)
The question that real people would ask about the class would be something like

Page 59
67 What class did he toll that the book was less difficult than it is?
Alternatively, if one wanted to pursue Chomsky's nonsense comparison of class/book one would have to say something like
68 What class is the book more more difficult than he told it (i.e. the class) that it was?
Chomsky is trapped in a vicious circle of his own making. He inevitably creates non-problems by the very method he uses; his
picture of language and how it works can lead to nothing else.
If one attacks linguistics in general, linguists and others 'up to date' in intellectual affairs will very likely tell one that Noam
Chomsky has, like Freud, changed the way we think about the human mind. Chomsky may have made mistakes, and may need
to be revised, but he has opened our eyes to the fundamental importance of language and its place in our nature. But if one
exposes the flaws in the Chomsky fundamentals, most linguists will immediately deny association with him. He is out of date,
superseded. What was true once is no longer true. Linguists are selective. Each linguist draws on various sources and combines
them into his or her own theory. Thus naively linguists have to resort to this effective admission of futility in order to keep their
self-respect and justify their existence.

Page 60
4
Unapplied Linguistics
One of the great illusions that modern linguists appear to have about their work is that it has led to a more flexible and generous
approach to the forms of language used by others, a less narrow view of what is 'correct' in grammar, expression and accent. But
in fact the only knowledge needed as the basis of a tolerant attitude is of the most elementary kind possessed by practically
everybody: nobody of any 'class' speaks or writes as people speaking the 'same' language did three or four or more centuries ago.
One needs only a very small amount of realism and logic to see that that means language changes and so there cannot be one
fixed 'right' form, or that, for example, there is no reason why conventions of spoken language should be inferior to those of
written language. There is one simple key to greater tolerance towards other people's language, and that is greater tolerance! It is
ironical, in view of the linguists' assertions, that language snobs and dictators do and can only condemn 'deviant' usage precisely
because they know about it. For linguists to claim the credit for an improvement in attitudes is irrational vanity. The only
antidote to authoritarian and exclusive attitudes is non-authoritarian and anti-exclusive attitudes. 'Science' can have nothing to
do with it. Marxism, I think, provides an instructive parallel. Marxists claim to possess a scientific political theory. I know
personally some very gentle and very tolerant Marxists, and there are certainly many such. But for all the 'scientific' knowledge
they claim, and in many cases probably precisely because of their pride in it, Marxists have been responsible for some of the
most intolerant savageries of this century. As for a broad-minded and respectful approach to other people's 'different' language, I
suggest that many came to such an attitude long before they knew anything of modern linguistics.
PhD's do not like such ideas. They prefer to think all progress, theoretical or practical, results from research. Indeed it seems
today as if almost everything, to be accepted as true, has to be the product of 'research'. Even if it is not really research it is
better for the author of the stuff to call it that. Much of the 'research' is nothing but speculation in order to help build a system.
Much of it seems to be done and used only to confirm a theory -more often than not a theory that contradicts other theories,
themselves claimed to be founded on research. Sometimes, even, the evidence gathered is used to demonstrate the opposite of
what it does demonstrate. Research is so much respected that not only is a lot of work done that is quite

Page 61
unnecessary - although it may come to perfectly valid conclusions; many people are led to believe in the truth of theories or
statements that just a few moments' thought would show must be false.
With rational thinking there is quite enough evidence around us and in us for one to settle very quickly many of the 'problems' of
language that the 'learned' have worried about for a long time. Research is not needed to establish, for instance, that practically
all human beings are physically able to make any language sound that any other human being can make. This is obvious from
the fact that children of many different races master the pronunciation of English in the United States, or of Portuguese in Brazil
etc., and that one has never heard of any child brought up in a 'foreign' country who cannot pronounce the native language
exactly as the natives do. There may turn out to be exceptions, but the general principle is clear from simple information that is
easy to get.
It does not need research to see that there is only one way anybody can ever learn a language, whether it is their own or a
foreign one, and that is by observing - listening or reading, or both. Speaking cannot play any direct part in learning a language,
because first one has to have something to speak; though of course speaking, either with one's mouth or inside one's head, can
help one not to forget, and the need to speak may make some people find out more energetically and observe more closely.
Yet this obvious truth is resisted almost everywhere. Perhaps people in our research obsessed age will only recognize it when it
is announced as the product of a grand program of scientific investigation - and so recognize it for quite the wrong reason.
In the same way, perhaps people will only now begin to admit various other obvious things in connection with language,
because researchers are apparently claiming to have 'discovered' them as a result of their labours. For example, that correcting
foreign language learners' mistakes does not work, that young children learning languages at school do not do better than older
people, that big (or small) institutions of experts providing intensive teaching of foreign languages (such as military
establishments, full of 'motivated' pupils) are ineffective, that courses in English, or any language, for 'special' or 'specific'
purposes are beside the point, that people cannot master language unless they understand it and cannot understand it unless they
recognize in it things in life they have experienced. (Linguists on whom a part of this last has apparently only recently dawned
talk of "comprehensibility input", as if humans were some sort of electronic device, and as if to reassure us that this is new
'scientific' discovery.)

Page 62
Such truths, with the contrasting fact that young children who go to live in a foreign country learn the new language very well
and easily, are obvious, and should have been obvious for decades or even centuries, without the help of any research, to anyone
who has just a little experience of the matter, who thinks a little, and who is honest. But they are uncomfortable truths for the
education powers that be, and so they have not been thought about and not recognized, or at least not publicly.
Nevertheless, supporters of linguistics insist that it makes, or can make, a great, even fundamental contribution to the teaching of
foreign languages. There is not the slightest evidence that it has done so or ever will. Linguists might say that linguistics has not
even now properly penetrated language teaching circles for practical purposes. I think that may in a sense be true, and it seems
significant. The reason is very likely that most teachers have found that in practice linguistics has nothing to offer.
In over thirty years of teaching English as a foreign language to students of many different nationalities from many widely
varying language and cultural backgrounds I have noticed no improvement at all in the mastery of it. Naturally, as English has
more and more become the world language, more and more people are taught it and use it. But they do not speak it or write it
or understand it better than people did before.
There is no objective way of testing that statement. Examination results cannot be used to show anything either way. Any
examinations or parts of them which were or still are of the old fashioned, traditional type were and remain subjective, and so
cannot be used for assessing trends over the years objectively. So called 'objective' tests (usually based on 'multiple-choice') are
equally useless, for at least three reasons. First, they were not used earlier, and have gradually been used more and more, so
there is no way of comparing exams 'now' and 'then'. Some might hope that in time, when 'objective' tests have formed the same
proportion - or even all - of certain exams for several decades, it will be possible to make a valid comparison. It won't.
'Objective' tests, by and large, do not test the things that really count in mastery of a language. Using a language in real life,
even 'just' comprehending it, is a creative activity; that is, one has to get the ideas of what the words mean entirely out of
oneself, spontaneously. Various possibilities are not put into one's head first. Nor is there any objective measure possible (even
if one accepted the validity of multiple-choice testing in general) of exactly what variants of 'objective' tests are truly significant
or of what weighting to give to each such test in an exam. Furthermore, it is fairly easy to train people to do well in 'objective'
tests, as I know from personal experience over several years. Conversely, many

Page 63
candidates fail abysmally merely because they are led astray by the 'suggestive' nature itself of the multiple-choice method.
Finally, there is the irony that some examining bodies go to great and conscientious lengths, using elaborate statistical methods,
to make sure their tests are appropriate - and use, perforce, contemporary students as their guinea pigs for testing their tests! So
that is yet another factor making it impossible to use exams as a means of comparing language abilities over the years. One is
left with nothing but personal impression to go on. My own is that, if anything, standards have fallen somewhat.
If I am wrong, though, and linguistics has in fact helped the learning of foreign languages, then one has to draw a strange
conclusion, because the different schools of linguistics disagree with each other, often fundamentally (transformational and
systemic linguistics, for instance). If there has been an improvement in foreign language learning as a result of linguistic
research, it must be for quite the wrong reasons, as it would be too uncanny if students and teachers had some wonderful
intuition which made them able to choose the right school of linguistics without knowing it.
What I am certain about, both on logical grounds and from personal experience, is that analysing structure is a very bad way to
try to learn a foreign language. The way Latin used to be (is?) taught, on the basis of structure analysis but with almost no
reference to meaning, is an extreme example of how useless and unreal the belief in structure is. The analytically minded student
nearly always finds it difficult to learn a foreign language properly. I am not saying that no person who believes in linguistics
can learn foreign languages. That is obviously untrue. People may learn a language well in spite of their theories, and a lot of
those who study linguistics are probably people who are already good at foreign languages anyway. What I do say is that a
person will fail hopelessly if at the practical level she stops thinking in a natural way about meaning and tries to apply an
anlytical method taught her by the linguists. I have seen it happen several times. Some advanced students who have expounded
transformational grammar to me, for instance, and applied it to their studies in English have not been able to get more than
about one sentence in ten 'grammatically' right. These are extreme cases. Most students have 'grammatical' difficulties, and
when they fail I believe they, too, often do so mainly as victims of formal structure teaching, but of a more old-fashioned,
cruder, and probably less harmful kind than the new cult mysteries.
It is not surprising that structural analysis tends to bog students down. A complicated system, in effect a new 'language', much
more difficult to master than the one they are aiming at, is put between students and their

Page 64
goal. The whole point of language is meaning, yet the linguistics approach virtually tells students they must only approach the
meaning through an interpreter, the analysis. This is madness.
Luckily everybody who uses their native language 'knows', without the benefit of linguistic analysis, that meaning depends on
meaning.1 When, for example, English-speakers distinguish between "can" and "am able" they are recognizing differences of
situation, not abstract structural differences. Compare
1 When I was a young man I could milk a giraffe in five minutes.
with
2 Yesterday I was able to milk the goat in nine minutes.
It is 'impossible' to use "could" in the second sentence. A student of English as a foreign language, too, can only appreciate the
distinction in terms of meaning. English-speakers know, without the benefit of linguistic analysis, that if in the sentence
3 I have not seen any football since I came to London.
one replaces "come" with "be" and wants to keep the same practical meaning the sentence must be
4 I have not seen any football since I have been in London.
Keeping the same tense would produce an opposite meaning:
5 I have not seen any football since I was in London.
The point is, of course, that the 'general' and 'specific' meanings of "came" do not combine to produce the same meaning as the
'general' and 'specific' meanings of "was" - at least, not in this particular situation in life. Similarly
6 When he comes we'll have supper.
describes a perfectly normal state of affairs. But with exactly the same tenses
7 When I wash the car we'll go for a drive.
suggests a rather wet and unusual trip. No doubt most people would regard the normal procedure as being:
8 When I have washed the car we'll go for a drive.
Again:
1 See Chapter 5.

Page 65
9 He admitted reading her letter.
has the same practical sense as
10 He admitted having read her letter.
But
11 He admitted reading very little.
is something quite different from
12 He admitted having read very little.
No elaborate analytical scheme is going to increase inner understanding of this - quite the contrary. The practical harm done by
the structural approach is perhaps shown up even more starkly when it comes to trying to master expressions like 13a-15a
below. Even really advanced students of English tie themselves up in knots, if they are structure-minded, when they attempt to
change the first group of sentences into the second; their native English teachers, too, become contortedly confusing if they try
to explain in structure terms. Yet relative beginners usually catch on quite easily if one gets them to see the 'problem' in practical
meaning terms - or there may never even arise any problem for some if they start off with a common sense approach.
13 They stole his raincoat.
14 Somebody jumped on his bicycle.
15 They shouted nasty things at him.
13a He had his raincoat stolen.
14a He had his bicycle jumped on.
15a He had nasty things shouted at him.
In 14a the usual difficulty is that "on" is left out:
14b* He had his bicycle jumped
All one needs to point out here is that while one does indeed "steal raincoats", one does not "jump bicycles" - 14b is the
equivalent of
14c* Somebody jumped his bicycle
which obviously doesn't work. For 15a one then often gets
15b* He had nasty things shouted at1
1 15b is not in fact an 'impossible' sentence. It makes perfect sense as an alternative to "He arranged for people to shout at
nasty things."

Page 66
- if the student is still playing word-juggling (or transformational?!) games despite one's attempts to bring reality into the
picture. Here one only needs to ask - about 15b - who he had nasty things shouted at. At his teacher? At his brother? At the
police? The meaning isn't finished. Ah - at him.
Students need to be aware that meaning is as decisive in foreign language as it is in their own, and that they should be guided by
it just as in their own. The only way to master any language is to master its meanings. One cannot do this by studying an
analytical scheme. That is truly to put the cart before the horse. Well, actually, not even the cart; just a large rock...If one knows
meanings one knows which fit together.
There was a strong tendency, even before the development of modern linguistics, to treat language as something completely
self-contained, and this is perhaps one of the main reasons why most people found foreign language learning so difficult.
Modern linguistics has made things even worse, because it presents a language, more than ever before, as something in a
vacuum, with arbitrary rules. I have experienced the effects of this approach constantly in my own work. For example, students
of English as a foreign language may ask whether one can ever use 'conditional ''would"' in an "if" clause. I answer that it
depends entirely on what they mean in real life, and explain the situation in which one would use "would" with "if". But so
often - and very often the more intelligent students, the ones most open to intellectual corruption - they are not satisfied. They
demand purely linguistic, formal rules inside a linguistic vacuum, and often I see that they are still anxious about "can one?",
that is to say, about whether "one is allowed to" by those arbitrary inflexible regulations of the language. Their anxiety often
persists, as I say, in spite of the much greater simplicity, and demonstrable 'rightness' of principle, of my explanation. In this
case the truth is just this: one uses "would" (or "will", which will have a different, less hypothetical meaning) with "if" if the "if"
action comes after the main action. For instance:
20 If it would make you happy I would give you a kiss.
21 If these people will run the country in 30-40 years' time, I don't want to be there. [From a letter to a newspaper.]
In other words, if one means "would" or "will" one says "would" or "will". It is not a matter of formal rules at all. A language
has its own conventions of meaning, and on the basis of those conventions one says what one means. It is seldom a question of
"can one?" If the situation in life is an "if...would" situation, then one must and obviously does say "if...would". "If...would"

Page 67
situations are rather rare in life; conditions are usually earlier in time than the actions dependent on them. But not always. As I
have already suggested (Chapter 2) there is really no such thing as a linguistic rarity -there are only rarities in life. When an
unusual situation arises native speakers use the language that represents that situation for them, and it is not difficult for them,
however unusual, because they know all the meanings they need to put together to produce their broader meaning, however rare
that meaning may be or even if it is unique.
And I think the example above also illustrates well how false it is in most cases to talk of 'rules of grammar' that native speakers
know but are not conscious of. The great bulk of 'grammar' in any language consists of meanings, and so when native speakers
are faced with situations like those expressed in 20 and 21, they do not need even my simple explanation about time - that is
merely a clumsy roundabout way of saying what the words of the sentence themselves say direct, so of course native speakers
do not think in the terms of that formal explanation. Whether learners of foreign languages ever need such formal explanations
is another question. I am not sure. As languages are taught now, possibly yes, because there is a vicious circle in which most
learners are trapped: they do not realize language is basically meaning, and this makes it all the more difficult for them to
appreciate direct the meanings of pieces of language in the same way as native speakers do; they concentrate on the wrong
things. For example, instead of simply absorbing the meaning of "would" direct from the context of real-life situations, they are
probably often corrupted in the first place by association with apparent equivalents in their own language, and then ache for
formal analysis of its use.1
The proficiency of students of English in fact varies in accordance with factors that can have nothing to do with the effects of
modern linguistic study. First, people from some countries are clearly, on average, better at foreign languages than people from
other countries. Some might argue that abilities in English merely vary according to how close it is to the students' own
language, and that one must allow for this when assessing the effects of modern linguistics.
It is not as simple as that. For instance, in my experience the Dutch are on average considerably more proficient in English than
Belgian
1 There are certainly patterns of word order and of phonology ("Seldom have I seen..." rather than "Seldom I have
seen...."; "an apple" rather than "a apple'') that language-learners have to master. They are conventions, 'unnecessary'
conventions (see Chapter 6) at least in the case of word order, but have to be followed for 'correct' speech. I do not believe
that even these should be called 'rules', at least not in the linguists' usual sense. I discuss this question further in Chapter 7,
p.147, and Chapter 9, pp.170 and 188.

Page 68
Flemish-speakers, although Flemish is essentially the same language as Dutch. The vocabulary of English should in the main be
easier for speakers of Latin languages than for those who speak Germanic languages. And there are wide differences in the
average standards of people from the various Latin countries. Japan claims the highest literacy rate (100%) in the world, and
modern linguistics has certainly penetrated there. Yet for such a sophisticated and in so many ways efficient society the standard
of English is quite astonishingly low, and that includes, decisively of course, a very big proportion of those responsible for
teaching it, even at the highest level. A Japanese scholar has made the interesting suggestion that the first generation of Japanese
English-speakers had a better command of the language than modern Japanese do because they learned direct from native
speakers - I should add, back in the 19th century long before the days of modern linguistic analysis.
The three main languages of the Middle East each belong to a different family: Semitic, Indo-European, and Turkish. Yet
speakers of all three tend to have the same sorts of difficulties, particularly in the early stages. Very often, especially in writing,
they produce English that for the native speaker is incoherent and incomprehensible. Yet there are many Arabic-and Persian-
speakers who leave institutions of instruction in the English language, and go to other institutions in Britain or elsewhere where
they use English in their daily work or studies, and then, only then, become very proficient in both speaking and writing the
language. A few of these are among my personal friends. They sometimes thank me for having taught them English. It is nice of
them to say so. I know that the truth is quite otherwise. They have taught themselves, after escaping from my and my
colleagues' ministrations.
It is thought-provoking to contrast the typical Japanese experience with that of the middle-easterners. Japanese-speakers are
more seldom incoherent and incomprehensible than the latter when they first come to Europe, even if they are at a fairly
elementary stage. Most of them - again unlike the middle-easterners - are able to pass first level exams in English after quite a
short time. Then they get stuck, and it is only a tiny minority that ever manage to succeed in higher level exams, even though
many make several attempts.
The factors that determine how successful people will be at foreign language learning are clearly very subtle. How subtle is
illustrated by an example from the field of pronunciation. Both Swedish and Norwegian have effectively the same "o" and "u"
sounds to choose from to represent the standard English vowel sound in words like "do'' and "too". Yet while Swedes

Page 69
usually choose the roughly 'right' "o" sound, Norwegians have a marked tendency to choose the 'wrong' "u" sound. So two
neighbouring peoples with very similar cultures and very similar languages react differently to the same problem of English
pronunciation.
Facts like those I have mentioned in the last few pages show, then, that being good at foreign languages is dependent on, among
other things, strong and so far intangible cultural factors that we do not understand at all. Linguistics, whatever validity it may
or may not have, has had no impact on these factors.
Even more striking and obvious, though, than the differences in foreign language learning ability between nationalities are the
differences between individuals. These too we know absolutely nothing about, although they are decisive. One can sometimes
point to basic things a language student is doing wrong, and point to basic things that she ought to do. And sometimes, though
unfortunately rarely, this can decisively affect the progress of the student. But usually far deeper processes are at work, and
certainly the analyses of linguistics are wholly irrevlevant to those processes. Why do some people have a 'feel' for how foreign
languages work (very often even if they have utterly benighted conscious ideas on the subject) and others not? What makes
some able to use their imagination far better than others to increase their vocabulary by reading? Why can some 'listen' properly
and so almost immediately have a good accent, but most not? Why can some at once apply good advice, or exercise a natural
sensitivity that largely avoids mistakes, and others, just as willing and eager to try, not? Why, simply, can some remember, and
others not? Nobody knows anything about these things, where they come from, what they depend on, how they 'work', even in
general terms, let alone anything about the causes of individual differences. Of course one should insist on good basic principles
of language learning - though no doubt the linguists and I will disagree as to what those are - but it is putting one's head in the
sand to ignore these fundamental realities that any number of new-fangled ideas will not have the slightest effect on.
I suspect, in fact, that at least indirectly it is the impossibility of doing anything about this deep-seated obstacle to foreign
language learning that is one of the reasons for the spawning of so many and various language teaching techniques, methods,
wheezes. There is certainly no lack of invention. They have gone as far as teaching grammar with coloured rods, and further.
(The 'language laboratory' is probably the most famous technique only because it is based on triumphant technology.) They
develop strategies. The word symbolizes for me the jargon-ridden and fallacious

Page 70
approach that appears to try to get students to learn almost without noticing it. Students are to be tricked into learning, in spite
of themselves. Above everything, at all costs, boredom is to be avoided. Learning resistance must be overcome. Everything is to
be made easy, painless, without tears. The student is to be manipulated into success by clever modern specialists or inspired
creators of new methods.
Unfortunately these 'strategies' often 'work' - particularly at beginners' and elementary stages. They 'work' in the sense that many
of the students learn a little, very slowly, and are not bored, may even be entertained. But in fact, coloured rods and all, they fail
totally to do the only really important things. They fail to give students curiosity about a language itself, for its own sake, they
fail to interest students in a language. Above all, they fail to show students how to find out about a language for themselves.
Essentially the language is avoided. In the end there is nothing to show for either all the ingenious new courses or all the
systems of celebrated educational thinkers.
Language 'course' books suffer from similar defects. Undoubtedly the best way to learn any language is unconsciously. But a
book on a language must rationally presume to work through conscious study; yet most modern course books proceed as if the
student is supposed to learn unconsciously from them. In fact the whole notion of 'course' books is misconceived. Do their
authors and users (students or teachers) really think that when students have gone through or 'done' a course book they have
learned the content, mastered the language presented therein? If the book has any content that can be learned, most students will
have forgotten by the fifth week - or earlier - most of what was in the first week of the course, however hard they worked at it.
If the book has no 'learnable' content students will be in even worse case at the end of it. An illusion of progress is often created
by the fact that students can do the attached exercises well; they do this because they are constantly alerted by the 'course' to the
precise nature of the problem they are being faced with at any given moment. In 'real' life, when they are off their guard,
unprompted, they will more often than not blunder and 'revert to form' as dictated by their own language. The only way a book
can be used properly is if students think about the problems of the language they are studying, constantly wanting to know and
find out how this, that and the other bit of the language works, above all constantly reminding themselves what the problems
are. A vocabulary large enough to be widely practical cannot be learnt from a course book or any sort of language book - only
from much listening and reading. Vocabulary aside, though, the problem for foreign

Page 71
language learners is not that of remembering the answer to the problem; the problem is remembering the problem. Language
teachers who believe that 'course' books work are suffering from an illusion. I repeat: studying a language through a book needs
thinking about the language, and so the book must be such that students can find the answers to questions they ask themselves.
Any decent language book must be, first and last, a reference book that students return to over and over again. The uselessness
of so many modern course books is symbolized by their lack of an index. Language book authors who do not provide indexes
are displaying horrifying complacency and lack of understanding of the real world of language learning. Course books are based
on the premise that the books can do it for the students. They can't. There is only one way. The students have to do it for
themselves.
If you think I am talking nonsense, or at least exaggerating intemperately, one must ask the question why at least one new
technique has not swept the field, so that we can all sit back with a sigh of relief and say "at last, what a pity somebody didn't
think of that before." The failure is confirmed by the fact that there is no let up in the stream of new devices for foreign
language learning and by the fact that 'grammar', for instance, is 'out' one decade and back 'in' again the next. There is the
absurd paradox that the more new 'strategies' are produced and fail to achieve what is needed, the more new 'strategies' appear
to be produced. One should not, however, be surprised at such a paradox. It is typical of humans' enormous ostrich capacity for
trying a way, failing, and coming back straight away again with twice as much of the same thing.
There has been considerable song and dance in foreign-language teaching circles about 'functional', 'notional', 'communicative',
'structural' etc. approaches. A very brief look at each of them shows that all that the first three of these mean is that people ought
to learn language that is useful to them. It should help them to say or write what they want to say or write, ask for a cup of
coffee, apply for a job, or explain why they are against nuclear energy, if those are the things they want to do. But what on earth
are language teachers about if they are not trying to do this already?
The functional, notional etc. 'movements' are typical parts of the new-systems manufacturing industry. All any of them do is
categorize and divide up language in yet another way. They tell us nothing new about the nature of language, nothing new about
what to teach, nothing new about how to teach it.
These new ideologies are also examples of that throwing-the-baby-out-with-the-bath-water process so dear to the human heart.
'Functional' is

Page 72
presented as an alternative, even a corrective, to 'structural'; 'student-orientated' is set up against 'language-orientated'; at the very
least it is suggested that the emphasis should be shifted.
It is quite obvious, as I have said, that when one learns a foreign language one should learn what is going to be useful. It is
equally obvious that one must learn 'grammar', since virtually everything said by native speakers of the language is 'grammar',
and much of what one says will be unintelligible if one's own speech is not 'grammar' too.
"The postman's daughter is less heavy than the chemist's cupboard" is rightly condemned if this is the sort of thing taught, and
nothing else. But it is a perfectly respectable and useful sentence, as it contains grammatical principles which can be applied in
other, new sentences. The mistake is not seeing that the practical and the grammatical are inseparable parts of the same thing:
language as it is in fact used. There is no conflict, and no choice to be made.
Unfortunately though, there are university departments, institutions of applied linguistics, and other institutions where the staff
have to justify their jobs and give themselves faith in their own value. So they busily invent pretentious new categories of
language, pretentious new names, pretentious new teaching philosophies. All it amounts to is either the affirmation in appalling
jargon of what any sensible teacher worth their salt will be doing anyway, or dogmas and 'isms that confuse teachers and make
them lose their sense of balance and reality.
What is so sad is that so many of the humbler folk in the language teaching world do not have the strength or self-confidence to
resist these empty ideologies. They feel they have to attach themselves to one or other of these profound insights handed down
to them by their betters; or - equally humbly, in truth - they say respectfully that each movement has its own valuable
contribution to make, and we must use the best of each. They should consider instead that it is the realities of everyday teaching,
and everyday inquisitive discusssion of those realities with their colleagues, that will serve them best and most honestly.
Ordinary teachers and ordinary language learners should have the courage to see that it is their experience that is the proper
source of truth and insight, not the contrivances of the theoreticians. The duty of good teachers is to think and talk daily - and
pleasurably and at leisure - about their work, not to be up in the latest fashions.

Page 73
This raises the question of professional qualifications for teachers of English as a foreign language, particularly to adults. There
certainly used to be (and still are) shady organizations, and some not so much shady as 'casual' organizations, that tried to make
fast money for their owners without concern for the quality of the teaching they provided. But there were also a few truly
conscientious language schools. Various sorts of people became worried about the state of affairs, and in particular the owners
or managements of schools who did not want to be lumped together with the more obviously fraudulent outfits and wanted to
present a good image to the world. So a demand started - it was in the spirit of the times, too - for courses and certificates that
would ensure that language teachers were 'properly' trained. Not surprisingly academic linguists were keen on the idea and
became directly involved. Now, without paper professional qualifications it is very, very difficult to get a job at a 'respectable'
language teaching institution, and virtually impossible to get one of the many English teaching jobs at institutions run by the
state or other authorities abroad - for these a university qualification in linguistics is more often than not required.
What makes the managements of language schools think that they now provide good and effective teaching? The simple fact
that all their teachers have gone through professional training courses and can put the 'appropriate' letters after their names? That
is truly making things too easy for themselves. Yet unfortunately most people are taken in by it. In fact it is quite possible that
teaching standards have gone down rather than up at many institutions, since having a 'qualified' staff can lull managements into
self-satisfaction, complacency, and a false sense of high standards. 'Qualifications' can do the same to the individual teacher.
With them safely achieved there may no longer be the urge towards constant curiosity and learning more, whatever one's age
and experience, that is essential for a good teacher.
There is only one way language schools can make sure their teachers are good, and that is the same way it always was, before
'professionalism' took over: by being conscientious. It should be a conscientiousness of every day. There should be an
atmosphere of unauthoritarian help, guidance, encouragement and co-operation, an atmosphere that makes new teachers feel
they can always, safely and gladly, talk openly and freely to any of their colleagues, including the most senior, about any
difficulties they have. There should be the constant discussion among teachers that I have mentioned above, discussion of the
language, of the work, of the individual students. Such discussion is far better done informally, spontaneously, at any and all
times of the working day, than at formal meetings. 'Old hands'

Page 74
should give the new the feeling that they will very soon become old hands too, and show that they, the old hands, are as willing
to learn from the ideas, inspiration and enthusiasm of the newcomers as they are to advise and help them. Whenever possible
they should try to inculcate the principles of good teaching by example rather than instruction.
All this applies particularly to the powers that be. Those worth respect do not spend most of their time behind their desks in their
offices, on occasion receiving or summoning people in audience, or going off to conferences. They teach. They share the daily
experience of all their colleagues. They know there is nothing more important than talking about the job over a cup of coffee to
one or more colleagues, new or old, on an equal footing and with everybody at their ease, being quick with intelligent sympathy
for both the successes and the troubles of their companions in a shared endeavour. Perhaps as important as anything is that those
with power should not have the attitude that new teachers are on trial and will be thrown out if they fail the test. That is just as
bad for good teaching as assuming any teachers with 'professional' qualifications are doing well, and morally less defensible. It
should be made clear to newcomers that if at first they do not make the grade, everybody, whether it be principal, director of
studies or any other colleague, is there to support them and help them do so. (I can assure you that if that is the spirit prevailing,
they practically all will.)1
The timetable should be undemanding enough to make these exchanges of ideas and experience something teachers can look
forward to, not grudge the time for. Language teaching establishments that will not arrange things thus are making too much
money at the expense of their teachers and at the expense of their students. That is one of the easiest and most common ways in
which they fail in conscientiousness, though that is rather a mild way of putting their ill-practice.
Whatever the merits of some aspects of the courses that would-be language teachers go through, instruction in modern linguistic
lore is not one of them. It is obviously essential for a teacher to have a practical knowledge of how the language works, and I
naturally mean a knowledge that the teacher can articulate in a way the ordinary native speaker cannot. But I believe this
knowledge is best obtained 'on the job', through a true apprenticeship in the sort of conditions I have described above. A good
language teacher needs confidence in the subject. But there are other things that come first for a good teacher, and partly
because they lead best
1 But there have been shocking cases where good teachers have been got rid of by a linguist-principal of professorial rank
on the grounds that they lacked sufficient professional qualifications.

Page 75
and most securely to that knowledge and confidence. Good teachers enjoy their students, and the students enjoy the teachers'
attitudes and personalities; students sense good teachers' dedication, enthusiasm, and genuine interest in the language, and that
they know about it for themselves; they make the students interested in the language too, and make them realize that they, the
students, have to do the work, and teach themselves; and the students enjoy this prospect and feel the teachers have helped them
to find the means to do it.
Most people do not do training courses in order to become like that, or to fulfil their interest in the English language. They do
them to get jobs.
If they do not have the qualities I recommend, qualifications will not give them to them. Only experience of the kind I have
described will do that.
For almost twenty years I advised and guided teachers of English as a foreign language - and in turn learned much from them. I
knew the work of my colleagues well. Without exception all the best teachers I have known, the really outstanding ones, had no
professional qualifications. Some of those who did have such qualifications were among the weakest teachers.
I believe educational authorities in many countries are to a large extent being conned or are deceiving themselves, or both, when
they insist on qualifications in linguistics in applicants for English-teaching posts at their institutions. Some countries pay a
great deal of money for their (I suspect) largely unquestioned assumptions.
Unfortunately things are unlikely to change. The university linguistic departments have a vested interest in making the customers
continue to believe that it is essential to get 'properly qualified' teachers, and the customers will continue to believe it because,
after all, the experts in the matter tell them so. They are unlikely to think seriously whether the substitution of 'genuine
specialists' for 'bungling amateurs' has really brought about the improvement in their citizens' English that it was supposed to. I
believe they would do much better if they chose among applicants by consulting their former students and possibly colleagues -
not their employers or the list of their diplomas.
One of the main purposes of studying languages is presumably to produce grammars. But who for and to do what with? In 1985
A comprehensive grammar of the English language was published. Its authors are professors Randolph Quirk, Sidney
Greenbaum, Geoffrey Leech, Jan Svartvik, and there are 1779 pages of it. I want to look at this book in some detail, because it
raises some fundamental practical questions.

Page 76
Quirk has said (Moorehead, 1985), "My book is of course a book of reference predominantly, but I am hoping that people who
care about the sensitivities of their expressions - writers, politicians, academics - will go to this book and take greater care how
they speak and write."
There can be no practical purpose in native English-speakers referring to it. Most of the book is devoted to aspects of English
which native speakers never get wrong. If there are parts of English grammar that some native speakers do get 'wrong', then
first, such speakers are very unlikely to go to such a book or even know of its existence, and second, even if they did they would
probably not understand it. As for the sensitivities of expression that Quirk talks of, I am not sure what he means. It is difficult
to find them in the book. And even if discoverable, I very much doubt whether any of the people he mentions would be
sufficiently humble to go to this book to sharpen their linguistic 'sensitivities'. In any case, his hope is surely contradicted in
principle by the authors' own emphasis that the grammar they present has been tested for 'acceptability' over large areas of the
English-speaking world. It surely does not make sense to go round asking people what they consider grammatical, write a book
on the strength of it, and then tell effectively the same people that they must read it in order to find out how to use their
language. So far as native speakers are concerned, all that seems left is the old academic pounding round in a circle that is
constituted by professors forcing their students to learn what they - the professors - have written.
The only other people this grammar might be thought useful for are those studying English as a foreign language, or those
teaching such students. I will not say that the book is devoid of all usefulness in this respect, but its faults and deficiencies
outweigh the usefulness.
Perhaps the least of the faults - but they are still very serious - are the things that have been left out of this heavy volume, and
the several rules or statements that are simply wrong. The things left out suggest that either Quirk and his colleagues have had
too little personal experience of teaching English as a foreign language, or they have forgotten that experience. Examples of
broad and important areas of grammar where they have not attended properly to the practical problems that trouble students of
English are the "-ing" form and the articles. They have also failed to deal with important points of detail, such as when one can
use "one" to replace a noun:
22 I was able to choose between studying a French dialect and a Spanish one.
but not

Page 77
23* Of the Latin languages I prefer the Spanish one.
They also make false assertions on points of detail, such as that "few", as opposed to "a few", has the sense of "not many".1
Everybody, of course, should be granted a few mistakes without suffering excessively severe strictures, even if one would
expect fewer from a team of four professors working for so long with numerous assistants. However, it is a tiny detail that
reveals something even more important, shows how wrong and impractical their whole approach is. On page 252 there is a note
[b]:
"It can be argued that some nouns, like weather, are neither count (*a weather) nor noncount (*a lot of weather), but
these nouns share features belonging to both classes. Noncount noun features include the premodified structures a lot of
good weather, some bad weather, what lovely weather. On the other hand, count noun features include the plural go out in
all weathers, in the worst of weathers."
This is typical of the mistake of applying abstract grammatical analyses instead of thinking naturally about meaning. They apply
an inflexible test -"a lot of" - and fall straight into the trap. There is no grammatical quirk in the word "weather". It is simply the
particular meaning of "weather" that makes "a lot of weather" somewhat uncommon. One does not often, in life, mean that. (But
it is by no means impossible:
24 A lot of weather is determined purely by temperature.)
Their, and most grammarians' attitude is strange. On page 288 Quirk et al. acknowledge that grammatical categories have
unclear boundaries. Yet they immediately go on:
1 One cannot, for instance, paraphrase
(a) He died after not many months.
or
(b) In not many days it will be New Year.
as
(c)*He died after few months.
and
(d)*In few days it will be New Year.
Notice, too, the difference between
(e) After not many (a few) games he was tired.
(f) After not many (few) games was he tired. In fact an essential distinction between "a few" and "few" is that "a few" has
the sense of "a small quantity, collection of', whereas "few" refers - negatively, indeed, as the inversion in (f) shows - to each
separate individual of the things it is attached to.

Page 78
''...we might wonder whether King's in King's College is to be classed as a genitive common noun which is part of a
composite common name."
They talk as if there was some absolute truth. We may wonder, but God knows the classes - if only we could find out too. If
they do not believe that God knows, they are effectively admitting that they are engaging in a ritual of human-made labelling for
its own sake, analysis for the fun of it, without meaning in reality.
The more they classify, the more categories they can establish, the happier they seem to be. Consider Quirk et al.'s treatment of
"the". Even before they have separated "the" from "a" they straight away divide the articles into "specific reference" and
"generic reference". By that they mean the difference between, for instance,
25 There are two giraffes outside the door.
and
26 Giraffes are strange looking animals.
where in 25 one means two 'specific' individual giraffes, but in 26 giraffes in general.
They do make an attempt to explain the meaning of "the" itself (p.265). They write:
"The definite article the is used to mark the phrase it introduces as definite, ie as 'referring to something which can be
identified uniquely in the contextual or general knowledge shared by speaker and hearer'."
This is a mess, for two reasons. It is written in the sort of imitation lawyer's language that grammarians use in case other
grammarians pounce on them for inconsistency or inadequacy or unscientificness. The result is that most of the people the book
might have been useful for won't understand a thing. Then, if any do, they will be misled by "...knowledge shared...". I guess
fewer than half the people who have trouble with their cars and are told
27 It's the constant velocity joint.
share any knowledge of the matter with the mechanic who tells them. But they will not question the mechanic's "the", because
they know which constant velocity joint it is - it's the one on their car. And in fact a vast

Page 79
amount of the "the" language that adults use to children refers to things that are not familiar to the children.
The confusion Quirk et al. have got themselves into is quickly confirmed, by a note on the next page (266):
"In practice, since a speaker cannot always be sure of the hearer's state of knowledge, use of the involves a certain amount
of guesswork. In fact, in some cases the assumption of shared knowledge is a palpable fiction. Notices such as Mind the
step and Beware the dog, for example, generally do not assume that the reader was previously aware of the hazards in
question."
"The" has nothing to do with the sort of "shared knowledge" Quirk et al. are worrying about, and so there is no problem
whatever with "Mind the step" or "Beware the dog". Just as ''the" in 27 tells you that "constant velocity joint" is the particular
one on your car, so "Mind the step" tells us of the particular step right there in front of us, and "Beware the dog" tells us of the
particular dog lurking around those parts. Such notices have to say "the" to make sense. Some people might get very alarmed if
they found notices saying "Mind a step" or "Beware a dog" and watch out furtively for steps and dogs for the rest of their lives.
But by this stage Quirk and his colleagues have started on something even worse, though it is an extension of the same
fundamental fault. Under one heading alone, "Specific reference...Uses of the definite article" (which is only the beginning of
their treatment of "the"), they establish no less than eight categories. (The examples are mine.)
(a) Immediate situation - e.g. Would you like to see the garden?
(b) Larger situation (general knowledge) - e.g. The world Cup was held in Mexico again.
(c) Anaphoric reference: direct - e.g. We've a dog and a cat. The cat's called Tonk.
(d) Anaphoric reference: indirect - e.g. This new cassette's no good. I'll take it back to the shop.
(e) Cataphoric reference - e.g. The history of Japan is almost unknown in the West. The wine that they produced last year is
considered very good.
(f) Sporadic reference - e.g. We're not on the telephone.
(g) The 'logical' use of the - e.g. Rico is the eldest son.

Page 80
(h) The use of the with reference to body parts - e.g. He was hit in the chest.
All this categorization is pernicious in two ways, and the second evil arises out of the first.
First is that the categorizing gives a wholly wrong impression of how language works. The truth is that "the" (when used in the
common way I am discussing here, as the 'definite article', though linguists waste energy debating that term too) has just one
simple basic meaning. All the other meanings of complete phrases or sentences arise merely as the result of that one simple
basic meaning in combination with other meanings. Exactly the same applies to, say, "a", or plurals, or words for uncountable
things without "the", and so on. (And I mean "things'', not "nouns". This is a crucial point. You can perhaps see what I have in
mind if you think of the example of "weather" again - p.77.) There is no such thing as an 'anaphoric' "the" as opposed to a
'situational reference' "the" as opposed to a 'cataphoric' "the" etc. They are all exactly the same "the".
It is not possible to describe any word accurately with other words. The only perfectly true description of the meaning of "the"
is "the". But one can give a pretty good and useful idea of what it means if one says it has the sense of "that particular limited
(thing), you know which I mean". Actually, the second part of that description is almost certainly unnecessary - the "you know
which I mean" idea must surely always follow inevitably from the first basic idea.
So in
28 We've a dog and a cat. The cat's called Tonk.
there is no "the" of 'anaphoric reference' in the second sentence. There's just plain "the". Of course the second sentence is about
the cat in the first sentence, but that does not excuse saying that the "the" is anaphoric. The second sentence is about the cat in
the first sentence because "the" - old, familiar, constant "the" - has been used together with "cat". Quirk et al. want us to believe
virtually the reverse: that a special sort of "the" (anaphoric "the") is being used because the second sentence is about the cat in
the first sentence.
The fatuousness of dividing "the" up into categories can be seen if one imagines a grammarian analysing "does" who says that
when a speaker uses "he does", "does" has male reference; but it can also have female reference, as in "she does", or "the hen
does", or "his wife does". But naturally

Page 81
the male/female distinction is in no way the foundation or starting point of the grammar of "does", and it is sheer befogging
analysing to present language on this basis. Meaning is the basis, and the result.
The categorizing approach gives the wrong impression of language in general; and it leads to wrong practical conclusions on
individual points of grammar. Here is a single example of the latter - under (c) Anaphoric reference: direct, Quirk et al. write:
"the first reference to an object will ordinarily be indefinite, but once the object has been introduced into the discourse in
this way, it can be treated as 'contextually known', and can thenceforward be referred to by means of the definite article."
(p.267)
Readers who speak a language without articles can draw quite the wrong conclusion from this, for it means that the following
conversation would be normal.
29a Bill: What are those?
29b Ted: A melon and a pineapple, of course.
29c Bill: It's the very small melon.
or
29d Bill: Yes, I can see it's the melon, but what's it doing here?
Now, I dare say the classifying grammarians will find a way of wriggling out of this. That will be quite beside the point,
because the only important issue is: What is the practical use of their grammars? Mistakes like those in 29c and 29d are
inevitable with their methods. But give students a simple explanation based on the meaning of the word itself, and no more, and
everything falls naturally into place. It will become clear, for instance, that one cannot use "the" for "melon" in 29c or 29d
because there one does not mean "that particular limited thing"; one means instead "one of the several (or many) that exist."
Certainly one will sometimes have to point out to a foreign student what a particular sentence or phrase means, and so one will
be involved in saying things like "Look, clearly we are talking about all tigers in general here, not particular tigers". But the
meaning of the words are all one needs, not some artificial system of categories which will only be another barrier to
understanding, insight and grasp.
For the analysing and categorizing system not only gives students false information; it imposes a completely unnecessary burden
on them.

Page 82
Naturally, describing words' meanings is not in itself enough. One has to give examples. But the larger part of Quirk et al.'s
explanations in their section on the articles, for instance, is concerned purely with an elaboration of their categories or trying to
get out of the tangles that their method must lead to. The whole of their section on Cataphoric reference (pp.268-69) is typical,
but I will quote only two parts of it in illustration:
"Sometimes the definite noun phrase can be contrasted with an equivalent indefinite phrase. In such instances, the definite
article is not in fact genuinely cataphoric but entails some degree of anaphoric reference."
"Note It is not necessary to postulate that the expression [5] presupposes some unspoken preamble such as [5a] but, rather,
such as [5b]:
the mud on your coat [5]
There's some mud on your coat. [5a]
You know there's mud on your coat. [5b]"
If students seriously get down to grasping and learning all this - and remember that it goes on and on for the greater part of
1779 pages - their heads will soon start to spin.
Again, under (d) Anaphoric reference: indirect, the authors are exercised by the problem of how much indirect anaphora is
combined with cataphoric reference, and so whether the use of "the" is explained by ellipsis. And so on. It starts all over again
when Quirk et al. get to "The articles in generic reference" (p.281). On page 283 there is:
"The Romans is thus a generalization (like the mathematicians, the teenagers, the birds, etc in similar use), whereas the
Roman implies a generic statement. Nevertheless, it will be convenient to apply the term 'generic' to both the singular and
plural uses."
Convenient for whom and why? Surely convenient only for analysers trying to fit everything into their classification scheme.
The authors of this tome are only the tip of the iceberg. Year after year after year grammarians publish articles in the linguistic
journals earnestly debating how generic 'generic' "a" and "the" really are. Naturally the debate is only a minute part of the whole,
of the constant analytical linguistic debate among academics.

Page 83
Yet, in the more than 30 pages of their section on the articles, with all their classification Quirk and his colleagues have no room
for two of the most important practical problems connected with "the" that face students of English as a foreign language: nouns
with modifying adjectives, and nouns modified by "of" phrases.
Without "the":
heavy industry
central Europe
tropical birds
But:
the electrical industry
the Upper Nile
the French impressionists
And without "the":
history of great interest
literature in the nineteenth century
But:
the history of music
the literature of the nineteenth century
(See Cook, Gethin, Mitchell, A new way to proficiency in English, 1967/1980, pp.79-81.)
All that Quirk et al. have to say on the two problems is on page 286 under the heading "The articles in generic reference":
"Normally the zero article also occurs when the...noun is premodified: she's studying European history."
(I will not dwell on the nonsense of the jargon; only explain that by "the zero article...occurs" they mean that the article does not
occur.)
"But when the same noun is postmodified, especially by an of- phrase, the definite article normally precedes it: She is
studying the history of Europe."
There is no explanation of why and when some nouns modified by adjectives take the article and others don't; and no
explanation of why and when some "of" (and other prepositional) phrases make "the" necessary and others don't. Yet these,
particularly the first, are 'mysteries' that cause difficulties for millions of students of English. It is hard to see what is the
purpose of a grammar that makes no attempt to clear them up. It is true that in an earlier version of their book, A university
grammar of English

Page 84
(1973/1979, pp.71-72), Quirk and Greenbaum had very slightly more to say on the subject:
"...English tends to make a liberal interpretation of the concept 'generic' in such cases, so that the zero article is used also
where the reference of the noun head is restricted by premodification. ...Canadian paper..."
What is the good of making these beds of Procrustes for themselves? The uselessness of their sort of classification and analysis
is shown up clearly by the vagueness of the weak attempts to get round difficulties. In this case, for example, they have got
themselves into a tangle only by insisting on the 'generic' classifcation in the first place. Quirk et al. talk of "the generic the" as
if the word "the" itself had different functions, and itself gave generic (or other) meaning to nouns - instead of it being the
meaning of each specific phrase or sentence in its context that gives the meaning - i.e. the meaning is the meaning! It isn't "the"
that is generic - it is the phrase that has generic meaning, or not, as the case may be. In any case, as we have seen, there are
different sorts of 'generic', and disagreement as to what has true generic meaning or not. This surely confirms that the linguists
have got the wrong end of the stick.
What students need to have shown and explained to them are the most basic meanings. Here are three of them:
a horse
the horse
horses
They can then be illustrated in varying contexts. But Quirk et al. and many other grammarians insist on lumping them together
under the heading 'generic' and then find they have to unravel them again. Quirk and his fellow authors fail to do this properly
and so only achieve confusion. All I can find in A Comprehensive Grammar... is:
p.281".....
"The bull terrier makes an excellent watchdog [1a]
A bull terrier makes an excellent watchdog [1b]
Bull terriers make excellent watchdogs [1c]
....It should not, however, be assumed that the three options
[1a], [1b], and [1c] are in free variation. One difference
between

Page 85
them is that, whereas the [1a] keeps its generic functions in
nonsubject positions in the sentence, a/an [1b], and to a lesser
extent zero [1c], tend to lose their generic function in these
positions:
the medieval mystery
play [3a]
Nora has been a medieval mystery
studying play [3b]
medieval mystery plays [3c]
Of these, only [3a] refers to mystery plays as a genre; [3b] refers to
only one play; and [3c] is most likely to refer only to a subset of
them.
...
The generic use of a/an picks out ANY REPRESENTATIVE
MEMBER OF THE CLASS. Thus any can be substituted for a/an
in examples like:
The best way to learn a language is to live among its speakers.
Generic a/an is therefore restricted in that it cannot be used in
attributing properties which belong to the class or species as a
whole. Thus:
p.282 The tiger is
becoming almost extinct.
Tigers are
BUT
NOT:
*A tiger is becoming almost extinct.
....
The generic use of zero article with both plural nouns and noncount
nouns identifies the class considered as an UNDIFFERENTIATED
WHOLE (cf 5.39 f):
Cigarettes are bad for your health.
...
The is rather limited in its generic function. With singular heads, it
is often formal or literary in tone, indicating THE CLASS AS
REPRESENTED BY ITS TYPICAL SPECIMEN:
A great deal of illness originates in the mind.
No one knows precisely when the wheel was invented.
My colleague has written a book on the definite article in Spanish.
Marianne plays the harp very well."
These explanations do not, I think, explain differences - and
'impossibilities' - like the following:
30 Bill trains a horse.

Page 86
31 Bill trains the horse.
32 Bill trains horses.
33 A horse gets its training at places like Bill's.
34 The horse gets its training at places like Bill's.
35 Horses get their training at places like Bill's.
30 certainly fits what Quirk et al. say about their example [3b], though Bill appears to have a rather limited life. But already
here one notes that they merely say "tend to lose their generic function" (my italics). That is not satisfactory for any student
hoping for clear guidance. 31 is impossible unless either the speaker is talking about one particular horse, or Bill is God. Yet
there is nothing in the exposition above to tell one that. Is it not the fullest generic sense that is needed and is not the "the horse"
form the very type of the generic? Students must be forgiven for believing both these things if they study Quirk et al., and they
will get decisive confirmation of their conclusion when they note [3a] -
Nora has been studying the medieval mystery play.
- and what is said about it. For 'syntactically' [3a] is effectively exactly the same as 31. The only difference that might be
significant is that 31 uses the 'simple' form rather than the "-ing" form of the verb. But any students who have read Quirk and
Greenbaum's earlier A university grammar of English (1973/79, p.68) will have their conviction strengthened further by
"There is considerable (though by no means complete) interdependence between the dynamic/stative [i.e. "-ing"/simple]
dichotomy in the verb phrase and the specific/generic dichotomy in the noun phrase..."1
So what is to stop students thinking that 31 is just what they need? The answer to that question is once more: don't think about
irrelevant classifications, but concentrate on the particular 'whole' meanings produced by the particular meanings of the parts -
"studying" is different from "training", and "the medieval mystery play" is different from "the horse", and will affect the
outcome accordingly.
1 Notice not only the familiar vagueness, but also the old mistake of turning what is simply more common in life into a
function of language. It is quite irrelevant that people do not so often want to express 'generic' with 'dynamic' as 'generic'
with 'stative', and it is utterly misleading to draw attention to non-existent linguistic associations. If the idea one wants to
express happens to be "Peanuts (in general) are doing something", that is what one has to say, irrespective of how rare
that situation may be in the world.

Page 87
Quirk et al. might of course be ingenious enough to show that their description of the "the horse" form as "indicating the class as
represented by its typical specimen" is precisely the right one for preventing the use of 31. But they can surely not seriously
pretend that anybody who did not know how to use the articles in English already could work out from such an abstract formula
that "Bill trains the horse" is not what they mean.
32 is the 'right' one. How are students to know this from Quirk et al.? Is "horses" specific or generic here? And which do we
want anyway? If generic, how generic? Presumably not anything generic in the sense of "Horses are beautiful". Perhaps their
reference to a subset (but how generic is a subset?) in non-subject positions is just what we should bear in mind here. But then,
why can't we say "Bill trains some horses", for isn't "some horses" a subset? And do we identify something as a subset first and
then choose the appropriate words accordingly? And if that is how we work, what happens to ideas like ''I like horses", where
"horses" is certainly not a subset. All this is a perfect example of how totally misguided the classificatory approach is, not only
to the theoretical problems of grammarians, but also, more seriously, to the practical problems of students.
From Quirk et al.'s examples [1a], [1b], [1c] one would think that 33, 34 and 35 are all perfectly acceptable. The various "horse"
phrases are all subjects. But again it is the particular meanings that prevent this, not any of the distinctions noted in A
comprehensive grammar... 34 is impossible in the same way as 31. If 33 and 35 are to be taken in a general or 'world' context,
they are patently untrue; in the limited context of horse racing both are a possibility, although "The horses" seems to me more
natural than merely "Horses".
Meanwhile bewildered students can only become more frustrated than ever by the failure to tell them the things they really need
to know.
Quirk and company's treatment of the articles is not at all an isolated case. For instance, when they come to the distinction
between "some" and "any" (and their compounds) they immediately start on their usual tricks. "Some" is declared to be
'assertive', "any" 'non-assertive'. This is exactly the same cart-before-the-horse approach as they applied to the articles. Once
more they fail to see that it is the basic, original meanings of "some" and "any" that produce, sometimes, sentences that can be
analysed or classified in this way. Once again, 'assertive' and 'non-assertive' and all the other analyses they make in connection
with "some" and "any" are simply some of the possible results, in meaning, of the use of and combination of various basic
meanings. "Some" and "any" have certain meanings; combine them

Page 88
with other meanings in various complete contexts and one gets various complete meanings as a result. (The result is very often,
for instance, that one cannot use "some" with a negative, simply because the basic meaning of "some" will not fit the particular
negative. Very often, however, "some" not only fits the negative, it is what one has to use with the negative to express one's
meaning - e.g. ''Somebody's not signed".) It is quite irrelevant and without interest to classify those resulting meanings. I have
already drawn a parallel with analysing "does" in male and female terms. Perhaps an analogy will help to make my argument
clear. Black paint does not mean paint that makes things invisible in the dark - it means black paint; but one can use it for,
among other things, making things invisible in the dark. Again Quirk et al. fail to put their finger on the essence.
In fact, one can probably best show the meaning of "some" and "any" with one's hands, and by waving one's arms about. If this
is impracticable, then one can perhaps say that "any" has the sense of "any at all, any you like, any whatever, any - it doesn't
matter what". "Some" is limited. "Some" is a certain amount, a certain number, a certain thing or person, although one may not
know how much, how many, or exactly who or what. You may consider these accounts of the words far too vague, but with the
help of good examples they work.
But Quirk et al., once they have imposed the artificial and rigid scheme of 'assertive' and 'non assertive' on "some" and "any",
and have associated those terms with the equally rigid classifications of 'positive' ('assertive' -"some"), 'negative', 'interrogative'
and 'conditional' ('non-assertive'-"any"), are forced to resort to all sorts of complicated contrivances to get out of all the
difficulties they have landed themselves in. Nearly all the detailed rules, convoluted analyses, conditions, qualifications, provisos
and reservations that they set up as a result are either false or unnecessary, or both. Most usage flows naturally from the basic
meanings, not from arbitrary rules prescribing a multitude of various usages. One example will suffice to sum up their approach
and its results:
"Note Whereas it is frequently impossible for a positive statement to contain nonassertive forms (*I have any ideas)1 it is
by no means unusual for assertive forms to occur in questions and negative clauses: Do(n't) you have some ideas? Our
use of the term 'nonassertive territory' does not exclude, and indeed anticipates, a more delicate stage of analysis (cf
10.61f) at which we acknowledge that assertive
1 This is not in fact an 'impossible' sentence.

Page 89
forms can give an assertive 'bias' to constructions which are predominantly nonassertive." (p.84)
They are pitching their descriptions at completely the wrong, far too superficial level, and not only getting it wrong - they are
making it much too daunting in the process.
These examples of the defects in Quirk et al.'s 'comprehensive grammar' are important, and those defects are serious, because
the authors are no doubt widely regarded as among the leading authorities on the subject, if not the world authorities. Such
authority needs to be challenged.
One of the few branches of linguistics that I believe may have practical use is phonetics. It is one of the very few that is
concerned with observable physical realities. Detailed knowledge of how sounds are produced by humans is, in principle at
least, useful or even essential for helping with various kinds of speech problem.
But let no one believe that a study of phonetics generally, or of the phonology of a particular language, is going to help anybody
pronounce a foreign language better. The ability or inability of a person to pronounce a foreign language well is founded deep in
the psychology of the individual. In the vast majority of people there seems almost to be a fixed limit to achievement as regards
accent that is built in, that is different in each individual, and that the individual never goes much or at all beyond after the first
few weeks of study of any new language. Among those who have studied phonetics a few are excellent pronouncers of foreign
languages and some are hopeless. Among those who have not studied phonetics a few are excellent pronouncers, and some are
hopeless. It so happens that none of the people I know personally whose accents are good enough for them to be taken for native
speakers have studied phonetics, and I do not think my experience is unusual. I once knew a professor of comparative phonetics
at a well-known European university; he was charming and modest but his pronunciation of English was a caricature of his
native accent.
It may be valuable to point out a few basic principles to students; in the case of English such points as the particularly important
part played by stress, and by the unstressed vowel sound as in the last syllable of "butter" or "station" or in the first and last
syllables of ''America". But students are quite misguided - and doomed to deadly tedium - if they believe that close study of
tongue positions or of intonation patterns is going to make any appreciable difference to their accents. Innocent teachers may
believe that they have

Page 90
achieved some success with the pronunciation of their students in class, unaware that the moment the classroom door is behind
them the students revert to normal. A good accent can in practice only be got by listening properly, and the ability usually either
comes very suddenly when the student opens her ears or loses her inhibitions, or both at the same time, - or not at all.
Linguists have also asserted that linguistics can be applied to translation. It is hard to see what possible use linguistics, of any
school, could be to human translators. Proper translation consists of recognizing the reality, quite free of any language, that the
original language refers to, and then converting that picture of reality into another language. The process at this second stage
should be exactly the same as the process when one puts anything of one's own thought or experience into one's own language.
Any proper translator knows that the real problem in translation work is not difficulty in understanding the original. If one
cannot understand the original language without much difficulty one has no business to be translating from that language at all.
What is often exhausting, however, is finding expression in the second language for the reality that one recognizes perfectly
well. The fact that this second language is usually one's own in no way gets rid of the difficulty. Translation becomes really
difficult, or impossible, when the original language is mere verbiage with little or no connection with reality, not even the reality
of fantasy. Linguistics can have nothing to do with these human translating processes of recognizing reality and converting it
back to language. The world is already full of good translations by humans which it would be impossible to improve by any
system of analysis.
The situation with regard to machine translation is ironical. Already some years ago it was discovered that Chomskyan
linguistics were useless as the basis of translation by computer. For once this might almost be something in favour of
Chomskyan linguistics. It certainly does not prove that Chomskyan linguistics are psychologically false as an account of human
language. What would have counted heavily against them would have been if they had turned out to be a sound basis for
machine translation, as computers do not work like human brains.1 If computers ever translate human languages really
successfully it will at the most only be by simulating what humans do. They cannot translate a language into the
1 See a reference to the proof of this in Chapter 12, pp. 247-48.

Page 91
sort of 'reality' or ideas that humans do, and then translate those back into another language. Analysis of language is of course
necessary, but it has to be analysis on a basis that suits computers, not one that has anything to do with actual human language
psychology. It is surely best done by computer scientists; if linguists contribute to the analysis and it proves successful, let them
not believe that they are making a contribution to the understanding of how natural languages work inside human beings.1
Linguists assert that their theories can be applied in several other fields, such as native language teaching, psycholinguistics,
sociolinguistics, literary studies. Every system of linguistics is an analytical system. It is quite obvious that any way of analysing
language can be used to analyse any particular sort of language, whether it is the language of schoolchildren, of particular social
groups or of novelists or poets. Any system of analysis can be used to analyse, no matter how fatuous and footling. And one
finds in practice that that is precisely what linguists nearly always mean when they talk about applying linguistics: analysing
language of some particular sort or in some particular context in accordance with their particular method.
There is no objective way in which one can ever measure the validity or otherwise of any mode of analysis - validity in the
sense of what is psychologically significant, important in individuals' experience of language. And one can only judge the
usefulness of a mode of analysis by experience of the actual results, that too, perforce, a subjective affair.
I believe that in fact however valid an analysis might be it could never help either teachers or students in practice. An analytical
system is an extra thing which has to be learnt and is interposed between enquirers and language, and so is nothing but an extra
barrier. It does not give psychological insight and it does not help learners.
It has been asserted that children use more complex language as they grow older. Some linguists want to use their analytical
system to codify the details of this astonishing fact. They think that then such 'knowledge' is useful for the teaching of English
(or, presumably, any other language) as a native language. Yet, the analysis-pottiness of the age notwithstanding, there seems to
be a lot of agreement that the standards of writing among school-leavers and university graduates are a good deal lower than
they were forty or fifty years ago.
1 For the unpleasant possibilities connected with a wedding of computers to language, see Chapter 12, pp. 248-49

Page 92
As for literary appreciation, once again any form of analysis has a good chance of welcome from those who like to do their
appreciating of literature by means of analysis. There's nothing new in that. But again, let them not think they are gaining any
new insight into the human psyche or the inspiration of the artist. That's not the way we are going to get more good books or
more genuine enjoyment of them.

Page 93
5
The Fantasy of Structure
In the sense of a single, unified system there is no such thing as structure in language.
Sadly, however, many academics and philosophers are unwilling to accept things as they seem to be. Some great hidden
principle has to be found beneath the appearance of things.
For language, the main myth that has been created is structure. The reality is simpler, but not so intellectually intoxicating, since
instead of being one profound unified system, it is a number of different rather humbler features.
As I have proposed in the earlier chapters, the basic feature of language is meaning. But the many pieces of a language
obviously have to be arranged in some order. So one aspect of language is what we can call word order. As I shall try to show
in the next chapter, however, word order certainly does not deserve the obsessive attention as a single fundamental and decisive
principle that it is given by many linguists.
Then there are other features that distinguish different languages.
For instance, some languages have types of meaning that others do not, e.g. the 'action-thing' idea that Japanese (-koto) and
English (-ing) have, and which many languages do not.
There are also combinations of meaning that one language will have, and another not. For example, many languages use
prepositions with infinitives (e.g. Spanish: "comenzó por hacer..") while English does not - it says "he began by making..", not
"...by to make.."
Again, a given combination of meanings may mean one thing in one language and something quite different in another. The
Italian equivalent of "He said he Would do it", for instance, is "Disse che lo avrebbe fatto", literally "He said he would have
done it."
But this simple identification of a number of different aspects of language does not satisfy the academics. Modern linguists think
themselves greatly in advance of traditional grammar. The irony is that in the end they are just as much victims of tradition
themselves, in fact more obsessed and enslaved by the formality of analysis than any before them. They scorn old-fashioned
parsing, but actually, in principle, are parsers gone mad. They are no revolutionaries. They are the most die-hard conservatives.
They take for granted the distinction between grammar and meaning, and few, if any, of them appear to question the assumption
that structure is a fundamental part of language. It is an asumption that they

Page 94
can surely have got nowhere except from their despised predecessors. I believe and hope that in a more enlightened future this
obsessive search for the perfect model of the 'structure' of language will be looked upon as something as bizarre and pointlessly
unreal as the centuries-long search for the philosopher's stone.
Chomsky and his followers have taken two sequences of words such as
1 colourless green ideas sleep furiously and
2 furiously sleep ideas green colourless
to demonstrate that English-speakers (like everybody else) have an unconscious awareness of the 'rules of grammar'. Both these
sentences will be new to most people, but, they say, although both are meaningless, we immediately recognize the first as
'grammatical' and the second as a random jumble. We see familiar structure in one, and not in the other.
If this is a fair account of the Chomskyan argument, then it says much that, in itself, is true (and obvious), but it suggests other
things that are serious mistakes. First there is an objection which is comparatively superficial and which no doubt others have
made before me. "furiously sleep ideas green colourless" is by no means an 'impossible' word order in English. It is the sort of
word order that one can often find in poetry - for instance, "Softly comes the dawn, pale, transparent". Looked at from this
angle, all one can say is that some word orders - "colourless green ideas sleep furiously", for instance - are more familiar than
others - such as "furiously sleep ideas green colourless". This seems to me to mean that at the very least there is no absolute line
between what is 'grammatical' and what is not.
But there are more fundamental objections to the whole structure built on the idea of 'structure'. Chomsky's case, in this instance
at least, hinges on the order of words; I shall take this up later. But first I want to look at what I think are broader questions.
If it is true that both sequences (1 and 2) are meaningless, then presumably meaning is to be left out of the argument altogether.
So if one says that both sequences are meaningless, but that we can recognize the first one as grammatical, then one is saying
that there exists an independent grammar, or 'structure', recognizable in its own right, quite irrespective of whether there is any
meaning or not. And unfortunately this seems to be precisely what Chomsky and others are trying to argue.

Page 95
But this is saying that there is structure because there is structure. Because if meaning is out of it, where else can the structure
come from?
This is the sort of basic mistake that tends to run right through the thinking of linguists. The fact that they are so convinced
structure is real and important shows them they must be right. Structure is fundamental because they say it is.
"colourless green ideas sleep furiously" is in fact not meaningless. Each of the words, even on its own, unconnected with other
words, conveys a meaning to people. Put together, the words convey a certain impression, albeit a strange one, and perhaps a
different one to different people. Linguists probably fail here to understand the difference between "meaningless" and
"nonsense". "colourless green ideas sleep furiously" is fantasy, nonsense, but full of meaning, in the same way as - though less
imaginatively and exotically -
3 He will come yesterday.
or a jingle we had at school,
4 One fine day in the middle of the night, two dead men got up to fight...
are nonsense but have meaning.
These sentences are nonsense because their meanings do not fit in the real world. In 4 the 'specific' meanings do not fit. In 3
there is a 'general' meaning - "will" - that does not fit the 'specific' meaning "yesterday". But this clash of a 'general' meaning
with a 'specific' meaning is also exactly what happens in, for instance,
5 He has come yesterday.
If the Chomskyans are consistent in their search for a 'complete', explicit grammar, then they must deal with these two sentences
- 3 and 5 -in exactly the same way. ("have done" is an action in the past, but it means an action not pin-pointed to a particular
time - such as "yesterday". To combine the two is universally, I think, regarded as a 'grammatical' or 'syntactical' mistake.) And
I imagine that they in fact do so. But they are surely arguing quite arbitrarily and artificially if they say that "colourless green
ideas" is a well-formed, grammatical phrase, but that "will come yesterday" is not. Why should the clash in meaning between
"colourless'' and "green" be acceptable and the clash in meaning between "will" and "yesterday" not be so? All one is doing if
one declares that "colourless green ideas sleep furiously" is grammatical is establishing that, although the

Page 96
'specific' meanings conflict, the 'general' meanings fit each other, and do not conflict with the 'specific' meanings. In other words,
one recognizes that these plural things ("ideas") are described ("colourless, green") and do something ("sleep'' - in the way
plurals normally do, not "sleeps") in a certain manner ("furiously").1
Here is the key factor. One cannot recognize these 'general' meanings if one does not understand, first, each 'specific' meaning,
and then the meaning of the whole. And 'understanding' means recognizing the meaning in reality, albeit sometimes a reality of
fantasy. Suppose one has a sequence
6 light fishes last
One doesn't recognize the meaning of "light" - it could be thing, quality, or the command to set fire to something. "fishes" could
be things or an action. And one can't identify "last" either, as it could be the opposite of "first", or a function, or something to do
with cobblers. One doesn't know if the sentence is 'grammatical' or not, or even if it is a sentence, because one doesn't know
what the meaning of the words is meant to be. And what that quite simply means is that one doesn't know the meaning if one
doesn't know the meaning.
But one can recognize the sequence as 'grammatical' if one decides that the words have certain 'general' meanings that fit each
other - but it must be the meanings that fit. Again, this is simply finding that there is meaning if there is meaning. There is no
grammar, or structure, which lives by itself, independent of meaning. Such an independent 'grammar' is only the linguists'
imagining.
Something that is probably fundamentally important, for practical purposes, is that most of the time people talk sense, not
nonsense. This means that their listeners can easily recognize the reality intended, and so quite easily recognize the 'general'
meanings, or 'grammar'. Here is a longer sequence where reality is not recognizable:
7 The prophets have prophesied the prophets the prophets the prophets have prophesied in prophesies prophesy the prophets
prophesy
1 I must emphasize again, though, - see Chapter 1 - that I am not trying to make any absolute distinction between
'general' and 'specific'.

Page 97
Once again, one cannot recognize that this is 'grammatical' because one doesn't recognize any meaning; while it is precisely the
meaning in "colourless green ideas sleep furiously" that one does recognize. But substitute other 'specific' meanings for the
words in 7, and one can get
8 The President has informed the various governments that the difficulties the energy bills have encountered in Congress require
that he stay at home.
In essential outline this sentence is 'grammatically' identical to 7, but is recognizable as 'grammatical' because of the realistic
meaning it contains, while 7 is not recognizable as 'grammatical' because of its lack of either sensible or even nonsensical
meaning. Here I note once more that all that 'grammar' is is 'general' meanings. If because of the confusion of 'specific' meanings
one cannot see the 'general' meanings - then one doesn't understand even the 'general' meanings! And the sequence is truly
meaningless. Put in a "to", some "that"s and a "who", and some meaning immediately begins to appear.
7a The prophets have prophesied to the prophets that the prophets who the prophets have prophesied in prophesies prophesy that
the prophets prophesy.
It is now nonsense - it wasn't even that before, only gibberish - because the extra words have given meaning to it, the
relationships have been made plain.
Perhaps I will myself be accused of arguing in a circle. In a sense I certainly am. But I think there is reality and sense in my
circle, for I am simply saying that meaning is meaning and without meaning there is no meaning. The whole reason of being of
language is meaning.
Some may also say that I have completely contradicted myself in my discussion of example 7; that the whole sequence rests on
the assumption that there is structure independent of meaning, as I could not make it up except on the basis of rules of structure;
that I took these abstract rules as applied in an 'actual' sentence (8) and applied them to my new one (7). But I have in fact done
nothing of the sort. I have only 'applied' meanings, meanings of a 'general' kind. In both the actual (8) and the invented sentence
(7) there are things and people and actions, linked by a logic that is not in any way abstract but entirely practical - actions are
performed by people or things, things happen in places, etc. etc. But as we have seen, in

Page 98
the invented sentence (7) 'general' meanings are not enough, and only produce an invisible theoretical sense, not an effective
one.
And to prevent any misunderstanding I think it is important to emphasize that throughout, whenever I talk of things that 'fit' and
those that do not, I am never talking of words, but precisely of things, that is, people, objects, feelings, actions and everything
else that goes to make up the real world - or the world of our imagination. In
9 out wooden milk forgets very also
the feeling, the 'reality' of out-ishness won't go with the feeling of woodenness, not even in a world of fantasy, probably not
even in a dream, nor will the idea of also-ishness go with the idea of very-ishness. The things, the feelings themselves won't fit.
But change "out" to "softly", and "also" to "certainly", and immediately there is a fantastic meaning, we can imagine the ideas:
10 softly wooden milk forgets very certainly
On page 66 of Modern linguistics Smith and Wilson (1979) write:
"Specifically we want to argue that syntax is prior to both semantics and phonology: prior in the sense that there are
phonological and semantic processes which depend for their statement on syntactic facts, but no syntactic processes which
depend on phonological or semantic facts."
And on page 134:
"...the difference between blúe-stocking and blúe stócking is due to their difference in syntactic structure; while both
consist of an adjective followed by a noun, in the former these unite to form a compound noun, in the latter an NP [noun
phrase], categories independently needed in the syntax. This last example is more significant than simple examples of
word stress, as the latter, being finite, could be handled entirely within the lexicon. The configurational differences
between blúe-stocking/blúe stócking or George has pláns to leave/George has plans to léave, however, can only be the
result of phonological rules which have access to syntactic information; the set

Page 99
of possible forms to which such rules can apply is infinite, and therefore in principle not treatable by lexical listing."
In their book The sound patterns of English (1968), in which among other things Chomsky and Morris Halle tried to establish
the stress rule for the compounds that Smith and Wilson refer to, there are phrases like "...kitchen towel-rack (in the sense of
'towel rack in the kitchen', not 'rack for kitchen towels'),...". These appear to make a reference to distinctions of meaning.
Chomsky and Halle also admit that there are exceptions to their rule that compounds are stressed on the first part only. Yet they
do not anywhere acknowledge any part played by meaning in determining stress. And while they confess they cannot explain the
exceptions to their Compound Rule it does not seem to occur to them that the exceptions might be because they have got it
wrong. They continue to insist that the problem is a syntactic one.
Once again they have things the wrong way round. One cannot put any piece of language into a 'syntactic' category until one
knows its meaning. When one knows the meaning, syntax is as irrelevant as it was before one knew the meaning. What is
a sleeping partner?
Is it a partner for sleeping with or a partner who is sleeping? When one knows which, and only then, one will know how to
stress the phrase.
The principle that almost certainly really decides the position of stress in 'double' (or 'triple' or more) expressions is simple, and
is a true principle rather than just an arbitrary technical rule. It is difficult to articulate precisely, because it involves the
description of meaning, and words are very bad at describing meaning, but the fact that millions of English-speakers apply the
principle shows that they understand it!
The principle is that both parts of a 'double' expression are stressed if the first part 'describes' the second part, or the second part
'belongs to' the first. Otherwise the first part alone is stressed, and a typical sort of 'double' where this happens is where the
second part is 'for' the first part. For instance, in "fúr cóat" and "réstaurant kitchen", "fur" describes "coat" and "restaurant''
defines the kitchen, whereas in "ráincoat" and "réstaurant owner" the coat is 'for' the rain and the owner owns the restaurant.
Although the principle is usually valid, it is certainly not applied invariably, and there are undoubtedly difficulties. But syntactic
distinctions offer no hope at all of unravelling the mystery of contrasts like "évening-

Page 100
dréss/évening-class/mórning-cláss" or "tówn plánning/defénce planning". Many exceptions are probably only apparently so.
False exceptions are almost bound to appear when it is so hard to express the principle perfectly.
I suggest the reason for the inconsistencies and variety in the way English-speakers stress some combinations is that people
simply follow personal, local, national, or even wider habits when it doesn't matter. Rather obviously it doesn't matter when it
doesn't matter. Humans using language tend to be pragmatic in this way; there is no point in bothering if it doesn't make any
difference. And so there will be changes, too. I have noticed recently, for example, how many people are changing to "wéekend"
instead of the "wéekénd" that I shall always insist on saying. All this is in great contrast to Chomskyans' rule-bound universe.
But where it does matter, I think people will always follow the stress principle based on meaning that I have tried to explain.
And it matters when otherwise there would be ambiguity, as with "sleeping partner", and where ignoring the principle would
produce absurdity. I don't think anyone would announce that although they had a running nose they were going to put on their
running shoes. It is clear too, surely, that the flexibility, or, if you like, the very vagueness, of the meaning hypothesis makes it
a better explanation of the facts, if not the only possible one.
In the English convention there are only two ways of doing it: strong-strong or strong-weak. All meanings have to be fitted into
one or other of these, so there are bound to be inconsistencies where there are several subtle variations of meaning to be dealt
with.
Add to this what is probably the key point. Different people can experience these combinations in different ways. One person
may feel the first part is 'describing', another may not, for example. The subjective meaning principle offers an explanation of
inconsistencies that can never be justified by the rigid syntax-based criterion.
When it comes to stress within a sentence, Smith and Wilson (1979) continue in the same way (p. 142ff). They insist once again
on an arbitrary abstract rule which ignores meaning, the Nuclear Stress Rule. It "assigns primary stress" in English "to the
rightmost constituent in a sentence". So we get, for instance
11 George has plans to leave.
12 Margaret left directions for George to follow.
But they recognize that the rule, put like that, is clearly wrong for sentences such as

Page 101
13 George has plans to leave.
14 Margaret left directions for George to follow.
(George is going to leave some papers, and Margaret provided instructions George could use.) The solution by Joan Bresnan
that they favour is a vivid example of the Chomskyan mentality. She apparently proposes that the rules of stress "refer to
information" that is only "available" at the deep structure level.1 This of course neatly and typically 'accounts for' the problem in
their descriptive code, but as usual does not really explain anything. When they run into yet further difficulties Smith and
Wilson try to get out of them by claiming a category of stress that is 'pragmatic', not 'linguistic'. This very suspect distinction is
surely an artificial device solely to get rid of inconvenient facts which would otherwise wreck their neat self-contained building
founded on syntax. It is one more example of the basic mistakes that linguists are forced into by the even more basic mistake of
their obsession with language as a separate autonomous system. Something does not fit so they have to declare that it cannot be
part of language proper.
Within sentences stress is, as all innocent people are aware, a means of expressing meaning through emphasis. So in 13 the
stress is put on "plans", because the most important thing is that George possesses plans that something is going to happen to. In
11 the most important thing is that George is going to leave, so the stress goes there. In 14 "directions" is the really important
thing. Notice how the practical meaning would not be affected at all if one left out "to follow" altogether. In 12, in contrast, one
could, for all practical purposes, say "Margaret told George to follow". Notice, too, how in the same sentence one could add
''her" at the end, and the stress would not move, which makes nonsense even on her own terms of Bresnan's attempt to get the
stressed word to the end of the sentence by appealing to 'deep structure'. The basis of stress really has nothing to do with
syntactic analysis. The basis is pure meaning according to the situation in 'real life'. Again:
15 George has a reason to go.
16 George has a reason to suggest.
1 See Chapter 8 for a discussion of 'deep structure'. See also above, Chapter 3, for an example of the fearful mistakes that
the 'syntactic' and 'deep-structure' view of language can lead to (wh-movement -p.44ff.) Here they say that 14, for
example, comes from the underlying:
s[Margaret leftNP[directionss[for George to follow directions]]].

Page 102
This pair of sentences has the same contrast in 'general' meanings as 11-12/13-14. But here "reason" is stressed in both
sentences because in most situations in life reasons are 'powerful' things and so are commonly emphasized. But in a different
context:
17 There's no reason to shout!
Quite simply one emphasizes the parts of a sentence that are particularly important, and one emphasizes most of all the parts that
are most important. Each language and each local variant has its own conventional rhythm; but this cannot override the effect of
meaning in a context. This does not mean that meanings intended through stress are always unambiguous, always understood
correctly by listeners. There are a limited number of words in a sentence that can be stressed, and there may be a larger number
of meanings - and contexts - that could be expressed.
All language is produced in a context, even if that context is very often 'from scratch', so to speak. Even in a 'from scratch'
utterance there is room for variation, which depends on how important the speaker considers the various things the listener does
not know. One might say:
18 There's a cat in the garden.
But equally one might say:
19 There's a penguin on the roof!
And where the speaker assumes the listener knows already that cats are around but now wants to emphasize the fact that one is
in the garden, one would get:
20 There's a cat in the garden.
In any given sentence the emphasis will be most commonly on one part and not another because that will express the emphasis
that is most common in life.
21 George has something to do.
But:
22 There must surely be something to see!

Page 103
But if there is a less common situation to express, one will put the stress in a less common position. In other words, one says
what one means, which is not very remarkable.
15a George has a reason to go.
would be a less common situation in life. But so would:
23 George has orders to go.
though here the stress is switched.
A final example perhaps puts the principle succinctly. Compare 24 and 25
24 The fire brigade's arrived!
25 The fire brigade's gone.
In the most common 'situation' that is the way I believe those sentences would be, are, stressed. I may be wrong that those are
the most common stress patterns. That would be irrelevant to the argument. It would merely mean that I am wrong about what is
the most common situation or real life context. It would not affect the basic principle: emphasis follows what is thought to need
emphasizing.
In the linguists' approach to pronunciation, as in their approach to all the rest of language, with its obsession with an analytical
scheme for its own sake, everything having to interlock neatly, connection with reality is apparently not important. Smith and
Wilson (1979, p.66) virtually affirm this in a footnote to the first of the two extracts I quoted from their book on page 98 above.
"Notice that this [claim that syntax is prior to both semantics and phonology] is not, and could not be, a claim about the
order of events that occur in actual speech production or perception. It would clearly be absurd to maintain that speakers
had to construct the full syntactic representation of an utterance before deciding what message they wanted to convey.
This is yet another reason for maintaining a radical distinction between the rules of a competence grammar and the
performance principles used for speech production and perception. While the two types of principle may interact, they are
nonetheless distinct."
But if this distinct 'competence' grammar exists it will have to be used for each particular sentence one produces. If it was true,
for instance, that

Page 104
the stress within any sentence depends on the syntax ('deep' or otherwise) then one would have to have the syntax of that
sentence clear in one's mind before one could decide on the stress. The same would apply to meaning.
The Chomskyan confusions and contradictions follow inevitably from the determination to arrive at a fixed schedule where in
one way or another (however contrived and far-fetched) the syntactic chicken lays the semantic and phonological eggs. The
truth is that meaning eggs and meaning chickens all depend on each other.
It is quite true that there are areas of convention where one cannot talk of any direct link with meaning. One is the convention of
gender in many languages. One can say, of course, that there is a direct dependence on meaning in that the combination in
French, for instance, is "la femme" (the woman), not "le femme". The masculine meaning "le" won't fit the feminine meaning
"femme". One could even maintain that many objects have been linked in the mind with the feminine or masculine idea: "le
soleil'' (the sun) but "la lune" (the moon) for example. But there are so many thousands of words in so many languages where it
is plain that the gender is pure convention that it would be straining absurdly to suggest any real link with meaning. This
becomes very obvious when, say, one comes upon the 'common' gender for "the war" in Norwegian - "krigen" - and the 'neuter'
gender for the same noun in a closely related language, Swedish - "kriget". The absence of any link with real life becomes even
more striking when one finds that - another Swedish example - "an important minister" is 'neuter': "ett viktigt statsråd". This last
sort of example is perhaps particularly important, because some would say, perhaps, that the 'neuter' "t" ending of "viktig" is a
clear case of structure, independent of meaning.
Originally there may well have been meaning attached to all gender forms. However, that meaning has long since been
forgotten, for the most part, and gender has become pure convention. But the gender convention must be seen for what it is -
certainly nothing to do with structure. "der Mann", "el dia" (the man, the day) etc. are essentially wholes, conventional
combinations no different in principle from the conventions, say, that in English it is "unfortunate" and "uneducated" but
"informal" and "inedible". 'Rules' such as that which says that adjectives linked to "statsråd" must have "t" on the end must be
seen in the same way. They are not really rules at all. They are all - "la femme, der Mann, el dia, unfortunate, vigtigt statsråd" -
automatic responses on the part of the speaker, learnt by imitation, as opposed to the logical decisions the native speaker has to
make

Page 105
much of the time when combining pieces of language, particularly when saying 'new' things.
This automatic response is certainly how the native speaker experiences these combinations. We should not be misled by any
attempt to find 'minimal syntactic units' or 'ultimate constituents' in a language, as if these formal abstractions were a reality. The
search of many linguists for these is pointless. It is the psychological reality, not the formal unreality, that is important. The old
distinction of something called 'accidence' is probably valid: the fixed convention for the combination of forms, and learnt by
familiarity and habit.1
Linguists say that some pieces of language are without meaning and have purely 'structural' or 'syntactic' functions - for example
"to" in
26 I want to go home.
They make this mistake partly, I think, for the same old reason that so often leads them astray: they are slaves to the ancient
grammatical tradition and so, associating "to" with the abstract concept of the 'infinitive', decide it must be purely 'syntactic'.
Perhaps even more important is that they find they cannot define the meaning of "to", and to their defining and analysing minds
that can only mean one thing - it has no meaning. Which is quite without logic.
"To" in fact, of course, has the meaning of to-ishness, and we can't define that, but we all (English-speakers) know exactly what
it means. "To" has its own special meaning just as, for instance, "-ing" has. "To'' only has a more 'subtle' meaning if we try to
define it.
Not all pieces of language are always necessary to meaning. That is quite another matter. It does not mean the 'unnecesary'
pieces have no meaning. "Hit" is enough for the past of "hit", but that does not mean the "-ed" in "asked" is meaningless.
English-speakers could, like Spanish-speakers ("soy") take away "I" from "I am" and everyone would understand just as well,
but that doesn't mean "I" in that phrase is meaningless.
As the last example shows particularly clearly, the use of a particular piece of language may often be pure convention. But if a
piece of language is only convention it can certainly not be structurally necessary. This is a more important point than it may
seem at first. Those who believe in
1 Linguists will no doubt insist on continuing to call these rules. However misleading a term, it suits the type of picture
they want to give of language. There does remain, however, the psychological mystery of of how 'familiarity' and 'habit'
work and what they work on. See the note on p.67, and Chapter 7, p. 147 and Chapter 9, pp.170 and 188.

Page 106
structure in language appear mostly to regard it as something quite independent of meaning, something with a life of its own,
which is absolutely essential to language.
But to return to "to". It is true that in several cases similar to "I want to go" there is a convention that "to'' is not used, as in, for
instance,
27 I must go.
or
28 She made me go.
while in the passive, where that is possible, the "to" appears once more:
29 I was made to go.
But this is further demonstration that "to" is not 'structurally' necessary, and it shows again how decisive meaning is. If "to" was
'structurally' necessary one would have to say
30* I saw her to go home.
as well as
31 I wanted her to go home.
In fact, it is only the 'modal' verbs (or most of them) such as "can", "should" etc., verbs of sense ("see, feel, observe, watch"
etc.) and "make" and "let" (with "help" undecided) that are used without "to". Notice, too, incidentally, how these words do not
have the same meaning as so-called synonyms:
32 The film made me cry.
is hardly the same as
33 The film forced me to cry.
There are many other examples one can think of to illustrate the meaning that this 'infinitive' "to" possesses.
34 Spurs to win the cup! / Spurs win the cup!
35 To be honest, it stinks. / Be honest - it stinks.
36 I wasn't trying to. / I wasn't trying.
37 He can play if he likes to. / He can play if he likes.

Page 107
The obvious falseness of the stimulus and response view of language should not drive people to the opposite extreme. A vast
and fundamental part of our use of language is pure imitation. The two basic elements of language are the conventions as to the
forms and the meanings of the pieces, which we copy; and the 'life' logic of meaning that determines how we combine those
conventional pieces. But the distinction between the two elements is not always clear. Why should it be? They often merge into
each other. Whatever something like the 'infinitive' "to" is, it is not part of some self-contained system, with a life of its own,
called 'structure'.
At this point some may object that while there may be meaning in "to" in the examples I have given, there is still no meaning in
"to" in ''I want to go home". This is to overlook a fundamental truth about language that seems to be largely ignored.
There are many contexts in which a difference in meaning has no practical importance, but this does not mean that the
difference is not still there. I shall show in Chapter 9 how serious the failure to see this is in a slightly different context. But
here, consider the following examples.
38a I have made the beds.
and
38b I have done the beds.
have, in the usual everyday context, the same practical significance; but that does not mean that "made" and "done" mean the
same thing, even in this context. They very definitely do not, as one can see easily if, for example, one considers that one would
not ask a child if he can do a bed; or if one considers
39a I have done your trousers.
where the practical equivalent is hardly
39b I have made your trousers.
in normal circumstances. "Washed", "ironed" or "mended" are all possible equivalents in 39a, and very likely only the owner of
the trousers will know which is indicated by "done", although all other listeners will understand exactly the meaning of "done".
There is no difference in practical meaning between
40a There are no cats in the house except mine.
and

Page 108
40b There are no cats in the house apart from from mine.
But "except" does not mean "apart from":
41 There are twenty-three cats in the house apart from mine.
"Except" is impossible here. In most contexts the phrases
42a music in the eighteenth century
and
42b the music of the eighteenth century
have the same practical meaning. But from that it does not follow that "the" does not change the meaning. Change "music" to
"religion" and the choice may be limited immediately:
43 Religion in the eighteenth century was tolerant.
but
44 The religion of the eighteenth century was rationalism.
Even if one assumes for the moment that there is never any practical difference in meaning between
45a South American horses
and
45b the homes of South America
(there can in fact be such a practical difference), there is still a fundamental difference in meaning, and this is simply, if nothing
else, that one phrase contains "the" and the other doesn't. One phrase is marked by a word that gives it the-ishness. The use of
such a basic word as "the", one that has such a basic meaning and is so much a basic part of native English-speakers'
consciousness, being absorbed almost with their mother's milk, must make a basic difference to the listener's apprehension of a
phrase. There is no need to develop complicated theories about its nature, as some have done. There is no mystery. "The" has a
simple, powerful meaning and colours deeply anything it touches.
Another word that some linguists apparently claim has no meaning and is just a syntactic element, is "some". Maybe there is not
much practical difference between

Page 109
46a There are books on the table.
and
46b There are some books on the table.
But surely, if asked what Smith does, few people would answer
47a He trains some horses.
as opposed to
47b He trains horses.1
However, perhaps one of the most important words to take up here is "it". In sentences like
48 It worries her that he smokes.
or
49 It is easy to boil an egg.
"it", they say, is a 'dummy'; it has no meaning, only a syntactic function. This, apparently, is because what these sentences
'really' are is
50 That he smokes worries her.
and
51 To boil an egg is easy.
So "it" is supposed to be superfluous as meaning, just a grammatical formality. Yet once again, it is surely clear, for instance,
that
52 To wait is best.
is a miserable and inadequate replacement for
53 It is best to wait.
even if one assumes that there is no important practical difference in meaning, which is very doubtful. In fact, the border betwen
practical and non-practical difference in meaning is often blurred. We can say, quite naturally
54 There are branches all over the world. One branch is in Scotland.
1 See also Chapter 4 above.

Page 110
and we can replace the second sentence by
54a There is one branch in Scotland.
but we cannot realistically replace the first, with its supposed meaningless 'dummy' "there", by
54b Branches are all over the world.
Readers can explore the quite significant implications of these examples by trying out slight variations for themselves:
54c Many branches are all over the world;... ?
The flaw in the 'dummy' argument is deeper even than this, though. The fact that one can replace "it" by "that" and "to"
expressions in the two types of sentence above (48, 49) is, I believe, only coincidence. All impartial observers would surely
agree that in
55 It seems he has left.
56 It is worth asking.
57 It is time to go.
"it" has the same 'function' (I would say 'meaning') as in 48 and 49, yet "it" cannot be replaced in the same way as it can in
those (50, 51).
Quite apart from this, the fact that it so happens that when one has a certain type of sentence one can always turn it round into
an "it" type sentence simply won't do as an argument that "it" is meaningless (even if there weren't many "it" sentences that can't
be replaced by other types of sentence). That some sentences can be interchanged without practical consequences does not prove
that they mean the same. Arguing that it does ends in absurdity.
58 There is a tiger in the garden.
can be turned into
59 A tiger is in the garden.1
1 Yet note that, as with many "it" sentences, not all "there" sentences can be re-cast either:
There is no point in waiting.
cannot be turned into
*No point is in waiting.

Page 111
but that does not prove they mean the same and that "there" is a 'dummy'. What one can reasonably claim is that "there" can be
used when one wants to express the idea that something exists somewhere. But one would not dream for that reason, surely, of
suggesting that the "exists" formula is an actual alternative to the "there"-type expression.
60 There's lots of bread. / Lots of bread exists.
Again, that it happens to be possible - sometimes - to use "to do" or "that he does..." as subjects of sentences does not mean that
"it" as subject merely replaces them and is a 'dummy'. And if the essence of their argument is that "it" is superfluous, they will
have to say that in, for example,
61 I need a bag for putting shopping in.
the words "for", "putting" and "in" are all meaningless 'dummies', since the sentence can be expressed as
62 I need a shopping bag.
So I think it is clear that "it" in the sort of sentences I have been discussing has its own independent meaning, a meaning in its
own right, the indefinable it-ishness. That is surely how native English-speakers experience it and feel it. The meaning, in fact,
is basically the same as in
63 It's raining.
And one has to ask why the "It is easy to boil an egg" form should ever have arisen if it simply means the same as "To boil an
egg is easy", if it does not meet a need for a different idea.1
What there is here is yet another example of the vicious circle that those who believe in the reality of structure get trapped in.
They can only see the language in formal abstract analytical terms, and so become blind to the real distinctions of meaning that
ordinary people are fully aware of.
There are other phenomena which suggest that 'structure', even if it existed as a distinct system, could not be a fundamental part
of language
1 For historical reasons, because languages change, there may be parts of a language which no longer have any important
meaning, but have been kept by habit. But what one certainly does not have to learn, nor need inborn, is any structural
system in its own right.

Page 112
and that, on the contrary, recognition of reality through meaning is the essence of language.
Where, as with many "-ing" form sentences, for example, there are no differences in objects or subjects or any other 'formal
functions', it seems to me impossible to claim that 'structure' has any part to play in showing us the differences in 'general'
meaning that we nevertheless immediately recognize. For instance:
64 Opening the door, he looked outside.
65 Hearing the door, he got up.
66 Holding the door, he waited there.
where 64 means that he looked outside after he opened the door, 65 that he got up because he heard the door, and 66 that he
waited at the same time as he held the door. Here are three sentences where all three "-ing" forms have a common 'general'
meaning, i.e. someone's activity, but the different 'specific' meanings of each sentence, in combination with that common
'general' meaning ("-ing") give different 'general' meanings to each sentence as a whole. Without those 'specific' meanings the "-
ing"s are ambiguous, but they are not ambiguous in the same way as Chomsky's "-ing" discussed in Chapter 2. That ''-ing"
could have more than one 'general' meaning; these have only one, and it is not susceptible to any structural 'transformation'.
On the other hand, while common human experience makes the meaning of the three sentences above (64, 65, 66) quite
unmistakable, where a common human experience or attitude is lacking meanings can be uncertain. For example, does
67 Feeling very unwell, he opened another bottle of whisky.
mean he opened another bottle because he felt unwell, or while he was feeling unwell? One's answer may well depend on one's
view of alcohol.
Pairs of sentences such as 68-70 are further examples of how decisive it is for language users to recognize situations through
their experience of reality.
68a If you like Bill we'll invite him to the concert.
68b If you like Bach you'll enjoy the concert.
69a When the doctor saw Mr Wagjaw he left hurriedly.

Page 113
69b When the doctor had seen Mr Wagjaw he left hurriedly.
70a As I was going past the pub I promised myself never to touch another drop.
70b As I was going past the pub anyway I promised to fetch the drinks.
Only real life experience of the distinction between Bill and Bach can tell one that "like" is probably future in 68a and present
in 68b: the 'specific' tells one the 'general'. In 69 the 'general' tells one the 'specific', that is to say, the past tense in 69a (but
again only in combination with the 'specificness' of "see") tells one that the meaning is equivalent to "set eyes on", while ''had
seen" in 69b indicates "attended to". In 70b the addition of the 'specific' "anyway" is alone enough to make it clear that "As I
was going past" expresses a plan that was the cause of the promise, in contrast to the same five words in 70a, which tell one
when the promise was made. But in all three pairs the essential factor is, as always, recognition of realities.
Again, no structural analysis (of surface or 'deep' structure - see Chapter 8) and no purely linguistic information - if such a thing
exists - is going to help one find out the meaning of simple sentences like
71 They hit me.
or
72 He blew it up.
Only knowledge of the real life context will do that. Any 'structural' analysis is impossible before one knows the meaning, and
so seems a total irrelevance.
I believe that if one thinks with an open and realistic mind about exclusively structural (or syntactic) elements one must find that
they are an impossible concept in a real natural language. Nobody, I think, could give a convincing explanation or description of
it. What is an exclusively structural or syntactic element? What does it mean to say such an element is unnecesary for the
meaning but necessary for the structure? It proposes something in a vacuum. But such a vacuum cannot exist in a natural
language. It is only in the formal mathematical abstractionizings of the linguists that it can.
But if the idea of isolated structural elements is difficult to justify, it is probably even more difficult to justify the idea of a
whole system of structure with an independent life of its own.

Page 114
6
More Fantasy of Structure
Many linguists appear to be obsessed with the order of words. So much of what they mean when they talk about 'grammar',
'structure' or 'syntax' turns out to be only this order of words. When Chomsky draws attention to the 'grammaticalness' of
colourless green ideas sleep furiously
as opposed to
furiously sleep ideas green colourless1
he can only, of course, be referring to the word order, since everything else is exactly the same in the two sequences.
It is difficult to see the reason for singling out word order for such an important and fundamental part. I have already pointed
out how, for a start, one can only recognize any significant order of the words in a sentence through first recognizing their
meaning.
Humans can only make sounds one after the other, in a straight line so to speak, so obviously words have to be spoken in some
sort of order. This does not mean that their order is always crucial.
There are many individual words that we are certain of the meaning of even when they are used quite alone; many more
become perfectly clear when they are used with other words, even if jumbled up and without the 'proper' 'grammatical' words to
link them. In a very large proportion of cases, probably, one can understand accurately all the important meaning of a group of
words. This is because "wind" is still wind, "house" is still house, "down" is still down, and "blew'' is still blew, and because
from one's experience of the world one knows the likely connection between these different things, actions, and places. Where
'grammar' becomes useful or even necessary is when there are several possible connections between the things and actions etc.
kiss girl nose boy
is not clear. But add the 'grammar' and one only adds more meaning; 'grammar' is meaning too, it means real things in life, in
experience - things like time, and what comes before what, and who does something to who. The English convention is that the
recognizable thing that is 'said' before an
1 See p.94 above.

Page 115
action is the 'do-er' and the thing 'said' after the action is the 'done-to'. (This makes it very difficult for English-speakers, among
others, to get used to a language like Latin, where the case system makes word order far less important - particularly if it is
taught, as it certainly was in my school-days, as a system of structure instead of a set of meanings about real life.)
Chomsky's basic arguments for all human language being 'structure dependent' appear mainly - if not all - to be concerned with
word order. In a discussion with Stuart Hampshire, broadcast in England in October 1968, he maintained that
"universal grammar - that means that set of properties which is common to any natural language, necessarily, by
biological necessity -is really a rich and highly articulated structure...with very explicit restrictions on the kind of
operations that can occur, restrictions that we can easily imagine violating."
He then suggests that if a mathematician were asked to design operations on sentences (that's a huge assumption for a start - that
'operations' on sentences are ever carried out in real life - but more of that later) he would think of elementary operations such as
reading (note "reading", not "saying") the sentence back to front in order to form a question; yet in fact there is no human
language that does this. Chomsky says it is not obvious why this should be so, as it is a much simpler operation than the one we
use to form questions in English, for example.
"Nevertheless, the principles that determine what operations may apply in a natural language preclude such simple
operations...in fact all these operations [have] the very interesting formal property...[that] they are...structure dependent
operations. They are operations which apply to a sequence of words, not by virtue of the internal content of that sequence
but by virtue of an abstract structure associated with them. Each structure has a set of phrases, and so on...all of the
operations that apply to sentences are structure dependent operations in this sense - that's an example of a simple
linguistic universal that you cannot explain on the grounds of communicative efficiency, or simplicity, or anything of that
sort, it must simply be a biological property of the human mind."
First, why should sentence operations ever be mathematically based? Chomsky's speculation can surely only come from his
imprisonment in his

Page 116
abstract cage. We are trying, with language, to represent the real world and our thoughts about it, and so naturally we follow, as
far as we can, the logic of life, the logic of things as they are - a practical logic. I return to this crucial fact below. But as an
example of how the word orders we use are not necessary for efficient communication Chomsky said to Hampshire that we
could perfectly well have a system in which
"the question associated with 'John saw Bill'"
- and at this point he momentarily hesitated and then said:
"you know, 'John saw Bill yesterday' would be 'yesterday Bill saw John'."
That added "yesterday" was essential to Chomsky's whole case. If he had kept to his original sentence of "John saw Bill" the
reversed form by his imaginary question system would of course be "Bill saw John", and that obviously wouldn't do at all. There
are very simple and basic restrictions on word order imposed by the needs of communicating efficiently, or perhaps one should
say, by the need to make meanings clear and distinguish between them.1
Furthermore, a scheme like reversing the word order to make questions would not be at all simple in practice when one had to
deal with long sentences. The brain would find it a difficult process, I think; it is only simple as an abstract principle.
There is what seems to me a very clear statement of Chomsky's position on 'structure dependency' in his Reflections on language
(1976, pp.30-33). Imagine, he says, a neutral scientist observing a child learning English. (Chomsky is dogged at every turn by
his abstract upbringing - a truly neutral scientist would be useless for investigating human language and might just as well be a
cow or a Martian; the investigator needs the human prejudices, that is, human experience and attitudes to even begin
1 That word order is essential to language is one of those things that half a minute's thought makes plain. For instance:
Sue gave Bill a kiss and Bill gave Sue a smile because Ted gave Joan a hug.
cannot be understood without conventions of word order, even if subjects and objects are indicated.
Sue-sub. Bill-sub. Ted-sub. Bill-ind.ob. Sue-ind. ob.Joan-ind.ob. gave gave gave a a a and because hug-ob. kiss-ob.
smile-ob.
could be interpreted in all sorts of ways. Yet people find employment pronouncing the obvious as if it was profound.
When asked "There are quite different word orders in different languages, aren't there - to what extent can you think of a
general rule that will cope with that?", an expert academic answered, "The general rule is that the order of words has
significance. There isn't any language that has totally free word order."

Page 117
the examination; all will be meaningless otherwise.) The scientist, says Chomsky, may discover that the child has learned to
form questions out of declarations as, for instance, with:
the man is tall - is the man tall?
From such facts the scientist might arrive at the hypothesis I that what the child is doing is processing the declarative sentence
from its first word until he reaches the first occurence of "is" and then preposes that "is" to produce the corresponding question.
This hypothesis is of course false as can be seen from the example
1a the man who is tall is in the room - is the man who is tall in the room?
1b the man who is tall is in the room - is the man who tall is in the room?
The correct hypothesis 2, Chomsky declares, is that the child analyses the declarative sentence into abstract phrases; he then
locates the first occurrence of "is" that follows the first noun phrase; he then preposes this occurrence of "is", forming the
corresponding question. Hypothesis 1 says the child is using a 'structure-independent' rule. Hypothesis 2 says the child is using a
'structure-dependent' rule,
"a rule that involves analysis into words and phrases, and the property 'earliest' defined on sequences of words analyzed
into abstract phrases. The phrases are 'abstract' in the sense that neither their boundaries nor their categories (noun phrase,
verb phrase, etc.) need be physically marked. Sentences do not appear with brackets, intonation boundaries regularly
marking phrases... or anything of the sort."
Chomsky maintains that the only reason the child unerringly uses the structure-dependent rule of hypothesis 2 instead of the
simpler structure-independent rule of hypothesis 1, and although he may never have experienced sentences like la, must be that
the rules must be 'structure-dependent'. This principle is inborn and is "part of the conditions for language learning". Chomsky
says all other rules of English are 'structure-dependent' too, and as humans are not designed to learn only English, the principle
of 'structure-dependency' must be universal, part of the human language faculty.

Page 118
One thing in Chomsky's argument is right, which is where he points out that children always get sentences like la correct
without any training. But that is in no way the mystery that Chomsky believes.
Children live in the real world and hear language about real things. How can purely abstract principles of analysis be applied to
actual language about real things? I do not see how Chomsky can explain this. My mind can have been born as full as you like
of x(y+z)-type analytical faculties, but that won't help me deal with "the man who is tall is in the room", precisely because one
is abstract and the other isn't. One has no meaning and the other has. If one is going to analyse the second abstractly it can only
be after one has understood the meaning, by which time there is no point in analysing it.
Further, Chomsky himself insists that children are not born ready for any particular 'grammar', so it is difficult to see how, even
if they were biologically prepared for analysing language in principle, they could in practice cope with the particular one they
are born into.
What in fact happens is simple and obvious, exactly what uncorrupted people think they are doing, though of course they do not
waste their time analysing and describing it.
Language users, including children - and indeed what I now describe is basic to the learning of language - manipulate 'lumps' of
meanings made up of 'smaller' meanings. Included in this 'manipulation' is the order of the words, which are arranged according
to conventions of word order that they have observed; this must be so - questions, for example, are asked in different ways in
different languages.
These lumps of meaning have nothing at all to do with analysis into abstract 'phrases', although unfortunately it is very easy to
make such analysis in most cases. The 'lumps' are lumps solely on the basis of meaning; we naturally think of the meanings "the
person on the left", for example, as one lump, and we will naturally keep that lump together (as long as we want to) when we
want to say something different about that person - a question about the person, instead of a statement, say.
When we listen we identify the 'lumps' by recognizing realities; that is, we recognize and understand things in the world, not
sentences or phrases or structures. In the same way, when we talk we talk about things.
It is easy to manipulate the 'lumps' 'correctly' because language users are simply using the language to refer to the pictures of the
real world they have in their minds. There is no problem, even for small children just beginning to learn English, in turning "the
man who is tall is in the room" into a question (assuming that we do 'turn' sentences like that), even if up to

Page 119
that point the children have never heard any questions longer than "Is daddy in the room?" (I say "longer", not "more complex",
because it is only a misunderstanding produced by the abstract approach to see the longer sentence as a more complex one.)
Children will recognize instantly, in the words "the man who is tall'', just one thing in reality, and so without any further thought
will reverse the whole of that lump of meaning with the right "is" to form the question: "Is the man who is tall in the room?"
One can be pretty certain that if children were not so firmly anchored in the real world, and tried to learn their languages in the
way some of their elders think they do, they would not get very far. Linguists only find abstract problems in language because
they mistakenly think language is a purely abstract phenomenon. If you are doubtful about my explanation of how we
'manipulate' words, think about the following sentences:
2 The man the man the man knew knew knew.1
This is very easy to analyse into an abstract 'phrase structure' -[x[x[xy]y]y] - but almost impossible to work out the meaning of.
This is simply because there are no sensible, realistic things to fasten on to, to give us our bearings; the sentence is, effectively,
just an abstraction. Yet
3 Did you realize that bomb a radical immigrant the finance minister that ass Smith appointed last year employs in his own
private bank managed to make in the small amount of spare time the minister allows him, and put under the self-important fool's
chair yesterday, was a toy?
is immediately comprehensible, although certainly one has to keep one's wits about one.
But 3 is in formal abstract terms a far more complicated sentence than 2. Like 2, it consists of what the linguists' jargon calls
'self-embedding' sentences, and it is a 'foursome', while 2 is only a 'threesome'. And 3 is not only more complicated in that
respect; it is all contained within a question, there is yet another 'relative' idea within one of the 'self-embedding' sentences, and
it is packed with different ideas. By 'structure'-thinking standards it ought to be horribly difficult to understand, yet it is not,
while 2 is almost impossible.
1 See also above, Chapter 3, p.53, and below, Chapter 7, p.137, and Chapter 8, p.155.

Page 120
The answer is once more that 3 is full of recognizable things. It doesn't matter a bit what sort of abstractions it can be analysed
into. Sentence 3 conjures up pictures of reality, and it is precisely the variety of vivid images that makes it so much easier to
understand than the sterile abstraction of sentence 2.
At one stage in his discussion with Hampshire (1968) Chomsky appears to become very muddled. In the course of insisting that
universal principles narrowly restrict the structures and processes possible in human language he suggests the example:
4 John kept the car in the garage.
This, he points out, is ambiguous. It could mean that John kept the car that was in the garage but not another, or that the place
where John kept the car was the garage. Chomsky then says that if he forms a question from that sentence, by transposing "the
garage" to the front of the sentence and replacing it by "what" -
5 What did John keep the car in?
- it is unambiguous, it can no longer refer to the car that was in the garage; but it is impossible to apply the same process to
form a question from the sentence
6 John kept the car that was in the garage.
One can't say
7* What did John keep the car that was in?
Chomsky maintains that English-speakers have no evidence to show them that they cannot form such a question, so it must be
one of the universal and unlearnable linguistic principles in all humans that
"there are certain complex noun phrases such as 'the car that was in the garage' from which a noun phrase cannot be
extracted by an operation of question formation or anything else."
Actually, Chomsky's original presentation of this argument was even more muddled than this; I have tidied it up. He brings up
the car and garage example three times in the discussion, and the first two times he makes the question form not "What did
John...?" but "What garage did

Page 121
John...?" Of course such a question has no connection with the original statement
4 John kept the car in the garage.
- it would instead be associated with a statement such as
8 John kept the car in the new garage.
This mistake is not in itself important, but it is a sign of the carelessness with which Chomsky associates pairs of sentences, in
such a way that trust in his argument is considerably undermined.
In the case of
6 John kept the car that was in the garage.
nobody wants to ask the question either "What garage did John...?" or "What did John...? "What" is quite irrelevant to the whole
idea. It is surely only Chomsky's intoxication with the formal mathematical abstract and his preconceiving obsession with
'transformations' that make him want to do such a thing. The associated questions every ordinary person will want to ask will be,
simply:
9 Did John keep the car that was in the garage?
or
10 Which car did John keep?
As I have shown above, it is no mysterious abstract principle of controlling brain that keeps the phrase "the car that was in the
garage" together. It is simply that that car is one thing in real life, and of course recognized as such by all English-speakers.
What Chomsky is doing is abusing an already established convention of meaning in order to try to show the existence of
general abstract principles in advance of that convention. It is no argument to claim that that very convention is conditioned by
those principles in the first place. The point is that
7* What did John keep the car that was in?

Page 122
already means something else, but is incomplete. It is the beginning of a sentence that might, for example, be
11 What did John keep the car that was in town in?
and if you want an associated statement it would be
12 John kept the car that was in town in the garage.
How Chomsky is trapped in analytical abstraction is shown by his 'A-over-A' principle; and at the same time Lyons' account of
the 'problem' is in fact a perfect demonstration of the 'real things' I have emphasized above. The 'A-over-A' principle
"means that, if a transformational rule makes reference to a phrase of type A and the string of elements to which the rule
applies contains two such phrases, one being included within the other, the rule will operate only upon the larger phrase.
(In the associated phrase marker the larger phrase of type A dominates the phrase of type A that it includes.) Obvious
examples of strings of elements that come within the scope of this principle are noun phrases which contain noun phrases.
For instance, the book on the desk is a noun phrase, and the desk, which is included within it, is also a noun phrase.
According to the 'A-over-A' principle, any rule that moves, deletes or otherwise operates upon noun phrases could apply
to the whole phrase the book on the desk, but not to the desk. There are a number of facts in the grammar of English and
other languages that appear to be satisfactorily explained in terms of this general principle. On the other hand, as
Chomsky points out, there are certain rules that violate the principle, although they appear to be otherwise motivated; and
in the present state of research it is not clear whether the 'A-over-A' principle should just be abandoned or whether it is
possible to modify it in such a way that it will cover the exceptions also. And this seems to hold for all the more specific
constraints that have been proposed so far: they are only partially satisfactory, since they explain only some of the
relevant data. Although the 'A-over-A' principle, as it is at present formulated, is, by Chomsky's own admission, probably
not valid, it will serve as an illustration of the kind of constraints upon the application of grammatical rules that Chomsky
has in mind when he talks of the formal universals of linguistic theory." (Lyons, 1977, pp. 130-31)

Page 123
This is a strange fuss about something so obvious to children (for instance), who do not enjoy a training in linguistics and so
can't be stopped seeing that the book on the desk is one thing, not two. It probably sums up as well as anything how skew-whiff
linguists' thought is.
In another plea (in Reflections on language, 1976, p.89) for the fundamental importance of innate structural principles Chomsky
makes exactly the same basic mistake. He points out that we can say
13 Who did John discover that Bill had seen?
and
14 John discovered who Bill had seen.
but that while we can say
15 Who did John discover pictures of?
we cannot say the equivalent of 14, that is
16*John discovered who pictures of
Chomsky has just before introduced the concept of 'complementizer' (such as the word "that", as in "they know that Bill shot
someone") and in 14, it appears, "who" occupies the 'complementizer' position. Chomsky then says that (unlike some earlier
cases he has cited) no change in verb will help 16. It is simply ungrammatical, and will remain so however much we change the
lexical items. It is the structure that is impossible, he says (Chomsky's italics), and the fundamental difference between 14 and
16 is that noun phrases do not have complementizers (my italics).
For a start, Chomsky is once more up to the linguists' old trick of arguing in a circle. He is assuming that "who" is structure"
while words like "discover" or "find" are mere 'lexical items'. This is an arbitrary distinction and an arbitrary assumption, and in
fact "who'' is in essence meaning just as much as any other word, and its meaning will make certain demands on other words
that go with it.
The simple and obvious truth is that there are a vast number of meanings that will not fit together sensibly. It is profitless to
subject that truth to detailed abstract analysis. It is false to suggest that it is that analysis which is the reason for things, when in
fact it is a mere labelling system only made possible by the reality of meaning. And the problems raised by that analysis are
artificial. No person in real life would ever want to make these 'ungrammatical' combinations, not at any level of their
consciousness or unconsciousness. There is no more point in declaring solemnly that we can't say "John discovered who pictures
of" than in declaring we can't say "John discovered no pictures of" or "John discovered of of of". The trouble is that

Page 124
the transformationalists take it all seriously and have their rules all ready for those eventualities too - or, if not, are doubtless
working on them. They want us to believe the engine is driven by the smoke. So Chomsky's statement that 16 is impossible
because noun phrases do not have complementizers does not really mean anything, in the sense that it provides no new
information, is only a formal and uninteresting acknowledgement of what every user of English knows. Declarations such as this
are parts of a system of definitions turning on each other in a vacuum.
It is perfectly true that one cannot make 16 sensible even by replacing "who". But nor can one make things any better by not
having any word at all in the place of "who": "John discovered pictures of". Something is missing, and this is the whole point.
Chomsky's 'impossible' 16 is the beginning of a meaning that is quite different from 15. The problem is confused by Chomsky's
linking of 15 with 16, and 13 with 14. He insists on 'deriving' (or, in the case of 16, trying to derive) these sentences from
common base structures, that is, 13 and 14 from "John discovered Bill had seen who" and 15 and 16 from "John discovered
pictures of who". (I have left out the analytical indications that Chomsky adds in his original examples.) This is a basic flaw in
Chomsky's system that I discuss further in Chapter 8. But here the immediate result is an example of the illusion that Chomsky's
system is. If one applies formal abstract operations to language as Chomsky does, it may be possible to make the gullible think
it all very scientific. It makes it very easy to come up with 'sentences' which can be presented as demonstrations of elaborate
formal structural restraints intrinsic to language. In fact, many of the sentences that result from such methods are bound to be
'ungrammatical'. Apply transformational rules, and absurdities are inevitable, so one has to think up further rules to neutralize
the absurdities. Actually, the whole procedure is in effect completely arbitrary. One might just as well pick - or invent - any
ungrammatical sequence of words at random and then claim that it shows how language is controlled by basic structural rules. In
the present case Chomsky gives us the impression that there are connections between sentences that actually have no logical
connection and give no justification for the conclusions he draws. He obscures the fact that 18 and 15 are questions, while 14
and 16 are statements, even going so far as to say that 13 is derived from the supposed base structure ("John discovered Bill had
seen who") through 14 by means of rules of wh-movement.1 (It is possibly
1 Moving words like "who", "which", etc. See above, Chapter 3, p.44ff.

Page 125
relevant to questions some people might ask at this point to note that the classification 'indirect question' is often wrong. In "I
know who did it" there is no indirect question. The sentence is pure statement, with the sense "I know (the identity of) the one
that did it." "who" has only relative meaning, no question meaning. So one should not be tempted to think that for instance 13
and 14 have something in common because they are both questions, one direct, the other indirect. )1
The sort of statement that ought to be paired with question
13 Who did John discover that Bill had seen?
is
17 John discovered that Bill had seen Tom.
and with
15 Who did John discover pictures of?
it ought to be
18 John discovered pictures of Tom.
'Sentence'
16*John discovered who pictures of
is the beginning of a completely different statement, quite irrelevant to 15. We are told that John discovered something - the
identity of the person that - but we are not told what the whole something was, and so the sentence is 'unacceptable',
'ungrammatical' if you like, but all that means in this case is that the sentence is not finished, i.e. that the meaning is not sensibly
complete. We need something like
16a John discovered who pictures of Bill were taken by.
in order to make it whole.
I am not saying his statement about complementizers is untrue; in a barren way it may well be the equivalent of part of what I
have said.2 But this is indeed no more than a statement, flat and lifeless, while what I have provided is an explanation, both
simpler and far closer to reality.
I am not offering alternative linguistics, proposing a new system of language analysis - quite the reverse. Even a whole body of
excellent
1 See page 126 below.
2 See above, pp. 123-24. I am puzzled, though, because I would have thought that "Believing that one does [is a mistake]"
and "the thought that she might [is exciting]", for example, are what linguists would classify as noun phrases. But is not
"that" in these phrases a 'complementizer? However, the only interest of this point is that it perhaps illustrates the difficulties
and contradictions that trying to erect an edifice of structural analysis is bound to lead to.

Page 126
explanations for innumerable cases would not give new knowledge or insight. It would merely be an articulation of what
everybody already knows, and would neither increase a single person's mastery of their language nor grant an iota more
understanding of human psychology. There would have been no need for my explanation above if it had not been for the
contrivances of the linguists.
My account, though, is superior not only because it is an explanation as opposed to a meaningless rule (it shows why something
must happen rather than just saying "it must happen"). It is important if only because it affirms once more the valid principle of
meaning.
Just before his discussion of sentences 13-16 Chomsky points out that we can say
19 The police think that the FBI discovered who Bill shot.
but not
20* The police think who the FBI discovered that Bill shot.
but that we can make 20 right by replacing "think" with "know":
21 The police know who the FBI discovered that Bill shot.
The difference, he says, is because the verb "know" takes an indirect question complement, but the verb "think" does not. First,
he has got his 'explanation' wrong, even in his own terms. 21 does not involve an indirect question.1 It has the sense "...know
the identity of the person that..." 20 is impossible because, as all English-speakers know, the meaning of "think" is not
susceptible to such use: ''...think the identity of the person that..." Chomsky's statement is wrong in another way as well, though.
Consider
22 I could not think who I should ask.
"who I should ask" probably can reasonably be called an indirect question -certainly much more so than 20 and 21. Now I
imagine that this was a slip on Chomsky's part, and he presumably knows as well as I do that in a sentence like 22 "think" has a
different sense ("get an idea about" as opposed to "believe") from the "think" he was talking about in 20. But I think his slip is
significant, because it is typical of an insistence on a foundation of mechanical structure rather than a foundation of meaning.
The fundamental explanation of the 'impossibility' of 20 is not a detailed formal rule but the much broader and simpler principle
that sentences have to
1 See p. 125 above.

Page 127
make sense. Please do not reject this conclusion just because it is simple and common sense, and does not involve a piece of
that formal definition that appears to be for many an unsuppressible itch and is so generally unquestioningly assumed to be the
essential mark of the scientific and the profound.
Word order, then - not structure - is very simply of two kinds. First there is the sort that is determined and limited by practical
necessity. Once the meanings of the pieces of a language are fixed, the variations of word order that are possible are restricted.
We have seen, for instance, that if it is 'established' that "cats" and "mice", say, can mean both 'do-er' and 'done-to' individuals,
and in statements the 'do-er' comes first, followed by action, followed by 'done-to', e.g.
23 Cats chase mice.
then we cannot possibly have questions formed by putting the words in reverse order
24 Mice chase cats?
because this gives exactly the reverse meaning.1 Obviously, however, if a language has cases, much greater flexibility of word
order is possible.
When sentences are longer the restrictions usually become even tighter. We have seen what happens with "The man who is tall
is in the room" for example. But we should not really even say "when the sentence gets longer", but ''when there are more
meanings to deal with". For it is ultimately the logic of reality, the situations in the real world, and our ideas about them,
initially independent of language, but then bound by the one-word-after-the-other limitation, which dictate our word orders.
Some restrictions on word orders are also essential, or at least convenient, when one would otherwise not know what belongs to
what. One can't, for example, have the third word always being the action, because in practice this would restrict the language to
saying so very few things. But one can always attach meaning to the beginning or the end because the beginning and the end
are, so to speak, absolute, a solid fixed base to rely on.
1 I gave another little example of these simple basic limitations on page 8; the Swedish "bok/en" as opposed to "en bok".

Page 128
That is why, for instance, one can put the preposition at the end, as in English, and yet have it attached clearly to the beginning.
25 This is the horse (that) I was telling you about.
It is easy to see where the "beginning" is here - the unspoken "that" - from the meaning of the whole. But here too, as always,
meaning is all important. One won't be able to find the 'beginning', or know what the end preposition belongs to, if one hears a
'sentence' without recognizable meaning in real life:
26 Rubbish is the research project computer operation records arguments for.
That's not only nonsense, but meaningless too; but I know what it means, because I made it up. You don't understand it because
it is unrecognizable as anything real, or even as fantasy, and in this as it were 'abstract' form it is impossible to find any 'ends' or
'beginnings'. But I can at least give it nonsense meaning if I put in a "that" somewhere to show what the "for" refers to. (There
are three places I could put it in.)
Then there is the second kind of word order, not dictated by necessity at all. Within the limits of possible word order the choice
of particular possibilities is purely a matter of convention; I have given one example in Chapter 1 (p.8). Another is the way
adjectives in French and English, say, are put in different positions in relation to nouns. In many cases adjectives are indeed in
the same position in the two languages, but the convention dictates that for certain senses adjectives come after nouns in French
and before them in English, although there seems to be no logical necessity for this.
So there is 'necessary' word order, and conventional word order, in human language. It is very misleading to give either of them
the name structure. Perhaps it is even dangerous, because of the grand, profound implications of the word.
It is perhaps interesting to consider what one does when faced with a jumble of words. For instance:
27 I the cat I went saw when
I do not think there is any evidence to suggest that when one tries to 'correct' such a sentence one looks for any structure. It is
true that one

Page 129
imagines various possible word orders. But the basic point is that meaning is both start and end; it is the meaning alone that
suggests those word orders. What one does essentially is imagine possible situations in real life that would fit the meanings of
the words; one does not refer to any structural pattern. There are in fact four possibilities for 27, I think:
27a I went when I saw the cat.
27b I saw the cat when I went.
27c When I saw the cat I went.
27d When I went I saw the cat.
But if one changes "when" into "so" one is immediately reduced to two possibilities:
28a I saw the cat so I went.
28b I went so I saw the cat.
This is purely because of the different logic involved in the meaning of "so" as opposed to that of "when". And if one uses
"sneezed" instead of "saw" the possibilities are reduced to none - again because of the logic involved in the meaning of all the
words.
Word order is very simply one of the elements of meaning that we observe in language as small children; and we can use it just
like the other elements to say things we have never heard or said before.1 If what one wants to say is
29a He stupidly began talking.
one does not say
29b He began talking stupidly.
because one instantly recognizes the implications of the particular combinations of meanings, even if it is the first time one says
such a thing.
I can see no theoretical case for a structural basis of language. But even if there was one, it is quite plain that in practice the
mind does not and cannot use such a structural basis to produce or understand meanings.
1 See below, Chapter 7.

Page 130
It has been a sad experience that occasionally when I have tried to explain my view of language to university students I have
met the criticism that I have failed to explain the hidden inner structure of language. I am trapped in their circular argument.
They have been brainwashed by their professors and writers of linguistic works into believing that these profound systems are a
law of nature. I am like a person who tries to show that there are no such things as ghosts, and is then told that he must be
wrong because he hasn't explained ghosts.
It may be possible to devise a logical and self-consistent structural system completely independent of meaning. But if it is
independent of meaning it is independent of language, that is to say, it has nothing whatever to do with language.
There is a logic of 'sentences' and 'clauses', but it is a logic of meaning. It is peculiar to say, all by itself,
30*When I got home
and one can say, if you like, that it is peculiar standing alone because it is a 'subordinate clause'. But it is a 'subordinate clause
because of the meaning of "when". Or, in fact, more directly and truthfully, it is peculiar by itself because of "when". Equally,
31 I am tired.
by itself is a perfectly natural thing to say because of the meaning.
32*But I am happy
though, is again peculiar in isolation, but peculiar in a slightly different way from 30 - once more entirely because of the
meaning of a particular word, "but". There is no need to set up any formal classification of the different types of sentence.
The linguists seem to assume without question that structure (or grammar, syntax etc.) and meaning are two different things. I
suspect the main reason for the continuation of this myth is simply that two different words were invented once upon a time.
Linguists still have to prove the distinction.

Page 131
7
Ungenerative Grammar
1 The happy little elephant with purple ears kissed the kind-hearted mouse with quivering whiskers on the tip of his friendly
nose.
I have never heard or read that sentence before, and you almost certainly haven't either, but naturally I can make it up and you
can understand it and recognize it as 'correct'.
A very great deal is made by linguists of this so-called 'creative' ability. Chomsky is spoken of almost as if he had discovered
this great truth. Although it is apparently recognized now that it was taken for granted from the beginning in western linguistic
theory, it is constantly emphasized as something that has to be appreciated as fundamental by students of language. It is
described as "nowadays almost a truism" as if it wasn't a truism before.
Yet every person, including small children, has always known this 'truth'. It is obvious; and has no special significance. We all
know we can say what we like. Nobody, I think, believes they are limited in what they can say to sentences they have already
heard.
So why do people who are told about this obvious truth for the first time often accept it as a new truth, a discovery? For two
connected reasons, I think.
First, they are told it by their teacher-experts, who they do not believe would make a great point of something that everybody
knows about already. Second, and ironically, they probably accept it as new truth because of the part of it that is false.
Linguists talk of 'generative' grammar. They present language as creative. They appear not to see that language itself is not and
cannot be creative. Language cannot 'generate' anything. Only human beings can create and generate in language, and they do
that from their experience and with their imagination. But because linguists transfer the creative ability from people to language,
people are indeed taken in, because this is a new idea to them, that language 'generates'.
What linguists say about 'generative' grammar and a number of connected themes is a fine example of their tendency to say
what is either new and untrue; or obvious or trivial, or both.
Not only do Chomskyans claim that the grammar of a language 'generates', which is clearly untrue if that means "creates"; and
say a language can be used to say an infinite number of things, which is obviously true. They say that "The grammatical
description of" a language "may be

Page 132
based upon a corpus of actually attested utterances, but it will describe these, and classify them as 'grammatical', only
incidentally as it were, by 'projecting' them on to the indefinitely large set" of potential sentences. (Lyons, 1977, p.38. Crystal,
1981 - p.223 - more coherently says that the finite rules that constitute the grammar of a language are 'projected' onto the
infinite set of possibilities.) This too, once one has worked it out, is unremarkable and obvious, just about as obvious as can be,
for all it means is that all sentences based on the rules of the grammar are grammatical. The grammatical is grammatical.
The trouble is that they also formulate this obviousness in another way. The grammar of a language, they say, is "a device which
generates all and only the grammatical sentences of a language" (Crystal, 1981, p.221 and Lyons, 1977, p.44). "Generate" is said
to mean "define as grammatical all the sentences of "a language, both those that have been recorded and those that have not
(Lyons, 1977, p.38). "A set of rules whose output is a particular language (or set of sentences) is said to generate that language
or those sentences" (Smith and Wilson, 1979, p.277). And so they say that a natural language ''consists of an indefinitely large
number of sentences", most of which have not "ever been uttered or will ever be uttered."
This is boloney, typical of the artificial contrivances of academic intellectuality. They will no doubt defend themselves
vigorously against such an accusation. But their defence will almost certainly only confirm the accusation.
As I have said before, a language consists of a set of meanings. It is a limited set, like the keys of a piano; the size of the set
varies from individual to individual; it includes 'specific' and 'general' meanings, which in turn include word order meanings.
This set of meanings can be used to make an infinite number of broader meanings.
So there is no mystery about our ability to use language to say new things. Once children have observed a fair number of
meanings they will quite obviously use those meanings to express both new experiences and what they imagine, as all of us do,
throughout our lives. When we have a new experience - and experiences includes wishes, reactions, etc. - we simply use the
particular but familiar pieces of our language that fit the new experience, and that will be a new combination. A child has heard
and understood
2 The dog's asleep.
and
3 The cat's tired.

Page 133
and then sees a third kind of situation for the first time and Says
4 The cat's asleep.
Quite simply the child says whatever means what she or he sees. And we use familiar words in new combinations in exactly the
same way to say our imaginings. It is all so natural and simple that one might find it embarrassing to expound it to a child.
I take up later the question of whether children have enough 'evidence' to build their language on, but the question is already
more than half answered by the fact that it is quite simply meanings that they 'collect'. They do not have to use anything they
have not already observed. And of course they cannot and do not use anything else. Chomsky says:
"Fortunately for us, we are rigidly preprogrammed with extremely rich systems that are part of our biological endowment.
Correspondingly, a small amount of rather degenerate experience allows a great leap into a rich cognitive system -
essentially uniform in a community and, in fact, roughly uniform for the species. We can say anything that we want over
an infinite range; other people will understand us though they have heard nothing like that before. We are able to do this
precisely because of that rigid programming; short of that, we would not be able to do it at all." (The Listener, 6 April
1978, p.435)
Quite apart from Chomsky's self-contradiction (he says in almost the same breath that because of our rigid genetic program
there are strict limits to what we can say and understand), that is very obviously absurd. If two children have learnt the numbers
up to ten, and know the words "cats" and "dogs", then one child can without any mystery see six cats and say "six cats" even if
she has never heard that phrase, and the other child can with as little mystery understand it. What is impossible to visualize is
how the children would produce such new phrases if they had nothing but Chomsky's preprogram to build on.
The Chomskyans agree that people do not always speak grammatically.1 They make mistakes. But they can all recognize that
they are mistakes, they know wrong sentences from right ones. So they know the rules of their grammar, and these rules are not
in the corpus (or
1 "Grammatical" in the broad sense, not in the traditional purist 'schoolmaster' prescriptive sense. (See the beginning of
Chapter 4.)

Page 134
samples) of attested utterances, they must be in the minds of the speakers (and listeners) of the language. Chomskyans make
much of this 'idealized' knowledge of the rules, of the organizing principles, and call this knowledge 'competence'; while the
production of actual sentences, mistakes and all, is 'performance'. They argue that this distinct concept of 'competence' is further
supported by the fact of 'creativity' in language - that we can all make up and understand new sentences.
All this is again a mass of the obvious; and that linguists make so much of these obvious truths is an example of that sad
phenomenon that academic and ideological absurdities are normally not exposed except by another academic or idealogue, who
is then credited with extraordinary insight (and almost inevitably produces new absurdities). This certainly seems the only
reasonable explanation for the portentous declaration that humans have knowledge of their grammar, and can distinguish
between right and wrong sentences. Apparently it "has often been maintained by linguists of an empiricist bent...that all the
utterances of a native speaker of a language are equally correct, and are proved to be so by the sole fact of their having being
uttered" (Lyons, 1977, p.39).
Just about everybody who has not been brainwashed by philosophers or social 'scientists' knows, has always known, that we
sometimes make mistakes when we speak. But academics and the like are often brainwashed - they are particularly prone to the
rubbish excesses of intellectual ideologies. So in such circles they get excited and tell us it is profoundly significant when, for
example, somebody says we make language mistakes, and recognize them as such. They make much of something which is no
more than that people notice when meanings don't fit.
Another sensational revelation that Chomsky is credited with is the insight that individuals' opinions about their language
(judgements about the grammaticalness, meanings of sentences etc.), their intuitions, are a proper and essential part of the data
that linguists should study. Chomsky and his followers have then been attacked for this unscientific attitude, accused of
unobjective 'mentalism', a term of abuse to 'true' scientists, and there is endless argument between various 'scientific'
philosophies.
This is a great fuss about something so easily clear to the uninitiated! But no doubt Chomsky was asking for it when he talked
about "intuitions". In fact what he is talking about has nothing at all to do with intuitions. What do we want to study? Our
knowledge of our language. Where is that knowledge? Inside us. So where do we look? Where in the world else would one look
but inside ourselves? Studying and thinking about our own

Page 135
language inside us, 'introspectively' if you like, is as empirical as it is possible to be. In fact, our own personal language is one of
the very few things we can study absolutely directly. Even in the physical sciences one cannot be sure that one has got all the
data right. One's senses may be making mistakes, one may have counted wrong, and so forth. There is no such intervening gap
between us and our language. Indeed, what we apprehend about our own language in a sense is that language. Introspection is
the only scientific course. We cannot examine knowledge of language in other people.
It is a very important but completely different point that we all to a greater or lesser degree have blindnesses, slownesses of
perception (as well as prejudices) about our own personal versions of language. An essential feature of our misjudgements is
that they nearly always err by being too restrictive: denying grammaticalness or distinctions of meaning which our actual use
shows we in fact appreciate.
Linguists talk much about the need for scientific principles of enquiry into the 'acceptability' of sentences. They are to use, in a
properly organized way, the 'intuitions' of both 'laymen' native speakers and professional linguists. It cannot be done. Both
laymen's and linguists' judgements are in practice useless for determining either fundamental 'acceptability' or the distinctions in
meaning that native speakers in fact make. I do not know if the reason is corruption by education or a deep incorrigible urge in
humans generally to establish rigid idealized models in the linguists' fashion, but laymen as well as the linguists are totally
unreliable - however the question is put to them - as formal judges of language. They will say "no, one can't say that" or "no,
there's no difference in meaning there" and be quite capable of in fact using that 'unacceptable' language or carefully making that
distinction in meaning when it comes to the practical point.
The truth is that there is only one way of definitely establishing 'acceptability', or distinctions in meaning - and that is through
observing specific, individual pieces of data as they arise. But - and here's the rub for the linguists - that's no use at all for
theoreticians intent on creating great generalizing systems. I guess there are few takers for the 'acceptability' or grammaticality
of
5 I saw the Smith today.
Yet, when I joined the staff of a language school many years ago, I was at first (I was later told) referred to by my colleagues - I
hope at least partly with a certain affection, even if mocking - as "the Gethin". In this one little

Page 136
phrase one can see the whole basic, simple principle of language at work. Words have meanings; it is the meanings that
determine how the words are used.1 Humans use those meanings to express whatever is appropriate to any particular situation;
in other words, again, they say what they mean.
There are of course certain, and many, meanings that cannot possibly be 'properly' combined, such as those I mentioned as
examples in Chapter 1 ("can kissed" etc.). But one has to be very careful whenever inclined to declare the 'ungrammaticality' of
an utterance. It is very easy to be caught out if one rushes into dogmatic statement, particularly if one wants to make
generalizations about a whole language community. One then naturally has to observe and check with the sentences of others,
and should then always give what may seem 'defective' sentences the benefit of any doubt, even if third parties disqualify them
too. Even speakers' (or writers') own subsequent condemnation of their sentences is not necessarily reliable.
Prejudices about language are another matter again. They may be aesthetic or social or of some other kind. (One may, for
instance, hesitate when one confuses one's own habits with those of others.) Yet even here I think one should avoid dividing
judgements about language into 'linguistic' and 'non-linguistic'. What one can say is that aesthetic or social judgements are not
connected - or at least not directly connected - with logic or practicality. But when someone complains about ambiguity or finds
an utterance unclear the borderline becomes indistinct once more. There is really no point at all in drawing up categories. All
one needs to do and should do if one wants to complain about a piece of language is to give one's specific reasons for that
particular complaint.
But to return to 'knowledge' of language and to sum up:
We have to have knowledge of our language in order to speak and understand it.
The only place we can examine that knowledge is inside ourselves.
It would be difficult to get more obvious than this. The lack of perception about it, followed by all the unwarranted song and
dance, can surely only be explained by the cultural burden of the past, the generations of philosophers and scholars with their
'isms' and programs, all taking up grand fixed theoretical positions, where they entrench themselves and cannot hear the sound
of simplicity for the noise of the battle.
As we have already seen, Chomsky and his followers make a lot, too, out of the distinction they insist exists between linguistic
rules and non-linguistic rules, between 'competence' rules and 'performance' rules; some
1 See also Chapter 4 above.

Page 137
sentences are 'acceptable' but not 'grammatical'; the 'perceptual strategies'1 (part of 'performance') that people often use for
understanding utterances are distinct from their knowledge of the language ('competence').
As I suggested in the previous chapter, there is no warrant for such an absolute division except the linguists' desire to establish a
separate independent and innate language faculty because their purposes require it. The argument is circular again. It amounts to
saying there is such an absolute division because there must be an absolute division.
In any case nobody can apply any of these faculties - distinct or not -until they have learnt meanings. This remains the
inescapable basis of the command of language, whatever arguments there may be over what follows afterwards. And people can
only learn those meanings from language that has been uttered. (Chomskyans appear to be unclear and undecided on the last
point.)
I suggest that the human control of language broadly works as follows: Once given meanings, once they have grasped meanings
(and in principle it does not matter whether the meanings are few or many) people combine them as they require in accordance
with their logical sense. They apply their logical sense to both 'general' meanings (syntax, sentence logic, if you like) and to
'specific' meanings, and to both mixed together. There is no clear separation between the two. And that logical sense has nothing
to do with formal logic, mathematical logic, abstract logic. It is 'life' logic, the logic of things in the world as people experience
them. Meanings are fitted together in a way that makes sense in that real world.2
When applied to 'general' meanings this 'life' logic is naturally rather strict. When applied to 'specific' meanings it is more
practical, and often people's practical sense will clash with the stricter logic. This is inevitable, since the only purpose of
meaning is meaning - practical meaning.
I have already given an example of a sentence which has strict 'general' logic but little sense:
1 For example, many people would interptret "The train left at midnight crashed" as meaning "The train left at midnight
and crashed", although it has a perfectly coherent meaning as it stands. This is an example of Smith and Wilson's (1979,
p. 41).
2 See Chapter 1. The way small children create new words that are 'incorrect', such as "I goed" or "foots", is unremarkable
too. Linguists sometimes make much of it and - in typical fashion - call it "constructing rules". This is one way of putting it
but it is a misleading way, implying some special extra faculty, some separate linguistic facility. It is surely in fact simply
that children apply their logic to meanings ("-d'' and "-s" in these cases) before they have heard or mastered the conventions
that override that logic.
3 See above, Chapter 3, p.53, and Chapter 6, p.119 and below Chapter 8, p.155.

Page 138
But the more abstract logic of such a sentence is useless in practical terms. That is, it is useless to most people. But there will be
some people perhaps whose 'general' logical perceptions are sharper than average, and who will therefore be able to make some
sense of 6.1 And of course there are many less outrageously unpractical sentences that will be manageable to many and not to
many others, simply because the power or training of people's logic varies greatly from individual to individual. Almost
everybody knows that humans vary strikingly - for whatever reason - in their ability to produce and understand sentences of
more complex logic; they vary in precision and in perception of meaning.
Practicality often overrides the strict logic of 'general' meanings. An example of this is an utterance produced by Hampshire
(1968) in his discussion with Chomsky:
7 ...inbuilt predispositions. Now, one thinks of them as placing a certain restriction on human thought, as if we were
wearing spectacles which if we took off we wouldn't see at all.
Between "which" and the end there are two serious illogicalities by the standards of strict 'general' meaning logic. (We wouldn't
see the spectacles if we left the ground in an aeroplane.) In cases like this, effective meaning, indeed one might say a rougher
kind of 'logic', takes over and is often accepted on both sides, speaker and listener. Now, Chomskyans acknowledge that this sort
of utterance is common; but as usual they insist on discriminating between the 'linguistic' and the 'non-linguistic'. But do they
seriously pretend that the psychological reality behind such sentences is an inner absolute knowledge of 'grammar' which is then
pushed aside by the work of extraneous totally non-linguistic faculties of mind, processes of 'performance'?
Is it not much more reasonable to suppose that the brain uses more general faculties which interact, look at the 'problem' as a
whole, take a mixture of factors into consideration - overlapping 'general' meanings and 'specific' meanings, strict logic and less
strict logic, contexts of various sorts -and often make compromises, sometimes for positive practical purposes, sometimes
through failing to see all the possibilities or implications. And it is not only that some people have a more developed logical
sense than others; the efficiency of the same individual's logic can vary greatly in different
1 Notice that at least in cases like this the problem cannot be connected with what psychologists and 'psycholinguists' call
'short-term' memory. My 'bomb' sentence (p.119) puts a far greater strain on short-term memory, but is far more
manageable.

Page 139
contexts, and according to the direction the speaker or listener approaches from. A good example of this and other fluctuating
and overlapping factors is provided by a sentence that Chomsky has used to make a false point -"Who does he wonder who
saw?" - which I discussed in detail in Chapter 3 (example 40q).
That basic human logic I am talking about is indeed a mystery. We know nothing of how it works. But we can find constant
evidence of it both within ourselves and in others.1 There is no such evidence of Chomsky's mental program of exclusively
linguistic analysis.
Scientists can naturally make mathematical analysis of language, as they can of virtually anything. It fits in very well with the
Chomskyan thesis of a special language program in the human brain. And so they can easily - as Chomsky and his followers
frequently do - refer to "highly technical" matters, so 'ordinary' people mostly dare not question them, believing that these are
expert affairs beyond their understanding. And indeed they are beyond the comprehension of the average non-mathematician.
But the terrible mistake that the outsider makes is to believe that these technical matters are relevant to the language problem;
the experts say they are, so they must be.
One of the basic flaws in the whole approach seems to be the pretension that this mathematical analysis reveals the active
psychological principle behind language. It is as if they mathematically analysed the different expressions on people's faces and
then claimed the analysis was the active principle behind the expressions. But it is worse than this even, far worse. For it is as if
they went on to say that expressions could not be satisfactorily described by analysis of facial movement alone; there are clearly
teams of gnomes secretly at work pulling the facial muscles about, so the gnomes must be mathematically analysed too to get at
the fundamental truth.
Chomsky and his followers arrived at the earth-shaking conclusion that humans know their grammar, they have their
'competence'. However, they were not trained as linguists for nothing, and so instead of looking for the equally obvious and also
sensible and likely basis of this knowledge, they had to describe, to analyse, and thrust the results on us as if they had explained
something.
1 In various experiments apes have been taught human sign language. If it could be shown that they spontaneously
produce what are for them new meaning combinations, the argument would be virtually settled. Apes are clearly not
'programmed' with a special language capacity; they do not spontaneously learn or acquire language. (Chomskyans always
stress this.) So if they can create new meaning they must be using a general logic possessed by apes and humans in
common, and perhaps by other animals as well. Unfortunately the evidence is so far inconclusive and controversial.
Needless to say, Chomsky will have none of it.

Page 140
Perhaps Chomskyans will deny that they are talking about an active psychological principle when they talk of 'generative'
grammar. But Chomsky really does talk about language - or grammar rather - as if it were some kind of living being. Even so,
perhaps what he means when he talks about it "generating" sentences really is no more than that from its parts (he would insist
'rules') an infinite number of combinations can be made by the speaker. If this is all he and his followers are emphasizing when
they talk of 'generative grammar', then they are indeed emphasizing the completely trivial, as trivial as emphasizing that from
the keys of a piano a player can 'generate' an infinite number of tunes.
If there is any mystery at all here, it is the mystery of the basic mental process that enables humans to see that they can attach
meanings to things, that is, refer to distinct experiences by producing distinct sounds. That is not a property of language even
assuming 'language' to be a special function of mind, but a property of the human brain that makes language possible. Once a
being has this faculty of putting meanings to experiences, it is not difficult to see how it will easily produce new meaning
combinations to refer to new experiences or imaginings. Imagining is another mystery, but again not part of language, although
it might well have had something to do with inventing the meaning 'system' in the first place.
Yet the pronouncements of Chomsky and other linguists do make it look as if they really do regard language, or grammar, as a
sort of organism with an independent life of its own. And it would be hard to understand otherwise why they should give
'generativeness' such a central position in their whole system. They refer constantly to the creative property of language, the
creativity of human language, generative grammar; not to creative or generative humans. It has been written that we must
possess, without realizing it, a concept of language which allows us to produce an infinite variety of comprehensible sentences:
"The idea of a grammar which embodies the principles of sentence formation and interpretation plays a crucial role in
explaining how novel utterances are produced, understood..." (Smith and Wilson, 1979, p.22)
"[Chomsky's] view of language as a creative, dynamic force, capable of making infinite use of finite means..." (Crystal,
1981, p.219)
"...the creative and dynamic potential of language..." (Crystal, 1981, p.224)
"The second general property of human language...is its creativity (or 'open endedness')." (Lyons, 1977, p.24)

Page 141
"...Chomsky considers that the creativity of language is one of its most characteristic features and one that poses a
particularly challenging problem for the development of a psychological theory of language use and language acquisition."
(Lyons, 1977, p.26)
"Chomsky lays great stress on the creativity...of human language." (Lyons, 1977, p.43)
"...it is this [generative] property of the grammar that reflects the creative aspect of human language." (Lyons, 1977, p.43)
We have already seen (p.133) how Chomsky himself says "We are able to do this [say anything we want] precisely because of
that rigid programming; short of that, we would not be able to it at all." It is perhaps difficult to understand how anybody could
ever take such ideas seriously, until one considers that the obvious truth would leave little to theorize about. The idea of
'generative grammar', as opposed to recognizing the power of our imaginations to make new sentences out of new experience
and out of pure fantasy even, is typical of the academic flight from reality into an abstract self-supporting system. They talk as if
language and 'grammar' create thought and imagination, as if the tail wags the dog. If they say that they of course understand
that it is imagination that creates new sentences, but that the grammar generates because the new sentences cannot be said
without it (i.e. the grammar), all they are saying is that one cannot express one's imagining without words. In other words, one
cannot use the words if one does not know the words. That is nothing to do with 'generating' or 'creating'.
The linguists have got it wrong either way. Either they claim that the creativity is a property of language itself or they do not.
They try to persuade us that the creativity is a fundamentally important reason for studying language as a separate self-contained
system. If they think that is because the creativity is a property of language itself they are wrong, since it clearly isn't. But if
they acknowledge that the creativity is a 'property' of human beings, the awareness of new experiences or the power of fresh
imaginings, then language does not come into it; what is significant is some much broader faculty or faculties of the human
mind. Creativity is in the player, not the piano, just as it is in speakers, not in the meanings they use. The notes on the piano and
the words of language are no more than the means available for giving form to what has already been created.

Page 142
In passing, it is worth noting how linguists give much too much emphasis to 'syntax' (that is, to 'general' meanings) in their
picture of 'generative grammar'. This is presumably because 'syntax' gives far more scope for the formulation of rules. In fact it
is much more important to know 'specific' meanings rather than 'general' ones if one wants to be able to say new things.
Chomsky claims that a generative grammar ('grammar' in the sense of a description of the rules) is a scientific theory. If
scientific theories are formulated in such a way that they cover the 'clear cases' then they can be used to decide the 'unclear
cases'. So if we are uncertain whether, for instance, the sentence
8 The house will have been being built.
is grammatical or not (the example is from Lyons) we formulate the rules of the grammar (for English) to cover all the definitely
acceptable sentences, such as
9 The house will have been built/The house is being built/They will have been building the house/etc.
but so that they exclude clearly unacceptable sentences like
10*The house can will be built.
Then we see if the rules generate 8. The rules in Chomsky's Syntactic structures in fact do so, Lyons says (1977, p.46).
Now, I am not saying that it is not possible to create a body of rules that 'work' in a superficial sense, that 'generate' a lot of
sentences that will be accepted as English. (Even the generative grammarians admit they are nowhere near achieving a complete
grammar even for English.) But the real problem is whether 'generative grammar' is what actually happens inside human beings.
You will understand by now that I regard any such system as completely artificial.
One of the first questions one must ask is why there should be any 'unclear cases' at all. Chomskyans insist that children do not
get enough evidence to form all the sentences they do form, and maintain therefore that the principles of 'generative grammar'
are inborn. In that case they

Page 143
must surely believe that possession of those principles should always lead everyone unerringly to 'well-formed' or 'acceptable'
sentences. This 'unerring' ability is greatly stressed at all other points in the Chomskyan argument for innate language faculties.
But if that is so, there should be no 'unclear cases'. We should all be using - albeit unconsciously - the rules of generative
grammar to determine 'correct' sentences and we should never be in any doubt about what is grammatical or not; it is, remember,
according to the Chomskyans, those very generative rules that are the criteria for grammatical and ungrammatical sentences.
Smith and Wilson (1979, p.234) acknowledge this difficulty. Once more they take refuge in an appeal to 'non-linguistic' factors.
"In short, the grammaticality judgements of speaker-hearers must be explained in some terms by a linguist who is
concerned with the psychological reality of linguistic knowledge. However, he is not committed to accepting every
judgement at face value; he may choose to ignore some of them in writing a grammar, and he is likely to make this choice
for one of two reasons. Either there is a clear non-linguistic explanation for them, and he has no need to account for them
in the grammar; or there is no clear non-linguistic explanation, but it is impossible to account for them within the type of
grammar he believes to be correct. Enough such cases might, of course, force him to revise his conception of grammar.
But a linguist who abandoned all hope of ever writing a grammar simply because of uncertainty about the grammatical
status of (4b)[A man picked up an umbrella who was smiling.]1 would be a rather irrational individual." (p.237)
This amounts to saying: "All cases that seem unclear are not really unclear, they merely seem so for non-linguistic reasons, even
if I cannot find a non-linguistic reason; all cases that seem unclear to people are really
1 Smith and Wilson have suggested on the previous page that changing the terms of their (4b) from past to future makes it
sound distinctly better;
"The way you will recognise our Agent will be as follows: you go into the airport, a man will be standing near a telephone
booth who will offer to buy you a drink."
This is a good example of the sort of thoughtless mistakes linguists can make (though I have exposed a much worse one by
the same authors in Chapter 3) and also illustrates well that basic principle that meaning, and the satisfactoriness of
sentences, depends so much on the specific things being talked about. Note first that in this case the improvement is not in
any way due to the tense.
I recognised the Agent: a man was standing near a telephone booth who offered to buy me a drink.
Secondly, it is one more sign of linguists' imprisonment within purely abstract structure that Smith and Wilson think that
what is to them a structural feature is the crucial factor, and that it doesn't matter a bit if one changes the real life situation.
(cf the original (4b) sentence.) See further p. 149 below.

Page 144
clear to them, although they don't know it. Their uncertainty is false, although I can't say why." Indeed, it amounts to a dogmatic
statement that people know their grammar because they know their grammar, and all apparently grammatical problems are not
really grammatical. And if one ignores the circularity of the argument and accepts it for a moment, everybody is still left not
knowing what is 'grammar' and what isn't. The pointlessness of the whole enterprise is again made clear.
The final sentence about the individual who might abandon hope of writing a grammar is of course wholly irrelevant; Smith and
Wilson seem to have forgotten which problem they were supposed to be addressing: Why should there be any 'unclear cases' of
grammaticality? Furthermore, in the physical sciences the 'clear cases' are facts. In linguistics the 'clear cases' are at best
sentences that are grammatical in the opinion of most people, while there will be many sentences that are 'clearly' grammatical in
the opinion of some but not to others.
And what are the generativists going to do about expressions like "It's me"? If large numbers of native English-speakers say
this, how are the linguists going to react? If they make a special extra rule to cover the usage, they are not sticking to the
principle of formulating a theory to cover the clear cases and then letting that grammar decide about the unclear ones. If on the
other hand they stick to that principle, they must find that it denies the grammaticalness of usage like "It's me" that undoubtedly
exists and is widely accepted. They are surely in a dilemma. They must either accept as grammatical everything that people
generally or widely actually say (apart from unintentional slips), in which case their system of using their 'generative' grammar
to decide on grammaticalness is useless; or they use that system and thus have to declare ungrammatical a vast amount of what
people in fact say, which is equally absurd.
Surely some of the sentences linguists choose as examples of 'unclear' cases are significant. Why just these? Why sentences like
11?
11 We will have been being tested for two hours soon.
I believe the linguists choose this particular sort of example of debatable sentences precisely because they know 'in advance' that
it is (or is not) a grammatical sentence. Now this is the very thing they deny (Lyons, 1977, p.46). Yet they will say that the
sentences of 9 are definitely acceptable. In other words, there is no doubt about such sentences; we know 'in advance'. But how?
Because Chomskyans always insist (quite rightly) that we recognize as correct sentences we have never heard, and so in
principle at

Page 145
least we may never have heard anything like "They will have been building the house". (They will say their 'competence' tells
them, and again raise the contradiction that it doesn't tell them about 8 or 11.)
Surely the answer is again simple and obvious. English-speakers (and linguists in particular) know 'in advance' that the
sentences of both 8 and 11 are grammatical because they know the meaning of all the words, and in those sentences the words
fit logically together. Many children will probably be able to say and judge these sentences, even if they have never heard "will
have been being + past participle", so long as they have heard and understood and appreciated (for example):
12a I am being chased by a tiger.
12b Tom will have arrived by now.
Such sentences will give them the 'general' meanings and logic that they need as a basis for longer verbs.
I have had conscious experience of this myself. In the course of drawing up verb tables for a grammar of English as a foreign
language, I was aware that I did not actually 'know' the 'future perfect continuous passive'. I had to 'work it out', and of course
that was not at all difficult. (Whether it is useful or desirable to publish such verb forms for the benefit of learners of foreign
languages is another question. In fact I think that so long as foreign languages are taught as they generally are now, it may be,
because it may reassure students that logic does really lead where it should.)
It is not always as easy as this. There is a lot of detailed work to do in learning a language. Smith and Wilson (1979, pp.22-23)
present the sentences
13a It is likely that John will leave.
b It is probable that John will leave.
They point out, in order to demonstrate that "novel utterances" are not "produced and understood 'by analogy' to sentences one
has already heard and understood", that from these sentences one cannot draw the conclusion that both
14a John is likely to leave.
b* John is probable to leave.
are grammatical, since 14b obviously is not.

Page 146
Smith and Wilson are quite right about the uselessness of this sort of analogy. It is hard to see why anybody should propose it as
a method of learning a language. There is no logical connection of any kind between the sentences of 13 and the 14 pair. The
'general' meanings are quite different in the two pairs. The irony is that 14a and b are excellent examples for the purpose of
showing the very opposite of what Smith and Wilson want to show. They wish to demonstrate that 'novel uterances' can only be
produced and understood (and judged grammatical or not) by people who have an innate unconscious knowledge of the
principles or rules of correct sentence formation. Well, even if people do ever have such knowledge, it won't help them here.
There is only one way to find out how people use "likely" and "probable" and that is to observe in detail how they do! Naturally,
there is a lot more than just ''likely" and "probable" to be mastered:
15 John is sure, certain, bound etc. to leave.
16* John is definite, clear, possible etc. to leave.
Further, as I pointed out in Chapter 5, some languages use a particular kind of 'general' meaning (such as this one) and some do
not. One does not know one can use it until one has heard it used. And in the languages where one can, apparently equivalent
'specific' meanings will fit the 'general' meaning in one language and not in another. There is much to observe about what
exactly each word means and how exactly it is used, and much to remember.
At the same time, though, notice that meanings are often in lumps of words together, and when one learns one's own language
one imitates these lumps as a whole and doesn't demand the same logic within the lumps that one follows otherwise. For
instance, "had better do", "It's me", "No wonder you're cold". Particularly interesting are examples such as the 'illiterate'
17 You should of told me.
18 It's too bigger job to do on my own.
It is probably only 'education' that makes it unlikely that this ."of" of 17 will be adopted as usage by all, like "had better". To
those who say it - "should of" - and they do say it, not just think they say it - it is just as grammatical as anything else they say.
The same applies to the "-er" instead of "a" in 18. One can check that some people think they are saying this and effectively are
saying it by the fact that they write it.

Page 147
I think examples such as these tell us something essential about the nature of practically the only things that might justify the
term "rules of grammar". I mentioned these in Chapter 4 (p.67, note). There is, for instance, inversion of subject and auxiliary
verb (or "be") in sentences beginning with 'negatives' or ''so" etc:
19 Only recently has he begun to appreciate how deeply sad Haydn's music essentially is.
The pattern goes further than this. Where there is no auxiliary (or the verb "be") already, one brings one in: "do, does, did":
20 Little do they realize what nonsense they are talking.
If one is to call patterns like these 'rules', what does one mean by the term? The only important question, in fact, is what the
human experience of such 'rules' is; what is the psychology? To begin with, English-learners' must surely again notice that the
change in word order is rooted in meaning, that is, it is meaning (in this case 'negatives' and "such", "so") that 'trigger' the
patterns. But how is "do", "does" or "did" added? I think a phrase like that in 18 - "too bigger job" - gives a clue. There is no
logic within the phrase; not of meaning, and certainly not of any abstract structure. Such a phrase cannot possibly be produced
by any analysis by any structural rules, transformational or otherwise. Yet phrases like this and "should of" etc. are produced. It
is fairly clear, I think, that what happens is that language-learners associate whole lumps of language (which have meaning for
them) with certain other meanings. And so, for example, the meaning "Never" at the beginning of a sentence triggers the whole
lump:
21 ...did he think for a moment that...
The fact that there happens to be logic within "did he think" and not within "too bigger job" is irrelevant at this point.
Psychologically they are exactly the same in principle for the language-learner. And neither are psychologically different from,
say, a single word on its own.
It is true that there is a certain 'arbitrariness' about the way these conventions of patterns are 'imposed' on language learners. But
really only in the same way as all the conventions of a language are, in the end, arbitrary. They are not arbitrary in the way
linguists propose, particularly 'transformational-generative' linguists. That is a proposal of an utterly
1 For the question as to what sort of faculty is needed for this learning, see Chapter 9, pp.188-89.

Page 148
dead arbitrariness, a system of restraints and 'permissions' imposed by a biological program with no anchor in logic or meaning
at any stage.
In actual language and real psychology there is always logic and meaning. An English-speaker will not say "So do I" when she
means "So I do!" And although strictly it belongs to the point I made in an earlier chapter (4) about the harmfulness of the
structural analysis approach to foreign language learning, I cannot refrain from one more example. Rule-bound students never
notice the distinction between
22 You're from Germany, aren't you? and
23 Ah! So you're from Switzerland, are you?
Although Chomskyans may stress variation and flexibility when it serves their distinction between 'competence' and
'performance', the linguistic and non-linguistic, their thought is so often inflexibly unrealistic. For example, Smith and Wilson
(1979, p.110) declare 24 to be an ungrammatical sentence.
24 There seemed to be the otter swimming effortlessly through the water.
They say that it is the "the" that is ungrammatical, and that the possibility of "the" occurring in this position is prevented by
transformational rules stating, among other things, that the subject in this position of a "there" sentence must be indefinite. Now
in fact 24 is not ungrammatical, that is, not illogical. The strangeness of the situation in life of 24 may make it hard for you to
accept its grammaticality. There is undoubtedly a tendency for this sort of "there'' expression to go together with indefiniteness.
That is entirely due to the meaning of "there is, there was," etc. Here once again is an example of the fact that in life there are
sorts of situation that are very common, and others that are not so common, or even very rare. One will find a certain form of
expression rare not because it is linguistically rare. To say that something is linguistically rare is, in a sense, meaningless; it is
only the things language refers to that are rare or not. Take the sentence
25 When we arrived at the village hall there only seemed to be the postman waiting for us.

Page 149
I think most unprejudiced English-speakers would accept that as a grammatical sentence. The linguists may say this is a small
matter. Perhaps they will concede a little mistake and say that all that is needed is a small adjustment to their transformational
rule in order to allow for "the". But this is not really the point. I believe it would be difficult to devise a consistent rule that
produced "the" sorts of ''there" sentences that most native speakers would accept. But even if it can be done, it cannot take away
the insight one gets into the unimaginative thinking of linguists, and into how insensitive they are to the decisive influence of
meaning - in these cases 'specific' meaning.
The Chomskyans want us to believe in a kind of absolute sealed world of its own in the brain which rigidly dictates the
individual's 'real' language, impervious to any outside influence of general faculties of mind; there may be such influences on a
person's actual use of language, they say, but they are outside, they never reach that secret privileged world within, the language
machine, they are essentially nothing to do with it. But this view of language raises so many problems, problems I have raised in
this book so far, with more to come.
I have suggested, in contrast, that words are simply the materials or tools - like a carpenter's wood and tools - to which language
users apply their logic and practical sense together. Such an idea of language surely accounts simply but effectively for the facts.
Among other things this picture accounts easily and clearly for differing judgements on grammaticalness and acceptability, for
those judgements are the results of individuals' subjective and so varying logics, and subjective and varying sense of the
practical.
And as I shall reiterate from another angle in the next chapter, knowledge of language, basic knowledge of language, is
inseparable from knowledge of the world. I have already given several examples at various points in this book of how
recognition of reality is essential for the coherent use of language, how it combines with both 'general' and 'specific' meaning
logic to determine meaning, how it is ultimately the very foundation of that logic, though of course it is not the faculty of logic
itself.
How can people differentiate between, say, "hair" and "a hair" if they do not know what "hair" and "a" refer to in real life? How
can people know whether to use "the" or "a", in utterances such as 26 and 27, and how can they understand such utterances
properly if they do not understand real situations?

Page 150
26 My wife is a nurse. It is useful to have a (cf. the) nurse in the (cf. a) family.
27 We have a cat. I love the (cf. a!) cat.
If linguists insist that descriptions of 'language proper', the rules of grammar, only reveal what is abstractly possible (and
impossible) that is indeed an admission of abstraction totally irrelevant to what language actually is, how it actually works.
I emphasize again that in my own account I am not defining the new linguistics. May we be preserved from any new linguistics,
as from the old and the current. But I see particular danger in Chomskyan linguistics. It pretends to reveal a kind of linguistic
divinity, immutable, the biological genius of humanity. If that is really how it is, then we are all helpless prisoners of the
language program. We are already in practice slaves to language. But if language is such an inescapable principle of human
brain, then there is no hope for us, we are a doomed race and doomed individuals.
Of course something is not untrue just because it is frightful. For all I know most of us are destined for eternal torment in a
Hieronymus Bosch hell, but it would be tragically absurd to live our lives on that basis when there is no evidence that such a
hell exists. It would in practice have even worse consequences if we persisted in believing in the unchallengeable authority of a
non-existent language gene.

Page 151
8
Deep Confusion
1 John is eager to please.
2 John is easy to please.
This pair of sentences is an example of Chomsky's, and has become very famous among linguists. Chomsky and the linguists
who follow him in this matter argue as follows: Both sentences have the same 'surface structure'; in traditional linguistic systems
they would get exactly the same analysis. But 'underneath' we are aware that while "John" really is the subject in 1, because he
is doing the pleasing, in 2 "John" is really the object - somebody pleases John, not the other way round. Thus, in 2, they say,
"John" is the 'underlying' or 'logical' or 'deep structure' object, although he is the 'grammatical' or 'surface structure' subject.
Similarly, often where the 'surface structures' are different, as in
3 The cat ate the mouse.
and
4 The mouse was eaten by the cat.
we know, they say, that the 'deep structures' are the same, the cat is the 'real' or 'logical' subject and the mouse is the 'real' or
'logical' object in both sentences, the relationship between cat and mouse is the same in both. So, Chomskyans maintain, we
know a grammar that does not appear directly in the sentences we actually hear and use, we know a grammar that lies beneath
the surface, and we also know rules for transforming 'deep structures' into 'surface structures' and back into 'deep structures'.
Chomsky, it is said, has drawn attention to the inner mental processes which alone can explain our competence to speak and
understand language. And a traditional analysis of the 'surface grammar', classifying the various parts of a sentence and
distributing them into various constituent groups or phrases, is completely inadequate, cannot explain the fundamental distinction
between 1 and 2 or the link between 3 and 4. A concept of 'deep grammar' has to be invoked; 'structuralist' (pre-
transformationalist) accounts of language have to be drastically revised.
Consider first some comparatively superficial, though nonetheless decisive, objections to this argument.
"John" is not the real or 'logical' subject in 2 ("John is easy to please"), they say; we are all aware he is really the object. So
suddenly they give us

Page 152
new definitions of subject and object. Subjects are now 'do-ers' and objects are 'done-to's'.
Actually, like most things, subjects and objects are impossible to define, and, as with most things, there is no point anyway in
doing so. (Dictionaries tend to say that a subject is that which has a predicate, and a predicate is that which has a subject.) But
everybody knows what a subject is, even if most people in history haven't known the term. Everybody knows that subjects not
only do things; they also are things, and also have things done to them (passive).1
The Chomskyans suddenly deny this and - for a brief moment and stage in their argument - limit subjects to do-ers, while
anybody or anything that has something done to them is an object. So John in 2 is an object, not a subject, things are not what
they seem. Then, having used this very un-'structural' notion of subject and object for their purpose of establishing the existence
of 'deep structure' they suddenly change their definitions again.
"The 'logical' subject is that NP [Noun Phrase] which is immediately dominated by S (=Sentence) in the deep structure;
the 'grammatical' subject is the leftmost NP which is immediately dominated by (the topmost) S in the surface structure."
(Lyons, 1977, p.81)
Furthermore, if they deny that they are playing around with practical notions of 'subject' when it suits their argument (although I
do not see how they can) they are in effect saying that there is a difference between 'surface' structure and 'deep' structure
because, for example, there is a difference between 'surface' structure subjects and 'deep' structure subjects!
But even if one ignores these illogicalities and accepts the existence of 'deep structure' for the sake of further argument, the
whole contraption is full of difficulties. For instance, what is the 'deep structure' analysis of 5 and 6?
5 That sound is beautiful.
6 That sound is frightening.
I cannot see that they can call "That sound" anything but their 'logical' subject in both sentences. If it isn't, what is it? But what
about 5a?
1 Some may feel that this amounts in fact to saying there is a universal grammar (see Chapter 9), at least as far as subjects
are concerned, a universal sense of the 'grammar' of the subject. But 'universal grammar' is supposed to be at the 'deep
structure' level, and it is precisely at the 'surface', in language as it actually is, that people are aware of subjects. As to why
everybody knows what a subject is, though, I have to say I do not know. Perhaps the idea of subject is a universal one,
based on human logic; or perhaps the practical limitations of language (see Chapter 10) make subjects necessary.

Page 153
5a That sound is beautiful to hear.
Is "That sound" now the 'logical' object? For here the logic of the sentence is exactly the same as in 2 ("John is easy to please").
But "to hear" or "to listen to" or something of the sort is inevitably implied in 5. So what is its 'deep structure'? On the other
hand, a frightening sound obviously frightens somebody. So perhaps it is unquestionably the 'logical' subject in 6. But that won't
do either, because how about 6a?
6a That sound is frightening to hear.
The same confusion arises with
7 That dog is eager.
and
8 That puzzle is easy.
One would think that "dog" and "puzzle" must both be 'logical' subjects. But perhaps not, after all, because the dog wants to do
something, while someone else does something to the puzzle, so "puzzle" must be the 'logical' object.
There are several very peculiar things about 'deep structure'. For instance, by Chomskyans' own affirmation, 'deep structures' are
arrived at by analysis of 'surface structures'. Not analysis of all types of 'surface structure', certainly, but of the "basic
combinations of syntactic categories...permissable within the sentence" (Smith and Wilson, 1979, p.97), and of course the
analysis of such 'basic combinations' (the 'phrase structure' rules) can only be based on language that actually exists, that is,
'surface structure'. How do they know these features of 'surface structure' are the 'basic' ones that appear in 'deep structure'? How
do they recognize the 'deep' from the 'surface' structure? What appears to make it all even more of a muddle is that some of the
specimens of 'deep structure' are perfectly normal, or correct, but others are not. For instance
"an elephant in the house"
but also
"that the heron eats frogs seems" (Smith and Wilson, 1979, p. 108-109).
At one stage at least Chomsky maintained that "the meaning of each sentence is derived...from its deep structure, by means of
the rules of

Page 154
semantic interpretation"; the "deep structure relations...are essential for the correct semantic interpretation of the sentence"
(Lyons, 1977, pp.80, 82). As we saw above, we are supposed to know the 'deep structures' because we know the rules for
transforming 'surface structures' back into 'deep structures' - as well as forwards from 'deep' into 'surface' structures.
For a start, it is difficult to understand why humans should go to all the trouble of turning 'deep structures' - which the
Chomskyans say are what we need for understanding anyway - into something else; and then back again.
And why should it be more difficult to understand the language as we really have it (their 'surface structure') than the hidden
language it is supposed to imply? It seems very odd that we should not understand directly what we hear, because we are not
equipped to do that; we cannot understand unless we know a whole set of transformational rules for turning it into something
else. It is surely more reasonable to suppose that it is far easier and more natural to learn to understand the direct connection of
the words with reality than to convert one language, by means of immensely complicated rules, into another language and then
convert that into reality.
Moreover, what determines the exact word order of 'deep structure'? For, as we have already seen, a particular precise word
order may be crucial according to Chomskyans. I cited an example of this in Chapter 3 in connection with the contraction of is
to 's. Their (false) rule depended on whether or not the wh-word followed immediately after is in the 'deep structure'. Do they
pretend that such, as they claim, significant word orders are universal genetically determined principles?
And what is it that tells the listener, in order to arrive at different 'deep structures', to apply a different set of transformational
rules to "John is easy..." and "John is eager to please"? Whatever else they say, Chomskyans too must affirm that it is the
difference between "easy" and "eager". What really happens is once more very simple and obvious, and is known by everybody
from childhood onwards. The moment people hear the word ''eager" they recognize a quality that tends towards doing
something; immediately they hear "easy" they recognize a quite different quality, a passive quality, and that whatever is said to
be easy is something that is easy for a person to perform an action on. I have suggested an example of this in 8 above: "That
puzzle is easy (to do, to solve, etc.)". To start talking about 'deep structure' is to invent pure fantasy.

Page 155
Put this obvious truth another way. We know something is done to John only because the meaning of "easy" tells us so; not the
other way round - it isn't any 'structure' that tells us what "easy" means! Nor does the difference in meaning of "easy" and
''eager" tell us anything about any structures1 - it simply tells us about a difference in the situation in life. These systems of
structure exist nowhere except in the minds of their creators and believers.
Linguists get into this tangle because they are trapped in a 'structural' vicious circle that prevents them seeing, or at least makes
them forget, the essential principle of language: the recognition (albeit imperfect recognition) of realities in life.
That recognition is not always so easy as it is when there are words like "eager" and "easy" to help one.
9 Kangaroos are best.
One doesn't know what this means unless one knows what matter, in real life, is being talked about. That is, only the situation
tells one what "best" means, and whether the kangaroos are being 'active' or 'passive'. The sentence may mean
9a Kangaroos are best for eating.
or
9b Kangaroos are best at jumping.
and thus for the deep structure addicts has two contradictory 'logical' structures at the same time, both that of 1 and that of 2
(above p.151). Once again they have it the wrong way round; no structure, deep or otherwise, can tell one what it actually
means.
Recall, if you are not tired of it by now, the sentence
The Chomskyans' 'deep structure' arrived at by our inner mastery of 'transformations' should interpret that rapidly and elegantly
for us. But of course it doesn't, because the sentence does not reflect any sane experience of life, and that 'deep structure' doesn't
exist.
1 But again only understanding of the meaning makes it possible for the linguists to construct their artificial structures.
2 See above, Chapter 3, p.53, Chapter 6, p.119, and Chapter 7, p.137.

Page 156
Why can the linguists not accept all the conventions of language as they are, at their perfectly clear face value? All meanings are
pure convention; the linguists themselves accept vast numbers of them without question, presumably without thought. Why
should they pick out just some of them and pretend they are obscure and difficult - and therefore need an infinitely more
obscure and difficult system to interpret them?
By 1975 Chomskyans had apparently abandoned the idea that "deep structure relations" were "essential for the correct semantic
interpretation of the sentence." In that year (see Reflections on language, 1976, pp.83-84) Chomsky wrote:
"Suppose it is a fact, as I now tend to believe, that a suitably enriched notion of surface structure suffices to determine the
meaning of sentences under interpretive rules...It may still be the case - I think it is - that initial phrase markers [the term
Chomsky now prefers to 'deep structure'] generated by the base have significant and revealing properties. It also remains
true that they enter, though now indirectly, into determining the structures that undergo semantic interpretation, and that
they play a role in the theory of performance. The thesis of surface-structure interpretation, if true, would constitute a
significant empirical discovery, which would in no way diminish the general interest of the results of linguistic inquiry.
Under this thesis, the structures called "deep structures" in the standard theory do not play the role formerly attributed to
them, but we may still agree with Strawson that "the central thesis of the transformational grammarians, the step which
conditions the whole character of their theories, is the insistence that any adequate grammatical theory must recognize a
distinction between the superficial syntactic structure of a sentence and its basic structure, between its deep grammar and
its surface grammar''. But we must now understand the terms "basic structure" and "deep grammar" to refer to
nonsuperficial aspects of surface structure, the rules that generate surface structures, the abstract level of initial phrase
markers, the principles that govern the organization of grammar and that relate surface structure to semantic
representations, and so on."
But if 'deep structure' is no longer necessary for understanding the 'surface structure' I do not see how Chomsky can continue to
claim that it even exists - at least, not on the basis of the "John is eager/easy to please" argument, which rests entirely on
meaning.

Page 157
The quotation above is most important, though, I think, because it naively reveals the unpractical, unreal, irrelevant abstract
intellectual games that linguists are playing. Does it not worry Chomsky that he is constantly shifting his ground as he finds
fresh awkward lumps that won't fit his concoction?1
The basic flaw in the argument for 'deep structure' is very simple, though.
Linguists have mixed up two quite different things; they think they are the same. They have confused meaning with what
actually happens. What they call 'deep structure' is simply what happens.
Meaning includes what happens - actual events. But it covers much more; it includes how people react to what happens, their
feelings about events. I have already pointed out that, for instance,
I have been to London.
and
I went to London.
refer to exactly the same event, but mean two completely different things: I am looking at the event from two different angles.2
The linguists' 'deep structure' analysis is not an analysis of language -it is merely an analysis of what actually happens. It has no
bearing of any kind on language, 'structural' or otherwise.3
2 John is easy to please.
does not mean the same as any 'deep structural' account of that 'event' or any other account, whatever it may be. It does not
mean the same as
11 It is easy to please John.
or
12 Pleasing John is easy.
1 Since I wrote this I have heard Chomsky saying on British radio that he would be "devastated" if it was discovered that
his hypotheses were in fact correct. He apparently feels it is against the spirit and fun of science to be right first time.
2 Chapter 1, pp.5-6.
3 And even if it did, one would still want to know why it was 'transformed' into 'surface structure' in all the particular ways it
is supposed to be in all the different languages. The 'transformations', even if true, would tell us nothing interesting, would
not explain the many varying conventions of the actual 'surface structure' we have. Why do the Japanese, for instance, say
the equivalent of "John is easy to be pleased"?

Page 158
or
13 To please John is easy.
or
14 John is easily pleased.
or
15 People can please John easily. etc.
All these different sentences have different meanings; they attempt to describe not just the fact behind "John is easy to please"
but also different attitudes to that fact. The deep structuralists are only commenting on the fact itself.
It is this confusion of fact with attitude to fact that leads them to make the quite untrue assertion, for instance, that the passive
means the same as the active. The fact behind both
16 I have cut my finger!
and
17 My finger has been cut by me!
is undoubtedly the same. They hardly mean the same. Again, if one asks "What happened to his leg?" the replies 18 and 19 do
not mean the same, and neither do the statements 20/21.
18 The ground broke it in a fall when he was six.
19 It was broken in a fall when he was six.
20 The airmen burned thousands of people alive in Dresden.
21 Thousands of people were burned alive in Dresden.
In the 'deep structure' system it is impossible to understand why passives should ever have been invented in any language. But if
one looks at language in the obvious, simple way that comes naturally to the uncorrupted it is easy to see why. This is not to say
that the use of passive rather than active is always significant; but there is always a difference in emphasis, sometimes important,
sometimes not.
22 There was nothing to do. (So we got bored.)
23 There was nothing to be done. (The situation was hopeless.)
Once more the meanings are quite different although we did nothing in both cases. It is not difficult to think of example after
example where there

Page 159
is no correspondence between active and passive, and what is interesting is to see how it is not some fixed principle that
separates them, but the varying realities of different situations in life that produce varying distinctions between the two forms.
24 We are expecting the result at any moment.
The passive way of expressing this is not
25 The result is being expected at any moment.
but
26 The result is expected at any moment.
The confusion of actual event with meaning leads 'deep structuralists' into so many difficulties. Lyons (1977, p.116) writes:
"...The book the man left is on the table. Here we have one sentence, The man left the book (strictly speaking the string
underlying this sentence) embedded in the middle of The book is on the table (and subjected to a variety of other
operations, including the deletion of the book in the embedded relative clause)."
This cannot be true. If there is an 'embedded' sentence (a 'deep structure' sentence?) it must be "the man left A book", because
"the" here would mean the book which is already limited to a particular one, which it isn't. The "the" of "book" is brought in
precisely by the relative clause, and only because and when we have the relative. That ''the" is not derived from any 'deep
structure' or 'embedded' sentence. It is 'derived' direct from the reality - the reality of a relative - that the speaker wishes to
express. There is no 'deep structure', no 'generative grammar', no 'transformation', no 'surface structure'. There is simply
meaning.
Perhaps they will say there is a little mistake here. The embedded sentence should indeed really be "The man left a book". This
won't do, because it may be that the speaker in fact wants to say "A book the man left is on the table". It then becomes purely
arbitrary whether they transform "a" into "the" or keep it as "a". Can they not see that even if one assumes that there are such
things as 'deep structures' and 'transformations', the right transformation must depend on the end product itself- quite simply,
does one mean "a" or "the"? In other words, one does not know the correct transformation until one knows what one is
transforming the sentence into. This is a circular absurdity. And if that does not worry them, and they say that here, for instance,
"a" or "the" are optional transformations, and that

Page 160
they are simply concerned with constructing a formal system of operations that can fit everything into itself (according to its
own rules), then that is simply a further confirmation that they are merely playing a formalistic game, with no bearing on reality.
Another example of the same mistake, in principle, is where they say that in a sentence like
27 John was persuaded by Harry to take up golf.
the deep structure is one sentence (S2)
27a John take up golf
embedded in another (S1)
27b Harry persuade John
Figure 1
"It will be seen that the logical subject of S1 (the matrix sentence) is Harry, and that of S2 (the embedded sentence) it is
John. Furthermore,

Page 161
the deep structure subject of S2 is identical with the deep structure object of S1 (that NP which is immediately dominated
by VP)." (Lyons, 1977, pp.81-82)
The flaw in this 'deep structure' is that "persuade", if it governs "John" alone, as object only, is meaningless - ''persuade" is used
precisely to mean that somebody persuades somebody else to do something. John has got to be object and subject at the same
time. The essential meaning of 27 (or of "Harry persuaded John to take up golf") is lost in the 'deep structure' version.
In a chapter called "Semantics and Meaning" Smith and Wilson (1979, pp. 148-49) give the following examples:
"(1) My son threw a brick at the window.
(2) a Someone threw a brick at the window.
b My son threw something at the window.
c I have a son.
d I have a child.
e My child threw a brick.
f My son did something.
g My child threw a brick at something.
h Someone threw a brick at something.
i Someone's son threw a brick at the window.
j Something happened."
Smith and Wilson say that (1) logically entails (2)a-j. "We may regard the meaning of a sentence as a set of propositions." And
"...part of the meaning of (1) is expressed by the propositions represented in (2a-j)..."
If this is true, then (1) also means, among many, many other things,
I am a parent.
A brick was available to an offspring of mine.
The window had a hard object propelled in its direction.
And further,
A male human, either in childhood or mature, the result of reproductive processes engaged in by me and another of the
opposite sex, having taken possession of a more or less solid object, the product of industrial activity normally utilized for
the construction of buildings,

Page 162
swung it in one of his hands through the agency of the arm that that one of his hands was at the extremity of, and then
released it in such a way that it travelled quite rapidly through the gas forming a large part of the normal environment, and
with the intention that it should strike a sheet of glass or other transparent material (also a product of industrial activity)
fitted vertically (or perhaps not) in an aperture in an edifice, probably for the purpose of letting in light but not cold, the
said sheet being a particular one that you, the listener, are supposed to know the identity of.
If you doubt whether all this, and much more, is involved by the argument of 'entailment' of propositions, on the grounds that
the information that I have suggested is inferred from (1) depends on knowledge independent of the sentence itself, then
consider that the information in (2)d and (2)e depends in exactly the same way on such independent knowledge. It may be very
basic, crude knowledge that practically everybody has, but one cannot understand that "son" indicates "child" unless one knows
that a son is a child! In other words, one cannot infer anything from any piece of language unless one knows what it means. A
person with more knowledge than another will obviously infer more. An eighteen-month-old child may associate the right things
with "window" or "pen'', but a six-year-old will have a much fuller and more precise appreciation of these words. A six-year-
old will almost certainly associate a pen with writing and a window with glass; an eighteen-month-old will probably not. It may
be argued that "son" involves "child" by definition. A "brick" equally involves by definition what I have suggested, although it
may be a more complicated definition that requires greater knowledge of the world.
Yet of course (1) does not mean all these things at all. What is perhaps even more important is that (2)a ("Someone threw a
brick at the window."), for instance, is not part of the meaning of (1) ("My son threw a brick at the window."). (1) and (2)a may
share certain events; but (1) does not include the meaning of (2)a, which is a whole new meaning in its own right. "Someone" in
place of "My son" produces a sentence that expresses a completely different idea. It implies I do not know who threw the brick,
or, if I do, I'm not telling you.
Smith and Wilson's (1979) chapter on semantics and meaning is a good example of how linguists tie themselves in knots trying
to make all the parts of their analyses fit together. The particular chapter I have mentioned reaches a climax when the writers
arrive at the sentences

Page 163
28 All parents are parents.
29 If John and Bill are here, then John is here.
According to the 'entailment theory' they have been explaining, "definable in logic", 28 and 29 are synonymous. Clearly they are
not, so the authors have set about contriving another set of operations to find out what the sentences mean. All the time they are
unaware of the essential truth: meaning and happening are not the same thing.
To be fair to the linguists, this seems to be a case where they are not corrupting non-linguists; non-linguists may have corrupted
them. Here linguists appear to be following those philosophers who have got hold of the idea that meaning, 'semantics', is a
matter of formal logic. Now, philosophers and linguists together, they confuse meaning with the situations or happenings that
follow from other situations or happenings. They recognize what they are doing by using such terms as "entailment", but believe
that that has something to do with meaning. Much of their logical calculation (I am not saying that their calculations are all
wrong) is based on the assumption that certain statements are true. But true or untrue has nothing whatever to do with either
meaning in particular, or language in general. True or untrue is, again, to do with facts, with situations or events themselves.
Meaning is words giving people recognition of, impressions of, reminders of things. Each person probably has their own slightly
different individual associations with most of the various words of their language; meaning is subjective. Nothing can be
concluded from meaning.
I have more recently come across another good example of this confusion of meaning with logic - formal logic, that is - in a
review of a book by J. Barwise and J.Perry1 who are said to have created the subject of "situation semantics", a confusion by its
very name. The utterance 1, "John kissed Mary", is said to "strongly entail" utterance 2, "John touched Mary'', since the set of
situations described by U1 is a sub-collection of the set of situations described by U2. This is undoubtedly and very obviously
true. John can't kiss Mary without touching her. But it has nothing to do with meaning. When people hear utterance 1 they will
almost certainly get the impression of somebody using his mouth; many people, but not all, will get the impression of mouth on
mouth. When people hear utterance 2, most of them will almost certainly get the impression of someone using his hand, not his
mouth. This is meaning, nothing to do with formal logic. Language
1Situations and attitudes, 1983, reviewed in Computer Weekly, March 1, 1984.

Page 164
is only what is used between or by people to refer to the things that logic is about.
It is odd that there should ever have been such a confusion. It may be yet another instance of that corrupting influence of
language itself which I try to make plain in a later chapter.1
An extreme example of the linguists confusing events with meanings is a development of the transformational system called
'case' grammar, an "early radical suggestion about the nature of the base component of a grammar" made by Charles Fillmore.
"By 'case', Fillmore did not mean the inflectional variations in nouns, which is the traditional sense of the term. Rather he
was referring to a set of concepts which identify types of judgements which human beings are capable of making about
the events going on around them (for example, who did something, who did it happen to, what got changed, where did it
happen, and so on). For example, in the sentences which follow, it is intuitively obvious that basically the same meaning
is being expressed in each sentence, but that the surface-structure relations are extremely variable.
(a) John opened the door with the key.
(b) The key opened the door.
(c) The door opened.
(d) John used the key to open the door.
(e) It was the key that opened the door.
(f) It was John who opened the door with the key.
1 This confusion may also be connected with the mistake that grammarians sometimes make of confusing the information
we get from a situation with the information we get from language. For instance, it has been repeated constantly that the
'present perfect continuous' in English indicates that the action is still going on at the time of speaking. This is not so, and
it is easy to show that it is not so. One may ask Bill as one sits down to supper:
(a) "What have you been doing this afternoon?"
"Tve been playing with Tom."
Clearly Bill is not still playing with Tom, but is sitting down to supper. On the other hand one has to say
(b) He has made (NOT: has been making) a considerable contribution to the stock of human myths and has not stopped yet.
This is because the continuous form, as in (a), expresses the activity itself, while the simple form, as in (b), expresses the
results of the action.
Grammarians have often been deceived, I suspect, by the situation in which the continuous form is used. For example, when
somebody says
(c) Look how awful! It's been raining for two days.
far more often than not it is actually still raining, and the grammarian confuses this fact with the meaning of the verb form.
Certainly (c) means it has been raining until recently. But that meaning of 'recent' is provided by the "has", not by the "-ing".
Compare:
(d) It has rained for two days, and the ground is still as hard as a rock.

Page 165
In (a)-(c), the key, John, and the door are all surface subjects of the verb open; but their deep relationship to the verb is by
no means the same in each sentence. In (a), the subject position is occupied by the 'agent', or 'actor', John; in (b), it is
occupied by the 'instrument' whereby the action was performed, the key; and in (c), it is occupied by the 'goal' of the
action, the 'thing which was opened', namely, the door. Putting this another way, the function of the underlying meanings
in relation to the verb does not change from sentence to sentence, despite the surface differences; and it is this
fundamental, semantic identity which is the important thing to recognize about these sentences, and the central fact which
a system of grammatical analysis should explain. It is always the key which is opening the door, and never the door which
is opening the key, for instance. It is always John who is performing the action (whenever he is mentioned). And so on.
The phrase key, that is to say, has the same underlying function in each sentence - Fillmore would call it an 'instrumental'
function - despite the variant surface realizations this notion has (in sentence (b) it is the 'subject', in (d) it is 'object', in (e)
it is 'complement', in (a) it is an adverbial phrase). In the same way, one could argue that John was 'agentive' in each
sentence in which it occurred. These 'meaning-relations', of agentive, instrumental, and the like, are what Fillmore called
his cases, and he hypothesized that they constitute a universal system in deep structure." (Crystal, 1981, pp.236-37)
All that this amounts to saying is that however we say something happens, that thing still happens. It has no bearing on
language. The examples given are just another demonstration of how humans use language to express different meanings about
one and the same event. But 'case grammarians' are going to another extreme as well; their work is an extreme example of
pointless analysis.
Another system connected with 'transformational' grammar, called 'generative semantics', needs even less attention. It merely
carries the confusion between 'event' and meaning a stage further. At least in the 'deep structure' of Chomsky's 'standard theory'
the so called 'terminal elements' are words like "kill" or "man"; but in 'generative semantics' they are "semantic primes like
cause, come about, not and existent or male, adult and human"; actual words are ''inserted into phrase markers by a
transformational rule of substitution".

Page 166
But the linguists' misunderstanding leads to what is probably one of the biggest mistakes at the heart of the generative and
transformational system:
"...transformational grammar reflects better the 'intuitions' of the native speaker and is semantically more 'revealing' than
phrase structure grammar. We can illustrate the deficiencies of phrase structure grammar from this point of view with
reference to the generation of corresponding active and passive sentences in English: e.g. The man hit the ball and The
ball was hit by the man. We have already seen how active sentences might be generated in a phrase structure grammar,
and we could easily add further rules to the system in order to generate passive sentences. What we cannot represent
within the framework of a phrase structure grammar, however, is the fact (and let us grant that it is a fact) that pairs of
sentences like The man hit the ball and The ball was hit by the man are 'felt' by native speakers to be related, or 'belong'
together in some way, and have the same, or a very similar meaning. ...this relationship between corresponding active and
passive sentences, as well as many other 'intuitive' and semantic relationships, can be accounted for in a transformational
grammar." (Lyons, 1977, pp.62-63)
"It does not take much thought to see that there are many important grammatical relationships which could never be
brought to light by IC [immediate constituent] techniques or which are taken for granted by them. The kinds and degrees
of relationship between active and passive sentences, for instance. Part of the reason is that IC analysis proceeds one
sentence at a time. It can suggest one analysis for That man saw John's mother; it can suggest another for John's mother
was seen by that man. But how does IC analysis tell us that these two sentences are closely related, that one sentence is
'the active of the other - that in this case they both mean the same thing? It cannot provide this information. All the
analysis tells us about is the structure of the two sentences, seen as isolates. And yet, it can be argued (and all linguists
would argue) it is important for a grammar to tell us about their relationship as well. Knowing a language means, amongst
other things, knowing when there are two ways of expressing the same thing grammatically. It means being able to
manipulate the rules of the language, so that we do not have to learn from scratch, but rather apply an already learned rule
to fresh material. Once I'have 'learned'

Page 167
that two such sentences as the above are structurally related, I can apply my knowledge and produce fresh pairs of
sentences. Given the sentence Seventeen elephants trampled on my country house the other day, I 'know' that it is possible
to say My country house was trampled on by seventeen elephants the other day. And the only way in which we can
explain how I know this is by postulating that I have learned a rule which tells me how to form passive sentences from
active ones, and vice versa. This rule, then, is a feature of the language's structure which any grammatical analysis should
make explicit...We can make a very strong case indeed for saying that the whole of language is a network of mutually
defining structural relationships - relationships between sounds, as we have seen, between morphemes, between words,
and between sentences. Similar examples in grammar to the active/passive one would be the relationship between positive
and negative sentences, between statements and questions, or, less obviously, between attributive and predicative
adjectives (cf. the good man and the man is good)." (Crystal, 1981, pp.206-207)
"The second component [of a generative grammar] consisted of transformational rules - rules which operated on the
strings produced by the phrase structure component, and altered them in various ways (e.g. by turning 'active' strings into
'passive' ones, by altering word order, by adding inflections, and so on), making the various relationships between
different types of sentence explicit. The passive transformation, for example, altered the order of elements in the active
sentence and added three further elements (a form of the verb be, in the appropriate tense; a particle by, to indicate the
agent following; and a past participle affix, symbolized as en, attached to the main verb, V). One formulation of this rule
could be as follows: NP1 + Aux + V + NP2 + NP2 + Aux + be + en + V + by + NP1. NP1 stands for the first noun
phrase in an active sentence; NP2 for the second. Aux stands for auxiliary verb. The rule says in effect: 'to form a passive
sentence, reverse the positions of NP1 and NP2, and introduce the be in one of its forms, a past participle affix, and a
particle by between V and the NP1'. (Exactly which form of be is required, and how the affix gets transferred from its
place in front of the verb to its place at the end of the verb, would be the subject matter of later rules in the grammar.) As
it stands, however, this rule goes a long way towards formally relating such sentences as The boy will kick the ball and
The ball will

Page 168
be kicked by the boy. The way in which these transformations were understood and formalized was a major contribution to
linguistics..." (Crystal, 1981, p.228)
This talk of the 'derivation' or 'association' of sentences really does show how muddled they are. The whole idea of linguistic
connections is of course false. There is no language connection between "The cat ate the mouse" and "The mouse was eaten by
the cat". The only 'connection' - if it is right to talk about connection at all - is that both sentences refer to the same event. So
however that event is appreciated there are likely to be references to a mouse, a cat, and eating. That obvious fact does not
demonstrate any language associations, derivations or processes of any kind.
Active and passive sentences are said to be 'felt' by native speakers to be related. I very much doubt if any unconditioned native
speakers feel anything of the sort. Innocent native speakers will not be aware of any sentence connection at all - but rather
naturally if one asks them what the connection is they will say that the sentences are about the same thing - if they actually are.
But what would they say about any connection between 30 and 31?
30 The cat ate the mouse.
31 My bicycle was mended by a friend.
It is also absurd to pretend great revelation of significant deep relationships by talking about 'learning' that two sentences are
structurally related so that afterwards if given the sentence "Seventeen elephants etc.", I 'know' I can say "My country house
was trampled on by seventeen elephants". In the first place, what is this 'given'? What does it mean? Does it mean that the
whole elaborate mental process cannot start until one is 'given' some sort of cue? And then is one unable to produce a passive
version of the elephant trampling event before one has heard the active? Certainly it appears that not only does one have to
know the active before one can form the passive but one has to know the passive to form the active - "vice versa" and "mutually
defining structural relationships'' says Crystal. This is silly, because it is obviously impossible in any psychological sense, which
is the only sense of any importance.
The truth is simpler. If one wants to express something 'passively' one just uses passive meaning! One 'applies' it direct to
whatever one wants to apply it to. One chooses the passive meaning in exactly the same way as one might choose "cold", say,
instead of "hot". And one has of course learned

Page 169
passive meaning in exactly the same way - direct - that one learned active meaning. It may well be that children learn active
before passive, but that does not mean that the one has some profound structural relationship with the other that demands some
sort of 'explanation'.
Naturally many forms are based on other forms. It would be very odd if it were not normally so when one wants to talk about
the same idea. It is not surprising and not a bit significant that in English, for example, plurals are accomplished by adding -s to
the singular, adverbs by adding -ly to adjectives, negative adjectives often by prefixing un- and so on. These are all on exactly
the same principle as the cases of 'structural relationship' that the linguists see as so significant.1 To call the understanding and
formalization of "these transformations...a major contribution to linguistics" is to expose linguistics for what it is: mere
'accounting for' code-devising dressed up as profound insight.
Such false principles as those I have quoted above, and the whole 'deep structure' confusion, are the result of seeing language as
an independent self-contained, self-supporting system that has to be explained with reference to nothing but itself, and of the
belief that the concoction of analysis has in itself great things to offer to understanding.
1 Sometimes the change in meaning - because that is all it is, a shift in 'general' meaning - is achieved very neatly, as in
Scandinavian passives, which merely add -s to the active form. But often the meaning is changed by using quite different
base forms, that is to say, different base meanings. "You saw" becomes "Did you see?" or "You were seen", for instance.
In Welsh the lack of connection in form between meaning variations can be even more striking.

Page 170
9
Being Able to Use Language
There can be no argument, I think, about at least some of the things that human beings need to be able to do in order to use a
language.
They must be aware of many different, separate parts of reality, things inside them as well as outside; they must be able to hear
and distinguish different sounds; and they must be able to understand the connection between the different sounds and the
different parts of reality - that is, they must understand that words mean things, and which words mean what. They must
understand that certain sounds made by humans have this connection while others, made by humans or not, do not (many
animals can do this a bit too). They must be able to remember all the different meanings, which means storing all the sounds
somewhere, with their connections, in a form which is not sound. They must be able to realize that if a word means something
in one situation it can be used to mean the same thing in another situation. They must be able to see patterns in the meanings of
sounds, such as that if "drop" and "look" and "reach" all just add a "t'' sound when they happened in the past, then "stop" and
"teach" ought to do the same - even if it doesn't always in fact work, and they have to notice when it doesn't work; and they
have to notice other patterns like "a potato, an apple" but "an old potato, a big apple'? They must notice the meaning of the
order in which the sounds are made. They must be able to carry out the whole process of getting the 'right' sounds, in non-sound
form, out of the memory, arranging them in the right order, and turning them into the right sounds in the mouth, and do much of
this in reverse when they listen and understand.
In their view of the ability to use language Chomskyans appear to give these facts little or no importance. They say that human
languages all have a lot in common (which is not strange), that children go through the same stages in learning whatever
language they do learn, and that this is because both the types of language they are able to learn and the form of their linguistic
development are restricted by biological inheritance. There is a 'universal grammar' known by all humans when they are born.
Language, Chomskyans say, is a separate and special faculty,
"different in kind from other cognitive systems, requiring different learning strategies and different genetic
programming...There are a number of rather obvious points that support the special-programming view of language
acquisition, and disconfirm the
1 See p.188.

Page 171
general-intellectual-ability approach. If we measure general intellectual development in terms of logical, mathematical and
abstract-reasoning powers, these powers are still increasing at puberty, when the ability to acquire native fluency in a
language is decreasing rapidly. A child of eight who can beat an eighteen-year-old at chess is something of a prodigy; if
an eighteen-year-old acquires native fluency in a language as quickly as an eight-year-old, simply by being exposed to it,
and without any formal training,it is the eighteen-year-old, not the eight-year-old, who is the prodigy. If it is thought
unfair to compare linguistic skills with powers of abstract reasoning in this way, the point has already been granted: there
is a difference between mathematical and linguistic abilities: linguistic know]edge can be distinguished from other types
of knowledge, which depend on different intellectual endowments, and are acquired at different rates." (Smith and Wilson,
1979, pp.33-34)
I have already shown, in Chapter 3, how much of the language evidence Chomskyans use to support their view is false. Here I
want to consider more general matters.
It is, first, strange that just the language program should 'wear out' so early, while most other faculties improve or at least remain
perfectly adequate for far longer than eighteen years. Moreover, if there is a special language program, and it wears out, one
would expect the whole language faculty to collapse and people to start finding in their teens that they were losing their powers
of speech - just as in fact happens to some people who suffer brain damage. This alone suggests that language does not depend
on a special program of its own but on other faculties.
At the same time it is well known that the ability to learn various skills does decline after childhood. It is not just language that
this decline affects. Playing a musical instrument, swimming, or driving a car, for instance, are better learnt early in life. If it
gets more difficult to master a language as one grows older, that is support for the idea that one has to learn language too.
But I suspect the main reason somebody of eighteen finds it more difficult than someone of eight to learn a new language is
precisely that it is new. An eighteen-year-old has a language already, and it gets in the way. People who had experience of the
problem and thought about it knew this a long time ago, but it has not been a fashionable explanation recently. Yet anybody
who has much experience of trying to teach a foreign language and is not corrupted by modern linguistics knows that it is so.

Page 172
Eight year olds have a language already too, of course. But the habit of the language is far more ingrained for eighteen-year-
olds. It is surely notorious that nearly everybody finds it more and more difficult as they get older to learn a new method of
doing something they have got used to doing another way; and this is just what learning a new language is.
I have observed closely the studies in English as a foreign language of several thousand adult students, from many different
mother tongue communities. Again, there can be no argument about what happens. The chief difficulty virtually all students
have is that they tend, in varying degrees, to turn what they hear or read in English into the nearest equivalent they can find in
their own language (but still more or less in English 'dress' of course), whether it is vocabulary, 'general meaning' (grammar),
word order, or pronunciation. "Is there any room?" becomes "Is there (any) place?", "arrive in Paris" becomes "arrive to Paris".
''I have been here for three weeks" becomes "I am here since three weeks" or perhaps "It is three weeks that I am here".
German-speakers, and others, will constantly put adverbs etc. between verbs and their objects since that is what they do in their
own language. Speakers of certain languages (e.g. Japanese) nearly always forget to use "a" or "the" in front of "book", "train",
"person", etc., etc., because they are used to talking about these things without any such preliminaries. Speakers of different
languages divide up reality and experience in different ways for the purpose of their language. German-speakers may constantly
refer to "she" when they really mean "they", simply because "Sie" means both "she" and "they" in German.1 Speakers of
languages belonging to different 'families' probably have more difficulty than average in mastering each other's language
because they have divided the world up even more differently, especially as regards many 'general' meanings basic to their
languages.
Probably ninety-five per cent, or more, of all mistakes in a foreign language are caused by learners' bondage to their own
language; all the rest are the result of individual confusions and misapprehensions. It is all perfectly natural to those who
recognize the truth about language; it is impossible to explain if one supposes that language is programmed. Apart from anything
else, if the difficulty in learning a new language was due to a breakdown in the 'program', mistakes would not be tied to the
native language; everybody would either make the same mistakes, irrespective of
1 German-speakers would also like to say: "Is the man who tall is in the room?"! See above, Chapter 6, p.117.

Page 173
nationality, or would make a mass of 'personal', apparently random mistakes.
But there are other reasons why it is difficult to learn a new language, and why it is more difficult the older people get. To begin
with, the younger a person is, the more 'pure', the more simple, the more basic is the language they are 'exposed' to. The
language a child is exposed to is large in amount but comparatively small in variation and vocabulary; it is largely concrete and
practical - physical objects, visible actions, direct and 'immediate' emotions - and repeated daily. Children get a really solid
foundation of language to build on in this way. Most eighteen-year-olds would never have a chance to experience such
exposure.
Perhaps more important still is that eighteen-year-olds do not have the same urge to get a new language as they had to get their
first one when they were two, or even a new one when they were eight. Eighteen-year-olds, and thirty-eight-year-olds, have a
language; they can already understand a language and express themselves in it. Their world is not going to fall apart, or cease to
function, or even - in most cases - become impoverished if they do not learn a new language. It is a fact - but one that most
people are probably not aware of- that desiring a nicer job, or a rise in salary or more profitable and efficient conduct of one's
business, or wanting to read the literature in the original or enjoy one's holiday more, is never enough in itself to make one learn
a new language well and quickly. Many businessmen, and others, have paid through the nose for this illusion. Only an urge to
have a command of a language for the sake of having a command of that language, and nothing else, no ulterior motive, will
give a person the power to get that command. One has to be interested in the foreign language for itself and for its immediate
rewards.
In adults this grasping at a language for its own sake, this saying "I want to do it just like the natives", is in a way a childish
itch. And children have this, and must have it. Language is an essential part of life for them. If they do not learn it, life does
stop, or at least does not develop as they feel it must. Without language children would feel surrounded on all sides by walls
they could not see or hear through. Yet language is not felt by children as a task. It is a natural part of growing up. For children
language is a matter of relaxed urgency.
This is also true for eight-year-olds learning a new language 'naturally', that is, by 'living' it. But it is not true for eighteen-year-
olds. They have become far too established members of their language community in most cases. They have no real urgency.

Page 174
This having or not having another language, other words, other grammar or 'general' meaning, to 'turn to' has a double effect. A
child learning a word in her first language has nothing else to turn to to represent the 'thing', so she must learn it; at the same
time, since there is nothing else to turn to, she cannot make a mistake for that reason, as adults do with new languages. When
adults, too, find themselves without anything to 'turn to' the results can be intriguing. One of the experiments I have made was
with a group of Spanish- and French-speaking students whose English pronunciation was atrocious. Reminding them that
English spelling was often bizarre - which they enthusiastically acknowledged - I presented to them a word that I was certain
they had never heard of. "Dingy," I said, "is particularly odd. It is spelt like this." And I wrote on the blackboard
"ZHUGHDEMB", at the same time repeating the word ''dingy" several times, fitting its two syllables to the two apparent
syllables in the 'word' on the board. I then asked the students to say "dingy" themselves. They all did very well. One French girl,
whose pronunciation was normally particularly French, got everything right, from individual sounds to intonation. I then
confessed the truth, and wrote "DINGY" on the board. The French girl threw her hands up in delighted recognition (not of the
word - she had never seen it before - but of the letters). "Ah!" she shouted. "Daingeee!"
This sort of prejudice about pronunciation, the urge to turn all sounds into something familiar, is just as strong when it comes to
meanings. This psychology of 'foreign' pronunciation I have just illustrated is significant. What is absolutely certain is that most
people, adults or not, can pronounce most 'foreign' sounds right if they listen, rather than just hear. In other words, the problem
of foreign pronunciation is connected with a general faculty that everybody knows of. (Though nobody knows anything about
the workings of the brain that make people listen or not listen.) So here is at least one thing, an essential and large part of using
a language, that looks pretty clearly as if it has nothing to do with any innate program or even a special language faculty.
In fact the difference in the abilities of people of different ages to learn a language is an area where academics and experts seem
to make particularly many confused and often contradictory statements, and where the true statements seem to be particularly
obvious.
The process of learning one's mother tongue cannot be speeded up, the experts say they have discovered - not a very remarkable
discovery, as ordinary humans would probably have noticed after all this time if it could be. They also say that the immature
nervous system is better equipped to

Page 175
learn a language. That is not a sensational finding either; as I have already pointed out, the principle applies to many other
activities as well.
However, some experts say that it is more difficult to teach foreign languages to children after they are 10 or 12. If they really
mean this they are talking nonsense. Other experts have recently arrived at the conclusion that young children at school are
worse at learning (i.e. being 'taught') foreign languages than older ones. This is something that ordinary people have been aware
of for a very long time. No sensible person would try to teach French, say, to an English-speaking four-year-old at school. But
they might well encourage the child to 'pick it up' in a French-speaking community. Trying to teach a foreign language to a
four-year-old can only come from cock-eyed theories, newfangled or otherwise. Equally, anybody who has been taught a foreign
language at school, and thinks clearly about their experience, knows that older children are nearly always better at it, and that
this is not only because they have been doing it longer - they are better at the actual learning process, the school type of learning
process, that is. And an eighteen-year-old starting a new foreign language at university usually makes so much more rapid
progress than an eight-year-old at school that most people would think it silly ever to compare them.
It is pretty clear, I think, that there are at least two ways of learning to do things. Humans are better at one way when they are
young, and better at another way when they are older. Or it may be that in later life they do not, for various reasons, get the
psychological opportunity to use their 'younger' method. The difference between eight-year-old language learners at school and
eight-year-olds 'in the wild' shows vividly how the different ways of learning are used in different circumstances, and alone
disqualifies any argument in favour of innate language faculties that is based on a comparison of language learning at different
ages.
How do people of different ages actually do at learning languages anyway? To begin with, eight-year-olds usually have a huge
advantage over eighteen-year-olds, because if they are living in the foreign country concerned - for instance - they are usually
left alone to really learn for themselves. Eighteen-year-olds are not as a rule so lucky. Even if everybody else leaves them alone
to get on with it, they won't leave themselves alone. They are full of ideas on studying languages that have been put into them at
school, and these usually slow them down. On the other hand, it is quite true that so-called 'uneducated' people often have great
difficulty mastering the language of a new country that they settle in. But in most cases they are immigrants who have practical
problems that join in a vicious circle with their language problem. To learn the new

Page 176
language properly they have to have constant ordinary contact with native speakers, and if they do not have any work, work that
gives them this contact, they cannot possibly learn.
But there is a simple, much more fundamental question. What do the linguists mean when they say older people cannot learn
languages as effectively as children? Many twenty-year-olds and thirty-year-olds, and older people, do in fact become fluent in
foreign languages in a few months with or without formal training, and learn a great many more meanings than eight-year-olds.
So there is actually only one way in which the children are superior; they are more accurate, and that is all. Adults are perfectly
capable of learning foreign languages effectively, but most of them make mistakes in speaking or writing, and they do that
almost certainly for the reasons I have suggested.
Adults who go about it the right way1 can acquire a far larger vocabulary in a foreign language, and far more quickly, than a
native child, for the reason that adults 'know' the world already, while children do not. Before a child can grasp a new meaning
('specific' or 'general') she has to learn the reality connected to that meaning - another of those simple truths easily discovered
without the help of experts. She has to do two things for each new meaning. Adults have to do only one: recognize the reality
they already know which the meaning refers to.
In fact, many adults who are able to hear and read a lot of a foreign language learn to understand it perfectly, both the spoken
and written language. Indeed, there are many adults who understand the foreign language much better than some less
sophisticated adult native speakers. Yet understanding is half of the whole business of mastering a language; so how can one
possibly say that adults are bad at learning language? The entire discussion of the respective language abilities of different ages
is more often than not based on false premises.
And what should make the language-program supporters stop and think is the obvious fact that it is people with agile brains (but
not necessarily 'intellectual' types) who are best at foreign languages. Some who are particularly adaptable and curious-minded
can even learn, as adults, to speak a foreign language perfectly. They are by no means prodigies; but most display quite plainly a
certain attitude of mind.
1 'Going about it the right way' means not wasting time looking up words in dictionaries, but reading and listening as
much as one can, not worrying about the words but about the 'story' and so discovering new meaning from the context,
just as children do with their own language; and not worrying if at first one understands less than a quarter. The fatal thing
is to spend hours with a dictionary getting through a third of a newspaper column or half a page of a novel. Most people
who do this are overtaken by tedium and despair and give the whole thing up after two days.

Page 177
Western schooling largely corrupts people for the learning of languages. But if one is not so corrupted, with age and experience
one gains ever increasing understanding of the principles of languages in general and of any particular language, and a sharper
intuition about how the language one is learning does or does not work. The same, or a very similar psychology is very likely
behind the fact that almost all bilingual people are very good at learning further languages as adults. Their ability compared with
the average is striking. This is probably due to the practical understanding they get at a very early age of how languages are not
translations of each other. They learn, instead, to make the direct link between words and experience. In this way with each new
language they can be free of the barriers of prejudice that a mother tongue usually puts between student and foreign language.
They can learn the new language much as a first language.
But where there is possibly a decline is in what one might call more 'practical' functions. (I want to avoid, though, dividing
functions up into any sort of categories, such as physical, intellectual, abstract, etc. That, I believe, is false.) I suspect that at
sixty, trying to learn some Italian in a desultory way in the depths of the Tuscan countryside where I am writing this, I do not
hear as efficiently as I did when I undertook learning a language in my twenties. I do not mean I am becoming deaf. There is no
evidence of this. I find I am simply unable, much of the time, to catch what seem to me words in a torrent when people speak at
the normal Tuscan rate; while some younger non-Italian friends who are also learning the language, who have almost certainly
been exposed to it less than I have, and probably have a smaller passive vocabulary - they apparently manage to understand,
very well for all practical purposes, most of what is said to them. I, on the other hand, spend a lot of my time gaping stupidly
and wondering, after the third attempt, whether I ought to pretend I've got it or not. "Si, si" I say hopefully.
To my chagrin I also find that I am prone, to some extent at least, to a failing which for years I have constantly criticized
students severely for, and which is one of the almost universal great obstacles to learning a foreign language. For instance, "We
ought to put the wine in the shade" I say to my student-of-English friend. Within less than thirty seconds he says "There's
shadow over here." It is right to get worked up about this fault, because the student understands perfectly what he hears; but he
does not observe what he understands.
This is the crux. Children do not fail to observe in this way. They may sometimes observe incorrectly. But they make the
observation. This is the essential difference between children and most adults. I have suggested

Page 178
some reasons why it should be so. There may be other even more basic reasons that I am not aware of. However this may be,
one thing is clear. Observation, not program, is an essential of mastering a language.
I realize, though, that many will not be satisfied with this sort of reasoning alone. They will demand an experiment. The only
valid experiment would be to find an eighteen-year-old (say) who had a completely normal upbringing in every essential
respect, except that she had no experience of language, and see how she then coped with a first language. But the experiment is
by definition impossible to make.
This has not stopped Chomskyans appealing to the case of "Genie" to support their argument.
Genie is an American girl who was locked up and heard almost no language until she was thirteen.
"Despite this horrifying background, Genie's intelligence turned out to be within normal limits in essential respects, and
thus her progress with language learning provides a useful basis for comparison with the language acquisition of more
ordinary children. Her early language acquisition was typical of all children in that it passed through stages of one-word,
two-word, three-word and then four-word utterances; however, Genie's three- and four-word utterances typically
displayed a cognitive complexity not found in the early speech of normal children, and her vocabulary was much larger
than that of children at the same stage of syntactic development. In general, her ability to store lists of words is very good,
but her ability to learn and manipulate rules has been minimal. This is reflected in the fact that whereas the 'two-word'
stage lasts for about two to six weeks with normal children, with Genie it lasted over five months:
e.g. Doctor hurt.
Like mirror.
Moreover, the kind of early negative structures which most two to three year old children use for a few weeks, where the
negative element is initial in the sentence, still persisted with Genie some one and a half years after she had first learned
to use negatives:
e.g. No more ear hurt.
No stay hospital.
Not have orange record.

Page 179
Indeed, no syntactic rule which is normally taken to involve the movement of a word or phrase from one point in a
sentence to another...has been consistently mastered by Genie as yet. But in contrast with this slow and partial linguistic
progress, Genie's intellectual development appears to be progressing extremely rapidly, and to be approaching the normal
for her age." (Smith and Wilson, 1979, pp.34-35)
Notice first the obsession with word order that I have complained about before. Word order is something comparatively
sophisticated, added to or imposed on the more fundamental business of language, which is sounds to refer to human experience.
Of all the parts of language, one would think the part for which one could least argue the need for a special faculty, rather than
a general faculty, would be a part that seems to involve the sort of grasp of relationships that is needed for the mastery of word
order.
In any case, it is a mistake to see in these 'incorrect' sentences of Genie or any other child nothing but mistakes in word order, or
word 'movement' as the Chomskyans would have it.1 Basic to these 'mistakes' is that the children do not immediately grasp the
full and exact meaning of pieces of language, such as "does, do, want, to, I." Their faculties are strong enough to work out
straight away the basic meaning of a piece, but not to master exactly how it is used. And, Chomskyans say, children go through
stages, each one more complicated than the last, in learning their language. For instance, children learning English first say
negative sentences like "Not Daddy come" and then after a few weeks change this to "Daddy not come". But, Smith and Wilson
maintain (1979, p.38), there is no "generally observable reason'' why the first form is simpler than the second, so there must be a
special linguistic notion of complexity which children apply. It is an unwarranted assumption, though, that children go through
stages in this order because one stage is simpler than another; nor does it follow that because a general psychological reason is
not observable, there isn't one. Nobody knows nearly enough about how the brain works to say that. It might be, for example,
that first or last positions in many sequences - music, numbers, rows of people - carry more emphasis.
But there are other serious flaws in the argument drawn from the story of Genie.
1 See above Chapter 2, pp.23-24, 28-29, Chapter 3, pp.44-46, and Chapter 6, and p. 186 below.

Page 180
Her slowness in 'grammar' development does not show that it must have been a specifically language faculty that was impaired
rather than a more general one. If her experience, or rather lack of it, could impair a special language faculty, it could just as
well impair a more general one used to work out 'grammar', that is, in this case, word order. Nobody knows how such a possible
general faculty works. But nobody knows how a supposed special language faculty works. Nor does anybody know exactly what
her experience did to Genie's brain, and anyway this is just one girl's experience and evidence. One is struck once more by the
strangeness of what the Chomskyans are in effect asking us to believe: that humans are born with a special language program,
which changes very rapidly through various phases into something more powerful around the age of two or so -but, as it is a
program, as a result of forces entirely within itself, not of the children's experience - and then virtually disappears in their teens.
Smith and Wilson's statement that Genie's intelligence "turned out to be within normal limits in essential respects, and thus..."
immediately arouses suspicion, even if one assumes we know what intelligence is.
The story of how Genie coped with learning English has been told by Susan Curtiss in Genie. A psycholinguistic study of a
modern day "wild child" (1977). On page 47 she writes:
"There were many problems involved in testing Genie. Standard psychological and intelligence tests could not be used at
first because such tests rely heavily on verbal comprehension and verbal responses. Other instruments, less dependent on
language abilities, had to be used in any attempts to assess Genie. Moreover, for a long time, there were the additional
problems of not being able to score Genie's performance in the way provided by the normed scoring data to be used in
accompaniment with these tests. Thus, in many cases, the resultant numbers represent merely the examiner's best attempt
at quantifying Genie's performance in a way compatible with both the test itself and normal scoring procedures."
This is a very different picture of Genie from the one suggested by Smith and Wilson. From the beginning one can see that it is
impossible to describe her intelligence as being within normal limits even by the standard of traditional testing.
Inherited characteristics and the effects of environment constantly interact in all beings. It is quite impossible to disentangle
them merely by observing behaviour, except perhaps in one egg twins. There is no

Page 181
conceivable way of distinguishing the two in Genie. It is impossible to find out what is the effect of her experience and what the
effect of her biological development past a hypothetical 'critical' stage for learning language, for instance. Genie had very
abnormal experience, and nobody can know all the effect on her. Indeed, being without language till she was 13 was itself part
of that abnormal experience and is likely to have had a fundamental effect on her mind. One can never discover the behaviour of
'normal' children deprived of language for many years, because by that very deprivation they will not be normal.
Even if one accepts the psychological tests done on Genie as valid in showing that she was like other children in many ways,
that is precisely all they can do. They are irrelevant to the attempt to isolate purely linguistic processes in Genie. She was also
obviously unlike other children in many ways, and there is no way, short of direct observation of her and other people's brains,
of discovering what that unlikeness did to her.
Again, even if it was shown that there really is a critical stage in language learning, that would not in any way prove that it is
the result of separate linguistic processes.
Curtiss has never actually seen this distinct language 'competence' in Genie or anyone else. We can only observe it within
ourselves. Yet when it fails to produce in Genie the results that it ought to, Curtiss persists in uncritically assuming that it exists
as a separate reality.1 She puts down all failures to conform with the community's principles and conventions of language to a
quite distinct "system of information processing", and says that Genie has an underlying ability to do much better (pp.203-4).
Curtiss cannot prove this by referring, as she does, to Genie's "exceptional utterances" (p. 197). There is another possible
explanation, just as worthy of consideration. A person's language 'competence' may be entirely the product of the 'performance'
of brain faculties that develop that 'competence'. Curtiss herself refers more than once to linguistic elements and rules being
acquired. If 'performance' or part of it is faulty, the acquisition will be imperfect and thus the 'competence' imperfect too.
'Competence' is merely a particular form of 'performance'.
It fact it is far better to abandon such categorizing abstractions as 'competence' and 'performance'. They are examples of the
terrible burden of words that I attack in Chapter 11. They set up myths behind whose banners academic experts typically array
themselves and bandy false
1 There are undoubtedly universal principles of human pronunciation, ways of making sounds that come naturally, and
other that do not, and sounds that are impossible, as with all species. This is another matter, not at issue here.

Page 182
profundities, blinding people to the only thing that is important: what actually happens.
I have suggested at the beginning of the chapter several of the things that have to happen for a person to be able to use language.
One of the things that perhaps went slightly wrong in Genie's case was that she did not notice patterns as accurately as ordinary
children do; and if she did notice them, perhaps she did not appreciate them 'enough', did not fully understand distinctions in
meaning, and did not appreciate how 'important' it is considered to conform exactly to the conventions.
I have seen it claimed that an IQ of only 50 is enough for a person to acquire a first language, as if this was something
remarkable and supported the thesis that language 'competence' is innate. It does not support such a thesis in any way. If IQ had
any bearing on the matter it would mean that intelligence is the same as various general faculties. "Intelligence" is one of the
most corrupt and corrupting words in language, but even so I do not believe many people would claim, after a few moments'
thought, that it is the same thing as distinguishing and producing different sounds, recognizing the connections between sounds
and experience, remembering the connections, calling them up when needed and coordinating them for one's own purposes, and
so on. It is clear that there are very wide differences in the ability to 'work things out' between different individuals of the same
species of mammal or bird. That does not appear to have any noticeable effect on the functioning of other general faculties in
those individual animals. Even if one accepts the validity of 'intelligence' testing, all that the statement about IQ and acquiring
language tells us is no more than that one can be of low intelligence and still acquire language. The same can no doubt be said
about riding a bicycle.
What he calls "degenerate evidence" is one of Chomsky's main supports in favour of the idea of children's inherited knowledge
of grammar, rather than knowledge gained from listening to other people. He maintains that human 'performance' of language,
what people actually say, is too full of mistakes, slips, and stumblings for children to be able to learn the rules of grammar
accurately from it. Yet, he says, children do rapidly master language, so they cannot be getting its principles from outside data,
but must be born already programmed with the principles of universal grammar, so that they are not misled by and can sort out
the insufficient and imperfect examples of language that they hear.
There are so many things wrong with this argument.

Page 183
For a start, why should one postulate a universal grammar after asserting 'degenerate' evidence? It would be just as reasonable to
argue the reverse: children's mastery of language shows the evidence cannot be degenerate. What exactly is 'degenerate'
evidence in this case? Whether incorrect data is confusing, and how much has to be wrong for it to be significant depends
entirely on the nature of the data and the use and purpose one puts it to.
Chomsky contradicts himself. He has emphasized again and again that children always get it right, that, as he once put it
(Reflections on language, 1976, p.32), "the child unerringly makes use of the structure-dependent rule..." Chomsky cannot have
it both ways. He is claiming that children get structures such as "Is the man who is tall in the room?" (p.117, 1a) right despite
the mistakes their elders make - and the adult mistakes he means must be wrong word orders in the case of sentences like 1b on
page 117, because otherwise there is no point in what he is saying. Yet this means that he is claiming that adults make mistakes
but children don't. The truth is that adults never make that sort of mistake, or, if they do very occasionally, certainly no more
than children do. So children in fact have perfectly good data to go on.
Chomskyans also claim that children learn language very quickly. That, too, is clearly untrue. From birth (and remember, they
are supposed to be born with universal grammar) it takes them many years to achieve anything even approaching adult
sophistication of expression, and many humans never use 'advanced' grammar throughout their lives.
Smith and Wilson (1979, pp.26-27) say that despite the "diversity of the utterances to which speakers are exposed in learning
their language" there is great similarity in the "grammars which result from the learning process". They imply that this means
there must be some innate universal grammar at work. It does not mean that at all. Any 'diversity' there is (and millions of small
children in fact hear hundreds of things that are the same for all of them) is diversity in the 'specific' meanings of whole
utterances that different children may hear, but not diversity in the 'general' way pieces of language are used. If there was such
diversity in this 'general' way of using language, then there would be great differences in the 'grammars' people learned. It is
rather obvious that if all parents use the same 'grammar' their children will all use the same grammar too; and if each family
uses a different 'grammar' each family's children will use a different grammar too. That is, you will learn whatever 'grammar'
you hear.
But the most fundamental and very elementary objection to the proposal that grammar must be innate and cannot be learned
from 'outside'

Page 184
because the data children are given is insufficient, is this: the language that we know children actually get right is not, in the
Chomskyan system, the 'deep structure', but language that has gone through 'transformations'. But the 'transformations' must be
different in each language, yet children learn them very efficiently and can only learn them from the data provided about their
own particular language by their elders. Chomskyans argue, though, by their 'meta-theory' of universal grammar, that children
are born with the knowledge of which transformations, among the many theoretically possible, human language actually
permits. There are, they say, constraints that apply to all languages.
This is irrelevant. The fact remains that children have to learn to use certain modes of expression and not others, even though
those others exist in other languages, and they can only learn which expressions are permitted in their language from experience.
No pre-programmed universal grammar or transformational restrictions can help them there.
One simple example should be enough. English does not normally invert main clause verb and subject after (say) an adverb:
1 Then the cat went to sleep.
But other Germanic languages do:
2 Then went the cat to sleep.
(Swedish: Då somnade katten; German: Dann ist die Katze eingeschlafen; etc.)
Such grammar can only be learned from the data;1 otherwise German and English children, for instance, would never know
which to do. Both ways are possible, both exist. But the children always learn the way their language does it. Very simply, one
can only speak any given language on the basis of the evidence one has of that same language. There is no other way.
The argument actually should stop here, because really no more should need to be said. Still, perhaps it is worth emphasizing
once more the obvious, that children must hear meanings and remember them. If they don't remember them, they will have no
language to use. No amount of pre-programming can have any effect on this simple necessity. Hearing and remembering several
thousand meanings, many with subtle nuances, is quite a task. If there is anything remarkable about human mastery of language
- though "remarkable" is a strange comment on something so
1 But see p. 188.

Page 185
normal - it would perhaps be that people all have such powerful and accurate memories.1
So it does take time. Remembering - or rather, collecting and becoming familiar with - thousands of 'specific' meanings is a far
longer business than mastering the 'general' meanings (grammar) of a language. It is ironical that it is grammar, or syntax, that is
regarded as the difficult part of language. This is the fault of the grammarians and linguists. In fact the essential grammar, the
'general' meanings, of a language are comparatively soon mastered by a child. Firstly, 'general' meanings are few compared to
the 'specific' ones, and a child can master all those that are important comparatively quickly for that reason. Secondly, precisely
because 'general' meanings are general, a child experiences what they refer to in life at a very early age - the distinctions
between things, actions, qualities etc., time, position, and so on.
It is not at all remarkable, though, that a number of 'specific' meanings come before 'general' meanings, because a child has to
have, first, something for the 'general' meanings to be about.
A child recognizes fairly early on, then, the realities in life that the 'general' meanings of a language indicate. This is where
'specific' meanings are so different from the practical point of view, because, as I have already commented, a child is not only
constantly meeting new words but also constantly meeting new things in life, and cannot understand any word until she has
experienced in some way the reality it refers to. This slows her down at the beginning. Later, learning new words takes time
also simply because there are so many of them. Because some older children hear more new words than others, the pace at
which they learn varies a lot, while when they are smaller and experience is more equal, they learn at roughly the same speed.
There is something essential about learning any language, whether one's 'own' or a foreign one, that - as far as I know -
grammarians and linguists largely ignore. This is the large amount of idiomatic usage, including the most common and everyday
usage, that cannot be worked out from grammatical or 'general' meaning principles. For example, the grammatical logics of pairs
of languages like Spanish and French or Swedish and Norwegian are very alike. But although the grammatical logic of the
language is not actually broken in most idiomatic usage, there
1 And with the meanings they have to remember all the conventions as well. For example, the Scandinavian variation I
have mentioned before; in Swedish "the war" = "krig/et", in Norwegian = "krig/en"; in Swedish "krig/en" = the wars. No
programming can have any effect on this 'problem' for a Swedish or Norwegian child. All that is involved, after
understanding, is memory.

Page 186
are many different ways in which it can be applied in these 'special expressions, and one culture will often apply it quite
differently from its sister culture, so it is quite impossible even for a Swede, for instance, to 'work out' a Norwegian idiom, or
for a Spanish-speaker to work out a French expression. Each language has its idiosyncracies, quirks, and spirit and one can only
observe and remember. The most basic ideas, like making requests or enquiring after health, are often expressed completely
differently in each language.1 Yet children (and foreign adults who can free themselves of their own language) are very good at
observing and remembering. There can again be no question of any 'universal grammar' operating here. Once again one finds it
comes down to individual and separate lumps of meaning, whether they consist of one or several pieces.
In fact, the more one thinks about the Chomskyan position the more perverse it comes to seem. It is difficult to see how
anybody could ever have been taken in by it and gone along with it for more than a short moment of lapse. It is so completely at
variance with simple facts and simple logic. By far the greater part of people's difficulty in learning a foreign language is that
they have to observe a very large number of new words and notice the 'life' situations in which they are used and the other
words they are used with. And then they have to remember all that. The words and their use are unique. No inherited program
can have any effect whatsoever on this task. It wouldn't matter whether people still had such a program or not. It is so utterly
clear that other abilities are required.
It does not seem to worry Chomsky and his followers that even with their assumptions they have continually run into difficulties
with the 'transformational' system. The transformation rules were felt to be too complex and difficult for children to acquire
quickly - though it is hard to see why complexity matters if their 'meta-theory'2 is correct. But anyway, Chomsky decided that
transformations can only move one part of a sentence at a time, so that children use, in forming passive sentences, say, two
simple rules one after the other instead of one 'complicated' rule. Then later again Chomsky decided to replace these two rules
with one very simple and general rule.
Chomskyans have become so obsessed with their would be 'universal grammar', with the idea of language as an innate faculty,
that it has
1 For instance: "She is six years" is how some languages put it, but in English it has to be "She is six years old" (or "She
is six"), while other languages again insist on "She has six years".
2 See p.184 above.

Page 187
become a blind and unshakeable assumption, proof against any doubts raised by its difficulties. "Ray Jackendoff...argues that
only by discovering the universal principles constraining choice, reducing possibilities, can we hope to explain why language is
natural rather than invented." (Campbell, Grammatical man, 1982, p.178. My italics. Notice again the tie up with
communication theory.) In other words, they are so determined that 'universal grammar' shall exist, that language shall be an
inborn faculty, that they take it as granted, and are prepared to go in for endless contortions to create a contraption of grand all-
ruling principles at the bottom of language. They naively assume what they still have to prove. They seem obstinately reluctant
even to consider that the idea of language as an invention to express human experience and thought fits both the facts and
common intuition far better and more naturally. They insist on setting up all sorts of non-existent and unnecessary problems and
then looking for impressive-sounding profundities to solve them.
Of course there are restrictions on human language. But this is so obvious that it ought to go without saying. To say that there
are restrictions on language is simply to insist on the fact that humans cannot express anything that is not their logic, their
feelings, their experience, their fantasies. One only needs to think of people who are colour blind to realize that not even all
humans experience reality in the same way. Some animals clearly do not have certain kinds of experience that humans have,
and some animals have kinds of experience that humans do not have. Perhaps even more fundamentally, just as there are more
colours than the colour blind see, there may be more than one kind of logic. But humans will never be able to grasp any logic
that is not the same as their own. There can be no insight into our own logic either, because however much we think about it we
can only see it in terms of the logic we already have, so we can only go round in a circle.
Among the many strange claims implied by the insistence on innate universal grammar as opposed to learning from 'degenerate'
evidence, must be that children 'acquiring' English (say) cannot progress from, for instance, the incorrect "Not Daddy come" to
the correct "Daddy hasn't come" by any sort of observation. But if that is true they must be programmed to go through 'wrong'
stages before they get to the right one. The Chomskyans will perhaps say that children are working through the transformations
away from the universal deep structure. That still leaves it all just as bizarre, because they do not allow that the data is good
enough to

Page 188
tell the children which of the many possible transformations is the right one for English.
What are they saying that transformational grammar is? Are they saying that it is a psychological reality, that is, that small
children and adults alike apply its rules whenever they use language (among other things to form those sentences they have
never heard before), despite the fact that those rules are immensely complicated, and that it is far more complicated to apply
them than to use, direct, the 'superficial' words and sequences as we have them?
If they do not claim transformational grammar as a psychological reality, they are merely proposing a series of hieroglyphics
without any significance or relevance of any kind.
Speaking of Chomsky, John Searle says:
". . .But, more importantly, the syntax that he came up with was extremely abstract and complicated and that raised the
question: 'How can little kids learn that?' You cannot teach a small child axiomatic set theory, and Chomsky showed that
English is much more complicated than axiomatic set theory. How is it that little kids can learn that? What he said was:
'In a sense, they already know it. It is a mistake to suppose that the mind is a blank tablet. What happens is that the form
of all natural languages is programmed into the child's mind at birth.'" (The Listener, 30 March 1978, p.397)
This is a good example of the circular argument. Chomsky has erected a frighteningly complicated and abstract system of
syntax, and instead of conceding that he has perhaps created something artificial and unreal, he uses its very difficulty to suggest
that therefore its mastery must be inborn. It reminds me of the impoverished community in Brazil who, when asked if their
sufferings did not prove that god could not exist, replied that on the contrary, their sufferings were so great that they proved that
god must exist, since only he could provide purpose for those sufferings.
It is true that humans - very young humans - must have some faculty that makes them able to see the patterns (Chomsky wants
to call it "form the rules") in their mother tongue. They have to be able to grasp many such patterns as:
a potato/an old potato
I have never heard such music./Never have I heard such music.

Page 189
But a faculty that can work out linguistic patterns does not necessarily have to be an exclusively linguistic faculty. It is not easy
to point to areas other than language where there is clearly recognition of patterns, but this may simply be because there are not
many activities demanding the faculty which are, as it were, exposed to view as they are in language. However, there are other
activities which plainly need that faculty for grasping patterns. Games are one sort; music is another. And in fact it seems clear
that some animals have the faculty too, though no doubt in a much less sophisticated form. A sheep dog, for instance, could
surely not be trained if it did not have some perception of patterns.
I cannot, of course, prove that it is not as Chomsky says: inheritance of mastery of a highly abstract system. But there is really
no evidence at all in the first place for the complicated system he has contrived. At the very best it is, as I suggested earlier,
merely an attempt to put into code form something that is in its reality in principle pretty simple. But it becomes apparently
complicated when violence is done to language by trying to force it into unnatural forms. Chomskyans cite many pieces of
language to demonstrate the special innate programmed language faculty. I have already tried to show, in Chapter 3, how in fact
these examples provide no basis at all for their assertions and merely make plain their confusions.
Chomskyan linguists' search for generalizations which apply to all languages and that therefore can be used in the creation of a
'universal theory of language' is meaningless and pointless, but they pursue it because they want to draw up a grand scheme
which they can fit everything into, rather like one of their 'phrase markers', with their 'universal theory' in the place of their
dominant ''S"(sentence). It is nothing more than draughtsman's itch. It cannot offer any enlightenment.
They go further. Just the rules themselves are not enough. They are also much exercised by the problem of 'evaluating'
grammars. They believe that there is more than one 'descriptively adequate' grammar of a language, but that an 'evaluation
measure' is needed to "explain the choice of grammar that the child eventually" makes (Smith and Wilson, 1979, p.252). For
Chomsky believes that "the language is 'reinvented' each time it is learned" (Lyons, 1977, p.112n); innate formal principles of
grammar "account for...the child's success in constructing the grammar of his language on the basis of the utterances he hears
around him." The innate 'language acquisition device' must contain such an 'evaluation measure'.
At the same time Chomskyans admit problems with their 'universals'. They wanted their universals to be 'absolute', that is, to
apply to all languages; a single rule in a single language which disobeys the universal

Page 190
obviously means that the universal is not a universal - not an 'absolute' one, anyway. The example from Fe'Fe' that I mentioned
earlier1 was quoted by Smith and Wilson to illustrate how universals can be spoilt; the evidence from many languages did not
fit Chomsky's system. So Chomsky invented a new sort of universal: 'relative universals'. He no longer claimed that the theory
of language should put an absolute ban on, for instance, extracting NPs from within co-ordinate structures;2 instead, it would
say: no rule may extract an NP from within a co-ordinate structure, unless otherwise stated. He is being serious. In fact, he has
used his new invention to provide an 'evaluation measure'. A rule which conforms to a 'relative universal' is to be regarded "as
simpler and better than one that does not."
"We could imagine this complexity reflected in the formulation of the rule itself. A rule which makes no explicit reference
to co-ordinate structure might be interpreted as unable to apply into such structures; any rule required to apply into co-
ordinate structures would then have to make explicit reference to all the types of co-ordinate structure into which it could
apply. Such a rule would automatically have a much more complex and lengthy formulation than a rule which obeyed the
relative universal.
This new notion of linguistic complexity makes a number of testable claims. In particular, since it is construed as part of a
language acquisition device, it will make predictions about the order of acquisition of certain rules and constructions by
the child. Rules which violate linguistic universals, since they will be more complex and less expected, should be learned
later than those which conform to them. Such a claim would be fairly easy to to test. Furthermore, a child who has been
exposed only to data which conform to the universals at a time when he is formulating a particular rule, should not
spontaneously produce new sentences which violate them. It will be recalled that a rule must now be treated as
conforming to the universals unless it contains explicit indication to the contrary: a child who has no evidence to the
contrary in formulating the rule, then, must treat that rule as conforming to the universal. Again, these predictions are
fairly easy to test. Children learning English, which obeys the co-ordinate structure constraint, should not normally
produce sentences which violate it: we are invoking here the notion 'impossible mistake for a child learning X to make'.
We might also
1 Chapter 3, p.57.
2 See Chapter 3, pp.56-58.

Page 191
predict that a child will find it harder to learn a rule which violates a universal, not just because it will be less expected,
but also because its exact formulation will not follow from universal principles, and the child will have to use a
considerable amount of trial and error before he gets it to cover exactly the right range of cases." (Smith and Wilson,
1979, pp.262-263)
There are several illogicalities in all this.
It amounts to a statement that the simplest and best rules are the ones that are most common among the world's languages. That
does not necessarily follow at all.
Then they say that "a child who has been exposed only to data which conform to the universals at a time when he is formulating
a particular rule, should not spontaneously produce new sentences which violate them." That is undoubtedly true - not of
universals, but of any rule at all! So that is not a very profound statement.
Further, the last part of the last sentence quoted above is a clear admission that children learn grammar that is not universal. If
that is so, why are any innate universals necessary for children's language learning?
But I think what most thoroughly demolishes the whole theory is the question as to what happens if the original universal rules
are wrong, if they miss the point. (Then of course, the 'constraints' on rules are only necessary because the rules are bad ones,
and the whole system is a mess.) I think I have already given enough examples to make it almost certain that a great mass of
Chomskyan rules are indeed completely wrong - if not the whole lot.1
And finally, the whole 'relative universal' wheeze once more gives the impression of that desperate search for a method of fitting
everything into a mere paper code or chart devoid of any link with what actually happens inside people's heads.
The only possible reason for trying to achieve a 'good', or 'better', grammar is to make foreign-language learning easier. If one
assumes that explanation is the right method, obviously the more simply the 'general' meanings (grammar in the narrower sense)
can be truthfully explained, the better. The dilemma that worries some linguists of whether it is better to have a few rather
complicated rules or more but simple rules, for example, is false and typically abstract and academic. In practice different
individuals need different explanations. And there are many 'general'
1 See Chapter 3.

Page 192
meanings that no analysis or rule can ever begin to convey the use of. What rule or analysis would teach foreign-language
learners how to use "-ing" in English, for instance?
Smith and Wilson's (1979) summing up of Chomsky's 'achievement' is a mixture of the contradictory and the obvious
"Language is a reflection of the human mind, not just in the sense that humans have produced it, can learn it, and do
speak it, but in the much more specific sense that language is as it is because the human mind is as it is. Nor is it just that
humans are intelligent enough to construct and learn languages; rather, they are specially designed to construct and learn
the sort of languages they do, and this ability is not a reflection of general intelligence at all. The human language faculty
is unique and innate. Chomsky's achievement has been to make this clear." (pp.265-66)
So language is a reflection of the human mind, language is as it is because the human mind is as it is, but the language ability is
not a reflection of general intelligence at all. Either that is a contradiction, or the first part of the statement is merely a statement
of the obvious. It is rather obvious that any product of the human mind is the product of the human mind - "as it is".
It is equally obvious that language is something specifically human. I do not count the experiments in teaching language to apes
as in any way modifying this truth. Apes don't learn language when they are together by themselves. One could say that perhaps
apes could invent language. The fact remains they haven't.
And finally, it is just as obvious that the human ability to use language is innate; if it wasn't, humans couldn't do it; for nobody
can do anything they are not born with the means to do. Already five thousand years ago people would have been very surprised
by a flying cat, a miaowing sparrow - or a talking mouse.
It is not good enough to state dogmatically, as some do, that the fact that all languages have much in common has nothing to do
with the world appearing basically the same to everyone everywhere. That is just a bald assertion, with no evidence to support it,
and with no logic to refute the very

Page 193
strong likelihood that common experience of life does indeed produce the common elements in languages.
One can sum up the whole business of supposed 'linguistic universals': Linguists have not yet shown, as far as one knows, that
there is a single unquestionable universal of language - as opposed to universal ways of thinking.
Even if one assumes that there are linguistic universals, they would not necesarily demonstrate a separate faculty inherited with
a restricting program, rather than universal ways of thinking and experiencing which limit what people can express and the ways
they do it.
Even if one assumes a whole mass of universals, linguistic or logical, or both, showing and describing them will not give the
slightest new insight into human thought and psychology, or anything else, but will be idle analysis for its own sake, without
importance or interest. I know of no interesting conclusion that anyone has ever tried to draw.
Even if one assumes linguistic universals, and that they are important and interesting, they cannot explain how humans learn
different languages. What is not universal can only be learnt by observation. Therefore faculties which are not separate or inborn
language faculties must be used.
There is just one simple universal truth to be stated about the form that any language, actual or hypothetical, can take: a set of
conventional meanings, once given, can be arranged in any way that makes sense to those who use those conventional
meanings. To make sense there will be, depending on what the particular conventions are in each language, various practical
limitations, which I have given examples of earlier.
The basic truth to be realized about language, though, is that studying it cannot tell us anything at all about the basic nature of
thought.

Page 194
10
Thinking and Language
Thought is not language. Thought is not based on language. Thought does not depend on language; language is not a condition
for thought.
There is no essential connection between language and thinking except in two senses: that language is a translating device for
the imperfect expression of thought or of the awareness of experience; and without thinking humans could not produce
language.
This does not mean that thought is not in practice influenced by language. Sadly, in most people, thinking is very much
dominated and shaped by language, even though it does not need to be. The corrupting effect of language is probably the most
important thing that needs to be shown about it.
And I am not talking about whether or not the existence of language leads to thought. It almost certainly does, since through
language people try to tell other people about their thought, and thought can arouse thought.
First, though, I must try to show how utterly different and separate language and thought are. People will never even begin to
understand either of them until they understand that; yet it is something people can work out or observe directly for themselves.
I was once chased down a Cornish lane by a duck. (We were both on foot -the duck and I, I mean. I can't remember flit was a
she or a he, but I think it regarded me either as a sexual rival or as a possible mating partner.) It was some years ago. What has
been kept inside me all that time, so that I can now tell you about it? Not words. Not the isolated word "duck", for instance,
because although that word is certainly kept inside me somewhere, it is obviously useless for keeping just this unique Cornish
occasion inside me -its whole function is to be ready for any 'duck' need that arises. Nor can it be a whole sentence, "A duck
chased me down a Cornish lane", because why not "I was chased by a duck down a lane in Cornwall" or "I ran down a Cornish
lane with a duck in pursuit" or ''I had to flee from a duck along a country road"? (I am free, when I speak to you, to use any of
these different meanings of the same event - see Chapter 8.) It is clear that it must be something else that is kept inside me,
something much closer in its nature to the original unique reality. And what 'comes up' inside me in the first place to allow me
to tell you about it? Again, it cannot be words, for if it was, there would be no reason why I should not say it was a tiger that
chased me, or

Page 195
that I chased the duck, or that I once sailed over Baghdad on a carpet. Something that is not words must first trigger the
particular words I actually use.
This point is fundamental, and is alone enough to make it clear that it is impossible for thought to be based on language. The
thought must come first, quite independently, on its own. For when I say something (or think or say words in my head) what is
it that decides what I say? Why do I say "duck" and not "tiger"? What 'chooses' that particular word? It is my thought, in this
case the picture in my head of a duck, and only a duck, not of something else. If thought depended on language, all thought
would be random, arbitrary - for what would decide the language?
But there are also many practical ways to see how thought is (or can and should be) quite independent of language. For
example, probably almost everybody knows that people often think about things that they do not know the word for. There must
be quite a few people, for instance, who have thought about the thing which has the word "pelmet" associated with it, without
knowing that word.
The philosopher Wittgenstein would have had to deny this, if I have understood his position correctly.1 He is supposed to have
shown that human concepts - or at least many of them - are impossible without language. He said that you can look at a triangle
and first see this as apex and that as base, and then look again and see that as apex and this as base. You have two different
experiences although what your eyes actually see is only one and the same thing. But without knowing the geometry words
"apex" and "base" you could not have these experiences.
This is quite untrue. There must be many people who, like me, often become intrigued by patterns on floors or wall tiles. I have
found many patterns that I can 'interpret' or 'look at' - experience - in several different ways. For most of them I have no words,
but the various alternative patterns are none the less clear to me for that. A fairly simple example is a chess board. One can see
this as lines of white diamonds tip to tip on a black background, sloping diagonally up from left to right - or right to left. Or as
1 Here and on p.208 below, where I quote Wittgenstein again, and Descartes, I am not quoting direct from their writings.
Here I take my information from John Searle in Magee, 1982, p. 155, and on p.208 from James Fodor in States of mind
(ed. Jonathan Miller), 1983, p.89. It is of no importance that I do not make sure I am quoting correctly from the originals.
As I suggested in the Introduction, it is the effect which is important, the use that academics are making of earlier writers
now, what Searle and Fodor, for instance, are trying to make us think now by quoting Wittgenstein or Descartes - not
what those philosophers did or did not actually say or mean.

Page 196
lines of black diamonds tip to tip on a white background, sloping diagonally up from left to right, or right to left. Or as vertical
columns of alternating black and white squares; or as horizontal bars of alternating black and white squares. I still remember
clearly how I noticed the different possibilities for the first time, on a floor something like a chess board in a hospital in Nairobi,
where I was visiting my mother, who had blackwater fever. I was at the most five at the time, and knew none of the 'technical'
words I have used above to describe the various patterns one can experience on a chess board. And of course there again arises
the problem: how do words like "base", "apex", "diagonal'', "vertical", "horizontal" etc. come about in the first place? Once
again, they can only be the response to something already experienced. This distinction between words and concepts is
confirmed by the cases of people with brain damage who have lost the name of an object but can still draw it.
Japanese has no plural. Do Japanese-speakers therefore not think plural? Japanese also has no articles. If a Japanese says the
equivalent of "I went into garden and saw big black cloud", is she not thinking of a particular garden and, on the other hand, a
new unfamiliar cloud?
In the same way, by the Wittgenstein argument it would be impossible for children to learn their native language at all. They
can only do it by recognizing, in the continually new words and expressions they hear, things they have already thought about. If
they hadn't thought about them the new words would be meaningless. Put it another way: if thought is language, or depends on
language, then children cannot think anything before they learn language; and even when they begin to do that, they cannot
think any particular thing they have not got words for yet. This cannot be true.
First, in order to understand or use any word, a child has to attach it to something, has to recognise something she can associate
the word with. I suppose people might argue that there things are, and a child is aware (but what is that?) of them and simply
connects the words to the things, and can then think, through the words, about the things. That could work, maybe, for the
things 'outside' that the child can see, or hear, or touch. The child sees the cat, and people say "cat", and so the child can
associate the word "cat" with what she sees. But suppose the cat goes behind the door, and the child cannot see the cat any
longer, but is aware that the cat is there, sufficiently aware to say "The cat is behind the door". What is happening then? Since
the child can no longer see the cat, there must be something else - not a word - inside her first which makes her say "cat". At
least with something like "cat", though, people could still say that the child learnt the

Page 197
word through seeing the cat and hearing the word together with it; and thereafter could also think about cats. But that won't
work for anything that is not outside the child but inside her. When she hears and understands "know" or "understand" or "hope"
for the first time what does she recognise, and where? How can she recognise what they stand for if she is not already aware in
some other way of the things they stand for? This is why, for example, a child of six, say, cannot understand an article on
politics in a newspaper, however advanced a reader she may be for her age and however simply the article is written. She is not
aware of the things she can associate the words with.
A person who has not lived at all, not experienced at all, cannot learn any language. This may seem a ridiculously unnecessary
thing to say, because it is so obvious. But perhaps one has to say it, because linguists give the impression that language is
something quite independent in its own right.
I have more than once heard people claiming that human awareness itself comes from language. It must be the other way round.
How could language come first? It is not possible for a system of expression to come first and cause the very thing - awareness
- that has, on the contrary, looked for expression. Where would language that caused awareness come from? Language, either in
its origins among humans, or in each individual, is impossible without the individuals knowing the things they want to convey.
Language is surely something far more superficial than awareness, logic, and the powers of co-ordination. I have heard someone
say that when a child learns the word for "spoon", for example, then, and only then, the child recognises that the spoon
continues to exist even when he can't see it. That is rubbish. This awareness is nothing to do with language. It is clear, for
example, that animals recognise that things exist even when they can't see or hear or smell them. Think simply of a dog who
goes to dig up the bone he buried. It is also pretty safe to assume that a child knows long before she talks that her mother
continues to exist.
Translating from one language to another shows clearly the difference between language and thought, particularly perhaps if one
is translating into one's own language. Practically everybody who has tried translating knows that one cannot translate direct
from one language to another. One first has to make that conversion from the original language into ideas independent of
language that I pointed out in Chapter 4 (p.90). The difficult part is then finding what one feels are appropriate words in one's
own language to describe what one is already conscious of perfectly clearly.

Page 198
Experiencing translating is in fact only another way of bringing one to the realization of what understanding actually is. The
essence of it is that one has to convert language into something else. And using a language is converting the something else into
that language.
This truth is illustrated in a slightly different way by an experience I was once aware of. I was looking at a gramophone record
and wondering if the stylus pressure on it was too great. I tried how I might say it in Swedish (a language I once knew well). I
was a little hesitant. But I decided on something that seemed fairly satisfactory, and then realized I did not know, had not
decided, what words I would use in English for what I was thinking. I had been looking at things and my ideas about them were
completely clear in my mind, but I had to find the language for the ideas, although the ideas were quite ready; and I had not
needed my native language, English, to think while I was deciding what to say in Swedish. I have met several people who have
told me how they have had basically exactly the same experience: trying to find words to express something in a foreign
language, and then realizing they had not formulated anything in their own.
Again, if thought depends on language, how can people distinguish, as they do in many languages, between, say, the present and
future meanings of exactly the same form? For example, the equivalent (in most European languages other than English) of
"She comes". From the context, you say. Exactly. But context is precisely the recognition of things that are not language. In
fact, the same problem would apply to any bit of language that is ambiguous if one did not have that non-linguistic recognition.
And when the non-linguistic recognition is missing, potentially ambiguous language becomes actually ambiguous.
Some of the most important evidence that thought is not language, though, must be individuals' awareness of how they think, or
at least of how they do not think. Short of directly observing thought going on in somebody else's brain, the only really proper
place to observe thought is in one's own mind. I know that I do not think in language or anything that could remotely deserve
the term language. The physicists Einstein and Sakharov have both said that in their work they did not think in language. (Nor is
mathematical notation the language of their thought. Mathematical notation was simply the language they expressed their
thought in afterwards.) Lordat, a 19th century professor of medicine, had a stroke and for a time was unable to read or express
himself properly in language. He maintained that there was no change in the working of his innermost intelligence, and that he
was certain he could think without the use of words. He could grasp, for instance, all the ideas relevant to the Holy

Page 199
Trinity, although he could not remember a single word. Since his time many others have repeated that evidence that their
thought was not damaged by the strokes that damaged their ability to use language. Such personal evidence does not prove
conclusively that thought is independent of language. A person may think she is thinking normally when she is not, and
precisely because she is not thinking normally. Yet the evidence does strongly suggest the separation of thought from language.
It was looking at the way I think myself that led me to see clearly what is the most absolute and most decisive gulf between
thinking and language.
Language is a single straight line.
Language can never be anything else, whether spoken or written. One word must follow another, and there is no escape from
that one miserable dimension.
Thinking is more than even three dimensional, for in thinking, two or more things can be in the same place at the same time.
I believe anybody who thinks without prejudice about their thinking for a few moments will know that this is so. And if you say
the more-than-three-dimensional thing that thinking is is a sort of 'language', you are forcing it into something it is not, forcing
it into preconceptions of what language is like; you will think of thought skew from the moment you call it 'language', and be
suffering some of that very corruption by language that is our curse.
Think of a mechanic who is trying to find out what is wrong with a car engine. He looks at, thinks about, parts of the engine,
and at the same moment may see the working and connection of as many as four or five different parts, and draws his
conclusion about the fault. Yet the mechanic may sometimes find it difficult or impossible to express in language what he quite
clearly sees; and if he had to think in language, one 'word' after another, he would probably never discover a fault in any engine
at all.
To say that thought is not language nor even at all like language is not to claim that one understands how thought works, or is
inspired or 'triggered'. Yet awareness that the products of thought are not the same as the processes of thought helps in
understanding that whatever else those thought processes are, they cannot be anything like language. Is that not an essential
beginning? Even if you are not convinced by your own experience that thought is not a string of 'ideas' one after the other like
language, it is surely impossible for any 'straight line' process of thinking to organize the

Page 200
straight line of language. If one can only 'think' one item after another, one could surely not, for example, organize the sentence
That's the cat I was telling you about.
The "about" at the end is only the most obvious example in the sentence of the need for something working in several
dimensions and by simultaneous actions to help one organize the expression of the idea. Straight line thinking could not in
advance arrange the position of "about", which belongs to "the cat"; but such anticipation is essential for arranging the words
sensibly, and would be impossible without what I can only call a 'whole view' in one's mind first.
I have no idea what this 'whole view' looks like or how it works. But there must be some such thing, so that when we think of
'whole' real things - situations, if you like - we also see whole sets of meaning at the same time. I do not see how we could
possibly use language otherwise. Whole groups of words are ready before we actually use them. This perhaps explains why
speech is usually less long-winded than writing, why we stumble more often when we talk than when we write (if we are used
to writing), and why, nevertheless, many people have experience from time to time of starting to write a sentence and finding
they cannot satisfactorily complete what they have started, precisely because they have not 'overseen' the whole sentence in their
mind first. It must also be the flexibility of thought as opposed to language that allows one to 'jump about' and work out from
the context what a word one didn't hear properly must have been.
Its straight-line nature is not language's only limitation, though. Drive a car, and try to say to yourself even half of what you
think as you drive, in the same time as you take to think it. You can't, but that is not only because of the 'time' problem - that is,
because you can only put one word after the other. Even if you had as much time as you wanted you still couldn't, because you
lack the means to express your thoughts.
Children, perhaps some so-called primitive peoples, and many others (including in practice quite a few academics) may
sometimes tend to think that words in a way are the things they symbolize. They may think that life and the world can only be
divided up in one way, and that way is the way of their own particular language. They may think that the world of reality has
absolute categories and classification, like their language, and so they may think words and realities fit neatly and exactly
together, that for every

Page 201
reality there is a corresponding word. This may be why some people and communities attach so much importance to word ritual.
They do not realize that the relation of language to reality is very different from this. First, different languages 'divide up' reality
quite differently. Meanings from different languages overlap. And what one language expresses by several different words may
be expressed by only one word in another. And some things that one can say in some languages are ignored by other languages
altogether.
Most important of all, there are huge areas of reality that are not expressed at all, by any language: there are many, many things
nobody can say whatever language they speak. Within the limits imposed by their nature as humans, the things that humans can
experience, imagine or think are practically endless in their possible range and variety. But language, both language as a whole
and each individual language, is a finite set of fixed meanings.
People sometimes say they can say some things in one language that they cannot say in another language. This does not mean
that people who have a particular language can think things that people who have another language cannot, and that therefore
thought is tied to language, that one cannot think things one's language does not allow one to. It emphasizes the opposite:
language is limited, thought can be almost infinitely varied. Thought is not limited to the means of expression that one has for it.
The idea of 'thinking' in English, or Portuguese, or Chinese is, basically, false; even though in fact so many people allow their
language to tie, to limit their thought.
And there are indeed many different ways one can try to say the thoughts or the experience one has. This is not because
language is so rich. It is for the opposite reason: language is poor, there are only a limited number of formulas which it has at its
disposal, and one may just as well use one as another, since none of them can come anywhere near the reality of the original
thought. It is as if there was an expanse of country, where one can wander on foot, or take the train. The land seems covered by
a pretty good network of railway lines. But in fact one is very severely limited in the places where one can go by these lines,
and one has to go from place to place in a prescribed order. On foot one can go to an infinite number of places, and range about
in any order one chooses. There may seem to be many different ways I can tell you about my Cornish duck; but really the
possibilities are very limited, and I have to take my pick from those presented to me. I cannot create exactly what I want. In this
deeper sense using a language is indeed making choices - poor ones. But if language was

Page 202
ever an accurate reflection of thought, or experience, there would be no choice. One would unhesitatingly use the 'right'
expression and no other.
It is true that I often turn my thinking into words afterwards. Language has become a strong habit inside many of us. Probably
in none more so than academics. There is a kind of nervous twitch that converts everything into words. In my own case I usually
find I am turning my thoughts into language only because I want to work out how I am going to express them, tell somebody
else about them, use convincing language, since language is all I have to convince with. But this link in the mind with language
does not mean that thought is or depends on language. One should not be confused by the fact that the brain can work very fast,
and turn thoughts into words so quickly sometimes that many would-be observers may be unable at first to distinguish the
thought from the language that follows. And of those determined come what may that thought should be based on language,
some may not be able to reach awareness of what is really their thought at all.
With almost every word I write here I experience the truth that thought and language are two distinct, in fact contrary, things.
The thoughts in my brain are completely clear, and I wish I could convey them in their full clearness. As it is, I struggle
constantly to find words to do them justice. Sometimes a thought is so clear, so simple, so obvious that I feel it must be very
easy to express it; but when it comes to the point I cannot do so. There is no creative joy in this sort of writing. The clean sight
of a truth is distorted, is done violence to. There remains only a sham, a caricature or just a vulgar, shabby reflection of the
original.
And that poor straight line of language makes organizing a book - and this part of this book especially - very difficult. I need to
show how my argument, my reasoning, my picture, my vision is made up of interlocking pieces; the pieces hinge on each other
in their more-than-three dimensions and need to appear all together, all at once. So for this reason too neither my arguments and
perceptions, nor anybody else's, can ever be seen as they really are. Both words and themes have to be strung one by one on
their long, long, long but pitiful string, and the depth is lost and the emphasis is twisted. Something has to come first and
something has to come last, and all the other things in between in their single file queue, stuck powerless between their
neighbours. Changing them round might make some things better. It'll make other things worse. So please bear with me, and
depend not only on my string of words but also on your imagination and your insight.

Page 203
In thinking, then, there is no word order, or at least nothing remotely resembling linguistic order. So there goes one of the most
important elements in what linguists call 'structure'. And although thought is so much richer than language, there are almost
certainly several things that languages need which thought doesn't. I will here only mention one group of words which I believe
have no counterparts in thinking simply because they are quite unnecessary to it. They are all what I might call 'reference'
words. For example, personal pronouns ("he, she, them, you" etc.), the articles "the" and "a", and relative pronouns like "who''
and "which". All these words are only necessary for telling things to other people. They are mere indicators, so to speak. I have
no need to think "him" or "it" myself, because in my mind I see directly the people or things to which these words refer. I do
not have to explain to myself what I am thinking about.
So normally I do not think anything that those words are associated with. I may think about what are associated with "men";
not, however, about "hims". But when I talk to somebody else about somebody, I have to use "him" to make sure my listener
knows who I am talking about, unless I am going to keep repeating a name.
Exactly the same principle applies to the articles and relatives. And this may be why it is just such types of word that are
missing from some languages. Many languages are without articles; maybe their speakers felt no need for references that they
did not think. And Japanese, for example, has neither articles nor relatives. On the other hand I imagine all speakers feel the
practical need for the references to people that pronouns provide.
Think, too, about pairs of words like "(I) saw it, experienced it, wrote it, entered it, stroked it, approved it, left it, frightened it,
questioned it." The 'general' meaning relationships between "it" and all the verbs above are apparently the same. But if one
thinks carefully about them, the relationships are not at all the same. In real life "it" is not an object in the same way for each of
these verbs; on the contrary, every verb or almost every verb above 'treats' "it" differently from all the others. On the other hand,
"heard it" perhaps involves the same relationship as "saw it"; "suffered it" as "experienced it"; and "made it" as "wrote it". I do
not know if the use of all the verbs as if they did have the same relationship to "it" is because of a common human logic which
insists that nevertheless there is something in common between all the relationships, or whether it is the result of the inevitable
weaknesses and limitations of an invented system, a system invented to try to cover the enormous variety of life. I think
probably both. Certainly there appears to be a common human idea of subject and object, for instance. Certainly too the fact that
languages vary

Page 204
in the way they make verbs 'treat' 'objects' suggests they are trying various solutions to the problem, the problem of varying
'reality'. It is not always the same verbs, that is to say, equivalent verbs, in different languages that take 'direct objects'. Far from
it. Many languages demand prepositions to govern the object after many verbs that in English take direct objects, and vice versa,
and so on. French and other languages, for instance, "enter in it" ("entrer dans"). The impression is very much of languages as
invented systems that 'make do' and come nowhere near the reality of experience.1
Could we ever understand more about our brains through studying language?2 As thought is not language, surely the answer
must be that we could not. We cannot even discover anything about the processes our brains use to produce language - any
more than one could tell how a carpenter made a bed only by looking at the bed. Many people could tell how he made it by
looking at it, but only because they already knew something of the processes of carpentry and could then recognize them in the
bed. But a Martian completely ignorant of Earthling carpentry would discover nothing about that carpentry however hard he
analysed the bed.
Chomsky is self-contradictory about the matter. He thinks language is "a mirror of mind". He says that by studying the
structure, organization and use of natural languages we may come to understand the specific characteristics of human
intelligence, to learn something about human nature. Yet he also constantly emphasizes that language is a quite separate and
special faculty, a distinct biologically inherited program, a unique 'competence'. These assertions can't both be true, unless
language and thinking are basically the same thing. I hope it is clear now that they cannot be. But what is not clear, in fact, is
what exactly Chomsky does think about the relationship between language and thinking. Rejecting some of John Searle's
criticisms, he writes:
"Language, it is argued, is 'essentially' a system for expression of thought. I basically agree with this view. But I suspect
that little is at
1 Other more obvious examples of how inexact languages are were given - unwittingly - by Smith and Wilson (see p.30
above); in these the inappropriateness of the verb-object relationship is so clear to speakers that there is no passive
equivalent. Languages also suggest they are invented by the way some lack certain ideas expressed by others; many, for
example, lack equivalents of the English "-ing" form, or articles. Perhaps even something like the Spanish use of the
preposition "a" before objects that refer to people is evidence that language is invented. This Spanish habit is quite a basic
modification of the usual human view of objects; maybe, indeed, a demonstration of language being controlled originally
by attitude.
2 Studying the effect of factors like brain damage on individuals' use of language is another matter.

Page 205
stake here, given Searle's concept of 'communication' as including communication with oneself, that is, thinking in words.
We do, I am sure, think without words too - at least so introspection seems to show. But insofar as we are using language
for 'self-communication', we are simply expressing our thoughts, and the distinction between Searle's two pictures
collapses." (Reflections on language, 1976, p.57. My italics.)
By "thinking in words" he may mean that we think by means of words. I repeat, that cannot be. But perhaps "thinking in words"
is another way of saying "expressing our thoughts''. Expressing our thoughts is surely not the same as mirroring them, exposing
them to view so that we can see their nature and how they work - unless he means that language is to thought as a tape is to
music, in principle an exact representation of the thought. That would still not explain why we should ever have to express our
thoughts to ourselves through language. He can surely not mean that we do not know what we are thinking until we have put it
into words.
And if, as Chomsky says, some thought is independent of words, how do we tell how much and what sort of thought is and is
not dependent on language? How do we tell what is mirrored by language and what is not? Why should any thought be?
Chomsky also says there are restrictions on human thought. Of course there are. We cannot think what our nature doesn't allow
us to think. But we shall not learn about those restrictions from language, even if only because of that one-word-after-the-other
line of language. One will have to look at thought, not language, to find the restrictions. Language cannot be a representation of
human thought, and cannot tell us anything about it, because of this basic disability.
In any case, I suspect we can never find out any more than we know already, through language or anything else, about the part
of thought we call logic or reasoning. We already know in a sense the way our logic works. If we didn't we couldn't use it. But
we do not understand its nature. We could only do that if we could stand outside ourselves and know other ways of reasoning,
and we can no more do that than a person born blind can conceive what it is like to see. No amount of analysis and theory-
making will ever help us there. We cannot gain any greater objective understanding of our own logic, or of the restrictions on
thought Chomsky mentions, because the only instrument we have for that understanding is the logic itself and the restricted
thought itself. It would be like trying to look at one of our eyes with that same eye. Certainly there is nothing significant

Page 206
we can learn about our logic by analysing the expressions in language of that logic. We might discover - to imagine an example
less trivial than many could be - that the past participle idea is universal1 and decide that therefore there was a common human
logic principle behind it. But that could not in any way give us insight - more than we have already - into that logic. Or if we
concluded that the "that" idea, as in "He said that...", is expressed by all the peoples of the world, we should not be any wiser.
I do not believe, either, that thought can ever be properly expressed or described - any more than say music can, for like music,
thought is simply itself.
On the other hand it may be possible, I suppose, to find out more about the processes of the brain that operate that logic, so to
speak. Again, though, that can only be done by looking directly at the brain, not by looking at the language that is one of its
expressions. Language is entirely dependent on thought; thinking is basically wholly independent of language.
What we do need to understand, and what I think we can understand better, is the psychological relationship between language
and thought, how they influence each other. I don't believe this will be done by academic work, and in any case it should not be,
because it is something everybody is personally and intensely involved in. It is important to have a practical 'everyday'
understanding of what thinking is, or at least what it is not, if people are to be able to think better and straighter and freer. Yet
straight, free thinking is itself so often the victim of that relationship between language and thought.
That humans are unaware of how they think, and unaware of what sets off any particular sort of thinking, is very obvious. It is
strange that philosophers apparently did not realize this fact for hundreds of years. It is hard to see why they were so
unperceptive. It could be that once again they were constricted through 'thinking' through language; that instead of thinking
about reality direct, they limited themselves, in what they imagined or perceived, to the associations made available by language.
Concepts are already themselves preconceptions, prejudices about the world. I suspect that using them and worrying about them
stops people looking at what really goes on; that trying to analyse and discover the nature of things like "thought", "knowledge",
"mind", "meaning'', "intention", "consciousness", "cognition", "representation", "symbol", "information" is no
1 It is not in fact.

Page 207
more than a pointless and meaningless debate about words, blinds enquirers to what is actually happening, and is itself a very
good example of the corrupting effect of language.
Even words like "what", "how", "why'' may be distortions. There are moments when I sense that the important question about
the 'nature' of 'thought' itself, for instance, is none of them. "Why" is ultimately a sort of 'religious' question; "how" may be an
interesting question, but when answered only reveals processes, does not solve the basic question; "what" is what it looks like,
consists of- this too leaves, for me, the fundamental question unanswered, a mystery about thinking, as of the universe, that will
probably never be truly understood. But something like what is suggested by the word "understanding" or "insight" is perhaps
what we really need and should strive for.
So far at any rate the debate about "thought" and "knowledge" and "meaning" and all the rest appears to have been a debate that
has got nowhere and looks suspiciously as if it is endless. What is perhaps worst of all, though, is that the abstractions debated
and the abstractions used to debate them will make most people think that it is only the professional clever persons at
universities who are capable of thinking and pronouncing on the subject of what goes on in our heads. Again it should not be so,
and again I believe almost any interested human being can usefully enquire into the problem.
Make the experiment of trying, as far as it is possible, to look at some of the facts that one knows, or can know, about what goes
on inside one, without using any of the sort of words I have mentioned. Or perhaps I should say, let me make the experiment,
because almost certainly some of the things that happen inside me are different from those that happen inside you.
Introspection will of course not tell me everything that goes on. It may only tell me a tiny part. But the fact that humans are
largely unaware of what goes on in their brains should not lead to a quite illogical decision to take no notice of what they are
aware of. These awarenesses are facts; some or most of them may be unnecessary extras, that is, awarenesses of things that do
not have to be there, accompaniments, only, of other things that go on which are the real things that make humans work as they
do. But the awarenesses, of necessary things or not, are there, and should not be ignored. Also, there are various truths about
what happens inside one that reason says must be so.

Page 208
In what follows I shall always be talking about things inside me, not words, unless I say otherwise, but I obviously have no way
of referring to those things other than through words.
The thing inside me that is expressed by the words "The cat walked towards me with his tail in the air" is a picture.
Many philosophers, it seems, have scorned the idea that one has inside one pictures, or resemblances, of things in the world. But
the objections I have seen are not impressive.
Wittgenstein1 apparently thought he had refuted the idea as follows: Form an image of a man climbing a hill with a cane.
Wittgenstein then says: How do you know what that's a picture of?. You answer that it looks like a man climbing a hill with a
cane, and Wittgenstein replies: Yes, but it looks equally like a man sliding back down the hill, dragging his cane after him.
The answer to that is quite simply that one's image is a moving picture: with one's inside eye one sees the man climbing up, not
sliding down. And further, if it wasn't so, one could not be understanding the original instruction to form an image of a man
climbing - one's understanding is, in fact, making that upward picture in one's mind.
Descartes apparently said that, for instance, he has an idea of a one thousand-sided figure, but that having that idea can't
possibly consist of having an image, because he couldn't tell the difference between an image of a one thousand-sided figure
and, say, a one thousand-and-two-sided figure. Yet the two concepts are quite distinct.
Well, for a start, one can't distinguish two such different figures in the world either. One cannot look at more than about ten
objects at a time and know there are ten - that's about my limit, anyway, even if the objects are close together - and I think most
people would count even ten one by one to make sure. I am not experienced as regards crowds, so I couldn't tell the difference
in real life between a crowd of fifty thousand and one of a hundred thousand. But with my inside eye I can see a picture of lots
of people with the number 50,000 attached, if that's the number I'm told - or a figure with lots of sides with the number 1,000 -
or 1,002 - attached. Just like that, more or less as the numbers are written on the page. For that is how I believe most people, at
least, experience numbers, both in the world and in their minds, that is to say as written symbols. I imagine that those that
cannot read 'see' inside them tens, say, grouped into hundreds, and hundreds grouped into thousands etc. very much like an
abacus, and most
1 See the note on p. 195.

Page 209
probably have a very vague 'picture' of numbers altogether, unless in fact they do use some system of signs. You may well deny
this and say that you, at any rate, do not see numbers 'written' inside you. Yet I think there is no doubt that some people are
much more aware than others of what is going on inside them. If you believe you do not 'see' numbers I will have to suggest to
you that you do, but don't realize it. Because there is nothing else you can do. Numbers are abstract in a sense that cats, say, are
not. There may, of course, objectively exist somewhere a community of 485 cats. But you and I cannot be aware of that
particular number of cats until we see a symbol of it. In fact, in one way the number is the symbol.
What is interesting, in fact, is that one's inside eye can do so much more than the outside eye. If I look at books in a shelf I can
see only their backs. But with my inside eye I can see, as well, both their side covers, and their front edges facing the wall, and
indeed, the wall itself, although all these things are hidden from me in the outside world. Cars are passing along the street in
front of the window where I am writing. When they go out of sight I can still see them inside me. Much more: I can get a
picture of a herd of mammals dashing down the street; they are largish furry animals with noses on the end of their tails, and
they stop to eat when daffodils suddenly sprout out of the asphalt in front of them, first sniffing each flower with their tail noses.
So my inside images are very much richer and fuller in some ways than the ones I get through my outer eyes. Yet at the same
time they are undoubtedly in some ways vaguer than the 'outer' ones.
I know that part, a very large part, of what goes on inside my head is what can only be described as pictures. I am quite clearly
aware of this. It is foolish to say this is a mere illusion. Such an 'illusion' is nevertheless there. I do not claim to know how those
pictures arise. I do not claim that they are themselves the basic elements of what goes on in my head. I do not, any more than
anyone else, understand the principles by which they serve, say, what is known by the word "reasoning", or anything else that
happens inside me. And I cannot describe, even to myself without words, the exact nature of my pictures. They are rather like
those spots in front of one's eyes that slink off to the side immediately one tries to look straight and clearly at them. And one of
the several mysteries is where I get the images from that my outer eyes do not give me, that only my inside eye sees.
But back to the thing inside me expressed by "the cat walked towards me with his tail in the air". It is a picture, a moving
picture. There is no doubt about that, and it does not matter whether the words are mine and the picture is inside me before the
words, or the words are somebody else's

Page 210
and the picture is inside me after the words. I would say that such a picture is one sort of thing at least that is absolutely
unavoidable, has to happen to everybody. No picture, nothing to express; no picture, nothing understood.
It may not be an accurate picture, and probably some people's pictures will be more 'lifelike' and detailed than others'. It will be
nothing like a photograph (it will have more than two dimensions for a start) and in each person it will have different details, the
details of each person's prejudices and experience. But picture it surely has to be. If it is not, then what is inside me is not about
the cat with its tail in the air. I do not know how I get that picture. But it is no good saying that I don't have a picture, only a set
of things in place of the original cat with his tail in the air. A gramophone record is no good to me till it is turned back into
sound. If the experience is seeing a cat with its tail in the air, what is inside me must be about that experience. What a blind
person has inside her is obviously something different - it must be about what she experiences, which is no doubt something to
do with touch and awareness of space.
If there is still any doubt about this, then one can think about what happens inside one with sound - tunes, for example. I don't
think anyone would deny that one can have tunes inside one. (I can even have instrument or voice tones inside me and that
surely goes for most people.) The actual sound 'outside' is replaced by 'sound' inside one. Words are not involved at all. What is
inside one must again be like the experience.
But back again to the cat who walked towards me with his tail in the air: even here there is already a mystery. What about the
time in it? The difference between past, present and future is inside me, but how? The things expressed by those words, or by
"last week, this week, next week", are perhaps positions in pictures, maybe different positions for different people. It's another
matter with "walked", and the other things "is walking", "will walk". I don't seem to be aware how these different things are
different inside me.
I like the cat. The cat likes me. I suspect that whatever goes on inside the cat is very much the same, in principle at least, as
what goes on inside me. But if someone says "The cat likes her", or I decide that as he likes her he will probably jump up on her
lap, then again what is inside me is undoubtedly at least partly picture. What happens, though, at the element of "likes"? It is
clearly not a picture of the same sort as the rest of the picture, or like the cat's tail in the air, for instance. Yet strangely I wonder
if even the things (not the words, remember) inside me connected to "likes" or, equally, "nice", or ''hope", or "interesting", say,
aren't some kind of picture, moving picture, too. I certainly have the impression that they can be -

Page 211
pictures of body movements, movements inside bodies, purring, vibrations, smiles, looks, and so on. (But it gets complicated
when it gets to things like looks; does a look contain things that are not themselves pictures, or are they simply more pictures
within a picture?) These pictures could be unnecessary accompaniments. Yet they could also suggest that many more sorts of
things than commonly accepted are 'seen' inside us, in the same way as the cat's tail in the air, even though the tail is something
we can see 'out there' with our outside eyes, and "liking" is not. It is interesting that languages have so many expressions of the
kind "My heart went out to him".
It is important to remember that there are two quite different things happening here. First there is me liking the cat inside me,
and the cat liking me inside the cat, and, as I have suggested, those are probably pretty much the same thing. Then there is
something that happens to the liking inside me, and that I have suggested may be a picture. Picture or not, does the same sort of
thing happen to the liking inside the cat? Possibly not, yet something must happen to the liking inside the cat when, for
instance, it leads to him jumping up onto a stranger's lap, something that he doesn't do every day.
One other thing is perhaps worth mentioning here. As I have been arguing what I have written - the distinction, for example,
between the 'liking' itself and what happens to it - I have been looking, quite clearly and unmistakably, at pictures inside me. My
argument, if it has not been the pictures themselves, certainly does not begin until the pictures appear inside me. (But then, in a
case like this, I looked carefully at what seems to happen, but what made me see what I saw?)
There is already a mystery about what happens inside me when someone says "walked" as opposed to "walks" (or that makes
me produce the word "walked" rather than "walks").
Perhaps a clearer example of the same sort of mystery is what happens if I decide I love the cat so I would try to protect him if
he was in danger or fear. A lot of this I see as 'picture'. But what I am not aware of as picture are (among others) the things I
have referred to as "if" and "would". The things "if" and "would" are not outside me like the cat. Nor are they inside me in the
same way as, say, the things liking or loving. But they are entirely mine. And while one can say that what happens inside me
when I tell you the cat had his tail in the air replaces the actual cat and his actual tail, and even that what happens inside me
when I am aware of my liking of the cat replaces the actual liking, or at least is not the actual liking itself, one cannot say that
about ''if" and "would". If-ishness and would-ishness are

Page 212
only inside me, only come from inside me, and are themselves alone. They do not replace anything and are not replaced by
anything. (And these particular two, "if" and "would", work on things that don't exist, yet which are things that I can see inside
me.)1
This I think may be a truth essential to understanding ourselves better - that is, the truth that there are things inside our heads
that are not instead of something else.
Mathematicians, for example, may work on things that are instead of 'reality'. But the work itself is only itself, it is the reality,
and it is only inside each individual mathematician's head. Equally, the 'work' I have done as I have argued all the above is only
itself, only inside me.
Or, again, what happens inside me when I decide that it is raining now, because the cat has just come in wet, but that it can't
have started raining before supper, because the cat came in earlier too, just as we were starting supper, and was dry then? Again
there are pictures inside me, quite a lot of different ones, of the cat coming in, his dry and wet coat, sitting down to supper etc.
But there are also 'pictures' of 'positions' which I shunt about to arrive at my conclusions. These 'positions' are tied in some way
to the pictures of the cat etc. Yet they do not replace anything outside me. And then there is inside me the thing that actually
does the shunting of the 'positions'. I am not aware of this thing, or at best it is another of those spots before the eye that slink
off when one chases them. One thing is plain, though. This 'shunting' too is entirely mine, entirely from inside me. Once more it
is not in place of anything. It is only itself.
People can know any amount about 'information' processing, about how 'messages' are shunted around in the brain, but that
won't explain the process of any particular sort of shunting (if-ishness, or deciding it must be raining, etc.), let alone what
inspires that particular sort of shunting. Thinking that it will is like thinking that if you know how a telephone works you also
understand the nature of the being making the telephone call and why it makes it.
I am not trying to argue for the abstract in contrast to the physical. I suspect it is a false distinction. For a moment I must
perhaps, reluctantly, abandon my resolution not to use conventional abstract terms and say that I am trying to make clear,
among other things, that thinking itself has nothing to do with symbols.
1 I think the cat must sometimes have something inside him very like this if-ishness and would-ishness. If he wants to go
into the road but sees the local tomcat bully waiting just outside the gate, something happens inside him to make him trot
off along to the hole in the fence which he knows about, but which he can't see from where he is just now.

Page 213
An old basic problem may be arising in some people's minds round about this stage. What does the shunting? What is inside me
that appreciates the results of the shunting, where does that come from? And so on. In the same way, what sees and appreciates
my inside pictures of the cat, and so on, endlessly - what observes whatever observes the picture, and what observes that
observer...? This problem worries some philosophers and the like. But that it is a troublesome problem is no excuse for saying,
as some do, that it is a false problem, that things can't be like that, and therefore the solution is to be found somewhere else,
such as in an analogy with computers, where different functions are distributed among different parts, and there is no one central
observer. Inconveniences don't go away just because they are inconvenient. And this problem of endlessness inside people is
only part of the basic .riddle of endlessness in all existence. In principle it is exactly the same as the problem of the ultimate
constituent of matter. Certainly the endlessness problem is not just tied to observers of pictures or representations. It will arise
whatever one claims ideas or experience or thought to be. Whatever they are they still need an 'appreciator'.
I am not quite sure that the endlessness problem applies here anyway. When the cat basks in the sun, his body receives the heat
and reacts to it. His body has to be something quite complicated, perhaps, to react in the way it does. But one does not demand
something further which is to react to his body. His body is just itself, and reacts in its own special way. So, surely, is and does
the thing inside me that appreciates and reacts to - and, often, produces - the 'pictures' and many other things that are
'appreciated' inside me. My 'appreciator' may be much more complicated than the cat's body, or at a different 'level'; but is there
any more reason than with his body to suppose there must be something else behind it, an appreciator of the appreciator?
Physiologists understand something of the nature of the cat's body, and do not ask for more beyond. One does not have to ask
for more beyond the appreciators in our heads just because we understand nothing about them.
In any case, the idea of several interacting sub-systems, on the computer model, each with a special task, is quite irrelevant to
the problem of an endless series of 'observers' or 'appreciators'. If that really is a problem, it remains just as much a problem
how something happening in part of me gets observed or appreciated as how something happening in 'all' of me does. And in
fact it has been obvious for rather a long time, without any computer analogies, that different things happen in different parts of
us. My toes, for instance, are not involved in appreciating music - unless I dance. The

Page 214
question of whether there is one central me that everything ends up in, or just several me's linked together, but each with its own
function, is quite another one. But for many people now, structure is an obsession. Suppose and describe a structure, and one
has, they believe, solved everything.
Where does all this that I have described as going on inside me get us? In one way not far at all. Basic mysteries remain, as
difficult as ever. But I hope that looking at things in the way I have perhaps helps to free one from the fixed and biased line of
sight that conventional terms can bind one to. There are so many different things that go on inside me. Almost certainly some of
them do not go on inside you, while you have others that do not go on inside me. But even if your and my 'inner' experiences
were all exactly the same, it would still mislead to lump them all together under the word "thought". Immediately one gives
things names one is in danger of distorting reality and forgetting what actually happens. It is, of course, much more inconvenient
to discuss things without names. It takes much longer - and forces one to think really properly. Think properly about the only
thing that matters, not about definitions, but about that thing I have just said: simply what happens.
There are present day neurologists and psychologists who very sensibly warn against assuming the existence of whole distinct
general faculties like "cognition" or "thought". Unfortunately, though, they do not apply this wise distrust of preconceived
categories to the concept of the language faculty. That there are distinct linguistic skills (if not just one language function) they
do not appear to doubt. Some of these language functions may overlap with 'thinking' or 'cognition' functions, they say; but they
seem to take for granted that there are special abilities in the brain just for language.
This mistake is probably closely connected to another basic mistake that as far as I know practically all the experts make. They
disagree among themselves on the relationship between language and thought, but seem to pose quite the wrong alternatives.
The neurologists, the psychologists, the philsophers, seem to think that the argument should be to decide whether there is a
connection between language and thought, or no connection - and, if there is a connection, how much connection.
But those are not the logical alternatives, because putting the question like that assumes that there is a distinct language faculty
and that that faculty is self-supporting. Yet it is precisely whether there is such a distinct self-supporting faculty that ought to be
at issue.

Page 215
I have already tried to make it clear, I hope convincingly, that thought cannot depend on language, because they are so different
from each other, and thought is so much more powerful and varied and flexible and must in any case come first. On the other
hand, the use of language must depend on particular parts of the brain. Perhaps there are alternative parts that can be used for
language. However that may be, it is obvious that, unique or not, there must be parts somewhere that language depends on. That
applies to everything we do. It is important to emphasize "depends on" as opposed to saying that the "language faculty" is
located in this or that part of the brain.
In The Shattered mind,1 a fine, fascinating, but sad book on the knowledge about the brain that has been got from studying
people with brain damage, Howard Gardner writes:
"First of all, it is clear that not all cognitive activity is dependent upon language, else severely aphasic patients [patients
with language disorders] would fail the full range of problem-solving tasks. By the same token, however, linguistic and
cognitive processes cannot be entirely divorced, for a high correlation has been demonstrated between specific language
disorders - for example, difficulty in carrying out logical commands - and specific nonlinguistic deficits, such as
performing calculations, or finding one's way about a maze. Whether an aphasic individual can perform the highest-level,
most demanding intellectual tasks is debatable..."
This must be true. It must be true because language simply cannot be produced without what one can call 'thought', or 'cognitive
processes', but what it would be much better simply to call work of a certain kind by the brain. The brain has to work in
particular ways to produce language! The evidence Gardner offers in no way means what he implies: that thought -or at least
some thought - depends on language. But it does mean the opposite. It is practical confirmation of what simple logic tells one is
almost certainly so: there are various general faculties, certain sorts of 'thinking' if you like, that are necessary for doing a
number of different things, among them calculating, maze travelling, and using language in various different ways.
It is strange that this has not always been obvious to enquirers, and that they should ever have conceived that it could be the
other way round -that 'thought' depended on language. I can only think that this blindness is
1 Vintage Books, 1976, pp.98-99.

Page 216
the result of the age-old reverence for language from which most humans suffer.
I have already pointed out that it is obvious that humans inherit the ability to talk. But talking ability is not the same sort of
ability as, say, the ability to move the fingers in precise and subtle ways. That ability is special and inherited direct. Then, one
may use that special - but in another sense, general - ability for the purpose of playing the piano; or picking pockets. The same
is true of the special but general ability which humans presumably inherit to turn experiences or ideas into outward signs. All
those other special yet at the same time general abilities of hearing, sound making, memorizing, logic, organizing, co-ordinating,
are necessary for a person to be able to use language and even to be able to invent it in the first place. The symbol-creating
power, together with an urge to express experience, was perhaps enough to lead people to invent language. It is the only thing
that could remotely be called a language faculty, and by itself would be useless, it could bear no fruit. And who knows. Perhaps
even symbolizing is a human invention too, rather than an inherited faculty.
Language is not a faculty, only an expression of a number of true faculties that combine in a particular way to produce this
particular result. Just because language may sometimes be the only outward consequence that one can observe of what goes on
inside certain areas of the brain, one should not be fooled into thinking that it is the only and basic activity of those areas.
Further, the surface appearance of language may give the impression it is one single and separate phenomenon. One could
believe the same of cooking for example: that is, that cooking is one single self-contained faculty - and uniquely human, which
of course is true. Yet obviously cooking is not a special separate faculty. Many general faculties are required for it, not least a
strong ability to have inside one things that aren't, an ability one also has to have for using language.
There is apparently already plenty of evidence that more than just the ability to use language depends on the same collection of
faculties. Gardner mentions some in the passage I have quoted above. Each ability that suffers when there is a breakdown in the
ability to use language in some way is evidence, not of the identity of language and thought and of a grand linguistic faculty
with broad authority, but that the same basic processes in the brain control different 'applied' activities. One more example
should be enough here. The composer Ravel became unable to speak or write properly. And at the same time although not all
his musical abilities disappeared, he found it difficult to read music, and to perform it (including

Page 217
his own) and to write it. He could no longer compose. Equally, though, there have been many cases of people who suffered
from severe language difficulties who could still draw or play an instrument or even compose. (Gardner quotes the case of the
Russian composer Shebalin.) And finally there are the people who still use language perfectly well but are no longer able to
draw or do mathematics or various other 'cognitive' tasks.
The way the brain works is clearly very complicated, and nobody really knows much about either the details or the basic
principles. But one can sum up one picture provided by the evidence and by reason very simply. There are a number of different
'thinking' processes in the brain; some of them are necessary for using language, and some of them are not. And one should
probably add that some of the processes actually used for language -and no doubt for several other abilities - are probably not
all the same in all individuals. It may even be that a process necessary for one individual to use language is not necessary for
another. There may be factors which make different individuals develop different habits that are in practice unbreakable for
certain aspects of their use of language.1
However, it does not follow automatically from what I have just been arguing that language is an invention. Yet again I think
elementary evidence and reason make it clear that it is.
Chomsky has referred to what he supposes to be a 'numbers faculty': the "capacity to deal with abstract properties of the number
system - and that's a distinctive human capacity, as distinctive as the capacity for language" (Magee, 1982, p.185). Yet he
himself emphasizes that through most of human evolution it was impossible to know that most people had the ability to handle
numbers, or, as he puts it, "deal with [their] abstract properties."
This I should have thought gives the game away and argues against specific distinct mental faculties and for the use of general
ones. If humans have an inborn special numbers faculty, why were they not all using it from a much earlier time? Because - and
there is surely no doubt about this -numbers were invented. And when they were invented (by a process I suspect had much in
common with the invention of language) humans found they had the general faculties to handle them, just as they had the
general faculties to handle other inventions: music, say, or football, or bicycles - and language. It is surely clear that music,
numbers, football and
1 For example, my own habit of 'seeing' words in my mind as I speak or listen - at least in certain kinds of conversation.
Or the way some people have to 'mouth' the words when they read.

Page 218
bicycles are all human inventions. We are not born with them inside us, but we are born with the ability to use them.1
Further, words are the very stuff of language, more basic to its nature than anything else. But can there be any doubt that words
are invented? Every language with its unique set of words must be simply a convention for expressing human experience. If
languages were not conventions there would presumably be just one inevitable language. It is surely impossible that all words of
all languages of all times should be programmed genetically into people. And even if they were, that would not account for the
changes and evolution that take place in words. But if words are invented, the relations between them, what the linguists call
grammar, must be invented too.
If the way I have argued about the relationship between thought and language is sound, it is really irrelevant to that argument to
know where in the brain the various faculties are, or are supposed to be. There must be some faculties of some kinds of
'thinking' in the areas that are associated with language. But my argument also allows in principle that there may be 'thinking'
faculties in areas unconnected with language. And in fact it appears that there are indeed several different areas of the brain with
'thinking' faculties not linked in any way to language; that simply confirms the difference between the two.
As to language disorders, they must be symptoms of something wrong at a much more basic level. I think it unlikely that
observation of language behaviour in either the healthy or the sick is going to lead to better or quicker understanding of the
human brain than the study of any other behaviour. It does not seem to make much sense either way. Even if language was
special and separate it would not reveal anything about the rest of the brain; language would only give information about
language. Think again of cooking. If cooking involves the use of various faculties, a deep study of just cooking is unlikely to
provide much insight into the working of the brain. But if cooking is a faculty all of its own, a deep study of cooking will, at the
very best, provide information only about the cooking faculty. To provide information only about language would be interesting,
and perhaps useful, if language was the same thing as thought. But it isn't. At least part of the trouble is probably that whereas
cooking appears as a
1Nearly everybody masters a minimum of language, although it is not true, as Chomsky and others say, that we are all
equally good at it. Not everybody masters music or even numbers; but nearly everybody learns quickly and easily how to
ride a bicycle, if they have one and start early enough. If bicycle-riding was part of an imposed curriculum, with
significant tests, but no further bicycle-riding at the end of it, it might be a very different story. The work of Suzuki in
Japan suggests that music can come to humans as easily as language.

Page 219
practical and so relatively inferior, even if much appreciated, activity, language appears to most people, and certainly to
academics, as a much more abstract activity, and so many people are deceived into thinking of it as, at the very least, a path to
the understanding of thought.
Instead, I believe that in order to understand the processes of thought in the brain physiologists and neurologists will have to
study the brain direct, however difficult that may be. Yet they cannot even begin to do that properly if they are not looking for
the right thing: that is, if they confuse thought and language.

Page 220
11
Language the Corrupter
Thinking, then, does not need language. But language needs thought, and words need experiences.
Some say that it is the other way round, that experiences need words.1 We couldn't have the experiences we have without the
categories made by language. Searle expresses it like this:
"I am not saying that language creates reality. Far from it. Rather, I am saying that what counts as reality - what counts as
a glass of water or a book or a table, what counts as the same glass or a different book or two tables - is a matter of the
categories that we impose on the world; and those categories are for the most part linguistic. And furthermore, when we
experience the world we experience it through linguistic categories that help to shape the experiences themselves. The
world doesn't come to us already sliced up-into objects and experiences; what counts as an object is already a function of
our system of representation, and how we perceive the world in our experiences is influenced by that system of
representation. The mistake is to suppose that the application of languages to the world consists of attaching labels to
objects that are, so to speak, self-identifying. In my view, the world divides the way we divide it, and our main way of
dividing things up is language. Our concept of reality is a matter of our linguistic categories." (Magee, 1982, p.156).
What Searle says here is true if it is a description of how things often or usually are in practice. But it is not basically true, and
cannot be true. Humans cannot apply these linguistic categories or representations, their names for things, unless and until they
are aware of the object or experience independently of language. I may not be aware of the same categories of things as our cat;
he may, for all I know, lump trees and poles together in the same category, while he no doubt has a whole world of smell that I
could never appreciate. But this is nothing to do with language. The awareness comes first, awareness of objective reality, even
if different creatures, and indeed different humans among themselves, are aware of different aspects of that objective reality.
Our cat can distinguish houses, dogs, mice, humans, and other cats for example. Free of language, he nevertheless has his
categories. His
1 I mentioned Wittgenstein's argument in Chapter 10, p.195.

Page 221
categories are determined by what he experiences, and his experience is determined by the combination of his nature as a cat
and the sort of life he leads. Eskimo have many different names for different sorts of snow and ice, not because they have many
different names for snow and ice (!), but because they actually experience, in the life they lead, objects which for them are
several distinct sorts of thing 'out there'. Neither Sahara desert dwellers nor the citizens of London have these experiences and so
they do not have the words for them.
This winter, in dampest Tuscany, we are getting heat from an old wood burning stove, the ''stufa". There is an iron flap which
one opens or shuts to increase or decrease the draught from the metal chimney pipe. I recollect now that there is a name for this
thing, but I cannot remember what it is. This does not stop me experiencing the thing quite distinctly, separately and
objectively! You might argue either that (1) I knew that there is a special term for the thing, or that (2) I designated it "a flap" or
"a draught regulator", and that therefore in either case I was experiencing it through a category imposed by language. I can
assure you that neither is true. Until now I had forgotten that there is a word, and I did not give the thing any name or even
description. Language did not enter into my experience in any way. This thing exists for me quite independently in its own right
- and, in this case, as something unique. I do not think even of other flaps in other "stufas".
Awareness, then, not language, determines our experience of reality in the first place. Or, at least, it does in a healthy mind.
Corrupted minds are precisely those that have become the victims of language. Nearly all humans are thus corrupted to some
extent.
So there is a whole world of things 'out there' which we can experience objectively. By objectively, I mean that we can
experience them direct, without any intervention of language. (Our experience is subjective, of course, in the sense that we can
only experience what is in our nature to experience. Humans cannot have the same same experience as moles, and moles cannot
experience what humans experience.) This ability to experience things objectively is a very obvious truth and would not need to
be stated if the language-mongers had not cast doubt on it.
It is when language comes in that things begin to go wrong, in more than one way.
Words are meanings. And meanings are not the real things they are connected to. Nearly everybody would agree about that. But
meanings do not represent things, or replace them, substitute properly for them, although most humans probably believe that
they do. Nor are meanings

Page 222
descriptions of things. Rather, meanings are invented symbols, tokens, references, associations, and so falsify and deceive from
the very first, by their very nature. Meanings do not even represent ideas, or pictures that people have inside them. Meanings,
words, are 'instead', in the worst way. The meanings are supposed to serve, to be instruments, merely, in the service of humans;
impartial. But these symbols that meanings are take on a life of their own. They do replace the real things, whether those real
things are outside us or inside us; but not to reveal those real things, make them clear, but to drive them out.
If language really reflected or represented reality, which is ever-varied, there would be no limit to the words in a language, and
people would constantly make up new words to represent reality - in fact, probably most words would be new words, and of
course it wouldn't work, because people wouldn't understand each other. So pigeon-holing had to be invented instead, an
artificial system of what one might call stylized symbols. As I have said, animals can recognise different sorts of things too, and
they can draw the proper practical conclusions from distinctions. But they don't suffer the disadvantage of having to squeeze
their recognition into the falsifying straitjacket of language the pigeon-holer, the classifier.
I have said that meanings, or words, take on a life of their own, and this is true. But it is a constricting, tyrannous life. Humans
have very exact associations with each word, to a large extent associations limited by contexts, contexts of both situation and
other words. In this situation you must use that word; with that other word you can't use this word, etc. The conventions are
rigid. Yet then most humans, most of the time, force the whole world and their own experience into those meanings, although
those meanings are inflexible, fossilized.
One has an experience, a thought, an imagining. Then very likely one is reminded by it of words, because the word associations
are there waiting; or one 'chooses' between various words or sentences that 'will do', as I did with my duck story (pp.194-95).
The choice is crude, limited. Immediately, the experience, the thought, the imagining is no longer pure, it is no longer itself and
true. The further one goes, the more one develops it, the more one tends to think in terms of the words instead of the things and
the feelings, and one forces the things and the feelings into beds of Procrustes. The process probably goes so far in many people,
particularly intellectuals, that the thoughts become limited in the first place to the words they have available. Humans become
blind to the variety of reality. They become blind to reality itself. Yet at the same time words make them able to create myths.

Page 223
The first examples I offer are in a sense rather superficial, but the confusion and crooked thinking is not less dangerous for that.
If a person asks some other people "Is Bristol north of London?" I think it is fairly certain that those other people must, before
they answer (if they are taking the question seriously) 'see' inside themselves something like a map and on that map blobs where
they think London and Bristol are. Most 'educated' people today think of things like "north" in this sort of way, I imagine,
though there might be a few realistic folk, I suppose, who would try to decide if one would have to go to Bristol tending away
from or towards the sun at noon. But even they would surely have to have inside them a 'picture' of the sun relative to where
they were. So far everything is all right.
However, I think it is very likely that there are quite a lot of people who, if asked "What is the capital of France?" may answer
"Paris" without using more than the mere words themselves. Such people have at some time learnt that the words ''France",
"capital" and "Paris" are linked, and so the words simply trigger each other and evoke little or nothing else. In a case like this
(France - capital - Paris) such automatic and unthinking association or linking of words alone probably does little if any practical
harm.
But the tendency is fateful. There are so many other combinations of automatically associated words that people use. For
instance, "the sanctity of life, great literature, human rights, linguistic structures, law and order, mob rule, sexual equality,
national liberation, the dignity of work, traditional values, bourgeois morality, class struggle, people's democracy". These are
just a few of the thousands of combinations that millions of people take for granted and use to fool themselves and most others
about the nature of the world. I will not explain why I think each of these phrases is false. But I ask sceptical readers to try to
make the experiment of thinking about what is really beyond them without using language at all, either these words or any
others. What happens to actual people? If that seems a tall order, notice what a good instrument prejudice can sometimes be for
clearing away the fog of words. I have deliberately arranged the phrases above so that, roughly, they get more controversial the
nearer they get to the end. I don't think many readers will have much difficulty doing what I suggest and seeing the fraud in the
case of "people's democracy". Many, though not so many and depending on their loyalties, will see how bogus the previous
three or four are. I ask them, though, not just to have a reflex action and say, "Oh yes, I've known all about that nonsense for a
long time." I ask them to rehearse and practice their wordless rationality and realism and see if

Page 224
they can apply them to the earlier phrases. For think how most of them would have been regarded in the past. Different
prejudices would have cleared away their fog. A hundred and fifty years ago Metternich would have mocked "national
liberation", for instance. (In view of what has happened since one should not assert he was wrong.) And there would have been
few takers for "human rights" in the Middle Ages. (Do not misunderstand me. I am not supporting the ancien régime or
medieval prejudices; only trying to show that the reality of phrases like "the sanctity of life" is not as self-evident as may at first
seem.)
Very often people associate words although they are not used together in fixed phrases. The association is no less fixed for that.
For instance, "freedom" and "democracy". How many people ever think of the reality beyond that knee-jerk coupling? And
combinations are joined in larger combinations: "Name a feature of free government." - "The people freely electing their own
rulers." Such a commonplace can only be trotted out confidently if the speakers or writers do not turn the words into something
else, perhaps at least partly 'pictures' in the head, which gets closer to reality.
A 'modern' example of the way words put blinkers on humans is the pair "terrorist" and 'freedom-fighter". Actions remain those
actions whatever is said about them. But most humans do not see this, because words have been attached. People hear of the
actions and, according to their prejudices, call those that have committed the actions terrorists or freedom-fighters. Instantly they
blind themselves to the reality of what has actually been done; neither do they try to understand the actions. The names become
an excuse merely to condemn or praise, and when they are said to others, they instantly blind the others too, and thinking stops,
and justification is given for continuing the horrors endlessly. If the words were miraculously abolished, the horrors would not
cease, because it is too late. The group mentalities, the hatreds, the prejudices are already there. But those prejudices could
never have arisen, certainly never have stuck fast, without the word-names to create and fix them. And the group mentalities
and hatreds too would perhaps never have taken the form they did. Now, as long as the names continue, the conflicts and
horrors will continue, and we shall often see, as we see today, how the actions of one day's freedom-fighters become those of
another day's terrorists, or the other way round. And it is not insight and realism, only the same blindness by word, that leads
some to discover "terrorist states". There are no such things, any more than there was such a thing as "Nazi Germany" or
"Stalin's Russia"; there are

Page 225
only particular individuals who do particular things on particular occasions - though the particular things may have a ghastly
repetitiveness.
At this point some people may be saying that this is all old stuff, we know all about that. Old stuff or not, the fraud still works
on most human beings, and the results for many of them are calamitous and piteous.
However, I do not believe it is only in the political and social fields that language works its confidence tricks; and do not think
that my purposes in attacking language are narrowly social or political. The nature of the 'broader' world around us affects all of
us fundamentally, but the way it affects us and the way our personal lives affect us depend on a psychology more basic than a
merely social or political one. It is our very way of thinking that language corrupts.
The basic corruption by language lies, I believe, in the nature of single words. I want to take a word which I hope is not
provocative and is a good example of one of the most fundamental ways in which words corrupt. "Intuition." Psychologists,
linguists, artists, people debating 'artificial intelligence', and academics generally and many others may often question what it is
and endlessly offer new descriptions and definitions of it. They admit their uncertainty about it perhaps; that does not stop them
using it in their arguments about all manner of things. That is bad enough, but is not the basic corruption by the word. The more
conscientious arguers will try to give sensible-sounding accounts of what "intuition" is. Such efforts expose the basic mistake.
This is that people assume without question that IT exists, that there is a thing, often working powerfully, and the only difficulty
is to pin it down. There is a definite truth as to its nature hiding somewhere, if only someone was clever enough to discover it.
There is a name for it, so it must exist as a distinct reality. The word has trapped people's thinking inside a circle, where it
dashes round and round fruitlessly. It does not seem to occur to them that this 'thing' survives only through the word, retains
human faith in it only through the word; and the word is so strong that people do not become suspicious when they cannot
firmly pin down its 'reality?
It is not that there is nothing. It is the opposite. It is that there is so much going on. Because of the barrier of the word people
cannot see that there is not just one fixed, immutable, strictly delimited lump sitting there waiting to be revealed. They do not
see that something quite different in fact took place. Nobody knows the details of the story, but at some time certain things have
gone on in the head of a human being. She does not understand
1 See Chapter 12, p.244.

Page 226
these happenings and even if she did she couldn't give any sort of realistic account of them; but she has language and so she
gives the happenings a name anyway. (This time: intuition.) The damage is done. It is too late to remember that the original
experience was complex and mysterious and not understood. It has acquired a bogus simple clarity. It is too late to consider that
the next time in the same person different things may happen; because the context seems rather similar the word will be trotted
out. Too late to consider that in other people the context may seem similar but different things may happen; the word will be
trotted out. And humans are left with a double evil. Not only do they not have any accurate account of their experience. As well,
the name-giving gives so much confidence that practically nobody really tries to find out what really happens. The variety of
experience is squeezed into a little tube of language. The terrible thing about words is that they limit humans' awareness of
reality, make them blind and deaf to it; in fact, words almost certainly limit experience itself.
If you believe this is impossibly gloomy, too impractical to be true, words can't work like this, then think what is demanded for
it to be otherwise. It would mean that once upon a time, in the beginning, the inventor of, say, "intuition", and everyone around
him, knew exactly what it was, really knew, really understood it. But now everyone has forgotten.
There is a variation on this kind of limiting distortion of reality and experience in words whose chief falseness is the giving of
the impression that the world is divided into contrasting extremes with nothing in between. I mean pairs of words like
"conscious/unconscious", "concrete/abstract". There is a vast amount in between. It is rather surprising that more people haven't
realized this, in view of the confusion and puzzlement caused by the constant attempts to fit all human experience into one or
other of the first pair, and the whole universe into one or other of the second. And the insight that these pairs of words do not
cover anywhere near everything should lead to the further insight that the extremes themselves do not exist in the absolute way
the words make us think they do.
Looked at from another angle one could express the corruption by language briefly as follows: People don't know what they are
talking about; but they think they do.
At this point readers may be thinking that I have contradicted myself. I have said that there can be no words without experiences
first to which the words are attached, and this seems to me clearly true, inescapable. I cannot understand "cat" or "tickle" until I
have seen a cat or been tickled; nor will any human invent a word for these things until she has experienced them.

Page 227
Yet now I say words are invented for experiences which are not at all clear to the inventor or to later users.
But in the case of thousands of words the experience is not direct in the way experiencing the cat is. One can understand the
word "volcano" without seeing a volcano. There is a combining in the head of various things that one probably has experienced
direct; hills, heat, smoke, bangs, bursting, etc. (A picture of a volcano is a short cut to combining these experiences.)
A word like "habit" one also understands, I think, by combining experiences. The cat sniffs the air carefully from the doorway
before he goes out in the morning. He does it today, he did it yesterday, and the day before, and the day before that... Here,
though, it seems to me that one has to produce something in one's head that wasn't before - neither outside nor inside one's head.
One cannot see, hear or touch habits, nor does one 'feel' them - like thirst or fright or longing. One combines the direct
experiences to produce something new. It is in one way only in one's head; yet I think "habit" is in another way firmly based in
reality. One can observe it, point it out to others, and there will seldom be dispute about it.
That is surely not true of a word like "instinct". As with "habit", one cannot directly see it etc. One has to 'create' it in one's head
by combining experiences, but this time one has to make a bigger leap, one's 'creation' comes from drawing conclusions; it is a
sort of conclusion. Or, if you like, one has to imagine something. (I hesitate to use this word, for all the reasons that I am trying
to explain right now. But maybe I will not make my thinking clear without it.)
One cannot imagine anything that is not ultimately based on experience. I can imagine - if you remember - my furry mammals
with noses on the ends of their tails. In the end, however fantastic,' my imaginings have to go back to real things that exist, have
to be combinations of such real things. Humans cannot imagine any sense, for example, which does not build on things we
actually experience. Even something like the idea of telepathy has to be based on things we are genuinely aware of. On the
other hand we can make 'false' combinations which result in imaginings of things that do not exist. Hence my tail-nose animals.
"Habit" is not the product of such false combinations. It suffers from the falseness and insufficiency that all words suffer from.
But it is not a bogus word. As the cat sniffs the morning air yet again we can say, "Look. There he goes again." And there is the
"habit". We can't do that with "instinct". That's the result of combinations too; there's more imagining needed for "instinct''; and
we can't check on it, test it for a link with reality in the same way as we can with "habit". I am not saying that "instinct" is a

Page 228
bogus word. I do not know. It is an 'imagined conclusion' from various things experienced, but one cannot observe it as one can
"habit". It is no good saying it must be real, just as it is supposed to be, because look, there is the eat stalking the bird; and birds
migrate etc. etc.; so there's instinct. That begs the question.
And then there is another type of combination of experiences that humans imagine. The combinations of experiences, first
imagined, are 'fixed' and given permanent existence by words: "nation, society, the state, fatherland, heritage, the faith, the
cause, the people, the working class, imperialism, freedom" and so on. (The experiences on which these 'combinations' are based
have themselves mostly already been distorted and fossilized into words: "soul, spirit, heart, life, country, land, love, protect,
sake, together, our, gratitude, duty, home" etc.) These are, I believe, definitely bogus. Bogus combinations, bogus words. They
produce bogus 'realities' that exist nowhere except in human brains. These 'things' are entirely the creation of human brains.
They are not and could never be experienced, not even to the extent that 'instinct' might be. I do not understand at all the details
of the way children (or adults) understand and come to use such words. But they do and the damage, usually, is done.
Language gives names and so a bogus but permanent reality to ideas that are the very reason of being of ideologies and
fanaticisms and hate. And language gives names and so bogus reality to many of the groups without which the hatreds and
fanaticisms and cruelty could not act. Without the names humans would never recognize those fearful unrealities which lead
them to commit and excuse savageries in those unrealities' name. The unrealities, existing only through words in human minds,
would never come into being. Whether the ideology or the group is religious or political or national or philosophical it could not
exist without its name. The name is a monstrous fantasy that nobody can see or hear or touch, that nobody can conclude the
existence of from the evidence of other things they can see or hear or touch, that nobody can feel as they feel love or hate -
themselves fossilized falsehoods that derive distorted power from the words; but it is a name fantasy which inspires those
feelings terrifyingly.
(Some years ago a then colleague had a group of Russian students in their early twenties who were doing a language course in
England. They complained about the West's lack of appreciation of the part played by the Soviet Union in the Second World
War: the Germans had killed twenty million people in the Soviet Union, but the Soviet Union had killed more Germans than
everybody else put together. My colleague said he was sorry they felt that killing people was the measure of virtue. They hadn't
killed

Page 229
people, they answered; they had killed Fascists. They were not amused when he pointed out that the Germans had not killed
twenty million people in Russia - only twenty million Communists.)
It is ironical that probably the sorts of word that are least dangerous, least corrupting, are those that many people would think
the most vague. the least exact and the least open, apparently, to definition. These are the words associated with mood and
sensation, at least when they are not used in some 'expert' sense: "happy, fearful, excited, gloomy, wary, passionate, lively,
angry, sensitive, warm, sleepy, sore, nervous, comfortable, bored, hungry..." Such words are inaccurate and inadequate, as all
words are, but in practice they are normally innocent, precisely because almost everyone senses indeed that they are subjective
words, not absolutes. So they are among the small proportion of words that people are in fact suspicious of, not to be trusted, or
at least not to be taken as solid bases of action. Part of the irony is that mood and sensation are supposed to be very difficult for
writers to convey well or realistically. But the words involved are nearer to what humans should be concerned with when they
think and try to appreciate and understand: what actually happens, what people do, what they feel.
There is no such innocence in words like "good, bad; right, wrong; beautiful, ugly; wicked, virtuous; dirty, clean; valuable,
worthless; deserving, guilty;..." These are as inadequate and untrue as any word. The real world is not divided up into words like
that. Again humans should look instead at what they and others do and feel - that is where the truth lies. And I do not think it is
fanciful to suggest that one way to clear the brain is to consider the reactions of animals to the actions of others. What, without
the 'benefit' of language, does the stray dog feel about it, what does our cat think about it - and what, for that matter, does the
mouse, prey to our cat, want to happen? (Morality is not always simple, even without language.) How will the wolf, or the
sheep, the whale, the sparrow or the starling react? If one can think of these things one may get into the direct presence of
reality.
The fateful paradox is that words of this sort are absolute, and humans know exactly what they mean. Uncertainty about their
meanings is not the problem, and there are few things more pointless than 'philosophical' dispute about their meanings, or the
suggestion that their meanings are 'relative' in some way. The real trouble is that so many of the wrong things are attached to
them, often with consequences of great suffering. I mean, simply, that the wrong things are called "good" or "bad", or
"beautiful'' and so on. It is usually not enough to say that people should be rigorous and make it clear to themselves and others
who or what something is "good" etc. for, though everybody ought to do this. People will not, in general, do it. So long

Page 230
as the words are there, available and respected, they will be used, they will have the wrong things attached to them, and there
will be intolerance, strife, control of young and old minds, fanaticism, war and torture, just because there is dispute about what
things are good, beautiful, right, and all the rest. It is the words' very existence that is the problem. Humans rally round them
and use them to egg themselves on, reassure themselves, justify themselves, fortify themselves, or as the basis for pride and
arrogance. The tragic irony is that they would need no reassurance if the words were not there to start anxieties and assertions in
the first place. The only escape from their tyranny is a universal distrust of them and all the rest of language.
It is striking that humans are the only mammals - with one possible exception - that have language, and no other mammal has
caused so much suffering.
The one possible exception is the whales. We do not understand their language, and perhaps never can, because it may be of a
kind and it may be used for purposes that we can never grasp. It may even be one of the corruptions by human language itself
that we ever use the word "language" in the first place for what the whales produce. Whale sounds might be some kind of
expression, for instance, of individual personality, have a function utterly unfamiliar, even meaningless, to us. It may have little
or nothing to do with the conveying of facts, but a lot to do with the conveying of sensitivities and awarenesses. That could
explain why whales seem unable to defend themselves against human cruelty or take measures against the human interference in
their world that sometimes leaves them dying stranded on the shore. This powerlessness gives an excuse for saying that they are
not really intelligent, and that in turn gives a circular justification, truly vicious, for continuing the cruelty against them.
As if lack of intelligence justified callousness. As if intelligence is a mark of superiority. As if we know what intelligence is,
and enjoy possession of the true, superior sort by which others are to be judged. AS if, above all, analysis by word and into
word justified conclusions, theoretical or practical. Here again language is, as so often, at its deadly work. Concepts -words, that
is - like intelligence or superiority form humans' attitudes and lead their actions; when in fact the concepts have no reality, and
in any case cannot justify cruelty. The only reasonable thing is to ask what whales actually do, not to make judgements based on
fictions produced by words like "intelligent, advanced, superior, sophisticated". Whales may have sensitivities far beyond ours;
they may have awarenesses far beyond ours. They may do things and enjoy things utterly beyond our understanding.

Page 231
And at the same time they may not be able to handle evidence as we do; they may not be able to analyse as we do, either in the
useful or - lucky them if not - the useless sense. They may not be able to gather, organize and use information as we do.
Information, we say, is power. It is, and what do humans do with the power? Use it to make each other and other living beings
miserable; while the whales do not have power and yet would be happy - there can be no realistic or sensible denial of this - if
they were free of us, free to live their gentle lives in peace.
Why is it that when people see Icelandic whale slaughterers or Japanese dolphin murderers at work they do not shoot the
humans down as they would normally try to do if the same humans started mass killing on the streets of their town?
Undoubtedly at least partly for the purely self-centred reason that people do not see the whale and dolphin killers (killers far
more cruel in their killing than most indiscriminate killers in street or bar) as a threat to themselves. Yet the savagery against
whales, and so many other living beings, would almost certainly come far less easily if it were not for the definition,
categorization, spurious rationalization made possible by words. Even the utter illogicality that supports that self-centredness
itself could perhaps collapse sometimes if not supported in turn by language. Without the terrible analysing work of words more
people, perhaps all people, would be able to feel the truth that whales are vulnerable beings able to suffer as intensely as we do,
perhaps - for who are we to know? - even more than we do. Without words perhaps we would not find it so easy to distinguish
between horrors simply because they are committed against what language can fix as different categories.
Words limit awareness, words distort. The associations people make with words also tend to be inflexible; as I have pointed out,
they have to be to make language function at all.
There may be associations that are 'private' for individuals, unbeknown to both the individuals and everybody else. That can
have disastrous results too. But the conventional associations with words shared by millions can also have fearful results.
"Acquisitive, aggression, anarchy, blackmail, competition, exhibitionist, genius, great, identity, intuitive, matriarchal, perversion,
program, property, psychopath, scientific, security, status, structure, intelligence", and so on. They can corrupt in varying ways,
but in the end nearly always because of those inflexible associations. Perhaps the point becomes quickly clear if one thinks
again of another example, that word "intelligence".

Page 232
Nobody in fact knows or can explain what intelligence is, yet it is a word that most people have very definite, or perhaps I
should say, strong or intense associations with, and practically everybody thinks it's a good thing, and tests have been thought up
to measure it, although we don't know what it is, and a lot of people's lives are seriously, perhaps crucially influenced by other
people who make the tests, although they don't know what it is, and the misgivings of sceptics haven't prevented the transfer, as
it were, from something vague and fairly harmless to something absolute and rigid, which it has no right to be, and all because
there is a word that can make the treacherous link.
But maybe things would have been even worse if there had been no sceptics. There aren't so many sceptics to guard against most
of the other words I have just mentioned. Most words tend to make people forget the reality that is the only justification for the
word's existence. One has fixed on a certain association for each word, usually early in life, and the word comes to replace the
reality. The word is a kind of generalization; it ignores the variety of every experience, the uniqueness of every experience. So
when somebody else says or writes any word, it is normally impossible not to allow one's generalization connected with the
word to determine one's understanding of whatever the other person is saying. The associations with a word usually remain
fixed even when other people start using the word in a new context. These other people are often 'experts', so most people do
not resist the new use, and become muddled - including the new usage leaders themselves - and come to believe that quite
different things have something in common, or even that they are the same things. They do not look carefully at the things
themselves to see what really happens, but are content to accept the link suggested by the word, and thus many people's attitudes
are coloured and dictated.
On more than one occasion a person has argued to me as follows: Small children grab for things. That's acquisitiveness. Adult
human beings in society are acquisitive. So what adults do in society is the way the small child grabs. So adult acquisitiveness is
natural, inborn... Thus the word props up the argument. I am not saying here that things that humans do are or are not inborn or
natural or inevitable. But it is utterly crooked thinking to use a word rather than what actually happens to argue either one way
or the other.
"Aggression" became a fashionable word, certain scientists gave it respectability, it became something natural, vital. Now sports
commentators talk of athletes needing it, having it. But people do not lose their fixed associations with the word; they do not say
to themselves, "Ah!

Page 233
We are talking here about something quite different which has nothing to do with our original associations." Instead, attitudes
and values change. I do not pretend that this single word "aggression" by itself has changed the behaviour of thousands. That
would be very fanciful. But behaviour has changed, both among 'sportsmen' themselves and among those that watch them.
Violence at games is probably due to many factors. But a 'transfer' of values through the word there sometimes is, and together
with other words I think it is possible that it has contributed to nastiness.
"Competition" can work in a similar way. Things that animals do are called "competition" by zoologists. (I am worried even at
this stage by the way zoologists, ethologists, foist words for human behaviour on animals.) Then people claim that
"competition" is a law of nature and is an inevitable mode of life for humans; but they do not think about what really happens
among humans, and whether that is relevant anyway.
The way words corrupt varies slightly according to the area of meaning they belong to. But a transfer in some way of
associations from one reality (or supposed reality) to another can happen through many different types of word. It can work in
such a way that one allows a generalization - the word - to replace or alter the reality of whatever one is experiencing oneself.
This turning of feelings and thinking into words and thus into something different from the original feelings and thinking, or
even the words thrusting themselves in before the feelings or thinking can mature properly, is something that probably happens
in all types of people, from illiterates to philosophers. I suspect it may be a good deal more extreme and fateful among the
philosophers than the illiterates.
But whoever it happens in, it blinds. Words such as "love", "like" and "hate" are perhaps good examples. Most people have, I
think, established rigid associations with the word "love", and when they have feelings that they reckon are roughly in that area
they quickly squeeze the feelings themselves into the generalization that is the word. And the distortion is not only in attaching a
particular word - here ''love" - to the reality; it can also be in insisting on making a rigid distinction between "love" and "like"
etc. where in fact there is an infinite variety of individual feelings. Those who go further and try to make precise analysis of such
words are making the confusion worse by making them even more absolute and cut off from reality.1 I can perhaps best
illustrate the confusion that people suffer as follows: my feeling=love; love=x; therefore my feeling=x.
1 Here the single track of language gets me into difficulty. The uselessness of definition is really an important part of the
argument against language. In my thinking at this stage I am always aware of
(Footnote continued on next page)

Page 234
Words can actually invent completely unnecessary problems and ordeals for human beings. Something associated with "identity"
may be a genuine problem for many people of whom perhaps most have never even heard of the concept; although I suspect
that in fact most of even these people have much less subtle and abstract problems. But I am certain that many individuals add
to their worries, or fail to see the true cause of them, as a result of believing in the reality of "identity".
Sometimes words 'change' people and things in a frightening way. Words are limited, so one uses each time the word that one
feels is least far off the mark. This, I imagine, does little or no harm when we say "cat" or "hot".
But humans not only associate words with things, they associate things with words. The word bounces back, so to speak, onto
the thing and, fatefully, changes the thing. One sees a man doing something one considers very bad. "Psychopath" one says. In
that instant one has invented a new truth, or rather, a new lie, because the word bounces back on its victim, and changes him.
Changes him, that is, in our minds, which is decisive. "He is a great man!" one exclaims, eager to find expression for one's
admiration; and the man is immediately transformed out of all recognition, and a whole new set of rules and a whole new set of
evaluations are applied to him. He has become a man apart. As for "scientific", that is a word which cons millions of people
every day, from the most non-intellectual to the supposedly clever, and fraudulently recommends all manner of things to them.
This sort of distorting word often has another characteristic, which one might call the 'apparently useful useless'.
"Exhibitionism." Who knows what this word puts into a person's head?! But it colours in all sorts of ways: sick, vicious,
abnormal, vulgar, sinister, disgusting, pathetic, insight, perception, scientific (all of them in their turn fearful distorters). One
should ask instead: Does it increase our knowledge to apply this word to anything? Does it tell us anything new? Does it give
us understanding? Instead of rushing in with a word to deal impressively with the whole matter, one should look in each case at
what is actually going on and what are the results. One finds a person who likes to tell everybody about his feelings. Well, how
exactly does he do it? Does it make him happy? Is it important whether he does it or not?
"Perversion." I have read: "If you have to see others do it every time, in order to get an orgasm, that is perversion." It is nothing
of the sort! It means you have to see others do it to get an orgasm, and no more! Then one
(Footnote continued from previous page)
that uselessness. But I do not want to clutter up the track at this point. So I am pushing it forward to the next chapter. If
you would like to bring definition into the picture already now, see p.242.

Page 235
can discuss that alone on its merits. At one time the experts will tell us that XYZ is a perversion. At another they will say it is
not. Nothing has changed - except the attitudes, the condemnations, the fears, the self-contempt.
"Aggression" once more. In a colour supplement article it is declared that women watching wrestling are "releasing their
aggression". Again, that tells us nothing, adds nothing to our knowledge. It merely repeats in 'scientific' jargon what most people
know already; it may even obscure what people know, muddle them. It certainly works against any real understanding. An
empty classification is produced instead of any attempt to understand the why, and real process, of what is going on.
In an interview on television an eminent thinker, an enlightened man, I had believed, is in next to no time talking about the
matriarchal approach, i.e. unconditional love, and the patriarchal approach, i.e. justice, earned love. This sort of classifying
analysis doesn't give us any more information. It doesn't -or rather, shouldn't - recommend the one or the other attitude or
behaviour any more than if we don't use these terms. The terms will probably confuse most of us, in fact, by making us think of
mothers and fathers instead of about the real attitudes and behaviour themselves. We should be arguing about the good or bad of
those, not about abstract and emotive classifications. These classifcations, labels, do not help us to understand why particular
attitudes have dominated, or why they change, or how they can be changed. And how can one get people in general to follow
this sort of talk, get them to change their attitudes by means of such words? I have chosen this particular example because it
shows how even apparently sensible people can waste their thought, and not just in obscure special journals, but in front of great
numbers of 'ordinary' people. Thousands of such pointless and distracting words are poured out every day in academic and other
intellectual circles. There are effectively two groups of people. One group is of people who go round in circles exchanging
language. Then there are the others who are not interested or don't understand, or both, perhaps the former because of the latter.
(I wonder if anybody has altered me yet by calling me populist.)
I have heard someone on the radio discussing the nature of dreams. He quoted someone else who has apparently declared that,
although the human brain is not exactly the same as a computer, it must, like a computer, have a program, and that dreams are
probably the rewriting of that program.
Such a pronouncement seems to me a good example of the corrupting influence of language working together with the barren
itch for analysis. The enquirer admits the brain is not the same as a computer. But he feels

Page 236
that brain and computer have something in common. He is almost certainly led to this conclusion, at least in part, by words that
are associated with both "brain" and "computer", and also very likely by definitions which "brain" and "computer" have in
common. Having thus decided that the brain has certain properties in common with computers, he starts drawing conclusions
based on further word associations. From "computer'' to "program"; or perhaps less directly, from "computer" to "input" to
"program". Now, it is almost certain that the same enquirer, if he had been speculating sixty years ago, before the days of
computers, would not have arrived at the word "program" in connection with the brain, and would have suggested, if he
suggested anything, a completely different concept, a completely different association. And if it had not been computers that
were developed but something else which could be seized on - again through definitions - as a model or partial model of the
brain, then we would have been offered a quite different word from "program", and one with quite different associations.1
But even if one assumes for a moment that the brain does have a program, what does that mean? In fact it tells us nothing. It
tells us nothing further about the nature of the brain. It is a typical piece of meaningless analysis for its own sake, going round
in a circle. It is decided that "brain" and "computer" have something in common, so the "brain" must have a "program". What
does that tell us? Merely that the "brain" is in some ways like a "computer". It does not tell us anything that would be really
interesting or useful to know, such as where the "program" comes from, what its nature is, how it works.
Further word associations lead the enquirer very easily to speculate that dreams are a "rewriting of the program". This, I think,
shows even more clearly the emptiness of this sort of 'thinking'. What rewrites this program, how and why? To say that a dream
rewrites the brain's program gives no increased insight into the nature of dreams. Even if true in any sense at all, all it says, in
different terms, is that dreaming is different from being awake. But of course saying that it is a "rewriting of the brain's
program" will fool those who think in the grip of language and therefore do not think genuinely at all.
Then there is a type of word which is particularly common in psychological and sociological language. It seems to me that most
of such words not only do not advance understanding; they corrupt by stopping people thinking properly about what they are
talking about. For instance,
1 See above, p.90, and below, pp.247-48.

Page 237
"distress, disturbed, inadequacy, depression, identify with, alienation, repression, reification, role play, cognitive
impoverishment, communicative competence, disadvantaged, sexism, racism, criminal, humanistic". There are political words
that work the same way: "liberalism, fascism, socialism, establishment, liberation, revolution..." These become, for experts and
non-experts alike, established, accepted terms. That is not the right way to think nor the right way to argue. It is very convenient
to have these terms all ready to do one's work for one. But it may distort reality, because the truth about the people one is
discussing may be rather or even very different from the idea associated with the word; and that is because of something that is
basically probably even more important: the convenient term, the convenience of a ready made token, excuses one from
explaining carefully and specifically to oneself and to others exactly what it is that is happening in this particular case, what
these particular people are doing and feeling. That is all avoided by means of the quick, 'efficient' generalization. It is easy to
illustrate the mistake with the word ''revolution". Whichever side people are on they tend to use it as their yardstick rather than
find out what actually happened or discuss what might actually happen. And both 'revolutionaries' and 'anti-revolutionaries' fall
into each other's trap by using the term and defining themselves by the term. Some find something fearful simply because it is
supported by the supporters of "revolution". They do not look at the realities involved. Others find something fearful simply
because it is not "revolution", it is the work or aspiration of mere contemptible "reformists". They do not examine the realities.
The result is often not just that failure to think about the realities that are the only true foundation of logic and morality, but also
the artificial intensifying of hostility and hardening of feelings.
It not so easy to see the danger in many of the other words I have cited as examples, particularly the 'psychological' ones,
probably very largely because they are supposed to be more 'scientific', and so are suspected less. But they work in exactly the
same way, and no doubt more insidiously precisely because of their status. And there is in fact no basic difference between this
sort of word and words like "wog, coon, whitey, bosses, lackeys, layabout, scroungers, exploiters, thugs, parasites, crooks,
hooligans, the pigs, drop outs," etc. The latter are openly abusive, the former have scientific pretensions. They both remove the
need for an honest account of exactly how things are.
Even if you agree with me in principle you may say that what I am suggesting - that we should do without such words - is just
not practical. Life is too short. We would never get anywhere. Well, it is true that what I

Page 238
am asking for needs more time, more thought, more trouble. But it seems to me that we have not got anywhere anyway. The
world is as full of social upheaval as ever - or even more so than ever. There appears to be as much personal unhappiness as
ever. Psychiatric institutions have millions of inmates, but there is no sign of a constant stream of relieved, smiling and
harmonious humans being discharged from them. We do not really understand ourselves better than we did. There are many
theories, but no calm confident appreciation of truth solidly established. What we do have is conflicting schools and conflicting
ideologies. It might turn out quicker and more trouble saving in the long run to make that effort to concentrate on the actual
happenings. That way people concerned with the problems might come to agree more easily about the realities and get down to
more of the compassionate practicalities.
I have tried to give some examples of how words, meanings, corrupt. They are inevitably superficial. Of the fundamental evil of
the work of meanings one cannot give examples, just because that work lies so deep. It is not just that a meaning appears to be
one thing but in fact is another, although that is often so, too, at one level. The evil is deeper than that; the meanings have taken
over. There are the real things; and there are the meanings that stand between humans and the real things. But the critical point
in the fraud, where the decisive 'twist' or 'turn' is - this one cannot, I think, illustrate in words. This impossibility lies in the
nature of the fraud itself and one can only have the insight that that is how it is. There are meanings instead. And that is why it
is perhaps sad, really, not a joy, when a child begins to talk, particularly when she talks to herself. For there she loses her
innocent realism, and the blinkers and the prescribed lenses take over. She has fallen for this treacherous game. She thinks it
gives her power. And the tragedy is that indeed, without it, such is the world, she could not cope. But the truth is, she has
become a slave.
I have tried to explain something of what one might call the basic 'psychology' of words, and how this may sometimes affect a
lot of people all together at the same time, or affect widespread attitudes. Language also has bad effects of a practical, 'everyday'
kind, which tend to affect people more singly, although I do not mean that there are two distinct, separate phenomena.
I have already said (p.202) how I cannot put clear thoughts into clear words. I believe most people, if honest, would admit that
they too at some time have had this experience. There must also be many people who, like

Page 239
me, sometimes find that when they put something into words it dismays by becoming something quite different. It now, in
language, arouses associations that were never there in the reality.
I have also often discovered that when I have turned ideas into language, whether to record them or in order to 'prepare' them
for explaining to others, and have left the words for some time, and then read them again, I can't get back to the 'true' ideas
themselves; they slip away, leaving me stranded with marks on a piece of paper. If I am to recover the original insight, I have to
forget the words and start all over again. And sometimes I find this very hard to do, even though I like to think I am a person
unusually free of words. They have done their corrupting work, and are still doing it, refusing to let me go, let me back to
reality.
It is not only that some writers - or speakers - are better at expressing themselves than others, although that in any case, as I
shall argue in a moment, is no justification of language. It is that language distorts reality the moment it touches it.
There is a widespread belief that when one puts one's ideas into words one clarifies those ideas. The opposite is true. Words are
not things; they are not even the thoughts about the things. Words are mere empty shells, repeating stereotypes. Turn one's ideas
into words, and one loses all their original solidity, richness, subtlety, variety, uniqueness and reality, and turns them into
substitutes, into dummies, into falsehoods. One may even, unwittingly - that's what's as bad as anything - lose the ideas
themselves altogether, and end up with something completely misleading to oneself as well as to others.
The maxim that unclear or confused language shows unclear or confused thought is false. Very often apparently clear language
does not express true thought at all; it is only a playing with and organizing of words, which are made on the surface to look as
if they fit. Often 'better' writers are simply fooling their readers. Their 'good' writing is no guarantee of their honesty, or of their
ability to convey what they are really thinking. On the other hand a person whose language is confused may be a person who is
too conscientious and not slick enough to do violence to her thought by distorting it to fit within the limits and conventions of
language. We should never assume that elegant and seemingly clear language better expresses thought and the truth. It may well
hide much irrational thought. It is the logic of reality we need, not the 'logic' of words.
There is an ironic contradiction about the way language works. Language cannot properly represent reality, but precisely
because its pieces are associations with reality, mere approximations, when humans want to

Page 240
express their experience of reality there are many alternatives of pieces, words, word orders that they can use to express that
experience. The result is that language gives the illusion of being richer than life itself; and what people believe is a more
accurate or better way of expressing something than another is often nothing more than more pleasant - seductive -fraudulent - a
con trick to control people with its seductiveness. So the very weakness and treacherousness of language makes it seem strong
and fine.
This confusion is probably encouraged by the great store set by articulateness in many modern communities. I suspect many
people cannot believe anything (thought) is there until it has been articulated. It doesn't seem real until it has taken form in
sounds, or, better still, on paper. It's not there until it's been communicated. Even so, you may say, there are different levels of
expression, of articulateness. There is no reason not to try to do one's best even if one cannot achieve perfection. What is wrong
if one speaks or writes well and effectively? One is only doing justice to one's thoughts. True. And what happens to the thoughts
of the people who cannot do justice to them? Doing 'justice to' one's thoughts is very likely the most essential accomplishment
for becoming a member of an elite. People can and do become part of an elite with power because they can express themselves
'well' and effectively. I do not mean that they necessarily use Hitlerian or Stalinist or Orwell Newspeak techniques. They may,
many of them, be fairly honest or even very honest people with 'decent' and rational thoughts. This cannot alter the fact that they
enjoy an advantage, wholly improper, over other people who may be equally or more honest, who may have equally or more
decent and rational and valuable thoughts.
If we could see into the heads of people and see their thoughts direct, whenever they wished us to do so, we could truly judge
thoughts on the basis of the thoughts themselves, not on the effectiveness of their expression; and thinkers could not deceive;
and thinkers would never be helpless. As it is, there is constant cheating through the word. If there have been any 'felicities' of
expression in this book - I suspect few, if any - then I have cheated; as I have undoubtedly cheated even without any felicities.
Language is used dishonestly to recommend thoughts.
On the other hand people often upset each other, or get angry, because they are using certain words, and each takes it for
granted that the other or others understand exactly the same by those words as she does herself. In fact each one may associate
quite different things with those same words and combinations of words; they recognize, or think they recognize, realities which
are in fact different. So they disagree and dispute. In other words, very often they are simply talking about different things.
People

Page 241
misunder stand and hurt each other needlessly simply because language is letting them down. Talking and listening to language
ought to be the using of signals to recognize thoughts and experience, but when those are at all subtle, when people's experience
and associations are different, they don't recognize, they misunderstand. The language misleads.
One can observe this in varying situations. Wife and husband for example, where one feels perhaps that it is particularly sad,
often hurt each other because they misunderstand each other's language. This is very common when they feel some sort of
abstract principle is involved. (They may introduce expressions like "rights, exploitation, responsibility, woman's role, duty,
independence, self-fulfilment, support, etc.) Each individual's experience and associations with the words are different, but
fixed, fossilized.
Yet this is not necessarily because people are using language carelessly. Often one can hear them using it meticulously, priding
themselves on their efforts to be articulate. And the more they try to be meticulous and articulate the more they insist on their
interpretation of the language they use; for they are being responsible and accurate. Sometimes it dawns on people that they are
being confused by clashing uses of language. This is where one must normally give up all hope of any sort of progress. The
discussion, the dispute, now becomes only a wrangle about words, the realities are forgotten. The participants become more
determinedly bent on self-vindication than ever. It is sad and silly. Nobody is using the language 'wrong' or 'right'. They are
only exchanging words about words.

Page 242
12
Trying to Speak the Truth
So what is to be done about it?
The solution is not in definition and linguistic analysis of meaning -neither in principle nor in practice.
Definitions are only words about words.
Definitions do not tell one anything about things - which are, or should be, what are important in life. A word is not the reality.
And not only that. As I have tried to show, a word is from the start a distorter of the reality. A definition, saying words about a
word, is something that takes one even further from the reality, carries the distortion one level further. It does not convert the
word back to the reality, it does not enlighten in terms of reality, it does not point to the reality. It is something entirely
artificial. It does not make the connection one really needs to that reality. Most people, I suspect, believe that definitions are
about reality, about things, and that there is something solidly scientific about them. In fact they are one more barrier between
humans and straight thinking.
Nor does definition and analysis of meaning tell one anything more about what a person actually means, and even less about
what a person thinks. It is useless, one might say meaningless, to talk about what the real, proper, correct meaning of a word or
sentence is. There is no independent truth about the meaning of a word or phrase or sentence which clever or learned people
know or can find out and some others don't know. Words only mean what their speakers want them to mean; only people
express meanings.
How does anyone decide whether a definition or an analysis is a good one? How does one make a definition in the first place?
Only through one's previous knowledge of another kind about the meaning, a knowledge without benefit of definition, that is,
knowledge of the associations of reality (or what passes for reality) attached to the word. There is nothing absolute or universal
about those associations; they may or may not coincide more or less with other people's.
People do indeed use words variously and often inconsistently, and perhaps many deliberately deceive others by manipulating
words. But to set up a huge, elaborate, systematic analysis of the meanings given to words is as pointless as if one was aware
how dangerous generals are with their bombs and so set up a technical examination of the details of the bombs instead of
concentrating on the mentality of the generals. In the same way we should be thinking about our own mentality in connection
with language and about the general principles - psychological principles -

Page 243
concerning the way we use words. Naturally this need not stop us drawing attention to the way specific words are used on
specific occasions when people are deliberately manipulated or accidentally confused. That is practical. But systematic analysis
of the meanings in language is theoretical in the worst possible sense of the word.
Earlier this century it appeared to dawn at last on some philosophers, the 'linguistic philosophers', that what most philosophers
had been debating up to then was largely a mass of illusions. The old philosophic discussions were discussions about things that
were thought to be realities in their own right - truth, beauty, justice - independent of words. This, the linguistic philosophers
said, was all wrong. There are no independent things that such words merely stand for. One knows what a word means when
one knows how to use it. "For example, with a notion like 'truth', when you fully understand how to use the word 'truth'
correctly - and its associate words like 'true', 'truthfulness', and so on - then you fully understand its meaning" (Magee, 1982,
p.111). So philosophers had previously largely been worrying about false problems, non-problems. What philosophers should do
was analyse human concepts, "the categories of human thought", by finding out "how they are actually used".
The linguistic philosophers got the wrong end of the stick, and drew the opposite of the logical and practical conclusion.
What they should have said is that words are not accurate or complete representations of realities - that they distort the realities.
It is the words that are wrong, the concepts that are false. But what the linguistic philosophers did was as if they wanted to
abolish the realities, instead of what they should have done - drawn attention to the defects of language. They appear to have a
partial insight that many so-called problems were really only questions of words; but seemed not to realize that the falsehoods
and illusions were the direct result of the words. They clambered out of one morass of words only to fall straight into another
even deeper one. They enslaved themselves to language perhaps more than anyone before.
There is a tendency in almost all humans to lose touch with the solid ground, float upwards, and instead of thinking about real
things they bounce words alone off each other up there in the air. This passes for thought; very often, the wordier, the airier, the
more profound or at least clever the thought is believed to be. Humans tend to grasp reality itself through words anyway, and as
a result fail very often to grasp it at all. Then the linguistic philsosphers came along and began bouncing words off words in
order to talk about bouncing words off words.

Page 244
I think a good example is the concept of "mental illness" (mentioned by Bernard Williams, a Cambridge philosopher, in Magee,
1982, p.114). I have much sympathy with those who reject this concept. But the answer is not to analyse 'it', redefine 'it'. That
way one simply comes up with something almost certainly equally false, a new concept, perhaps, or squeezing 'it' into some
other 'old' concept, that too false. That is useless and possibly just as harmful as the old 'misconception'. It is no better to decide,
for instance, that 'it' is unhappiness or criminality or anything else. Instead one should look once again at what actually happens.
Certain people do certain things that - for example - some other people find strange or irritating or frightening, perhaps. What is
the effect on other people? What do the certain people want? What should other people - and which people -do about it, if
anything? That way people can deal direct with what really is. But once one starts giving names, creating concepts, establishing
analyses, the damage is done, because people will act on the basis of the name, the concept, the analysis, the fossilizing words,
not on the basis of what actually is. People may then act very foolishly, or cruelly, or both - they often do -and the damage may
be fearful and irrevocable.
The linguistic philosophers' dissection of language is not about either real things or real problems. Their words about words do
not make moral conclusions or decisions easier. And it goes almost without saying that they do not all agree with each other
anyway. One has to ask to hear of a single unmistakable advance linguistic philosophy has provided humans with.
Linguistic philosophers and definers generally are muddled and inconsistent. To say that the meaning of words is their use is to
state the trivially obvious. That words mean what the people who use them mean. (It is only people with what one might call a
French Academy mentality who could ever really believe otherwise. If there is ever any doubt that 'people mean what they
mean' one need only think that if it is not so, then modern English-speakers are talking gibberish a lot of the time because the
words they use 'really' mean what Shakespeare meant by them, and Shakespeare in turn was in even worse case, because the
words he used meant what Chaucer meant by them.)
It is perfectly true that people are often muddled and self-contradictory in their use of words. That is one of the inevitable
corrupting effects of language. But what do the language analysers and the definers wish to do about it? If they say that when
Smith says such and such he, Smith, does not really mean what he thinks he means, that and that, he really means this and this,
they are patently wrong - Smith does not mean anything of the sort. Definitions and language analyses are incorrect and

Page 245
pointless unless they are universally - or at least very widely - accepted. A definition is only 'true' if it is accepted as true. And
even then it is not really true in the way most people believe, because it remains mere words about words: it is not a truth about
the world beyond language.
The purposes of the definers and language analysers are suspect whichever way one looks at their activity.
If their intention is not a reforming one, then definition and analysis is pure trivial and self-indulgent intellectual play.
If they wish to reform and improve the rest of us, they either succeed in getting us to agree and practise their use of words, or
they don't. If they don't, then their definitions and analyses are clearly wrong, as I have pointed out above. I hope they will find
it difficult to persuade people to apply their prescribed definitions consistently, and that even if willing, people would not be
'skilful' enough to do so.
The most frightening result would be if they succeeded. If in some way the definers and analysers - men and women of 'science',
naturally -managed to impose their regulations for using language correctly - and "correctly" is their term, not mine (see p.243
above) - if they imposed their regulations, we would become slaves of the word more than ever, since we would no longer have
any suspicion of it, we would believe that we always understood each other perfectly. People would be lulled into a more
complete illusion than ever of the accuracy and clarity of language.
This might lead to even more confusion and muddle than we suffer now. That would be bad enough. The alternative would be
far worse, and I fear more likely. The imposition of universal definitions, meanings, would probably destroy the remains of free
and clear thinking, as all thought would be imprisoned within the limits set by the definitions. Tight definition of language is the
last thing we want, because language is not thought. Language distorts, distorts our thought, and moves us away from reality.
The analysers and definers, with their language fossilizing activities, seem intent on removing us yet one more stage away from
reality into an artificial world where we are even less able to think directly about what actually happens. 'Thinking' in language,
already so harmful, would become the only accepted, proper way of thinking and humans would become blinkered robots living
in a strange world of bloodless fantasy devised by the language-mongers. The tyranny of the word would be complete.
There really are forces, although diverse, that try to bring all human thought, perhaps indeed all human experience, under the
rule of language. The attempts are frightening because they come from academics and

Page 246
experts who claim science as their warrant and are likely to win respect for their view of humanity.
These forces are at their extreme in movements such as 'structuralism'. I am indebted to an article by Richard Webster in the
Observer of 1 February 1981 for some information on the writings of Tony Tanner. In a book called Adultery in the novel
Tanner "treats adultery not primarily as a sexual act involving love, lust, jealousy and betrayal but as a 'category confusion'."
Tanner's interpretation of the Old Testament law on adultery is that it did not prescribe that adulterers should be killed; it
maintained that "violators of the social space" must be "excorporated into non-being". Tanner defines a napkin-ring as "a
material encirclement of nothing" and snoring as "the unspeech of nonconsciousness, a drowning of semantic utterances in
involuntary bodily noise''.
I do not know if one can afford simply to regard this sort of thing as hilarious. I should like to. But some influential people
support it, and I fear one may have to take it seriously. Not all linguistic movements are so obviously grotesque, though.
One of the main themes of Chomsky's system is that humans are genetically preprogrammed to use language, so all languages
have a common basic structure, so nothing that cannot be fitted into that structure is expressible, and so there are strict limits to
humans' ability to understand the world.
This sort of thinking is one of the effects of the obsession with language. It can surely only be based on the assumption that
thought, or most of it, depends on language. There are of course restrictions on human understanding, as there are in the nature
of all creatures. But language is not, biologically, one of them, and it is a terrible thing to encourage the idea that we must
remain forever in thrall to it. Apropos the constraints imposed by the human genetic program, Chomsky declares: "Fortunately
for us, we are preprogrammed with rich systems that are part of our biological endowment. Because of that, and only because of
that, a small amount of rather degenerate experience allows us to make a great leap to a rich cognitive system..." Fortunately?
Both for us and the rest of the world "disastrously" or "cruelly" would be a more appropriate comment. As for language,
preprogrammed or invented, it should be the object of regret, not the cause of self-congratulation.
Yet the faith in, the awe of, the affection for language appears incorrigible. On the occasion of the publication of his and his
colleagues' grammar Quirk declared in an interview published in the Times (May 20 1985): "...language is the most important
attribute of humanity. Without

Page 247
language we wouldn't be out of the trees. Language is the basis of our wooing, our thought, our friendships, our planning, our
everything." Well, we and everybody else would almost certainly be much better off if we were still in the trees, like the
gorillas. And I have tried to show how so much of what we suffer is the result of allowing language to be the basis of thought.
It is interesting Quirk mentions friendships. I believe there are millions of humans who would warmly confirm that friendships
with dogs or cats or other animals are often truer, deeper, more genuine, more disinterested and more rewarding, less
problematical and less anxious, than friendships between humans. This, it is true, is probably largely for the self-centred reason
that the relationship of a human being with a cat or a dog, say, is perfect because it is for the human exactly what she wants it to
be, what she wishes to make it. The same is probably largely true for the animal too. But why? Because there is no language to
distort, to make foolish demands, and to interfere with the spontaneous interaction of the personalities of animal and human.
I suspect that much the same goes for our wooing as well. The long term results of wooing would almost certainly be far more
satisfactory for far more people if it was freed from the bad inspiration, bad reasons, bad judgement and irrelevancies induced
by words.
Humans, indeed, spend a great deal of their time using language to put false ideas into themselves and others.
The illogic of would-be supreme logic is dangerous. The definers, scientific-thinking people all of them, define. They squeeze a
thing (or that is what innocents believe) into the limits of language - and as like as not find terms that will suit their particular
purpose, that will allow them to control the argument. And when they have finished, they have a definition, not the thing itself.
And then, they will often claim to equate things or differentiate things by definition, although in fact they are playing a trick
with words about words.
For example, it is just so, it seems to me, by such argument by definition, by talk of functions such as 'problem solving', etc. that
some people have persuaded themselves and try to persuade us that computers are a model of human brains; thus 'artificial
intelligence'. By such arguments John McCarthy, the inventor of the term 'artificial intelligence', can claim that even a
thermostat has beliefs. John Searle reports that McCarthy once told him: "My thermostat has three beliefs - it believes it's too
hot in here, it's too cold in here, and it's just right in here."
Searle has demonstrated very clearly the absurdity of equating computer and human brain: Computers can never think like
humans

Page 248
because they work (by the affirmation of the proponents of 'artificial intelligence' themselves) solely on the basis of formal or
syntactical processes. There is no meaning content in a computer program.
I think Searle may be unhappy about the use I have put his argument to here. But for me the most attractive part of his case
against 'artificial intelligence' is his 'Chinese room' experiment. This is just the sort of practical reasoning and inquiry into what
actually happens that we need in place of words about words. See, for instance, the Listener, 15 November 1984. I have long
held the same view of computers and human thinking as Searle, but without the inspiration of his 'Chinese room' or the
completeness of the rest of his reasoning. I have argued that computers experience nothing, cannot experience anything in the
human sense. So they cannot understand anything, as words fed into them do not tell them anything, as they do us, about the
world they have experienced, or rather, not experienced. They cannot recognize anything in human language as they have no
experience of the human sort to recognize. For this reason it is my guess that however advanced computers or their programmes
become they will never be able to translate satisfactorily from one natural language to another; by that I mean that they will not
be able to make translations from any and every text one chooses to give them, without restriction, and produce versions that do
not have to be checked by humans.
Where I think I would disagree with Searle is on his point that the crucial lack of the computer program is meaning. 'Meaning'
is only words. That is not the crucial thing missing. The crucial thing is awareness.
The Russian students I mentioned earlier argued against me, through definition, that a factory and my toothbrush were both
"property", and that therefore they were basically the same thing. My point is not that people outside the Soviet bloc were going
to fall for that particular argument, nor that it is orthodox Marxism, nor even that the view was the official Soviet position. What
is important is that it is so easy for people to be fooled by that kind of defining method of 'thinking' into taking absurdities
seriously. Without the language of definition I believe few people would even be tempted to make the sort of connections and
disconnections so many in fact do.
Unfortunately it is not just as part of an example of the defects of argument by definition that computers are relevant to the
dangers I suggest. There are those who quite seriously wish and plan to harness both language and emotions to computers. To do
this what is needed, as far as language is concerned, are "closely spaced and carefully defined meanings". Language must be
framed "with the precision essential if the mathematics

Page 249
are to be of any use in analysis". AS to emotions: "Human intentions and feelings can now be quantified mathematically and
modelled and defined precisely on any computer." It may be that this last is as yet treated with some scepticism even by most of
those in 'artificial intelligence' circles. But what is sinister is that emotions can only be defined and quantified through the words
for them, so in this way too language is the system that threatens to dominate us - now in a new way.
Joseph Weizenbaum, a computer scientist, has pointed out the danger that people will come to consider that what is not
'computer readable' does not exist.
So it is possible to imagine that as a result of the beliefs that
1 Language is the measure of human awareness, and
2 What cannot be computerized is not real,
computers and language might be brought into a fraudulent but plausible and convincing alliance (it's all science) which could
develop into something just as frightful as any monster authority imagined by Huxley or Orwell. But even if things did not turn
out quite such a disaster as that, it could only harm us for the computer to be used to help language in general and definition in
particular to confine ideas and give us false pictures of ourselves.
So what are we to do? I have so far only partly answered my question - and that only in passing - by urging that we should try
to think of what actually happens instead of hunting for and being led around by words.
But first, humans must change their attitude towards language. They need to be aware, always, that language is treacherous;
treacherous not just sometimes, but always and fundamentally. They should never trust words, try never to think in words. They
should think only about what is, outside and inside themselves.
Defining, attempting linguistic precision, is not only misguided in principle, not only leads, at the very best, to yet more
confusion and irritation. It also favours the 'clever' and articulate, because only a small minority of humans have the privilege of
being able to play these word-manipulating games. On the other hand, everybody can distrust language.
Language remains, however, the only thing we have got for most of what we want to communicate to each other. What we can
almost all do, though, is to be careful, and helpful to each other, about the reality we are

Page 250
talking about, careful to help break through the artificial barrier of words to recognize the actual experience or thought itself. To
do this we must, apart from anything else, use the simplest language we can find.
I think it is important here to make sure I am not misunderstood. I do not say that we should use simple words (and sentences)
because in some way simple words have a 'truer' meaning than complicated ones. Indeed, if you have been able to follow me up
to this point you will realize that this is certainly not what I think. It is not that there is a false language, the complicated one;
and then, in contrast, the simple accurate one. Liars, deceivers, manipulators, propagandists, 'scientific' frauds, or even plain
pretentious wafflers can use simple language as effectively as complicated language for the purpose of pulling the wool over
eyes; probably more effectively in most cases.
No, it is that in practice, on the whole, and assuming honesty and good intentions - a genuine wish to convey one's real thought
as true as possible, to be aware of what really is real, to find out what actually happens - simple language is better for the
purpose. This is for several reasons.
More people will be able to understand it more easily. People need not learn 'sophisticated' words. They need not feel inferior
because they do not know 'sophisticated' words. More people can use simple language.
I think also that the more 'physical' the words we use the better. We should as far as we can use words with physical
associations because these are the ones closer to a real world we are more sure of and understand better. They may stir better
the realistic imagining we need. They are easier.
But we should not use simple words and then expect others to analyse the words strictly. As listeners we should never ask
"What exactly does this piece of language mean? Instead, as speakers we should say "Can you recognize what I am thinking?"
Bernard Williams does not agree. He had the following exchange with Magee (1982, p.122):
"Magee I know from previous discussions with you that you're very much against the idea, so popular because self
indulgent, that expresses itself in some such words as: 'Don't worry about what I actually say, it's what I mean that
matters'.
"Williams That's right. That's what linguistic philosophy of all kinds was good at stopping people saying and, what is
much more important, stopping them feeling. That and the idea that somehow I have my meaning here - my little sentence
will try to convey it to you - but if it

Page 251
doesn't convey it to you that's through some failure of imagination on your part. We have a responsibility to our words
because, in the end, we don't have these meanings just inside ourselves, independent of what we're disposed to say. Our
sentences are our meanings."
This is quite irrational.
What is Williams talking about when he says "meanings"? If there is any point at all in what he says, he must mean that our
sentences = our meanings = our thoughts. But then even my own arguments against thought being language are unnecessary.
For if he means that, then that "responsibility to our words" must mean responsibility to our thoughts. How are we to fulfil that
responsibility to our words-thoughts? How do we adjust our sentences? By what mental processes do we get them 'right'? By
some more language, further sentences, inside our heads? And if so, how do we get those right, and so on...? And then, what is
the criterion by which we should judge whether we've got our sentence right or not?
We need the opposite of what Williams demands - not responsibility to words. Let us not handcuff words to words (definition,
analysis) and so imprison and blind our minds. Let us try with sympathy and insight to see what the other is getting at, not slap
her down because her version of the language is not ours. It can only mislead if we try to analyse the other's words, instead of
trying to get some insight round behind the words into what she is really thinking. We need to see what is in the other's mind,
and so we need to encourage her to explain in different ways, to find some way of conveying her picture.
In the same way, as speaker, writer, I should try by experiment to help others recognize my thought, see my picture. And if they
cannot, then I must try again, differently, patiently, imaginatively if possible, trying to jog recognition, trying to help their
imagination. Not by pedantic, irritating word-splitting that is only about words, not sincerely directed towards helping the others
towards insight but towards self-assertion as the umpire. Not by that, but by trying ever new ways of letting the light in on my
picture. This too is an important reason for using simple language. It is only in simple language that there is enough flexibility,
only by means of simple language that I can vary richly enough my account of my idea, that is, find enough different ways of
saying where I am trying to point.
There is an irony here that, because words are not to be trusted and it is not by an exact equivalence of each word to each idea
that we can convey our ideas, we may need to use very many more words than the champions of language would approve. If
you have found much in this book repetitive,

Page 252
this is at least part of the reason. I may have failed completely with many or most readers. But perhaps some of those who have
been content to persevere this far, and who have tried to follow me each time I put my thought another way, have caught my
mind's picture the second, the third, the fourth time round, where they missed it the first time.
In the comparatively unimportant work of trying to explain grammar to students of English (as a foreign language), I discovered
long ago that this is the way things work well. There is an absolute or near absolute truth about the use - that is to say, the
meaning - of most pieces of English grammar (and all other grammar): the articles, the use of the -ing form, the distinction
between the did and have done ideas, and between some and any, to give only four examples. But, contrary to what most
linguists believe, there is no absolute or 'correct' way of expressing these truths in either language or mathematical formulae, no
rock-like rule. (Those who think there is have either never tried to do it with real live students of a foreign language or have the
very widespread capacity for self-deception.) So one has to try different approaches, lead along different paths, until suddenly,
with luck, something sparks understanding in the student's mind. The successful account, the successful phrase, the successful
appeal to recognition, the successful example will not be the same for every student. Far from it, although certainly one may
find certain approaches more often successful than others, simply because there is a way of thinking more common than others,
a way of thinking neither necessarily better nor worse than others, merely different. This, I believe, is the way we should go
about trying to reveal any sort of thought that we want to convey.
There are many far more important thoughts than grammatical thoughts. Language can be a barrier twice over to good thinking,
to proper reasoning. Proper reasoning must be about 'realities' in the brain, not about words! Next, when humans need to
exchange their reasoning, and lead each other on to further reasoning, they must again discuss the realities, not words.
I have not much more to say about language. Just one thing which may at first seem irrelevant. Human problems are at bottom
nearly all moral problems. Morality and moral decisions must be based on facts. Surely nobody would claim that they should be
based on myths and falsehoods. And morality must be based on reason. Not many would claim, I think, that it should be based
on unreason. But it should not be the task of a learned elite alone either to reveal the facts to us or instruct the rest of us in
reasoning. Nor should 'moral' philosophers tell people with power or influence - such as politicians - that they (the politicians
etc.) should listen

Page 253
more to them (the philosophers). That would only give the powerful an even better and falser excuse for dominating our lives.
Instead it should be open to everybody to try to see carefully how things really are, what they are really talking about; and open
to everybody to use their reason to come to moral decisions about those realities.
So I return to my urging. Distrust language. In particular distrust names and descriptions and general terms. For the sake of both
understanding and morality use simple and unambitious words to discover and tell what actually happens to actual people.
Which real individuals do what? What do real individuals wish that we should do to them?

Page 254
13
Slavery to Authority and the Word
You may feel I have made something sensational, or at least something greatly exaggerated, out of the problem of language.
And then proposed a very much less than sensational solution to that problem.
Yet I think the failure of language is perfectly plain. There has been an improvement in the life of humans in many parts of the
world. But in the main that has not been the result of ideas, which have done a great deal of harm and very little good.
Where human lives are more comfortable and pleasant it is because of the success of mathematics and technology. In
mathematics and technology if one gets it wrong things stop, fall to bits, or blow up - or never work in the first place. And one
discovers these failings rather quickly.
Where human lives are now less miserable than cruelty, indifference and selfishness used to make them, it is because of the
growing of concern, pity, tolerance and responsibility. These feelings have been supported by the technology which has helped
so many people, from the eighteenth century onwards, to know about the misery of others. For this reason and other reasons
which are unfathomable, attitudes have changed for the more tender in many communities.
What is noticeable is that where facts work directly and immediately and incontrovertibly, humans have made undoubted gains.
I do not have to face the alternatives of a rotting mouth or repeated agonies as my teeth are drawn one after the other. I can
enjoy, any day I like, by means that would have been miraculous to the contemporaries of the men who made it, large amounts
of music that enriches my life. I can visit loved people hundreds or thousands of kilometres away swiftly and in comfort, if not
in complete safety. I do not mean that the result of technology is always happiness. It is often sickening disaster. But it is human
greed, ambition or stupidity that produces those disasters. The technology itself is not to blame. And the greed, ambition and
stupidity have fed on words, and excused themselves by words.
Where there is only supposed 'knowledge', particularly where that supposed 'knowledge' claims to be about humans, human lives
have not been advanced one whit. Here there are no incontrovertible and uncontroversial facts, and so it has always been. Only
schools and abstract systems and dogmas and pretensions and propaganda and polemics. Here there are only words. This is the
realm of words and all in it has failed.
The manipulation of physical objects and of numbers has succeeded; the manipulation of words has failed, brought no good and
much agony.

Page 255
That is the truth, and no exaggeration. It should be acted on, even if it is not easy to think of immediately effective action. But at
least we can recognize where the illness is and do our everyday best to resist.
An earlier faith in an inevitably improving world has perhaps disappeared. But there is still faith in ever-increasing knowledge
about ourselves and the 'skills' it is going to give us. We are to achieve happiness by the researcher. The researcher will tell us
the rules of how we are and the rules by which we should live. This bogus word-dominated knowledge -and its half truths are
probably more insidious and so more dangerous than the full nonsense - threatens to become the monster master of us all.
Even any real knowledge of ourselves that we achieve may not bring us more happiness. It is no doubt true that knowledge is
power. What will we use that power to do? I suspect mainly to manipulate each other and ourselves. That will almost certainly
be what happens unless we are moved first by sympathy, sympathy which gives the sort of 'understanding' we really need. And
the irony is that with that sympathy we probably do not really need formal, scientific knowledge of ourselves.
I am not attacking rationality. I want to defend rationality as well as I know how. But true rationality is straight thinking based
on the facts of all the sorts of experience. Today too many people with influence base their 'rationality', or their attempts at it, on
abstract formality without regard to what really happens to actual individual human beings, without looking to see what is the
actual effectiveness, if any, of their theories.
Language has long been popular with the learned and 'cultivated'. Now many such people believe that 'knowledge' about
language in particular will and should bring about an intellectual revolution. They believe the study of language is fundamental
to the study of ourselves. They believe, or at least tell us that they believe, that detailed analysis of language is a wonderfully
fruitful activity. They believe language will reveal the secrets of human thought. They will be very unwilling to give up
language as the great and wonderful illuminator.
Every year young men and women at universities in many parts of the world are enslaved to such beliefs. They feel that mastery
of the linguistic lore is essential to intellectual coming of age; they must prove themselves, show themselves worthy to be
initiated and given membership, perhaps even leadership and prophethood, in the cult.
It comes back again, as it almost always does, to the problem of authority and power. The academics around the world have a
virtual monopoly in forming our view of language and its psychology, our view of practically every other aspect of ourselves,
and the monopoly is threefold.

Page 256
They control the work on language, they control publication of ideas on language, and they control the reaction to any ideas on
language.
It would perhaps not be quite so bad if this monopoly applied only to the subject of language. But it in fact controls most of
intellectual life.
All over the world students are persuaded by teachers, lecturers, professors, and fellow students that if you are to be intelligent
and learned -or at least if you are convincingly to show you are intelligent and clever -you must study profound systems of
abstract analysis. You are a simpleton if you believe that things in society, in psychology, in history, can ever be what they seem
to be. A thing is apparently not right, not worth considering by the intelligent or the world in general, unless it has been
presented as a great analytical structure. Everyday observation of the simple or the obvious won't do. The truth, the teachers and
their students think, is only arrived at by the study of the 'literature', by knowing about the works of other profound
systematizers and knowledge builders - of whom you might even become yet another, with luck a better and profounder one
than those before you, a person with an even more deeply penetrating and truth-revealing analysis, an analysis of revolutionary
importance.
You can criticize any system, question any profundity, pick holes in any analysis, so long as you remain one of the brotherhood
at heart and create better systems, greater profundities, more earth-shaking analyses. But almost nobody seems to doubt whether
we should have this system-, profundity- and analysis-making at all.
Ordinary unpretentious statement of what is there for any interested person to see with straight thinking goes unnoticed. Usually
it is not the truth but the origin of its expression, the status of the person who says it, and how, that is the important thing. And
so the truth is, more often than not, ignored.
It is not surprising. There is so much to make students think like this. There are their careers. There are their incomes, there is
their security.
Perhaps above all there is the need to share the secrets of the brotherhood, to belong to it by showing the ability to master the
skills and the lore. If you questioned or rejected the whole structure you would be ignorant, pitiable, ridiculous. You would be a
failure and an outsider -morally an outsider, which is perhaps the worst, but literally too, because the student would cease to be
a student.
And once you have joined the brotherhood you have what is probably often the strongest motive a human being can have, for
not admitting that it is all a sham: the fear of admitting that you are wrong, that the ideas and work on which your life is based
are meaningless. The self respect of most

Page 257
academics and theoreticians is probably tied to their intellectual attitudes. The self-respect and the attitudes are prisoners of each
other.
You may feel that all these learned people, all these clever professors at universities, this cream of the intelligence of our
communities, all these people that governments give money so that they can continue their important work of profound
discovery - you may feel that these people cannot be wrong. But they can and they often are. I hope I have already given enough
examples of this in the field of language.
Humans follow great illusions and suffer. The worst of the illusions is authority, because it supports the other illusions. Humans
accept authority and so good people go to war and kill and maim and cause and suffer agony. They believe in authority, so
people subdue children and women. They believe in the inevitability of authority, and so see financiers and employers as part of
the natural order and as our benefactors and providers of work. Or they respect authority when their leaders send violent men
against other leaders who they say are evil, though the violent men mostly do little harm to the other leaders but terrorize many
ordinary and powerless people. Or they accept authority and so also the philosophy their leaders teach them, even if the leaders
tell them that thousands or hundreds of thousands must die so that the philosophy is accepted by everybody else as well.
Humans respect authority and believe that the picture of the world and of themselves that their teachers and the experts show
them is true.
It is true that those in authority often hate each other and compete desperately; rulers and those who want to be rulers often kill
or torture each other. It is true that experts, the learned, scholars, academics often attack each other jealously and bitterly. But
nearly all accept the foundation of their position without question and expect everybody else to do the same. However savage
their competition, rulers all accept the principle of rule. However jealous some may be of each other, practically all scholars and
experts join together in presenting the sort of work they do as the only right sort and their way as the only right basis of
knowledge and truth.
In less savage communities we can see parliamentarians or other notables, from left to right, supposed to be bitter opponents,
sitting round a table together. They scorn each other's pronouncements, but they call each other David, or Robin, or Roy, or
Joan, or Tony, and they are all mates and civilized, and it is they and their kind who decide the way things are to be done,
always essentially the same way, because that is the way that puts

Page 258
them and keeps them where they are; and if occasionally someone from outside says something that questions the very basis of
their position, they all ignore it or mock it quickly into oblivion. In some parts of the world rulers have even got together and
passed resolutions that when a new ruler throws out an old one he should not kill him and should show him compassion, a
compassion that these same rulers do not often show to the ruled.
Academics and experts are at the moment not anywhere near as harmful and dangerous as politicians. But in the end they could
become the most dangerous people of all, because of their claims as creators of knowledge and keepers of the truth. Rulers tend
to be guided more and more in their view of man by what 'scientists' tell them, and the ruled perhaps accept rule more easily
than ever because they too believe what the 'scientists' say about their nature. In the end, even, the 'scientists' may themselves
become the rulers.
I want to repeat something I said at the beginning of this book. What do the experts expect and intend? They presumably neither
expect nor intend the whole population to understand their knowledge. Most experts do not even seem to respect argument and
debate based on their own or others' general or 'popular' presentations of their work. They say such popularizations cannot fully
present their work in its proper complexity. This is really saying: this is how things are and you must just accept what we say
about them because unless you are one of us experts you can never really understand; we are the ones who know; all we are
doing in books and articles and talks for the general public is just telling you. That does not entitle you to have valid opinions of
your own on any subject involved. So even assuming that any of the experts' findings are true, what is to happen? Are we all
simply to believe everything we are told and lead our lives according to the pattern these wise people draw up for us? Because
we are not intellectually equipped to follow their 'science', must we obediently follow the directions of our betters? This is a
question that anyone who insists that important truths are complicated and only attained by scholarly analysis must answer.
Today the only counter to academic pronouncement, in both West and East, is religious fanaticism. There is no alternative
rational force. We need one.
It would be very reasonable for you to comment that in this book I have by no means always followed my own principles for the
use of language. As I have already admitted, I have often cheated. I have not always limited

Page 259
myself to dispassionate rationality, or at least an attempt at this. I have sometimes been polemical and rude; I have used emotive
words.
If I have appeared contemptuous of persons, I am sorry. I am not contemptuous of any person. I am only contemptuous of ideas,
attitudes, and practices. It is not my wish or purpose to attack anybody as a human being. Tenderness is due to every human
being, whoever they may be, and whatever they have done. Every one of us is the victim of the way things are in our
communities. No-one has the right to judge an individual for the way they react to the things done to them and around them.
But the authority that protects falsehoods and the consequences of those falsehoods has no right to respect.

Page 260
Bibliography
Below I list the printed or spoken material that I have referred to in the text, together with a number of the relevant books,
articles and radio talks that have provoked or interested me most. I have not included any of the very many letters to newspapers
or non-specialist periodicals that I have often found at least as stimulating as anything else, nor, obviously, the many
conversations I have had over more than thirty years with people holding a variety of opinions on language and linguistics,
although those conversations have often been as revealing and important as any written material for my understanding of these
things.
I have deliberately also not included any of the considerable number of articles in learned journals that I have gone through. I
have not included them because although they are nearly always revealing - monotonously so - I do not want to lure readers into
suffering the same tedium and irritation I suffered. Even more important, I do not want them to think I feel it is necessary to
study such stuff in order to have valid opinions.
Artificial intelligence [talks on], BBC radio broadcasts, 28 February, 7, 14, 21, 28 March 1983.
Baby talk, BBC radio broadcasts, ca.1984
Berry, M., 1975. An introduction to systemic linguistics. 1.Structures and systems, B.T.Batsford.
Burgess, A., 1979. The metabolism of words [review of N. Smith & D.Wilson, Modern linguistics], in The Observer, 29 July
1979.
Butterworth, B., 1983. Biological twists to grammar [review of D.Lightfoot, The language lottery], in New Scientist, 20 October
1983.
Campbell, J., 1982. Grammatical man: information, entropy, language, and life, Allen Lane.
Chomsky, N., 1959. Review of B.F.Skinner, Verbal behaviour, in Language, 35.
Chomsky, N., 1976. Reflections on language, Fontana.
Chomsky, N. [talking to John Maddox], BBC radio broadcast, 21 October 1981.
Chomsky, N. & Halle, M., 1968. The sound patterns of English, Harper & Row.
Chomsky, N. & Magee, B., 1978. Noam Chomsky on the genetic gift of tongues, in The Listener, 6 April 1978.

Page 261
Cook, J.L., Gethin, A. & Mitchell, K., 1980. A new way to proficiency in English, Blackwell, 2nd edn.
Coulthard, M., 1985. An introduction to discourse analysis (Applied linguistics and language study), Longman.
Crystal, D., 1981. Linguistics, Penguin Books, revised edn.
Curtiss, S., 1977. Genie. A psycholinguistic study of a modern day ''wild child", Academic Press.
Davy, J., 1969. The Chomsky revolution, in The Observer Review, 10 August 1969.
Fowler, R., 1968. The rat box philosophy [talk on Skinner and Chomsky], BBC radio broadcast, 29 October 1968
Gardner, H., 1976. The shattered mind: The person after brain damage, Vintage Books.
Geschwind, N., 1972. Language and the brain, in Scientific American, April 1972.
Geschwind, N., 1979. Specializations of the human brain, in Scientific American, September 1979.
Hampshire, S. & Chomsky, N., 1968. Noam Chomsky [talk followed by discussion], BBC radio broadcast, 17 October 1968.
Laidler, K., 1980. Giving apes a voice of their own, in The Guardian, 26 June 1980.
Lyons, J., 1977. Chomsky, Fontana, revised edn.
Magee, B., 1982. Men of ideas, Oxford University Press.
Melberg, A., 1973. Strukturalismen är ingen humanism [Structuralism is no sort of humanism], in Dagens Nyheter, 3 January
1973.
Miller, J., 1983. States of mind: Conversations with psychological investigators, British Broadcasting Corporation.
Moorehead, C., 1985. The most important attributes [interview with Randolph Quirk], in The Times, 20 May 1985.
Pragnell, M.O., 1984. Whacking the sacred cows of logic for six [review of J.Barwise & J.Perry, Situations and attitudes], in
Computer Weekly, 1 March 1984.
Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J., 1985. A comprehensive grammar of the English language, Longman.
Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J., 1972. A grammar of contemporary English, Longman.
Quirk, R. & Greenbaum, S., 1973/1979. A university grammar of English, ELBS/Longman.

Page 262
Searle, J., 1984. 1984 Reith Lectures - Minds, brains and science. 2. Beer cans and meat machines, in The Listener, 15
November 1984.
Simons, G., 1982. Whatever turns you on [machines with emotions], in The Guardian, 8 July 1982.
Smith, N. & Wilson, D., 1979. Modern linguistics: The results of Chomsky's revolution, Penguin Books.
Webster, R., 1981. Structuralism and dry rot, in The Observer, 1 February 1981.

Page 263
Index
In this index many topics will be found listed as sub-entries under the following headings: grammar; language; learning foreign
languages; learning language; linguistics; logic; meaning; meanings; pronunciation; sentences; stress; structure; teaching foreign
languages; thinking; word order; words. n = note
A
"a" 77, 81, 84-87, 203
A-over-A principle 122
abilities, see language; learning foreign languages
abstract phrases 116-19
academics and academic world (see also universities) 1-2, 37, 206-07, 255-59
monopoly 255-59
solidarity 37, 257-58
acceptability of sentences, see sentences
accidence 105
accounting for 16-17, 25-27, 29-30, 169
African languages 56-58, 191
ambiguity 25-27, 32
American girl 178-82
analysis (see also grammar; language; structure) 1, 91-92, 169
invention through 26-27
as proof of reality 38
"and" 49-50, 56-57
angels and pins 39
animal diagram 38-39
animals, relationships with 247
animals' categories 222-23
"any" 87-89
apes 139n, 192
Arabic-speakers 68
articles 77-87, 108, 135-36, 203
articulate people, unfair advantage enjoyed by 240, 249
Aspects-type grammar 21-22
atrocities 224-25, 228-31, 257-58
authority 1-2, 37, 255-58
solidarity of those in 37, 257-58
awareness 197, 248
objective 195, 220-21
B
Barwise, J. 163
"because/for" 48-50
behaviourism 4, 8n
Belgians 67
"believe", double meaning of 51-52
Berry, Margaret 19-20, 37
bicycle riding 182, 217-18
bi-lingual people 177
biology 3
brain
areas for thought and language 218
and the use of language 9-11, 138, 140, 192, 214-18
study of, through language study 34, 36, 151, 204-06, 218-19
Bresnan, Joan 101
Brazil 61
Britain 68
Burgess, Anthony 33-34

Page 264
"but" 49-50
C
Campbell, Jeremy 40-41
case grammar 164-65
cases 8
categories, grammatical 77-89
categorization, danger of 214, 231-38
children (see also learning language) 10, 91, 137n, 145
enslaved by words 238
Chinese room 250
choice 20, 201-02
Chomsky, Noam 3-4, 13-14, 21, 24-26, 40-43, 51-55, 58-59, 94, 99, 114-126, 131-36, 139-42, 151, 153, 156-57, 165, 184-85,
188, 190-92, 194, 206-07, 219, 220n
absolutism of 53
classification 6, 38-39, 77-89
by words 214, 231-38
clauses
logic of 130
subordinate 48-50, 130
co-ordinate 48-50
codes and codifying 25-31, 37, 191
competence, see language, non-linguistic and linguistic factors in
complementizers 123-25
composers 216-17
compound stress rules 98-100
comprehensibility input 61
A comprehensive grammar of the English language 75-89
computer model of thought 213
computers 20-21, 90-91, 247-49
concepts dependent/independent of language 195-97
constituent analysis 17, 166
context 49-50, 53-55, 67, 84, 88, 101-03, 107, 112-13, 138-39, 155, 176n, 198, 200, 222, 226, 232
continuous, present perfect 36
cooking 216, 218-19
co-ordinate structures 56-58, 190
Coulthard 36
Crystal, David 15-17, 38-39, 132, 140, 164-68
Curtiss, Susan 182-83
D
deception through attractive language 239-40
definition 5n, 6-7, 38, 124, 233-34n, 242-51
of language, see language, principles of
degenerate evidence 173, 182-88
Descartes 197n, 210
descriptive adequacy 34
dictionaries 31, 176n
discourse analysis 36
dolphins 230-31
dummies 105-11
Dutch 67-68
E
"eager" 151-55
"easy" 151-55
echo-questions 45
effects of experts and writers (see also experts) 2-4, 195n, 232, 257-58
Einstein 200
emotions created and maintained by

Page 265
words 224, 228-31
emphasis, see stress
English 8, 44, 62, 89, 93, 114-15, 128, 186, 188n, 206
English as a foreign language, see learning foreign languages; teaching foreign languages
English-speakers 76
entailments 161-63
environment, effects of 180-81
Eskimo 223
event and meaning, distinction between 5-6, 157-68
evidence, simple 61-62
examinations in English 62-63
experience and language (see also language; thinking) 5, 220-21
experiences
combinations of and new words 227-28
new 132
experts 1-4, 139, 174, 232, 255-58
F
facts, see event
faculties (see also language faculties; language faculty)
breakdown of 217
general and special 216-18
fantasies, grammatical 26-27
Fe'Fe' 57-58, 191
"few" 77
Fillmore, Charles 164-65
fingers 216
Flemish-speakers 68
"for" 42-43
"for/because" 48-50
Fodor, Jerome 197n
formal systems 11-37
French 44, 56, 104, 128, 176, 187-88, 206
Freud 59
future tense sentences 64
G
games 189
Gardner, Howard 217-19
gender 104-05
general meanings, see meanings, general
generative semantics 165
Genie 180-84
German 174, 186
Germanic languages 8, 68, 186
gerund, see "-ing"
grammar (see also language; meaning; meanings; meanings, general; sentences; structure; word order)
abstract analysis of, see analysis
acceptability of, see sentences
arbitrariness of rules of 147-48
Chomskyan, see linguistics, Chomskyan
classificatory 77-89
complicated and simple rules of 186, 190-91
explaining 34, 36, 48, 65-67, 125-27, 252
fantasies 26-27
as general meanings, see meanings, general
generative (see also grammar, transformational) 21-34, 131-50, 166-69
immediate constituent 17, 166

Page 266
individual judgements on 53, 134-39, 144
knowledge of one's own 34, 36, 94, 133-36, 139-40
mathematical analysis of 139
and meaning (see also meanings, general) 67, 94, 130
phrase structure 22-25, 28-29, 166
rules of (see also language and situations) 32, 66-67, 94, 105n, 132, 137n, 142, 147, 188-92
structuralist 151
systems of, various 14-21
traditional 93, 105
transformational 14, 21-34, 120-24, 148, 151, 153-56, 159, 166-69, 183-84, 186-88
universal 12, 14, 26, 40-59, 115-22, 152n, 170, 182-84, 186-93
grammars
complete 12, 142
of English 75-89
evaluating 189-90
as reference books 76
grammaticality, see sentences
Greenbaum, Sidney 75-89
H
habit 105n
"habit" 227-28
Halle, Morris 99
Halliday 17
Hampshire, Stuart 115, 138
happens, thinking about what actually 182, 206-07, 214, 223, 226, 229, 232-33, 237-38, 244-45, 248-50, 253, 255
"have something done" 65-66
Hebrew 44
helping each other to understand 249-52
Holy Trinity 200-01
human nature 1
humans, as language users and cause of suffering 230-31
I
ideas distorted by (good) language 201-02, 238-40
ideologies created and maintained by words 228-31
idiomatic usage 185-86
"if" clauses 66-67
if-ishness in thought 211-12
illogical expressions and sentences 9, 55, 138, 146-47
imagination 140-41
imitation 8, 107
inborn language faculty, see language faculty
incomprehensibility 37
Indo-European languages 68
infinitives 32-33, 40-41, 105-07
information theory 20
"-ing" 25-26, 32-33, 65, 76, 93, 112, 192
inherited characteristics 180-81
inside eye 209
"instinct" 227-28
intelligence 178, 180-82, 230-32
"intelligence" 230-32
"intuition" 225-26
intuitions about language 134-35
invention through analysis 26-27
inversion 8, 67n, 147-48, 184
IQ 184

Page 267
"is" contraction 44-48
"it" 109-11
Italian 93, 179
"it's" 9
J
Jackendorff, Ray 189
Japan 68
Japanese 41, 44, 55, 68, 93, 174, 198, 205
jargon 36, 78, 83
K
knowledge 1-2
elementary 61-62
failure of and danger from 254-55
L
labels, avoiding 214, 236-38
Lamb 16
language (see also grammar; learning foreign languages; learning language; logic; meaning; meanings; pronunciation; stress;
sentences; structure; thinking; word order; words)
abilities at different ages 171-82
analysis of 11-39, 77-78, 91-92, 169
and the brain 9-10, 11, 34, 36, 138, 140, 151, 192, 204-06, 214-19
and computers, alliance of 249
constraints on 40, 53, 115, 120, 122, 124, 133, 148, 184, 187, 246
conventions 8, 67n, 105, 128, 147, 184, 218
'correctness' in 60
corpus of 137
corruption by, see thinking; words
creation of by individuals 37
creativity of 14, 131-42
damage to happiness by 246-47
definition of, see language, principles of
deprivation 178-82
description of 5, 11, 29, 31-32, 34, 37
disorders 11, 218
disputes about the use of 241
distrust of 230, 249, 253
domination by 2-3, 245-47, 249
and events, distinction between 5-6, 157-58
explanations of 34, 36, 48, 65-67, 125-27, 252
faculties, general, controlling use of 138, 141, 215-18
faculty, independent, innate, programmed 40-59, 115-23, 133, 137, 146-50, 169-70, 180-93, 214-18
failure of and harm done by 3, 254-55
features of 93
good, no proof of good ideas 239-40
habits, individual 217
ideas changed and obscured by 201-02, 238-40
infinite character of 131-33
as instrument 9-10
intuitions about 134-35
as invention 217-18

Page 268
knowledge of, where to study 134-36
learning, see learning language; learning foreign languages
limited nature of 201-03
limiting effect of 222, 225-26
meaning alone fundamental to, see meaning, primacy of
mistakes in the use of 133-34, 182-83
misunderstandings caused by 35, 240-41
models of 19, 37
and moral problems 35, 223-25, 228-41
non-linguistic and linguistic factors in 34, 101, 103-04, 133-34, 136-39, 143, 151, 181
origin of 11
people good at 240, 249-50
pigeon-holing nature of 222
poorer than life 239-40
power of 2-3, 245-49
principles of 3, 5-10, 63-66, 93-94, 114-30, 132-33, 193
program, see language faculty
psychology of 20-21, 26-27, 29-31, 34, 36, 138-40, 146-51, 170-93
and reality (see also meaning and reality) 5, 13, 200-02, 220-29, 232-38, 242-53
as reflection of mind 34, 36, 151, 192, 204-06, 218-19
respect for 2-3, 215-16, 246-47, 249, 255
restrictions on, see language, constraints on
rules confused with situations, see language and situations
simple 250-51
and situations, confusion between, in formulating rules 13-14, 66-67, 77, 86n, 135-36, 148-49, 164n
source of information about 134-36
as a straight line 114, 199-200, 202
study, purposes of 11
theory of 5, 9-10, 12, 14
and thought, see thinking
Latin 8, 63
languages 68
learning foreign languages (see also teaching foreign languages) 34, 61-90, 145, 170-78, 185-86
abilities in, varying 67-69
conditioning by native language 171-74
immigrants 175-76
principles of 63-66, 69, 75
for special or specific purposes 61
standards in 62-63
and structure analysis 63-67, 148
learning language 8-9, 45-46, 49, 107, 116-19, 123, 137n, 147, 170-93
by analogy 145-46
data on which based 133, 145-46, 173, 182-88
general faculties needed for 170, 180-82, 188-89, 193, 216-18
Page 269
motivation for 173
observation in 8, 49, 107, 132, 146, 177-78, 184, 186, 193
stages in 178-81, 187
Leech, Geoffrey 75-89
lexicon 30-31
life, human, improvement in conditions of 254
"likely" 145
linguisticians, see linguists
linguistics (see also language) 2-3, 11-39, 59-75, 89-92, 156
applications of 89-92
artificiality of 19-20, 58-59
'choice' in theories of 20-21
Chomskyan 3-4, 12-14, 21-34, 40-59, 90, 94-104, 112, 114-26, 131-34, 136-63, 166-69, 170-71, 178-93, 204-05
complexity of rules of 37, 45-46, 63-64, 154, 187-89
controversy within 37
generalizations made possible by 31-32
illogical defence of 59
influence of on tolerance 60
and language teaching 61-75
physical sciences, compared with 144
as psychology 36
schools, systems and theories of 14-21, 37-38, 63
uselessness of 19-21, 34-39
linguists 3, 11-39
changes of mind of 156-57
circular arguments of 137, 143-45, 151-55, 159, 188
double standards of 59
mistakes by 40-59, 143n, 191
lists 13
literary studies 91-92
literature, scholarly 2, 258
logic
capacities, individual variations in 138
of children in creating new words 137n
formal 161-64
'inside', of phrases 9, 146-47
of life and reality in language (see also language) 5, 8, 21, 64-66, 95-98, 103, 107, 116, 123, 127, 129-30, 137-38, 148-49
practical versus abstract 5, 137-39
of sentences and meaning 53, 55, 130, 145
Lordat 200-01
Lyons, John 20, 24-25, 33, 122, 131-32, 134, 141-44, 152, 154, 159-61, 166, 191
M
Magee, Bryan 52, 245, 252
Marxism 1, 60, 250
mathematical analysis of language 139
mathematicians 212
mathematics 254
McCarthy, John 249
meaning (see also grammar; language; logic; meanings; sentences; structure; word order)
as basis of language, see meaning, primacy of

Page 270
versus classification in grammar 77-89
Chomskyan refusal to
acknowledge importance of 40-59, 99
combinations of 6-8, 64, 80, 84-87, 93-98, 137, 145
combinations of, determined by life logic 64-66, 95-98, 137-38
combinations of, in different languages 93
dependent on deep structure 31-32, 153-54
differences in, practical and/or basic 33, 49-50, 64-65, 107-08, 157-59
and event, distinction between 5-6, 157-58
and grammar (see also meanings, general) 67, 94, 130
and logic, confusion between 161-64
lumps of 118-23, 146-47
primacy of 5-8, 40-59, 63-67, 80, 95-104, 106, 114-30, 132, 136-37, 148-49, 154-55
and pronunciation 44-48, 98-104
and recognition of reality 64-66, 96-98, 112-13, 118-23, 128-29, 149, 154-55, 170, 185, 196-97
attaching of to sounds 140
versus structure 25-27, 47-48, 93-130, 151-69
system, invention of 140
in systemics 18-20
in teaching languages 63-67
unnecessary words for 105
wholes of 9, 146-47
word order as form of 8, 31, 129
meanings (see also grammar; language; logic; meaning; sentences; structure; word order)
common and rare in life 13-14, 66-67, 77, 86n, 135-36, 148-49
general (= grammar) 6-8, 64, 93, 95-98, 112-13, 137-38, 142, 146, 149, 185
in different languages 93, 146
particular, decisive for whole meanings 6-8, 33, 54, 64-65, 77, 80, 84, 86-88, 112-13, 143n, 148-49, 159
specific 6-8, 64, 95-98, 112-13, 137-38, 142, 146, 149, 185
meaningless structural elements (dummies) 105-11
meaninglessness 94-98
memory 170, 184-86
short term 138n
mentalism 134
meta-theory of universal grammar 184, 186
Metternich 226
Middle Ages 226
Middle East 68
mind, see brain; language, reflection of mind
minimal syntactic units 105
misunderstandings 35, 240-41
moral philosophers 252-53
moral problems and morality 35, 223-25, 228-41, 252-53
muscles 9-10
music 189, 206, 217

Page 271
N
name giving, evils of 214, 224-38
native language, conditioning by 171-74
native-language teaching 91
neurologists 214
neutral scientist, hypothetical 116
new utterances, see language, creativity of
new ways of doing things 172
Newton 3
non-linguistic contrasted with linguistic factors in language, see language, non-linguistic and linguistic factors in
Norwegian 68-69, 104, 187-88
noun-phrase-and-noun-phrase structures 56-58
Nuclear Stress Rule 100
numbers and numbers faculty 208-09, 217-18
Nupe 56
O
object
logical/underlying and grammatical/surface 151-53
relationship to various verbs 203-04
observation, see learning language, observation in
''one" 76-77
order of words, see word order
Orwell 14n, 251
P
paradigmatic relationship 14-15
parsing 93
passives 30-31, 158, 168-69
patterns, recognition of 67n, 147-48, 170, 182, 188-89
perceptual strategies, see language, non-linguistic ...
performance principles, see language, non-linguistic ...
Perry, J. 163
Persian-speakers 68
PhD's 60
philosophers 4, 9, 163, 206, 208, 214, 243-44, 252-53
philosophy 3
false problems of 243
linguistic 243-45, 250
phonetics and phonology, see pronunciation
phrase markers 23-24, 29
phrase-structure 22-25, 28-29, 166
phrases, abstract 116-19
Pike 15-16
politics 3, 257-58
popular works 258
pragmatic factors in language, see language, non-linguistic ...
preconceived models 19
present participle, see "-ing"
priests 1
"probable" 145
professors 4
pronouns 203
pronunciation (see also stress)
conventions of 67n
and meaning 44-48, 99-104
of English "oo", Norwegian and Swedish 68-69
of foreign languages 174, 176
by different races 61
and structure 28, 45, 110-19
study of phonetics 89-90
propositions 161

Page 272
psycholinguistics 91
psychology and psychologists (see also language, psychology of) 3, 36, 214,
Q
question forming 52-55, 115-25,
questions, indirect 125-26
Quirk, Randolph 75-89, 248-49
Quirk et al.'s grammar of English 75-89
R
rationality and reasoning 1-2, 252, 255
as alternative to research 61-62
Ravel 218-19
reactions to events included in language 5-6, 157-68
reality, see language and reality; meaning and recognition of reality
recognition of reality and meaning, see meaning and recognition of reality
of the thought of others 249-52
reference words unused in thought 203
relative clauses in various languages 43-44
relative pronouns 203
religion 1
religious fanaticism 258
repetition, so-called 'recursive structures' 33
research and researchers 2, 39, 60-62, 255
rigidness and rule-boundness 13-14, 148-49
rules, see grammar, rules of; language and situations; word order
rulers 307
Russian students 230-31, 250
S
Sacks 36
Sakharov 200
Saussure 14, 17
scale and category grammar 16-17
Scandinavian passives 169n
schools, language 72-75
science and scientists 1, 144
scientific theory, generative grammar as 142
Searle, John 190, 197n, 206-07, 222, 249-50
semantic space 38
semantics, see meaning; meanings
Semitic languages 68
sense, making 53, 57-58, 193
sensitivities of expression 76
sentences
acceptability of 13-14, 44-59, 76, 94-98, 131-50
deriving 23-24, 124, 166-68
'impossible' 13, 28-30, 40-59, 65, 77, 94-95, 117, 120-26, 130, 142, 146
operations on 115-24
subjective judgement of 53, 134-39, 144
related 25-26, 30, 124, 158-59, 166-69
restrictions on, see language, constraints on; logic of sentences
Shebalin 219
"since" 64

Page 273
skills, learning at different ages 171
Skinner 4, 8n
Smith, Neil 28-34, 36, 43-46, 48-49, 56-58, 98-103, 132, 137n, 140, 143-46, 148, 153, 161-63, 172-73, 180-82, 185, 191-94,
206n
"so" 148
social science and scientists 1
sociolinguistics 35, 91
"some" 87-89
sound in the mind 210
The sound patterns of English 99
Soviet Union 230, 250
Spanish 41, 93, 105, 176, 187-88, 206n
specific meanings, see meanings, specific
stimulus and response theory of language 107
stratificational grammar 16
stress (see also pronunciation)
importance of for meaning 54-55
and "is" contraction 47
rules 34, 100-01
shift 34
structure and meaning 98-104
structure (syntax) (see also grammar; language; sentences; word order) 8, 11-39, 45, 93-130, 142
analysis and foreign-language learning 63-67, 148
analysis versus meaning 93-130, 147-48
deep 22, 26-27, 31-32, 101, 151-69, 184
dependency 115-17
dependent on meaning 48, 96-97, 118, 123, 155n
meaningless elements of (dummies) 105-11
obsession with 40-59
passim, 93-169
passim, 143n
surface 151-56
structures, identical, with different meanings 112-13
subject
logical/underlying and grammatical/surface 151-53
missing 40-41
Suzuki 220n
Svartvik, Jan 75-89
Swedish 8, 68-69, 104, 186-88
symbol-creating 216
symbols, thought is not system of 211-12
synonymity 32, 106-08
Syntactic structures 142
syntagmatic relationship 14-15
syntax, see structure
systemics and 'systems' 17-20
T
tagmemics 15-16
Tanner, Tony 248
taxonomy 14
teaching foreign languages (see also learning foreign languages) 30-31, 61-90
course books 70-71
practicality and psychology 31
principles and qualities needed 72-75
techniques and approaches 69-72
training courses 73-75

Page 274
teaching the native language 91
technical matters 139
technology 254
tests 62-63
"the" 78-87, 108, 135-36, 203
categories of 79-80
students' problems with 83-84
"there" 109-11, 148-49
things contrasted with words 98, 207-12
thinking (see also words) 2, 27, 194-252
computer model of 213
corruption of by language 5n, 20, 194, 201-02, 206-07, 220-41, 243-44
endlessness problem in 213
experiences of without language 198-99
introspection about processes of 198-99, 207-14
language, parts of unnecessary in thought 203
and language (see also words) 11, 27, 194-207, 214-41, 243-44
in a language 201
multi-dimensional 199-200
as pictures 208-13
precedes language 194-97, 220-21
'shunting' in 212-13
symbols 211-12
'transformed' into language 27
trying to express accurately 242-52
unlimited nature of 201-02
thought, see thinking
"thought", distortion by the word 214
time in the mind 210-11
"to" 105-06
tolerance of language variants 60
transformational-generative grammar, see grammar, transformational
transitive verbs 30-31
translating, thought, and language 197-98
translation 90-91, 248
true and false utterances 4
truth 1
confused with comprehension 38
trying to express 242-53
truths, plain 2, 4
Turkish 68
U
ultimate constituents 105
unacceptable sentences, see sentences
unclear cases in generative grammar 142-45
underlying strings 23-26, 32
ungrammaticality, see sentences
United States 61
universal grammar, see grammar, universal
universals, relative 190-91
universities (see also academics) 1
linguistics departments 72-75
A university grammar of English 83-84, 86
university students, conditioning of 255-56
unnecessary words 105

< previous page page_275
Page 275
V
vocabulary learning 176
W
war games 27-28
"weather" 77
Webster, Richard 248
Weizenbaum, Joseph 251
Welsh 169n
wh-movement 44-46, 51, 124
whales 232-33
"who...?" 51-55
Williams, Bernard 246, 252-53
Wilson, Deidre 28-34, 36, 43-46, 48-49, 56-58, 98-103, 132, 137n, 140, 143-46, 148, 153, 161-63, 172-73, 180-82, 185, 191-
94, 206n
Wittengenstein 197-98, 210
word forms 169
word order (see also grammar; language; sentences; structure) 8, 31, 44-45, 48-50, 56-57, 94, 114-29, 178-79
conventions of 8, 67n, 128, 147, 184
inversion 8, 67n, 147, 184
and logic of life 127
as form of meaning 8, 31, 129
necessary 127-29
obsession with 93, 114, 179
restrictions on 53, 115-28
rules 56-58
words (see also language; thinking) anticipating and distorting thought and feelings 233
associations, false, created and transferred by 231-35
attitudes determined and modified by 232-36
basic evil of 238
classificatory, as substitutes for thought 214, 234-38
combinations of that prevent thought 223-24
emotions created and maintained by 224, 228-31
freedom from 249-51
as easy and false generalizations 232-37
ideas changed and obscured by 201-02, 238-40
ideologies created and maintained by 228-31
inflexibility of 231-33
new and false, invented through combinations of experiences 227-28
'physical' 211, 250
problems created by existence of 234
as relacements for thought 243-44
responsibility to 250-51
seeing behind 250-52
single, corruption by 225-26
would-ishness 211-12
writing
good, no proof of good ideas 239-40
standards 91
< previous page page_275

Amorey Gethin - Antilinguistics - A Critical Assessment of Modern Linguistic Theory and Practice - Intellect Books (1990)

Uploaded by

Copyright:

Available Formats

You might also like

Amorey Gethin - Antilinguistics - A Critical Assessment of Modern Linguistic Theory and Practice - Intellect Books (1990)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Amorey Gethin - Antilinguistics - A Critical Assessment of Modern Linguistic Theory and Practice - Intellect Books (1990)

Uploaded by

Copyright:

Available Formats

cover next page >

title: Antilinguistics : A Critical Assessment of Modern Linguistic

cover next page >

< previous page cover-0 next page >

First published in Great Britain in 1990 by

< previous page cover-1 next page >

< previous page page_i next page >

intentionally left blank

< previous page page_iii next page >

< previous page page_1 next page >

< previous page page_2 next page >

< previous page page_3 next page >

< previous page page_4 next page >

< previous page page_5 next page >

< previous page page_6 next page >

< previous page page_7 next page >

< previous page page_8 next page >

< previous page page_9 next page >

< previous page page_10 next page >

< previous page page_11 next page >

< previous page page_12 next page >

< previous page page_13 next page >

< previous page page_14 next page >

< previous page page_15 next page >

< previous page page_16 next page >

< previous page page_17 next page >

< previous page page_18 next page >

< previous page page_19 next page >

< previous page page_20 next page >

< previous page page_21 next page >

< previous page page_22 next page >

From these 'phrase-structure rules' they get 'underlying strings', e.g:

With such a 'string' they 'associate' a 'phrase marker':

and claim that they have 'derived'1 a passive 'string':

< previous page page_23 next page >

< previous page page_24 next page >

< previous page page_25 next page >

< previous page page_26 next page >

< previous page page_27 next page >

< previous page page_28 next page >

< previous page page_29 next page >

< previous page page_30 next page >

< previous page page_31 next page >

< previous page page_32 next page >

< previous page page_33 next page >

< previous page page_34 next page >

< previous page page_35 next page >

< previous page page_36 next page >

< previous page page_37 next page >

< previous page page_38 next page >

< previous page page_39 next page >

< previous page page_40 next page >

< previous page page_41 next page >

< previous page page_42 next page >

< previous page page_43 next page >

< previous page page_44 next page >

< previous page page_45 next page >

< previous page page_46 next page >