MARIANA NEAGU CLAUDIA PISOSCHI

FUNDAMENTALS OF SEMANTICS AND

PRAGMATICS
2
MARIANA NEAGU CLAUDIA PISOSCHI

FUNDAMENTALS OF SEMANTICS AND

PRAGMATICS

EDITURA UNIVERSITARIA
Craiova, 2015

3
Referenți ştiinţifici: prof. univ. dr. Ioana MURAR
prof. univ. dr. Adriana COSTĂCHESCU

Copyright © 2015 Editura Universitaria

Toate drepturile sunt rezervate Editurii Universitaria

Descrierea CIP a Bibliotecii Naţionale a României

NEAGU, MARIANA
Fundamentals of semantics and pragmatics / Mariana
Neagu, Claudia Pisoschi. - Craiova : Universitaria, 2015
Bibliogr.
ISBN 978-606-14-0939-6

I. Pisoschi, Claudia

811.111

4
 FOREWORD

The intention of the authors was to design a book for the

background of students in Philology who study English Semantics and
Pragmatics and also for all those who want to get (more) acquainted with
the essential aspects of communication from the perspective of the intricate
relationship between form and meaning, on the one hand, and between
these two facets of language and the user, on the other.
The fields of Semantics and Pragmatics are very broad and diverse
and the duration of the course required a strict selection of the topics.
Therefore, the book is structured in two complementary parts, one dealing
with the major topics of English semantics, and the other with the traditional
domains of pragmatics.
Part I. Semantics is broken up in five chapters and can be
conveniently used in a one-semester course on English Semantics.
Chapter 1 outlines the scope of semantics and presents an overview of the
main semantic theories: diachronic semantics, structuralist semantics,
semantics in generative linguistics and cognitive semantics. Chapter 2
looks into the relationship between language, thought and reality and
presents two major models of the sign: the Saussurean model and the
Peircean model. Chapter 3 focuses on types of linguistic meaning and
emphasizes their importance in practice (e.g. translation). Chapter 4 deals
with the paradigmatic sense relations studied by the two major branches of
semantics: semasiology and onomasiology. Chapter 5 looks at semantic
organization by discussing the issues of mental lexicon and semantic field.
Part II Pragmatics comprises six chapters dealing with the traditional
domains of pragmatics: deixis, conversational implicatures, presuppositions
and speech acts, politeness. Chapter 1 is an introductory one, approaching
the study area of pragmatics, explaining its main concepts and terms.
Chapter 2 explains the concept of deixis and its basic types, providing
cross-linguistic examples. Chapter 3 approaches the domain of
implicatures, focusig on conversational implicatures and conversational
maxims. Chapter 4 discusses presuppositions as pragmatic inferences,
their types and triggers. Chapter 5 enlarges upon speech acts (concept,
levels, felicity conditions) and performativity. Chapter 6, the last one is an
analysis of the relationship between the concept of politeness as a social
and linguistic phenomenon and the previously discussed pragmatic
domains: deixis, pragmatic inferences and speech acts.

5
 Combining theoretical considerations and relevant examples
(sometimes allowing cross-linguistic comparisons), the present volume is
synthetic, but dense, clear and systematic, but requiring a deep
understanding and challenging the reader to reflect on the issues
approached. If it succeeds in doing that, it means that the authors have
achieved the most important purpose: to arouse and/or increase the
interest for language in general and the desire to be equally accurate and
creative in communication.
Considering the various types of requirements specific to such a
book, we are trully endebted to our peer-reviewers and express our deep
gratitude for their accurate reading and pertinent observations.

The authors

6
 CONTENTS

PART I. SEMANTICS
1. 11
INTRODUCTION............................................................................................
1.1. Scope and beginnings of semantics.................................................... 11
1.2. An overview of main semantic theories.............................................. 12
1.2.1. Diachronic semantics............................................................ 12
1.2.2. Structuralist semantics........................................................ 15
1.2.3. Semantics in generative linguistics.................................... 19
1.2.4. Cognitive semantics............................................................. 21
Conclusions.................................................................................................. 24

2. THE RELATIONSHIP BETWEEN LANGUAGE,

THOUGHT AN REALITY............................................................................... 25
2.1. Intension and extension and related dichotomies............................ 25
2.2. Sign – sense – reference....................................................................... 27
2.3. Models of the sign................................................................................. 30
2.3.1. The Saussurean model......................................................... 30
2.3.2. The Peircean model.............................................................. 31
2.4. Types of signs........................................................................................ 32
Conclusions.................................................................................................. 33

3. LINGUISTIC MEANING: TYPES AND DIMENSIONS............................... 35

3.1. Lyon’s classification: descriptive, social
and expressive meaning ............................................................................ 35
3.2. Leech’s seven types of meaning......................................................... 36
3.3. Descriptive meaning............................................................................... 38
3.4. Non-descriptive meaning....................................................................... 40
3.5. Social meaning...................................................................................... 43
3.6. Evoked meaning................................................................................... 44
Conclusions.................................................................................................. 46

4. SENSE RELATIONS................................................................................. 49
4.1. Semasiology and onomasiology- two basic
approaches to the study of words and their senses ............................... 49
4.2. From word to concept: polysemy and homonymy ............................ 50
4.2.1. Polysemy ............................................................................... 50
4.2.2. Homonymy.............................................................................. 52
4.2.3. Polysemy vs. Homonymy...................................................... 52
4.2.4. Polysemy in Cognitive Linguistics....................................... 55
4.2.5. Polysemy vs. Vagueness....................................................... 56
4.2.6. Polysemy and semantic change .......................................... 57
4.3. From concept to word: synonymy and antonymy ............................. 58

7
 4.3.1. Synonymy ............................................................................. 58
4.3.2. Antonymy .............................................................................. 60
4.4. Hierarchical sense relations: hyponymy and meronymy ................. 63
4.4.1. Hyponymy ............................................................................. 63
4.4.2. Meronymy .............................................................................. 65
Conclusions ................................................................................................. 67

5. SEMANTIC ORGANIZATION.................................................................... 69
5.1. The lexicon............................................................................................. 69
5.1.1. Views of the lexicon ............................................................. 69
5.1.2. Lexical vs conceptual knowledge ....................................... 70
5.1.3. Lexical item, lexical unit and lexical entry .......................... 71
5.2. Semantic fields ...................................................................................... 71
5.2.1. Field theories ......................................................................... 71
5.2.2. Lexical gaps .......................................................................... 73
5.2.3. Conceptual field, lexical field and semantic field .............. 74
Conclusions ................................................................................................. 78

PART II. PRAGMATICS

1. THE DOMAIN OF PRAGMATICS ……………………….………………….. 79
1.1. Why do we need pragmatics?
Correctness vs. acceptability ……………………………………………… 79
1. 2. The domain of pragmatics ………………………………………………. 83
1. 3. Connections to the other linguistic branches ……………………….. 85
1. 3.1. Connections form - pragmatic meaning …………………... 85
1. 3.2. Connections sense - pragmatic meaning ………...………. 87
1.4. Subdomains of pragmatics. Some of
their basic terms and concepts …………………………………………. 88
Conclusions ……………………………………………………………………… 95

2. THE CONCEPT OF DEIXIS. TYPES OF DEIXIS ...................................... 97

2.1. The term and concept of deixis. 97
Types of deixis .................................................................................
2.2. Person deixis ........................................................................................ 99
2.2.1. Means of expressing person deixis.
Personal pronouns system ............................................................ 99
2.2.2. The reference of pronouns. Deixis vs anaphora................. 99
2. 3. Time deixis ........................................................................................... 102
2.3.1. Definition ................................................................................ 102
2.3.2. Linguistic markers ................................................................ 102
2.3.3. Time vs tense ........................................................................ 103
2.3.3.1. Absolute and relative tenses .................................. 104
2.3.3.2. Tense and Aspect .................................................... 104
2.3.4. Coding time, event time, reference time.
Deixis and grammar ............................................................. 105
2.3.5. Lexical time markers............................................................. 106
2.3.6. Time vs space deixis............................................................. 109
2.4. Space deixis........................................................................................... 110

8
 2.4.1. Definition and importance .................................................... 110
2.4.2. Classification of space deictics ........................................... 110
2.4.3. Main values ............................................................................ 112
2.4.4. Combined values .................................................................. 114
2.4.5. On time and space deixis (again). Which was first? ......... 114
2.5. Social deixis .......................................................................................... 115
2.6. Discourse/ textual deixis ...................................................................... 118
2.7. Empathetic deixis ................................................................................. 120
Conclusions ................................................................................................. 121

3. CONVERSATIONAL IMPLICATURES……………………………………… 123

3.1. Logic of Conversation. To say vs to imply. From logical
connectors to Cooperative Principle ……….............................................. 125
3.2. Classification of implicatures ……………………………………………. 125
3.3. Conversational implicatures. The Cooperative 129
Principle and conversational maxims……………………………….….
3.4. Domains of manifestation of conversational maxims ………………. 132
3.5. Types of non-observance of conversational maxims ………………. 133
3.6. Tests for conversational implicatures. Drawbacks ………………….. 140
Conclusions ……………………………………………………….……………... 143

4. PRESUPPOSITIONS …………………..……………………………………... 145

4.1. A philosophical approach of presuppositions …………..…………… 145
4.1.1. Presupposition and reference ……………….………………. 145
4.1.2. Presuppositions and entailment …………………………….. 146
4.2. Presuppositions as pragmatic inferences ………………..…………… 149
4.3. Linguistic triggers of presuppositions ………………………………… 150
4.4. Types of presuppositions ………………………………..………………. 152
4.4.1. Existential presuppositions ……………………….…………. 152
4.4.2. Factive presuppositions …………………………………..….. 152
4.4.3. Lexical presuppositions ……………………….……………… 153
4.4.4. Structural presuppositions ………………..…………………. 153
4.4.5. Non-factive presuppositions ……………..………………….. 154
4.4.6. Counterfactual presuppositions …………………………….. 154
4.5. The Projection Problem …………………………………………………… 154
4.6. Tests for presuppositions …………….………………………………….. 157
Conclusions ………………………..…………………………………………….. 160

5. SPEECH ACTS ……………………………………………………..…………. 161

5.1. Introduction …………………………………………………..…………….. 161
5.2. From performative utterances to SAs or vice versa
The Performative Hypothesis ………………………..………………………. 162
5.3. SA Levels. Speech Act Schema …………………………………………. 166
5.3.1. SA Levels …………………….………………………………….. 166
5.3.2. Speech Act Schema (SAS) …………………….……………... 168
5.4. The concept of Speech Act in communication …………….…………. 168
5.5. SA classification: from Austin to Searle ………….…………………… 170
5.5.1. Direction of fit ……………………..……………………………. 170
5.5.2. Illocutionary point …………..………………………………..... 172

9
 5.5.3. Felicity (Happiness) Conditions ………………………..…… 173
5.5.3.1. Felicity Conditions with Austin ……………………. 173
5.5.3.2. Felicity Conditions with Searle …………………….. 174
5.5.3.3. Counterarguments to Felicity 177
Conditions frame …….…………………..………………
5.6. Indirect Speech Acts ………………………..………………..…………… 177
Conclusions ……………………………………………………………………… 180

6. POLITENESS ………………….………………………………………………. 181

6.1. The concept of politeness ……………………………………..…………. 181
6.1.1. Politeness − a social phenomenon …………..……………... 183
6.1.2. The conversational maxim view on politeness …….…….. 183
6.1.3. The face-saving view …………..………………...……………. 185
6.2. Politeness ‘in its own right’. Politeness Principle
and its maxims ……………………………………………………………… 185
6.3. Politeness and deixis ………………………………………….………….. 189
6.3.1. Person deixis markers and politeness……………………… 189
6.3.2. Social deixis and politeness................................................. 190
6.4. Politeness and conversational implicatures ………………………….. 192
6.5. Politeness and presuppositions ………………………………………… 194
6.5.1. The problem of negative ambiguity……..………….…….…. 194
6.5.2. Structural presuppositions and hearer’s manipulation …. 196
6.6. Politeness and (indirect) speech acts ………….………………………. 196
Conclusions…………………………………………………………….……...…. 200

BIBLIOGRAPHY …………………………….…………………………………… 201

10
 1. INTRODUCTION

1. Introduction
1.1. Scope and beginnings of semantics
1.2. An overview of main semantic theories
1.2.1. Diachronic semantics
1.2.2. Structuralist semantics
1.2.3. Semantics in generative linguistics
1.2.4. Cognitive semantics
Conclusions

1.1. Scope and beginnings of semantics

Semantics is the major branch of linguistics which studies meaning

communicated through language (words and sentences). It is concerned
primarily with lexical meaning, grammatical meaning and sentence
meaning. Generally, it is recognized that one cannot account for lexical
meaning without accounting for sentence meaning and viceversa. Thus the
meaning of a sentence depends upon the meaning of its constituent
lexemes and the meaning of some, if not all lexemes depends upon the
sentence in which they occur. Lyons (1995: 144) discusses the importance
of grammatical meaning as a further component of sentence meaning.
The term 'semantics' is of relatively recent origin, being coined in the
late nineteenth century from a Greek verb meaning 'to signify'. This does
not mean that scholars first turned their attention to the investigation of
meaning of words less than a hundred years ago. On the contrary, from the
earliest times down to the present day grammarians have been interested
in the meaning of words and frequently more interested in what words
mean than in their syntactic function. Lyons (1985/1968:400) argues that
the practical manifestation of this interest is the production of innumerable
dictionaries throughout ages, not only in the west but in all parts of the
world where language has been studied.
In spite of the interest in meaning manifested by philosophers,
logicians and psychologists, linguists doubted that meaning could be
studied as objectively and as rigorously as grammar and phonology and
thus semantics came to be neglected and received proper attention only
since the 1960s.
The beginnings of semantics as an independent linguistic discipline
go as far back as early 19th century, to the works of the German linguists
Ch. C. Reisig and Hermann Paul. Reisig was the first to formulate the

11
object of study of the new science of meaning which he called semasiology
and conceived the new linguistic branch of study as a historical science
studying the principles governing the evolution of meaning. Hermann Paul
also dealt extensively with the issue of change of meaning.
The ‘birth date’ of semantics as a modern linguistic discipline was
marked by the publication of Essai de sémantique (1897) where the French
linguist, Michel Breal, defines semantics as ‘the science of meanings of
words and of the changes in their meanings’. However, in 1887, that is ten
years ahead of Michel Breal, Lazăr Şăineanu published a remarkable book
called Încercare asupra semasiologiei limbei române. Studii istorice despre
tranziţiunea sensurilor (Essay on Romanian Semasiology. Historical
Studies on the Transition of Meanings). This is one of the first works on
semantics to have appeared anywhere. Şăineanu amply used the
contributions of psychology in his attempts at identifying the semantic
associations established among words and the logical laws and affinities
governing the evolution of words in particular and of languages in general.
Ferdinand de Saussure's distinction between the two basically
different ways in which language may be viewed, the synchronic or
descriptive and the diachronic or historical approach introduced a new
principle of classification of linguistic theories. The next section will make
an overview of the major theoretical trends in semantics, trying to show
how linguists have been doing word meaning in the last century and a half.

1.2. An overview of semantic studies

Starting from the presence or absence of the referent in discussing

the linguistic sign, semantic theories can be grouped in two major general
approaches: (1) language-intrinsic or language–immanent approaches to
semantics that exclude extra linguistic objects (referents) and relations and
(2) referential or denotational (language-extrinsic) approaches to semantics
that focus on the properties of the referents denoted by the linguistic signs.
In what follows, the former group will be illustrated by structuralist and
generative semantics while the latter group will be represented by
diachronic and cognitive semantics.

1.2.1. Diachronic semantics

Diachronic or historical semantics developed through the literature

on semantic change which had a golden period between the last twenty
years of the 19th century and the 1940s (the 1880s and the 1940s). One of
the longest treaties on semantic change is Gustaf Stern's book Meaning
and Change of Meaning, published in 1931. Stern's principal aim was to
establish a theoretically tenable and practically workable system of
classification comprising all known types of sense change.

12
Stern’s classification of semantic changes

Stern defines change of meaning as

“the habitual modification, among a comparatively large number of
speakers, of the traditional semantic range of a word...to denote one
or more referents which it has not previously denoted or to express a
new manner of apprehending one or more of its referents”. (Stern
1968/1931:162)

Stern starts by classifying a large number of authentic sense changes and

then formulates a theory to account for the existence of the different
classes. In other words, the classes were established inductively rather
than deductively. He analyses historical instances of sense change mainly
with regard to the psychic processes involved and identifies seven main
classes of change: substitution, analogy, shortening, nomination, (regular
transfer), permutation and adequation.
Substitution is a change of meaning due to an external, non-linguistic
cause. For instance, alterations in the design of ships have brought about
changes of meaning in the word ship. It once meant only a sailing vessel;
now it can mean a steam-driven vessel of quite different appearance.
Therefore, the referents of a word undergo some change so that new
referents are added to or substitute old ones.
Analogy occurs when a word assumes a new meaning on the
analogy of some other word with which it is connected derivationally (e.g. the
adjective fast has borrowed the sense “quick” from the middle English adverb
faste), semantically (e.g. the special meaning of low, “non-dogmatic” in Low
Church on the analogy of High Church where high means “dogmatic”.
Shortening is he omission of a word from a compound expression,
the remaining words carrying the total meaning that formerly belonged to
the whole expression: e.g. private is a shortening of private soldier
(common soldier), periodical is a shortening of periodical paper/ review.
Nomination is a change of meaning in which a name is intentionally
transferred from one referent to another. Stern gives as example of
nomination the convention of using proper names for units of
measurement, inventions, or discoveries (e.g. volt, sandwich). Other
examples include place names for products (e.g.champagne, a jersey),
article of dress for person (e.g. mackintosh), habitual expressions for
persons (e.g. jingoes “music-hall patriots who sing jingo songs”).
(Regular) transfer is the unintentional transfer of a word from one
type of referent to another one resembling it. Examples are root as in root
of hair and bed as in river bed.
Permutation is the unintentional shift from a referent to another
brought about by the possibility of interpreting a word in two ways in some
context. Beads in He is counting his beads can mean either “prayers” (the
original sense) or “little balls on a rosary”.

13
 Adequation is the change of meaning resulting from the adaptation
of the meaning of a word to the actual characteristics of the referents.
Stern’s main example is horn, which, in order of historical development of
meaning, denotes (i) “animal horn”, (ii) “animal’s horn used for music”, (iii)
“musical instrument made from animal’s horn” and finally (iv) ”instrument for
producing a certain kind of sound”. The change from (ii) to (iii) is an
instance of adequation. Adequation differs from substitution in that the
immediate shift does not lie in the referent, as in the change from (i) to (ii)
or in the change from (iii) to (iv) but in the speaker’s apprehension of the
referent. As can be noticed, adequation occurs after other sense changes
(e.g. substitution) have taken place.

Ullmann’s classification of semantic changes

Stephen Ullmann (1962) proposes a "better" version of Stern's

classification of semantic changes. Concerning the causes of semantic
changes, Ullmann distinguishes two main approaches: A. Meillet's theory
and Sperber' s theory. In his article Comment les mots changent de sens
(1904-1905) Antoine Meillet maintains that there are three main causes of
semantic change, viz. linguistic, extralinguistic and social. Sperber's
approach is different from other approaches in that he emphasizes the role
of emotion. By seeking in emotive forces the clue to changes in meaning,
Sperber (1923) focuses exactly on what the French philologist had
disregarded. Although Sperber neglected the non-expressive functions of
language, he introduced a new perspective for the understanding of
changes of meaning and their spread. Following Ullmann (1962) we
conclude that the two theories mentioned above are mutually
complementary rather than exclusive.
Ullmann (1962) distinguishes between semantic changes due to
linguistic conservatism and linguistic innovation. When we keep a word, in
spite of the the fact that the character of its referents has changed, we have
- in Ullmann’s terminology - an instance of linguistic conservatism. Warren
(1992: 9) rightly notices that Ullmann’s linguistic conservatism corresponds
to Stern’s substitution.
The semantic changes due to linguistic innovations are grouped into
three main subclasses: transfers of names, transfers of senses and
composite changes. Considering the word a union of name (form) and
sense (content), Ullmann assumes that there are two possibilities: either
the name or the sense of the word may change or be transferred. Both
transfers of names and transfers of senses occur due to contiguity or
similarity relations.
A case of name transfer through sense similarity is overlook which
is related to the sense of oversee. Instances of sense transfer through
sense similarity are antropomorphic transfers like leg of a table, eye of a
needle, bridge head, etc. Sense transfers through contiguity are sail

14
meaning ”ship”, town meaning “its inhabitants”. Such sense transfers based
on similarity and contiguity correspond to the ancient categories of
metaphor and metonymy.
The third subgroup, composite changes, includes all asssociative
links that can be conceived: composite name transfers, composite sense
transfers and sense-name transfers.
In general, more recent approaches to sense change list
generalization (Hughes1989, Berndt 1989, Ungerer and Schmid 1996),
specialization (Hughes 1989, Warren 1992, Ungerer and Schmid 1996),
figurative use (Berndt 1989, Warren1992) and substitution /semantic shift
(Berndt 1989).
Specialization of meaning can be illustrated by the Old English word
fugol which referred to all kinds of birds. Gradually fugol was replaced by
the word bird whose meaning underwent the process of generalization (OE
bryd ‘young bird’ -> Mod. E bird ‘any bird’)
Concerning figurative use we further consider the category BIRD and
the attribute ‘locked in a cage’ characterizing parrots, budgerigars, and the
attribute ‘exotic appearance’ which applies to ostriches, flamingoes, peacocks.
As can be noticed, the metaphorical uses of bird ‘prisoner’ and rare
bird ‘strange person’ rely predominantly on peripheral attibutes rather than
on central attributes such as ‘can fly’ and ‘has wings’ which are at the basis
of bird ‘aeroplane, missile, spacecraft’.
The major attributes of good examples of the BIRD category (‘can
fly’ and ‘has wings’) in Anglo-Saxon times were probably similar to what
they are today. However, there are instances where central attributes of a
category are replaced, normally as a result of extralinguistic changes. This
type of meaning change is traditionally called substitution or semantic shift
or, in cognitive linguistics terms, prototype shift.

1.2.2. Structuralist semantics

Besides diachronic or historical semantics, we also have to consider

synchronic or descriptive semantics which displays the applications of the
principles of structuralist linguistics to the study of meaning (hence the name
of “structuralist semantics”). The central idea of linguistic structuralism is that
natural languages are symbolic systems with properties and principles of
their own and that the form of lexical items is generally arbitrary; the arbitrary
nature of the linguistic sign impels the linguist to describe language as a
conventional system of rules. As the language sign is part of the language
system, it is characterized within the system, in its relations to other signs in
the system. The main strands that have been identified within structuralist
semantics (Geeraerts 2009) are lexical field theory, componential analysis
and relational semantics.

15
Lexical field theory

The most influential book which was published in lexical field theory
is Der Deutsche Wortschatz im Sinnbezirk des Verstandes. Die Geschichte
eines Sprachlichen Feldes (1931) by Jost Trier. Jost Trier’s books, as well
as his other studies, which are influenced by W. von Humboldt’s ideas on
language, represent an attempt to apply some of the Saussurean principles
to semantics. Analyzing a set of lexical items having related meanings and
thus belonging to a semantic field, Trier concludes that they are structurally
organized within this field. For the first time words were no longer
approached in isolation, but analyzed in terms of their position within a
larger unit - the semantic field - which in turn, is integrated, together with
other fields, into an even larger one. The process of subsequent
integrations continues until the entire lexicon is covered. So, the lexicon is
seen as a huge mosaic with no piece missing1. Although Trier's ideas are
enveloped in a mystical cloak (Chițoran 1973: 17) and he believed that
there would be no gaps or overlaps in a lexical field, his book is a valuable
starting point for subsequent lexical field theories.

Componential analysis

The second strand in structuralist semantics, componential analysis

(CA) or the (semantic) feature approach2 goes as far back as 1953 to the
Danish linguist Louis Hjemslev and has a long tradition both in European
structuralism and American anthropology3. Hjemslev and Jakobson were
the first to maintain that componential analysis could and should be
extended from phonology into both grammar and semantics.
This approach assumes that all meanings can be analysed into
semantic components, distinctive semantic features or semes. Semes are
the minimal semantic features which work within a particular semantic
field and serve to structure the field in terms of different kinds of
opposition. They can be expressed with the help of a binary feature
notation using ‘+’ and ‘-‘. Classemes are very general sense-components
shared by lexical items belonging to different semantic fields, e.g.
“animate” / “inanimate”, “male” / “female”4.

1 See Chapter 5. Semantic organization where the issues of semantic field and the lexicon
are discussed in detail.
2 The name ”feature approach” comes from the basic idea that word meanings are given by

sets of features.
3 In the 1950s American anthropologists devised a technique for the analysis of kinship vocabulary.
4 The distinction between semes and classemes concerned scholars that noticed the

interdependence between field theory and componential analysis: Coșeriu, 1967; Pottier,
1974; Lyons, 1977.

16
 Analyzing chaise, fauteuil, canapé, tabouret (roughly equivalent to
chair, armchair, sofa and stool) Bernard Pottier (1964) identifies the following
semes5, i.e. “for sitting upon”, “with legs”, “with a back” , “with arms” and ”for
one person” .
The common components of the set under discussion are “artifact”,
“piece of furniture” and “for sitting upon”. The features that provide the
contrasts, i.e. diagnostic features refer to:
(a) the occurrence or lack of “legs”
(b) the occurrence or lack of a “back”
(c) the occurrence or lack of “arms”
(d) the number of “persons” for which the piece of furniture is designed.

The seme/classeme distinction drawn by European structuralists can

be associated with the distinguisher/marker distinction that has been
postulated by the American generative linguists Jerold J. Katz and Jerry A.
Fodor. Semantic features like (HUMAN), (MALE) express general semantic
properties, enter into the analysis of very many items in the vocabulary, and
are involved in the statement of syntactic rules and of semantic restrictions.
They are usually referred to as markers. Distinguishers represent ‘what is
idiosyncratic about the meaning of a lexical item’ (Taylor 1989: 34). According
to Lyons (1987/1977: 327) they are merely ‘the residue of lexical meaning’ that
is not accounted for in terms of markers.
Roughly markers would correspond to what Nida (1975) had called
diagnostic components and distinguishers to what he had labeled as
supplementary components. A famous example of analysis in terms of
semantic markers and distinguishers is the noun bachelor:
bachelor {n}
a. (human) (male) [one who has never been married]
b. (human) (male) [young knight serving under the standard of another knight]
c. (human) [one who has the first or lowest academic degree]
d. (animal) (male) [young fur seal without a mate in the breeding season]
As can be noticed we can separate out the feature basis of the different
senses of bachelor listing the markers (enclosed in parentheses) and the
distinguishers (noted within square brackets) for each sense.
The similarity between classemes and markers is stressed by Lyons
(1987/1977: 327) who shows that they both represent the part of the meaning
that is systematic for the language. For instance, some of Katz’ markers
(e.g.”male”) are systematic for the language because they play a role in
selection restrictions. Thus, “pregnant” cannot combine with a noun whose

5 Pottier (1974:28) proposes an interesting distinction between specific semes (sèmes

spécifiques), generic semes (sèmes génériques) and connotative semes (sèmes connotatifs).
He uses the term sémantème for the set of sémes spécifiques, classème for the set of sèmes
génériques and virtuème for the set of sèmes connotatifs.

17
meaning contains the component “male”: *That man is pregnant6. In a
comparable way, Pottier and Coșeriu emphasize that it is the classemes that
determine the semantically-based interdependences between nouns and
adjectives or nouns and verbs. For instance, it is the classeme “male” that
determines the selection of the Italian ammogliarsi (rather than maritarsi) or
Romanian a se însura (rather than a se mărita).
Consequently, one of the apparent advantages of the componential
approach has to do with the semantic acceptability of syntagmatic
combinations of words and phrases, i.e whether a given combination is to be
generated as significant or excluded as meaningless. The second advantage
refers to semantic relations reformulated within a componential theory of
semantics. Lyons (1985/1968: 476) rightly maintains that “componential
analysis is a technique for the economical statement of certain semantic
relations between lexical items and between sentences containing them”.
In other words, componential analysis is especially effective when it
comes to representing similarities and differences among words with related
meanings (synonyms, antonyms) or between related meanings of the same
word (polysemantic words). The most comprehensive approach to polysemantic
items in terms of the componential analysis of meaning is provided by Nida
(1975). He views the relations between various types of meanings as being
systematic and argues that there is a close relation between:
a) an instrument and the activity associated with it: brake, chain, hoe,
hammer, knife, motor, rake, rope, saw.
b) a place and the activity associated with it: board a ship, tree a racoon,
bank money.
c) an entity (affected) and the activity of which the entity is the goal: to fish, to
barbecue, to handle, to behead, to go birding.
d) agent- action: governor, learner, student, etc.
The limitations of the CA originate in the fact that in some cases
semantic features cannot provide any insignt into the nature of the meaning
they are supposed to represent. For instance, the meaning of dog can be
characterized in terms of the features “animal”, “canine” but there is no further
analysis of the concept underlying the feature “canine”.
In Coşeriu’s view, lexical field theory has to be supplemented with
the functional doctrine of distinctive oppositions. According to him,
structural analysis should focus on the functional language that is
homogeneous, leaving aside geographic (diatopical), social (diastratal) and
stylistic (diaphatic) variation. Lexical items are opposed to each other and
this oppositional contrast yields specific distinctive features or semantic
components. He uses the term Archisemem to characterize the semantic

6 Similarly, it cannot combine with an item which has the component “inanimate” such as

table ( * pregnant table).

18
features common to all members of a word field and Archilexem for cases
of lexical units representing such features
The third approach in structuralist semantics, relational semantics, was
believed to provide a different type of meaning description apparatus, i.e. one
that is more purely linguistic7, thus being more adequate to the structuralist
conception of meaning. Relational semantics, introduced by John Lyons
(1963), considers that sense relations can provide a more independent
description of linguistic meaning, thus realizing the structuralist intentions:

“a theory of word meaning will be more solidly based if the meaning

of a given linguistic unit is defined to be the set of (paradigmatic)
relations that the unit in question contracts with other units of the
language (in the context or contexts in wich it occurs) without any
attempt being made to set up ‘contents” for these units.”

In our view, relational semantics can also be called ‘paradigmatic’

(Bidu-Vrânceanu et al. 1997: 435) because it studies meaning through
paradigmatic sense relations (i.e. relations between two or more linguistic
elements belonging to the same grammatical category that are mutually
exclusive within one and the same linguistic sequence) such as synonymy,
antonymy, homonymy, polysemy and hyponymy.
Like the previous two strands, i.e. lexical field theory and
componential analysis, relational semantics has the disadvantage of not
being able to systematically make a distinction between the linguistic level
and the referential (encyclopedic) level of content description. For instance,
hyponymy is a genuine sense relation, because it can be defined in terms
of inclusion between senses, while meronymy (partonymy) holds between
entities (e.g. hand-finger), not between senses. Another semantic relation
whose understanding relies on situated (contextual) knowledge that is
“encyclopedic and textual rather than structural and purely linguistic”
(Geeraerts 2009: 84) is oppositeness of meaning. For instance, nature may
contrast with art in one context or with civilization in another. Consequently,
the explicit aim of providing a truly structuralist account of meaning,
involving an independent level of description, is hard to attain.

1.2.3. Semantics in generative linguistics

A turning point in the history of lexical semantics is the formulation

of a semantic theory by Jerrold J. Katz and Jerry A. Fodor (1963), who

7 In the Componential Analysis model, descriptive features such as “gender” and

“generation” in the kinship semantic field are real world features, describing characteriscs of
the referents, not of the language structures.

19
introduced componential analysis into generative grammar8. Their paper,
entitled “The structure of a semantic theory” raised significant questions
about the use of formal methods in the description of meaning and about
the cognitive (psychological) reality of meaning. Besides, the incorporation
of lexical semantics into a formal grammar adds syntagmatic relations (i.e.
relations that hold between members of different grammatical categories
which are simultaneously present in a single syntactic structure) to the set
of phenomena to be considered (e.g. the lexical properties and the
semantic relations between the terms in a lexical field).
As initially there was no place reserved for semantics in generative
linguistics (See Chomsky, Syntactic Structures, 1957), Katz and Fodor’s
incorporation of a formalized semantic description into the formal theory of
grammar brought about a major shift of perspective for generative
linguistics. Geeraerts (2009: 103) observes that

“Katz and Fodor demonstrated that a formalized semantic description

could be incorporated into the generative framework. They were
successful to the extent that, next to the traditional syntactic and
phonologic component, Chomsky explicitly incorporated a semantic
component in his Aspects of the Theory of Syntax (1965), the so-
called Standard Theory of generative grammar.”

Interpretive (syntagmatic) semantics holds that the basic structure of

a sentence is a syntactic one and semantics only comes in as an
interpretation of those syntactic structures. The main focus of interpretive
semantics is syntagmatic semantic relations; this is why interpretive
semantics is also called syntagmatic semantics. The representatives of
interpretive/syntagmatic semantics (Jerrold J. Katz and Jerry A. Fodor
1963, Noam Chomsky 1965 and Ray Jackendof 1986 are concerned with
possible combinations of particular words and with restrictions on possible
combinations of meaning, the so-called “selection restrictions”, i.e.
semantic restrictions on the choice of individual lexical units in construction
with other lexical units (e.g. pregnant will typically ”select” a subject
referring to someone or some animal that is female). Syntagmatic
semantics also deals with the meaning of complex linguistic expressions,
including sentences. This explains why some scholars refer to this type of
semantics as ‘sentence semantics’.

8In his Dictionary of Linguistics and Phonetics, David Crystal (1992: 151) defines generative
grammar as „a set of formal rules which projects a finite set of sentences upon the
potentially infinite set of sentences that constitute the language as a whole and it does this in
an explicit manner, assigning to each a set of structural description.”

20
 Generative semantics, by contrast with interpretive semantics,
assumes that the underlying representation of a sentence is a semantic
one. One of its main concerns is lexical decomposition, i.e. the analysis of
word meanings into smaller units which are seen as standing to one
another in constructions like those of syntax.
Generative semanticists such as James D. McCawley (1968, 1973),
and George Lakoff (1970, 1971) argue that lexemes have an internal
structure like the syntactic structure of sentences and phrases. For
example, the sense of kill can be analysed into CAUSE, BECOME, NOT
and ALIVE; these elements are not simply conjoined but are combined in a
hierarchical structure which may be represented as (CAUSE (BECOME
(NOT ALIVE))).
As in their concept of meaning they do not normally look beyond
language, the schools of interpretive and generative semantics must also
be considered language-immanent approaches to semantics.

1.2.4. Cognitive semantics

Cognitive semantics has developed in the 1980’s on the basis of
findings in cognitive psychology. The main difference between structuralist
semantics and cognitive semantics is that the former analyzes meaning
from a purely language-internal perspective (i.e. on the basis of semantic
networks connecting lexemes), whereas cognitive semantics explains
meaning primarily in terms of categorization (i.e. the grouping of similar
phenomena into one class). Concerning the structure of word meaning, the
classical approach advocates that lexical concepts9 are well delineated
entities whose definitions are expressed in terms of an invariable set of
necessary and sufficient features applicable to all instances in that concept.
Cognitive semantics does not exclude that some lexical concepts may be
analysed in terms of necessary and sufficient features. For example, the
necessary and sufficient features of the concept square are: “closed figure”,
“four sides”, “sides equal in length”, and “equal angles”.
On the cognitive view instances of a concept may not be linked
because they all share the same features, but because they share different
sets of features with each other. The features linking the various instances
of a lexical concept have been called ‘the family resemblance relationship’.
Cognitive semantics is the semantic approach of linguists who see no
separation between linguistic knowledge and general thinking, or cognition.
Cognitive linguists tend to adopt a functional view of language, as opposed
to the more formal accounts favoured by Chomsky and similar generative

9 Evans, Vyvyan (2006), in his article 'Lexical concepts, cognitive models and meaning

construction' published in Cognitive Linguistics (17: 491-534) defines lexical concept as “a

linguistically encoded concept, conventionally associated with a wide range of forms that
provides access to conceptual knowledge. Lexical concepts exhibit polysemy and have their
own selectional (semantic and grammatical) requirements”.

21
linguists. They argue that no adequate account of grammatical rules is
possible without considering the meaning of elements. As such, the
difference between language and other mental processes is viewed as one
of degree rather than kind.
In cognitive semantics, meaning is considered to be inextricably
linked to human cognition, to the way we perceive the world and group
phenomena into conceptual categories. All linguistic meaning is conceptual
in nature and the structure of linguistic categories is held to reflect the
structure of conceptual categories (e.g. in the sense that the meaning of a
word is the cognitive category connected with it). On the cognitive view,
word meaning is not determined by the language system itself, but reflects
how people interact with, perceive and conceptualize the world. Cognitive
semantics suggests that all conceptual information associated with a lexical
item is broadly encyclopedic, that is, it is part of and needs to be
understood against broader cognitive structures. These cognitive structures
have been labeled by using a diverse range of terms: prototype (Rosch,
1978), schema (Barlet, 1932), script (Schank and Abelson, 1977),
experiential gestalt (Lakoff and Jonhson, 1980), global pattern (de
Beaugrande and Dressler, 1981), frame (Fillmore, 1982), idealized
cognitive model (Lakoff, 1983), mental model (Jonhson-Laird, 1983),
cognitive domain (Langacker, 1987).
For Langacker (1987) a prototype is a typical instance of a category,
and other elements are assimilated to the category on the basis of their
perceived resemblance to the prototype; there are degrees of membership
based on degrees of similarity; a schema is an abstract characterization that
is fully compatible with all the members of the category it defines (so
membership is not a matter of degree). Schemas are generalizations
extracted from specific instances, involving elaboration rather than extension,
as in the case of prototypes. Script is usually applied to a stereotyped
sequence of actions that constitute a global event, such as a visit to a
restaurant or dentist, a birthday party, etc. Experiential Gestalts are,
according to George Lakoff and Mark Johnson, “the natural products of our
bodies, our interactions with the physical environment, and our interactions
with other people in our culture” (Lakoff and Johnson 1980:117). Taylor
(1989: 87) defines frame as “the knowledge network linking the multiple
domains associated with a linguistic form” and maintains that frames are
static configurations of knowledge. An Idealized Cognitive Model (ICM) is a
theoretical construct developed by George Lakoff and refers to a
conventionalized mental representation of reality as perceived and
interpreted by our senses or as determined by culture. They are called
idealized because “they abstract across a range of experiences rather than
representing specific instances of a given experience” (Evans 2006: 104).

22
Johnson-Laird (1983)10 views mental model as “a form of semantic
representation that plays a central and unifying role in representing objects,
states of affairs, sequences of events, the way the world is, and social and
sociological actions of daily life. It relates words to the world by way of
conception and perception” (Johnson-Laird 1983: 397). Finally, a cognitive
domain or conceptual domain has been defined as a conceptual area
relative to which a semantic unit is characterized.
Through the introduction of new models of description and analysis
of meaning such as prototype theory and frame semantics, and through the
renewal of metaphor studies by the Conceptual Metaphor Theory, cognitive
semantics has had a wide appeal among lexical semanticians. An idea
central to Cognitive Linguistics is that of construal, a cover term referring to
the nonobjective facets of linguistic meaning that has come to be used for
different ways of viewing a particular situation. Arie Verhagen (2007)11
considers that

“At a very elementary level, construal is a feature of the meaning of

all linguistic expressions, if only as a consequence of the fact that
languages provide various ways for categorizing situations, their
participants and features, and the relations between them. Speaking
thus always implies a choice. … The fact that a particular situation
can be construed in alternate ways should, from a cognitive linguistic
perspective, not come as a big surprise or require extensive
justification. What is more important linguistically is that languages
systematically provide means for different kinds of construal.”
(Verhagen 2007: 49)

Prototype semantics

According to prototype semantics, word meanings contain the

properties of cognitive categories, i.e. we can distinguish central and more
peripheral meanings of a lexeme; word meanings are not rigid, there are
often gradual transitions between word meanings. Prototype semantics is
closer to psychological reality than traditional feature semantics (or all or
nothing semantics).The advantages of prototype semantics over feature
semantics does not diminish the usefulness of the latter for the description
and comparison of word meaning, especially for identifying semantic
structures like lexical fields and sense relations. Prototype theory and the
family resemblance model cannot do without a feature-based classification.
Ultimately, prototype and feature semantics complement each other, in the
sense that feature semantics receives a sounder psychological basis.

10 Johnson-Laird, Philip. 1983. Mental Models. Cambridge: Cambridge University Press.

11 Verhagen Arie. 2007. “Construal and Perspectivization” in Oxford Handbook of Cognitive
Linguistics (eds. Dirk Geeraerts and Hubert Cuyckens). Oxford: Oxford University Press, pp. 48-81.

23
 Although prototype semantics is particularly adequate for the
description of concrete (extra linguistic) objects, especially those in which
shape and size are relevant, it cannot capture connotative feature or deal
with deictics, relational words and syntagmatic relations (restrictions or
transfer of features). Nevertheless, it has clear advantages in comparison
with feature (Aristotelian) semantics. Thus, the prototype approach can
explain: (1) vague, fuzzy category boundaries; (2) gradual category
membership; (3) categories with prototypical kernels; (4) the different
importance attributes.
What is ultimately needed in semantic theory is an integration of
both language-intrinsic and denotational /referential approaches. It is only
in this way that the limits and boundaries of either traditional structuralist
semantics or prototype semantics can be transcended.

Conclusions

In this chapter we have taken a brief look at some major

developments in the history of linguistic semantics. We noted that, in its
initial stage, semantics focused mainly on changes of meaning; thus,
historical semantics has a lasting theoretical significance as it highlights the
dynamic nature of meaning: meanings do not stay the same, as language
is applied in new circumstances and contexts.
The second major approach in the history of semantics, structuralist
semantics, was discussed in terms of three main strands or methodologies:
lexical field theory, componential analysis and relational semantics. Shifting
the focus on semantic change, specific to historical/diachronic semantics, to
the description of synchronic phenomena, structuralist semantics had a
major impact on lexical semantics through its change from a semasiological
(from expression to meaning) to an onomasiological perspective (from
meaning to expression). It viewed the vocabulary of a language as a network
of expressions that are reciprocally related by various semantic links.
The third development presented in the chapter, semantics in generative
linguistics, showed that the incorporation of semantics into generative
grammar could lead to discussions of issues regarding both cognitive and
formal adequacy.
Finally, the approach emerged in the 1980s and known as cognitive
semantics, successfully transferred research results in cognitive
psychology on the internal structure of categories onto the structure of
lexical categories. Two of its main tenets, the first related to the internal
structure of lexical categories (prototype structure, family resemblance
structure and radial network structure) and the second related to the
polysemous nature of lexical items and the cognitive principles motivating
the relations between the different senses of lexical items will be discussed
in more detail in Chapter 4.

24
 2. THE RELATIONSHIP BETWEEN LANGUAGE,
THOUGHT AND REALITY

2.1. Intension and extension and related dichotomies

2.2. Sign – sense – reference
2.3. Models of the sign
2.3.1. The Saussurean model
2.3.2. The Peircean model
2.4. Types of signs
Conclusions

Long before linguistics existed as a discipline, thinkers were

speculating about the nature of meaning. For thousands of years, the
question ‘what is meaning?’ has been considered central to philosophy.
More recently it has come to be important in linguistics, as well. The next
subchapter discusses three pairs of key concepts in semantics, which are
more or less overlapping: intension and extension, sense and reference,
denotation and connotation.

2.1. Intension and extension and related dichotomies

Contributions to semantics have come from a diverse group of

scholars, ranging from Plato to Aristotle in ancient Greece to Putnam and
Frege in the twentieth century.
The impossibility of equating a word's meaning with its referents has
led to a distinction between intension and extension or Sinn (sense) and
Bedeutung (reference), a distinction introduced by Gottlob Frege (1892).
The extension of a term corresponds to the set of entities that it
picks out in the real world while the intension of an expression is those
properties which define it, its mental content independent of context. For
instance, the extension of "tiger" is the set of tigers in the real world, while
its intension corresponds to the inherent sense of the term, to the concept
that is associated with it: „large carnivore of the cat family, having an
orange-yellow coat with numerous back stripes”.
The term reference designates the relation between an entity in the
external world and the word/expression that is used to pick out this entity (e.g.
person, object, event, place, point in time, etc.). The referent is the entity
referred to by an expression in a particular context; it is a member of the class
of objects that constitutes the word’s extension12.

12 Sometimes the term extension is used interchangeably with the term denotation.

25
 The sense13 of a linguistic expression is its content without
reference, those features and properties which define it. For example, the
sense of “girl” is a bundle of semantic features: /+human/, /-adult/,
/+female/.
The reference of a sign may differ from the sense as in the example given
in Frege (1892): The Bedeutung of 'Evening Star' would be the same as
that of 'Morning Star', but not the sense. The expressions morning star,
evening star and planet Venus refer to the same object, i.e. they have the same
reference. Their sense, however, is different, because they do not have the
same defining properties. Thus,
- morning star could be defined as „bright planet seen in the eastern sky
when the sun rises”;
- evening star as “bright planet seen in the western sky when the sun
sets”;
- Venus as “the planet second in order from the Sun and nearest to the
Earth”.

Types of reference
Non-referring expressions
There are cases when to the sense does not correspond a
reference, i.e. in grasping a sense one is not certainly assured of a
reference: e.g. sign words such as unicorn, Santa Claus, hobbit, dragon,
elf, fairy, World War III, have no referents in the real world even though
they are far from being meaningless.

Expressions with constant reference

Some expressions have the same referent across a range of
utterances: e.g. the Eiffel Tower and the Pacific Ocean. Expressions of this
type are called „expressions with constant reference”.

Expressions with variable reference

Other expressions have their reference entirely dependent on
context: e.g. the Pope, The President of Romania, my neighbour, I, you,
here, there, now, tomorrow). These expressions are called „expressions
with variable reference”. To identify who is being referred to by pronouns
like “I”, “you”, etc., we certainly need to know a lot about the context in
which these words were uttered. These words whose denotational
capability requires contextual support are called deictic words14. Actually,
most acts of referring rely on some contextual information. For example, to
identify the referent of the nominal the President of Romania, we need to
know when it was uttered.

13 Kortmann (2005: 197) believes that the sense of an expression is „its descriptive meaning.”
14 The term deixis comes from Greek and means roughly ‘pointing’.

26
Denotation and connotation

‘Denotation’, like ‘reference’, is the relationship that exists between

an expression and entities, properties and situations external to the
language system. However, unlike ‘reference’, ‘denotation’ is not bound to
the context, and “holds independently of particular occasions of utterance”
(Lyons 1987/1977: 208). For example, the expression the cow is context-
dependent, and ‘refers to’ a particular cow in the way the expression cow
alone, which ‘denotes’ the class of all cows, cannot. It follows that
reference’ indicates the actual persons, things, places, properties,
processes and activities, i.e. ‘referents’, in a particular context of utterance
whereas ‘denotation’ indicates “the class of persons, things, etc., generally
represented by the expression” (Palmer 1986/1976: 18).
Connotation is the variable, subjective, often emotive part of the
meaning of an expression. For example, the connotations of the lexical item
night might include romantic, lonely, uncanny. Connotations are secondary
meanings which can vary according to culture, religion, social class and
which are often restricted to particular contexts.

2.2. Sign-sense-reference

In their book, The Meaning of Meaning (1923), Charles Ogden and

Ivor Richards represent meaning as a model that shows how linguistic
symbols are related to the objects they represent. Ogden and Richards argue
that the symbol corresponds to the Saussurian "signifiant" (E. signifier). They
use the term reference for the concept that mediates between the symbol/
word/expression and the referent. The triadic concept of meaning was
represented by Ogden and Richards in the form of a triangle.
thought/reference

symbol referent

Most linguists agree that a sign (word, expression or symbol)

expresses its sense, stands for and designates its reference. Coşeriu
(1981) stresses the importance of the distinction between signification or
meaning (“Bedeutung”) and designation (“Besiechnung”). Designation, for
him denotes the relationship between the full linguistic sign, combining
signifier (French signifiant) and signified (French signifié) and the
extralinguistic object or referent. The signifier is the "shape" of a word,
provided by the sequence of phonemes (e.g. /d/, /o/, /g/ or the sequence of
letters that make it up (d-o-g). The signified is the ideational component, the

27
concept that appears in our minds when we hear or read the signifier, e.g. a
small domesticated feline. The signified is not to be confused with the
"referent". The former is a "mental concept", the latter the "actual object" in
the world. As signification (meaning or Bedeutung) alone is believed to be
significant for structural semantics, Coşeriu’s theory excludes extralinguistic
objects and relations and is therefore restricted to language itself; therefore,
it can be characterized as a “language-intrinsic” or “language- immanent
approach to semantics” (Lipka 1990: 99).
The regular connection between a sign, its sense and its reference
is of such a kind that to the sign there corresponds a definite sense and to
that in turn a definite reference, while to a given reference (an object) there
does not belong only a single sign. The same sense has different
expressions (lexicalizations) in different languages (E. table, Fr. table,
G. Tisch, It. tavola) or even in the same language (pass away, die, kick the
bucket). The association of two or more forms with the same meanings
(synonymy) and the association of two or more meanings with one form
(homonymy and polysemy) show that one can hardly find an ideal language in
which words are defined by a one-to-one relation between signified (Fr. signifié)
and signifier (Fr. signifiant).
The arbitrariness of the link between the signifier (Fr. signifiant) and the
signified (Fr. signifié) or the arbitrary nature of the sign, was, for Ferdinand de
Saussure (1916), the first principle of language. This principle states that there
is no inherent, essential, transparent, self-evident or natural connection
between the signifier and the signified – between the sound of a word and the
concept to which it refers15: „the process which selects one particular sound
sequence to correspond to one particular idea is completely arbitrary”
(Saussure 1983/1916: 111).
Along with Chandler (2007: 25), we believe that the arbitrariness of the
sign is a radical concept because it establishes the autonomy of the language
in relation to reality.
However, if linguistic signs were to be totally arbitrary in every way,
language would not be a system and its communicative function would be
destroyed. While the sign is not determined extralinguistically, it is subject to
intralinguistic determination in the sense that signifiers must constitute well
formed combinations of sounds which conform with existing patterns within the
language in question. In the same Cours de linguistique générale, De Saussure
introduces the idea of degrees of arbitrariness:

Not all signs are absolutely arbitrary. In some cases, there are factors
which allow us to recognize different degrees of arbitrariness, although

15 For example, there is nothing „treeish” about the word tree.

28
 never to discard the notion entirely. The sign may be motivated to a
certain extent.
(De Saussure 1983/1916: 130)

As an example we can mention onomatopoeic words that are not completely

arbitrary, although different languages use different words for the sounds made
by familiar animals.

The famous triangle of meaning of Ogden and Richards (1923) stands

for a model of an analytical and referential definition of meaning; it has been
referred to in hundreds of subsequent works and has had a powerful
influence on semantic thinking.
Nevertheless, Ullmann (1962: 56) contends that "for a linguistic study
of meaning the basic triangle offers too little or too much". As a diachronic
semanticist, he observes that the meaning of words may change as new
knowledge is generated without a corresponding change in the referent or
real world entity (for example, atoms remain unchanged while our knowledge
of their structure has increased considerably in the present century).
Ullmann indirectly advises linguists to confine their attention to the left-
hand side of the triangle, i.e. on what he calls name and sense, corresponding
to the set 'lexeme-concept' (Magnusson and Persson16 1986: 257) or 'form -
content' (Warren 1992: 76). The implication is to neglect the right-hand
element, i.e. the thing (Ullmann 1962: 56), entity (Magnusson and Persson
1986: 257) or referent (Warren 1987: 76), leaving us with a simplified model.

sense

name thing

This model corresponds to an intra-linguistic attitude to the study of

meaning where there is no room for extension, i.e. the relation between the
symbol and the real world entities to which it refers. More recently, cognitive
linguists have shown that the various beliefs that people may have about real
world entities are crucial to their understanding of word meaning.
It is possible that we can use different expressions to identify the
same referent. For example, William Shakespeare can be identified by the

16 Magnusson, U. and G. Persson.1986. Facets, phases and foci: studies in lexical relations
in English. Stockholm: Almquist and Wiksell.

29
expressions the writer of the play „Hamlet” or the Elisabethan playwright born
at Stratford-upon-Avon. These two expressions share the same referent but
have different meanings. A description that uniquely describes an individual
was called a definite description. In the description theory associated in
various forms with the philosophers Bertrand Russel (1967), and Gottlob
Frege (1989) a name is taken as a label or shorhand for knowledge about the
referent. In this theory, understanding a name and identifying the referent are
both dependent on associating the name with the right description. For a
description to be a definite description, it almost always has to be
accompanied by the definite article. In addition, the description needs to be
singular. For example, the description “the children” is not a definite
description.

2.3. Models of the sign

Any communication, whether it is between animals or humans, takes

place by means of signs and is studied in semiotics. Semiotics is the systematic
study of signs which analyzes verbal and non-verbal systems of human
communication as well as animal communication. The adjective semiotic was
coined by the English philosopher John Locke in the early 17th century. In its
widest sense, a sign may be defined as a form associated with meaning.
Present-day semiotics arises from the independent work of two
linguists, one in Switzerland, the other in the United States. In Switzerland,
Ferdinand de Saussure (1857 - 1913) coined semiology as part of his interest
in language as a system of signs, while Charles Sanders Peirce (1834 - 1914)
used the term to describe the study of signs and symbolic systems from a
philosophical point of view.
Semiotics is a comprehensive discipline, in that almost anything can be
a sign: clothes, hair-styles, type of house or car owned, accent and body
language. All send messages about such things as age, class, and politics.
Therefore, signs may take the form of words, images, sounds, odours, flavours,
etc., and become signs only when we invest them with meaning. The two major
models of what constitutes a sign are those of the structuralist linguist
Ferdinand de Saussure and of the American philosopher Charles Sanders
Peirce.

2.3.1. The Saussurean Model of the Sign (1916)

Focusing on linguistic signs, De Saussure defined a sign as being

composed of a signifier (Fr. signifiant), i.e. the form that a sign takes and a

30
signified (Fr. signifié), i.e. the concept to which it refers. De Saussure referred
specifically to the signifier as a sound pattern (Fr. image acoustique)17.
He stressed that sound and thought (or the signifier and the signified)
were as inseparable as the two sides of a piece of paper. Saussure’s model
brackets the referent, i.e. it excludes reference to objects existing in the world.
His view of meaning is structural (any sign has two structural levels, that of
signifier and of signified) and relational rather than referential: the meaning of
signs was seen as lying in their systematic relation to each other rather than
deriving from any inherent features of signifiers or any reference to material
things. For example, the meaning of tree depends on its relation to other words
within the system, such as bush, therefore his relational conception of meaning
is differential as he emphasizes the differences between signs. Saussure’s view
of the relational identity of signs lies at the heart of linguistic structuralism.
With De Saussure, the principle of arbitrariness, already discussed
in section 2.2, does not mean that an individual can arbitrarily choose any
signifier for a given signified. The relation between a signifier and its
signified is not a matter of individual choice; if it were, then communication
would become impossible. What has to be stressed is that the relationship
between the signifier and the signified is conventional, i.e. it depends on
social and cultural conventions that have to be learned.

2.3.2. The Peircean Model of the Sign (1931)

Unlike the Saussurean model that is dyadic, i.e. has a twofold

structure, the Peircean model of the sign is triadic18, i.e. it has three parts: (1)
the form which the sign takes; (2) the sense made of the sign and (3) a
referent. The sign is a unity of what is represented (the referent or the object),
how it is represented (the form or the representamen) and how it is interpreted
(the sense or the interpretant). The representamen is similar in meaning to
Saussure’ s signifier (Fr. signifiant) while the interpretant is analogous to the
signified (Fr. signifie):

interpretant

representamen object

17 Roman Jakobson and other subsequent theorists refer to the form of a sign as either
spoken or written.
18 Prior to Peirce, a triadic model of the sign was employed by Plato, Aristotle, the Stoics,

Francis Bacon etc.

31
The broken line at the base of the triangle is intended to indicate that there
is not necessarily any observable or direct relationship between the sign
vehicle and the referent. What is also interesting to note is that Peirce’s object
is not confined to physical things; it can include abstract concepts and fictional
entities. A variant of the Peircean model of the sign was adopted by Charles
Ogden and Ivor Richards (1923). As we have seen in the previous section, it is
presented as the semiotic triangle.

2.4. Types of signs19

According to Charles Sanders Peirce, the relationship between a form

and what it represents (or, in Saussurian terminology between a signifier and its
signified) can be of three types: (1) a relationship of similarity (e.g. between a
portrait and its real life object or a diagram of an engine and the real life
engine); (2) a relationship of close association, not infrequently causal
association (e.g. the smoke as an indication of fire) and (3) a conventional link,
an arbitrary relation. Starting from these types of relationship that may hold
between a sign and the object it represents, C. S. Peirce makes a distinction
between iconic, indexical and symbolic signs.
An iconic sign or icon (from Greek eikon 'replika') resembles the
referent and provides a perceptual, e.g. visual, auditory, or any other perceptual
image of what it stands for: a portrait, a cartoon, onomatopoeia, metaphors,
sound effects in radio drama, a dubbed film soundtrack, imitative gestures. This
type of sign is a highly motivated one.
An indexical sign or index (from Latin index 'pointing finger') stands for
what it points to; this link can be observed or inferred: medical symptoms (spots
indexical of a disease like measles, fever indexical of flu, a rash, pulse-rate),
smoke, thunder, footprints, echoes, measuring instruments (weathercock,
thermometer, clock, spirit level), signals (a knock on the door, a phone ringing),
pointers (a pointing index finger, a directional signpost), recordings (a video, an
audio-recorded voice), personal trademarks (handwriting, catch-phrases). An
index is partially motivated to the extent that there is a connection, usually of
causality, between sign and referent.
A symbol (from Greek symbolon 'a token of recognition') or symbolic
sign does not have a natural link between the form and the thing represented,
but a conventional link. Peirce's symbol is the most arbitrary kind of sign: the
word in language, the formula in mathematics and chemistry, a military
emblem, the dollar sign, a flag, red circles in television, the traffic sign of an
inverted triangle, etc. The last example does not have a natural link between its
form and its meaning “give right of way.”

19 A distinction between conventional signs i.e. the names we give to people and things and
natural signs (e.g. pictures resembling what they depict) dates back to ancient Greece (Plato).
Later, St. Augustine, distinguished natural signs from conventional signs on the basis of an
immediate link to what they signified, e.g. smoke indicating fire.

32
 In language, the notion of arbitrariness holds true for most of the simple
words; however, new words (compounds, derivatives) built on already existing
linguistic material are partially motivated. The notion of motivation refers to
non-arbitrary links between a form and the meaning of linguistic expressions.
In terms of their degree of abstraction, the three types of signs can be
ordered from the most 'primitive' to the most abstract. Indexical signs, which are
said to be the most 'primitive' (Dirven and Verspoor 1998: 3) are restricted to
the 'here' and 'now' and are based on a relation of contiguity between form and
meaning. Body language, traffic and advertising are areas providing examples
of such signs.
Iconic signs are more complex in that their understanding requires the
recognition of similarity between form and meaning. Road signs picturing
children, animals or various vehicles or scarecrows in the fields which birds take
for real enemies, onomatopeic words such as cock-a-doodle-doo, ding-dong are
some instances of iconic signs.
Symbolic signs, based on a relation of convention between sign and
meaning, are the exclusive prerogative of humans. As it has been
acknowledged, people have more communicative needs than pointing to things
and replicating things; we also want to talk about things which are more
abstract in nature such as events in the past or future, objects that are distant
from us, hopes about peace, etc. This can only be achieved by means of
symbols which humans all over the world have created for the purpose of
communicating all possible thoughts (Dirven and Verspoor 1998: 4).
Besides language which stands for arbitrary symbolism, as pointed out
by De Saussure, mathematics is also an instance of the symbolic mode as it
does not refer to an external world; its signifieds are concepts and mathematics
is a system of relations.
The three types of signs are not mutually exclusive, that is, a sign can
be an icon, a symbol, or any combination. For instance, the fact that there are
three elements in the Roman numeral III is an example of iconism, but their
vertical orientation (as opposed to horizontal) is arbitrary. Roman Jakobson
(1968) argues that many deliberate indexes also have a symbolic or indexical
quality, such as traffic lights or the pointing gesture that is not always
interpreted purely indexically in different cultural contexts. Jakobson notes that
the dominance of one mode (or type of sign) is determined by context, in the
sense that the same signifier may be used iconically in one context and
symbolically in another. A sign may be treated as symbolic by one person, as
iconic by another and as indexical by a third. Signs may also shift in
mode/category over time, that is, the relation between signifier and signified is
subject to dynamic change.

Conclusions

In this chapter we have looked at three central pairs of concepts used in

semantics, which are more or less overlapping: intension and extension,

33
connotation and denotation and sense and reference. The first mentioned
terms in these pairs (intension, connotation and sense) relate to the conceptual
side of meaning and to the issue of how to provide a language-intrinsic
definition of meaning. The second mentioned terms (extension, denotation,
reference) relate to extralinguistic reality, to the relation between language and
the world and to the issue of how to provide a language-extrinsic definition of
meaning. For a better understanding of the difference we have presented the
two major models of the sign, the Saussurean model and the Peircean model.
Last, but not least, we have seen the distinction between iconic, indexical and
symbolic signs.

34
 3. LINGUISTIC MEANING: TYPES AND DIMENSIONS

3.1. Lyon’s classification: descriptive, social and expressive meaning

3.2. Leech’s seven types of meaning
3.3. Descriptive meaning
3.4. Non-descriptive meaning
3.5. Social meaning
3.6. Evoked meaning
Conclusions

In the context of language the issue of meaning requires reference

to nonlinguistic factors, such as thought, situation, knowledge, intention and
use. These factors might explain why other disciplines are involved in the
study of meaning along with linguistics: philosophy, psychology, logic,
sociology, theology, etc. David Crystal (1992: 214) rightly observes that
“within linguistics, the role each level plays in the interpretation of a
sentence is often referred to as the meaning of that level”. Thus, we can
identify lexical meaning, the meaning of lexical items or lexemes that
belong to one of the four lexical classes: nouns, verbs, adjectives, adverbs.
Grammatical meaning is the meaning of grammatical structures
such as inflections, constructions or any other units that form closed
classes. For instance, of, in a wall of silence marks the syntactic relation
between wall and silence; or, -s in walls marks the plural of the noun wall.
By contrast with lexical meaning that is frequently concrete, grammatical
meaning in general is abstract: e.g. the meanings of case or tense
morphemes or definiteness marked with the help of a and the.
Sentence meaning is the meaning a sentence has by virtue of the
words it contains and their grammatical arrangement, and which is not
dependent on context. By contrast, utterance meaning is “the meaning a
sentence carries when it is used in a particular context, with referents
assigned to all referring expressions…” (Cruse 2006: 164). The present
work will focus on the first and last types of meaning, i.e. lexical meaning
and utterance meaning, respectively. Several linguists (Lyons 1987/1977,
Leech 1987/1974, Cruse 2004) have proposed ways of classifying meaning
into types and the various proposals are more or less overlapping.

3.1. Lyons’ classification: descriptive, social and expressive meaning

Speaking of language functions, semanticists have mentioned that

language is used for (1) the communication of factual information (2) the

35
establishment and maintenance of social relationships (3) the expression of
our attitudes and personality. Lyons (1987/1977) correlated each of these
functions20, with different kinds of semantic information that is encoded in
language utterances: descriptive, social and expressive:

“Descriptive information or descriptive meaning is factual in the sense

that...it can be explicitly asserted or denied and in the most
favourable instances, objectively verified. An example of an utterance
with descriptive meaning is the statement It is raining here in
Edinburgh at the moment.” (Lyons 1987/1977: 51)

Other terms listed by Lyons used to refer to this type of meaning are
referential, cognitive, propositional, ideational, and designative.
As the distinction between expressive and social meaning is far
from clear-cut, and many authors subsumed both under a single term
(emotive, attitudinal, interpersonal, expressive, etc), Lyons (1987/1977:
51) proposes the term interpersonal for what is common to the social and
expressive functions of language.

3.2. Leech’s seven types of meaning

While Lyons (1987/1977: 51) distinguishes descriptive meaning from

social and expressive meaning, Leech (1987/1974: 23) separates conceptual
meaning from various types of associative meaning (connotative, social,
affective, reflected, collocative) and from thematic meaning.
Conceptual meaning, sometimes called ‘denotative’, ‘cognitive’ or
‘descriptive’ is widely assumed to be the central factor in linguistic
communication. We will discuss this type of meaning in more detail in
section 3.3.
Associative meaning is the meaning which becomes attached to a
word because of its use, but which is not part of its core sense. The
principal types of associative meaning are connotative meaning, social
meaning, affective meaning, reflected meaning and collocative meaning.
Leech (1987/1974: 12) defines connotative meaning as „the
communicative value an expression contains by virtue of what it refers to,
over and above its purely conceptual content”. Connotation is the variable,
subjective, often emotive part of the meaning of an expression. Connotations
are relatively unstable, i.e. they vary considerably according to culture,
historical period and the experience of the individual. For example, the
connotative meaning of woman embraces the putative properties of the
referent according to the viewpoint adopted by an individual (e.g. a feminist

20 Roman Jakobson refer to these functions as referential, phatic and expressive.

36
or misogynist) or a group of people and varies from age to age or from
society to society.
Social meaning refers to what is communicated of the social
circumstances of language use. It relates to the use of language to
establish and regulate social relations and to maintain social roles. This
type of language use is alternatively described as social or phatic
communication. The notion of phatic communication emphasizes
experiences of social interaction and the participation in social linguistic
rituals such as greetings, apologies, condolences, etc. In phatic
communication the verbal interaction has little information value, but
instead plays an essential role in handling social interaction. Examples of
words with social meaning include greetings like hello, goodbye and forms
of address such as sir, madam, pal, mate, love.
Closely related to social meaning, affective meaning consists in
what is communicated of the feelings and attitudes of the speaker/writer,
including his attitude to the listener, or his attitude to something he is talking
about. In a manner comparable to social meaning, affective meaning is only
indirectly related to the conceptual representation. Affective meaning is
more directly a reflection of the speaker’s personal attitude or feelings
towards the listener or the target of the utterance. Affective meaning can be
noted in gosh! and in the differences between father and daddy, policeman
and cop, horse and nag, very small and tiny, etc.,
Reflected meaning is that type of meaning which arises in cases of
multiple conceptual meaning, when one sense of a word forms part of our
response to another sense. We sometimes find that when we use a word
with a particular sense, one or more of its senses is reflected in it. Reflected
meaning allows speakers to indulge in innuendo, ambiguity and the
generation of puns as in I have the body of an eighteen year old. I keep it in
the fridge.
Collocation is the habitual co-occurrence of particular lexical items,
sometimes purely formally (e.g. eke out), sometimes with some semantic
implication (e.g. slim chance). Collocative meaning is the type of meaning
that “consists of the associations a word acquires on account of the
meanings of words which tend to occur in its environment” (Leech
1987/1974:17). For example, strong has a completely different meaning in
strong coffee than it does in strong language where it is usually a
euphemism for swearing.
Therefore, the meaning of an expression containing more than one
meaningful element can be worked out by combining the meanings of its
constituents21. This is what the principle of compositionality actually states:
the meaning of a complex expression in natural language depends on (and

21 Idioms seem to be an exceptional case.

37
can be reconstructed from) the meanings of its parts and the syntactic
relations holding between these parts. This important principle of syntagmatic
(sentence) semantics is attributed to the German philosopher and
mathematician Gottlob Frege (1848-1925); hence the name of Frege’s
Principle.
The last meaning type in Leech’s classification is thematic meaning,
namely, the type of meaning that is communicated by the way in which a
speaker or writer organizes the message, in terms of ordering, focus and
emphasis. Its name points to the notion of theme, used in linguistics as part
of an analysis of the structure of sentences. Theme refers to the way
speakers identify the relative importance of their subject matter and is
defined as the first major constituent of a sentence, seen as a string of
constituents. The process of moving an element of the sentence to the front
of the sentence (fronting) to act as theme, is called thematization or
topicalization22. For example, John was sacked last Thursday and What
happened to John last Thursday was that he was sacked are
propositionally identical, but thematically different, i.e. they differ in terms of
thematic meaning.

3.3. Descriptive meaning

Descriptive meaning refers to those aspects of meaning which

relate directly to denotations of lexical items and the propositional content
of sentences and thus corresponds to an intellectual level of interpretation,
as opposed to one where emotional and subjective interpretation is
involved. This type of meaning has been given various labels such as
logical, propositional, referential, objective, conceptual, denotative,
cognitive and ideational. Each of these labels seems to be accounted for by
the defining characteristics identified by Cruse (2004: 44), in his book,
Meaning in Language. An Introduction to Semantics and Pragmatics.

a. That aspect of the meaning of a sentence which determines whether or

not any proposition it expresses is true or false justifies the labels logical
and propositional. For example, in the utterance Somebody has turned the
bloody lights off which contains both descriptive and non-descriptive
meaning, bloody makes no contribution to the truth or falsity of the
statement. However, in a situation where Somebody’s turned the lights off
is true, Somebody’s turned the lights on would be false; therefore, what off
signifies is part the descriptive meaning of the utterance.

22The distinctions topic vs. comment and given vs. new information are other ways of
analyzing the sentence structure of a message.

38
b. That aspect of the meaning of an expression which constrains what the
expression can be used to refer to, or, from another point of view, which guides
the hearer in identifying the intended referent, motivates the label referential.

c. Descriptive meaning is objective, that is, it is not bound to the here-and -

now of the current speech situation.

d. It is fully conceptualized in the sense that it provides a categorization

which effectively “describes” aspects of experience.

e. The descriptive meaning of a sentence can potentially be negated or

questioned.

Dimensions of descriptive meaning

In discussing descriptive meaning, David Alan Cruse (2004: 47)

assumes that the dimensions along which descriptive meaning can vary are
quality, intensity, specificity, vagueness, basicness and viewpoint.
The dimension of quality can be seen in the differences between red and
green, dog and cat, apple and orange, run and walk, hate and fear, here and
there. Pure differences of quality are to be observed only between items which
are equal on the scales of intensity and specificity which are discussed below.
Intensity characterizes items that designate the same area of
semantic quality space such as, warm-lukewarm-hot-boiling, dirty-filthy, etc.
Variation in intensity is not confined to the domain of qualities, it is also
possible in other areas: scare-fright-horror-terror, mist-fog, beat-thrash.
Specificity shows up when one term (the more general one)
designates a more extensive area of quality space than the other: animal-dog,
kill-murder-execute. Specificity is a property which distinguishes a hyponym
from a hypernym23: the hyponym is more specific, the hypernym more
general. The hyponym gives more detailed information and denotes a
narrower category. Thus, scarlet is more specific than red, sprint than run,
slap than hit, etc. Besides type-specificity, when the more specific term
denotes a subtype included within the more general term, Cruse (2006:
197) believes that there is also part specificity which holds between a
meronym an a holonym24: finger, for instance, is more specific than hand.
Vagueness can be noticed in terms which designate a region on a
gradable scale such as middle-aged in She’s middle-aged vs. She’s in her
fifties. Similarly young in Jane is a young woman vs Jane is in her twenties.
For the notion of basicness, Cruse (2004: 50-51) provides several
interpretations. First, he relates it to concrete vocabulary items whose
meanings „are fixed by their relations with observable properties of the

23 For a full discussion of hyponymy see Chapter 4, section 4.4.

24 The topic of meronymy is discussed in more detail in section 4.4.

39
environment”. Second, basicness involves the distinction between
independence and dependence: dependent meanings, being more
complex, build on more basic meanings. For example, acceleration
depends on the notion of speed, which in turn presupposes the notion of
movement, which builds on basic notions such as physical object, location,
change and time. And thirdly, basicness, viewed in terms of cognitive
psychology, corresponds to basic level category, i.e the level of the
ordinary everyday names for things (e.g. chair), creatures (e.g. cat), actions
(e.g. eating) and properties (e.g. tall)25.
The last dimension of descriptive meaning mentioned in Cruse (2004)
is viewpoint (or vantage point, according to Ronald Langacker) which refers
to the way something is described, depending on the position of the speaker
relative to the thing being described. For example, a box next to a tree, can
be viewed in a number of ways, depending on the posion of the speaker: The
box is in front of the tree, The box is behind the tree, The box is to the left of
the tree and The box is to the right of the tree. The linguistic expressions
which encode as part of their meaning the viewpoint of the speaker at the
moment of the utterance are deictic expressions such as this, that, here,
there, now and then26.

3.4. Non-descriptive (affective) meaning

Affective meaning is that type of meaning that shows how language

reflects the personal feelings of the speaker, including his attitude to the
listener, or his attitude to something he is talking about. Alternative terms
for affective meaning are attitudinal, emotive or expressive meaning. This
type of meaning is often held to fall within the scope of stylistics or
pragmatics. It can also be found in (suprasegmental) phonology as
intonation is one way of conveying attitudes and emotions27. As English
does not have a rich system of grammatical moods (subjunctive, optative,
dubitative), it encodes expressive meaning in much of its vocabulary and in
the prosodic structure of spoken utterances. For instance, words that are
not necessarily expressive, such as still, yet, already, may become
expressive if appropriate intonation and stress are added:

Does she still live in Manchester?

Has the postman been yet?
The railway station had already been closed when we came to live here.

25 For more details on levels of categorisation, see Chapter 3 in Mariana Neagu. 2005.
Cognitive Linguistics. An Introduction. București: EDP.
26 For a full discussion of the topic of deixis see Chapter II in Part II. Pragmatics.
27 The attitudinal function of intonation is described in Mariana Neagu and Roxana Mareș.

2013. Contemporary English Language. Phonetics, Spelling and Vocabulary. București:

ProUniversitaria, pp. 108-110.

40
Cruse (2004: 58) maintains that these sentences seem to be expressively
neutral, but feeling can be expressed prosodically:

Are you still here?

You mean you haven’t done it yet?
Surely she hasn’t gone already?

However, what still, yet and already basically express is not an emotion
proper but an expectation or a set of expectations on the part of the
speaker. Similarly, implicit superlatives such as huge, tiny, beautiful,
brilliant, which are expressively neutral if not stressed, seem to be able to
acquire an expressive element if stressed:

It was absolutely huge.

It was absolutely tiny.

However, there are cases when not all the members of a synonymic series
can be expressively stressed: e.g. baby vs. infant, child, neonate:

Mother and baby are doing well.

Oh, look! It’s a baby! Isn’t it lovely?
? Oh, look! It’s a child/infant/neonate! Isn’t he lovely?

As can be noticed, baby is capable of quite neutral employment and can

also be invested with emotive expressive meaning, usually prosodically. In
contrast, child, infant and neonate are incapable of expressive use,
although their denotative meaning is very close to that of baby.
According to the type of meaning they possess, words may be
divided into (1) those that have only expressive and no descriptive meaning
- the so-called expletives and (2) words that have both descriptive and
expressive meaning.
Expletives can be interjections

Wow! Oops! Ouch! O, hell! Hell’s Bell’s! Bother! Ace! I’ll say!

or they may have an emphasing role within a sentence, expressing anger:

Get that damn dog off my seat!

It’s freezing – shut the bloody window!
You can blooming well put it back where you got it.

Taboo words lend themselves readily to expletive use :

Holy shit! Balls! My arse! Piss off! Bugger me!

41
Some expletives are historically merely euphemistic alterations of taboo
items: Gosh (God), Heck (Hell) Gee whiz (Jesus)

Lexical items that have both expressive and descriptive/propositional traits

are daddy, mummy, paw (in the sense of “hand”), mug (in the sense of
“face”), blubber (in the sense of “weep”), damn (in the sense of
“extremely”), rag (in the sense of “paper of poor quality”):

It was damn cold.

Stop blubbering!
Don’t read that - it’s a rag!

In the last example, rag expresses contempt for the paper in

question. It is fairly common to find pairs of words whose meanings differ
only in that they express different evaluative judgments on their designated
referents28 (or one expresses a judgment while the other is neutral): horse-nag,
car-banger, clever chap-a smart alec, careful with money-mean. That some
of the evaluative meaning may well be expressive is obvious in the following
example sentences:

A: Arthur tried to sell me an old nag.

B: No, he didn’t - it was a perfectly good horse.

A: I hear Arthur’s very mean.

B: No, he isn’t - he’s just careful with his money.

A: Arthur’s a smart alec.

B: No, he isn’t – but he is clever.

It seems that lexical items characteristic of informal style and slang

are more likely to have expressive meaning than items belonging to more
formal styles. Propositional and expressive meanings are considered the
most important types of meaning in language and we can think of them as
what a speaker principally utilizes and directly manipulates in order to
convey his intended message.
Cruse (1986: 274) rightly believes that “every communicative
utterance must transmit as part of its meaning an indication of intended
propositional attitude. Without this, an utterance would be communicatively
dead - it would resemble a proposition ‘entertained’ by a logician”.

28 In French linguistics (Pottier, Kerbrat-Orecchioni), these differences are called

“connotative”

42
3.5. Social meaning

Many semanticists consider expressive meaning and social

meaning not to be clearly separated. The interconnection between
expressive meaning and social meaning can be understood if we realize
that the rules of conduct constraining the expression of feelings or attitudes
in certain social situations and the use of expressive terms, in particular
swear words as terms of address may have severe social consequences.
When dealing with meaning and social interaction in his book
Understanding Semantics, Sebastian Löbner (2002) looks at social
meaning as being „part of the lexical meaning of certain words, phrases or
grammatical forms”. He further states that „an expression or grammatical
form has social meaning if and only if its use is governed by the social rules
of conduct or, more generally, rules for handling social interaction (Löbner
2002: 29). Expressions with social meaning include forms of address,
phrases of greeting or saying good-bye, phrases of apologizing,
acknowledging or answering the phone. We believe that the importance
attached to the use of these expressions depends on the speaker’s
background and the culture he belongs to.
In today’s European languages, with the exception of English where
the you form has come to dominate the entire spectrum of addressing,
most languages possess a distinct deferential form used in addressing
people of higher social status or in order to mark distance. Languages that
use respectul forms of address, identical with the second person plural are:
French (vous), Romanian (dumneavoastra) Czech and other Slavic
languages (vy), Finnish (te).
In other languages, the forms of respectful address are based on a
third person plural form, e.g. German Sie and Danish/Norwegian De or on
frozen paraphrases of an original honorific such as Spanish Usted, plural
Ustedes. In Italian, the 3rd person singular (male or female) pronoun Lei
and the 3rd person plural pronoun Loro are used in more formal situations
to address strangers, acquaintances, older people, or people in authority.
The so-called ‘majestic’ plural is commonly used by cardinals,
popes, (Your Eminence, Your Holiness), the royalty (Your Grace),
governors of states, ambassadors (Your Excelency). These address forms
that indicate social standing in addition to identifying the person addressed,
represent a form of social deixis29, to use a term coined by Levinson (1983:
89). The informal variants tu (in Romanian) and du (in German) have the
same descriptive meaning (i.e. they designate the hearer/ the addressee)
as Dumneavoastra and Sie but differ in social meaning. By the choice of
the pronoun the speaker indicates his social relationship to the

29 A detailed presentation of social deixis is given in Section 2.5., Part II. Pragmatics.

43
addressee(s). The distinction between the two kinds of relationship relevant
for choosing either dumneavoastra or tu in Romanian and Sie or du in
German is also relevant in other respects: it coincides with the use of
surnames with titles vs. first names as vocative forms of address.
In American English, idiomatic, colloquial speech is heavily used on
most occasions, except for public events and fairly formal situations when
they use formal speech. When meeting strangers for the first time
Americans use first names; even the simple greeting Hi is a badge of
informality.
In most Latin American and European societies there are levels of
formality attached to status differences. In Asian cultures, formality is
demanded by greater age as well as by higher status. High formality is a
characteristic of the teacher-student relationship in countries such as Egypt,
Turkey and Iran. The use of personal titles is a way the Germans and the
Mexicans show their position in the social structure, show respect and mark
formality.
Two further expressions with social meaning are please and thanks
(thank you, containing you might be considered as referring to the
addressee and to this extent it also has descriptive meaning). Please marks
a request as polite (it is a formality marker) and indicates, similar to the
forms of address, a certain kind of social relationship between speaker and
addressee(s). Interestingly, phrases like I’m sorry and Nice to meet you
which literally represent descriptions of attitudes, are primarily social and
not expressive.

3.6. Evoked meaning: dialect and register

Approaching the issue of synonymy in his Lexical Semantics, David

Alan Cruse (1986: 282) finds that a potential source of variation among
cognitive synonyms is provided by evoked meaning which is the result of
the existence of different dialects and registers within a language.
Dialectal variation, which is variation in language use according to
speaker, can be of three types: geographical, temporal and social.30 (Cruse
1986: 282). Lexical items that exemplify geographical dialectal variation are
the Scots words glen (valley), loch (lake), wee (small) and dram
(melancholy) that are familiar to most speakers of English outside Scotland
and recognized as Scottish. Other lexical items that have the power of
evoking images and associations of their home surroundings are
Americanisms such as fall (autumn), elevator (lift), apartment (flat). Temporal
dialectal variation is illustrated by the synonymic pairs wireless – radio and
swimming bath - swimming pool. Social dialectal variation involves variation

30 In Romanian, the term dialect is mostly applied to the geographical varieties of language.

44
according to the social class of the speaker. The phrase “U and non-U” has
been coined to refer to upper-class and non-upper class words:

U non-U

dessert afters, pudding

relation relative
potatoes spuds
perspire sweat
sitting room lounge
writing paper note paper
dinner tea
dinner lunch
sofa settee

Kate Fox, in her famous Watching the English. The Hidden Rules of
English Behaviour makes a very interesting remark about the relation
between linguistic choices and social status in England:

“The linguistic codes we have identified indicate that class in England

has nothing to do with money and very little to do with occupation.
Speech is all important. A person with an upper-class accent, using
upper-class terminology will be recognized as upper-class even if she
is earning poverty line wages, doing grubby menial work and living in
a run-down council flat. Or even unemployed, destitute and
homeless. Equally, a person with working class pronunciation, who
calls his sofa a settee, and his midday meal dinner, will be identified
as working class even if she is a multi-millionaire living in a grand
country house. There are other class indicators such as one’s taste in
clothes, furniture, decoration, cars, pets, books, hobbies, food and
drink but speech is most immediate and most obvious. … Words are
our preferred medium, so it is perhaps significant that they should be
our primary means of signaling and recognizing social status. “ (Fox
2004: 82)

The second type of variation which contributes to what Cruse (1986) calls
‘evoked meaning’ is register variation, that is, variation (within the speech of
a single community) according to situation31. Whereas dialects are
language varieties associated with different characteristics of users, (e.g.
regional affiliation, age and class), registers are varieties of language (used

31 Hence, the association of register with context of situation, a key concept in Halliday’s

(1985) approach, defined as the immediate environment in which a text is actually functioning.

45
by a single speaker) which are considered appropriate to different
occasions and situations of use.

Components of register

Register is usually divided into three main components: field, mode

and style32. Field refers to the topic or field of discourse: there are lexical
(and grammatical) characteristics of, for instance, legal discourse, scientific
discourse, advertising language, sales talk, political speeches, football
commentaries, cooking recipes and so on. The difference between expert
(technical) terms and their correspondents (synonyms) in ordinary
language is that the former may have stricter definitions (e.g. extirpate)
while the latter are more loosely defined (e.g. take out). Terms that differ
only in respect of the fields of discourse in which they typically appear are
cognitive synonyms. For instance, matrimony may be considered a field-
specific synonym most frequently encountered in legal and religious
contexts of one of the senses of marriage (state of being married); wedlock
overlaps with matrimony, but is more likely to be heard in church than in a
court of law.
The second dimension of register, that is, mode, is concerned with
the manner of transmission of a linguistic message – whether it is written,
spoken, telegraphed or emailed. For example, further to is specific of
written language, wheras like is used in the spoken language (e.g. I asked
him, like, where he was going.)
The third dimension of register, that is, style, is a matter of the
formality/informality of an utterance. Style spawns the most spectacular
proliferation of cognitive synonyms, especially in taboo areas such as
death, sex, excretory functions, money, religion, power relations, etc. For
instance, pass away belongs to a higher (more formal) register than die and
kick the bucket and croak belong to a lower register. The synonymic series
of die contains items that may be differentiated in respect of field as well as
style: kick the bucket, buy it, snuff it, cop it, pop off, peg out, expire, perish,
die, pass away, decease, etc.

Conclusions

In this chapter we have identified a number of meanings that were

distinguished in terms of various criteria: (1) language level (e.g. lexical,
grammatical, sentence, utterance meaning) and (2) language function (e.g.
descriptive, social, expressive). One main difference between lexical
meaning and grammatical meaning is that the former is generally more

32 Or tenor in Halliday’s terminology.

46
concrete than the latter. Another essential difference relates to the fact that
the number of grammatical meanings expressed in a language is by far
smaller (and finite) in comparison with the number of potential lexical
meanings. The distinction drawn between different types of meaning can be
useful for translators since one of their most difficult tasks is to perceive the
meanings of words and utterances very precisely in order to render them
into another language.

47
48
 4. SENSE RELATIONS

4.1. Semasiology and onomasiology - two basic approaches to

the
study of words and their senses
4.2. From word to concept: polysemy and homonymy
4.2.1. Polysemy
4.2.2. Homonymy
4.2.3. The distinction between polysemy and homonymy
4.2.4. Polysemy in Cognitive Linguistics
4.2.5. Polysemy vs. vagueness
4.2.6. Polysemy and semantic change
4.3. From Concept to Word: Synonymy and Antonymy
4.3.1. Synonymy
4.3.2. Antonymy
4.4. Hierarchical Sense Relations: Hyponymy and Meronymy
4.4.1. Hyponymy
4.4.2. Meronymy
Conclusions

4.1. Semasiology and onomasiology - two basic approaches to the

study of words and their senses

The terminological pair onomasiology/semasiology is a traditional one

in European lexicology and lexicography. This pair is generally regarded as
identifying two different perspectives for studying the relationship between
words and their semantic values. A brief look at the etymology of the terms
semasiology and onomasiology points to a distinction between meaning and
naming.
Semasiology comes from the Greek sema, "sign" and takes its
starting point in the word as a form and describes what semantic values it
may have. Onomasiology comes from the Greek onoma "name" and
accounts for the opposite direction in the study of meaning, that is, it starts
from a semantic value and investigates by which expressions a particular
concept can be designated.
Actually, semasiology studies the semantic structure of single
expression (e.g. polysemy and homonymy), while onomasiology is
concerned with sets of related concepts (expressed by sense relations
such as synonymy, antonymy, hyponymy). Dirk Geeraerts, Stefan

49
Grondelaers and Peter Bakema (1994) in their book “The Structure of
Lexical Variation. Meaning, Naming and Context” include semasiological
and onomasiological variation among the four main kinds of lexical variation
they identify: semasiological, onomasiological, formal and contextual. The
first two types are placed under the general heading conceptual variation.
Semasiological variation involves the situation that one particular
lexical item may refer to distinct types of referents. Onomasiological
variation involves the situation that a referent or type of referent may be
named by means of various conceptually distinct lexical categories.
While the poststructuralist phase in the history of lexical semantics
had a predominantly semasiological focus (concentrating as it did on the
changes of meaning of individual words), the structuralist stage stressed
the necessity of complementing the semasiological perspective with an
onomasiological one. A number of scholars, Kurt Baldinger among them,
emphasized the importance of a semasiological perspective next to an
onomasiological one. In his 1964 article entitled “Semasiologie et
onomasiologie” he concludes that diachronic semantics should neither be
based exclusively on a semasiological, word-oriented method, nor
exclusively on an onomasiological, structure-oriented method. Later, in his
“Semantic Theory” he stresses the complementarity of the onomasiological
and semasiological perspectives as follows:

"Each linguistic evolution is produced on the one hand within the

framework of a semasiological structure and on the other within the
framework of an onomasiological structure." (Baldinger 1980: 308)

An important idea that should receive more systematic attention

nowadays is a contextualized, pragmatic conception of onomasiology,
based on the view that onomasiology approaches problems form the
viewpoint of the speaker, who has to choose between different names of
expression33.

4.2. From word to concept: polysemy and homonymy

4.2.1. Polysemy

Polysemy is the phenomenon whereby a linguistic unit exhibits

multiple distinct yet related meanings. Being a common feature in any
language, polysemy is justly considered to be a necessary means of
language economy. This idea is stressed by Steven Ullmann in Principles

33 Semasiology, on the other hand, can be viewed as approaching problems from the
viewpoint of the listener, who has to determine the meaning of the words he hears, from all
the possible meanings.

50
of Semantics (1964:118): “polysemy is an indispensable resource of
language economy. It would be altogether impracticable to have separate
terms for every referent”. The topic of polysemy has attracted linguists’s
attention and interest and has posed special problems both in semantic
theory and semantic applications, such as lexicography or translation.
Thus, it has been found that dictionary entries for some words tend to
inflate the number of sense categories beyond those normally distinguished
by speakers. The difficulty people will have in using the dictionary is in
distinguishing major and minor senses because most dictionaries treat all
senses as equally important, which is confusing.
Besides lexicographers, translators may also face some difficulties
when polysemy is used as a source of ambiguity and is explored in various
forms of humour (e.g. jokes, puns). Apart from these cases, polysemy is
seldom a problem for communication among people. In fact, language
users select the appropriate senses of polysemous words effortlessly and
unconsciously because they perceive analogies and make natural
associations using the cognitive tools of metaphor and metonymy.
In the book Categories in Natural Languages: the Study of Nominal
Polysemy in English and Romanian (Neagu 1999) the hypothesis is that
language users find it easier to learn an extended meaning than learn a
meaning that is unrelated to a familiar one. In other words, the
psycholinguistic function of polysemy is to facilitate the acquisition of lexical
categories. Starting from the premise that polysemy is not the result of a
random process, but of systematic meaning extensions based on metaphor
and metonymy I analysed meaning extension in nous belonging to three
different semantic fields: (1) animal nouns (2) deverbal nouns (3) social
status nouns. In what follows we will consider only the first and the second
semantic field.
The main finding relative to the first group is that animal nouns
favour metaphor as a polysemy creating mechanism. For instance, some
nouns denoting animals have metaphorical uses (based on similarity of
appearance) in technical domains: mouse ”pointing device invented by Dong
Eglebart, used in computing), cat „a nautical term denoting the contrivance by
which the anchor is raised out of water to the deck of a ship”, spider „a part of
a machinery, instrument or apparatus having radiating arms or spokes”, crab
„a machine with claws used for hoisting or hauling heavy weghts”.
For the second group, i.e. deverbal nouns (labeled as the least
nouny nous), I demonstrated that they develop metonymic meanings where
a semantic role/participant such as RESULT, INSTRUMENT, AGENT,
LOCATION may stand for another one. For example, the AGENT-for-
INSTRUMENT metonymic pattern can be noticed in a deverbal noun like
reader where S1 denotes “a person who reads, especially one who spends
much time in reading” and S2 designates “a book intended to give students
practice in reading”.

51
4.2.2. Homonymy

Unlike polysemy that is property of single lexemes, homonymy is

the semantic relation that holds of two or more distinct lexemes. It may be
regarded as a proof of the Saussurean prlnciple of conventionality. The
term homonymy is used in semantic analysis to refer to “lexical items which
have the same form but differ in meaning” (Crystal 1992: 166). Homonymy
exists in many languages but in English it is particularly common especially
among mono-syllabic words: e.g. fair, bow, ball, sew, row, plot, match, etc.
As causes or sources of homonymy, Steven Ullmann (1962)
acknowledges the following: (1) divergent sense development and (2)
convergent sound development. The first cause, i.e. semantic divergence,
is illustrated by flower and flour that originally were one word. The second
cause, i.e. phonetic convergence (when two or three words of different
origin coincide in sound), is exemplified by ear34 and case.
John Lyons (1986/1981: 43), in Language Meaning and Context
draws a distinction between two types of homonymy: (1) absolute
homonymy and (2) partial homonymy. Absolute homonyms must satisfy the
following conditions:
(1) their form must be unrelated in meaning
(2) all their forms must be identical
(3) identical forms must be syntactically equivalent.
Examples of absolute homonyms include bank1 „financial institution”, bank2
“sloping side of a river”; sole1 ‘bottom of foot/shoe’, sole2 ‘kind of fish’,
tattoo1 ‘an ink drawing in the skin’, tatoo2 “a military drum signal calling
soldiers from their quarters”.
Partial homonyms fail to satisfy the second and third requirements
for absolute homonymy: found, last, lie. Finally, homonymy can be related
either to the pronunciation of the lexemes (homophony) or to their spelling
(homography). Examples for words that are homophones are tail/tail,
story/storey, cue/queue, threw/through, write/rite, thei/there, whole/hole and
so on. Examples of homographs include bow, row, bear, tear, etc.

4.2.3. The distinction between polysemy and homonymy

In traditional approaches of semantic analysis represented by

Ullmann 1962, Weinreich 1963, 1966, Lehrer 1974, Leech 1987/1974, Lyons
1987/1977, 1995 and Lipka 1990, polysemy is usually discussed in
conjunction with homonymy. If two lexical items have either 1) etymologically
distinct meanings or 2) semantically unrelated meanings, they are regarded
as homonyms. In contrast, if the meanings concerned are related by

34 The ear example is discussed at length in the next subsection.

52
metaphorical or metonymic extension, they are considered to be one single
lexeme with two senses.
Several criteria have been suggested to distinguish polysemy from
homonymy, such as the formal identity or distinctness, etymology and close
semantic relatedness, but none of them seems to be satisfactory.

Etymology and spelling

Ullmann (1962)35 has proposed two criteria for distinguishing

homonymy and polysemy: etymology and spelling36. As has already been
shown in the previous section, he rightly notes that it is impossible to
imagine a language without polysemy. By contrast, a language without
homonymy is not only conceivable; it would in fact be a more efficient
medium.

The maximized homonymy view

Uriel Weinreich, in his “Explorations in Semantic Theory”, more or

less side-steps the issue of the difference between polysemy and
homonymy by treating every sense as a separate word. In fact, this stand
corresponds to the maximized homonymy view or ‘lexicographic approach’
as George Miller (1978)37 calls it. Considering the possibility of metaphor,
Miller shows that there is no limit at all to drawing conceptual distinctions
among the senses of a word. That is why, he says, “sooner or later it will
become necessary to stop drawing distinctions and to start grouping
minimally different senses together.” (Miller 1978: 101). The disadvantages
of the maximized homonymy view are also presented by John Lyons
(1987/1977)38 who argues that, by following this approach, which has been
proposed by certain linguists, we will end up with many more lexical entries
than are recognized in the standard dictionaries of the language described.
And, most importantly, many of these entries will duplicate the phonological
and grammatical information that is attached to other entries.

35 Ullmann, Stephen. 1962. Semantics: An Introduction to the Science of Meaning. Oxford:

Basil Blackwell.
36 These criteria rely on diachronic structure and are not workable for languages that are

unwritten or for which the history is unknown (cf. Adrienne Lehrer.1974. “Homonymy and
Polysemy. Meaning. Similarity of Meaning”. In Language Sciences. 25: 33- 38).
37 Miller, George. 1978. “Semantic Relations among Words”. In George Miller, ed. Linguistic

Theory and Psychological Reality. Cambridge, Massachusetts: The MIT Press, pp. 60-118.
38 Lyons, John. 1988.[1977]. Semantics. 2 volumes, Cambridge: Cambridge University Press.

53
The maximized polysemy view

The opposite alternative, that is, the maximized polysemy view is

also discussed in Lyons (1987/1977: 554) who maintains that, out of the
two radical maximizing views it is preferable to maximize polysemy. This
will have the effect of producing a lexicon with far fewer entries than are to
be found in our standard dictionaries. Finally, John Lyons favours polysemy
because “polysemy - the product of metaphorical creativity is essential to
the functioning of languages as flexible and efficient semiotic systems.
Homonymy, whether complete or partial, is not.” (1987/1977: 567).

Etymology and relatedness of meaning

The criteria that are invoked traditionally in the literature to

distinguish between polysemy and homonymy are etymology and
relatedness of meaning. In terms of the former criterion, that is etymology,
lexical items with the same origin are considered polysemic, whereas if
they have evolved from distinct lexemes in some earlier stage of the
language then they are regarded as homonymous.
This condition is not always relevant and therefore decisive,
because the history of the language does not always reflect its present
state: there are instances of words that come from the same source and
cannot be considered polysemantic, but homonymic. For instance, in
present-day English, the lexemes pupil1 "student" and pupil2, "iris of the
eye" are not semantically related but they both come from Latin pupillus,
pupilla "ward, orphan-boy" which is a diminutive of pupus "child". The
opposite case is also fairly common, namely when two lexemes derived
from different roots in an earlier state of the language are seen as related.
For example, ear1 "organ of hearing" comes form Latin auris 'ear', while
ear2 "spike of corn" is derived form Latin acus, aceris 'husk'. Synchronically,
most people take these two lexemes for one polysemous word and explain
their relation by means of metaphor, i.e. the ear corn was felt to be a
metaphor of the type "the eye of a needle", "the foot of the mountain", etc.
Therefore, the etymological criterion can be misleading when deciding
between homonymy and polysemy.
The latter criterion, that is, relatedness vs. unrelatedness of
meaning is questioned by Lyons (1987/1977) who argues that relatedness
of meaning appears to be a matter of degree, together with the fact that
sometimes native speaker's intuitions are far from being the true
interpretations as has been seen with the ear example. The criterion of
relatedness or similarity of meaning is sometimes associated with another
one, i.e. comparing semantic components. Unfortunately, componential
definitions of the type [physical object], [concrete], [inanimate] for the
description of lexemes such as bank or mouth are not sufficient for the

54
polysemy - homonymy problem. In Katz’ theory39 all features are of equal
value and thus it is not clear how to count differences - whether they should
be ignored or substracted from the similarities. For instance, the two
senses of bank “bank for money” (bank1) and “bank of a river” (bank2) are
classic homonyms, but they share the features [Physical Object],
[Concrete], [Inanimate]. Since these features are generic, high-level
components, one cannot consider them instances of polysemy. Besides,
the two meanings cannot be traced to a common etymology: bank1
“financial institution” comes from Italian ‘banco’ through the French
‘bangue’, while bank2 “slope, elevation in sea or river” is of Scaninavian
origin. By contrast , two meanings of mouth - of a person and of a river -
which seem to be related semantically, also share the feature [Physical
Object], [Concrete] and [Inanimate]. In the case of mouth relatedness of
meaning is based on similarity that is metaphorical.
Therefore sense relatedness should be viewed in terms of conceptual
connections rather than as a matter of shared properties. This is why we next
turn to discussions of polysemy in the context of cognitive linguistics.

4.2.4. Polysemy in cognitive linguistics

While in traditional approaches polysemy is assumed to be a

property of lexical categories only, in cognitive linguistics40 the notion of
polysemy is essentially extended and is applied to both lexical and
grammatical language levels. Linguistic categories that have been shown to
be polysemic include:

 nouns (e.g. head)

 verbs (e.g. climb, turn)41
 prepositions (e.g. over)
 verb particles (e.g. in, out)42
 constructions (e.g. there is)
 the diminutive suffix (e.g. -ling)
 the grammatical case

In cognitive linguistics, polysemy is viewed as a category and is

analysed by means of models such as: prototypes, radial categories and
schematic networks, all of these sharing a concern for the flexibility of
meaning. The prototype model has been used in the discussion of the

39 Katz, Jerrold J. 1972. Semantic Theory. New York: Harper and Row.
40 See, for example, Lakoff 1987, Taylor 2003, Croft & Cruse 2004 and Evans & Green 2006.
41 The polysemy verbs of perception has also been investigated (e.g. Sweetser 1991).
42In Neagu M. 2007. „English Verb Particles and Their Acquisition” (published in Revista

Espanola de Linguistica Aplicada, vol. 20, pp. 121-138) I demonstrate that English verb particles
disclose figurative related meanings derived from a central/prototypical locative meaning.

55
concept lie and the concept bachelor. Coleman and Kay (1981) discuss the
concept lie in terms of (a) falsehood, (b) deliberateness and (c) intent to
deceive. As these three elements may possess different degrees of
importance, there may be prototypical lies, when a statement is
characterized by properties (b) and (c) and partial lies that include
instances of social lie (e.g. 'Drop in any time'), white lie, exaggerations,
joke, etc. A social lie is a case where deceit is helpful and a white lie is a
case where deceit is not harmful.
Fillmore (1982) analyses bachelor that is usually defined as an
unmarried adult man by bringing into discussion less typical examples of
bachelors such as male participants in long-term unmarried coupling, boys
abandoned in the jungle and grown to maturity away from contact with
human society, some priests or homosexuals. Such cases are regarded as
deviations from prototypical bachelorhood, marginal examples of bachelors
that stand for ‘prototype effects’, namely “asymmetries within categories
and gradiations away from a best example”. (Lakoff 1987: 59)
Lakoff (1987) describes the radial set model in the following terms:
(a) polysemic words consist of a number of radially related categories; (b)
the central radial category member provides a cognitive model that
motivates the noncentral senses and (c) the extended senses clustered
around the central category are related by a variety of possible links (e.g.
metaphor, metonymy, etc). Lakoff (1987: 76) analyses the concept mother
and concludes that it cannot be defined "in terms of common necessary
and sufficient condition approach" that can be associated with
Componential Analysis (CA) in structuralist semantics. His argument is the
existence of marginal or less typical cases of mother: biological mothers,
donor mothers (who donate an egg), surrogate mothers (who bear the child
but may not have donated the egg), adoptive mothers, unwed mothers who
give their children up for adoption, and stepmothers.

4.2.5. Polysemy vs vagueness

Another polysemy-related issue that is approached in studies devoted

to polysemy concerns the distinction between polysemy and vagueness, that
is, the question of how different two usages of the same form need to be in
order to count as distinct polysemous senses rather than different
instantiations of a single underlying sense. One of the cognitive linguists who
dealt with the issue is David Tuggy (1993)43 who believes that homonymy,
polysemy and vagueness form a continuum and the place on the continuum
depends on two factors: (i) the presence of a subsuming schema and (ii) the
relative conceptual distance of such a schema from the structures.

43 Tuggy, David. 1993. “Ambiguity, Polysemy and Vagueness”. In Cognitive Linguistics, 3-4,

273-290.

56
 Vagueness involves meanings which are not well entrenched such
as the gender distinction (female/male) in the English words student,
doctor, neighbour, etc., but whose schematic meaning is relatively well
entrenched and elaboratively close. Other examples of vague words are
lexical forms that profile parts of different domains in their respective
semantic base. For instance, the adjective fast in a fast car as opposed to
fast in a fast drink or the noun window understood either as a glass pane or
a wooden frame44 evoke different domains and profile different attributes of
the things they refer to. Such examples as fast or window involve profiling
of parts associated with an object within one conceptual domain, hence the
name ‘‘(partial) segment profiling’’ for vagueness.

4.2.6. Polysemy and semantic change

The notions of polysemy, homonymy and vagueness cannot be

clearly defined and understood without bringing into discussion the issue of
diachronic change. The importance of the relation between polysemy and
semantic change has lately been emphasized by Dirk Geeraerts (2009) in
his valuable book Theories of Lexical Semantics where he compares
cognitive linguistics with diachronic semantics:

“To what extent, in fact, can we say that cognitive semantics is a

return to the fundamental position of historical-philological
semantics? First, cognitive semantics and traditional historical
semantics share, by and large, a psychological conception of
meaning. Second, both approaches start from an encyclopaedist
conception of meaning, in the sense that lexical meaning is not
considered to be an autonomous phenomenon, but is rather
inextricably bound up with the individual, cultural, social, historical
experience of the language user. Third, both are specifically
interested in the flexibility and polysemy of meaning and the
mechanisms underlying those phenomena; in the case of historical-
philological semantics, the perspective is almost exclusively
diachronic, whereas cognitive semantics also considers polysemy and
flexibility from a synchronic point of view”. (Geeraerts 2009: 243-4).

This last idea regarding the close interrelation between synchronic

and diachronic linguistic phenomena has been observed and explained by
Eve Sweetser (1990: 9) who rightly notes:

“synchronic polysemy and historical change of meaning really supply

the same data in many ways. No historical shift of meaning can take

44 See the topic of automeronymy, a variety of polysemy, discussed in section 4.4.

57
 place without an intervening stage of polysemy. If a word once meant
A and now means B, we can be fairly certain that speakers did not
just wake up and switch meanings on June 14, 1066. Rather, there
was a stage when the word meant both A and B, and the earlier
meaning of A eventually was lost.” (Sweetser 1990: 9)

When analysing the semantics of social status nouns (Neagu 1999:

200-243) I start from the premise that polysemy is a synchronic
phenomenon that cannot be understood without referring to the relations
between historically earlier and later senses of a word. For instance, sense3
of knave „an unprincipled man, given to dishonorable and deceitful
practices” developed as a result of the implication/connotation [dishonest
servant], contained by sense 2 „a boy or lad employed as a servant”.

4.3. From Concept to Word: Synonymy and Antonymy

4.3.1. Synonymy

As stated earlier, onomasiology deals with cases in which the same

concept or similar concepts are expressed by different words or
expressions. According to one definition (usually attributed to Leibniz), two
expressions are synonymous if the substitution of one for the other never
changes the truth value of a sentence in which the substitution is made. By
that definition, true or absolute synonyms are rare, if they exist at all.

Absolute and partial synonymy

Absolute synonyms are defined (Lyons 1986: 51) as "expressions

that are fully, totally and completely synonyms" in the sense that

(a) all their meanings are identical (full synonymy)

(b) they are interchangeable in all contexts (total synonymy)
(c) they are identical in all relevant dimensions of meaning (complete
synonymy)
Actually the very terms 'absolute synonymy', ''full synonymy", ”total
synonymy" and "complete synonymy" (not to mention exact synonymy) are
themselves used as synonyms whether absolute or partial in standard
works in semantics or lexicology, usually without definition. Without favoring
the hair-splitting terminological distinctions, Lyons (1986/1981: 51) insists
upon the importance of (a) not confusing near synonymy with partial
synonymy and (b) not making the assumptions that failure to satisfy one of
the conditions of absolute synonymy necessarily involves the failure to
satisfy either or both of the other conditions.

58
 To exemplify the first condition, i.e. same range of meanings,
required by absolute synonymy or full synonymy, we will consider the pair
big - large, where the former term has at least one meaning that it does not
share with the latter one. If we compare the sentence "I will tell my big
sister" with "I will tell my large sister" we notice that the polysemy of big
does not perfectly overlap with the meaning of large.
The second condition for absolute synonymy, i.e. interchangeability
of terms in all contexts (total synonymy) refers to the collocational range of
an expression (the set of contexts in which it can occur). For example, the
members in the pairs busy-occupied, decoration-ornamentation, liberty -
freedom do not always have the same collocational range. There are many
contexts in which they are not interchangeable without violating the
collocational restrictions of the one or of the other. For instance, freedom
cannot be substituted for liberty in 'You are at liberty to say what you want'.
In his approach of cognitive synonymy, i.e. the relation defined in
terms of truth conditional relations, D.A.Cruse defines collocational
restrictions as “co-occurrence restrictions that are irrelevant to truth
conditions – that is to say, those in respect of which lexical items may differ
and still be cognitive synonyms” (Cruse 1986: 279). He assumes that
collocational restrictions are not logically necessary, unlike selectional
restrictions, i.e. semantic co-occurrence restrictions which are logically
necessary45. Examples of cognitive synonyms that carry the same
propositional traits, have the same selection restrictions, but differ in terms
of collocational restrictions are die - pass away, grill - toast, customer -
client. The difference between die and pass away in My grandfather died
yesterday and My grandfather passed away yesterday lies in the greater
semantic cohesion of the latter sentence, i.e. its subject is more predictable
from the rest of the sentence. Cruse (1986: 281) believes that generally,
collocational restrictions behave as presuppositions46 of the selecting item.
Relative to the third condition for absolute synonymy, i.e.
identity/similarity of all dimensions of meaning (complete synonymy), Lyons
(1986/1981: 55) distinguishes descriptive synonymy and expressive
synonymy. Two expressions are descriptively synonymous, i.e. they have
the same descriptive propositional/cognitive/referential meaning, if and only
if statements containing the one necessarily imply otherwise identical
statements containing the other and vice versa. For example, big can be
substituted for large in 'I live in a big house'.
However, in particular instances, synonymous expressions may
differ in terms of the degree or nature of their expressive meaning.
Expressive (affective/attitudinal/emotive) meaning, already discussed in

45 For example, the semantic features “organic”, “alive” and “mortal” are logical pre-requisites of
the meaning of die.
46 The topic of presupposition is approached in section 3.2 of Part II.Pragmatics.

59
Section 3.4 is the kind of meaning by virtue of which a speaker expresses,
rather than describes his beliefs, attitudes and feelings. For example, words
like huge, enormous, gigantic, colossal are more expressive of their users'
feelings towards what they are describing than very big or very large with
which they are perhaps descriptively synonymous (Lyons 1986: 54).
As languages seem to vary considerably in the degree to which they
grammaticalize expressive meaning, to choose the right word /expression
out of a wide range of synonymic terms differing in their degree of
expressivity is a very demanding task for translators. It is the expressive
rather than the descriptive component of meaning that is dominant when we
decide to use terms that imply approval or disapproval: statesman vs.
politician, thrifty vs. mean/stingy vs. economical, stink/stench vs. fragrance
vs. smell, crafty/cunning vs. skillful vs. clever. In order to attract the reader
and listener's attention headline and advertisement writers have to be very
skillful at using expressive synonymy. Knowing the expressive meaning of a
lexeme is just as much a part of one's competence in a language as
knowing its descriptive meaning.
Although synonymy is fairly irrelevant for the structure of the lexicon of a
language, i.e. a language can function without synonymy, language learners
cannot use the language properly without knowledge of all its synonymic
resources.

4.3.2. Antonymy

Although there is no logical necessity for languages to have lexical

opposites at all (English would be just as efficient as semiotic system if
there were such pairs as good - ungood, wide - unwide, far - unfar),
antonymy reflects the human tendency to think in opposites, to categorize
experience in terms of binary contrast (Lyons, 1987/1977 : 276).
Antonyms have received a good deal of attention from linguists
such as Sapir (1944), Duchacek (1965), Bierwisch (1967), Lyons
(1985/1968, 1987/1977), Cruse (1976, 1986), Bolinger (1977), Lehrer
(1982).
Lyons (1987/1977) replaces the term “antonymy” in the wider sense
by "oppositeness" (of meaning) and distinguishes three different types of
oppositeness: a) complementarity b) antonymy (in the narrower, restricted
sense) and c) converseness.

Complementarity

Complementarity can be exemplified by pairs of words) like male

and female, single-married. It is characteristic of complementaries that the
denial of the one term implies the assertion of the other and vice versa. For
instance, John is not married implies that John is single and also John is
married implies that John is not single.

60
 Although complementaries (or binary antonyms) are not gradable
opposites there are instances that do not cover all possible cases in real
life. Thus there may be other possibilities besides complementaries, e.g.
male and female namely hermaphrodite.
Cruse (1986: 202) claims that complementaries are not normally
gradable, that is, they are odd in the comparative or superlative degree or
when modified by intensifiers such as extremely, moderately or slightly (e.g.
extremely true, moderately female, etc). Nevertheless, he states, there are
instances where one member of the pair lends itself more readily to grading
than the other. Thus, alive is more gradable than dead (very dead,
moderately dead, deader than before vs. very alive, moderately alive, more
alive than before). For example, if someone says to us ‘Is X still alive
then?’. And we reply ‘Very much so.’ or ‘And how!’ we are not thereby
challenging the ungradability of dead : alive in the language system. What
we are grading, Lyons (1987/1977: 278) assumes, are various secondary
implications or connotations of alive. Finally, David Alan Cruse (1986)
maintains that

“the relation between dead and alive is not at all affected by medico-
legal uncertainty as what constitutes the point of death. Such
referential indeterminacy afflicts all words, without exceptions. The
point about complementaries is that once a decision has been
reached regarding one term, in all relevant circumstances a decision
has effectively been made regarding the other term, too.” (Cruse
1986: 199)

The idea that complementarity is to some extent a matter of degree

is supported by examples such as ghosts and vampires that existed in a
state which was neither death nor life. Similarly, Cruse (1986) says, the
existence of hermaphrodites and animals of totally indeterminate sex
weakens the relationships between male and female. An even weaker
relationship would hold between terms such left-handed and right-handed.
Complementaries are, generally speaking, either adjectives or
verbs. According to Cruse (1986: 200), an interesting feature of verbal
complementaries which distinguishes them from adjectival
complementaries is that the domain within which complementarity operates
is often expressible by a single lexical item: e.g. the verb command sets the
scene for the complementarity of obey and disobey.
Further examples are born: live - die, start: keep on - stop, learn:
remember - forget, arrive: stay- leave, earn: save- suspend, request: grant -
refuse, invite: accept - turn down, greet: acknowledge - snub, tempt: yield -
resist, try: succeed- fail, compete: win- loose, aim: hit - miss.
A final example of lexical triplets involving verbal complementaries
are attack: defend - submit, change: refute - admit, shoot (in football): save-
let in, punch: parry - take.

61
 As can be noticed, the members of the complementary pair
represent an active and a passive response to the original action or
perhaps more revealing, counteraction or lack of counteraction.

Antonymy proper

Antonymy in the narrow, restricted sense of Lyons (1987/1977) is

the second subclass of oppositeness of meaning. The logical relationship is
based on the fact that the assertion of one member does imply the negation
of the other, but not vice versa. In other words, for pairs of antonyms like
good - bad, big - small, high - low, only one of the relations of implication
(entailment) stated for complementarity holds.
Thus, John is good, implies John is not bad. But John is not good does not
necessarily imply John is bad. Therefore, the negation/ denial of one term
does not necessarily imply the assertion of the other.
In the case of antonymy proper, a third possibility exists. The two
expressions involved in gradable antonymy constitute opposite poles of a
continuum. The great majority of gradable antonyms are pairs of adjectives
that behave like comparatives, i.e. they are fully gradable. Some of these
pairs reveal a certain asymmetry in the sense that one member appears in
more contexts than the other. For instance, if we want to know a person’s
age, or the length of an object we use old (How old are you?) and long
(How long is it?) rather than young and short. The members of the
antonymic pairs old - young, long – short or tall – short differ in
markedness. The term with the wider range of uses is called unmarked
(old, long, tall) and the one with a more limited range marked (young,
short).

Converseness (relational oppositeness)

Converseness is the third subclass of oppositeness of meaning

which describes the same situation from different perspectives or
viewpoints. Converses are expressions which represent the same event or
the same relation from contrasting perspectives. Thus, if we are told that
John owns this book we automatically know that This book belongs to
John.
Lyons (1985/1968) argues that the logical criterion used for the
sense relation of converseness is the possibility of permuting noun phrases
functioning as arguments (semantic roles) in sentences which remain
otherwise equivalent; the sentences imply each other and thus have the
same meaning. For example, John bought the car from Bill implies Bill sold
the car to John and vice versa. Schematically, the sentences may be
represented in the following way:

62
NP1 bought NP3 from NP2.
NP2 sold NP3 to NP1.

As can be noticed, the substitution of lexical converses causes a

permutation of NPs functioning as arguments.
Examples of converses (relational opposites) include pairs of nouns
such as teacher - student, doctor – patient, grandparent - grandchild,
deverbal nouns ending in –ee and –er (employer - employee, examiner –
examinee, interviewer – interviewee), pairs of verbs like give - take, buy –
sell, rent – let, pairs of prepositions like above – below or comparative
forms of adjectives (older - younger, taller – shorter).
The three types of oppositeness of meaning (i.e. complementarity,
antonymy proper and converseness) proposed by Lyons (1985/1968) are
based on the relation of lexical implication or entailment, a notion that will
be defined below, in section 4.4.1.

Directional oppositeness (reverseness)

This fourth subclass of oppositeness of meaning discussed in more

recent works of semantics (Saeed 2000) is based on the notion of contrary
motion (i.e. motion in opposite direction): up - down, in - out, right – left come
- go, arrive - depart, ascend – descend, rise- fall, pull – push, marriage –
divorce.

4.4. Hierarchical Sense Relations: Hyponymy and Meronymy

4.4.1. Hyponymy

Hyponymy, like incompatibility and antonymy has been one of the

topics of lively interest for lexical semantics since the structuralist period.
Although Lyons (1985/1968) declared that all sense relations were context
dependent, they have almost universally been treated (by Lyons himself) as
stable properties of individual lexical items.
Traditionally, sense relations are defined in terms of entailment, i.e.
of the logical relation between two sentences, such that the truth of the
second sentence follows from the truth of the first. On this approach, a
sentence like It’s a dog unilaterally entails It’s an animal so dog is a
hyponym of animal. Similarly, I always avoid the red skirts unilaterally
entails I always avoid the scarlet skirts and John punched Bill unilaterally
entails John hit Bill. As can be noticed, the normal direction in the
entailment is from hyponym to superordinate.
Hyponymy is one of the most fundamental paradigmatic relations,
corresponding to the inclusion of one class in another. For example, terms

63
 Living things

Animal Vegetable

Animal
Bird Fish Insect

Animal Human

such as daisy, daffodil and rose all contain the meaning of flower. That is to
say, they are all hyponyms of flower.
The set of terms which are hyponyms of the same superordinate
term are co-hyponyms; for example, red, black and yellow, in the colour
system, or ox, bull, calf that are covered by the superordinate term cattle.
Another way of describing the relationship is to say that the individual
colours are sisters of the parent term colour or sisters of the parent term
cattle.
A hyponym is a word that is more specific (less general), which has
more elements of meaning and is more marked than its superordinate. For
example, it can be marked for age (puppy, kitten, calf, piglet, duckling and
cygnet are marked, while dog, cat, cow, pig, duck, swan are unmarked) or
for sex (bitch, drake, bull, hog, sow, cob, are marked, while dog, duck, cow,
pig, swan are unmarked). Hence, we can define hyponyms in terms of the
hypernym plus a single feature, as in stallion = ’male horse’, kitten = ’young cat’.
The more general term with reference to which the subordinate term
can be defined, as is the usual practice in dictionary definitions (‘a cat is a
type of animal…’) is called the superordinate or hypernym. Sometimes a
word may be superordinate to itself in another sense. This is the case with
animal, as shown in the figure below. The first occurrence, opposed to
vegetable, is the sense contained in the phrase ‘the animal kingdom’. The
second occurrence is synonymous with mammal, and the third with beast.
Superordinate terms in turn may become hyponyms in relation to a
more general superordinate term: e.g. cattle is a hyponym of animal. Pairs
of lexical items related by hyponymy are far more frequently found among

64
nouns than among adjectives or verbs. Hyponymy is a vertical relationship
which is fundamental to the way in which we classify things. Most
dictionaries rely on it for the provision of definitions where the superordinate
or hypernym corresponds to „genus proximum” and the specific properties
are described in „differentia specifica”. For example, flower is the hypernym
which appears in the definition of daisy „a flower which is very common,
small and white with a yellow centre”.
Hyponymy offers a good organizing principle for vocabulary learning
and teaching. Most language coursebooks use this feature of organization
implicitly or explicitly in grouping names of flowers together or garnments or
articles of furniture.

Autohyponymy

Autohyponymy is a variety of polysemy (Cruse 2004: 108) and

occurs when a word has a default general sense and a contextually
restricted sense which is more specific in that it denotes a subvariety of a
general sense. For example, dog has two senses, a general sense,
‘member of the canine race’ as in Dog and cat owners must register their
pets and a more specific meaning as in That’s not a dog, it’s a bitch.
What is interesting to note is that in the lexicalization of a distinction
of sex, for some species it is the lexeme denoting males (e.g. drake), and
for other species the lexeme denoting females (e.g. sow) that is
semantically marked. An instance of generalization of a feminine term is the
use of cow as in those cows over there or a field full of cows to refer to
bovines of both sexes, especially when there is a mixed group.

4.4.2. Meronymy

Meronymy is a term used to describe a part-whole relationship

between lexical items. For instance, cover and page are meronyms of book
(the holonym). We can identify this relationship by using sentence frames
like X is part of Y, or Y has X, as in a page is a part of a book or a book has
pages.
The lexical relation of meronymy, sometimes referred to as
„partonymy”, is usually informally described as ‘part-whole relation’. Croft
and Cruse (2004: 151) claim that meronymy is a relation between
meanings, whereas the part-whole relation links two individual entities and
generates chains of elements: A is a part of B, B is a part of C, C is a part
of D and so on.
For instance,
A fingertip is a part of a finger.
A finger is a part of hand.
A hand is a part of arm.

65
An arm is a part of a body.
An important point is that the networks identified as meronymy are
lexical: it is conceptually possible to segment an item in countless ways, but
only some divisions are coined in the vocabulary of a language. Every
language has a range of ways of referring to parts of things. Many of these
ways involve specialized lexical items.
Meronymy is similar to hyponymy because it reflects a hierarchical
and asymmetrical relationship between words, represented by the ‘less
than’ sign. For example, stanza is a meronym of poem, but poem is not a
meronym of stanza. Or, sonnet is a hyponym of poem but poem is not a
hyponym of sonnet. However, unlike hyponymic relations, meronymic
hierarchies are less clear cut and regular. Meronyms may vary in how
necessary the part is to the whole. Some are for normal examples, for
example, nose is a meronym of face, others are usual but not obligatory, like
collar, as a meronym of shirt, still other are optional, like cellar for house.
Meronymy also differs from hyponymy in transitivity, a relational
property that can be described like this: if a relation R is transitive, then the
truth of aRb and bRc guaranteses the truth of aRc. Hyponymy is always
transitive, but meronymy may or may not be. A transitive example is nail, a
meronym of finger and finger of hand. We can see that nail is a meronym of
hand as we can say A hand has nails. A non-transitive example is: pane is
a meronym of window (A window has a pane) and window of room (A room
has a window); but pane is not a meronym of room, for we cannot say A
room has a pane. Or hole is a meronym of button and button of shirt, but
we wouldn’t say that hole is a meronym of shirt (A shirt has holes).
Meronymy and hyponymy involve completely different types of
hierarchies. While meronymy relates to individual referents of meronymic
terms, hyponymy involves a relation of inclusion between classes: the
extension of the hyponym is included in that of the hypernym.

Automeronymy

Cruse (2004: 104) argues that automeronymy, like autohyponymy,

is a variety of polysemy. While in the case of autohyponymy the more
specific reading denotes a subtype, in the case of automeronymy the more
specific reading denotes a subpart. For instance, door can refer to either
the whole set-up with jambs, lintel, threshold, hinges and the leaf panel as
in Go through that door or just to the leaf, as in Take the door off its hinges.
Further, a sentence such as We took the door off its hinges and walked
through it illustrates zeugma, a type of semantic anomaly which appears
when a single occurence of an expression has to be interpreted in two
distinct ways simultaneously.

66
Hyponymic and meronymic enrichment

The effects of context on the meaning of a word can be seen in

what Cruse (2004: 119) calls ‘contextual modulation’ that can manifest itself
in two forms or varieties: enrichment, i.e. the addition of semantic content to
the meaning of a word, and impoverishment, i.e. the removal of semantic
content from the meaning of a word.
Hyponymic enrichment arises when the context adds features of
meaning to a word which are not made explicit by the lexical item itself:

Our maths teacher is on maternity leave (gender is determined)

My brother always bumps his head when he goes through the door
(height is determined)
My coffee burnt my tongue. (temperature is determined)
Our house was burgled while we were away. They only took the
video, though (legality is determined)

Sometimes the context points to a specific kind of the class normally

denoted by the lexical item employed, rather than adding a feature, like in I
wish that animal would stop barkink/miaowing or John is going well in the
1500–metres freestyle.
Meronymic enrichment arises when someone specifies the part of
what the lexical item used normally refers to. This part may be definite and
identifiable (e.g. a tyre as in A car has a puncture) or less definite (e.g. a car’s
damaged area as in The car was damaged when John drove it into a tree).
Cruse (2004: 120) argues that this kind of narrowing down to a part,
that is, meronymic enrichment, is widespread in language use and
speakers are not usually aware of this. For instance, a red book has red
covers, not red letters, whereas a red warning sign most likely has red
letters. Further examples include noun phrases made up of a colour
adjective and a head noun; very often the colour does not apply globally to
the object denoted by the head noun but only to a part: a red apple (a
significant portion of outer skin is red), a yellow peach (inner flesh is yellow),
a pink grapefruit (inner flesh is pink), red yes (white of eyes is red), blue eyes
(iris is blue). In all these examples the colour adjective indicates that the
referent of the head noun is distinctive by virtue of its possession of an area
with certain perceptual properties.

Conclusions

This chapter contains a discussion of two opposite ways of studying

meaning: one which proceeds from a given form and asks for its meanings
(semasiology) and the other which starts out from a given meaning and

67
asks for the kinds of forms that are used to express this meaning
(onomasiology). Regarding the paradigmatic semantic relations studied by
each of these two branches of semantics, we have seen that semasiology
studies polysemy and homonymy, while onomasiology is concerned with
synonymy and antonymy. The last paradigmatic semantic relations
approached in the chapter, hyponymy and meronymy, involve hierarchies
in the vocabulary, i.e. super- and subordination. Hyponymy, “a type of”
relation, differs from meronymy “a part of“ relation, mainly because the
former is a relation of inclusion between classes, while the latter relates to
individual referents.

68
 5. SEMANTIC ORGANIZATION

5.1. The lexicon

5.1.1 Views of the lexicon
5.1.2 Lexical vs. conceptual knowledge
5.1.3 Lexical item, lexical unit and lexical entry
5.2. Semantic fields
5.2.1. Field theories
5.2.2. Lexical gaps
5.2.3. Conceptual field, lexical field and semantic field
Conclusions

5.1. The Lexicon

5.1.1. Views of the lexicon

In linguistic terminology, the lexicon of a language is a collection of

all the words, more precisely, lexemes, of that particular language.
Lehrer (1974:190) maintains that the lexicon is an unordered set of
lexical entries and as such it can be arranged in a number of ways –
alphabetically, as a dictionary, by semantic fields, as a thesaurus.
According to Lyons (1987/1977: 516), the information that is found in a
typical lexical entry in a conventional dictionary is of three kinds:
morphological, syntactic and semantic.
In generative grammar, the lexicon has a special status and it refers
to the component containing all the information about the structural
properties of the lexical items in a language. Thus, a lexical entry includes
phonological, semantic and syntactic information. Certain syntactic theories
ascribe a more significant role to the lexicon, some claiming that much of
the syntax is projected from the lexicon (e.g. Chomsky 198147). In other
words, the semantic organization of the lexicon can predict and explain at
least some regularities.
The sum of all the lexemes an individual speaker has in his mind is
often referred to as the mental lexicon. Jean Aitchinson (1994), in Words in

47 Chomsky, Noam.1981. Lectures on government and binding. (Studies in Generative

Grammar.9) Dordrecht: Foris Publ.

69
the Mind. An Introduction to the Mental Lexicon, points to the difference
between an ordinary dictionary and the mental lexicon:

”Unlike ordinary dictionaries which are limited in scope, have an

oversimple organization, a fixed and outdated content and contain
only a relatively small amount of information about each item, the
mental lexicon has a large and complex content organized as a
structured network of intersecting defined classes and governed by a
number of rules”. (Aitchinson 1994:15)

The idea that the mental lexicon has an internally structured nature
and contains a number of rules for creating new lexical items or for extending
the meaning of given lexical items is also present in Dirven (1985)48. New
lexical items are formed by the rules of compounding, derivation, borrowing,
the creation of neologisms, acronyms. The meaning of given lexical items can
be extended by processes such as metaphor and metonymy.
Starting with the 1990s there has been a surge of interest in the
lexicon. The demand for a fuller and more adequate understanding of
lexical meaning required by developments in computational linguistics,
artificial intelligence and cognitive science has stimulated a refocused
interest in linguistics, psychology and philosophy.

5.1.2. Lexical vs. conceptual knowledge

The basic problem that distinguishes the different views of the lexicon
relates to the nature of the information in the lexicon. Murphy49 (2003) argues
that knowledge about words (i.e. lexical knowledge) does not always overlap
with knowledge about the things words denote (conceptual knowledge). The
lexicon contains information that is necessary for linguistic competence, i.e.
our capacity to produce grammatical and interpretable sentences.
The fact that we can fail to make the association between things
that we recognize and words that we know for those things indicates that
our means of storing and/or accessing the name of that thing is the same
as our means of storing and/or accessing other knowledge about the thing.
The piece of evidence for this is tip-of-the tongue syndrome, i.e. the case
when we have complete access to the concept, because we can picture it,
reason about it and describe it, but we are not able to access its name.
Other evidence for the separation of lexical and conceptual information is
related to the lack of the one-to-one relationship between words and

48 Dirven, René (1985): “Metaphor as a basic means for extending the lexicon”, in: Wolf

Parotté & René Dirven.eds. The Ubiquity of Metaphor (Current Issues in Linguistic Theory
29). Amsterdam: Benjamins, 85-119.
49 Murphy, M Lynne (2003) Semantic relations and the lexicon: antonymy, synonymy, and

other paradigms. Cambridge: Cambridge University Press.

70
concepts proved by the existence of polysemy and synonymy in language.
Words can be used to indicate more than a single concept, and the name
that we attach to a thing may vary by context. To use the examples given
by Murphy (2003:14), in the first case, the word knife can refer to things like
scalpels, daggers, butter knives and letter openers; in the second, a single
kind of furniture may be referred to by a variety of terms like table,
bedstand, and chest of drawers.
Although they are two distinct types of knowledge, lexical
knowledge and conceptual knowledge interact in the processes of
language production and comprehension.

5.1.3. Lexical item, lexical unit and lexical entry

The lexicon contains both linguistic expressions that are greater

than words and ones that are smaller than words. Phrasal expressions like
throw up or paint that town red and morphemes such as – ness and pre –
are also to be included in the definition of lexical item or lexeme (Murphy
2003: 14). A lexical item in the lexicon is an abstract representation that is
instantiated as a lexical unit in language use, which has a particular form
and a particular sense. For example, highest in the phrase the highest note
in the song and high in I threw the ball high are both lexical units
instantiating the lexical item high. The term lexical entry denotes the
collection of information (phonological, morphological and semantic) about
a lexeme that is included in the lexicon.
Most linguists agree that the lexicon is the repository of what is
exceptional and idiosyncratic in language (the part that has to be learned),
while grammar expresses the regularities of a language. Psychologically,
the lexicon is a more tangible entity than grammar because speakers are
aware that they know and use words, but they are hardly aware that they
know and use rules of the grammar (Cornilescu 1995: 95).

5.2. Semantic fields

Unlike diachronic linguistics in which a language was seen primarily

as a collection of individual elements such as words, syntactic patterns,
etc., structuralist linguistics views language as a system of relations
between interdependent elements. Semantic field theory applies
structuralist ideas to the study of the lexicon of languages.

5.2.1. Field theories

Trier’s view of lexical fields

Semantic field theory derives very largely from the work of German
and Swiss scholars in the 1920s and 1930s. Among the German linguists,

71
Jost Trier was the most important and influential50. He postulated that no
item in the vocabulary can be analyzed semantically unless one takes into
account the relationships and oppositions it enters with the other words in a
given subsystem or system. Trier advanced the idea that vocabulary as a
whole forms an integrated system of lexemes interrelated in sense, a huge
mosaic51 with no loopholes52.
Semantic fields with a more restricted number of terms are
incorporated into larger ones, the latter are themselves structured into even
larger ones, until the entire lexicon of a language is integrated into a unitary
system. In Trier's opinion, therefore, semantic fields act as intermediaries
between individual lexical entries, as they appear in a dictionary, and the
vocabulary as a whole.
According to field theory, meanings of words cluster together to form
fields of meaning, which in turn cluster into even larger fields until the entire
language is encompassed. So, for example, we can identify a semantic
field of madness containing words like insane, demented, batty,
schizophrenic, paranoid, some of which are synonyms of mad, and others
which are types of madness. This field belongs in turn within a larger one of
mental states, which includes a wider selection of words. Similarly we can
identify a field of running including words such as sprinting, running and
jogging, which itself clusters into the field of human motion and so on.
One of the procedures followed by Trier was to compare the
structure of a lexical field at time t1 with the structure of a lexical field at
time t2. He pointed out that the slightest change in the meaning of a term in
a semantic field brings about changes in the neighbouring terms as well.
Therefore, a word acquires its meaning by its opposition to its neighbouring
words in the pattern.
Although Trier opened a new phase in the history of semantics
(Ullmann 1962: 7) he has been criticised for a number of assumptions that
are highly controversial. First he has been challenged for assuming that
lexical fields are closed, well-defined sets. The disagreement is founded
especially if one considers peripheral items in a field53. For example, in the
semantic field of cooking verbs, we have bake, boil, fry, but scald,

50 Jost Trier’s most significant contribution is his 1931 monograph Der Deutsche Wortschatz
im Sinnbezirk des Verstandes. Die Geschichte eines sprachlichen Feldes.
51 Here’s is Dirk Geeraert’s comment on Trier’s idea: „ His use of the mosaic image was not

a happy one. To begin with, the image suggests that the mosaic covers the whole surface of
the field, i.e. that there are no gaps in the lexical field, that no pieces are lacking in the
mosaic. (Geeraerts 2009: 66)
52 Trier distinguished between lexical and conceptual fields, whereby the lexical field divides

the conceptual field into parts, like a mosaic.

53 Relative to clearcut boundaries between fields, Geerarets (2009) argues that „it is often

difficult to indicate exactly where a field ends; discreteness will usually only be found in the
core of a field, whereas there is a peripheral transition zone around the core where field
membership is less clearly defined. (Geerarets 2009: 67)

72
caramelize (e.g. caramelize fruits), render (e.g. render fat) and clarify (e.g.
clarified butter) are peripheral. Second, he has been criticized for
maintaining that there are no gaps or overlaps in a lexical field (we will turn
to the issue of lexical gaps in section 5.2.2). The third and last objection
regards his concentration upon paradigmatic relations of sense to the
exclusion of sintagmatic relations.
Nevertheless, to a greater or lesser extent, Trier’s original ideas
certainly contributed to the development of the subsequent semantic field
theory.

Lehrer’s view of semantic fields

Lehrer (1974) believes that the study of linguistic field should prove
to be a rich source about human conceptualization and that the correct or
at least the best semantic analysis is one that describes a speaker’s
conceptual structure. In Semantic Fields and Lexical Structure, Adrienne
Lehrer (1974: 8) defines a semantic field as a group of words closely
related in meaning, often subsumed by a general term.
For instance, the words in the field of colour in English fall under the
general term COLOUR and include red, blue, green, white, scarlet and
dozens other. In their study of colour terms (1970) Brent Berlin and Paul
Kay54 found that speakers disagree among themselves as to where to draw
the line between colours, e.g. red and orange. Moreover, the judgments of
a single speaker differ at various times. The solution the two American
scholars have proposed is that of focal points for colours, e.g. the most
typical red or the best example of yellow. The prototype – based model has
to be more useful for the analysis of semantic fields because it allows for
fuzzy borders among lexical items. The study by Berlin and Kay also shows
that there are some parts of the colour spectrum that are not happily
covered by any term or at least by any basic term. Lehrer (1974) rightly
states that a very interesting question to investigate is what speakers do
when they want to express some concept not covered by any lexical item in
the language. Her analysis of cooking verbs (1974: 100) reveals lexical
gaps in the field: some of the systematically present conceptual possibilities
are simply left unfilled: for instance, there is no word for the preparation of
food in a pan without water and oil, nor for cooking with oil on a flame.

5.2.2. Lexical gaps

The absence of a lexeme at a particular place in the structure of a

lexical field is generally referred to as a lexical gap. For instance, in English

54 Colour categories have been investigated by Brent Berlin and Paul Key (1969), two American

cognitive anthropologists who contributed to the development of prototype theory.

73
there is a word corpse meaning roughly ‘body of a dead human being’ and
a word carcass meaning ‘body of a dead animal’, but no word which is
applied to dead plants. In general, conceptual fields are heavily lexicalized.
When part of a field is unlexicalized, it constitutes a lexical gap. For
instance, it can be argued that there is a gap for a term superordinate to
aunt and uncle and another for niece and nephew. Another instance of a
lexical gap, occurring when the coverage of the conceptual field by the
lexical field is not complete is the absence of a cover term for bull and cow
(for stallion and mare such a cover term exists: the hypernym horse).
A natural consequence of field theory is the idea that words, or
more particularly the senses of words, define themselves against each
other. So, for example, in the field of medical personnel, part of our
understanding of doctor is ‘not nurse/surgeon/matron or orderly’.
Therefore, the meanings of words must be understood, in part, in
relation to other words that articulate a given content domain. The goal of
the analysis of semantic fields is to collect all the words that belong to a
field and show the relationship of each of them to one another and to the
general term.

5.2.3. Conceptual field, lexical field and semantic field

For John Lyons, (1987/1977: 253) a conceptual field is a structured

conceptual area while a lexical field is a set of lexical items that covers a
specific conceptual field. In his definition of “lexical field”, Geeraerts (2009)
also includes the idea of semantic relatedness and mutual interdependence
of the items:

“A lexical field is a set of semantically related items whose meanings

are mutually interdependent and that together provide conceptual
structure for a certain domain of reality”. (Geeraerts 2009: 56)

The distinction Lyons makes between lexical field and semantic field
is based on the absence or presence of other linguistic units (besides
words) in the field, i.e. whether the set of expressions that covers a
conceptual field consists only of words or also contains other units, such as
idiomatic expressions. For instance, if the field of anger terms includes
expressions like to look daggers or to boil over besides rage, fume, seethe,
etc. the field could be called semantic rather than lexical.
Basic to field theory is the view that words occupy a certain amount
of semantic space within the language, which is distributed among the
specific lexical items available. So, for example, the field of residences is
divided up into castle, maisonette, home, bungalow and flat, to name just a
few. These terms constitute the lexical set, or lexical field which realises the
semantic field. The meaning of any one of them is affected by the other
terms to which it is related. As a consequence, fields are constantly

74
expanding and contracting. If the term maisonette were removed from the
set, then one of the others, possibly house, or flat, would expand to occupy
the space.
Field theory is very useful in the contrastive analysis of different
languages. Languages differ quite widely even in apparently basic lexical
divisions, and fields such as temperature, kinship, colour, parts of the body,
and animal and vegetable worlds, divide the semantic space differently with
respect to them. For instance, some languages like English use eleven
colour terms which name the following colour categories: BLACK, WHITE,
RED, YELLOW, BLUE, BROWN, PURPLE, PINK, ORANGE, and GREY.
Other languages use only two basic colour terms (black and white), three
basic colour terms (black, white and red), etc. Actually, when there are
fewer than eleven basic colour terms55 in a language, one basic term
names a union of basic colour categories; for example, BLUE + GREEN.
According to the cognitive linguistics view the words of a language
reflect conceptual distinctions made by a particular culture. Dirven and
Radden (1997:4) illustrate how the Anglo-Saxon culture and the German
culture carve up the conceptual continuum atmospheric conditions for
which the German culture provides two categories:

Anglo-Saxon culture fog mist

haze
German culture Nebel Dunst

As a result, speakers place their experience of visibility and air moisture

under one of the categories provided by their culture.
The cognitive approach claims that meanings do not exist
independently of human perception and cognition but are created by the
way in which humans experience and think of the phenomena that
surround them. The cognitive view could account for the flexibility of word
meaning and explain why definitions of words are often too difficult to make
precise. It concentrates on how language is shaped by human experience
and cognitive processes. Cognitive linguists argue that categories are
conceptual in nature and that many, if not all of our conceptual categories
are laid down in language as linguistic categories.

55 For a colour term to be basic it must meet the following requirements (Lakoff 1987: 25):
- it must consist of one morpheme, like blue, rather than one, as in dark-blue.
- the colour denoted by the term must not be contained in another colour. Scarlet, is,
for example, contained within red.
- it must not be restricted to a small number of objects; for example blond.
- it must be common and generally known, like yellow as opposed to saffron.

75
An illustration: the semantic field of cooking terms

Lehrer (1974) illustrates the theory of semantic fields with words

from two lexical fields: cooking and sounds. One of her arguments for this
choice is that the sets seem to contain many of the subtleties, asymmetries
and indeterminacies which are characteristic of other lexical fields.
The basic words in the field of COOKING are cook, bake, boil,
roast, fry and broil (or grill for British English) and for some speakers,
steam. Grill and toast denote the same action or process from the point of
view of the agent, but different patients are involved. Grilling is a method of
cooking, whereas toasting is not; things that get toasted are normally
already cooked, whereas items for grilling are raw. The set also includes
simmer, stew, poach, braise, sauté, French fry, deep fry, barbeque and
charcoal. The most general are cook and bake; words such as deep-fry
sauté, parboil, plank, shirr, scallop, flamber, rissoler or compounds like
steam-bake, pot-roast, oven-poach, pan-broil, pan-fry and oven-fry are
considered peripheral.
The first three basic cooking terms, i.e. cook, bake and boil have
both general and specific senses (they represent instances of
autohyponymy, the semantic relation discussed in 4.4). It is interesting to
note that only basic words show this characteristic. Cooking words can be
placed in a chart like in the figure below:

cook1
bake1
cook2
steam boil1 fry broil
roast bake2
simmer boil2 sauté deep-fry grill grill
barbeque
French-fry
charcoal
poach stew braise

As can be noticed, words are synonyms if they appear in the same

square and hyponyms appear directly under the superordinate term. Thus,
steam, boil fry, broil, roast and bake2 are hyponyms of cook2. French fry
and deep fry are synonyms, etc.; cook1 and bake1 differ from the rest in that
they refer to human activities – in one case the preparation of food for
meals and in the other the preparation of a number of items commonly
called bakery products – bread, pastry, cookies, etc. Only cook1 and bake1
freely occur intransitively with human subjects. I cook and He bakes are
more acceptable than *John simmered yesterday or Helen is frying.

76
 Cook2 and all the words under it are process words which can be
analysed grammatically as causatives. Boil1 and its subordinates differ from
others in the semantic field in that water or some water-based liquid must
be used (wine, stock milk) while the absence of water is necessary for fry,
broil, roast, and bake. Simmer differs from boil2 by specifying that the liquid
is just below the boiling point, without the rolling bubbles that characterise
boil2. The hyponyms of simmer bring in highly specific aspects of meaning.
Poach specifies that the food is slowly cooked in water carefully so that the
shape is preserved. Stew is applied when the food is to be cooked slowly
for a long time usually until it is soft. Braise is even more complex – the
food is first browned (quickly fried on the outside) and then allowed to cook
slowly in a tightly covered pot with a small amount of water.
In general, the more specific the meaning of the word, the fewer
collocational possibilities there are: boiled meat, boiled eggs, boiled
vegetables are linguistically acceptable, but poached vegetables and
stewed eggs are less so (Lehrer 1974: 33). Steam and boil are closer in
meaning than to any other basic term. Steam contrasts with boil in that the
food, which must be a solid, is not submerged; it is cooked by the rising
vapours. Fry and its hyponyms contrast with other in the field by requiring
the presence of fat or oil in the cooking process although the fat can be in
the food itself. Like bacon. Deep-fry and its synonym French-fry require a
large amount of oil or fat – enough to cover the item being cooked. Sauté,
on the other hand, refers to quickly cooking something in a frying pan with a
small amount of fat. Fry is used when food is cooked in a frying pan
whether or not fat is added (in the latter case there is some fat in the food
cooked, e.g. steak, or a non-stick frying pan is used). Broil and its
hyponyms refer to cooking something directly under a heating unit or over
or under an open fire. Grill has a range of meaning that overlaps with fry
slightly, since grilled cheese sandwiches are fried, not broiled. Grill also
applies to cooking food on an open grill, but sometimes it is used
synonymously with broil. Barbeque, in one of its senses is synonymous to
charcoal, and both refer to cooking food over hot coals. Bake2 is applied to
cooking food in an oven, such that the heat is indirect, rather than direct as
in broiling. Roast and broil are close in meaning.
The semantic field of cooking verbs can finally be set up to look like a
series of +/- features as in the table below, where 0 means that the feature
does not apply distinctly one way or the other. For example, frying as a kind
of cooking that involves the use of fat in contact with a flame and is not
usually gentle.

77
 water fat oven flame
gentle
Cook 0 0 0 0
0
Boil + - - +
-
Simmer + - - +
+
Fry - + - +
0
Roast - - + -
0
Toast - - - +
0
Bake - - + -
0

Metaphorical extension

Most of the terms in the field of COOKING may have metaphorical

extensions in other semantic fields. They may be used for states of
emotions (boil, burn, simmer, steam, stew) or temperature (bake, steam,
roast).
Sentences like She was boiling with rage and He was burning with
excitement prove that some verbs of cooking develop metaphorical
meanings in the field of EMOTIONS. Other verbs of cooking may have
metaphorical extensions in the semantic field of TEMPERATURE: Our
apartment on the top floor bakes in the summer or It’ s steaming/roasting in
this room.

Conclusions

In this chapter we have seen that the mental lexicon has an

internally structured nature and contains a number of rules for creating new
lexical items or for extending the meaning of given lexical items. In
structuralist theories of semantic fields the meaning of lexical units is
specified in terms of lexical relations (hyponymy or antonymy) to other units
constituting the same field. Fundamental to field theory is the assumption
that words can belong to more than one field. This is possible due to the
polysemy creating devices of metaphor and metonymy.

78
 PART II. PRAGMATICS

1. THE DOMAIN OF PRAGMATICS

1.1. Why do we need pragmatics? Correctness vs. acceptability

1. 2. The domain of pragmatics
1. 3. Connections to the other linguistic branches
1. 3.1. Connections form - pragmatic meaning
1.3.2.Connections sense - pragmatic meaning
1.4. Subdomains of pragmatics. Some of their basic terms and
concepts
Conclusions

1.1. Why do we need pragmatics? Correctness vs. acceptability

‘Pragmatics is the waste-basket of linguistics;’ that is: ‘whenever you

cannot explain a phenomenon in language using regular, accepted
linguistic theories, then you must have recourse to something else,
something that is supposedly as undefined as it is tangible, namely,
pragmatics.’ (Mey 1993: 5)

Starting from Mey’s half-ironical statement, it is obvious that

linguists were at a deadlock when trying to interpret some linguistic realities
strictly within the frame offered by the ‘regular, accepted linguistic theories’,
whether they pertained to phonetics or morphology, syntax or semantics. It
means that a new frame had to be created in order to adequately interpret
those realities. Therefore, what looked rather vague, in spite of being
‘tangible’, i.e. very concrete and present in our daily life but with hardly
clear-cut borders, turned out to be a necessity.
The analysis of the following example would show that:

‘I brought some sushi home and cooked it; it wasn’t bad.’ (Chicago
alternative cultural weekly reader, 21 August, 1992, in Mey (1993: 4))

The information necessary in order to interpret the utterance

adequately is that it is an ad to a cocktail lounge. Is this relevant? Is the text
effective? Actually, it is essential since otherwise the whole reply would
seem just an example of stupidity or ignorance in terms of general
knowledge, on the part of the speaker. The message sent would be
unfavourable to the speaker. Of course, a person can do what he/she

79
wants as long as that does not affect negatively other people, but the type
of behaviour mentioned above is totally illogical according to the laws of
reason: sushi means raw fish possibly combined with other ingredients and
foods, but definitely ‘not cooked’. Cooked sushi is a contradiction in terms
and cooking sushi would seem odd if not a sign of some mental disorder.
All these considerations remain true but totally inadequate when
someone says such a reply after visiting that cocktail lounge. The text
makes implicit reference to a cocktail lounge and such an establishment
should advertise its alcoholic drinks. Doing that directly would not have any
effect on a reader: Our drinks are the best! Try them! At best, the reply
would be No kidding! Do you take us for fools? You would say anything to
sell your stuff!
So, the message of the text is much more effective if conveyed
indirectly: [After having a drink at that cocktail lounge, I couldn’t stop any more,
that good the drink was. I got so drunk that] I brought some sushi home and
cooked it; it wasn’t bad (= I have enjoyed it). Not everybody would enjoy such
an ad, but its effectiveness cannot be denied, since people wouldn’t go to a
place where food and drinks are bad, just for the atmosphere.

‘I just met the old Irishman and his son, coming out of the toilet.’ (David
Lodge, Paradise Lost, in Mey 1993: 7)

This first part of a dialogue can be the beginning of some type of

joke, since it is based on the structural ambiguity of the complex sentence,
resulted from the fact that it is elliptical. One interpretation is I just met the
old Irishman and his son when I was coming out of the toilet, in which case
the time subordinate can be reduced to a participial clause that is better
placed at the beginning of the sentence, to avoid ambiguities: Coming out
of the toilet, I just met …. It is obvious that the subjects of the two clauses
are I, in other words, they are co-referential and therefore, the subject of
the first can be omitted.
Another interpretation is I just met the old Irishman and his son
when they were coming out of the toilet. By reducing the finite clause to a
participial one, we obtain the initial structure quoted by Mey.
Our considerations might lead to the conclusion that the word order
chosen prevents any ambiguities, but a reply of the type When were you at
the toilet? shows that the interlocutor might not always be competent or
attentive enough to interpret the sentence correctly simply on the grounds
of its syntactic pattern, especially since we are dealing with the deep
structure of that sentence. Syntactic ambiguity was commented on by
Chomsky in relation to the different deep structures having the same
surface structure. In our case, the surface structure corresponds to the
elliptical sentence quoted and the deep structure corresponds to the two
possible interpretations.

80
 The identification of the right deep structure is a process which
involves the context of utterance. If the interlocutor knows that the speaker
went to the toilet, he might disambiguate the sentence more easily.
Ambiguities imply a cognitive cost, i.e. the interlocutor makes an effort to
understand the meaning of an utterance.
Following Kroeger (2005) this means that ‘the form of an utterance by
itself (ignoring context) does not determine its function’ (Kroeger 2005: 2). At
the same time, the need of expressing a certain meaning or communicative
function does not necessarily lead to a unique linguistic form.

‘In other words, we cannot fully explain the form of an utterance while
ignoring meaning and function; at the same time, we cannot account
for the form of an utterance by looking only at its meaning and
function’ (Kroeger 2005: 2).

It is necessary to find the devices which allow a fuller, deeper

account of these replies. If we must observe the grammar rules when
structuring a sentence, in order to obtain a logical combination of the
linguistic elements, attention must also be paid to the conversational
restrictions and tendencies which make the users combine certain linguistic
elements in a particular way to express a particular message.
For instance, in the example

Don’t concern yourself with that, things are settled, and anyway, what are
friends for?

the speaker has to observe the rules of forming the negative form of the
Imperative Mood, use the right preposition after concern, apply the concord
subject-predicate in the second and third sentence, and form the
interrogative of the third sentence verb. All these syntactic rules are not
optional, but in a concrete situation of communication, if the speaker
chooses to speak in the colloquial register, a double negation or the use of
the invariable Present Tense negative form of be, ain’t, are accepted even
if they are not correct in standard language: No sweat, it ain’t no problem,
you’re my buddy.
Grammatical correctness often collides with the users’ perception as
to what is correct. To describe the grammar of a language essentially
means trying to explain why speakers recognize certain forms as being
“correct” but reject others as being “incorrect”. Correctness is an umbrella
covering both syntactic structures (formed according to the rules specific to
a language) and logical semantic combinations (according to semantic
rules): the latter are expressed by the former. Both require the linguistic
competence of the user.

81
 Acceptability is at the border between correctness and incorrectness,
and only the criterion of frequency tips the scales in favour of the
acceptance, integration and generalisation of a certain use, or, on the
contrary, of its rejection.
In terms of linguistic domains, acceptability brings together
semantics, socio-linguistics, psycho-linguistics and pragmatics. If the latter
is to be defined as implying the speaker’s choice, then all the other
domains are subsumed to it. The user’s competences remain the basic
criterion in his choices.
Of course, some would argue that the domain of pragmatics is thus dilluted.
Our aim is not to favour one or another of the viewpoints, but to show the
interfaces of pragmatics with other linguistic domains. The conclusions
remain to be drawn accordingly.
Kroeger (2005: 2) stresses on the acceptability of the form itself,
rather than on the meaning or function which it expresses. A native speaker
of a language will often be able to understand a sentence perfectly well
even if it is not grammatically correct [...].The examples provided by
Kroeger involved elliptical sentences, non-observance of subject-predicate
agreement or of Past Tense Interrogative forming rules:

Me Tarzan, you Jane.

Those guys was trying to kill me.
When he came here?

Conversely, the form of a sentence may be accepted as correct

even when the meaning is obscure or absurd: His wife is fine, actually he is
not married. We will discuss that in the chapter about presuppositions.
Semantically, the two sentences are contradictory, but, pragmatically, the
utterance makes sense in a context such as:

‘How is Bill’s wife?’ ‘His wife is fine, actually he is not married.’

The first reply implies that Bill is married. This implied meaning is
contradicted explicitly by the content of the last sentence.
Similarly to morpho-syntactic choices, at lexical level the range of
choices is even greater, depending on the competence and communicative
intention of the speaker: friend and buddy are synonyms, bother, concern,
sweat, worry are also partial synonyms having various degrees of intensity
and formality.
The conclusion to be drawn from the analysis of the examples
above is that grammar is rule-dependent, even if at formal level there is to
be made a distinction, too.There are some rules about using language that
must be consciously learned, the kind of rules learned in school. They are
prescriptive rules and define a standard form of the language, which some

82
authority must explicitly state for the benefit of all speakers. There are also
rules that the native speaker is usually not aware of because they comprise
that kind of knowledge about the language that children learn from the
speakers they are surrounded by and interact with, in a natural and
unconscious way.
According to Leech’s assertions (1983: 5), there is a contrast
between grammar and pragmatics: grammar is based on rules,
conventional in nature, generally allowing no exception, prescriptive,
focusing on form, ideational and describable in terms of discrete and
determinate categories; pragmatics is governed by principles, non-conventional
in nature and motivated by conversational goals, interpersonal and textual,
describable in terms of continuous and indeterminate values.
In Kroeger’s view, studying the interdependence among form, meaning
and function means defining grammar as the totality of these non-prescriptive
rules rather than ‘all the structural properties of the language except sound
structure(phonology56)’, i.e. the structure of words, phrases, sentences, texts,
etc. Word-order facts within any given language tend to show interesting
patterns of correlation, and the patterns observed in different languages tend
to vary in limited and systematic ways (Kroeger 2005: 7).

‘All languages, whether standardized or not, have rules of this kind

[non-prescriptive], and these rules constitute the grammar of the
language. The term grammar is often used to refer to the complete
set of rules needed to produce all the regular patterns in a given
language’ (Kroeger 2005: 5).

It is interesting to notice how during a timespan of almost 30 years,

the concept of grammar was re-defined and broadened so as not to be
equated simply to correctness and formality.
In his introduction to Principles of Pragmatics (1983), Leech also
refers to pragmatics as the study of the use of language that is distinct
from, and complementary to, the language itself seen as a formal system.
But, in spite of the differences between grammar and pragmatics, ‘one
must develop theories and methods of description which are peculiar to
pragmatics itself, and show that these have to be different from those which
are appropriate to grammar’.

1.2. The domain of pragmatics

To circumscribe the domain of pragmatics means considering the

stages of its development, from its beginnings to present times, including

56 If emphasis is linked to language use, then phonetics too is linked to pragmatics, and

some linguists consider the phonetics-pragmatics interface.

83
the recently-appeared areas of research. The foundation of any attempt of
defining pragmatics should be Morris’s taxonomy having semiotics as
hyperordinate term that comprises syntax, semantics and pragmatics as
components. The terms syntax and semantics are used considering a
logical and philosophical perspective: syntax comprises morphology,
because logical languages, unlike natural languages, have no morphology;
semantics refers to denotational meaning (see Part I Semantics, 2.1.; 3.2.)
i.e. to the relationship sign-referent. According to Morris, pragmatics means
‘the study of the relation of signs to interpreters’ (Morris 1938: 6).
Becoming more specific in referring to linguistic signs combined to
form words and emphasizing the concrete character of the elements
studied, Mey’s definition is a clear reflection of Morris’s:

‘Pragmatics is the science of language seen in relation to its users.

[…] the science of language as it is used by real, live people, for their
own purposes and within their limitations and affordances’. (Mey
1993: 5)

Pragmatics is equally based on understanding and on the

cooperation of the interlocutors. It is a theory of the users comprising

‘everything that characterizes the user as a person whose use of

language depends on the rules and norms that are valid at any time,
in any place, in the community in which he or she is living.’ (Mey
1993: 37)

Limitations and affordances are linked to objective and subjective

factors. Objective factors include the potential represented by the syntactic
rules and the lexicon of a given language. Subjective factors comprise the
users’ competences: linguistic competence, encyclopedic competence and
communicative competence. All these types of competence depend on the
users’ context of culture and life. Therefore, pragmatics ‘tells us it’s all right
to use language in various, unconventional ways, as long as we know, as
language users, what we’re doing’ (Mey 1993: 4). Language is a form of
behaviour and our idiolect (the specific, unique linguistic variant each of us
uses) represents the sum of our particular forms of linguistic behaviour in
various contexts.

‘Pragmatics is the study of the conditions of human language uses as

these are determined by the context of society.’ (Mey 1993: 42)

Levinson’s definition of pragmatics seems to give a description at a

more general level, by implicitly acknowledging the role of variable factors
and considering just those relevant, grouped together as context; a similar
process of further generalization and essentialization is operated at the

84
level of the relations language − context; Levinson views pragmatics as
inevitably linked to linguistic structures and accounting for the infinite
potential of languages in point of expressivity:

‘Pragmatics is the study of those relations between language and

context that are grammaticalised, or encoded in the structure of a
language’. (Levinson 1983: 9)

1.3. Connections to the other linguistic branches

According to Morris (1938), syntax, semantics and pragmatics are

included into the domain of the general science of signs, semiotics. Signs
have origin, uses and effects within the communicative behaviour they occur
in (see Part I Semantics, 2.3., 2.4.). That implies psychological, biological
and sociological phenomena studied by general pragmatics and indirectly by
linguistic pragmatics.
Language is a communicative complex able to perform various
functions, but should be seen as made up of concrete means for concrete
users at concrete moments in time. That cannot be accomplished in the
absence of a structure having a sense. Consequently, the basic relations to
be discussed are between linguistic form and pragmatics, and between
semantic meaning/sense and pragmatics.
Verschueren’s opinion (2001: 83) summarizes the relationship
among linguistic branches:

‘If it is the case that pragmatics approaches language as a form of

behavior, there must indeed be a pragmatic way of looking at any
aspect of language at any level of structure. In other words, unlike the
traditional components of a linguistic theory, pragmatics does not
have its own privileged object of investigation. And if this emphasis
on behavior implies psychological, biological, and sociological
considerations, pragmatics must by definition be highly
interdisciplinary, thus rubbing against the hyphenated disciplines
(psycholinguistics, sociolinguistics and the like) while differing from
them by its lack of a correlational object of its own (the mind, society,
and the like).’ (Verschueren 2001: 83)

1.3.1. Connections form - pragmatic meaning

Linguistic forms are the material carrying semantic and pragmatic

meaning, even if some would agree with Green (2004) that pragmatic
information depends not on linguistic forms in themselves, but on their use
as an act resulted from their uttering.

85
 To Ariel (2008) the connection grammar-pragmatics ultimately
means the connection code − inference. Language is a code and its signs
are combined into structures which in a certain situation make the users
infer implicit meanings. Of course, when it comes to deictic elements or to
speech acts, they don’t have implicit meaning, but are to be interpreted in
the context of use.

‘Codes and inferences do make contact, or we wouldn’t be able to

explain many cases of historical change. The fact that so much of the
linguistic code can be pragmatically motivated stems from the origin
of grammar, the extralinguistic factors guiding interlocutors. Codes
commonly develop out of (salient, recurrent) speaker-intended
inferences associated with specific forms.’ (Ariel 2008: 306).

Furthermore, Ariel (2008: 117) considers that

‘(some) interpretative aspects traditionally classified as pragmatic do

not merely complement the grammar’ [...]. ‘Rather, some
extralinguistic generalizations regarding forms must constitute part of
the grammar (e.g. definite NPs encode Given/identifiable information,
Left Dislocated sentences introduce discourse-new entities) [...]’.

In other words, the semantic content of the referring expressions is

identified only as a result of considering the situational context, and both
the processes involved and the result led to the formation of a pattern. Ariel
argues that grammars routinely evolve as a response to extragrammatical
forces and, consequently, there are linguistic conventions associating
linguistic expressions with extralinguistic factors, which leads to the finite
number of grammatical forms. Generalising, all of grammar could be
motivated pragmatically, since discourse patterns are so. Hyman’s
conclusion, quoted by Ariel (2008: 117) might be agreed to by some
researchers, “when pragmatic factors become part of grammar, the result is
syntax and morphology” (Hyman 1983: 71–72). In the same line, Kuno
(1987) believes that the syntax chosen for a given sentence corresponds to
the perspective of the speaker.
The patterns referred to in the previous paragraph imply similar
cognitive processes and linguistic uses of the whole community of language
users. As early as 1933, Bloomfield explicitly acknowledged the fundamental
role played by two concepts: speech community and speech utterance. The
members of a speech community produce speech-utterances which, if
repeated in rather similar situations, acquire a certain degree of
conventionality and are almost automatically decoded by the users precisely
because of their conventional character: for instance, the use of declarative

86
affirmative sentences, which apparently offer information on the speaker, to
express a refusal:

Shall I bring you something to eat?

I’m not hungry. = I don’t want anything.

Ariel (2008: 118) firmly stated that

‘we must distinguish between functionality and extralinguistic

motivation. Grammar is functional if it is adequate for its speakers for
their communicative purposes. Arbitrary grammatical conventions
pose no obstacle for successful communication’.

Any linguistic change is determined by some communicative needs and this

is what makes a user innovate at structural level. The motivated character
of grammar is determined by its being ‘a natural historical product’
(Ariel 2008: 118).

1.3.2. Connections sense - pragmatic meaning

Discussing semantic meaning/sense (we chose Ariel’s use of the

term semantic meaning which is synonymous to the concept of sense; see
Part I Semantics, 2.2., 3.2., 3.3.) presupposes the presence of a phonetic
form. If we agree to Bloomfield (1933) who follows Saussure, Humboldt,
and continues an idea stated by the antiquity grammarians, more precisely,
Aristotle, and, later, medieval speculative grammars, a linguistic form is a
phonetic form which has a meaning, therefore meaning is necessarily
attached to form. In this light, semantics has to establish what meanings
are attached to phonetic forms. Bloomfield implicitly makes no difference
between the two types of meaning (semantic and pragmatic) when defining
the meaning of a linguistic form as ‘the situation in which the speaker utters
it and the response it calls forth in the hearer’ (1973: 138-139).
The interdependence between semantic and pragmatic meaning is
the very idea underlying the definition of pragmatics: following Grice, we
typically say (i.e. mean) more than we say. In other words, literal meaning
(i.e. denotative lexical meaning) is the foundation on which various layers of
pragmatic meaning, having nothing to do with the words themselves, but
with the situation in which they are uttered, are added. The example I’m not
hungry is an utterance which contains some information about the
speaker’s state at the moment of uttering, but linked to the first utterance
(Shall I bring you something to eat?) whose reply it is, counts as a refusal.
Outside this situation, such an utterance would seem out of place.
A conclusion of this argumentation can be considered Leech’s
definition of pragmatics as an approach attempting to link the two types of

87
meaning: ‘general pragmatics relates the sense (or grammatical meaning)
of an utterance to its pragmatic (or illocutionary) force’ (Leech 1983: 5).
The selection of certain linguistic elements and the rhetorical
structure of a sentence, the latter accounting for the semantic-pragmatic
interface, depend on the characteristics of the user and contribute to
conveying the intended meaning. The connection between pragmatics,
semantics and socio-linguistics is, thus, obvious. Register plays an important
part in this respect (for more on register see Part I, Semantics, 3.6.).
Let’s consider the next three examples:

1. No sweat, it ain’t no problem, you’re my buddy.

2. Don’t worry, it’s ok, we’re friends!
3. Don’t concern yourself with that, things are settled, and anyway, what
are friends for?

Syntactically, example 1 is made up of three sentences which are

juxtaposed, but that is of no direct relevance. It is more relevant what each
of them expresses and the result of combining them from the perspective of
the interlocutor. The first two sentences are synonymous, the first
containing a slang expression no sweat, which has a figurative meaning
‘don’t worry’ (for more on synonymy, see Part I, Semantics, 4.3.). Its
meaning is resumed in the second sentence by using a more neutral
structure but still in the colloquial register, since it contains two negations.
The register used connotes [+closeness], [+informality] [+trust], [−vanity]
and is justified by the content of the third sentence, in its turn containing a
colloquial term, buddy.
The first two sentences reinforce each other, expressing the
message to be conveyed to the interlocutor, objectively justified by the third
sentence: S1 + S2 ← S3. This choice of expressing the message is not the
only one possible, but it is the one selected by the speaker because it
seemed the most appropriate to him.
Examples 2 and 3 are variants of 1 which express the same message,
even if the register used is different; 2 is rather neutral and 3 uses more formal
structures.
Many more variants can be added to the ones above, the sentence
meaning remaining basically the same, but the connotations implied
varying to a great extent. Generalizing from the analyses of the examples
above, a strictly descriptive frame based on formal reasoning cannot fully
explain the concrete meaning of the previous examples.

1.4. Subdomains of pragmatics. Some of their basic terms and

concepts

‘[...] definitional efforts, therefore, have been stranded either on vague

and impracticable distinctions (such as semantics studies meaning

88
 out of context, while pragmatics studies meaning in context) or on ad
hoc lists of topics that were supposed to belong to the province of
pragmatics (in particular: deixis, conversational implicatures,
presuppositions, speech acts, and conversational structures’
(Verschueren 2001: 83).

In spite of the implicit simplification characterising any attempt of

systematization, for pedagogical reasons, it is useful to approach pragmatics
by discussing its major domains mentioned by Verchueren and agreed upon
by most linguists (deixis, conversational implicatures, presuppositions, speech
acts). On the other hand, it remains true that these domains intertwine and in
many cases we are confronted with the circularity of definitions and
explanations.
Describing the domain of pragmatics as a science in its own right
includes mentioning its sub-domains, and they reflect the main concerns of
interlocutors during and regarding, the verbal exchange: focusing on some
elements of discourse (grammaticalised in the form of deictics), implying a
meaning beyond the semantic one, but not independent of it (inferences, be
they implicatures or presuppositions), the concrete message sent to the
interlocutor (speech acts).
In the following example,

It’s too late for you to go now.

you designates the interlocutor who is also the addressee, the person to
whom the message initiated by the speaker is oriented. The Present Tense
short form of the verb to be, ’s , and the adverb now are deictics expressing
the focus on the present moment, the moment of uttering; the personal
pronoun you is a person deictic which makes reference to the interlocutor in
the concrete situation of communication. Together with the specific word
order, deictic markers used in the sentence help the interlocutor make the
inference that there was a favourable previous moment which was not
taken advantage of. Therefore, the message conveyed by the speaker is
that of refusal.
Any pragmatic study has the concept of utterance as the basic dual
(structural & functional) unit. While a sentence is the basic unit of analysis
in syntax, where the focus is equally on form and semantic meaning, i.e.
logical meaning, utterance is a key term in pragmatics, since it is the result
of the process of uttering. The structure of the utterance is irrelevant, since
it is its function, the communicative goal underlying it, that matters. The fact
that we are dealing with an elliptical sentence or with a complex sentence is
secondary, what matters is their appropriacy to the intended message:

89
‘No.’ ‘Not me.’ ‘I would never do such a thing if I were you and you should
know better than that.’

Whether we talk about an utterance which structurally is made up of

a sentence or a complex of linguistic forms larger than the single sentence,
a discourse, which in the written mode of a language constitues a text,
language use is action, and investigating it requires attention paid both to
language and to action (Hanks 1996). ‘Real language’ (Brown and Yule
1983; de Beaugrande and Dressler 1981) means linguistic structures
actually used by people. This conception of discourse, broadly speaking,
underlies the development of contemporary linguistic pragmatics
(Blommaert 2005: 2).
But language operates differently in different environments, and to
understand how language works, we need to contextualise it properly, to
establish the relations between language usage and the particular
purposes for which and conditions under which it operates (Blommaert
2005: 14).
The essential role of the context is transparent from Mey’s assertion
that ‘ambiguity exists only in the abstract’ (Mey 1993: 8). Due to the
linguistic and mainly situational context, any utterance is disambiguated, at
least partly. The concept of context is to be invoked to determine what an
ambiguous sentence means, but there is an indefinite number of elements
playing a role in the production and interpretation of utterances.
The context should be established in connection to the ongoing
interaction between the interlocutors, since only the dynamic development
of the conversation offers the necessary clues to its understanding.
Consequently, the context is dynamic (an environment in steady
development) and pro-active (oriented towards the future).
The context is not ‘just a widening of the sentential perspective: it is
the total social setting in which the speech event takes place’, what Mey
(1993: 31), following Bilmes, calls context of use. Mey distinguishes
between societal context, primarily determined by society’s institutions, and
social context, primarily created in interaction.

It’s a long time since we visited your mother. (Mey 1993: 39)

Context 1: uttered by the husband to his wife, in the married couple’s living
room, in the morning.
Context 2: uttered by the husband to his wife, at the zoo, in front of the
hippopotamus enclosure.
Context creates the conditions for the interpretation of an utterance,
depending on the competences of the interlocutors, and also on their
motivation. Referring to agency and cause-effect relationship, Hanks (1996)
concludes that

90
 ‘if a feature of the context …is encoded, then speakers using the
language are habituated to take notice of the corresponding aspect of
context.’ (Hanks 1996: 179)

The first context leads to a neutral, literal interpretation (in the

absence of any other elements proving the contrary), whereas, the second,
has a definite ironical connotation.
The very presence of a linguistic element in a sentence means that it
had been selected by the user because it is important in the context, i.e. it is
relevant. Paralinguistic devices, i.e. stress, intonation, pause, can highlight that
importance.
In the previous example, the relevant phrase is your mother, since it
is to be interpreted in relation to the interlocutor and in relation to the
situational context.
Starting from the literal meaning of a sentence, which is standard,
conventional, (‘type’ or ‘token’ meaning in Grice’s terminology), the user
draws on inferences based on his/her best guesses as to what the speaker
intends to convey (i.e. speakers’ meanings, in Grice’s terminology). They
are non-literal or indirect meanings, impossible to be deduced only by the
sum of the meanings of the linguistic elements making up the utterance.
Pragmatic inferences are to be calculated by both the speaker and the
interlocutor according to their common ground, i.e. their shared knowledge
regarding the topic of the discussion.
In Mey’s example

It’s a long time since we visited your mother.

the interlocutors’ relationship is crucial, and so is the place where the

utterance is made. But, irrespective of that, it is obvious that the interlocutor
has a mother, and that the two persons talking visited her previously, which
in its turn, implies that they are not living together.

Indeterminacy denotes the indefinite value of a referential

expression, whose reference is ambiguous in the linguistic context, and is
disambiguated only if decoded against the background represented by the
situational context. What linguistic elements have the feature
[+indeterminacy]? How is the problem solved?

Betty’s father gave her a present. Cf Both Mary’s and Jane’s father gave
her presents.

In the example above, the personal pronoun in the Dative Case, her, can be
coreferential with Betty, or not, depending on the situational context which
disambiguates the reference. Of course, the first interpretation implies their

91
coreferentiality. If they are not coreferential, her will be stressed in speech
and will imply a contrast (to her, not to somebody else). In the example
where the synthetic Genitives determine identical nouns coordinated it is
obvious that the Dative Case pronoun is not coreferential with any other
nominal element in the sentence.
In the example

Yesterday’s events troubled the President.

the time deictic yesterday and the nominal phrase the President have
variable reference, depending on the relative time of reference (the
absolute time corresponding to now), and on the time and place of the
utterance (in a certain place/institution etc, at a certain time, there was a
unique referent having the function of president). (see also Part I, 2.1.)

Deixis is a Greek term, synonymous to the Latin term

demonstrative, which refers to those elements making up the situation of
communication (participants and the relationships among them, time, place,
topic) on which the interlocutor’s attention is drawn by means of various
linguistic devices. Deictics are the concrete elements pointed out in a
certain utterance. Such elements are implicitly stressed because they are
considered relevant by the speaker. Let’s consider the example

I’m here now.

In this case, all the elements making up the utterance are deictics. The
speaker points out his/her presence by the use of the first person personal
pronoun I, at the moment of speaking (both grammatical and lexical means
are employed: the former represented by the use of Present Simple, the
latter by the use of the adverb of time now), in a certain place, where the
uttering process occurs (by using the adverb of place here).

Appropriacy refers to the competence of the users to make the most

adequate choices in point of word selection and combination, so as to best
express their communicative goals.

You are not Mrs. Brown, you are my daughter.

This example proves the connection between the deictic you and the
address term Mrs. Brown, on the one hand, and the referential expression
my daughter. The referent denoted by you can be described from various
perspectives, and expressed by various linguistic expressions but the
speaker stresses the NP which relates him/her to the interlocutor, making the
most appropriate choice. The structure Mrs. Brown connotes distance,

92
independence, even estrangement if interpreted in opposition to the NP my
daughter.

My dear, you should listen to these people, because they certainly mean well.

The deictic you is correlated to the address term my dear, a term of

endearment, which connotes a close relationship and thus marks the
authority of the speaker. The speaker has to use his/her authority because
the whole utterance functions as a request. The modal verb should used
with deontic value enhances the desired effect. The deictic these and the
personal pronoun they designate the same referents, who are obviously
present in the situation of communication, and the speaker can point at
them during the speech event. Their use, corroborated which the content of
the adverbial clause of reason, has a cumulative persuasive effect (or are
meant to).
One should notice that all the elements enumerated as basic for a
pragmatic analysis coexist at the level of the utterance, they don’t exclude
one another, but together, contribute to the adequate expressing of the
message intended by the speaker and to its understanding by the hearer.
For example, in the utterance I’m here now there are three elements whose
reference is variable, i.e. indeterminate, depending on the situational
context: I, here, now. I can be any speaker, any member of the linguistic
community, here can have a narrower or wider area of reference (part of a
building, building, part of a locality, locality, geographical area –county,
region, district, country, continent), now can refer to the exact time of the
day, to the day of the week, date, week, month, year, or to any moment in
time which is relative in relation to some events known by the interlocutors.
Therefore, the utterance can be interpreted in many ways by the
interlocutor, depending on the circumstances of communication; some
variants are:
a. I am in my hotel room today.
b. I am in Bucharest at present.
c. I am at the supermarket at the moment.
d. I am at my parents’ this week.
If the interlocutors are aware of the reference ascribed to each
deictic, there is no need to be so specific as in the above utterances. If the
interlocutor does not have the necessary shared information, then the
speaker should be more specific and give all the necessary details
explicitly. Subjective reasons, on the other hand, may prevent the speaker
from being specific enough, and he/she might avoid precision on purpose,
in order to hide that relevant information from the interlocutor (imagine a
discussion parent-child). Therefore, deictics can be intentionally used with a
certain degree of indeterminacy (which is in itself relevant) in a certain
context, to make the interlocutor draw a certain inference (the parent would

93
know that the child is in a place he/she is not supposed to be, a friend/ a
business partner would know that the speaker is not able to meet him/her
or to speak on the phone).
Uttering something, no matter what linguistic structures are used,
means action, the action of expressing a certain message to be understood
by the interlocutor. There is no one to one correspondence between a
linguistic structure and a certain type of speech act. A speech act implies
using some words in structures which are correct and meaningful in the
language used, in order to convey a certain meaning and to obtain a certain
effect on the interlocutor. The communicative goal or illocutionary force of
an utterance is the message conveyed by the speaker and, hopefully,
adequately interpreted by the interlocutor. The role of the context is
essential in the process of message understanding. The variants of
illocutionary values associated to the utterance below are a proof in this
respect. The speech act is part of the thematic deictic center included in the
situation of communication (with all its components and features). From this
perspective, the verbal exchange consisting of successive speech acts is
(the core) part of a speech event.

I’m exhausted.

Structurally, the utterance above is a simple sentence made up of the first

person singular personal pronoun I, designating the speaker in the situation
of communication, the link verb be, connecting the grammatical subject to
the property assigned to him/her, and the adjective functioning as a
predicative, exhausted. The intensity of the adjective helps conveying the
message of:
a. reproach if the interlocutor caused the subject to overwork;
b. excuse/refusal if the interlocutor has some expectations of the
subject;
c. request if the speaker wants the interlocutor to perform a certain
action on his/her behalf.
etc.

Beyond the domains encompassed traditionally by pragmatics,

Verschueren proposes a more flexible and open-minded view on
pragmatics as an approach, rather than a domain, whose boundaries and
area are to be strictly delimited:

‘An alternative view, going back radically to Morris and gaining

ground slowly, is to treat pragmatics as a specific identifiable
perspective on language, in particular a functional perspective
studying language from the point of view of its usage phenomena and
processes (see Verschueren 1999). Since language use involves

94
 human beings in all their complexity, the perspective in question is
necessarily interdisciplinary, touching on aspects of cognition,
society, and culture in a coherent and integrated approach, without
privileging any of these specific angles. For the same reason, the
perspective in question must pay attention to flexible processes of
making linguistic choices, both in production and in interpretation,
from a variable (and in principle infinite) range of options, in a manner
that is negotiable and dynamic rather than mechanical, thus betraying
a high degree of adaptability’ (Verschueren 2001: 83-84).

Conclusions

Pragmatics offers a new and distinct perspective on utterance

interpretation, representing a necessary direction of analysis. Any utterance
has a literal meaning, a structural meaning, i.e. the sum of the meanings of
the linguistic items that make it up, and that is the basis to which the
speaker adds his/her own meaning (speaker’s meaning), depending on the
context (linguistic and situational) that plays a disambiguating role, solving
the issues linked to indeterminacy and relevance. In their turn, deictics are
essential in offering the interlocutor the frame necessary to interpret an
utterance and to draw the appropriate inferences.

95
96
 2. THE CONCEPT OF DEIXIS. TYPES OF DEIXIS

2.1. The term and concept of deixis. Types of deixis

2.2. Person deixis
2.2.1. Means of expressing person deixis. Personal pronouns system
2.2.2. The reference of pronouns. Deixis vs anaphora
2. 3. Time deixis
2.3.1. Definition
2.3.2. Linguistic markers of time
2.3.3. Time vs tense
2. 3.3.1. Absolute and relative tenses
2.3.3.2. Tense and Aspect
2.3.4. Coding time, event time, reference time. Deixis and grammar
2.3.5. Lexical time markers
2.3.6. Time vs space deixis
2.4. Space deixis
2.4.1. Definition and importance
2.4.2. Classification of space deictics
2.4.3. Main values
2.4.4. Combined values
2.4.5. On time and space deixis (again). Which was first?
2.5. Social deixis
2.6. Discourse/ textual deixis
2.7. Empathetic deixis
Conclusions

2.1.The term and concept of deixis. Types of deixis

Starting inductively, we can consider an example such as:

I told you to come here tomorrow.

All the elements that we should point out in this utterance are related to the
speaker (I) and make up an ensemble calle deictic center. I is always the
speaker, in the same way as you is always the hearer/listener/interlocutor.
The infinitive to come and the adverb of place here refer to an action of
moving towards the speaker, and to the place where the speaker is,
respectively. The adverb of time tomorrow refers to a moment in time which
is posterior to t0, the uttering moment. As can be seen, all these elements

97
depend on the context to be interpreted concretely, beyond their
person/space/time dimension.
Consequently, it can be said that

Deixis is a discourse structure property,

‘the single most obvious way in which the relationship between

language and context is reflected in the structures of languages
themselves.’
(Levinson 1983: 54)

‘Aspect fundamental al organizării pragmatice a discursului,

desemnând faptul că referinţa anumitor componente ale unui enunţ
nu poate fi determinată decât prin raportare la datele concrete ale
situaţiei de comunicare./ A fundamental aspect of the pragmatic
discourse structure, which designates the fact that the reference of
some utterance components can be determined only by relating it to
the concrete data of the situation of communication. (our translation)
(Ionescu-Ruxăndoiu 2001: 157)

Both definitions point out the essential value of deixis in

communication, as an inherent and necessary aspect of discourse
structure/language structure. As a process, deixis can be said to be the
encoding of the utterance context.
We can conclude that ‘deixis abounds in language use and marks
one of the boundaries of semantics and pragmatics’ (Fromkin & Rodman
1998: 201).

Types of deixis

Types of deixis can be grouped into the basic subgroup including

person, space and time deixis, and a newer subgroup, rather 'parasitical' in
nature, including social deixis, discourse deixis and empathetic deixis. The
last three can be considered parasitical since they rely on the deictic
markers specific to the other types of deixis and even their existance
implies the manifestation of the basic ones.
Person deixis encodes the reference made to the participants in the
speech event.
Place deixis encodes the reference made to the place where the
speech event takes place.
Time deixis encodes the reference made to the time when the speech
event takes place.
Social deixis encodes the relationships among participants in the
speech event.

98
 Discourse deixis encodes the manner in which the components of a
text acquire coherence.
Empathetic deixis, connected to social deixis, expresses some
affective meaning involving the attitude of the speaker towards a certain
referent, be it positive or negative.
Any type of deixis is expressed by some linguistic markers,

‘ways in which languages encode or grammaticalize features of the

context of utterance or speech event.’ (Levinson 1983: 54)

They are deictic words or pointers (Mey 1993: 92), which in many
cases can acquire multiple functions, which made Jakobson call them
shifters. Such deictic markers can be specific or non-specific to a certain
type of deixis (conjunctions are typical discourse deictic markers but
personal and demonstrative pronouns can function as markers of various
types of deixis).
Regarding their nature they can be lexical or grammatical, i.e. they
are words or simply morphemes.

2.2. Person Deixis

This type of deixis takes into account the praticipants in the speech
event. They can be active (speaker, hearer) or passive (by-standers,
eavesdroppers – voluntary participants, the former visible, the latter
apparently not present during the dialogue; over-hearers – involuntary
participants).

2.2.1. Means of expressing person deixis. Personal pronouns system

Person deixis is instantiated by indexical expressions

‘a particular kind of referential expressions, where the reference is not

just “baldly” semantic, but includes a reference to the particular
context in which the semantics is put to work’. (Mey 1983: 91)

For person deixis, the basic indexical elements are personal

pronouns.

2.2.2. The reference of pronouns. Deixis vs anaphora

Pragmatically, what matters with personal pronouns is their

reference and role in communication. Pronominal reference is variable and
recoverable either due to the linguistic context or to the situational context:
first and second person pronouns are deictic by nature, their reference

99
being identifiable strictly within the situation of communication, third person
pronouns have a reference which can be recovered only within the
linguistic context.

I was walking, you were running.

With I the role in communication is a stable feature, whereas its

reference can be variable; the same is true for you:

‘Everybody can say I, and whoever says it points to another object

than everybody else; one needs as many proper names as there are
speakers, in order to map the intersubjective ambiguity of this one
word into the unambiguous reference of linguistic symbols.’
(Bühler 1934: 103 apud Mey 1993: 90, Mey’s translation)

First person personal pronoun always refers to the speaker, second

person personal pronoun refers to the interlocutor, hence their
communicative role57 is automatically established: that would be their
reference (cf Saussure’s signifié); the referent’s recoverability can pose
some problems, whenever there isn’t a face to face type of communication,
in written texts or when people talk on the phone etc. The speaker is also
the source of the message or just the messenger, the addresser; in the
same way, the hearer is the addressee or just the receiver of the message:

You are to wait here. This is the rule. (the addresser is not the source)
The teacher to the students: ‘Everybody, Mary must come at once.’ (the
hearer is not the addressee)

Third person pronouns have no role in communication attached,

since they refer to people generally not present, or not involved in the
communicative event. It is only their reference which can cause some
ambiguities in interpretation. The person referred to by using a third person
pronoun may be present in the situation of communication and the speaker
can refer to him/her/them with various intentions, depending on whether
he/she can hear the speaker:

People don’t know when to stop, do they?

He is rude to his guests.

57 Fromkin &Rodman (1998: 201) state that ‘The pronoun I certainly has a meaning

independent of context – its semantic meaning, which is “the speaker”; but context is
necessary to know who the speaker is, hence what “I” refers to.’

100
 Third person pronouns have anaphoric or cataphoric value,
anticipating or resuming the reference made usually by means of a
noun/noun phrase in the linguistic context, the two, noun and pronoun,
being co-referential; in other cases the condition of co-referentiality is not
met, the interpretation of the pronoun relying on the situational context (see
Pisoschi 2012: 156-157 ):

Ann1 is here and she1 is waiting for you. (Ann and she are co-referential) cf
Rom. ‘Ana e aici şi te aşteaptă.’

Ann1 is here and she2 is waiting for you. (Ann and she are not co-
referential) cf Rom. ‘Ana e aici şi te aşteaptă şi ea (de acolo).’

She1 is here, Ann1.

In this case, the pronoun has a cataphoric value since it anticipates the
referent expressed by the noun.

When she is home, Ann is always painting, it is her passion. cf Rom.

‚Când e acasă, Ana întotdeauna pictează.’

Considering the example

John loves hisi mother.

we notice that the pronoun his can be coreferential with the antecedent John
(this is what Recanati (2005) called anaphoric use), or can refer to another
referent (this is what Recanati called free use) recovered from the situational
context. If this sentence is embedded in a quantificational context, the
pronoun doesn’t have a definite value, but a course of values, since the
antecedent is a quantifier. The values of the pronoun vary with the individuals
introduced by the quantifier every. This is what was defined by Recanati as
the bound use: it is similar to the anaphoric use, the referent being not a
definite value assigned, but a course of values (Recanati 2005: 287).

Every boyi is such that John loves hisi mother.(every boy has a mother and
all the mothers of the boys refered to by every boy are loved by John)

Second and third person pronouns can have a generic value in

contexts where the reference is made to a group of people or to the whole
class of human beings; pragmatically, universal truths are thus expressed
in the form of proverbs, sayings etc.

They believe it. = It is believed.

He who laughs last, laughs best.
You never know.

101
 Personal pronouns in the plural have some specific referential
characteristics. Even if, typically, we doesn’t always correspond to a plural I
(Kerbrat-Orecchioni 1999: 46), we is ‘more than one I; its plural’ (Wales
1996: 59), the exceptions are represented by:
- We = I.
The two extremes are the plural of modesty and the plural of
majesty:

We, the Queen of the United Kingdom...

We, the author, would like to thank all our readers.

The three subtypes of the plural of modesty, authorial we (Crystal

1985), lecturing we and doctor we, have each some distinctive features.
Authorial we has the feature [- addressee] and [+ ego] (Wales 1996: 66),
collocating with declarative or mental verbs:

We discuss the problem of pollution in contemporary society.

- We = you when implying the moral authority of the speaker, in

interactional utterances, and when the communicative goal is persuasion,
empathy or reproach:

Shall we go now?
Are we better today?
Aren’t we extremely unfair?

2.3. Time deixis

2.3.1. Definition

As a subtype of deixis, time deixis encodes the reference made to

the time when the speech event takes place.

2.3.2. Linguistic markers of time

Linguistic markers can be classified according to their nature

(grammatical or lexical) and to their character (absolute or relative).
Absolute time markers are not deictic since they are not related to the
speaker and the situation odf communication.
Time deixis is expressed by grammatical means (tense, aspect)
and lexical means (adverbs of time, noun phrases, prepositional phrases
functioning as time modifiers):

I like flowers.

102
(like is a verb in the present tense, therefore it is implicitly marked
deictically – since there is no explicit marker of tense-)

They were walking [when I saw them].

(the verb in the main clause is in the past continuous, expressing an action
in development at the moment when the action in the time clause took
place. The two actions were simultaneous in the past and implicitly anterior
to present, the speech time. The time clause marks the time frame for the
action in the main clause (hence its name: adverbial clause of time).

I saw them yesterday, not two days ago.

(yesterday, two days ago are lexical deictic markers). The former is an
adverb, the latter a noun phrase, both being syntactically adverbial
modifiers of time. These two deictics are relative in nature, since they
depend on the moment of speaking (now). Yesterday can be defined only
in relation to today, which is in itself a relative notion. The same is true for
the structure two days ago.

Other lexical time markers are absolute in nature, i.e. they

correspond to calendrical time divisions, even if a certain level of relativity is
implicit in their case, too (consider the various calendars specific to various
cultural spaces − the criterion was either the duration of the solar year,
hence the difference between the Julian calendar and the Gregorian
calendar in use at present, or a religious event, hence the creation of the
Christian calendar, the Hebrew calendar, the Mayan calendar, each having
a different origo point):

See you on Sunday.

(i.e. next Sunday, the closest Sunday to the day when the utterance is
made)

If relativity is clear in the previous example, no ambiguity arises in

the following examples where the past action and the future actions,
respectively, are implicitly anterior or posterior to the present moment when
the utterance is made:

See you on the 3-rd of March next year.

They will move out on the 15-th of May 2015.
In 1821 the former emperor died.

2.3.3. Time vs tense

Time is a philosophical category, a continuum which is rendered

linguistically by means of the category of tense, hence Comrie’s concise

103
definition: “Tense is grammaticalised expression of location in time”
(Comrie, 1985). Time is one-dimensional and moves in only one direction.
What is generally defined as morphological tense is marked by
affixes, auxiliaries, particles. English has just two morphological tenses, past
and non-past.

2.3.3.1. Absolute and relative tenses

Some tenses are absolute, i.e. they don’t need to relate to other
time deictics to be interpreted correctly, others are relative, i.e. they depend
on their connection to the previous category of elements. Absolute tenses
(present, Past Tense, Future Tense) are deictic, relative tenses (Present
Perfect, Past Perfect, Future Perfect, Future-in-the-Past) are non-deictic.
The following examples illustrate either a past action implicitly
anterior to present, or actions that follow after the present moment of
speech. They are absolute in that they don’t require a time anchor because
that anchor is implicitly present time. The three examples below illustrate
deictic tenses.

We started our journey that day.

You deserve to get the prize.
They will win.

The perfect tenses in the following examples are linked to their

anchors (past - arrived, present- is tired, future – midnight). The anchors
are expressed grammatically in the first two examples or lexically in the last
one. All the actions in the perfect tenses express anteriority in relation to
the events which are the anchors.

We had started (non-deictic) our journey when you arrived (deictic).

He has worked (non-deictic) hard and is tired (deictic).
They will have won by midnight. (non-deictic)

2.3.3.2. Tense and Aspect

‘The terms Tense, Aspect, and Modality refer to three kinds of

information that are often encoded by verbal morphology. Tense
marking indicates, to varying degrees of precision, the time when an
event occurred or a situation existed. In other words, it specifies the
situation’s “location” in time. Aspect relates to the distribution of an
event over time: is it instantaneous or a long, slow process?;
completed or ongoing? once only or a recurring event? [...]In many
languages, we find that a single affix actually encodes information
from more than one of these domains, e.g. tense and aspect; or
tense and modality. For this reason, many linguists prefer to treat

104
 Tense–Aspect– Modality (TAM) as a single complex category.’
(Kroeger 2005: 147)

Tense and aspect combine in order to fully characterize an event,

i.e. action, or state, considering that an event implies that something
happens, there is a development in the state of affairs, whereas states
imply lack of change.
Aspect refers to the ‘internal organization’ of an event (Bybee 1985)
or state. Events and states are included in the concept of eventuality (Bach)
or predicative situation (Smith).The series of features characterising an
event cover the concepts of continuity, anteriority, completion58/
perfectiveness, result, frequency. Perfective aspect is subdivided into
progressive/continuous and habitual. Tense and aspect are considered
independent categories, allowing combinations between them:

‘a language with three tenses and four aspectual categories could

potentially have twelve distinct tense–aspect combinations. In
practice, however, thereare often restrictions on which combinations
are possible in a particular language’ (Kroeger 2005: 161).

He has been working hard and is tired. [+continuative],

[+completion]/[+perfective], [+resultative] [+anteriority]

I’m waiting for them. [+continuative], [- completion]/[-perfective]

There is disagreement among linguists over classifying the perfect

as tense or aspect, since it shares features of both completion (subsumed
to the aspect) and location relative to some temporal reference point
(subsumed to the concept of tense). Kroeger (2005: 159) quotes Comrie
who identifies four major uses of the perfect: (a) perfect of result (a perfect
verb is used to describe a result state); (b) experiential perfect; (c) perfect
of persistent situation; (d) perfect of recent past. These cases describe the
values of Present and/or Past Perfect. In point of terminology, perfect
should be used to refer to tenses (Present Perfect, Past Perfect, Future
Perfect), whereas perfective is synonymous to completed.
The latest research in the field considered only the opposition
perfective ≠ non-perfective as pertaining to the domain of aspect.

2.3.4. Coding time, event time, reference time. Deixis and grammar

The interpretation of an utterance in point of its time reference

implies considering four basic concepts: Coding Time (the time when the

58 Completion should not be interpreted as synonymous to the completed aspect which is

marked lexically by the aspectual verb finish, marking the end of an action/state.

105
utterance is uttered, i.e. present), Receiving Time (usually identical to
coding time, unless the text is recorded), Event Time (there can be several
events referred to in an utterance), Reference Time (the point of reference,
the absolute time t0 in relation to which all the other events are interpreted
in point of their location on the time axis). The importance of the last
concept results from the fact that it is implicitly part of the definition of tense
as Bybee (1985) puts it: ‘Tense refers to the grammatical expression of the
time of the situation described in the proposition, relative to some other time.’
The system was somehow simplified by preserving the opposition
deictic ≠ non-deictic and the concept of reference/anchor time.
The examples below illustrate how deixis and verb mophology
intertwine.

I’m waiting for them.

Coding time (CT) Present
Receiving time (RC) usually Present
Eventuality time (ET): am waiting deictic tense
Reference time (RT= CT=ET) Present

‘He has been working hard and is tired.’

Coding time (CT) Present
Receiving time (RC) usually Present
Eventuality time (ET): ET1 has been working, non-deictic tense; ET2
is tired, deictic tense
Reference time (RT) is tired (simultaneous to Coding Time; ET2 = RT.
It is normal for one of the events/states described in the utterance to
be the point of reference used to mark linguistically the chronological order
of the events.
Sequence of Tenses applied in case of Indirect Speech is a clear
case of deictic shift from ‘near speaker’ meaning (direct Speech) to ‘away
from speaker’ meaning (Indirect Speech) (Yule 1996: 16):

He said he had been working hard and was tired.

Coding time (CT) Present
Receiving time (RC) usually Present
said Reference Time, t0
had been working ET1 anterior to t0 non-deictic tense
was tired ET2 simultaneous to t0 deictic tense

2.3.5. Lexical time markers

Lexical time markers can be subdivided according to formal or

semantic criteria:

106
- formally, they can be nominal expressions (nouns, numerals) sometimes
preceded by prepositions (prepositional phrases) and functioning as time
modifiers, or adverbs/adverbial phrases having the same syntactic function;
calendrical time markers are non-deictic, but here we intended to consider
them strictly in point of inventory:

See you on Thursday/ next month/ in January/ on the 31-st.

We’ll meet tomorrow. (the future actions described are posterior to the
Coding Time –Present)

The battle took place in 1600 A.D. (the underlined expression is non-
deictic, even if the event described is implicitly anterior to the Coding Time -
Present)

- semantically, they can be either absolute (depending strictly on the origo

specific to a certain calendar) or relative (depending on another time
anchor: for instance, it can be the CT coding time now, i.e. the time of
uttering). We give below some examples of relative time deictics:

See you tomorrow (i.e. the next day from today).

Sometimes, two variants are possible, the latter being more emphatic and
formal, the emphasis being meant to convince the interlocutor that the
timespan is rather short, which could add a positive or a negative
connotation to the whole utterance, depending on the context:

See you the day after tomorrow./ See you two days from now.
See you next month (i.e. the month following this one, that today, the CT,
is part of)/ in May.

In the example above both variants are possible, the latter is more exact,
whereas the former might give the interlocutor the impression that the
timespan is shorter, which can have a positive effect on him/her. The same
happens in I’ll be back soon. The speaker is not informative enough,
maybe because he/she cannot give the precise time when he/she returns,
therefore the focus is on giving the interlocutor the impression that the
timespan referred to will be extremely short in relation to the CT.
If the CT is not clear from the situational context of utterance, such a
notice is not informative enough and can cause various reactions, from
smile to anger, if we find it on a shop window, for instance:

Eng. I’ll be back soon. vs. Rom. Vin imediat. Fr. Je reviendrai
immédiatement.

107
 Calendrical time divisions (proper names designating months and
weekdays) can be used both as absolute and relative time deictics,
whereas time units proper can be used only as relative deictics:

Eng. This Sunday/Next Sunday/ we’re staying at home.vs Rom. Duminica

asta/Duminica viitoare stăm acasă. (relative calendrical deictic)

See you on Sunday, April 12. (absolute calendrical time marker, non-deictic)
We meet on Sunday. (the time adverbial is interpreted either as a relative
deictic meaning ‘this Sunday’ (the closest to CT) or as a non-deictic if it
means ‘every Sunday’.
In Romanian there is no ambiguity because the two meanings are
lexicalized differently, i.e. the noun gets the definite article or not,
depending on the meaning intended: Ne vedem duminică (prima duminică
care urmează)/ duminica (în fiecare duminică).
In English the ambiguity (resolved anyway at the level of the situational
context) can be avoided structurally by adding the grammatical affix –s to
the noun making it acquire a repetitive value and a permanent character:

Eng. We meet on Sundays. vs Rom. Ne întâlnim duminica/ în zilele de

duminică.

The time adverbial can be associated to a past tense verb, in which

case the indefinite value of the adverbial narrows, referring to a past
interval. All the argumentation above remains valid:

We met on Sunday(s).

In the following examples time divisions are used either for precision
(the first example) or as interpersonal pragmatic markers meant to assure
the interlocutor that the delay won’t be long (the second and the third
example). In examples 2 and 3, the perlocutionary effect is that of making
the interlocutor benevolent in relation to the delay, getting his approval etc.
The last two time markers are [-definite], since the speaker cannot possibly
be back in a minute or the interlocutor won’t be asked to wait just a second:

See you in one hour (from now).

I’ ll be back in a minute. vs Rom. Mă întorc într-un minut.
Wait (just) a second! vs Rom. Aşteaptă o secundă! Stai o clipă!

Greetings are a type of relative deictic which clearly depends on the

situational context:

108
 Good morning! is used till midday, Good afternoon! between 12 p.m.
and 6-7 p.m. etc. Other greetings can be considered in our view discourse
or empathetic pragmatic markers:
Good evening! is a relative time deictic but it also marks the
beginning of a possible discussion. Good night! is its counterpart marking
the end of the conversation. The analogy principle was at work in this case
too, and the speakers tend to use a symmetrical greeting, saying Good
evening! both at the beginning and at the end of the encounter. If a
discussion is to follow, or took place, such greetings could be considered
discourse markers.
Good day! is both a discourse and an empathetic deictic, its
features being [+ end of a conversation], [+authority],
[+annoyance]/[+boredom] etc.
An extension of the previous structure is the expression Good day
and good riddance! cf Rom. Drum bun şi cale bătută!, the negative
connotation being the same.

2.3.6. Time vs space deixis

Kroeger (2005: 148) points out the thinking patterns specific to a

number of languages which illustrate that we find it easier to think to the
spatial dimension, since spatial location is someting that we can observe
directly. The next step is to extend the linguistic means of referring to it so
as to include the time dimension. The conclusion would be that time
expressions are metaphorical extensions of space expressions: ‘when we
talk about time reference we often use the vocabulary of spatial location: on
the table vs. on Tuesday; in the house vs. in ten minutes; at school vs. at
midnight; next door vs. next week; plan ahead, think back, etc’.
We provide a series of examples to illustrate the previous idea:

We go back a long time.

They want further explanations!
What do you intend to do further?
See you shortly.

Short basically refers to distance in space but it can be used

metaphorically to refer to distance in time. A series of expressions include it
as a modifier of nouns denoting time: for/ after a short while , in a short
time, to make it short etc. All mean ‘a small amount of time’.
In temporal deixis, the remote/distal form can be used to
communicate not only distance from current time, but also distance from
current reality or facts (Yule 1996: 15). This is the explanation of metaphors
such as: the coming winter, the past week.

‘Our experience of space is normally three-dimensional, with no one

direction having a specially favored status. Time is one-dimensional

109
 and moves in only one direction. Picture yourself traveling down a
one-way street with no turn-offs, and you will have a good spatial
analogy for thinking about tense systems. Another possible analogy,
reflected in the words used to refer to time in some languages, is to
picture yourself sitting on the bank of a river facing downstream. Time
flows past in one direction, like the water of the river. You can “see”
what has flowed past, but not what is flowing toward you (Kroeger
2005: 148).’

2.4. Space deixis

2.4.1. Definition and importance

Space deixis means the encoding of the spatial location of the

elements referred to in the discussion. Its importance derives from the fact
that spatial location is the basic constitutive part of any utterance canonical
deictic dimension (Lyons 1977: 637-638; Levinson 1983: 63). Spacial
information specifies the position of an object in relation to an element
which constitutes its space anchor (Costăchescu 2013: 81). Given the
egocentric organisation of deixis, space reference indicates the location of
the referents designated by various linguistic expressions, always in
relation to the participants in the speech event (i.e. the discussion).
Consequently, space markers are as important as definite descriptions
(nouns preceded by the definite article) or proper names. In other words,
definiteness is an essential deictic notion. Morphologically, this is reflected
by the mutual exclusion between demonstatives and the definite article:

There is a man waiting for you in the lobby.

Meet them in the park/ in that park.
I want the/ this pen.

2.4.2. Classification of space deictics

Space deictics can be classified according to a series of factors,

having to do with their form and role. We attempt to order them rather
traditionally, starting from their form and continuing with their functional
characteristics:

a. they belong to various morphological classes: they are demonstrative

pronouns and adjectives, adverbs and adverbial particles, nouns
sometimes preceded by a preposition (NPs and PPs), verbs of movement:

This does not belong here. (pure deictics)

110
Move on! (i.e. from this state, figuratively from the location we are now;
[+proximal])

Put it away! [+proximal]

You come and go. (deictic) vs They come to London every week. (non-
deictic since the location is not related to that of the speaker; proper names
are by nature non-deictic, unless they become part of addressing terms)

b. pronouns and adverbs share the feature [+/-proximity] i.e. being

[+proximal], or [-proximal/+distal], generally, in relation to the speaker,
therefore they can be considered ‘pure’ deictics; nouns and verbs are rather
‘impure’ since their deictic dimension is not so obvious or pattern-observant:

This matter is not to be discussed here. [+indirect and metaphorical

proximity to the speaker, i.e. to a topic known by/familiar to/shared by the
speaker]; here can refer strictly to the place or to the stage of the discussion.

Take this away (from me). (the Prepositional Phrase is optional and has
strictly an emphatic value, since it is implicit anyway)

Come and fetch me some milk. It’s on the top shelf on the right.
Come implies movement towards the speaker, and fetch movement away
from the speaker; the PP on the right has the feature [-proximity], too– the
right side is defined strictly in relation to the position of the speaker, and to
any person facing the fridge.

The cat’s behind the car. (Levinson 1983: 82)

Behind can refer to the rier part of the car, which remains so irrespective of
the location of a speaker, or can refer to the position of the cat relative to
the car seen as a whole, not as a sum of part, sides etc; depending on that
behind the car can be interpreted deictically or non-deictically, according to
Levinson; we think the most frequent interpretation has in view the direction
of movement.

Most deictic systems are binary, expressing the opposition proximity

to vs. distance from the speaker: here ≠ there. In some other languages,
i.e. Italian or Romanian (at dialectal level) display a tertiary system which
differentiates between proximity to the speaker ≠ proximity to the hearer ≠
distance from them both. Generally, languages tend to simplify the system
of space deictics, operating with a binary opposition. In English, yonder
expresses somew distance from the speaker, hither ‘to this place’ implies
movement toward the speaker, and thence ‘from that place’ implies
movement away from the speaker.

111
c. in point of their autonomy, space markers can be absolute or relative,
similarly to time markers. Relative ones are deictic and imply the reference
to the location of the speaker, whereas the others express a space location
whose point of reference is generally familiar to the community of users,
thanks to their general knowledge. Absolute spatial relations express either
direction or distance from the point of reference (the Equator, the
Greenwich Meridian). Cardinal points, the geographical position of any
place on the planet, according to the latitude and longitude corresponding
to it are elements marking space in an absolute way:

They live in the south.

The city is located at 45 degrees east longitude.

Direction is expressed in the example:

You come and go. vs. They come to London every week. (the speaker is
in London, otherwise he would say They go to London.)
In the first example the verbs are deictics: come expresses [-proximity] and
go [+proximity]). These features are intrinsic to the verbs, therefore this
type of reference is relative intrinsic reference. The meaning of the
utterance is ‘You come and go from me/ my place/ my location.’ In the
second example come preserves its feature but it is combined with the
absolute space marker to London, defined as non-deictic.

Distance from the point of reference is expressed in the examples:

They live fifty miles from here. (deictic expressing the implicit reference to
the location of the speaker)
The bag is behind me, on the table. (the absolute location expressed by
the NP on the table is completed with the explicit linguistic marking of the
relation between that location and the speaker’s location; this what makes
the NP on the table acquire the feature [-proximity].

d. space deictics are generally subdivided into gestural/ostensive and

symbolic. The second types is maybe less obvious but much more frequent
because the situations of communication require such uses.

That pen is not working. (ostensive use, the uttering is accompanied by the
gesture of pointing towards the object)
This atmosphere is fascinating. (symbolic use)

2.4.3. Main values

Most values have already been mentioned throughout the previous

subchapter, therefore we will just summarize them. The major opposition is

112
between deictic (ostensive or symbolic) and anaphoric values, though they
can coexist in some contexts.

Relative deictic use

‘I want that’, she said pointing at the doll. (deictic ostensive value)
This week is great! (deictic symbolic value, since not the seven day time
span is referred to, but the events taking place etc)

Around the corner there is a gas station. (deictic reference, since it is

obvious that the speaker refers to a place far from him/her; non-anaphoric
reference since there is not a previous mention of the PP around the corner
in the linguistic context, the interlocutors relying on the situational context to
interpret the structure)

Absolute/non-deictic use

Look at this side of the tree! (Levinson 1983: 82) (intrinsic non-deictic
spatial orientation)

At the corner of 53rd Street and 10th Avenue there is a gas station.

They visit London frequently.

According to Levinson and not only, such an example illustrates a non-deictic
use, if deixis is restrictied to the deictic center represented by the speaker.
However, we consider that, broadening the view in point of the
interpretation, London can implicitly express either [+proximity] to the
speaker, if the speaker is in London, the meaning of the utterance being
‘they come here frequently’ or [-proximity] if the speaker is not in London.

Deictic + anaphoric value

A (handling the album to the hearer): I wanted you to have the album
because this album is everything to me. (ostensive deictic and anaphoric
value)
They have bought several roses: these ones. (ostensive deictic value and
anaphoric value
We were born in Paris and have lived here/there ever since. (anaphoric
value and symbolic deictic value; the speaker is away from London)
I called him at the office but he wasn’t there. (in this case, the point of
reference is the home-base, i.e. the office)

Following Levinson’s examples (1983: 67) it is obvious that deixis and

anaphora are not always mutually exclusive.

113
2.4.4. Combined values

In many cases, space deictics can cumulate two or even three

values which merge:
- place + empathetic deictic:

Move on! I’m fed up with it.

- social+ empathetic deictic (Levinson considers this example a non-deictic

occurrance in point of place deixis:

I met this weird guy. (Levinson 1983: 66)

this [-definite], [-pleasant]; combined this and weird enhance the negative
connotation.

I did this and that.

this and that (Levinson 1983: 66) [-definite], [+informality], [+annoyance at
the interlocutor because of his/her curiosity]/ [+neutral attitude]. The
structure becomes an idiomatic expression and only in that case the above
interpretations are appropriate. I did this and that can be interpreted as a
deictic use of the demonstratives if an ostensive use is implied (the speaker
points at two referents), at the same time, the empathetic use is added, the
speaker being either proud of or disappointed at his/her accomplishment.
The structure can be used as a ostensive deicitc if the speaker points
towards two objects etc.

2.4.5. On time and space deixis (again). Which was first?

That question cannot be answered since both time and place

deictics can replace each other metaphorically in various contexts, taking
over each other’s values. The use of the pure place deictics this and that as
time deictics, and also that of time modifiers expressed by NPs in the same
way (but adding the time adverb ago) as in

See you later this week.

The shop is ten minutes from here.
You should have got off ten miles ago/ two bus stop ago.
I’ll see you shortly/ in a short time.

makes Lyons (1977: 609) take sides in favour of localism, a theory which
reduces non-spacial to spacial expressions, therefore, considering the latter
more basic. Actually Lyons considered the argument that demonstratives
can be used as discourse markers, making reference to a previous or
following part of discourse. On the other hand, Levinson (1983: 85) argues

114
that place deixis always incorporates a covert time deictic, while the
converse is not true, i.e. mentioning the time related to an event does not
necessarily imply that the interlocutors will share the knowledge about the
place of that event; in conclusion, time deictics seem more basic:

We met at school. (implicitly when we were there)

In conclusion, time deictics might seem to some more basic, even if

at an implicit level, whereas space deictics seem more easily perceptible by
users because of the specificity of space dimensions.

2.5. Social deixis

Social deixis comprises all the linguistic elements which mark the
relationships which are being built among the participants in the verbal
exchange constituted as a speech event. Social deixis is concerned with
direct or oblique reference to the social status and the role of those
participants in the verbal exchange. They cannot be analysed without
considering the notion of politeness as a linguistic phenomenon, i.e. all the
aspects of the discourse which are rule-determined and are meant to
preserve the harmonious character of the interpersonal relationships.
Consequently, it is equally important to discuss the type of distance,
horiyzontal and vertical, established between the interlocutors and the
means by which politeness functions.
The distance characterising the verbal exchange among
interlocutors depends mainly on:
a. the number of participants in the verbal exchange;
b. individual characteristics (age, sex, health state etc);
c. mutual relations, which in their turn, depend on the level of
knowledge (stranger, acquaintance, friends, relative etc) and also on the
type of relation developed (professional, personal, i.e. a family relation or a
relation based on mutual interests and affection).
a. The number of participants should be correlated with their
typology to explain the linguistic choices in point of social deictics. Goffman
(1967) mentions the active and passive participants in a conversation, their
roles not being obligatorily unchangeable, but definitely influencing the
linguistic behaviour of the speaker at any moment in the course of the
verbal exchange:
- interlocutors: they are active participants, the speaker having the feature
[+/- source] and the hearer being the direct or indirect receiver, i.e. having
the feature [+/-addressee];
- bystanders: passive participants who can become active at any moment if
they choose to, since their presence is obvious and acknowledged at the
moment of the verbal exchange;

115
- overhearers can also become active participants unless they transgress
politeness norms, since they are not hiding their presence at the place of
the discussion;
- eavesdroppersare not legally present in the situation of communication
and, consequently, they are not making their presence known to the others
and they don’t normally become active participants.

They never stop plotting, some people.

Uttered by the speaker in front of the hearers who are the receivers,
but not necessarily the addressees, and meant for bystanders or
overhearers, such an utterance might trigger the verbal reaction of the
addressee(s).

Look at her! (the person can hear and is meant to hear the utterance)
Somebody didn’t clean up after himself. (Yule 1996: 11)

b. The individual characteristics of the interlocutors, i.e. sex, age,

status, determine the speaker to adopt a certain attitude towards the hearer
and that results in the linguistic choices made in order to express respect,
familiarity, superiority, etc.
c. A horizontal relationship means that the interlocutors are
relatively equal. This is reflected in the addressing formulas (first name use,
polite reference to the sex and status of the person – Mr. Brown, Miss
Smith, Mrs. Johnson −, reference to the professional status and the
surname – Dr. Johnson, Professor Brown –), in turn-taking (none of the
participants dominates the conversation). It is essential to mention the fact
that these quasi-equalitarian relationships do not automatically presuppose
informality, but can be associated with various levels of politeness.
In many contemporary societies, what we call interpersonal
distance is not set once and for all: its boundaries are variable and,
frequently, within a conversation the tendency might be towards a
progressively increasing level of informality.
The equal status of the interlocutors associated with informality may
result in conversational overlaps, when one participant in the conversation
does not wait for his/her turn and speaks before the speaker has finished
his/her utterance. The interlocutors may share the same points of view and
they interpret correctly the deictic elements used by them both, and they
make the correct inferrences, or, on the contrary, they might jump to
conclusions, and interpret each other’s sayings inappropriately in point of
deictic markers, inferences or communicative goals.
On the other hand, vertical relationships are based on a position of
power, of authority, hold by one of the participants who might tend to
dominate the verbal exchange. Such relationships are usually hierarchical

116
and assymetrical, since one participant adopts a [-distant] attitude and an
adequate linguistic behaviour, while the other(s) are distant, polite. They
allow interruptions and overlaps because of thier inferior status, and also
conversational gaps when they don’t dare to express their ideas, to
disagree or to change the topic of discussion etc.
Proximity is associated with somebody placed lower on the social
scale or in a social hierarchy, younger, or less powerful, and that is
linguistically encoded in the form of informal, familiar style structures. In
opposition, distance is associated to a high position, to an older age and to
more power (Yule 1996: 10-11). In definig what he calls the Principle of
Politeness, Leech (1983) mentions social distance and power/authority as
the the extra-linguistic variables, which, together with interferences in the
speech act, result from our simultaneous wish for autonomy and approval.
Conversational principles to be followed for good communication
can be both general and culture-specific. Also, they are the result of a
systematic acquisition; we involuntarily acquire them in social practice
unless we fail on a certain occasion. Then learning starts. Not only
politeness strategies may be different, but also the linguistic means that a
language can use in order to expresss various relationships: personal
pronouns, demonstratives, addressing terms, verb agreement.
As far as addressing terms are involved, English uses nominal phrases
that cover the range of attitudes from extreme politeness to the total lack of
politeness:
- honorifics: Your Majesty, Your Highness,Your Excellency, Your/His
/Her Grace, Mr. President, Mr. Brown, Dr. Smith etc; indirect address can
be made by using third person reference: We invite Her Majesty the
Queen59 to share Her impressions...;
- kinship terms, sometimes in variants which connote affection,
becoming also terms of endearment: Mother,Uncle, Auntie, Sis60etc;
- terms of endearment: dear, darling, Billy, love etc;
- insults: you fool etc.
Goffman introduced for the first time in linguistics a notion used
initially in sociological research, face. It refers to the self-image that we all
want to be taken into account in verbal/non-verbal interaction. We can lose
face, maintain it or enhance it, all the result of a constant concern. Face, our
public image, has a dual nature (Brown & Levinson 1987). It is what Goffman
called Politeness Conflict: we manifest a positive face, including the desire

59 A case of acknowledging the monarch as a country’s ruler is represented by the English

referential expressions designating the official name of a British ship/submarine: HMS
Queen Elizabeth.
60 The addressing term is also used to denote equal status and, implicitly, solidarity in point

of social or ethnic group/community cf Rom ‘frate’ ‘soro’, and not a sibling relationship: Lasă-mă,
frate! Ce vorbeşti, soro?

117
that our self-image should be appreciated by the interlocutors and a negative
face which means acting according to our intentions. Goffman mentions that
if the strategies used to keep a positive face are explicit and include keeping
distance to gain appreciation and approval, those specific to the negative
face mean apparent integration and are implicit.
Polite addressing terms used below are examples of positive face,
whereas familiar ones are negative face markers:

George/ Johnson, you are expected to finish in time. Do you think you can
do that?
Mr. President, we are delighted to have you here.
Professor Brown, can you give us some details about your future plans?
Mr. Ambassador, you have repeatedly expressed your opinion on this
matter.
Lady, mind your step!

Addressing terms become deictic in nature whenever they are used

for direct communication and the referent is [+definite] in every concrete
context: we address the person in office, unless, out of politeness, we
preserve the addressing structure when the referent is out of office (for
instance, in case of former country presidents, former ministers etc).
Successful reference means that an intention was recognized via
inference61, indicating a kind of shared knowledge and, hence, social
connection (Yule 1996: 24). Similarly, a successful interpretation of an
addressing term by the interlocutor indicates shared knowledge and
(non)acknowledgement of the social relationship established.
For instance, he last example Lady, mind your step! can be interpreted as
corresponding to various levels of politeness, cf Rom. translations:

Doamnă, fiţi atentă pe unde mergeţi!

Fii, madam, atentă, pe unde mergi!
Uită-te, cucoană, pe unde mergi!

2.6. Discourse/ textual deixis

Another name for it is textual deixis, which brings us to the discussion

about the concepts denoted by the terms discourse and text, regarding the
use of the former when referring to speech, and of the latter when referring to
the wrriten mode, or, the use of discourse as a superordinate term including
both speech and writing. No matter what position we adopt, whether we
consider the oral or written form of a language, it remains true that, in many

61 a pragmatic deduction.

118
situations, the speaker finds it necessary to make reference to a whole part
of a discourse /text and, thus, he/she points to that part. Therefore, discourse
markers are deictic elements, since they express the emphasis placed on
them by the source.
At the same time, they help creating and preserving the coherence
of the text, i.e the logical ordering and progress of the ideas expressed.
Discourse/textual deictics are related to the deictic center in point of
the discourse topic dimension, i.e. the speaker links every utterance to the
topic under discussion.
In point of their nature, such deictics cannot be specific, ‘pure’,
because this type of deixis is not a basic one, therefore, its devices won’t
be either: the job of making reference to a certain part of a context is
performed by personal and demonstrative pronouns, by adverbial phrases
and prepositional phrases:

In the following, we willl refer to....

Next, we will be visiting another interesting sight.
The author enlarges upon that aspect in chapter two.

Technically, pronouns will send us back or forth to the part of the

discourse/text which is brought back to the interlocutor’s attention. They
ensure the coherence of the whole discourse/text, by avoiding a repetition
which would be in fact, almost impossible if one intends to keep
communication concise and clear:

[They came to us to ask for our help]S and the fact [that they came to us to
ask for our help]S forces us to act responsibly.
[They came to us to ask for our help]S and thatS forces us to act
responsibly.

Typically, they will have an anticipatory value and their function will be
cataphoric: it62 and this are such instances, even if plural forms are possible, too:

It was incredible that everybody agreed on the new plan.

It was Ann who asked for you.
This was our reason for leaving: the hotel was fully booked and no spare
room could be found in the neighbourhood.
These are our arguments:the hotel was full, there was no vacancy, we
were too tired to remain there.

62 For a detailed account of the values of it, see Pisoschi (2012: 158-160).

119
 It acts as the formal subject of a complex sentence whose logical
subject was extraposed, in order to topicalize it, since it carries relevant
information. Cleft constructions are included in this category.
Demonstratives behave similarly, in that they need a clear specification of
their reference in the form of a subject clause.
In other cases, the pronoun anticipates a direct object after
cognition verbs (see Pisoschi 2010: 86):

We found it hard to believe such a story.

Accordingly, some grammarians consider it to be a semantically void

category, and treat it simply as a grammatical morpheme filling a thematic
role, and not as a pronoun in its own right (Fromkin &Rodman 1998: 200). A
discussion of Bolinger’s view on the deictic value of it is in Pisoschi (2010:
86-87).
The personal pronoun it can anticipate a finite or a non-finite subject
clause and Larreya (1993) points out the cases of subject repetition for
emphasis and enumerates the basic possible patterns: it anticipating a
gerund and an infinitive (ibidem):

It’s nice, sitting around and talking.

It was nice meeting you.
It was nice to meet you.

Pronouns as discourse/textual deictics can also have an anaphoric

value, when they resume the reference made by means of a whole
sentence/clause; in English it and that share the resuming role and
anaphoric function typically, even if the plural demonstrative those is
possible:

They left abruptly. It/ That annoyed everybody.

He says he saw you. I don’t believe it. cf Rom. Spune că te-a văzut. Eu nu
cred (asta)/Eu n-o cred. (Pisoschi 2010: 88)
It was late and the hotel was full. Those were our arguments.

2.7. Empathetic deixis

Empathetic deixis, as it is apparent from its name, is a category

referring to the attitude of the speaker in showing his/her empathy, i.e.
sharing the state of mind of the interlocutor. There are no pure empathetic
deictics, personal pronouns and demonstratives being used to convey the
intended attitude. Some linguists consider that we are dealing with indirect
anaphora cases, since the basic value of pronouns is deviated from, being
performed a metaphorical extension.

120
 If for personal pronouns the positive or negative attitude is to be
inferred from the context, with demonstratives a metaphorical extension of
their characteristics is made: the feature [+proximity] connotes positive
attitudes, whereas [-proximity] connotes negative states and feelings:

That’s it! I don’t want to hear anything more. (negative connotation)

This is the book! (positive connotation)

To Chaika (2000: 27) that refers to a less immediate circumstance;

that does not imply focus on it, the event referred to may be over, whereas
this implies reference to the deictic center - here and now:

That bothers me. vs This bothers me.

Chimombo and Roseberry (1998), too, show that demonstrative

pronouns can indicate emotional closeness or distance. Their example was

This was a very naughty thing to do.

uttered while telling a story to a child. In this case, emotional closeness is

associated to a negative connotation, the proximal demonstrative is meant
to draw the hearer’s attention. Psychologically, this implies emotional
closeness, how vivid something still is in the speaker’s mind. Haiman
(1998) considers that language may be alienated from the emotions that
produced it: people control their emotions enough to be able to use
language in that context, and, by doing so, they both express and describe
those emotions. Thus, in his opinion, they are no longer, primarily,
participants, but observers and exorcists of their emotions.
In some idiomatic expressions, the demonstratives are markers of
opposite empathetic values (see also Pisoschi 2012: 186-187):

a. That’s it! Enough is enough! (negative connotation)

That’s it! I’m done! (positive connotation)
b. This is too much!
This I really want on my birthday.
c. You are leaving tomorrow and that’s that.

For examples of the merged values of linguistic markers, see 4.4.

Conclusions

Deictic elements are indispensable in establishing the ‘coordinates’

of an utterance; only in this way can the utterance make sense; small in
size, highly frequent, and having a wide range of possible uses, deictic

121
expressions always communicate more than is said (Yule 1996: 15);
specific and non-specific deictics are to be used and interpreted within the
linguistic and situational context; when analyzing person deixis markers,
reference (variable by definition) and role in communication (stable in its
basic coordinates) are of paramount importance. As social deixis markers,
personal pronouns are decoded in relation with other nominal structures,
expressing a certain level of politeness. Time deictics take into account t0,
the speech time, while space deictics set up the spatial position of various
objects according to the position of the speaker/ addressee. Both time
deictics and space deictics can be relative or absolute in relation to the
space and time deictic center, generally presupposing each other and, in
most cases, behaving as each other’s complements or, frequently,
replacing each other metaphorically. Empathetic deictics associate
reference with a certain attitude of the speaker and interpret its
proximity/distance metaphorically.

122
 3. CONVERSATIONAL IMPLICATURES

3.1. Logic of Conversation. To say vs to imply. From logical

connectors to Cooperative Principle
3.2. Classification of implicatures
3.3. Conversational implicatures. The Cooperative Principle and
Conversational Maxims
3.4. Domains of manifestation of conversational maxims
3.5. Types of non-observance of conversational maxims
3.6. Tests for conversational implicatures. Drawbacks
Conclusions

3.1. Logic of Conversation. To say vs to imply. From logical

connectors to Cooperative Principle

In his lectures delivered at Harvard in 1967, Paul Grice aimed at

proving that human action and rationality result in a system of principles
which determine, in their turn, a system of conversational principles that
account for the differences in the meaning of logical connectors (and, or, if
…then) in logic and in ordinary language.
Logical connectors analysis offered the starting point in extending
the area of linguistic analysis. Conversation analysis shows that the speaker
often means more that he (explicitly) says. Token (literal) meaning, i.e. the
denotative meaning of the linguistic elements making up an utterance,
together with speaker’s meaning (a term introduced by Grice to refer to the
speaker’s communicative goal), constitutes the unity represented by the
utterance meaning. It is not an abstract meaning, but a meaning depending
on a particular context and directed towards a certain listener.

I’m tired is an example which can be associated to many speaker’s

meanings, but in a conversation, only one of them will be appropriate:
a. as an answer to the question How are you? is an assertion meant to
inform;
b. as a reply to a request is a refusal ‘I can’t do what you ask (because I am
tired);
c. uttered by the speaker while sitting down is an excuse ‘(let me sit
because) I am tired’ etc.
Describing the state is the basic meaning, the literal meaning, but in
different contexts the description of the state is the reason for denying,
refusing, excusing oneself etc.

123
 In everyday communication, interlocutors are rarely very direct and
explicit so that their communicative goal is fully understood. Such a
strategy would be perceived as impolite and chances are that the
interlocutor would have an opposite reaction to the expected one. Indirect
meaning is the choice preferred by speakers in many situations, the
interlocutor having to make a pragmatic inference.
The distinction to be made is between what is said and what is
implied, suggested, meant. A meaning which is conveyed indirectly,
distinctly from what is said directly, is called an implicatum. What an
utterance conveys in context falls into two parts: what is said (the logical-
semantic explicit content) and the implicatures (the cognitive, intended
meaning). (for more on types of meaning, see Part I, 3.1.)
Sometimes, the intended and the explicit meaning coincide (‘How
are you?’ ‘ I’m tired.’)- this is what Grice called natural meaning. But, in
most cases, the two types of meaning don’t coincide and Grice referred to
this situation as being characterized by non-natural meaning: (I can’t do
what you ask because) I am tired; Can you open the window? meaning
‘I require you to open the window.’
Grice’s example is: A and B are talking about a mutual friend C, now
working in a bank.

A: How is C getting along in his job?

B: Oh, quite well, I think he likes his colleagues, and he hasn’t been to
prison yet.

Implicatures: C is likely to yield to temptation; C’s colleagues are

unpleasant and treacherous, or, on the contrary, C is a bad person and his
colleagues are all right, etc.
Any utterance is to be analysed as a means of conveying two types
of information:
- a set of propositions deducible from the sentence uttered by linguistic, i.e.
semantic, rules;
- a further set of propositions deducible from the sentence uttered, together
with some items of non-linguistic knowledge shared by the speaker and the
hearer, plus a set of shared inference rules (Smith & Wilson 1990: 174)
In order for the hearer to infer the right (i.e. relevant) implicatures
the speaker intended him to, the interlocutors must be cooperative at every
moment during their verbal exchange63. Any implied meaning can be
misunderstood by the hearer, if the two don’t have the adequate types of
competence and don’t share a common ground (the necessary types of
information able to facilitate communication, implicature inference
included). The hearer is able to hypothesise about the speaker’s meaning,

63 Our point of interest is verbal communication, even though The Cooperative Principle is
the necessary condition of any type of human interaction.

124
based on the literal meaning of the sentence uttered, on contextual
assumptions and, on general communicative principles.
Grice expressed that in the form of the Cooperative Principle:

Make your conversational contribution such as is required, at the stage at

which it occurs, by the accepted purpose or direction of the talk exchange
in which you are engaged.

This is a principle, not a rule, ‘a convention of communicative

practice’ without which there would be no communication; of course, the
user can choose not to give a reply, i.e. to be non-cooperative. The
Cooperative Principle is also a norm against which violations (lying is a
case of violation) or deviations (exaggerations) can be measured (Finch
2003: 157). Without the underlying cooperative convention there would be
no way of registering deviations, and, also, the general manifestations of an
‘unspoken pact to cooperate in communicating’ so as to understand and be
understood (Finch 2003: 159; 157).
In conclusion, communication, and pragmatics itself, as a science
which studies the process of communication, can be correlated with the
process of inferring, since to understand an utterance means to produce
inferences, i.e. to appropriate the interpretation of the utterance to the
context of communication (basically, to the participants, time and place, but
not, necessarily, only to them). According to Levinson (1983), context is
everything that can trigger inferences.
On the other hand, Armengaud (1985: 60 in Dragoş 2000: 23) refers
to four types of context64, including presuppositional context consisting of
all the presuppositions of the interlocutors - their beliefs, expectations,
intentions. It represents a common ground, either preexistent to the verbal
exchange, or in making throughout it.

3.2. Classification of implicatures

Grice distinguishes two types of implicatures: conventional and

conversational.
The former are much more dependent on the linguistic structure of
the utterances, being more easily inferred due to the presence of certain
lexical elements: adverbs, conjunctions, and syntactic constructions such as
still, moreover, however, although, so, therefore etc. Being related to the

64 The types of contexts are: the circumstantial, factual, existential or referential context –
including the identity of the interlocutors, the time and place, i.e. indexicals; the situational or
paradigmatic context, culturally shared – it includes the situation, the purposes associated to it
and the sense mutually acknowledged by the interlocutors; the interactional context consisting
of the speech acts organized within a coherent discourse; the presuppositional context,
explained above. (for more on context see 1.4.)

125
words used, they are also related to the conventional meanings expressed
by the latter.
Conventional implicatures do not contribute to the truth conditions of
the sentences, but have a constant component, i.e. they convey the same
extra meaning regardless of contexts.
They don’t even need to appear in a conversation, a single reply is
enough:

Some people believe in God. (the word some means ‘an indefinite number
of’, so, it is partially synonymous to not all. The implicature would be Not all
people believe in God. Depending on the context, an implicature could also
be I am among those, even if you aren’t.

Drink responsibly. The additional premises are: most people drink without
considering their health, drinking alcohol heavily affects health. The
implicature is: don’t ruin your health by drinking in excess.

Conventional implicatures are always lexicalized, i.e. associated to

specific words, the additional conveyed meaning requiring no calculation.
Yule (1996: 45) provides examples with even and yet as implicature
triggers. The adverb even is also a trigger of presupposition and Yule
mentions the fact that this type of implicatures are not unlike lexical
presuppositions:

Even John came. The implicature is that the fact was contrary to our
expectations and, according to the speaker(s) beliefs, John was the less
likely to come; the presupposition is someone came.
He didn’t come yet. The implicature is that the fact is contrary to our
expectations, he should have been here by now.
In the same line, the conjunction and can be explained as meaning
either ’in addition, plus’ or ‘and then’:

She entered the room turned on the light.

Moreover, this type of implicatures are not cancelled by the

following linguistic context, and are detachable, i.e. they are not preserved
in case of using synonyms in the original utterance.

John is an Englishman, therefore he is brave.

Premise 1: John is an Englishman.
Premise 2: Englishmen are brave.
--------------------------------------------------------------------------------------------
Conclusion (implicature): John is brave.

126
 We exemplify below non-cancellability and detachability:

John is an Englishman, therefore he is brave, even if he might be

overwhelmed by the circumstances.(non-cancellable implicature – the
implicature John is brave is maintained, even if we add another premise -
he might be overwhelmed by the circumstances)
? John is an Englishman, and, as a result of that he is brave. (the
implicature is detachable, if a synonym is used. The conjunction would imply
a logical relationship cause-effect, which is not confirmed in this case.)

He is poor but honest.

Premise 1: He is poor. (premise to be corroborated with the presupposition
poverty is valued negatively and poor people try to do anything to improve
their situation)
Premise 2: He is honest. (premise to be corroborated with the
presupposition honesty is positively valued.
--------------------------------------------------------------------------------------------
Conclusion (implicature): Honesty is a virtue, more important than richness,
and he is among the few poor people who are not willing to do anything to
become rich, therefore he is a worthy man.
But adds the contrast dimension between the two adjectives; this is what
differentiates the example from the variant He is poor and honest.

Yule (1996) comments on such examples stating that the

interpretation of any utterance of the type p but q will be based on the
conjunction p & q , plus an implicature of ‘contrast’ between the information
on p and that in q.

John, my colleague, and, nevertheless, my friend (cf Armengaud 1985: 71

in Dragoş 2000: 26)
The lexical marker of implicature is the adverb nevertheless conveying the
idea that the feature [+colleague] does not necessarily include the feature
[+friend]. We add that the original relationship concerning the two linguistic
elements between which a relation of reciprocal presupposition is established
could have been initially in relation of opposition, i.e. of mutual exclusion:

John is my rival, and, still, my friend.

Conversational implicatures rely on objective elements (the context)

and on subjective factors (the degree of cooperation between the
interlocutors). They follow from sentence meaning, but considering the
background of the suitable circumstances (including knowledge and

127
beliefs), which give relevance to the utterance. Smith &Wilson (1990: 179)
define conversational implicatures as

‘pragmatic implications65 which follow from a remark only on the

assumption that it was intended as relevant, together with additional
premises which do not form part of shared knowledge, but which the
speaker expects the hearer to construct for himself’.

The additional premises themselves will become part of the

pragmatic implications of the remark.
In their turn, conversational implicatures can be generalized or
particularized (Yule, 1996). The latter are the more frequent, therefore they
are usually simply called conversational implicatures. They can be inferred
on the basis of a single reply:

I’m a woman.
Uttered by a male professor in front of his students, or by an actor in
front of his public, the additional premise is let’s imagine that I’m a female
character… . The implicature, exploiting the apparent false assertion, is that
he thinks and feels like a woman, for the sake of demonstration (assuming
the qualities and flaws typically associated to women, considering a social
and cultural feminine prototype); the sexual orientation of the
professor/actor is irrelevant, unless some incident occurred and he
declares publicly that he is psychologically a woman.

You are here now.

The additional premises that can be inferred are: in the past, maybe
when I needed you, you weren’t here, but you are here now. In this case,
the implicature is that present is more important than past, we start anew,
let bygones be bygones. This could be considered a generalized
conversational implicature.

Generalized conversational implicatures don’t require any special

knowledge in the context to calculate the additional conveyed meaning.
Yule (1996: 40) gives examples which:
-at first sight, seem to illustrate the lack of sufficient information:

‘Did you invite Bella and Cathy?’ ‘I invited Bella.’

The implicature is I didn’t invite Cathy.

65 The quotation contradicts Grice’s view; he created the term implicature, precisely to distinguish

them from an implicatum; an implicatum p  q is always true when q is true, irrespective of the
truth value of p; what is called an implication should be called a possible inference.

128
-oppose the indefinite article to the definite article or possessive adjectives,
in terms of identifying the referent in the context:

I was in a garden one day. A child looked over the fence.

Yule (1996: 40) explains that a garden, a child exclude the variant
my garden, my child. Still, if the speaker intends to indirectly remind the
interlocutor of their first encounter when they were kids, then maybe the
structures under discussion could be interpreted as a garden = ‘a known
garden (yours etc)’, a child = ‘you’. Of course, our counterargument is just
an illustration of the variety, complexity and indeterminacy of utterance
interpretations.
Scalar implicatures can be included into the same category of
examples: some is more than a little, or less than all, possible is less than
probable, cool is less than freezing:

I have borrowed some money (not all the sum we need). vs I have
borrowed all the money we need.

3.3. Conversational implicatures. The Cooperative Principle and

Conversational Maxims

Grice formulates certain specific maxims of conversation, referring

to the quantity, and quality of information, to the coherence of the dialogue
and to the manner of expressing oneself to get the results expected.
In point of their characteristics, conversational maxims are not
arbitrary conventions, but describe rational means for conducting
cooperative verbal exchanges and, more generally, relational human
behaviour. It is to be expected that they would govern non-linguistic
behaviour, too, and, indeed, they do so: if two people are fixing a car, and
one of them requires a certain tool and is given another, or a number of
spare parts and is given a different number, or after a while, all these are
cases of non compliance with the maxims. If a wrong tool is handled, then
the maxim of quality is not observed, because the tool is not the one
required; if the number of spare parts is wrong, then the maxim regarding
the quantity is not observed.
Implicatures can be established by envisaging the four
conversational rules or ‘Maxims’ complementing the CP:

I. Maxim of Quantity
1. Make your contribution as informative as is required for the
current purposes of the exchange.
2. Do not make your contribution more informative than is required.

129
 When speaking to someone, one feels obliged to give enough
information so as the interlocutor should understand the message. At the
same time, one should avoid providing too much information and obscure
the intended meaning. Setting the boundary between too little and too
much is a measure of our communicative competence. This type of
competence is not innate, it must be learned. Once we learn it, as adults,
we also learn how to manipulate this convention about quantity to our
advantage. Sometimes, we also think not to retain one’s attention for too
long. We are ‘economical with the truth’ (Finch 2003: 158).

II. Maxim of Quality

Try to make your contribution one that is true.
1. Do not say what you believe to be false.
2. Do not say that for which you lack adequate evidence.

Finch (2003) considers this maxim as a supermaxim, all the others

being logically subsumed to it. ‘White lies’ are an exception, in that they are
precisely meant to preserve cooperation, politeness ensuring the
willingness of the interlocutor to remain a participant in the verbal
exchange. This position is opposed to that of Sperber and Wilson, among
others, who consider relevance maxim as the supermaxim. If being
cooperative is the most important thing in communication, politeness is a
way of ensuring cooperativeness and, sometimes, out of politeness, the
speaker avoids the truth; then truthfulness ranges the second after
relevance.

III. Maxim of Relation

Be relevant.

Relevance is to be linked to the topic of the conversational

exchange. Changing the subject abruptly or divagating can mean rudeness
or uncooperativeness. Relevance takes many forms, the users trying to
make sense of any apparently ‘odd’ reply in relation to the verbal exchange.
Sperber and Wilson (1986) considered this maxim as being the
most important, subsuming all the others, since the quantity, the quality of
the information and the manner of organizing it are meant to make the reply
relevant in the situational context, directly or indirectly.
Relevance plays a very important role in pragmatic interpretation.
Remarks can be of two types: informative, adding new information to the
stock of shared knowledge, assumptions and beliefs, and uninformative,
which convey a meaning that is not directly deducible from their structure.
Two remarks are relevant to each other, if, combined, they convey new
information, not derivable from them separately:

‘A remark P is relevant to another remark Q if (and only if) P and Q,

together with background knowledge, yield new information not

130
 derivable from either P or Q, together with background knowledge’
(Smith &Wilson 1990: 177).

Let’s consider the example:

A: ‘If you drive, I’ll join you’.

B: ‘I’ll drive.’

The implicature is A will join B.

There is considerable debate about the role played by shared
knowledge, on the ne hand, and inferences, on the other hand, in the
process of turning from an overt message reply to a tacit message one.
Smith and Wilson (1990) consider that judgements of relevance are
essential, besides the systems of knowledge and inference. The
interpretation of an utterance will depend on the hearer’s judgement about
how relevant it was intended to be:
a. If so, the hearer will make the necessary efforts to interpret it as
such, even if this means reading into it a certain amount of information
which it does not overtly convey (Smith &Wilson 1990: 175):

A: Where is the chocolate?

B:I was feeling hungry. [so I ate it.]/ The children saw it [and ate it].

b. if the hearer decides that the utterance was intended as

irrelevant, he may still gather some relevant information from it (Smith
&Wilson 1990: 175), as long as it is seen as directed at the basic remark:

A: Where is the chocolate?

B: I’ve got a train to catch. →’I don’t have time to talk about such
trifles.’/’I’m guilty and I prefer to run away.’

c. the third possibility is that A’s utterance is accidentally irrelevant:

the first utterance was misheard, objectively irrelevant, unworthy of
consideration etc (Smith &Wilson 1990: 176). We interpret Smith and
Wilson’s statement as meaning that speaker A’s utterance was accidentally
irrelevant in relation to the expectations of speaker B (but it was relevant
from A’s point of view). Consequently, B changes the subject, introducing
new information unconnected to the previous remark of A:

A: Where is the chocolate?

B: I’ve got a train to catch. (→I need to focus on that. Help me prepare!)
The children saw it, (→Maybe we should be more careful with the kids,
they shouldn’t be allowed to enter your room.)

131
 IV. Maxim of Manner: Be perspicuous (‘comprehensible’, ‘clear’,
‘unambiguous’)
1. Avoid obscurity of expression.
2. Avoid ambiguity.
3. Be brief (avoid unnecessary prolixity).
4. Be orderly.

An orderly manner of organizing information can be seen as a

condition for its assimilation by the listener. Nevertheless, flouting it (Finch
uses the term violation66 even if it is clear that he refers to cases of flouting)
is a way ‘in which strength of feeling is communicated’ (Finch 2003: 159),
and not only in literature.
Horn (1972) simplifies CP and its maxims, reducing them to two: the
Quantity Principle and the Relation Principle. The Q(uantity) Principle – say
as much as you can; we believe can is to be interpreted both objectively, in
point of the ability and possibility, and subjectively, in point of permission,
according to the norms of politeness- say as much as you can afford.
The example of Horn quoted by Mey (1993: 79)

I cut the finger yesterday.

illustrates the speaker’s truthfulness, but, also, tact and consideration for
the interlocutor, by making the utterance informative enough, but not as
informative in point of scaring details as to worry the hearer.
Referential expressions, definite or indefinite, also illustrate the
intention of the speaker to be polite by being accurate about the identity of
the referents, marking the distinction [+definite] ≠ [-definite] reference: (the
finger = my finger, a woman ≠ his woman):

I cut the finger yesterday.

John met a woman last week.

We consider that Horn adopts implicitly a viewpoint similar to that of

Grice, considering that, essentially, truth ranks first for the speaker.

3.4. Domains of manifestation of conversational maxims

It is not the case that people follow these guidelines to the letter.
Rather, in most ordinary kinds of talk, these principles are observed by
interlocutors at a deeper level, contrary to appearances. Whenever possible,

66 He makes the difference between apparent and real violations (2003: 160).

132
people will make sense of what they hear, i.e. they will interpret what is said
conforming to the maxims, at least on some levels. It is precisely because of
the need to make assumptions contrary to superficial indications that
conversational implicatures appeared. Let’s consider the example:

‘Where is Bill?’ ‘There’s a yellow BMW outside Sue’s house.’

Beyond its literal semantic interpretation, provided by the grammar,

an utterance has three main types of pragmatic implications (Smith &
Wilson 1990: 179):
- those following from the utterance itself together with a preceding
remark and any item of shared knowledge needed to establish a
connection between the two; in our example, the shared knowledge means:

Bill is our common friend/acquaintance.

Bill has a yellow BMW.
We know Sue and the house she lives in.
The implicature is Bill may be visiting Sue.

- those additional premises, not already part of the shared knowledge,

needed to establish the connection; if the speaker has knowledge of none,
he will construct one:

Even if he is our friend, we have no idea where Bill is, he didn’t tell us.
Sue and Bill know each other. Bill knows Sue well enough to visit her at
home and park his car in front of her house.

- those inferences that follow from the additional premises, together with all
the above, under the first entry:

Bill and Sue might be involved (in something).

All those elements contribute to establishing the relevance of the

utterance as intended by the speaker, and none will follow unless the
utterance is treated by the hearer as relevant (Smith & Wilson 1990: 179).
A hearer will attempt to get out of a remark ‘just what he believes
the speaker put into it’, and the judgement made about the intended
relevance of the utterance ’will crucially affect the amount of work he is
prepared to do to get a message out of it’ (Smith & Wilson 1990: 176).

3.5. Types of non-observance of conversational maxims

‘The idea that someone may be misleading us in some way, either

intentionally or unintentionally, seems alien to this concept [of
communication]. Of course, people often do mislead us in all sorts of

133
 ways, but the fact that we recognize this as a misuse of language is
an indication that communication has as its raison d’être a strong
social, and moral, basis.’ (Finch 2003: 154)

To establish the major types of non-observance of conversational

maxims, one should consider the following features: /intentionality/ and
/misleading/.
Unintentional non-observance accounts for infringement; it is
caused by objective causes making the interlocutor unable to comply with
all the maxims. Generally infringing stems from imperfect linguistic
performance (in the case of a young child or a foreigner) or from impaired
linguistic performance brought about by nervousness, drunkenness,
excitement, disability:

‘Hello!’ ‘Nice weather indeed!’ cf Rom. “Bună ziua, babă surdă!” „Duc la
piaţă nişte urdă!”

Grice (1999: 82; 84) includes into this category clashes; for
instance, there is a clash between the maxim of quality and the maxim of
quantity; to observe one of them would mean to infringe the other one. The
maxim of quality is considered to be more important, therefore the speaker
will choose not to observe the maxim of quantity:

‘Where does C live?’ ‘Somewhere in the south of France.’

The hearer is not opting out, but his answer is less informative than
required, simply because he does not have the necessary information; of
course, there is also the possibility of intentionally misleading the first
speaker, pretending not to know the right address, because going there
contrasts with the interlocutor’s interests.
Intentional non-observance can be misleading and this accounts
for the violation proper of one of the maxims. The speaker is deliberately
and secretly subverting the maxim and the CP, usually for some self-
serving purpose. Violation is defined as the unostentatious or ‘quiet’ non-
observance of a maxim. Violating a maxim is quite the opposite of flouting a
maxim because it rather prevents or at least discourages the hearer from
seeking for implicatures and rather encourages their taking utterances at
face value.

Violation of the Quantity Maxim:

Supervisor: Did you read the articles and write up the review of literature?
Supervisee: I certainly read the articles. Weren’t they captivating!

A: Who’s eaten the biscuits?

B: I’ve had some. (Finch 2003: 158)

134
 If some does not mean ‘all’, then the speaker is cooperative, if not
he withholds information and violates the quantity maxim.

Violation of the Quality Maxim

A: You stained the tablecloth all over! What will the guests think?
B: Nobody will notice.

Violation of the Relation Maxim

A: Did you like my article?
B: That page of the newspaper was impressive, wasn’t it?

Violation of the Manner Maxim

Pierce: Major Frank Burns, M.D., manicdepressive. It’s an honourary title.
Trapper: He’s also schizoid.
Pierce: He sleeps in two bunks. (M.A.S.H.)

The interpretation involving the figurative meaning of schizoid, the

meaning accounting for its specialized use in psychiatry is overtly
cancelled, the receiver being encouraged to consider the literal meaning of
the word, with humoristic effects.
Intentional non-misleading non-observance of the maxims
comprises opting out of a maxim, suspending it and flouting (exploiting) it.

A speaker opts out of observing a maxim whenever s/he indicates

unwillingness to cooperate in the way the maxim requires. Examples: a
suspect exerts their right to remain silent or a witness chooses not to impart
information that may prove detrimental to the defendant.

Detective: Has the defendant ever told you she hated her father and
wanted him dead?
Shrink: Such information is confidential and it would be unethical to share it
with you.

Journalist: Have you accepted bribe?

Suspect: No comment.

Suspending a maxim implies that the interlocutors have no

expectations regarding its observance. Under certain circumstances, there
is no expectation on the part of any participant that one or several maxims
should be observed (and non-fulfillment does not generate any
implicatures). In many cases, conversational maxim suspension is culture-
dependent, i.e. some words, structures are considered taboo, their uttering
signifying an ill omen: Macbeth is ‘The Scottish Play’ in the British acting
community. In the same way, one should not tell an actor ‘Good luck!’, but
‘Break a leg!’

135
 Let’s consider funerals, obituaries, poetry, telegrams and jokes and
notice what maxim is suspended.
In case of funerals, the deceased is praised and the priest and the
family and close friends deliver laudatory speeches, even if the referent
could not have been flawless. Only good things should be said about the
deceased people, it is a cross-cultural belief. Therefore, the maxim of
quality is not observed, if not by lying, at least by omission (in which case
the maxim of quantity is not observed). The same considerations are true
for obituaries. Telegrams obviously don’t observe the maxim of quantity,
whereas poetry and jokes allow for any type of maxim non-observance.

Flouting a maxim means exploiting it, adding new tinges of

meaning, and giving the interlocutor clues in order to be able to interpret
the utterance correctly, i.e. to make the right inferences.

a. Flouting the maxim of quality results in the intentional

non-misleading non-observance of the maxim, which is actually exploited
at a deeper level, the result being:
- a metaphor
Queen Victoria was made of iron. (the interpretation can involve a positive
connotation, iron meaning [+determination], or a negative one, iron
meaning [+inflexibility], [+insensitivity]
You are a lion. (cross-culturally, lion is the symbol of bravery and majesty)
He is a Casanova. (cross-culturally, the proper name came to designate
any womanizer, the negative connotation becoming part of the denotative
meaning of the name)

The examples above are nominal metaphors (Levinson 1983: 152), which,
together with predicative metaphors, are categorically false; unlike them
sentential metaphors are irrelevant to the surrounding discourse when
literally construed:

‘What kind of mood did you find the boss in?’

‘The lion roared.’ (Levinson 1983: 153) (the lion roaring is like the boss
displaying anger)

- a piece of irony
‘What if Russia blockades the Gulf and all the oil?’ ‘Come on, Britain rules
the seas!’ (Levinson 1983: 109)
The reply is an example of using a famous poem title and song lyric
ironically, to express the contrary of what is stated literally.

- meiosis
He was a little intoxicated. (of a man known to have broken up all furniture;
this is an example of meiosis) (Grice 1999: 86)

136
- hyperbole
Every man must have written a poem.

All the examples above require background information, shared by

the interlocutors, the information necessary for an appropriate interpretation
being historical, cultural or political.
The previous examples prove the maxim flouting can give rise to
figures of speech, many of which have a cross-cultural character, becoming
standardized in point of meaning, and , thus, losing some of their originality,
of their power of suggestion, but, on the other hand, being more easily
understood by all the members of that linguistic community. They are
motivated and calculable, i.e. whenever a speaker utters such a structure
and it is not meant literally, but a certain figurative meaning is intended.

b. Flouting the maxim of quantity

War is war.
Either John will come or he won’t.
If he does, he does it.
Flouting this maxim can result in tautologies, which, not only that
aren’t annoying, but convey a variety of meanings.
War1 is war2.
The first term is used with its denotative meaning [+battle,
confrontation], while the second, identical in form with the first, can refer to
any connotation of the basic term: [+death], [+destruction], [+damage],
[+sufferings] etc. The semantic richness of such structures accounts for the
existence of proverbs of this type in many cultures: Fr. À la guerre, comme
à la guerre. Rom. La război, ca la război.

Either John will come or he won’t. ‘There isn’t much we can do

about it, we have to take things as they are.’
If he does, he does it. ‘Even if he does it (and we can’t stop him), it’s
not the end of the world.’

c. Flouting the maxim of relation/relevance

‘Have you finished your homework, John?’ ‘Mum, can I have a piece
of cake?’ (the kid changes the subject, meaning that he hasn’t)
‘Your son is really taken to Annette.’ ‘He used to play with snails.’
(Smith &Wilson 1990: 178)

There is indeterminacy regarding the implicatures: one implicature is

‘a normal person wouldn’t like Annette, but his son has strange tastes,
Annette being just an example’; another is ‘Annette is like a snail.’
Therefore, ambiguities arise regarding implicatures, since several
inferences may be made, even if all unfavourable to the referent Annette.

137
The hearer cannot be certain: ‘the hearer has to supply additional premises
of his own, which he does not necessarily believe, yielding pragmatic
implications, which, he also need not necessarily believe’ (Smith &Wilson
1990: 178).
Grice (1999: 86) gives an example of a real, not apparent,
flouting/violation (the two terms are synonymous to him, in this case) of the
relevance maxim:

‘Mrs. X is an old bag.’ ‘The weather has been quite delightful this
summer, hasn’t it?’

To us, the term ‘real’ seems to contradict the very nature of the
phenomenon of flouting; moreover, the example looks rather as an
illustration of uncooperativeness, since the linguistic contribution should be
made considering the purpose or direction of the verbal exchange;
changing the topic abruptly, when the speaker expects for an answer to his
reply, might not be the best illustration of the Cooperative Principle
definition.
On the other hand, the speaker should be as gentle as possible with
the positive and the negative face of the hearer. We put this idea in
connection with what Kerbrat-Orecchioni called interactional sincronization
(1990: 20-25), since Grice’s framework expresses the harmony between
interlocutors. If the above example, which illustrates a topic change, is also
the illustration of the speaker’s attempt to be polite and to preserve
harmony, then the perspective on the CP should be changed.
Costăchescu (2014) makes such an attempt: starting from the
analysis of what she calls ‘inadequacy markers’, i.e. elements which
express the speaker’s opposition to the discourse or to the behaviour of the
person talked to, she discusses the question of the infringement of the
Cooperative Principle in those cases when the speaker does not refuse to
carry on the talk, but rejects the topic (Costăchescu 2014: 50). The
conclusion of the author is that the CP is too strong and it should be
changed as follows: the contribution of the speaker must be as ‘required’ by
the direction of the talk exchange ‘if the speaker agrees’ (Costăchescu
2014: 59). The interlocutors do not violate the CP if their disagreement is
explicitly expressed.
The example above does not contain an explicit inadequacy marker,
therefore it is debatable if the change of topic unmarked linguistically can be
considered a case of uncooperativeness, or just a case of CP infringement.

d. Flouting the maxim of manner

-flouting the submaxim ‘avoid obscurity’:
In the following examples the answers to the question are, at the
same time, obscure since the speaker uses a variant of the word which the

138
interlocutor or the by-stander doesn’t recognize, and, too long, since the
words not to be recognized are spelled:

‘Let’s get kids something!’ ‘Ok, but I veto I_C_E C-R-E-A-M/ il gelato.’

The kids don’t know the spelling of the word ‘icecream’ and neither
the Italian for the English word, so, the parents are safe if using them,
without the kids protesting against their decision.

‘Where are you taking the dog?’ ‘To the V-E-T.’

The last example is similar to the one given by Yule (1996) in order
to explain that, in the local context of the speakers, the dog is known to
recognize the word ‘vet’, since he hates to be taken there, the speaker
produces a more elaborate, therefore, less brief, and more obscure version
to the dog, so that the latter won’t know the destination.
-flouting the submaxim ‘avoid prolixity’

Miss Singer produced a series of sounds corresponding closely to the

score of an aria from Rigoletto.

In this case, the formal register, vague meanings and rather long
sentence convey the negative meaning intended, but expressed
euphemistically, and with a tinge of irony.
Therefore, the maxims of quantity, quality and manner, submaxim
“be precise” are flouted.
Flouting the maxim of manner, submaxim ‘be precise’67:

Is the Pope Catholic?

Does a bear shit in the woods?

Such structures mean ‘Isn’t it obvious?’ and they can be associated

as replies to any utterance functioning as an interrogation.

A distinct category of examples of maxim flouting is represented by

the use of hedges (Yule 1996). Some structures partly contradict the
implicature that the speaker tells the truth (all of it), flout the maxim of quality:

To my knowledge, nothing happened.

As far as I know, nothing happened.

67 Such an example could also exemplify the non-observance of the maxim of relevance,

since there is no obvious connection between an utterance such as “would you like
something to drink?” and the replies under discussion.

139
I’m not sure it’s (entirely) true, but she was seen in town.

Other hedges denote the non-observance of the maxim of quantity:

We don’t intend to bother you with all the details….

The maxim of relation can also be flouted:
Not that I change the subject, but…
To stick to the point…
Well, anyway, …
I’ll try to be clear and brief….

Yule (1996: 37) mentions three purposes of hedge use, all circumscribed to
the condition of cooperation: they show that the speaker is aware of the CP
and its maxims, that he/she observes them, and that he/she is aware that
the hearer considers him/her cooperative in conversation (has such
expectations from the speaker).

3.6. Tests for conversational implicatures. Drawbacks

Such tests are meant to characterize implicatures as a distinct

class, compared to entailments (logical inferences) and to presuppositions.
The tests are designed as to prove that the inference ahs the features
specific to conversational implicatures.
Conversational implicatures are:

-calculable (predictable)
This property is the result of the fact that logical meaning
corroborated with the Cooperative Principle lead to inferring a certain
implicature.

We present below the steps in calculating an implicature (Levinson

1983: 113-114). The line between the first three steps and the following
was introduced by us for pedagogical reasons, in order to make the
distinction between the stages involving just the speaker and those which
concern both interlocutors and the relationship between them:

1. S said that p
2. There is no reason to think S is not observing the maxims, or at
least the cooperative principle
3. In order for S to say p and be indeed observing the maxims or the
cooperative principle, S must think that q (he is aware of what his words imply)
4. S must know that it is common knowledge that q must be
supposed if S is to be taken to be cooperating
5. S has done nothing to stop me, the addressee, thinking that q

140
 6. therefore S intends me to think that q, and in saying that p has
implicated q.

A counter-example given by Lewis (1973) is the class of standard

metaphors and all clichés which become rather opaque, non-calculable in
the linguistic community. Non-calculability means that their meaning is
almost automatically deducible cross-culturally because of the overuse of
the structures. Their meaning is so general, without being obscure, that
they can be use din any context:

You are a lion.

Is the Pope a Catholic? (‘Isn’t it obvious? cf Rom. “Mănânci , calule, ovăz?”)
To be or not be. (There are only two possible solutions, there isn’t a third
one.)

-cancellable/defeasible (by addressing further premises); a

necessary, but not sufficient condition (if we consider grammatical
homonymy)
He is seeing a woman tonight, but it might well be his mother.
‘What did you do?’ ‘Wake up, get angry, read my mail, had a coffee, but not
necessarily in that order’.

-non-detachable – the implicature is attached to the semantic

content, not to the linguistic form. The synonyms used in an utterance don’t
change the conversational implicatures:

You are a genious/ a mental prodigy/ a big brain.

Esti un geniu/ o minte sclipitoare/ Ai un creier…! Geniule!

In addition to the implicature features implicitly assigned to this type

of inferences by Grice (calculability, defeasibility, non-detachability), Ariel
(2008: 13) adds three more: indeterminacy, universality and reinforceability.

- non-conventionality (indeterminacy)
The speaker associates the structure to various pragmatic
meanings, depending on the background knowledge and situational
context. Ambiguity can be intentional and can contribute to the preservation
of the positive public image of the speaker:

He is a machine.

Two implicatures can be calculated, i.e. two pragmatic meanings can be

inferred in relation to the utterance, depending on the context. The maxim
of quality is flouted and the result is a metaphor, in its turn ambiguous. It

141
can imply the feature [+efficiency], because the major characteristic of a
machine/device is that it can work for a long time at a constantly high level
of precision. Such an interpretation is preferable, since it does not damage
the image of the speaker. Secondly, the metaphor can imply the feature
[-human], including the seme [+cold heart], in which case, the speech act
threatens the public image of the speaker, and, of course, aims at doing so
with the referent’s public image, too.
Ariel’s example (2008: 10) is an originally Hebrew verbal exchange:

Boss: You have small children. How will you manage long hours?
HD: I have a mother. (Hebrew, June14, 1996)

The woman’s reply can imply two things: ‘My mother can help me with the
kids whenever I need it’ or ‘My mother can help me with the kids when
I work late.’
So much of utterance meaning depends on implicature, that one
can never be entirely certain of the full extent of meaning (Finch 2003:
156). It is the users who decide on the pragmatic meaning, they are the
arbiters, therefore it is impossible to prove linguistically who is right and
who is wrong.

- universality
Conversational implicatures are universal because no specific
linguistic form is involved in the triggering of the inference (see non-
detachability). We consider that Ariel makes a step forward compared to
Grice, in that she might refer to the cognitive patterns underlying the
uttering process. Other than that, non-detachability and universality are
quasi-synonymous.

-reinforceability
Inferences may be reinforced explicitly, since they are implicit,
without causing the speaker to sound redundant (Ariel 2008:14):

He passed away in the arms of a woman, not his [wife]. (originally Hebrew,
Hair, March 13, 2003) (a woman means indefinite reference, therefore, an
unknown referent, this implicit meaning being reinforced explicitly by the
structure not his wife)
According to Yule, the opposite of explicit reinforceability is explicit
denial (Yule 1996: 44). Sometimes, we add, there isn’t actually a question
of denial, but a suspension of the inference deduced from the first
utterance:

A: ‘I won 5 dollars.’ (I won only five dollars and no more)

142
B: ‘You won at least 5 dollars.’ (My conviction is that you won more, so I
am not committed to the truth of your assertion. If we want to keep using
the term denial in this case, what B denies is the existence of just one
winning variant; B considers there is a whole range of possibilities
regarding the sum won.)

-suspendable (Yule 1996)

‘Where is Bill?’ ‘There’s a yellow BMW outside Sue’s house. Or am I

wrong?’
The interlocutor is uncertain regarding the implicature to be inferred.

Conclusions

Communicative competence means interpersonal intentionality

(Parret 1976 in Dragoş, 2000: 25), which is subject to the restrictions
imposed by conversational implicatures. It implies differentiating between
literal and figurative (metaphorical) meaning, i.e. between what is explicitly
meant (and stated) and what is implicit.
The way we choose to cooperate in communication is responsible
both for the rhetorical strategies we employ as addressers, and for the
interpretative difficulties we experience as addressees (Finch 2003: 160);
such difficulties result in ambiguities, indeterminacy, uncertainty.
Relevance is the key element in initiating and supporting the
process of interpretation, considering the literal meaning of the replies,
shared knowledge, additional premises and their implicatures.
Cooperative Principle and its maxims are observed at a deeper
level, their flouting being a source of linguistic creativity.
What a speaker implicates is distinct from what he says and from
what his words imply (shared knowledge and a series of inferences are
added); a speaker can make an utterance and intend to mean something
else or something more by exploiting the fact that he may be presumed to
be cooperative, i.e. to be speaking truthfully, informatively, relevantly, or,
otherwise, appropriately. The hearer relies on this presumption to make
contextually relevant inferences from what the speaker says to what he
means. Taking the utterance at face value may be incompatible with this
presumption, which means that the speaker intends the listener to figure
out what he means by searching for an explanation regarding the reason
why the utterance was made.

143
144
 4. PRESUPPOSITIONS

4.1. A philosophical approach of presuppositions

4.1.1. Presupposition and reference
4.1.2. Presuppositions and entailment
4.2. Presuppositions as pragmatic inferences
4.3. Linguistic triggers of presuppositions
4.4. Types of presuppositions
4.4.1. Existential presuppositions
4.4.2. Factive presuppositions
4.4.3. Lexical presuppositions
4.4.4. Structural presuppositions
4.4.5. Non-factive presuppositions
4.4.6. Counterfactual presuppositions
4.5. The Projection Problem
4.6. Tests for presuppositions
Conclusions

4.1. A philosophical approach of presuppositions

Historically speaking, presuppositions were first discovered as a

result of a philosophical approach, the purpose being to differentiate them
from semantic entailments. Hence, the relationship presupposition-
reference and presupposition-entailment are to be paid full attention, in
order to give an adequate description of presuppositions.
Discussing presuppositions from a pragmatic perspective makes
sense, since they are to some extent related to conversational implicatures,
and, to some linguists, the border between the two types of inferences,
and/or the relationship between them, are not clarified. Subchapter 4.2.
illustrates that.

4.1.1. Presupposition and reference

Presuppositions imply that something is implicitly true; that implicit

assertion means assigning a property to an existing entity. Frege (1892)
was the philosopher who discussed referential expressions and time
clauses from that point of view (see Levinson 1983: 169 for the discussion
of Frege’s example – the first below):
Kepler died in misery. ≫ Kepler existed.

145
The existential presupposition is independent of the meaning of the
sentence, which is proven by the fact that it resists the negation of the
sentence verb:
Kepler didn’t die in misery. ≫ Kepler existed.

The discussion continued after you left. ≫ You left. (the presupposition
remains true if the sentence is negated)
The discussion didn’t continue after you left. ≫ You left.
Even in this case, there is an existential presupposition, you exist,
otherwise, there wouldn’t be an agent to perform the action of leaving. (see
4.3. Linguistic triggers of presuppositions)

In his paper ‘On Denoting’ (1905), Russell proposed the theory of

descriptions, holding that the definite descriptions occurring as grammatical
subjects in natural languages are not logical subjects, too, but correspond
to conjunctions of propositions (see 4.1.2.). Almost 50 years later, Strawson
(1950) rejects the complex logic from underlying definite descriptions as it
appears with Russell, but he cannot explain an example such as

The king of France is not wise, because there is no such person.

Strawson showed that the sentence The king of France is wise is

neither true nor false, if there is no such person (in 1905 when Russell
wrote his paper, France was a Republic). The truth value of a sentence
presupposes the existence of the entities spoken of. At the same time,
Strawson acknowledges that one of the main purposes for which we use
language is that of stating facts about things, persons and events
(Strawson 1990: 228).
As a conclusion to this part, we mention Levinson ‘s observation
(1983: 173) that presuppositions and theories on them are in line with our
linguistic intuitions, therefore, if we utter something, there is a foreground
assertion that the referent about whom we assert something exists (or
existed, we should add,) in the context set (of worlds). In conclusion,
existential presuppositions are background assumptions.

4.1.2. Presuppositions and entailment

Knowing the truth of a sentence S1 necessarily implies the truth of

another sentence S2. It would be logically impossible for S1 to be true and
S2 false. If S1 is the first sentence below, and S2 the second, the
relationship is clear.

146
He is a bachelor68. ≫ He is a man. (Fromkin & Rodman 1998: 180)
S2 is an entailment of S1. Knowing the entailments of true sentences is
how the knowledge about the world is acquired. A contradiction is a false
entailment: for the previous example, it would be He is not a man.
The sense of a declarative sentence allows us to know under what
circumstances it is true. The ‘circumstances’ are the truth conditions of the
sentence. For declarative sentences, they are the same as their senses.
Sense facilitates the study of the world and finding out facts. […] the truth
or falsehood of a sentence S is its reference.
Presuppositions are facts whose truth is required for an utterance to
be appropriate (Fromkin & Rodman 1998: 198). They presuppose a
relationship between two propositions:

Mary’s dog is cute (=p)

Mary has a dog. (=q)
p≫ q
The relationship manifests constancy under negation (Mary’s dog is cute≫
Mary has a dog) and the same is true for a proposition expressed by an
embedded clause: I think Mary’s dog is cute. ≫ Mary has a dog.
Therefore, a presupposition means double entailment: p ≫q, non-p ≫ q.
Levinson (1983: 168) discusses presuppositions defined as ‘any
kind of background assumption against which an action, theory, expression
or utterance makes sense or is rational’, a definition which is to be
contrasted with the perspective of presuppositions as pragmatic inferences
‘built into linguistic expressions and isolated using linguistic tests’.
Strawson referred not to sentences, but to statements, i.e. to the
use of sentences to express something true, but defined presuppositions in
the same way, as double entailments. He clarifies the conceptual, and,
consequently, terminological ambiguity existing with Frege in the discussion
on presuppositions: sentence vs. assertion. Strawson points out that a
sentence is different from its use (for example, an assertion), in point of
their basic feature, truthfulness: sentences can’t be true/false, statements
are. The truth/falsity of an assertion depends on conventions about the use
of referring expressions, conventions which lead to pragmatic inferences
(presuppositions), in their turn, preconditions to judging an assertion as
true/false.
Statement A presupposes statement B iff B is a precondition of the
truth or falsity of A (Strawson 1952: 175 in Levinson 1983: 172).

68 He replaces a common noun or proper name designating a male referent: John/the man
is a bachelor. Of course, bachelor is used here with the meaning ‘unmarried man’, not
‘owner of the first university degree’; in this latter case, the noun bachelor is a common
gender noun.

147
 Considering Russell’s point of view, he refers to assertions, i.e. his
p.o.v. inclines towards the concept of statement, rather than towards that of
sentence. To him, an assertion is a conjunction of conditions and negating
it may imply negating one of those conditions: negation has a wide and a
narrow scope, i.e. it is ambiguous – a wide scope negation means denying
that an entity x exists; a narrow scope negation only denies that the
predicate applies to that entity. The nature of this ambiguity remained to be
discovered, even if Levinson disagrees with this idea, mentioning that there
is no proof of ambiguity regarding negation in natural languages.
Nevertheless, the merit of Russell consists in discussing what are now
called scope ambiguities. We illustrate Russell’s theory with his example:

King of France is wise.

There is a king of France.
There is a unique king of France.
If there is a king of France, and he is unique, then he is wise.

Negating the basic sentence, means negating any of its three

components. Generalising,

The F is G is the conjunction of the following three assertions:

X has property F
There is no other entity y, y ≠ x, which to have property F.
X has property G.

So far, we referred to linguists who associated presuppositions to

sentences as structural and semantic units, or to assertions/statements as
logical-semantic entities. To other linguists, presupposition is something
that the speaker assumes to be the case prior to making an utterance.
Speakers, not sentences, have presuppositions, while, by contrast, an
entailment is something that logically follows from what is asserted in the
uttterance (Yule 1996: 25-26). Yule’s example is meant to show that
entailments follow from the sentence, irrespective of the speaker’s beliefs,
be they right or wrong, whereas presupppositions can be false if the
speaker’s beliefs are wrong; a counterargument would be that He has a lot
of money is an implication, not a presupposition:

Mary’s brother bought three horses.

Presupp. Mary exists.
Mary has a brother.
Mary has only one brother.
He has a lot of money. (this presupposition can be false; some
would consider this an implication, not a presupposition)
Entail. Mary’s brother bought something.

148
4.2. Presuppositions as pragmatic inferences

Though presuppositions were first approached in literature from a

philosophical (and, implicitly, semantic) perspective, we continue with the
pragmatic approach of this class of inferences, in order to point out their
similarities with conversational implicatures, previously discussed in
chapter 3.

‘Presuppositions are the assumed or shared knowledge which exists

between speaker and hearer. […] Presupposition and inference are
part of the logical machinery we use to interpret utterances. […] they
need the raw material of shared knowledge and cultural
understanding on which to operate’. (Finch 2003: 156)

Presuppositions are best treated as conversational implicatures: this

idea was first stated by Karttunen & Peters (1975, 1979) and mentioned by
Levinson (1983: 131), who disagrees to it. But even Levinson agrees that
both presuppositions and implicatures are pragmatic inferences which
cannot be thought of as semantic (in the narrow sense), because of their
sensitivity to contextual factors. Those factors make such inferences
relevant.
We further provide two examples illustrating how presuppositions
are calculated:
I went THERE. ≫ (presupposes I went somewhere. The focus is on the
adverbial of place, stressed by phonological means, which makes the
presupposition relevant in the context)

I imagine you are/do. (depending on the context, the verb imagine means
‘making a mental projection including a certain referent & action’ ≫ you
arent’/don’t etc. etc, or simply means ‘believe’, in which case, the verb
refers to the speaker’s commitment to a belief ≫ you really are/do etc.

Stalnaker (1974) pointed out that the speaker presupposes

something about the addressee and/or the context, i.e. takes it to be
understood that presupposition is true, since it is assumed, believed. There
is a permanent acceptance by the interlocutors regarding the informative
presuppositions, assumed to be true; the updating is not necessary
previously to the sentence uttered.
According to Stalnaker, ‘a speaker pragmatically presupposes that p
by uttering an expression e in a certain context just in case:
a. the speaker assumes or believes that p
b. the speaker assumes or believes that in a given
context his addressee assumes or believes that p

149
 c. and the speaker assumes or believes that in the
context his addressee will recognize that the speaker is making
these assumptions or has those beliefs
d. or the speaker acts as if or pretends that all the
above conditions are true’.

Moreover, as already mentioned, the context includes the shared

knowledge of the interlocutors. Pragmatic presuppositions imply shared,
common ground information (Lakoff 1972) and any further inferences which
are taken to be understood by the speaker and the hearer to contribute to
the relevance of an utterance. Presuppositions must be true in a context if a
sentence is felicitously used. Making an assertion means assuming
pragmatic presuppositions, i.e. imposing requirements on the common
ground. A sentence cannot be used to update a common ground unless it
has a determinate semantic value in all of the worlds of the context set
described by the common ground. (Burton-Roberts 1989; Gauker 1998)
It should be said that the common ground of a conversation at a
particular time is the set of propositions that the participants in that
conversation at that time mutually assume to be taken for granted and not
subject to (further) discussion. The common ground describes a set of
worlds, i.e. the so-called context set: those worlds in which all propositions in
the common ground are true, and the actual world is such a world. Any
assertion (I’m tired.) can be true or false in a certain context (state of affairs).
Assertions are meant to update the common ground and are added
to it, if the sentences expressing them are accepted by the participants. In
other words, the context set is updated by removing the worlds in which this
proposition is false and by keeping the worlds in which the proposition is
true. (von Fintel 2000)

4.3. Linguistic triggers of presuppositions

Since it is generally acknowledged that presuppositions are derived

from the linguistic meaning of a sentence containing a lexical element
acting as a presupposition trigger, there are a series of such elements that
we enumerate below, following Levinson (1983: 181-182), who himself
selected the triggers from a comprehensive list of Karttunen:

- definite expressions (proper names, NP – definite article/ possessive adj.

(+ adjective) + noun):
The new neighbour has come. ≫ there is a new neighbour known to us
John is here. ≫ John exists
My car is broken. ≫ there is a car that belongs to me

150
- factive verbs (which presuppose the truth of the content of the subordinate
clause following them): regret, realize, know, be aware of, be odd, be
proud/ glad/ sad/ indifferent that…:
I know what you did. ≫ You did something.
They will realize their mistake. ≫ They made a mistake.

- implicative verbs: manage, try, forget, happen, avoid etc:

We managed to come earlier. ≫ We tried to come earlier.
She forgot talking to us. ≫ She talked to us.

- change of state verbs: stop, begin, start, continue, cease, leave, enter,
come, go, arrive etc:
They stopped working. ≫ They worked.

- iteratives: again, another time, the x-th time, come back, restore, repeat
etc:
The students repeated the question. ≫ They asked the question before.

- verbs of judging: accuse of, criticize for etc:

He was accused of theft. ≫ Something was stolen. Stealing is bad.
- temporal clauses and constructions; clauses are introduced by after, before,
while, since, as, whenever, whereas time adverbials are introduced by during
etc:
I’ve been waiting for you since you called. ≫ You called.
During the party, everybody talked to everybody. ≫ There was a party.

- cleft and pseudo-cleft constructions:

It is him who saved them. ≫ Somebody saved them.
What they want is freedom . ≫ They want something.

- implicit clefts with stressed constituents:

She was called by MARY. ≫ Somebody called her.

- comparisons and contrasts:

I am better than you. ≫ You are good.
HE called HER, not the other way round. ≫ Somebody called somebody else.

-non-restrictive relative clauses (they are not semantically closely linked to

the main clause, and are separated from it by a comma:

151
Her friend, who graduated in France, owns a restaurant in our town. ≫
Her friend graduated in France.

-counter-factual conditionals:
If he had done it, we would have known. ≫ He didn’t do it.

- questions; yes/no questions presuppose an either... or… inference,

whereas wh- questions introduce the presupposition obtained by replacing
the wh- word by somebody/something/somewhere etc:

Did you finish? ≫ Either you did, or you didn’t.

When did you finish? ≫ You performed an action and you finished it (at a
certain moment in time).

As a conclusion and justification of this enumeration, it remains true

that, as Levinson pointed out, the list of presupposition triggers depends on
one’s definition of presupposition.

4.4. Types of presuppositions

The interdependence between the linguistic meaning of a sentence

and the situational context of use makes Yule (1996) refer to all
presuppositions triggered by linguistic expressions as potential
presuppositions, actualized in concrete situations of communication (Yule
1996: 27). The linguistic triggers were enumerated in the previous
subchapter, so, at this point, we discuss only the corresponding type of
presupposition introduced by each class of triggers.

4.4.1. Existential presuppositions were analysed in the previous

subchapters, due to their importance in establishing the basis of any
assertion; linguistically, such presuppositions are triggered by structures
containing Noun Phrases made up of proper names, or of common nouns
preceded by determiners, such as definite article or possessive adjectives.
The speaker is assumed to be committed to the existence of the entities
named:
Mary finished her job. ≫ (presupposes) Mary exists.
My cat left. ≫ My cat exists.
The window is broken. ≫ There is a window in that room/ nearby.

4.4.2. Factive presuppositions are triggered by verbs expressing a high

level of certainty regarding the information stated in the clause following the
main clause verb. Such verbs are, thus, factive verbs, i.e. strong epistemic

152
quantifiers, according to Caton (1981), they are placed high on a scale
measuring the degree of propositional certainty expressed by the clause
governed by them:

I’m aware that you are tired.

(to be aware presupposes the truth of the statement you are tired; that is
considered certain, a fact, therefore, the structure in the main clause – to
be aware- is a factive one)

It’s odd that they left.

(it’s odd represents the speaker ‘s qualifying of the event described in the
subordinate clause; it presupposes that the event happened, it was a fact)

We regret telling you that your reservation is no longer valid.

(the factive verb to regret presupposes performing the action of telling, and
the action of telling is determined by the state of affairs; therefore the verb
regret takes scope over the subordinate THAT Clause, too)

4.4.3. Lexical presuppositions are triggered by aspectual/change of state

verbs. The lexical form (i.e. the verb) and the asserted meaning
presuppose that another non-asserted meaning (another concept) is
understood, by convention:
You stopped laughing. ≫ You laughed.
They started laughing. ≫ They did not do that previously.
She laughing again. ≫ She laughed before.

4.4.4. Structural presuppositions

Certain sentence structures conventionally and regularly

presuppose that part of the structure is already assumed to be true. The
information is assumed to be true by the speaker and to be accepted as
such by the hearer. Temporal clauses are such structures, since they
provide the information implicitly assumed to be true about the
circumstances of the event described. Time coordinate is essential in
interpreting an utterance, as it was presented in the chapter 2 about deixis.
They left before we returned. ≫ We returned.
The information is assumed to be true even if it can lead the interlocutor to
believe that they are necessarily true. Wh- questions assume the truth of
the event referred to in point of circumstances. Questions (during an
interrogation, but not only) can be asked so as to subtly manipulate the
hearer to necessarily believe what the speaker believes (see 6.5.2.
Structural presuppositions and hearer’s manipulation):
Where were you when you heard the noise? ≫ You heard the noise. (it’s to
my interest to make you believe so)

153
4.4.5. Non-factive presuppositions are triggered by verbs which
semantically denote a state of affairs which is possible, but not actualized:
dream, imagine etc.

I dreamed I was at the seaside. (obviously, at present I am not there)

I imagined the house was ready. (the house is not ready, otherwise,
I wouldn’t have needed to imagine that)
Still, the verb imagine can be interpreted as a synonym of believe, in which
case it is no longer a non-factive presupposition trigger: I imagine you can’t
do that.

4.4.6. Counterfactual presuppositions are a special type of non-factive

presuppositions, only that they cannot be actualized any more, since they
refer to a past moment and events that could have happened then, but
didn’t. Conditional sentences type 3 illustrate syntactically and semantically
this kind of presuppositions:
If I had known you were in trouble, I would have helped you. ≫ I didn’t
know you were in trouble.

4.5. The Projection Problem

Technically, the presupposition of a whole utterance is the sum of

the presuppositions of the components (Langendoen & Savin 1971). As a
result of that, presuppositions survive when part of a more complex
sentence, it is a basic expectation on the part of the users. But, sometimes,
the meaning of some presuppositions as ‘parts’ doesn’t survive to become
the meaning of complex sentences as ‘wholes’. This means the Projection
Problem (Yule 1996: 30).
Some presuppositions don’t ‘project’ because they are weaker than
entailments (necessarily following from what is asserted). This is one
reason of presupposition cancellation. The power of entailment also
cancels existential presuppositions:

‘Jane was here, in your house.’ ‘I know nobody called Jane!’

‘Santa Claus is coming to town!’ ‘There is no Santa!’

All presuppositions should be treated as potential presuppositions,

actualized only if the speaker intends them to be recognized as such
within utterances.
Yule asserts that an existential presupposition introduced by a NP
containing a possessive adjective can be suspended, i.e. turned into a
potential presupposition, by adding the structure or something
(Yule 1996: 32) cf Rom. sau ceva de felul ăsta/ceva de genul (ăsta)/ sau
cam aşa ceva:
He is mending his car. ≫ He has a car. cf Rom. Îşi repară maşina. ≫ Are
maşină.

154
He is mending his car or something. ≫ Maybe he has a car to mend or he
does a similar job. cf Rom. Îşi repară maşina sau ceva de genul ăsta. ≫
Poate are maşină şi o repară sau face vreo muncă de acest fel.
Yule’s example is questionable, since the structure or something
can refer to the action of mending, not to the relation of possession.

Another reason of presupposition cancellation is the semantic

nature of the embedding predicate (main clause predicate). The first to
discuss the matter was Karttunen (1973, 1977). At this point, it would be
useful to enumerate the types of presupposition triggers which, depending
on their linguistic meaning, behave as operators with a variable potential to
preserve presuppositions in case of embedding. Levinson (1983) talks
about presupposition ascending, i.e. the presuppositins of the parts (of the
subordinate clauses) become presuppositions of the whole, i.e. of the entire
utterance:
- holes (factive verbs, modals, aspectual verbs) that allow
presupposition ascending; presuppositions and entailments survive in
different contexts:
I regret that you are mistaken. ≫ You are mistaken.
You must have seen Jane, (because) only she knew that information. ≫
Jane knew the information.
They continued looking for a solution. ≫ They began looking for a solution.

Existential presuppositions certainly survive under negation, and in

modal contexts. A modal context means the presence of a grammatical or
lexical marker of modality. In case of epistemic modality (expressing
various degrees of certainty), modality markers are to be seen as
quantifiers (Caton, 1981): strong- know, must, be certain, be sure etc-,
moderate – think, be likely etc- or weak- be possible, maybe etc-,
depending on the degree of certainty expressed:

The manager hired ten employees. ≫ there is a manager

The manager didn’t hire ten employees. ≫ there is a manager
It’s possible that the manager hired ten employees. ≫ there is a manager
He should have hired ten employees. ≫ there is a manager

- plugs are the opposite of holes, they may block the ascending of
the presupposition; declaratives and attitudinal verbs are included in this
class: say, declare, be convinced etc:
He said/ assured us that everything will be ok. ≫ Maybe it will, maybe it won’t.
I was convinced that his friend was right.

155
The presupposition introduced by the subordinate clause is turned into a
belief of the speaker; if we agree that presuppositions characterise the
speaker, and not a sentence, then the presupposition is his friend was right;
if we define a presupposition as deriving from a sentence, then, objectivelly,
his friend was right does not become a presupposition)!!

- filters are actually connectives that behave selectivelly in point of

presupposition preservation; such connectives are and, if, either... or..., but,
alternatively, suppose that. ‘Presuppositions may be filtered in specifiable
contexts when they arise from sentences that are part of compounds
formed by the use of the connectives or, if...then and others.’ (Levinson
1983: 198)

You’ll help her and you’ll pay for it.

The first sentence explicitly states what the second presupposes – so, of
course, the presupposition is preserved)

Either she will not go abroad, after all, or she will regret doing it.
To regret is a factual verb, technically presupposing she went abroad, but
the connective either…or… takes scope over the second sentence and the
presupposition is not preserved.

If she takes up Chinese, she will regret it. (the same explanation as before,
if cancels the presupposition triggered by regret – she took up Chinese)

But in a sentence of the type

If they were caught cheating again, they will be expelled. ≫ They were
caught cheating before.
the lexical element is stronger than the conditional structure; again takes
scope over if.
Nevertheless, we may enconter a presupposition suspension even
in such a case:

‘If they were caught cheating again, they will be expelled.’ ‘If, indeed, they
were ever caught cheating before.’

‘In a sentence of the form if p then q, and perhaps, p & q, the

presuppositions of the parts will be inherited by the whole unless q
presupposes r and p entails r. The same is true for a structure of the
type p or q: the presuppositions of the parts will be inherited by the
whole unless q presupposes r and non-p entails r’ (Levinson 1983: 197).

For example, in the utterance

John is going to do linguistics or he is going to regret it.

156
the first sentence asserts what the second presupposes. A sentence can
both entail and presuppose the same proposition. Thus, a presupposition of
the second clause is filtered by entailing it. Presuppositions are not entailed
by the context, but must be consistent with it.

4.6. Tests for presuppositions

Such tests involve, actually, assessing presuppositions in point of

their preservation or non-preservation, in various linguistic and discourse
contexts (or including making assumptions contrary to them).
Consequently, the presupposition properties and their nature itself can be
established.

Non-detachability (see also 4.5.1., 4.5.2., 4.5.4.)

Factive verbs, modals and time clauses illustrate presupposition

preservation in embedded clauses.
Nevertheless, there are counter-examples proving that the test is
not entirely reliable:

Presupp. Presupposition preservation Presupposition

trigger non-preservation
Factive I regret telling him. I regret to tell him…
verbs He won’t have to regret that…
He knows it’s too late. I don’t know that it’s so late .
Modal He should leave. ≫ he exists It should rain but you can never
know…
Time She left before I came. ≫ I She left before we could talk to
clause came. her.

‘[…] it can be reasonably claimed that the positive sentences

constructed with holes (factives, modals), in fact entail their alleged
presuppositions, and it is only in the negative, disjunctive or
conditional contexts that the uniquely presupposition survival
behaviour manifests itself’. (Levinson 1983: 193)

Presuppositions are preserved under negation, in questions and

imperatives:
a. existential presupposition maintained in case of negation or
exclamation:

The manager didn’t hire ten employees. ≫ there is a manager cf The

manager hired ten employees. ≫ there is a manager
The manager hired ten employees! ≫ there is a manager

157
b. presupposition preserved in a question:

When did you finish? ≫ You finished.

Presuppositions are also preserved, at least, sometimes, if

synonyms are used, which partly contradicts their dependence on certain
lexical triggers:

I regret telling /am sorry that I told/am unhappy about telling/etc him the
truth. ≫ I told him the truth.

Defeasibility/Cancellability

This property cannot be analysed independently of its counterpart,

non-detachability, that is why the basic ideas regarding the topic have
already been stated. Technically, any presupposition can be cancelled, if
there is a linguistic or discourse context which causes that (there is an
entailment which cancels the presupposition). Potential presuppositions are
cancelled if they conflict with entailments, implicatures, or other
presuppositions whose truth has already been established in the context.
The discourse context can cancel the presupposition:

‘How is your brother?’ ‘Fine, but I have no brother.’ (this verbal exchange is
possible if the two interlocutors don’t know each other very well; the reply
‘Fine, but I have no brother.’ could express the interlocutor’s intention to be
polite, i.e. not to contradict the speaker from the beginning, and, also,
depending on the context, to add a positive connotation of joyfulness and
humour, or, on the contrary, of irony. The tone and the intonation make the
difference, in this case.)

The linguistic context can also cancel the presupposition:

He won’t have to regret that he left, because he didn’t leave at all.

Suspension

Presuppositions can be not only denied, but also suspended (Horn

1972), and the most frequent and efficient way of doing it is by adding an if-
clause, or a coordinated sentence containing a weak epistemic quantifier:
We resume an example given under 4.6, pointing out that the last
conditional clause suspends the presupposition they were caught cheating
before, making it uncertain:

‘If they were caught cheating again, they will be expelled.’ ‘If, indeed, they
were ever caught cheating before.’

158
‘Did he call any more?’ ‘No, and I doubt that he called in the first place.’/
‘No, and maybe he never did, in fact.’

The description of presuppositions in their basic features and

theoretical approaches would not be complete without making a contrastive
analysis of them and of conversational implicatures:

Characteristic Implicature Presupposition

Their nature Pragmatic in nature (types of inferences)
Semantic/pragma Felicitous/infelicitous True/false*
tic general feature
Connection to No connection to More connected to
sentence sentence meaning, but to sentence meaning
meaning/lexical sentence usage and/or (speaker’s
item meaning meaning); it is
conventionally
associated to the
meaning of some
lexical items

There may be lexical There are

elements triggering presupposition
conventional implicatures carriers/triggers
Type of source The speaker (source) The speaker and the
(internal–the gives it to be understood hearer take it to be
speaker/external) that… understood that…
Context- Variable They are context-
dependency (generalised/standard bound, deriving from
implicatures are context conventions about the
independent, while use of referring
particularized expressions
conversational
implicatures are context-
bound
Properties Calculable Identifiable starting
from the hypothesis of
common ground
Non-detachable (presuppositions only in
some cases –under negation, in questions,
imperatives, embedding)
Cancellable
Indeterminate meaning Suspension →
for conversational uncertainty
implicatures
*assumed to be true in a context, consistent with it (i.e. not contradicted by it)

159
Conclusions

Whether presuppositions are considered as pertaining to sentences,

sentence uses or speakers, it remains true that they are pragmatic
inferences to a great extent linked to the linguistic structure of the basic
sentence;
As a result of that, presuppositions are triggered by certain linguistic
carriers, whose list depends on how presuppositions are defined; the type
of linguistic carriers determines the type of presupposition inferred;
Presuppositions have a series of features which become tests, in
order to determine if the inference under discussion is a presupposition or
not; such tests are also a result of the relationship between presupposition
carriers and presuppositions and they include cancellability, non-detachability
and, last but not least, suspension;
The contrastive description of conversational implicatures and
presuppositions may give an idea about the reason why there are
controversies regarding their illustrating a paradigmatic, i.e. taxonomic, type
a relationship.

160
 5. SPEECH ACTS

5.1. Introduction
5.2. From performative utterances to SAs or vice versa? The
Performative Hypothesis
5.3. SA Levels. Speech Act Schema
5.3.1. SA Levels
5.3.2. Speech Act Schema (SAS)
5.4. The concept of Speech Act in communication
5.5. SA classification: from Austin to Searle
5.5.1. Direction of fit
5.5.2. Illocutionary point
5.5.3. Felicity (Happiness) Conditions
5.5.3.1. Felicity Conditions with Austin
5.5.3.2. Felicity Conditions with Searle
5.5.3.3. Counterarguments to Felicity Conditions
frame
5.6. Indirect Speech Acts
Conclusions

5.1. Introduction

A basic model of verbal communication must involve the following

factors, shared by the interlocutors (Smith & Wilson 1990: 173):
- a body of linguistic knowledge (grammar);
- a body of non-linguistic knowledge and beliefs (an encyclopaedia);
- a set of inference rules (a logic).
Linguistic knowledge makes the interlocutors deduce a set of
propositions from the sentence uttered by purely linguistic (semantic) rules;
they will be entailments. Up to this point, formal syntax and ‘truth-functional’
semantics can account for linguistic behaviour; but these domains study
predominantly declaratives, which express a logical-semantic content (a
proposition) that can be True/False.
But language is more than sound and meaning; the shared non-
linguistic knowledge and the shared set of inference rules account for the
propositions deducible directly or indirectly from the sentence uttered (Smith
& Wilson 1990: 174). Speech, i.e. uttering, but also the oral/written
expression of it, i.e. the utterance, mean action, conveying a certain
message to the interlocutor. This is how the concept of speech act appeared.

161
5.2. From performative verbs to performative utterances or vice
versa? The Performative Hypothesis

Austin (1962) made the distinction between constative utterances,

which describe an existing state of affairs and performative utterances, by
which the speaker performs the act expressed by the verb, at the moment
of uttering. Performatives are self-referential, they have a token (i.e.
occurrence)-reflexive nature:

The door is open.

I promise to come back soon.

The first utterance is descriptive, it presents a state of affairs which

is not to be changed, whereas the second refers to an action performed by
the speaker at the very moment of uttering, the action of promising. This
example contains explicitly a verb whose meaning is compatible with
expressing the action while uttering that the speaker is doing it: promise. It
is a performative verb, and the utterance is an explicit performative
utterance. The verb has to be in the first person singular, Indicative Mood,
Active Voice; such conditions of form are obvious, since only so the
speaker can perform the action he states. If the form of the verb is
changed, then its performative property is annulled:

She promises/promised to be back soon. (the past tense form turns the
utterance into a descriptive one)

Therefore, the property of a verb to express a performative value is a

function of its sense, but also a function of its form. Of course, it is a
function of its use: the utterance is performative only if the speaker utters it
in order to make a promise (see 5.5.3.).
As far as sense is concerned, it can change in time, the
performative dimension, inherent at first, disappearing in some cases; but
the change of the sense will be accompanied by an obligatory change of
morpho-syntactic behaviour:

I bet you six pence that you will lose.

I bet you still love her.

As a performative, the verb has an obligatory direct object, whereas,

as a non-performative, it becomes a strong epistemic quantifier, expressing
a high degree of certainty: ‘I’m sure you will lose’. Moreover, in the first
sentence, for the speaker to claim his money, in case the prediction is
correct, the interlocutor has to ‘take on’ the bet, by saying ‘You’re on’. This

162
is performing a corresponding SA expressing ‘uptake’. Otherwise, the
parties will not have had the right intentions and/or conduct.
In his book, Austin tried to classify verbs according to this property,
but such a list would be impossible to make, therefore, Searle (1989)
preferred to discuss the case of performative utterances, and not of
performative verbs. More so as there were some reasons for that: a
performative utterance needn’t necessarily contain a performative verb:
some utterances are what Austin called primary performatives, not explicit
ones, as in the previous case, I promise to come back soon:

Come back! vs I order you to come back.

You may come! vs I allow you to come.
You have come! vs I rejoice at your coming!
Could you have come? vs I am asking if you could have come.

As a conclusion, Austin made a deductive analysis, starting from

performative utterances and getting into the deeper level of performative
verbs analysis. Searle actually focused on performative utterances, since
the same utterance can have various values in the context. Moreover,
every utterance can be turned into a clause containing an explicit
performative which makes the illocutionary force, i.e. the value of the
utterance, explicit, as in the second example of each of the pairs above.

I (hereby) Vperformative you [(that) U]

As mentioned at the beginning at this subchapter, the subject has to

be in the first person singular in order for the subject to have the semantic
role of agent performing the action expressed by the Main Clause verb; the
verb must be in the Present Tense Simple or Continuous, Active Voice.
Optionally, the discourse marker hereby, an adverb, can be added to
indicate the certainty that the utterance counts as the action denoted by the
main clause verb. Hereby is, thus, an interpersonal theme marker, i.e. it
aims at establishing a relationship with the interlocutor, relationship which
involves full trust in the action expressed by the main clause verb. Such
adverbs are attached mostly to explicit performatives. Searle’s classification
mainly rests on features of representation, i.e. a semantic trait (Mey 1993:
132 cf Lyons 1977: 5), other features rely on morpho-syntactic traits.
Exceptions from these remarks and cases are given by Levinson
(1983). We adapted them:

The company hereby undertakes to make the changes… (the subject is not
in the first person singular)
We regret that we are forced to hereby request you… (the adverb in not
placed in the main clause, the one containing the performative verb)

163
 Nevertheless, explicit and primary performative versions might not
be equivalent:

Have a sandwich! (offer) vs. *I order you to have a sandwich. (impossible

pragmatically, unless the relationship is a vertical one and the speaker
chooses this variant to force the hearer to perform the action to his own
good; it could be uttered like that by a parent to his child who is fussy about
food.)

Could you do that? (request) vs. I’m asking you if you could do that. (the
use of an explicit performative could express the speaker’s irritation about
the interlocutor’s avoiding to give an answer).

The fact that any utterance could be turned into a performative one
made Gazdar (1979) and Levinson (1983) generalize and formulate this
idea as The Performative Hypothesis, which states that all sentences have
a performative clause as the highest clause in their deep structure. This
highest clause can be deleted, and the result is an implicit performative.
The message to be conveyed (the illocutionary force) is simply the
performative clause, which is true by the simple uttering of it.
Truth/falsehood is not an issue.
As a result of that, Austin’s constatives become primary
performatives with the illocutionary force (i.e. the communicative goal) of a
statement:

It’s raining. → I state/assert that it is raining.

The Extended Performative Hypothesis (Sadock 1974) refers to the

fact that the illocutionary force of an indirect speech act can be appropriately
formalized in a performative deep structure (Leech 1983: 193-194):

Can you close the window? means ‘I request you to close the window.’
(see 5.6.)

To quote Mey (1993: 133)

‘there is a certain asymmetry in the relationship between SAVs

(speech act verbs - our note) and speech acts (SAs) proper: first of
all, not all SAs are represented by a specific SAV, but may be
represented by several […]second, and conversely, nor every SA has
a corresponding, custom-made SAV of its own’

I promise to help you (PF) ≠ I/He promised to help you./ He promises to

help you.(non-PF)

164
 The same can be said about the value of the adverb hereby which is
not always performative and does not render a sentence a performative
value by its simple presence. The examples are adapted from Mey:

I hereby declare the session open.

*I hereby adore you.
*I hereby declare that two and two is four.
I hereby declare my innocence.

Mey’s conclusion is that, at best, the adverb hereby is an indicator of SAVs

in general, not of performativity.
Also, there are verbal expressions which deny what they are doing
or do what they are denying:

1. I don’t want to bother you, but could you help me with the luggage?
2. I hate to disturb you but I need to know…
3. I don’t want to scare you, but it’s too late to leave now all by yourself.

All the examples contain a first sentence with verbs performing the action of
not disturbing or presupposing it (ex. 2), which are then denied by the
second sentence of each example.
A SA can be performed without a SAV and, in some cases, the
speaker cannot even properly perform that very SA technically expressed
by that verb if he/she explicitly mentions it:

*I threaten you that I’ll come again.

Therefore, performativity is a property that is not specifically bound

up with SAVs; Mey refers to Verschuren’s opinion that we are dealing with
a performativity continuum, running all the way from ‘institutionalised’ SAVs
such as ‘to baptize’, to everyday verbs that occasionally can take on a
performative character’ (Mey 1993: 137).
Katz (1977) explained SA force in terms what he called, following
Chomsky, grammatical competence and pragmatic performance; in other
words, contextual factors interact with the embodied illocutionary force to
create speaker’s meaning (Holtgraves 2002: 19). That is why Levinson
thought there is no need for a SA dimension of the illocutionary force, and
Leech agrees with the following idea:

‘given a set of conventional maxims, a literal interpretation of the

utterance, and a specification of its context of use, it is in principle
possible to infer its illocutionary force.’ (Miller &Johnson-Laird 1976:
636 in Leech 1983: 172)

165
 The question to be asked is if the (rather) conventional character of
those inferences can be linked to Austin’s belief that IF is conventional by
nature.

5.3. SA Levels. SA Schema

5.3.1. SA Levels

John Longshaw Austin (How to Do Things with Words, 1962), was

the first who, in his lectures delivered at Harvard in 1955, described the
three levels of a performative utterance. He was the adept of the study of
ordinary language philosophy, considering the distinctions and connections
worth drawing by people throughout time.
To him, any SA has three levels:
a. the locutionary act, the act of saying;
b. the illocutionary act, the act performed in saying;
c. the perlocutionary act, the act performed by saying.

a. According to Austin, in its turn, the locution means:

-a phonetic act, that of uttering noises;
-a phatic act – uttering words made up of those noises, words used as
pertaining to a vocabulary and combined according to rules as pertaining to
a grammar;
-a rhetic act – the word was used having sense and reference, i.e.
meaning.
John Searle (1976) preferred to replace the term locution by
proposition, which included reference and predication, i.e. the use of a word
to designate an object in reality and to assign a property to it. He didn’t
think it necessary to make reference to phonetics, grammar or vocabulary,
probably considering that these levels were implicit, the user being a
competent one.
In conclusion, a SA is proposition + illocutionary force (IF); a SA is
constructed according to some constitutive rules based on a set of
necessary and sufficient conditions which make it successful, i.e. felicitous.

b. illocution is a functional category, related to a sentence form. It

refers to a type of act performed in saying something. By it we assess the
function of what is said (the illocutionary force of an utterance). The
illocutionary force is what turns literal meaning into a SA. To Searle,
SA meaning = propositional content + Illocutionary Act (illocutionary
force – IF)
This led to formulating the Principle of Expressibility: it is an
analytical truth about language that whatever can be meant can be said. IF
contains the complex communicative intention/goal, ‘the securing of the

166
uptake’, or, according to Grice, the speaker’s meaning. Two basic
properties characterize an illocutionary act: its communicative character
and its conventional character. These two properties made Bach & Harnich
(1979) formulate The Communicative Presumption: following the CP, there
is a mutual belief that S (the speaker) states a proposition p to H (the
hearer) meaning q.

I’ll come can count as

-a promise: ‘The party won’t be the same without you.’ ‘I’ll come.’
- a threat, which can be subsumed to promises, since it is a promise that
something detrimental to the hearer will happen: ‘If you don’t behave
yourself, I’ll come and get you.’
-an offer: ‘I have to leave and find someone to take aunt Mary home.’ ‘I’ll
come.’
- an assertion, if everything has been arranged and the action has high
chances to happen: ‘Who comes to the meeting?’ ‘I’ll come.’

c. According to Austin, saying something produces consequences

(therefore the act of uttering functions as a cause), this being a frequent
and normal phenomenon. The consequences involve the thoughts, feelings
and/or actions of the audience, speaker, or other persons, and the speaker
has some design, intention or purpose of producing them (Austin 1962: 101
in Cobley (ed.) 2001: 235). Austin referred to the illocutionary uptake
meaning that the hearer is supposed to hear, register and react to what has
been said for a SA to ‘come off’:

A: ‘I promise I’ll come.’

B: ‘Ok, I trust you. (positive reaction)/ You always say that.(disbelief,
negative reaction)

A: ‘I’ll come.’
B: ‘Do you promise?’ (disbelief)/ Do you mean that? (idem)/ What do you
mean by that? (checking understanding or reacting negatively)/ Don’t
bother.’ (refusal or irony)

To Levinson, the perlocutionary aspect deals with ‘the bringing

about of effects on the audience by means of uttering the sentence, such
effects being special to the circumstances of the utterance’ (1983: 236).
Searle (1969) left perlocutionary effects largely undiscussed for two
major reasons: they are not part of the linguistic system and their
characteristics cannot be considered stable and predictable, but random.
Other pragmaticians tried to preserve the role of the notion because
all illocutionary act types must have certain effects typically associated to
them, though their emergence is not predictable; those effects can be

167
regarded as illocutionary points: assertives inform, questions elicit answers,
promises generate trust, directives make the hearer do something.
The same utterance can have different illocutionary force values;
getting the perlocutionary effect intended means getting the hearer
recognize the illocutionary force. This can be done by means of the
Illocutionary Force Identification Devices (IFID) and Felicity Conditions.
IFIDs include lexical items, such as PF verbs, adverbs etc, and also
grammatical means such as moods (indicative, imperative) and modal
verbs; also, paralinguistic means, intonation and stress can be indicators of
the illocutionary force of an utterance. Last, but not least, sentence type
can be considered an IFID.
It shouldn’t be overlooked that the context is the disambiguating
element which helps the users decode the utterance appropriately (see
also 5.4.)

5.3.2. Speech Act Schema (SAS)

Speech Act Schema was formulated by Bach & Harnisch (1979) to

account for the way in which the hearer derives the IF of an utterance; it is
an inferential chain:
L1 S is uttering some expression e
L2 S meant such and such by e L1 –L3 locutionary act
L3 S is saying that so and so

S is doing such and such in uttering e L4 illocutionary act

The levels involved are partly based on the SA structure as

described by Austin: they comprise the formal level (the sentence) + the
literal meaning level (semantic level) making up the phatic act –L2, the
speaker’s meaning – the rhetic act (L3) and the illocutionary level (L4).
The schema is based on semantic rules, but the IF of an utterance
depends on the speaker’s intentions and on the context of use, it can’t be
just a function of semantic rules. (see also 5.5.)

5.4. The concept of Speech Act in communication

A speech act was defined by Searle as the minimal and basic unit of
communication, based on constitutive (semantic) rules, leading to
conventional realizations. It is not the symbol, the word or the sentence, or,
even the token (occurrence) of those, which is the minimal unit of
communication, but rather the production or issuance by the speaker of the
symbol, the word or the sentence in accordance with the intentions of the
interlocutors (Searle 1969: 16 in Mey 1993: 111).

168
 In other words, generally, the speaker has the literal meaning as the
basis in producing a SA. Throughout time, some SAs have been
conventionalized in point of the correspondence structure-pragmatic meaning:

I promise to come back soon.

Some patterns have been automatically associated to a SA even when

we talk about indirect SAs, for example, questions used to make a request:

Can you open the window?

Mey agrees with Searle that it is the context which determines

whether an expression counts as a SA (Searle 1969: 52). Something counts
as something only within a specified set of rules. In producing a SA, the
speaker must have appropriate communicative intentions. This fundamental
idea is to be found in the description of the Felicity Conditions determining a
successful SA. The essential role of the speaker, makes the study of SAs the
study of langue, i.e. of language as used by the members of the linguistic
community, it is a functional study of language. Reference is made to
speaking, but in writing, the basic principles of successful communication
remain the same; speaking is a rule-governed form of behaviour. In
discourse analysis, the speech act is the minimal unit of the taxonomy:

encounter → phase → exchange → movement → speech act

The encounter can be said to represent the speech event,

comprising the replies and the circumstances of the event. Phases
represent the stages of the verbal interaction, the general term used for that
being (verbal) exchange. It implies all the replies and the properties of turn-
taking as part of cooperativeness.
A speech act (SA) should be defined against the background of the
speech situation (SS) and speech event (SE).
The speech situation is the situation of communication, the context
of utterance, which comprises the scene (the cultural context) and the
setting (the physical context).
A speech event, part of a speech situation, is the functional
dimension of the verbal exchange and is made up of speech acts. The SE
means the utterance, plus the circumstances (including other utterances);
the nature of a SE determines the interpretation of an utterance as
performing a certain SA (Yule 1996: 48). A SE is an activity in which
participants interact via language in some conventional way to arrive at
some outcome. Levinson quotes Bauman & Scherzer (1974) in referring to
the SE as a culturally recognized social activity in which language plays a
specific, and often rather specialised role (1983: 279). It can comprise a

169
central SA and other utterances ‘leading up and subsequently reacting to
that central action’ (Yule 1996: 57).
For instance a speech event involving complaining:

A: ‘This is the restaurant.’

B: ‘It looks strange. There are few people in here. The food smell is
unpleasant.’
A: ‘Now, you’re exaggerating things.’
B: ‘I’m just tired and I don’t think I can put up with it.’
A: ‘Give it time.’
B: ‘I don’t like it.’

The central SA is represented by the last utterance, but all the others
contribute to the interpretation of the descriptive part, either negatively (the
food smell is unpleasant, it looks strange) or adopting a neutral attitude
which conveys negative implicatures (there are few people because it’s a
bad restaurant). Initially, the sentences making up the circumstances are
indirectly expressing the message intended (both complaint and reproach),
for reasons of politeness, but B’s replies are interpreted in relation to the
speaker A’s intended message, as non-acknowledgements, not as
instances of cooperative behaviour.
Therefore, a speech act is the act of uttering something with a
certain intention and expecting a certain reaction from the interlocutor.
We give below some examples of corresponding speech situations,
speech events and speech acts:

SS SE SA
supermarket transaction offer/demand
conversation story assertion
wedding ceremony prayer invocation

5.5. SA classification: from Austin to Searle

A speech act classification is necessary because ‘we must ask, first

of all, when exercising our power of speech, what effects our speech acting
has, or can have, when performed in actual social, institutional or other
surroundings.’ (Mey 1993: 148)

5.5.1. Direction of fit

Austin (1962) refers to the onus of match and the direction of fit. The
onus of match involves matching two categories, X and Y: X can be

170
matched to Y, or Y to X. ‘Fit’ expresses the fact that our words both match
the world we live in, and that they, at least potentially, though not always
visibly, are able to change that world. (Mey 1993: 131)
The concept of direction of fit is to be used when discussing the
situations of matching words to a state of affairs, or vice versa.
Within the philosophy of the mind, the opposed terms are mind vs.
world (mental states are to be changed in accordance with the state of the
world), and world to mind (the state of the world is to be changed in
accordance with mental states). The first opposition is more general, the
second is narrower.
Beliefs, perceptions, hypotheses and fantasies are states with a
mind-to-world direction of fit, states that exist currently. In other words, if
the mental state is in some sense false or wrong, it should be changed in
accordance with reality.
Intentions and desires are states with a world-to-mind direction of fit,
states that don’t exist currently, but can exist in the future. Therefore the
existing state of the world should be changed sometimes, to fit the states of
the mind.

‘The term direction of fit refers to two ways in which attitudes can
relate propositions to the world. In cognitive attitudes [such as belief],
a proposition is grasped as patterned after the world; whereas in
conative attitudes[such as desire], the proposition is grasped as a
pattern for the world to follow. (Velleman 1992: 8)

Searle & Vanderveken (1985) assert that there are four possible
directions of fit in language:
- the word to world direction of fit: the propositional content of the
utterance fits an independently existing state of affairs in the world:
assertions illustrate that: They are tired.
- the world to word direction of fit: the world must change to match
the propositional content of the utterance; it is a possible state of affairs
which is envisaged; linguistically the conative attitudes are expressed by
using modals, and attitudinal verbs: You’d better sell the house! I want to
sell it! Will you sell it?
- the double direction of fit: the world is thereby altered to fit the
propositional content by representing the world as being thus altered;
Searle calls such sentences declarations: I declare you man and wife. I
declare the meeting open. We find the defendant not guilty.
An existing state of affairs, including the necessary circumstances,
cause the speaker to utter a certain sentence intended to change that state
of affairs (word to world direction of fit) and the state of affairs is changed by
the simple uttering of those particular words (world to word direction of fit).
- the null or empty direction of fit: there is no question of achieving
success of fit between the propositional content and the world, because the

171
success is presupposed by the utterance itself, which reflects the speakers’
emotional reaction to something real, to a fact: I’m happy that they came.
We summarize the above considerations in a table; it contains the
basic types of illocutionary acts:

SA type Fit direction Message Examples

intended
Declarations Words change S causes x I pronounce you
the world man and wife.
Representatives Words are S believes x I think you are
/assertions made to fit the right.
world

Expressives Words are S feels x I’m impressed!

made to fit the
world
Directives The world is S wants x Close the
made to fit the window!
words
Commissives The world is S intends x I promise to do
made to fit the that.
words

5.5.2. Illocutionary point

Searle (1979) introduces the concept of illocutionary point.

Illocutionary point is a parameter to distinguish classes of IAs:

SA type Illocutionary point

Declaratives To bring something about in the world

I declare the session open.

Assertives To represent a state of affairs

It’s raining.

Expressives Express a psychological state

I feel sad.

Directives Make the hearer do something

Come here!

172
Commissives The speaker commits himself to do something
I will come.

Interactants (i.e. interlocutors) need not recognize the illocutionary

point of each utterance. People are viewed as rational agents possessing
goals and plans designed to achieve those goals. Mutual awareness of this
fact prompts interactants to attempt to recognize each other’s plans and
goals, as well as possible obstacles to those goals (Cohen & Levesque
1990 in Holtgraves 2002: 20). Recognition involves general plans of action
designed to achieve particular goals and the discrete illocutionary forces.
There are SAs with the same illocutionary point, among which the
interlocutors make distinctions: the criteria involved to do that are:
-felicity conditions (see 5.5.3.);
-the degree of strength (i.e. the sincerity condition of the SA –it
refers to the speaker’s belief /desire strength) and the differences in the
specific means used for achieving a particular illocutionary point:
command/require/beg. (Holtgraves 2002: 15).
A small set of conditions, including ability and desire on the part of the
speaker, determine the illocutionary force recognition: implicit performative
verbs are activated during comprehension and are part of the
comprehender’s representation of the remark (Holgraves & Asheley 2001 in
Holtgraves 2002: 19).

5.5.3. Felicity (Happiness) Conditions

Descriptive utterances are true or false, performatives are felicitous

or infelicitous. Felicity Conditions are expected or appropriate
circumstances which ensure that the performance of a SA is recognized as
intended.

5.5.3.1. Felicity Conditions with Austin

Considering conventional procedures, Austin (1962) introduced

three types of felicity (happiness) conditions meant to make a SA
successful. The first type concerns the elements involved in the procedure,
the second refers to the manner of executing the procedure and the third
concerns the sincere intention of the participants to perform the act:
(A.1) There must be an accepted conventional procedure having a certain
conventional effect, that procedure to include the uttering of certain words
by certain persons in certain circumstances, and further
(A.2) the particular persons and circumstances in a given case must be
appropriate for the invocation of the particular procedure invoked.
(B.1) The procedure must be executed by all participants both correctly and
(B.2) completely.

173
(C.1) Where, as often, the procedure is designed for use by persons having
certain thoughts or feelings, or for the inauguration of certain consequential
conduct on the part of any participant, then a person participating in and so
invoking the procedure must in fact have those thoughts or feelings, and
the participants must intend so to conduct themselves, and further
(C.2) must actually so conduct themselves subsequently.

Therefore, normally, within a conventional speech event, there is a

conventional procedure which implies certain participants, certain words to
be uttered and certain circumstances characterizing the speech event.
Locutionary acts without illocutionary force occur when the speaker lacks
the intention to perform the respective illocution, if the speaker or hearer
are not of the right kind, or other circumstances do not meet the
conventions of the act.
If these conditions are fulfilled, then the SA has certain intended
effects; if not, it becomes a misinvocation:

e.g. during a marriage ceremony, the priest is not a person who has the
right to serve or one of the future spouses is already married; the baptizing
ceremony of a ship does not take place in an office or by uttering the
sentence ‘Your name is X’; during a trial a defendant is not officially
declared guilty or not guilty by his neighbours.

Generally, by convention, the procedure must be executed by all

participants completely and correctly, otherwise the act is a misexecution.
Both misinvocation and misexecution lead to a void act.
Last, but not least, the persons must have the designed feelings and
thoughts, otherwise the act is insincere, which makes it unfelicitous, but not
void. An exception, referred to by Mey (1993: 114-115), is the intention of a
spouse-to-be, if the marriage is to take place according to the rules of the
Catholic Church. The intention (expressed or not before or during the
ceremony) that includes the desire not to consummate the marriage or not
to have kids, or the withholding of one’s intention regarding the marriage
counts as a speech act which makes the marriage null and void. Another
example of an insincere act is a false promise that the speaker does not
intend to keep.

5.5.3.2. Felicity Conditions with Searle

For a SA to happen ‘felicitously’ or ‘happily’, and to prevent it from

‘misfiring’ there have been proposed some necessary and sufficient
conditions. Sas can be classified according to their basic properties
resulting from these conditions, which are also tests of their validity, i.e. of
success, since it is not a question of them being true /false.

174
 In analyzing the SA of promising, Searle considered two basic
questions, valid for any SA and cross-linguistically:
1. What conditions can be formulated for a SA to count as a promise?
2. What rules govern a successful use of this SA?
We took the liberty of adapting Searle’s conditions for promises so
as to account for any SA:

C1 Normal conditions must obtain for input and output (no jokes, no
acting, linguistic competence, no disabilities)
C2 The SA must have a content –the propositional content
C3 At the moment of uttering, the content must refer to a future,
possible action of the speaker
C4 What is performed refers to an action to the advantage of the
hearer
C5 The content of the utterance must not be something which will
happen anyway
C6 The sincerity of the speaker in carrying out the act
C7 The speaker intends to put himself under the obligation of
carrying out the promised act (recognition of an obligation to perform the
action of carrying out the act)
C8 The uttering of the words making up the sentence functioning as
a SA determines the hearer’s understanding of it as a SA. This is explained
by the fact that the circumstances of uttering must be conventionally right:
‘the speaker assumes that the semantic rules which determine the meaning
of the expression uttered are such that the utterance counts as the
undertaking of an obligation (Searle 1969: 61)
C9 The sentence is the one which by the semantic rules of
language is used to make that SA.

It is obvious that:
C4 and C5 are the preparatory conditions involving the participants,
the words chosen and the appropriate circumstances;
C6 is the sincerity condition
C7 is the essential condition, involving the commitment of the subject
C1-C3 are the general conditions mostly referring to the
propositional content of the utterance

These conditions made Searle formulate a series of rules characterizing the

act of promising, since Searle analysed this type of SA, but which can be
generalized. Searle’s intention was to link the performative verbs to felicity
conditions systematically. He proposes the following types of rule:
propositional content rules, preparatory rules, sincerity rules and the
essential rule. The propositional content rules concern the fact that many
acts seem to be about a proposition that something has happened
(thanking, reproaching etc) or will happen in the future (promising , ordering

175
etc). Preparatory conditions include all the circumstances necessary for the
act to be performed. The sincerity rule specifies to which respect the
speaker (and, maybe, the hearer, too) are required to be sincere. The
essential rule is in line with Austin’s observation that an utterance of a
certain type, under the appropriate circumstances, counts as a certain SA
for the members of the language community, as a result of a convention.

R1 The act is to be uttered only in the context of a sentence x and it

predicates some future act of the speaker.
R2 The uttering is determined by the preference of the hearer for
the speaker to do the act.
R3 The act is performed only if the action will not take place in the
natural course of events.
R4 The speaker must intend to act.
R5 Uttering counts as the undertaking of an obligation to do the
action.

Semantic rules determine what a sentence may mean, not what it

must mean. The IF is determined in context. Most conditions enumerated
by Searle can be criticized as being irrelevant because the speaker has no
control on them: there can’t really be made a differentiation (by the hearer)
between a possible action and one which will be definitely performed by the
speaker. The users should be able to make the distinction between a real
performative, on the one hand, and polite, conventional, ritual expressions,
on the other. It is a difficult task, and, ultimately, we should rely on the
psychological profile of the speaker, on his/her relationship with the
interlocutor, and on a whole series of contextual factors (time, place,
obstacles in performing the act etc):

‘Don’t postpone your visiting us, again! ‘I promise.’

‘I declare open the plenary session of the 15-th Congress of….’ (verbs
expressing a verdict, a decision are ambiguous as to their performativity)
I welcome you. (welcome is described as a semi-descriptive verb, unlike
the structure to bid welcome, considered performative)

Even a performative verb such as promise can be used as a non-

performative if it refers to a habitual action. We adapted an example from
Levinson (1983: 133):

‘How do I make you convince me to throw all these parties?’ ‘I [always]

promise to come./ [Each time] I promise to come.’

Generally, criticisms of Felicity Conditions are linked to the

performativity of utterances, therefore the arguments are somehow common to
those used in discussing performativity and PF verbs (see 5.3. and 5.4.)

176
5.5.3.3. Counterarguments to Felicity Conditions frame

Leech (1983: 171-172) believes that it is unnecessary for

pragmatics to include special speech act rules, such as those stipulated by
Searle: essential conditions, preparatory conditions and sincerity conditions
of illocutions do not have to be independently stated, they can be arrived at
just corroborating the sense of the utterance and conversational maxims.
We present some counterarguments to felicity conditions frame:
-felicity conditions may characterize non-performative examples, too:

Jane’s brother is brown-haired.

When Jane has no brother, it is neither true, nor false, but void.

- a non-performative utterance can be insincere, including the case when

insincerity is determined by the speaker’s desire to be polite, the result
being the ‘white lies’ meant to protect the face of the hearer:

Jane’s dress, the one that you bought for her, suits her fine. (when the
speaker knows that the dress looks awful on Jane)

-truth/falseness considerations may affect performatives, too:

I warn you that the box will fall off that wardrobe.
If the speaker does not appreciate distances well, his warning can prove
false if the box does not fall off; we don’t think that it is relevant that the
misleading effect is unintentional (the speaker is sincere in expressing the
warning, but it’s a false warning, unless reality proves the speaker to be
right, which is not the case.) The issue here is that the truth or falseness
concerning the proposition referred to by the SA is not in contradiction with
the speaker’s sincerity in performing the SA.
Felicity Conditions are a consequence of applying semantic
concepts (the propositional aspect of meaning, or the literal meaning) in
defining both descriptive and PF uses. The felicity conditions of various
SAs are part of the meaning of the PF verbs. The IF is semantic in the
truth-conditional sense and is fully specified by the meaning of the PF
clause; there appear problems with assertions and declaratives, though.
PF utterance theory is sustained by syntactic and semantic reasons,
but has its flaws.

5.6. Indirect Speech Acts (ISAs)

The notion of ISA makes sense only if one subscribes to the notion
of literal force, i.e. to the view that IF is built into sentence form (Mey 1993:
263). Literal force will be a consequence of Performative Hypothesis (PH).

177
 An indirect illocution is a case of a sentence ‘masquerading’ as a
sentence of a different type (Austin 1962). Following this line of reasoning,
a sentence has a literal force determined by semantic rules and an inferred
indirect force.
Continuing in Austin’s line, for Searle, indirect SAs have a double
illocutionary force. For instance in the example Can you do that? the
primary illocutionary act is that of request, it is what the utterance stands
for, and the secondary illocutionary act, functioning as prep condition for
the primary one, is that of question. According to Grice, the former is the
non-natural meaning and the latter the natural meaning. The pre-condition
is: the hearer can perform the action. The content condition is: reference is
made to a future action of the hearer. Asking about preconditions doesn’t
come as a direct request, but makes the hearer react as if that had
happened: the variant chosen is more tactful, more polite and saves the
face of both interlocutors.
(see also 6.6. Politeness and (indirect) speech acts)

Leech (1983: 195) disagrees with Austin, analyzing the latter’s

examples and considering that:
- representing the indirect force by means of a PF verb is a simplification
which does not account for the complexity and subtlety of human
communication;
- indirect IF is a matter of degree:

Can you help me?

Can you help me, please?
Would you mind closing the window?
Close the window, would you?
Can’t you close the window?

- the relationship sense-force should be considered from a functional

perspective;
- direct and indirect force are not an illustration of grammatical ambiguity,
but a matter of two coexisting meanings.
Levinson claims that most usages are indirect (Levinson 1983: 264)
and tries to analyse critically the solutions proposed. ISAs often have
syntactic or, at least, distributional reflexes associated not only to their
surface-sentence-type, but also with their indirect or effective IF.
The distribution of please, obviously, I believe, etc proves that they
are restricted to utterances sharing the IF of an assertion, no matter the
sentence type:

I want you to leave, please.

Can you, please, leave?
Please, leave!

178
In the three examples above, the assertion is I want you to leave; that is the
IF indirectly conveyed in the form of an interrogation or of an imperative
sentence.

May I tell you that I believe that you’re wrong? (tell is a PF verb, but, in this
case, it is used in an interrogative main clause containing a modal
expressing permission to help expressing an indirect IF; the assertion
which represents the IF of the utterance is I believe that you’re wrong; it
appears as a subordinate clause. That is the proposition, i.e. the logical
semantic content that the SA refers to. These comments remain true for the
following example:

Let me tell you that I believe you’re wrong.

The modifier in the last two examples (the parenthetical clause I believe)
seems restricted, according to Levinson, to utterances that have the force
of an assertion, whatever the sentence type of the linguistic expression that
performs the assertion. The sentence types in this case were interrogative
and, respectively, imperative.
Two basic kinds of theories have been proposed ‘to rescue Literal
Force Hypothesis’ (postulating the existence of the literal force of a
sentence, based on its literal meaning): the idiom theory and the inference
theory (Levinson 1983: 268). For example, the former implies that a
structure of the type Can you VP? is a standard format for indirect requests
and shall be treated as an idiom, i.e. as a whole, not analysed into
components. According to this theory, there would result an infinite list of
such idioms, and, also, structures can be ambiguous, or both readings
could be possible and the hearer might not know which to choose:

‘Can you fetch me a glass of water?’ ‘Yes, I can. (Here you are.)’

The inference theory states that the indirect illocutionary force

should be inferred whenever the literal meaning is blocked by the context:
Can you fetch me a glass of water? (the speaker is aware that the hearer is
physically apt to fetch the water, that there is running water in that place,
there are glasses, the hearer has the linguistic competence to understand
the words, etc).
Obviously, the conclusion is that there is a pragmatic conditioning of
some syntactic or distributional processes. Levinson argues that:

‘Illocutionary force belongs firmly in the realm of action and the

appropriate techniques for analysis are therefore to be found in the
theory of action, and not in the theory of meaning, when that is
narrowly construed in terms of truth-conditional semantics.’ (1983: 246)

179
 An argument in favour of the pragmatic approach of SAs is a quotation
from Mey: ‘if we want to be secure in our expectations, we should concentrate
on the people who promise, than on what they say’ (Mey 1993: 115).
Interlocutors need not recognize the illocutionary point of each
utterance, but each other’s plans and goals, possible obstacles to those
goals (Cohen & Levesque 1990 in Holtgraves 2002: 20) and the discrete
illocutionary forces. (see also 5.5.2.)
Levinson (1983: 274-275) refers to conversational postulates and to
the process of mapping SA force onto sentences in context. For example,
to state or question a Felicity Condition on a SA where the Literal Force of
such a statement or question is blocked by context counts as performing
that SA:

I want more ice-cream. (this is a speaker-based felicity condition, the literal

force is not to inform in the speaker’s wish, it is not an answer to the
question ‘What would you like to have in life?’; it is a condition of making a
request, a request implies that the speaker desires the required thing.)

Can you pass me the ice-cream? (it is a hearer-based Felicity Condition;

the ability of the hearer is implicit, it is a condition, not the topic of a genuine
question; see the example Can you fetch me a glass of water? for
similarities)

IF can be viewed as entirely pragmatic and has no direct and simple

correlation with sentence-form and sentence-meaning: there are no isolable
necessary and sufficient conditions for questionhood, for instance, and
explicit performatives can be assigned truth-conditions that are as general
as is consistent with their actual use.

Conclusions

A SA is generally acknowledged by linguists as the basic minimal

unit of analysis in pragmatics, and its three levels (locutionary, illocutionary,
perlocutionary), are the starting point in discussing the link sentence form -
sentence meaning - sentence IF. The form of the sentence and its literal
meaning are realities from which the IF can be derived, according to a
series of Felicity Conditions, or the IF is to be distinguished from the literal
meaning and linked to the speaker’s competence and intentions and to the
entire context. The strategies of verbalizing a SA vary not only from culture
to culture, but also from user to user, changing conventional patterns being
a proof in this respect.

180
 6. POLITENESS

6.1. The concept of politeness

6.1.1. Politeness − a social phenomenon
6.1.2. The conversational maxim view on politeness
6.1.3. The face-saving view
6.2. Politeness ‘in its own right’. Politeness Principle and its maxims
6.3. Politeness and deixis
6.3.1. Person deixis markers and politeness
6.3.2. Social deixis and politeness
6.4. Politeness and conversational implicatures
6.5. Politeness and presuppositions
6.5.1. The problem of negative ambiguity
6.5.2. Structural presuppositions and hearer’s manipulation
6.6. Politeness and (indirect) speech acts
Conclusions

6.1. The concept of politeness

6.1.1. Politeness − a social phenomenon

Politeness can be seen as a social phenomenon based on the

social values of a particular community. This corresponds to the social
norm view on politeness, which reflects the historical understanding of the
phenomenon; that is what is called ‘first-order politeness’, different from
‘second-order politeness’ which is a theoretical construct: ‘This normative
view considers politeness to be associated with speech styles, whereby a
higher degree of formality implies greater politeness’ (Fraser 1990: 221).
Social values determine the social status of the participants in the
verbal exchange, better said, their perception and assigned value. In its
turn, social status shapes the influence (established prior to interaction) that
external factors, such as age and power, have on the verbal exchange.
Other factors, such as amount of imposition, or friendliness degree, are
often negotiated during an interaction (Yule 1996: 59). Both types of factors
determine what we say and how we interpret utterances.
In spite of the importance of the concept of social values, Mey
draws attention on the fact that language (and, we would add, the language
user, by extension) is not supposed to be virtuous in the moral sense (Mey
1993: 67). Certain forms of behaviour are preferred and rewarded, others
are subject to sanction, but these are not moral issues (Mey 1993: 74-75).

181
The social values talked about are not rigid moral values, but the complex
result of the type of culture characterizing a community, itself subject to
changes in time.
Social values are, to a large extent, cultural values, even if the moral
criterion might be the cause of some cross-cultural values. At least in some
respects, cultures can be seen as a macroscopic family-like structure, its
members behaving much like members of a family, and cultural values can be
understood like points on a continuum, in the sense that the degree of
importance of a common value and the (linguistic) behaviour it triggers varies.
Nevertheless, each culture has some systematic and repetitive
characteristics which make up a cultural pattern, reflecting the conditions
that contribute to the perception, way of thinking and lifestyle of a
community.
Depending on the type of culture, several factors shape the social
values and, thus, also the profile of a verbal exchange. Hofstede (1980)
enumerated four such factors, organized as antonymic pairs:
- individualism vs. collectivism: concern for individual rights vs.
concern for duty, for what is owned to the group; collectivism triggers
indirectness, individualism triggers directness and explicitness (see Hall’s
criterion below);
- power distance: there may be a preference for unequally
distributed power (tendency for lower power distance or for large power
distance); in large power distance cultures, it is believed that people are not
equal in this world and that social hierarchy is prevalent. In these countries’
organizations and institutions there is a greater centralization of power,
great importance being placed on status and ranks. At the other pole, in
small power distance cultures people believe they are close to power and
should have access to that power. Hierarchy is a convention implying
inequality of roles, but subordinates consider superiors to be the same kind
of people as they are, and vice versa.
- uncertainty avoidance: it means avoidance of the nervousness at
unstructured, unclear and unpredictable situations, situations which tend to
be avoided by maintaining strict codes of behaviour and a belief in absolute
truths. Low uncertainty avoidance countries accept uncertainty in life and
tolerate the unusual; members of these cultures are more willing to take
risk, are more flexible, are less tense and more relaxed.
-masculinity vs. femininity; male oriented traits are ambition,
achievement, acquisition of money; in a masculine society, men are taught
to be domineering, ambitious and assertive. Cultures that value feminity
tend toward a feminine world view by stressing caring.

To these criteria, Hall (1959) adds the role of the context (including
gestures, silence, and the use of space): there are high context cultures and
low context cultures. For high context cultures, meaning does not need to be

182
conveyed by words. High context culture members are more reliant on and
tuned in to, non verbal communication. Status (age, sex, education, family
background, title, affiliations) and an individual’s friends and associates also
add meaning to what is said. In low context cultures most of the information
is contained in the verbal message and very little is embedded in the context
or in the participants; communication style tends to be as direct and explicit
as possible.
Three other criteria cannot be overlooked:
- formality vs. informality; formality is manifested at the level of
verbal communication, by using titles, honorifics; on the other hand,
informality concerns postures, way of dressing, and, most of all, a verbal
style in accordance with non-verbal communication: avoidance of the use
of titles and honorifics and preference for idiomatic, colloquial speech, first
name address, informal greetings;
- assertiveness has to do with masculine societies: being assertive,
i.e. expressing oneself by presenting events as factual, certain, or almost,
gives the impression of control, of competence, of safety. Politeness, on the
other hand, makes people be more tentative and use non-assertive
markers (moods, lexical items) in order to win the interlocutor’s
benevolence and willingness to act in the direction intended by the speaker:

It is best if you do your homework now! cf It would be maybe better if you

thought to do your homework now rather than later.

-interpersonal harmony has to do with the equilibrium to be

maintained throughout a conversation, by permanently negotiating meaning
and attitudes, beliefs, impressions, etc.
The Anglo-Saxon society tends to be rather individualist, assertive,
masculine, informal - except for official events -, favouring small power
distance, at least apparently, and showing low uncertainty avoidance. The
British society tends to behave more like most European cultures,
emphasizing more formality and power distance.
The Romanian society, on the other hand, is rather a mixture of
individualism and collectivism, is masculine, and favours high power
distance, high uncertainty avoidance and formality.
To summarize, from this perspective, politeness is a concept
comprising a set of principles regulating the communicative behaviour in a
certain cultural space.

6.1.2. The conversational maxim view on politeness

It is based on the work of Grice (1975), on his foundation of The

Cooperative Principle (CP) and on Leech’s formulation of the Principle of
Politeness (PP).

183
 Some politeness phenomena have been considered indirectly by
Grice, whose Cooperative Principle and maxims of conversation were
formulated on the assumption that the main purpose of conversation is ‘the
effective exchange of information’ (Grice 1989: 28). CP may not be directly
related to politeness but its formulation has constituted a basis of
reference on which other principles, such as The Politeness Principle and
its maxims, have been stated.
Leech adopts Grice’s construct of conversational principles and
elaborates a thorough analysis of politeness in terms of principles and
maxims, seeing politeness as a regulating factor in interaction, and the PP
as a necessary complement to the CP. His PP is constructed in a very
similar format to the CP and is analysed in terms of maxims: the Tact
Maxim, the Generosity Maxim, the Approbation Maxim, the Modesty
Maxim, the Agreement Maxim, the Sympathy Maxim. Leech’s maxims are
related to the notions of cost and benefit and each of them is stated as a
pair of submaxims (see 6.2.). Politeness appears as the key pragmatic
phenomenon for indirectness in order to obtain the perlocutionary effect
intended. That would be one of the reasons why people deviate from the
Cooperation Principle. Leech (1983) considers that we should minimize the
effects of impolite statements (negative politeness) and maximize the
politeness of polite illocutions (positive politeness), bearing in mind the
intentions that accompany all conversations (Leech1983: 80).
Moreover, the PP and the CP can conflict. If the speaker sacrifices
the PP in favour of the CP, he will be putting at risk the maintenance of ‘the
social equilibrium and the friendly relations which enable us to assume that
our interlocutors are being cooperative in the first place’ (Leech 1983: 82).
But Mey considered that ‘it is not at all plausible that PP is able or ever
needed to rescue CP; CP may not even need to be rescued’ (Mey 1993:70).
Mey also criticizes Levinson in point of the general rationality of
cooperation between humans as a general, inviolable and indisputable
maxim; also, cultures can be very different with respect to cooperative
behaviour, and, last, but not least, certain forms of behavior are preferred,
or even rewarded, others, subject to sanctions, beyond moral principles
(Mey 1993: 74-75). (see also 6.1.1.)
Another criticism (Wierzbicka 1985) is that studies on English
should not be considered as proving universal principles of politeness.
Features of English, which have been claimed to be due to universal
principles of politeness, are shown to be language-specific and culture-specific.
Linguistic differences are associated with cultural differences such as
spontaneity, directness, intimacy and affection vs. indirectness, distance,
tolerance and anti-dogmaticism.

184
6.1.3. The face-saving view

It was proposed by Brown and Levinson (1987) and has been, up

to now, the most influential politeness model. It converges in many ways
with the conversational-contract view presented by Fraser (1990):

‘Upon entering a conversation, each party brings an understanding of

some initial set of rights and obligations that will determine, at least
for the preliminary stages, what the participants can expect from the
other(s). During the course of time, or because of a change in the
context, the two parties may readjust what rights and what obligations
they hold towards each other’. (Fraser 1990: 232)

Returning to the face-saving view, considering the concept of face -

‘public self image’, politeness comprises the means used to show
awareness of another person’s face that is to be described in relation to
social distance/closeness (Brown & Levinson 1978). Respect and
deference are opposed to friendliness, camaraderie and solidarity. Relative
social distance is connected to face wants. As a result of that, within
everyday social interaction, the speaker may say something which
preserves his face and/or that of the interlocutor, or something that
threatens it.
(see also 2.5.; 6.3.2.)

6.2. Politeness ‘in its own right’. Politeness Principle and its maxims

Leech (1983) stated the Politeness Principle as having two

submaxims:
Minimize the expressions of impolite beliefs.
Maximize the expression of polite belief.
The principle was rather general, cross-cultural, and meant to
complement the Cooperative Principle. The PP is supposed to operate on
the same level as, and to collaborate with, the CP. Politeness maxims are
neither theoretically, nor practically, on the same level as the CP, but can
be subsumed to it, since they ensure cooperation, the basis of
conversation. (Mey 1993: 70)
Its component maxims give a better view of what Leech (1983)
intended to explain in this way:

The Tact Maxim

The Tact Maxim (TM) consists of (a) Minimize cost to other (b)
Maximize benefit to other.
This maxim is applicable in impositives (e.g. ordering, requesting,
commanding, advising, recommending) and commissives (promising,
vowing, offering).

185
 The first aspect to be considered is the size of imposition to the
hearer, and that can be reduced by using minimizers such as just, a
second, a bit of, as in these examples:

Wait a second!
We’ve got a bit of a problem here.
I think you are a little unfair.

Imposing less can mean offering optionality. Allowing options or,

better said, giving the appearance of allowing options is a measure of
politeness. Otherwise, the interlocutor would feel under pressure and might
react contrary to the speaker’s expectations:

Please, teacher, could you check my spelling mistakes?

The third aspect of the Tact Maxim is the cost/benefit scale: if

something is perceived as being to the hearer’s benefit, it can be
expressed politely without employing indirectness:

Have a beer! cf May I offer you a beer? or I authorize you to order a beer
which will be paid by me.

The next example combines various linguistic devices in order to

observe the tact maxim:

You know, I really do think you ought to move out. It’s costing you too much
money to keep this apartment.

The speaker minimizes the cost to the addressee by using two discourse
markers, meant to diminish the imposition on the interlocutor: one appeals
to solidarity (you know) and establishes a common ground: the speaker
tries to convince the interlocutor that his opinion is, in fact, similar to the
way the interlocutor thinks. The formal modal verb (ought) gives options to
the interlocutor and stresses the idea of necessity, not of subjective
imposition; the modifying hedge (really) and the attitudinal predicate (think)
belong to the epistemic domain and are correlated to the explanation
explicitly uttered by using an assertive utterance: the speaker maximizes
the benefit to the addressee in the second part of the turn by indicating
that (s)he could save a lot of money by moving out.

The Generosity Maxim

The same application characterizes the Generosity Maxim (GM)

which consists of submaxim (a) Minimize benefit to self and submaxim (b)
Maximize cost to self.

186
 Leech points out that some cultures attach more importance to the
Generosity Maxim than others, suggesting that expressing generosity
linguistically is particularly important in Mediterranean cultures.
This maxim prevails over the indirect illocutionary value considered
more polite in invitations; this justifies the use of the imperatives in the
examples below:

You must come and spend the night with us if your house is being
repaired.
Help yourself with some cakes.

or when the speaker gives a piece of advice:

It’s none of my business, really, but you’d better change the reservations.
That restaurant is by no means good.

In the first part of the utterance, the speaker reduces his benefit by using a
parenthetical clause, this strategy being stressed by the use of the hedge
really. His authority, i.e. his benefits, are again diminished by the use of a
hypothetical form ’d better; the objective reasons are presented explicitly in
a separate assertive sentence: That restaurant is by no means good.

The Approbation Maxim

The Approbation Maxim (AM) is applicable in expressives such as

thanking, congratulating, pardoning, blaming, praising, condoling, etc., and
in assertives like stating, boasting, complaining, claiming, reporting. Its
submaxims are: (a) Minimize dispraise of other and (b) Maximize praise of
other:

It was very thoughtful of you to invite us here. (the speaker maximizes

praise of the addressee)
I was wondering if you can prescribe me something else. I’m finding it very
hard to tolerate this medicine. (the speaker minimizes dispraise of the
addressee)

The action of this maxim is obvious if we think that we prefer to

praise others; if not, we can change the subject, give some minimal sort of
response (Well, …) or simply remain silent.

The Modesty Maxim

The Modesty Maxim (MM) only applicable in expressives and

assertives, consists in submaxim (a) Minimize praise of self and submaxim
(b) Maximize praise of others.

187
Excellent! What a wonderful painting! I wish I could paint as well as that.
(the speaker belittles her/his own abilities in order to highlight the
achievements of the addressee)

This maxim varies in its application from culture to culture: in

English-speaking societies, it would be customarily more polite to accept a
compliment graciously (e.g. by thanking the speaker for it) rather than to
go on denying it. English speakers would be inclined to find some
compromise between violating the Modesty Maxim and violating the
Agreement Maxim.

The Agreement Maxim

The Agreement Maxim (AM) is only applicable in assertives: (a)

Minimize disagreement between self and other and (b) Maximize
agreement between self and other.

I know we haven’t always agreed in the past and I don’t want to claim that
we are friends, or anything, but I believe we can solve this matter together
if we want to.

The Sympathy Maxim

The Sympathy Maxim (only applicable in assertives) points to the

speaker making an effort to minimize the antipathy between himself and
the addressee and to maximize sympathy:

Despite our very serious disagreements, I have done my best to

understand your points of view, but, so far, I see that we haven’t been able
to find any common ground.

The maxims can be viewed as making up three pairs, since each

pair of maxims applies to the same type of utterances, is based on the
same criterion and represents the perspective of the speaker/hearer, with
the exception of the pair agreement maxim - sympathy maxim, which
regard both the speaker and the hearer, equally; the table below illustrates
those aspects:

188
 Maxims Criterion Perspective Tendency Type of
utterances
where it
applies
Generosity cost-benefit benefit of self ↓ Impossitives
Tact scale benefit of other ↑ Commissives
Modesty praise scale praise of self ↓ Assertives
Approbatio praise of other ↑ Expressives
n
Agreement Agreement agreement of ↑ Assertives
+sympathy* other
Sympathy scale disagreement ↓
of other
sympathy of ↑
other
antipathy of ↓
other

* sympathy can be seen as a step further once agreement is reached;

sympathy is more subjective, and, if genuine, not achievable if there is
obvious disagreement.

6.3. Politeness and deixis

6.3.1. Person deixis markers and politeness

Personal pronouns can be interpreted as politeness markers in

association with verb forms and with addressing terms. Such combinations
convey the whole structure they are part of a certain degree of politeness.
English does not have politeness pronouns and does not use the plural
verbal forms in order to mark distance in relation to the interlocutor.
Therefore, pronominal forms are to be interpreted strictly in relation to the
other nominal phrases whose semantic structure includes politeness
semes; in the example below the pronominal forms of the emphatic
pronouns themself, ourself were created to express the plural of modesty,
which can be interpreted as an expression of politeness in relation to the
hearer:

‘You won’t be the first or last man or woman who gets themself involved in
a holiday romance. We find ourself…’ (Biber 1999: 340)

189
 The following examples can be translated into Romanian in various
ways, which denote the differences and similarities between two structurally
different languages:

Who are you? cf Rom. Cine eşti tu? Cine eşti dumneata?Cine sunteţi
dumneavoastră?

Only the linguistic context (nominal phrases denoting the referent)

and the situational context, meaning the relationships among the
interlocutors can provide the necessary information in order to interpret the
utterance appropriately in terms of formality level. Formality features are
not explicitly marked in English in such a sentence and they are to be
inferred considering the elements mentioned.

6.3.2. Social deixis and politeness

Social deixis is ‘that aspect of sentences which reflect or establish or

are determined by certain realities of the social situation in which SA occurs’
(Fillmore 1975: 76 in Levinson 1983: 89); it is concerned with the meaning and
grammar (e.g. honorific concord) of linguistic expressions (Levinson 1983: 93).
The rather constant element is the social identity of the participants (see 2.5.
social deixis). Besides that, social deixis is concerned with the social
relationship between the interlocutors or between them/one of them and
persons or entities referred to; Levinson considers relevant only those which
are grammaticalised: polite pronouns and titles of address. All titles of address
and vocatives ‘seem invariably marked for speaker-referent relationship: there
is no such thing, it seems, as a socially neutral summons or address’
(Levinson 1983: 93).
Social deictics are absolute or relational. Absolute deictics include
authorised recipients (Your Honour, Mr. President) and self-reference
(≠ authorized speaker, to express self-irony):

Our Majesty is sad today.

Relational social deictics have the speaker as the source of the

attitude of social distance or closeness which manifests itself in relation to a
referent which may be: the addressee (but the addressee is not necessarily
the target of respect), bystanders or the setting. Social deictics range from
honorifics, to dishonorifics and intimacy markers. Only this type of
relationship can be encoded in the form of referring expressions and
morphological agreement with them. Honorific concord can and does
become a topic of morphology.
Sometimes, the same expression can express different politeness
degrees:

190
My lady, please, do us the honour of accepting our present! (the term of
addressing expresses deference cf Rom. ‘Excelenţă/ Alteţă, faceţi-ne
onoarea de a acepta darul nostru!’)

You are talking to a lady. (a ‘standard’ level of politeness is expressed

here, where the sense of the lexeme lady contains the semes [+human],
[-male] and the seme [+respect] cf Rom. ‘Vorbeşti cu o doamnă.’)

This lady doesn’t know when to stop. (the lexeme lady becomes a
dishonorific cf Rom. ‘Femeia asta/ cucoana asta nu ştie când să se
oprească.’

Lady, mind your businesss! (the degree of disrespect increases compared

to the previous example, to express the opposition between the sense of
the lexeme lady and the context of use, including the referent’s behaviour in
relation to the speaker’s expectations and background, cf Rom. ‘Vezi-ţi de
treaba ta, cucoană!’)

I’m listening, my lady, said the girl, looking at the cat. (the addressing term
becomes a term of endearment, an intimacy marker, when associated to a
[-human] referent, the cat, for whom the speaker feels an attachment and
whose authority she also ironizes affectionately, cf Rom. ‘Ascult, stăpână, …’).

Diminutives can be a marker of positive politeness, of solidarity, in

some cultures more than in others, as Wierzbicka (1985) states:

‘Rich systems of diminutives seem to play a crucial role in cultures in

which emotions in general and affection in particular is expected to be
shown overtly. By contrast, the Anglo-Saxon culture does not
encourage unrestrained display of emotions and this could explain
why expressive derivation has not developed to that extent in
English.’

Hey, kitty, kitty, kitty!

Who’s your daddy?

The problem of the relation between presuppositions and social

deixis is tackled by Levinson in the line of Keenan (1971). Levinson defines
pragmatic presupposition as the relation between a sentence and the
appropriateness of a sentence in a context, adding that this type of
inference is an aspect of social deixis encoded as a conventional
implicature (1983: 177):

191
Tu es Napoléon.
Vous êtes Napoléon.

The two examples have the same presupposition (Napoléon exists/

existed, since we don’t know what referent Levinson had in mind, the
Napoléon, i.e. the emperor, or any other male referent bearing the same
name), but the variant chosen is appropriate in the context or not, i.e. it is
relevant or not, depending on the relationship between the interlocutors,
and, consequently, on their communicative expectations.
Social distance or closeness can be considered as informative
elements necessary to observe the maxim of quantity: making one’s
contribution as informative as necessary would mean including linguistic
elements which to mark the acknowledgement of the social
distance/closeness. Failure to do so might cause the lack of
cooperativeness of the interlocutor.

6.4. Politeness and conversational implicatures

Green (1989) notices that the maxims have various weightings in

people’s minds. A greater moral value is attached to the maxim of quality
than to the others: ‘violating it amounts to a moral offense, whereas
violating the others is at most inconsiderate or rude.’ (Green 1989: 89 in
Mey 1993: 77)
When discussing Grice’s maxims, we provided an example for
generalized conversational implicatures in Yule’s view (see under 3.2. the
example ‘Did you invite Bella and Cathy?’ ‘I invited Bella.’). The same
example can prove that observing the quality maxim is a proof of
politeness, and, to maintain the level of politeness expected in point of not
saying something that you know it’s not true, the speaker is prepared to
flout the maxim of quantity:

‘Did you remember to buy bread and milk?’ (= Did you buy …) ‘I brought
bread all right.’

Horn’s example quoted by Mey (1993: 79), (see 3.3.),

I cut the finger69 yesterday.

is, in our view, linked to the Politeness Principle in point of not bothering the
interlocutor with disturbing details, avoiding a more elaborate and
impressive variant, such as I accidentally hurt myself awfully while cutting

69 The finger, i.e. my finger; the exceptions (the finger = his/her finger) are to be treated as such.

Anyway, the definite article means definite reference, the referent being recovered in the context.

192
… Horn’s R(elation) Principle – say no more than you must – can be
interpreted as corresponding to politeness requirements, in some situations.
An example such as

I saw John with a woman at the opera last night, but maybe she was his
sister.
can be an illustration of flouting the maxim of quantity, if the woman was,
indeed, John’s sister, in which case the NP a woman is inappropriate in the
context. The first sentence implies the presupposition the woman was not
his girlfriend/ fiancée/ wife etc, therefore John presumably had a date. This
presupposition has a positive or a negative connotation, depending on the
context, but, irrespective of that, it is anyway suspended by the second
sentence. Nevertheless, the connotative value of appraisal, in case John is
not married and the speaker is a friend encouraging John to go on dates,
means neutrality in point of politeness. But, if John is married, the whole
sentence threatens John’s positive face.
The link between politeness and conversational implicatures is also
established by discourse markers. The following example is from Smith
&Wilson (1990: 180) and we generally took over their comments regarding
the values of the parenthetical structures placed in front position, because
they are native speakers of English, exposed to the patterns of use of such
elements. Nevertheless, there can be other points of view regarding the
possible interpretations. (see 6.2. politeness maxims).

A: I really disliked that man you introduced me to.

B: 1. Actually, he is your boss.
2. Anyway, he is your boss.
3. After all, he is your boss.
4. Still, he is your boss.
5. Well, he is your boss.

1: ‘reconsider your attitude because the perspective is different, you are

compelled to do so’; it is a discourse marker which functions as a
disagreement minimizer, focusing on the objective reasons: he is your boss;
2: ‘it’s irrelevant’; technically, it expresses the same pragmatic message as the
previous utterance, only that the authority of the speaker seems stronger,
since it is an element which minimizes antipathy; the idea of the irrelevance of
the reaction is present in both first two examples, but speaker B tries to save
his face;
3: reproach: ‘you say you dislike him, but it is because he is your boss; you
have to tolerate him, why do you expect to like him?’ Speaker B is losing
face by using that discourse marker, even if he uses an antipathy
minimizer. He threatens A’s face by being more authoritative.
4: ‘make the most of it’; the hedge still is a tact marker;

193
5: ‘you’ve got problems if you think like that’; the speaker is rather neutral,
i.e. he does not express disagreement as strongly as in the previous
variants; the last two variants are milder in our view: the speaker tries to
save both his face and that of the hearer. The last variant is the mildest.
Our comments are just suggestions; we are aware that discourse
markers can express some conventionalised attitudes of the speaker within
the linguistic community, but there may be variations due to the context, too.

6.5. Politeness and presuppositions

6.5.1. The problem of negative ambiguity

Returning to the problem of existential presuppositions and their

resistance to negation, theoretically, they can be cancelled, but practically,
it is background assumed knowledge that they are true. Asserting
something about something else has as a precondition (presupposition) the
existence of the entity referred to.
In real life communication, politeness can determine the
confirmation of a state of affairs implying the presupposition, followed by its
cancellation:

‘Have you visited Fairytown?’ ‘No, actually, there is no such place70.’

We go further and dare assert that there may be cases of affirmative

sentences whose existential presuppositions are cancelled similarly:

‘How is your brother?’ ‘Fine, but I have no brother.’

The reply contains two parts: the speaker makes a FSA, seeking for
acceptance fine, followed by the FTA expressed by the second sentence; it
is not the relevant point that, in fact, the first part of the reply is neither true,
nor false, since it does not correspond to reality, because at the moment
when it is heard by the first speaker, for a second or two, it is taken as
such, i.e. as corresponding to a real situation. The intention of the speaker
may be, in some cases, to be ironical, even sarcastic, in relation to the
interlocutor who asks an inappropriate question (under the circumstances)
that he might not be aware of – in this case, that the interlocutor does not
have a brother and he misses that, or he had one, but he is dead, etc.

70
We refer stricly to a real place, not to fictional worlds, to possible worlds in
general, or to metaphors of the actual world.

194
 To illustrate cases of affirmative sentences whose existential
presuppositions are cancelled, we adapt an example from Yule (1996: 36):

A man and a woman are sitting on a bench in a park. Nearby, there

is a dog sitting rather close to the woman. The man asks:
Man: Utterance 1 ‘Does the dog bite?’
Lady: Utterance 2 ‘No.’ (the dog bites the man.)
Man: Utterance 3 ‘You said it doesn’t.’
Lady: Utterance 4 ‘It doesn’t, but this is not my dog.’

U1 Shared knowledge: dogs are not always calm in the presence of

strangers.
U1 Context circumstances: the dog sits next to you and it is calm.
U1 Referent recoverability: Considering the above, I presuppose it’s
yours.

The first speaker doesn’t observe the maxim of quantity, because in

this case, the (i.e. this dog) ≠ your dog. Linguistically, the definite article
and the possessive adjective belong to the class of central determiners,
having anaphoric situational function in this case; pragmatically, it is a
background assumption that when the speaker uses the in such a situation,
he means the closest referent, which, under the circumstances, is identified
as yours/my/his/her etc. Otherwise, the utterance would be irrelevant.
U2: the lady replies considering that U1 is the result of observing the
maxims; the woman is polite and assuring; she doesn’t contradict the first
speaker by saying: that is not my dog.
In such a dialogue, the man favoured relevance, the woman
favoured the truth (referring to her own dog, as a result of the wrong
referent recoverability). The linguistic effect is humoristic, but in real life it
can cause ambiguity and misunderstanding.

‘Can you lend me the car?’ ‘What car?’ cf Rom. ‘Îmi împrumuţi maşina?’
‘Care maşină?’

In the example above, we refer to contexts in which it is common

knowledge that the referent has a car, and just one, therefore, the reply
should not be interpreted as a genuine question implying selection (which
of them?). Apparently, this is a case of non-recoverability of the referent.
The presupposition is there is a car; there is a car that you can lend; for you
to lend me something, you must own it.
Therefore, in this case, the car means your car. The interlocutor’s
linguistic knowledge should ensure him the correct recoverability of the
referent designated by the definite description the car. The reply seems to
buy the interlocutor some time, but, in fact, more important, it is a more
indirect, and, thus, a more polite way of denying the request.

195
6.5.2. Structural presuppositions and hearer’s manipulation

The speaker can manipulate the hearer, if the role of the hearer
makes the speaker resort to this strategy: a witness can be made to believe
what the speaker presupposes to be true because this serves the latter’s
purposes. The imperative mood is a direct linguistic manner of imposing the
presupposition on a felony witness:

Police investigator: ‘Where were you standing when you saw the suspect
attack them and steal the car?’
Witness: ‘I didn’t quite see that.’
Police investigator: ‘Just answer the question. Where were you standing?’
Witness: ‘At the bus stop.’

The relationship existing between interlocutors must be a vertical-

type one, with the speaker having authority and exerting his authority, on the
hearer:

Parent: ‘What did I tell you about being late for school?’
Child: ‘But I wasn’t, honestly.’
Parent: ‘I didn’t ask you that. I asked you to tell me if you remember what I
said about that.’

The parent is in a position of authority and he takes it to be

understood that his child was late. This simple strategy functions as a
reproach and the focus is on the possible consequences.

6.6. Politeness and (indirect) speech acts

Politeness is an important missing link between Cooperative

Principle and the problem of relating sense to illocutionary force (Leech
1983: 104). Illocutionary facts relate to politeness in their contribution to
establishing and maintaining cooperation throughout the verbal exchange.
The illocutionary force, i.e. the communicative goal, and the social goal can
be in one of the following types of relationship:
-convivial: illocutionary and social goal coincide in case of offers,
invitations, congratulations, greetings, thanks;
-collaborative: illocutionary and social goal are indifferent to each other in
case of assertions, information notes, reports, instructions,
announcements;
-competitive: the two types of goals compete, as in orders, begging etc;
-conflicting: the two types of goals are opposed, as in case of threats,
accusations, curses.

196
 Mey (1993: 68) argues against Leech’s remark ‘some illocutions
(orders) are inherently impolite, and others (offers) are inherently polite’
(1983: 83).
Politeness is asymmetrical: what is polite with respect to the hearer
or to some third party will be impolite with respect to the speaker and vice
versa. This piece of criticism turns out to be against the perception of
politeness as an abstract quality. Mey provides two counterarguments:
1. the social position of the speakers relative to one another may
indicate different politeness values for individual cases: e.g. the existence
of a social hierarchy may preempt (i.e. act contrary to) the use of
politeness. An order in the military can’t be said to be polite, but it is
vindicated (proven true) if it conforms to the demands of the military
hierarchy; commands are neither polite, nor impolite;
2. the politeness of an order may depend on the positive/negative
effects on the person who is given the order. Kunst-Gnamuš has shown
that the cost-benefit scale is decisive in assigning politeness value to ‘bald’
imperatives (Kunst-Gnamuš 1991: 59 in Mey 1993: 68).
Offers (we include here invitations, which, in our opinion, are
basically offers to receive somebody as a guest, to give a person
something etc) can be expressed by an imperative, in two cases: when the
relationship between interlocutors is a vertical one, i.e. the speaker has
some authority on the interlocutor, be it only moral, or the interlocutors have
a horizontal type of relationship implying equality, but the speaker takes
advantage of the interlocutors attitude/ feelings towards him. Either way,
imperatives can be a form of manipulation:

Have another cake!

In all the examples above, the tact and generosity maxim are
involved, and the speaker makes a FTA to himself and a FSA in relation to
the interlocutor, maybe anticipating his desires.
Unlike the previous examples, the utterances below illustrate cases
when the action referred to is not necessarily to the benefit of the
interlocutor. This speech act threatens the positive face of the speaker, i.e.
his need to be accepted socially, and the negative face of the interlocutor,
i.e. his wishes:

Peel the potatoes! (this example is in favor of the speaker, he expresses

his negative face, avoiding to perform the action himself)
Come and visit us!
You must see my new apartment! (in this case, the imperative is enhanced
by the use of a strong deontic verb, must, which intensifies the illocutionary
value, apparently increasing the costs to the speaker’s face; but in the long

197
run, the effects of the actions of inviting the interlocutor to visit him will
improve his positive face). (for more on this see Leech 1983: 104-128)

Levinson (1983: 274) refers to the concept of interactional

pessimism, grammaticalised in an example such as

I don’t suppose that you would by any chance be able to lend me some
cash, would you?

The speaker uses the negative form of the verb in the main clause, this
form being contradicted by the implicature I suppose it, obvious as a result
of the message of the utterance, which is a request. The same mitigating
value is held by the epistemic quantifiers, would, by any chance, which
induce the idea of strong uncertainty. Would be able to lend is another
structure which focuses apparently on the objective conditions necessary
for the hearer to perform the action: ability and possibility (favorable
circumstances).The example is relevant for the strategies used by the
speaker in order to get the intended effect.
The elements discussed in this chapter can be illustrated in point of
their use in a dialogue:

A: Mr. Brown, have you read our project? What do you think of it?
B: Ann, I’m not interested in something like that for the moment.
A: Mr. Brown, Mr. President, I’m aware of your priorities, I can understand
your attitude, but, still, I would reconsider my position, if I were you. I’d
really appreciate it if you could take the trouble of reading it, sir.
B: Ann, don’t insist. I made a decision and that’s that.
A: Please, sir, it would mean a lot to all those who contributed to this
project. We worked a lot on it.
B: Ok, Annie, I’ll look it over.
A: Thank you, Mr. President. You won’t regret it. May I suggest you to start
reading it as soon as possible, so that we could discuss it at our 3 o’clock
meeting?

The dialogue exhibits the features of Anglo-Saxon society:

individualist, assertive, informal (partly), favouring small power distance
(from the perspective of the person with authority) and showing low
uncertainty avoidance, and, last but not least, masculine.
Speaker A, the subordinate woman, uses a formal address term, to
minimize the costs of the hearer, counterbalancing the directness of the
questions. She risks to lose her face, but observes the tact maxim by the
address term. Speaker B, the man in a position of power, favours small
power distance because it is to his advantage. Using the first name as an
address term is a minimizer of the effect of non-observing the Agreement

198
Maxim. He bluntly refuses to discuss the matter, and the address term can
be interpreted not only as a minimizer, but as a maximizer, too, in the
sense that the man emphasizes his position of authority in relation to her.
He might want to make her feel inferior, not liked. The Anglo-Saxon culture
does not encourage unrestrained display of emotions, therefore, it would be
more likely for the address term to function as a FTA (Face Threatening
Act) for the man and a FSA (Face Saving Act) for the woman: ‘I appreciate
you, we know each other for some time, but its’s my call and I say no’. Still,
there are some politeness devices: another minimizer is the time adverbial
for the moment, which somehow acts as a tact maxim marker, creating the
illusion of a further discussion (and agreement ) on the topic.
Speaker A, uses a double address term, getting from individual
reference to social reference – an absolute deictic Mr. President-; they are
combined with sentences expressing the praise (she observes the
Approbation Maxim) and agreement, at the same time. She tries not to lose
her face by being tactful: she uses indirect linguistic expressions –
conditional forms, If clauses. In the end she reiterates her respect by using
a polite address term, sir. The reply Ann, don’t insist. I made a decision and
that’s that reiterates the same message as his first reply: the address term
is the minimizer, the tact marker is meant to reduce the effect of the
imperative, reinforced by an assertive sentence containing an impolite
structure that’s that, which should not allow for a continuation of the
discussion
Out of despair, speaker A adopts an attitude of submitting to the
authority of the interlocutor in point of accepting his point of view; she
observes the Generosity Maxim by using the softener please and the
formal address term. She continues by applying the Modesty Maxim using
the pronoun we with reference to the collective effort; thus, she turns what
seems to be a confrontation between two people into a possible act of
acknowledgment and reward towards a group of people. She takes
distance from the group (they) – it is a manifestation of the Generosity
Maxim, only to emphasize the effort put into it by all of them, including her.
This complex strategy pays off, the interlocutor gives in, the term of
endearment used (Annie) is a marker of the Approbation Maxim: I
acknowledge your efforts in order to convince me. The reply could also be
an illustration of the sympathy gained by speaker A throughout the
discussion.
But the discussion is not over; speaker A risks once more losing her
face, in order to speed things up. The Approbation Maxim is observed by
the thanking, and the use of the address form, sir, contributes to that. The
next sentence, You won’t regret it is an expression illustrating the Tact
Maxim, and also, indirectly, praises her own efforts, and those of the group.
The Maxim of Modesty is, thus, flouted, but for a good reason, since it
tactfully refers to the benefits of the other in the long run. This first part of

199
the utterance is counterbalanced by the second part, which, similarly to the
reply that begins the discussion, is an indirect request. The manipulation of
the interlocutor is done by using modal verbs (may), attitudinal verbs
(suggest), a false question, a solidarity we.
The text above was intended as an illustration of the complex
strategies used by interlocutors to obtain a certain perlocutionary effect
without losing their face.

Conclusions

The concept of politeness is, above all, a social phenomenon, but,

not only: it is based on the interlocutors’ identity and types of relationships.
During the verbal exchange, a series of strategies can be used and they
were systematized in the form of the Politeness Principle and its maxims; at
the same time, the main concern of the interlocutors is their public image,
their face, which they manage to save or, on the contrary, they lose during
the verbal exchange. Linguistically, politeness interferes in all the domains
of pragmatics, it is the super-ordinate term, in a way, since all linguistic
devices pertaining to pragmatics are exploited for the speaker to remain
polite, against the background of his cultural space specificity: deictics,
inferences, illocutionary values.

200
 BIBLIOGRAPHY

PART I. SEMANTICS

Aitchison, Jean.1994. Words in the Mind. An Introduction to the Mental

Lexicon. 2nd edition. Oxford: Basil Blackwell.
Allan, Keith. 1986. Linguistic Meaning. London and New York: Routledge
and Kegan Paul.
Baldinger, Kurt. 1990. Semantics. Towards a Modern Semantics.
Translated by W.C. Brown and R. Wright. Oxford: Basil Blockwell.
Bidu-Vrânceanu, Angela et al. 1997. Dicţionar general de științe. Știinţe ale
limbii. București: Editura știintifică.
Bréal, Michel. 1900 [1897]. Semantics: Studies in the Science of Meaning.
London: W. Heinemann.
Brown, Keith and John Miller. eds. 1999. Concise Encyclopedia of
Grammatical Categories. Amsterdam: Elsevier.
Brugman, Claudia. 1988. The Story of Over: Polysemy, Semantics and the
Structure of the Lexicon. New York: Garland.
Carnap, Rudolf. 1956. Meaning and Necessity. Chicago: University of
Chicago Press.
Chandler, Daniel. 2007. Semiotics. The Basics. London: Routledge.
Chițoran, Dumitru. 1973. Elements of English Structural Semantics.
București: Editura didactică și pedagogică.
Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge,
Mass: MIT Press.
Coleman, Linda and Paul Kay. 1981. Prototype semantics: The English
verb lie.
Language 57.1: 26-44.
Cornilescu, Alexandra and Dumitru Chițoran. 1986. Elements of English
Sentence Semantics. Bucuresti : Tipografia Universitatii Bucuresti.
Cornilescu, Alexandra. 1995. Concepts of Modern Grammar. București:
Editura Universitățtii din București.
Coşeriu, Eugenio and Horst Geckeler. 1981. Trends in Structural
Linguistics. Tubingen: Gunter Narr Verlag.
Croft, William and David Allan Cruse. 2004. Cognitive Linguistics.
Cambridge: Cambridge University Press.
Cruse, David A. 1986. Lexical Semantics. Cambridge: Cambridge
University Press.
Cruse, David A. 2004. Meaning in Language. An Introduction to Semantics
and Pragmatics. Oxford: Oxford University Press.
Cruse, David A. 2006. Glossary of Semantics and Pragmatics. Edinburgh:
Edinburgh University Press.

201
Crystal, David. 1992. A Dictionary of Linguistics and Phonetics. Oxford:
Blackwell.
Dirven, René and Verspoor, Marjolijin. 1998. Cognitive Exploration of
Language and Linguistics. Amsterdam, Philadelphia: John
Benjamins.
Evans, Vyvyan and Melanie Green. 2006. Cognitive Linguistics. An
Introduction. Mahvah, NJ: Lawrence Erbaulm Associates.
Fillmore, Charles J. 1982. Frame Semantics in Linguistics in the Morning
Calm, (Linguistic Society of Korea.ed.). Seoul: Hanshin, pp. 111-138.
Frege, Gottlob. 1970. On Sense and Reference in Translations from the
Philosophical Writings of Gottlob Frege (P.Geach and M.Black,
eds.). Oxford: Basil Blackwell.
Geeraerts, Dirk. 2009. Theories of Lexical Semantics. Oxford: OUP.
Geeraerts Dirk and Hubert Cuyckens (eds.). 2007. Oxford Handbook of
Cognitive Linguistics. Oxford: Oxford University Press.
Geeraerts, Dirk and Stefan Grondelaers, Peter Bakemo. 1994. The
Structure of Lexical Variation. Meaning, Naming and Context. Berlin
and New York: Mouton de Gruyter.
Greimas, Algirdas J. 1966. Semantique structurale. Recherche de
methode. Paris: Larousse.
Jackendoff, Ray. 1986 [1983]. Semantics and Cognition. Cambridge, MA:
MIT Press.
Katz, Jerrold J. 1972. Semantic Theory. New York: Harper and Row.
Katz, Jerrold J. and Jerry A. Fodor. 1963. 'The structure of a semantic
theory'. Language 39: 170-210.
Kortmann, Bernd. 2005. English Linguistics: Essentials. Berlin: Cornelsen.
Lakoff, George. 1970. 'A note on vagueness and ambiguity' in Linguistic
Inquiry 1, 357-359.
Lakoff, George. 1971. 'On generative semantics' in Semantics. An
Interdisciplinary Reader in Philosophy, Linguistics and Psychology
(Danny D. Steinberg and Leon A. Jakobovits, eds.). Cambridge:
Cambridge University Press, 232-296.
Lakoff, George and Mark Johnson. 1980. Metaphors We Live by. Chicago:
University of Chicago Press.
Lakoff, George. 1987. Women, Fire and Dangerous Things. What Categories
Reveal about the Mind. Chicago and London: Chicago University
Press.
Lakoff, George. 1988. ‘Cognitive semantics’ in Meaning and Mental
Representations (U.Eco and M. Santambrogio, P. Violi, eds.).
Bloomington: Indiana University Press, 119-54.
Langacker, Ronald W. 1987. Foundations of Cognitive Grammar 1.
Theoretical Prerequisites. Stanford: Stanford University Press.
Leech, Geoffrey. 1971. Meaning and the English Verb. London: Longman.
Leech, Geoffrey. 1987/1974. Semantics: the Study of Meaning.
Harmondsworth: Penguin.

202
Lehrer, Adrienne. 1974.a. Semantic Fields and Lexical Structure.
Amsterdam and London: North Holland.
Lehrer, Adrienne. 1974.b. ‘Homonymy and Polysemy. Meaning. Similarity
of Meaning’ in Language Sciences, 25, 33- 38.
Lehrer, Adrienne. 1990. ‘Polysemy, Conventionality, and the Structure of
the Lexicon’ in Cognitive Linguistics, 1.2, 207-246.
Lehrer Adrienne and Eva Kittay E.1992. Frames, Fields and Contrasts.
New Essays in Semantic and Lexical Organization. New Jersey:
LEA.
Levin, Beth and Steven Pinker, (eds.). 1992. Lexical and Conceptual
Semantics. Oxford: Blackwell.
Levinson, Stephen. 1983. Pragmatics. Cambridge: Cambridge University
Press.
Lipka, Leonhard. 1990. An Outline of English Lexicology. Lexical Structure
Word Semantics and Word Formation. Tubingen: Max Niemeyer
Verlag.
Löbner, Sebastian. 2002. Understanding Semantics. London: Arnold.
Lyons, John.1963. Structural Semantics. Oxford: Blackwell.
Lyons, John. 1985 [1968]. Introduction to Theoretical Linguistics.
Cambridge: Cambridge University Press.
Lyons, John. 1986 [1981]. Language, Meaning and Context. London:
Fontana.
Lyons, John. 1987 [1977]. Semantics. (2 vols.) Cambridge: CUP.
Lyons, John. 1995. Linguistic Semantics. An Introduction. Cambridge:
CUP.
McCawley, James D. 1968. The Role of Semantics in a Grammar in
Universals in Linguistic Theory (Emmon Bach and Robert T. Harms,
eds.). New York: Holt Rinehart and Winston, pp.124-169.
McCawley, James D. 1973. Grammar and Meaning. Papers on Syntactic
and Semantic Topics. Tokyo: Taishukan.
Meyer, Paul G. 2002. Synchronic English Linguistics. Tubinger: Narr.
Miller, George. ed. 1978 a. Linguistic Theory and Psychological Reality.
The MIT Press.
Miller, George. 1978 b. Semantic Relations among Words. InMiller,
George, ed. 60-118.
Murphy, Lynne M. 2003. Semantic Relations and the Lexicon. Antonymy,
Synonymy and Other Paradigms. Cambridge : CUP.
Nida, Eugene. 1975. Componential Analysis of Meaning. Mouton: The
Hague.
Neagu, Mariana.1996.Elements of Lexical Semantics and Pragmatics. Braila:
Evrika.
Neagu, Mariana.1999. Categories in Natural Language. The Study of
Nominal Polysemy in English and Romanian. Buzău: Alpha.
Neagu, Mariana. Cognitive Linguistics. An Introduction. 2005. București:
Editura Didactică și Pedagogică.

203
Palmer, Frank R. 1986 [1976]. Semantics. A New Outline. Cambridge:
Cambridge University Press.
Panman, Otto. 1982. Homonymy and Polysemy. Lingua. 58.
Persson, Gunnar. 1990. Meanings, Models and Metaphor. A Study in
Lexical Semantics in English in Acta Universitatis Umensis. Umea
Studies in the Humanities. 92. Stockholm: Almqvist & Wiksell
International.
Pottier, Bernard. 1964. Vers une sémantique moderne. Travaux de
linguistique et de littérature. 2, 107-37.
Pustejovsky, James and Branimir Boguraev. 1996. Lexical Semantics. The
Problem of Polysemy. Cambridge: Cambridge University Press.
Putnam, Hilary. 1975. The Meaning of Meaning in Minnesota Studies in the
Philosophy of Language. vol. VII. Minneapolis: University of
Minnesota Press.
Saeed, John. 2000 [1997]. Semantics. Oxford: Blackwell Publishers.
Saussure, Ferdinand de. 1965 [1916]. Cours de Linguistique Générale:
Payot.
Saussure, Ferdinand de.1983 [1916]. Course in General Linguistics. (transl.
by Roy Harris). Chicago: Open Court.
Stern, Gustav. 1968 [1931]. Meaning and Change of Meaning.
Bloomington: University of Indiana Press
Sweetser, Eve. 1990. From Etymology to Pragmatics. Metaphorical and
Cultural Aspects of Semantic Structure. Cambridge: Cambridge
University Press.
Şăineanu, Lazăr. 1887. Încercare asupra semasiologiei limbii române.
Studii istorice despre tranziţiunea sensurilor. Bucureşti: Tipografia
Academiei Române.
Taylor, John. 1995 [1989]. Linguistic Categorisation. Prototypes in
Linguistic Theory. Cambridge: Cambridge University Press.
Tsohatzidis, Savas L. (ed.). 1990. Meanings and Prototypes. Studies in
Linguistic Categorization. London and New York: Routledge.
Tuggy, David. 1993. Ambiguity, polysemy, and vagueness. Cognitive
Linguistics 4: 273-90.
Ullmann, Stephen. 1962. Semantics: An Introduction to the Science of
Meaning. Oxford: Basil Blackwell.
Ullmann, Stephen. 1964 [1957]. The Principles of Semantics. New York:
Philosophical Library.
Ungerer, Friedrich and Hans Jorg Schmid. 1996. An Introduction to
Cognitive Linguistics. London: Longman.
Vasiluţă, Laura. 2001. Semantica lexicală. Timişoara: Mirton.
Warren, Beatrice. 1992. Sense Developments. A Contrastive Study of the
Development of Slang Senses and Novel Standard Senses in
English. Stockholm: Almquist & Wiksell International.

204
 PART II. PRAGMATICS

Ariel, Mira. 2008. Pragmatics and Grammar. Cambridge: CUP.

Armengaud, Françoise. 1985. La pragmatique. Paris: Presses
Universitaires de France.
Austin, John L. 1972. How to Do Things With Words (J.O. Urmson,
M. Sbisà, eds.). Cambridge: Harvard University Press.
Austin, John L. 1972. How to Do Things With Words (J.O. Urmson,
M. Sbisà, eds.). Cambridge: Harvard University Press.
Biber, Douglas and Stig Johansson, Geoffrey Leech, Susan Conrad, Edward
Finegan. 1999. Grammar of Spoken and Written English. London:
Longman.
Bidu-Vrânceanu, Angela et alii. 2001. Dicţionar de ştiinţe ale limbii.
Bucureşti : Nemira.
Blommaert, Jan. 2005. Discourse. A Critical Introduction. Cambridge: CUP.
Bloomfield, Leonard. 1973 [1933]. Language. London: George Allen &
Unwin Ltd.
Brown, Gillian and George Yule. 1983. Discourse Analysis. Cambridge:
CUP.
Brown, Penelope and Stephen Levinson. 1978. ‘Universals in language
usage: politeness phenomena’ in Questions and politeness:
strategies in social interaction, (Goody, Esther, ed.). Cambridge:
Cambridge University Press.
Burton-Roberts, Noel. 1989. The Limits to Debate. A Revised Theory of
Semantic presupposition. Cambridge: CUP.
Bűhler, Karl. 1990 [1934]. Theory of Language. The Representatinal
Theory of Language. Amsterdam: John Benjamins.
Bybee, Joan. 1985. Morphology. A Study of the Relation Between Meaning
and Form. Amsterdam: John Benjamins.
Chaika, Elaine. 2000. Linguistics, Pragmatics and Psychotherapy. A Guide
for Therapists. London: Whurr Publishers.
Chimombo, Moira and Robert L. Roseberry. 1998. The Power of Discourse:
An Introduction. Mahwah, New Jersey: Lawrence Erlbaum
Associates, 108.
Comrie, Bernard. 1985. Tense. Cambridge: Cambridge University Press.
Costăchescu, Adriana. 2013. La Pragmatique Linguistique: théories,
débats, exemples. Lincom Studies in pragmatics. Munich: Lincom
Europa.
Costăchescu, Adriana. 2014. ‘On disagreement markers in French and
Romanian dialogues’ in Pragmatic markers form Latin to Romance
languages (Chiara Ghezzi and Piera Molinelli, eds.). Oxford: Oxford
University Press, 41-59.

205
Crystal, David. 21985. A Dictionary of Linguistics and Phonetics. NY: Basil
Blackwell.
Dragoş, Elena. 2000. Introducere în pragmatică. Cluj-Napoca: Casa Cărţii
de Ştiinţă.
Finch, Geoffrey. 2003. How to study linguistics. London: Palgrave.
von Fintel, Kai. 2000. What is Presupposition Accommodation?
Massachusetts: MIT, http://web.mit.edu/fintel/fintel-2000-
accomm.pdf/August 10, 2015
Fraser, Bruce. 1990. ‘Perspectives on politeness’ in Journal of Pragmatics
14, 219-236.
Fromkin, Victoria and Robert Rodman. 1998. An Introduction to Language.
Thomson & Heinle.
Goffman, Erving. 1967. ‘On facework: an analysis of ritual elements in
social interaction’, in The Discourse Reader (Adam Jaworski and
Nicholas Coupland, eds.). London: Routledge, 306-321.
Goffman, Erving. 1967. Interaction ritual: Essays on face-to - face
behaviour. NY: Doubleday Anchor.
Green, Georgia. M. 2004. Pragmatics and syntax. Handbook of pragmatics.
Oxford: Blackwell Publishing.
Grice Paul H. 1975. “Logic and conversation” in Speech Acts (Cole, P., and
J.L. Morgan, eds.). New York: Academic Press, 41–58.
Grice, Paul H. 1989. Studies in the Way of Words. Harvard: Harvard
University Press.
Grice, Paul H. 1999. ‘Logic and Conversation’ in The Discourse Reader
(Adam Jaworsky and Nicholas Coupland, eds.). London: Routledge.
Haiman, John. 1998. Talk is Cheap: Sarcasm, Alienation, and the Evolution
of Language. New York: OUP.
Hall, Edward T.1959. The silent language. New York: Doubleday.
Hall, Edward T.1966. The hidden dimension. New York: Doubleday.
Hanks, W. F. 1996. Language and Communicative Practices. Boulder,
Colorado: Westview Press.
Hofstede, Geert. 1991 [1980]. Cultures and Organizations: Software of the
Mind (2-nd edition). London: McGraw-Hill.
Holtgraves, Thomas M. 2002. Language a social action: Social Psychology
and Language Use. Lawrence-Erlbaum & Assoc.
Hyman, Larry M. 1983. ‘Form and substance in language universals’ in
Linguistics 21, 67–85.
Ionescu-Ruxăndoiu, Liliana. 2001. Deixis în Dicţionar de ştiinţe ale limbii.
Bucureşti: Nemira, 157.
Karttunen Laurie and Stanley Peters. 1975. ‘Requiem for Presupposition’ in
Proceedings of the Third American Meeting of the Berkley Linguistic
Society, Berkeley, University of California Press, 360-371.

206
Karttunen Lauri and Stanley Peters. 1979. ‘Conventional Implicature’, in
Syntax and Semantics 11: Presupposition, (Choon-Kyu Oh and
David Deenen, eds.). New York: Academic Press, 1 - 56.
Karttunen, Lauri. 1973. ‘Presuppositions of Compound Sentences’ in
Linguistic Inquiry, 4, 169 - 193.
Kerbrat-Orecchioni, Catherine. 1992 [1990]. Les interactions verbales. Paris:
Armand Colin.
Kerbrat Orecchioni, Catherine. 1999 [1980]. L'enonciation. Paris: Armand
Colin.
Kroeger, Paul R. 2005. Analyzing Grammar. An Introduction. Cambridge:
CUP.
Kuno, Susumu. 1987. Functional Syntax, Anaphora, Discourse, and
Empathy. Chicago: University of Chicago press, 203-267.
Langendoen, D. Terry and Harris Savin.1971. ‘The Projection Problem for
Pressupositions’ in Studies in Linguistic Semantics (Charles Fillmore
and D. T. Langendoen, eds.). New York: Holt, Rinehart and Winston,
55 - 62.
Levinson, Stephen. 1983. Pragmatics. Cambridge: CUP.
Leech, Geoffrey. 1983. Principles of Pragmatics. London: Longman.
Lewis, David K. 1973. Counterfactuals. Cambridge: Harvard University
Press.
Lyons, John. 1977. Semantics. 2 vol. Cambridge: Cambridge University
Press.
Mey, Jakob. 1993. Pragmatics. An Introduction. NY: Blackwell.
Morris, Charles W. 1938. Foundations of the theory of signs. Encyclopedia
of Unified Science, vol. I, no. 2, Chicago: The University of Chicago
Press.http://www.scribd.com/doc/51866596/Morris-1938-
Foundations-of- Theory-of-Signs#scribd
Parret, Herman. 1976. ‘La pragmatique des modalités’ in Langages, 43,
pp. 47-63.
Pisoschi, Claudia G. 2010. Deixisul personal: valenţe ale pronumelor
personale şi de politeţe. Aspecte contrastive română-engleză.
Craiova: Editura Universitaria.
Pisoschi, Claudia G. 2012. The Basics of Nominal Reference.
Craiova:Universitaria.
Recanati, François. 2005. ‘Deixis and Anaphora’ in Semantics and
Pragmatics (Z. Szabo ed.). Oxford: Clarendon Press.
Quirk, Randolph & S. Greenbaum, G. Leech, J. Svartvik. 1985.
A Comprehensive Grammar of the English Language. London & NY:
Longman.
Searle, John R. & D. Vanderveken. 1985. Foundations of Illocutionary
Logic. Cambridge: CUP.
Searle, John R. 1979. Expression and Meaning: Studies in the Theory of
Speech Acts. Cambridge: CUP.

207
Smith, Neil & Deirdre Wilson. 1990. Modern Linguistics. The Results of
Chomsky’s Revoluton. London: Penguin Books.
Sperber, Dan and Deirdre Wilson. 1986. Relevence Communication and
Cognition, Harvard MA, Harvard University Press.
Stalnaker, Robert C. 1974. ‘Pragmatic presuppositions’ in Semantics and
Philosophy, (Milton K. Munitz & Peter K. Unger, eds.). New York:
New York University, 197-213.
Stalnaker, Robert C. 2002. ‘Common Ground’ in Linguistics and
Philosophy, 25, 701 – 721.
Strawson Peter. 1950. ‘On Referring’, in Mind, 59, 320 - 344.
Strawson, Peter. 1952. Introduction to Logical Theory. London: Methuen.
Verschueren, Jef. 2001. ‘Pragmatics’ in Routledge Companion to Semiotics
and Linguistics (Paul Cobley editor). London &NY: Routledge.
Vincent Marrelli, Jocelyne. 2002. ‘Truthfulness’ in Handbook of Pragmatics,
(Jef Verschueren, Jan-Ola Őstman, Jan Blommaert & Chris
Bulcaen, eds.). Amsterdam: John Benjamins.
Wales, Katie. 1996. Personal Pronouns in Presentday English. Cambridge:
CUP.
Velleman, J. David. 1992. ‘The Guise of the Good’ in Noûs, Vol. 26, No. 1,
(March 1992), 3-26.
Wierzbicka, Anna. 1985. ‘Different cultures, different languages, different
speech acts: Polish v. English’ in Journal of Pragmatics, 06/1985;
9(2):145-178.
http://www.researchgate.net/publication/240418078_Different_cultur
es_different_languages_different_speech_acts_Polish_v._English
[accessed Aug 12, 2015].
Yule, George. 1996. Pragmatics. Oxford: OUP.

208