Essays On Freges Basic Laws of Arithmetic First Edition Edition Ebert Full Chapter

Essays on Frege’s : basic laws of
arithmetic First Edition. Edition Ebert

Visit to download the full and correct content document:
https://ebookmass.com/product/essays-on-freges-basic-laws-of-arithmetic-first-edition
-edition-ebert/
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 1 — #1

i i
Essays on Frege’s Basic Laws of Arithmetic
i i
i i
i i

i i
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page iii — #3

i i
Essays on Frege’s Basic

Laws of Arithmetic
Edited by
P H I L I P A . E B E RT A N D M A R C U S R O S SB E R G
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page xii — #12

i i
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© the several contributors 2019
The moral rights of the authors have been asserted
First Edition published in 2019
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2019945694
ISBN 978–0–19–871208–4
Printed and bound in Great Britain by
Clays Ltd, Elcograf S.p.A.
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page v — #5

i i
Contents
Foreword vii
Contributors x
1 The Basic Laws of Cardinal Number 1

Richard Kimberly Heck
2 Axioms in Frege 31
Patricia A. Blanchette
3 When Logic Gives Out. Frege on Basic Logical Laws 57
Walter B. Pedriali
4 The Context Principle in Frege’s Grundgesetze 90
Øystein Linnebo
5 Why Does Frege Care Whether Julius Caesar is a Number?
Section 10 of Basic Laws and the Context Principle 115
Joan Weiner
6 Grundgesetze and the Sense/Reference Distinction 142
Kevin C. Klement
7 Double Value-Ranges 167
Peter Simons
8 The Proof of Hume’s Principle 182
Robert C. May and Kai F. Wehmeier
9 Frege’s Theorems on Simple Series 207
William Stirton
10 Infinitesimals, Magnitudes, and Definition in Frege 235
Jamie Tappenden
11 Frege’s Relation to Dedekind: Basic Laws and Beyond 264
Erich H. Reck
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page vi — #6

i i
12 Frege on Creation 285

Michael Hallett
13 Mathematical Creation in Frege’s Grundgesetze 325
Philip A. Ebert and Marcus Rossberg
14 Frege on the Real Numbers 343
Eric Snyder and Stewart Shapiro
15 Frege’s Little Theorem and Frege’s Way Out 384
Roy T. Cook
16 “How did the serpent of inconsistency enter Frege’s paradise?” 411
Crispin Wright
17 Second-Order Abstraction Before and After Russell’s Paradox 437
Matthias Schirn
18 Formal Arithmetic Before Grundgesetze 497
19 Definitions in Begriffsschrift and Grundgesetze 538
Michael Kremer
20 A Brief History of English Translations of Frege’s Writings 567
Michael Beaney
21 Translating ‘Bedeutung’ in Frege’s Writings: A Case Study and
Cautionary Tale in the History and Philosophy of Translation 588
Michael Beaney
22 Contemporary Reviews of Frege’s Grundgesetze 637
Philip A. Ebert and Marcus Rossberg
Index 653
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page vii — #7

i i
Foreword
Gottlob Frege published his Grundgesetze der Arithmetik in two volumes; the
first appeared in 1893, the second in 1903. Grundgesetze was to fulfill Frege’s
ambition to demonstrate that arithmetic and analysis are reducible to logic,
and thus establish a view that we now call ‘logicism’. In previous work, in
particular in Begriffsschrift (1879) and Die Grundlagen der Arithmetik (1884),
Frege provides some of the formal and philosophical foundations for his logi-
cism. More specifically, his first book contains the initial formulation of the lo-
gical system—the eponymous Begriffsschrift—while his second book eschews
formulae altogether and offers a philosophical foundation for his logicist po-
sition in the philosophy of mathematics. Both publications are, in their own
right, groundbreaking.
Begriffsschrift constitutes “perhaps the greatest single contribution to lo-
gic ever made and it was, in any event, the most important advance since
Aristotle.”1 On the other hand, the philosophical methodology of Grundla-
gen, in particular its analytic writing style, led many to think of Frege as one
of the founding fathers of analytic philosophy. Michael Dummett considers
Grundlagen “the most brilliant piece of philosophical writing of its length
ever penned”.2 In this work scholars locate the origins of the linguistic turn
in philosophy, which shaped much of twentieth-century philosophy.
However, Frege’s standing in the history of analytic philosophy does not
simply derive from these two groundbreaking books, but it is also based on
a series of lectures and articles that Frege wrote just before the publication
of the first volume of Grundgesetze. Function und Begriff (1891), ‘Über Sinn
und Bedeutung’ (1892), and ‘Ueber Begriff und Gegenstand’ (1892) make
for a trilogy of Godfather-like proportions with part II being, of course, one
of the most widely read articles in philosophy. Its influence in the philosophy
of language and in linguistics would be hard to exaggerate.
How then does Grundgesetze, a work that Frege without doubt intended
to be his magnum opus, fit into the list of Frege’s philosophical and logical
achievement? Most scholars have, until recently, viewed Frege’s Grundgesetze
1 Alexander George and Richard Kimberly Heck (2000), ‘Frege, Gottlob’, in Edward Craig
(ed.), Concise Routledge Encyclopedia of Philosophy, London and New York: Routledge, page 296.
(Orig. publ. under the name “Richard G. Heck, Jr”.)
2 Michael Dummett (2007), ‘Intellectual Autobiography’, in Randall E. Auxier and Lewis
Edwin Hahn, eds. The Philosophy of Michael Dummett, vol. XXXI of The Library of Living Philo-
sophers, Chicago and La Salle, Ill.: Open Court, page 9.
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page viii — #8

i i
viii Foreword
as containing for the most parts, merely the formal details of his logicist pro-
ject which we now know to have failed. Frege’s derivations of fundamental
arithmetical principles, which cover about half of either volume, are based
on an inconsistent axiom: the infamous Basic Law V. It is thus perhaps no
surprise that little attention was paid to Frege’s predominately formal work.
Interestingly, however, it is Grundgesetze that contains Frege’s most force-
ful and most famous rejection of psychologism in logic, and the relevant pas-
sages from the Foreword of the first volume were the first pieces of Frege’s cor-
pus to be translated into English. Moreover, Wittgenstein, Russell, Jourdain,
Peano, and other philosophers and mathematicians of the time who seriously
engaged with Frege’s work paid close attention to Grundgesetze. Nonetheless,
throughout the rest of the twentieth century more and more emphasis was
placed on Frege’s other writings—writings that are independent of the fail-
ure of his logical foundations, or indeed independent of his logicism more
generally.
In 1983, Crispin Wright published Frege’s Conception of Numbers as Ob-
jects, in which he reconstructed Frege’s derivation of the axioms of arithmetic
from what would come to be known as “Hume’s Principle”, or “HP”. Wright
also conjectured there that HP, unlike Basic Law V, is consistent.3 This in-
spired George Boolos to look more carefully at Frege’s informal derivations in
Grundlagen, and also at the formal work in Begriffsschrift.4 Boolos’s student
Richard Kimberly Heck then began investigating Frege’s technical work on
arithmetic in Grundgesetze.5 Around the same time, Peter Simons and Michael
Dummett also published their discussions of Frege’s formal work on the real
numbers in Part III of Grundgesetze.6 The increasing interest in Frege’s proofs
was accompanied by a renewed interest in Frege’s philosophy of mathematics,
given also the rise of a view labelled ‘neo-logicism’, ‘neo-Fregeanism’, or simply
‘abstractionism’, as championed by Crispin Wright and Bob Hale,7 and the
3 For the first published consistency proofs of HP see two reviews of Frege’s Conception of
Numbers as Objects: one by John P. Burgess (1984, Philosophical Review 93:638–40), and the other
by Allen Hazen (1985, Australasian Journal of Philosophy 63:251–4); and see Boolos (1987), as
cited in note 4 below.
4 George Boolos (1987), ‘The Consistency of Frege’s Foundations of Arithmetic’, in Judith
Jarvis Thomson (ed.), On Being and Saying: Essays for Richard Cartwright, Cambridge, Mass.:
MIT Press, pages 3–20; George Boolos (1985), ‘Reading the Begriffsschrift’, Mind 94:331–44.
5 Richard Kimberly Heck (1993), ‘The Development of Arithmetic in Frege’s Grundgesetze der
Arithmetik’, Journal of Symbolic Logic 58:579–601 (originally published under the name “Richard
G. Heck, Jr”). See also their more recent books Frege’s Theorem (2011) and Reading Frege’s Grund-
gesetze (2012), both Oxford: Clarendon Press.
6 Peter Simons (1987), ‘Frege’s Theory of the Real Numbers’, History and Philosophy of Logic
8:25–44; Michael Dummett (1991), Frege: Philosophy of Mathematics. London: Duckworth.
7 Compare here, e.g., Crispin Wright (1983), Frege’s Conception of Numbers as Objects, Aber-
deen: Aberdeen University Press; Bob Hale (1987), Abstract Objects, Oxford: Basil Blackwell;
Bob Hale and Crispin Wright (2001), The Reasons’ Proper Study: Essays towards a Neo-Fregean
Philosophy of Mathematics, Oxford: Clarendon Press; Philip A. Ebert and Marcus Rossberg, eds.
(2016) Abstractionism: Essays in Philosophy of Mathematics, Oxford: Oxford University Press.
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page ix — #9

i i
Foreword ix
work by Michael Dummett. Yet a broader engagement with Frege’s magnum

opus faced the obstacle that only parts of the two volumes had been translated.
Together with Crispin Wright, we started work on a new translation of Grund-
gesetze in 2003, and, with the help of many distinguished Frege scholars, the
first complete English translation, Basic Laws of Arithmetic, volumes I and II,
appeared in 2013.8
The present volume is the first collection of essays that focuses on Frege’s
Grundgesetze and aims to highlight the technical as well as philosophical rich-
ness of Frege’s major work. The companion brings together twenty-two Frege
scholars, whose contributions discuss a wide range of topics arising from both
volumes of Grundgesetze. And so may this collection, even if belatedly, con-
tribute to a renaissance of Grundgesetze and help to establish this work as one
of Frege’s many masterpieces.
We would like to thank our editor, Peter Momtchiloff at Oxford Uni-
versity Press, for his support and his patience. Thanks to Denise Bannerman
for meticulous proof-reading. Typesetting a nearly 700-page volume in LATEX
proved to be yet another big challenge, and we would like to thank Colin
McCullough-Benner and Andrew Parisi who provided invaluable assistance
with this arduous task. We would also like to thank Colin for his work on the
Index and his additional help with copyediting and proof-reading.
And thanks to Crispin, for getting it all started, and for everything.
Philip Ebert, Stirling, Scotland

Marcus Rossberg, Storrs, Conn., USA
8 For more details on earlier translations as well as the first complete translation of Grundgesetze,
and how the latter came about, consult Crispin Wright’s Foreword and our Translators’ Introduc-
tion in Gottlob Frege (2013), Basic Laws of Arithmetic. Derived Using Concept-Script, volumes I
and II, trans. and ed. Philip A. Ebert and Marcus Rossberg with Crispin Wright. Oxford: Oxford
University Press.
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page x — #10

i i
Contributors
Michael Beaney is Professor of History of Analytic Philosophy at the Hum-

boldt-Universität zu Berlin and Professor of Philosophy at King’s Col-
lege London.
Patricia A. Blanchette is Glynn Family Honors Chair of Philosophy at the
University of Notre Dame.
Roy T. Cook is CLA Scholar of the College and John M. Dolan Professor of
Philosophy at the University of Minnesota, Twin Cities.
Philip A. Ebert is Senior Lecturer in Philosophy at the University of Stirling.
Michael Hallett is John Frothingham Professor of Logic and Metaphysics at
McGill University, Montreal.
Richard Kimberly Heck is Professor of Philosophy at Brown University, in
Providence, Rhode Island.
Kevin C. Klement is Professor of Philosophy at the University of Massachu-
setts Amherst.
Michael Kremer is the Mary R. Morton Professor of Philosophy and in the
College at the University of Chicago.
Øystein Linnebo is Professor of Philosophy at the University of Oslo.
Robert C. May is Distinguished Professor of Philosophy and Linguistics at
the University of California, Davis.
Walter B. Pedriali is Associate Lecturer in Philosophy at the University of St
Andrews.
Erich H. Reck is Professor of Philosophy at the University of California at
Riverside.
Marcus Rossberg is Associate Professor of Philosophy at the University of
Connecticut.
Matthias Schirn is Professor Emeritus of Philosophy at the University of Mu-
nich and a member of the Munich Center for Mathematical Philosophy.
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page xi — #11

i i
Contributors xi
Stewart Shapiro is O’Donnell Professor of Philosophy at The Ohio State Uni-

versity, Distinguished Visiting Professor at the University of Connecti-
cut, and Distinguished Presidential Fellow at the Hebrew University of
Jerusalem
Peter Simons, FBA, MRIA, is Professor Emeritus of Philosophy at Trinity
College Dublin.
Eric Snyder is Postdoctoral Fellow at the Munich Center for Mathematical
Philosophy, LMU Munich, and Associate Professor of Philosophy at
Ashoka University.
William Stirton works as an administrative assistant for Edinburgh Leisure
Ltd.
Jamie Tappenden is Professor of Philosophy at the University of Michigan,
Ann Arbor.
Kai F. Wehmeier is Dean’s Professor of Logic and Philosophy of Science and
of Language Science at the University of California, Irvine.
Joan Weiner is Professor of Philosophy at Indiana University.
Crispin Wright, FBA, FRSE, FAAAS, is Professor of Philosophy at New York
University, Professor of Philosophical Research at the University of Stirl-
ing, and Regius Professor of Logic Emeritus at Aberdeen University.
i i
i i
i i
“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page xii — #12

i i
i i
i i
i i

i i
1
The Basic Laws of Cardinal Number
Frege begins his Grundgesetze der Arithmetik as follows:

In my Grundlagen der Arithmetik, I aimed to make it plausible that arithmetic is a
branch of logic and needs to rely neither on experience nor intuition as a basis for its
proofs. In the present book this is now to be established by deduction of the simplest
laws of cardinal number by logical means alone. (Grundgesetze I, 1)
The plausibility of what is now called ‘logicism’ was supposed to have been
established not only by the philosophical arguments in Die Grundlagen but,
more importantly, by the proofs of basic arithmetical principles that Frege
sketches in §§70–83. But the character of those arguments left a large lacuna:
I do not claim to have made the analytic character of arithmetical propositions more
than plausible,1 because it can always still be doubted whether they are deducible solely
from purely logical laws, or whether some other type of premiss is not involved at some
point in their proof without our noticing it. This misgiving will not be completely
allayed even by the indications I have given of the proof of some of the propositions;
it can only be removed by producing a chain of deductions with no link missing, such
that no step in it is taken which does not conform to some one of a small number of
principles of inference recognized as purely logical. (Frege, 1884, §90)
But how can we be sure that no link is missing? That problem was the one
that had led to Frege’s interest in logic, as he makes explicit in Begriffsschrift:
[W]e divide all truths that require justification into two kinds, those for which the
proof can be carried out purely by means of logic and those for which it must be sup-
ported by facts of experience. … [W]hen I came to consider the question to which of
these two kinds the judgments of arithmetic belong, I first had to ascertain how far
one could proceed in arithmetic by means of inferences alone, with the sole support of
those laws of thought that transcend all particulars. … To prevent anything intuitive
from penetrating here unnoticed, I had to bend every effort to keep the chain of infer-
ences free of gaps. In attempting to comply with this requirement in the strictest way
possible I found the inadequacy of language to be an obstacle; no matter how unwieldy
1 Austin translates this as ‘probable’, but I have altered the translation, since Frege uses the
same word here as in Grundgesetze.
i i
i i
i i

i i
2 Richard Kimberly Heck
the expressions I was ready to accept, I was less and less able, as the relations became
more and more complex, to attain the precision that my purpose required.
(Frege, 1879, 5–6)
What Frege needed to do, then, to fill the lacuna, was to provide formal proofs
of the various propositions he had only proven informally in Die Grundlagen.
Frege seems already to have achieved something along these lines even
before he wrote Die Grundlagen. In a letter written in August 1882, he says:
I have now nearly completed a book in which I treat the concept of number and
demonstrate that the first principles of computation, which up to now have generally
been regarded as unprovable axioms, can be proved from definitions by means of logical
laws alone, so that they may have to be regarded as analytic judgements in Kant’s sense.
It will not surprise me and I even expect that you will raise some doubts about this
and imagine that there is a mistake in the definitions, in that, to be possible, they
presuppose judgements which I have failed to notice, or in that some other essential
content from another source of knowledge has crept in unawares. My confidence that
this has not happened is based on the application of my concept-script, which will not
let through anything that was not expressly presupposed… (Frege, 1980, 99–100)
Nonetheless, it would be more than a decade after Frege wrote those words,
and nine years after the publication of Die Grundlagen, before he actually
would provide the gap-free proofs he had promised.2
Those proofs are contained in Part II of Grundgesetze. Part I of the book is
devoted to the ‘Exposition of the Concept-Script’, that is, to the explanation
of the formal language in which Frege’s proofs will be stated and of the formal
theory in which they will be developed, that is, of the basic laws and rules of
inference of his system (see Heck, 2012, Part I). Part II contains the ‘Proofs
of the Basic Laws of Cardinal Number’. All of the Dedekind–Peano axioms
for arithmetic are proven there, including the statement that every natural
number has a successor, whose proof Frege had sketched in §§82–3 of Die
Grundlagen.
It is therefore clear that Part II of Grundgesetze plays an important role
in Frege’s philosophy of mathematics. That makes it really quite astonishing
that it has only recently been published in English translation (Frege, 2013).3
There are, of course, several reasons why that is. Frege’s formal system is, as is
well-known, inconsistent, since Russell’s Paradox is derivable in the concept-
script from Frege’s Basic Law V. One might therefore suppose that Frege’s
proofs can be of little interest, since anything can be proven in an inconsistent
system. There is also the problem of Frege’s notation, which is utterly unlike
that used by any other author and which has a reputation for being difficult to
2 I discuss some of the reasons for the delay in my other contribution to this volume (Heck,
2019).
3 There were no translations at all available when I started working on Grundgesetze in the
early 1990s, until Jason Stanley and I did a very (very) rough one in the summer of 1992. That
was used in a seminar George Boolos and I taught together in 1993. It was one of the first things
I put on my web site when I got one, around 1996, and at least a few other people used it in
seminars of their own.
i i
i i
i i

i i
The Basic Laws of Cardinal Number 3
read.4 In fact, however, those of us who have learned to read it know that it is
not difficult to read. Rather, its unfamiliarity makes it something of a challenge
to learn to read. And we have known since the mid-1980s that Frege’s system,
though inconsistent, is not irremediably inconsistent. As was first observed
by Peter Geach (1955, 570), and emphasized shortly thereafter by Charles
Parsons (1995, 198), Frege’s own arguments in Die Grundlagen make very
limited appeal to Basic Law V, which is the source of the inconsistency. Law
V is used only in the proof of what is now known as “Hume’s Principle”, or
HP: The number of F s is the same as the number of Gs if, and only if, the F s
are in one–one correspondence with the Gs. The remainder of the argument
appeals only to HP. And, as Crispin Wright (1983, 154–8) conjectured and
several people then proved (Burgess, 1984; Hazen, 1985; Boolos, 1998a), HP
is consistent. So, as Wright (1983, §xix) showed in detail, Frege’s proofs in Die
Grundlagen can be reconstructed in a consistent sub-theory of the inconsistent
theory he implicitly assumes.
The obvious question, which George Boolos directed to me in the summer
of 1991, is whether something similar but stronger is true of Grundgesetze. In
Frege: Philosophy of Mathematics, which was published that year, Sir Michael
Dummett seems to assert that there is:
Crispin Wright devotes a whole section of his book … to demonstrating that, if we
were to take [HP] as an implicit or contextual definition of the cardinality operator,
we could still derive all the same theorems as Frege does. He could have achieved the
same result with less trouble by observing that Frege himself gives just such a derivation
of those theorems. He derives them from [HP], with no further appeal to his explicit
definition. (Dummett, 1991, 123)
What Boolos asked me was simply whether this is true. I set to reading Grund-
gesetze and soon discovered that, if it was, it was going to take work to show
it. It is easy enough to verify that, after proving HP, Frege makes “no further
appeal to his explicit definition”. But that is not enough. The crucial ques-
tion is whether Frege makes no further appeal to Basic Law V, and he most
certainly does. Hardly a page of Part II lacks terms for value-ranges, of which
extensions of concepts are a special case, and the logical law governing value-
range names is, of course, Basic Law V. More precisely, due to the details of
how Frege formalizes various notions, almost every result he proves depends
upon his Theorem 1, which is a generalization of the principle known as naïve
comprehension:
a ∈ {x : F x} ≡ F a
And that principle leads directly to Russell’s Paradox, once we take F ξ to be:
ξ∈
/ ξ and a to be: {x : x ∈/ x}.5
4 Not to mention typeset. I scanned the formulas for the translation Jason and I did. How
they were handled in the new translation is well-related by Ebert and Rossberg (Frege, 2013,
xxx–xxxii).
5 Theorem 1 itself is proven from Basic Law V and Frege’s definition of the analogue, for
value-ranges, of membership (Heck, 2012, §1.2).
i i
i i
i i

i i
But it also quickly became clear to me that many of the uses Frege makes
of value-ranges can easily be eliminated. For example, Frege almost always
quantifies over the extensions of concepts instead of over concepts, so that we
find things like:
∀f (. . . a ∈ f . . . )
rather than things like:
∀F (. . . F a . . . )
But, as just illustrated, this is easily remedied. And Frege’s other uses of value-
ranges proved to be eliminable as well. So Dummett turned out to be right, in
spirit if not in detail: Modulo uses of value-ranges that are essentially just for
convenience, Part II of Grundgesetze really does contain a formal derivation
of axioms for arithmetic from HP.
And there is much more in Part II. Frege’s proof of axioms for arithmetic
comprises only about a third of it. In the remainder, Frege proves a number of
results concerning finitude, infinity, and the relationship between these two
notions. When I examined those proofs closely, it turned out that Frege used
Law V in them, too, only for convenience. And there was much of philosoph-
ical interest both in Frege’s formal arguments and in the informal discussion
of them contained in the “Analysis” sections.
My goal in this chapter, then, is to provide a brief overview of what Frege
accomplishes in Part II and to give some indication of the philosophical and
historical interest this material has. Further details, and actual arguments for
the interpretive claims to be made below, can be found in Part II of my book
Reading Frege’s Grundgesetze, of which this chapter is a kind of précis.
And since this chapter is meant to provide an introduction to Frege’s
formal work on arithmetic, I will present his results using modern notation,
so as to make the discussion more accessible. I will also silently translate away
Frege’s reliance upon value-ranges, since that serves only to obscure his accom-
plishments.6
1.1 THE PROOF OF HP

Frege’s first task in Grundgesetze is to prove HP, which may be stated, in mod-
ern notation, as:
Nx : F x = Nx : Gx ≡ ∃R[∀x∀y∀z(Rxy ∧ Rxz → y = z) ∧
∀x∀y∀z(Rxz ∧ Ryz → x = y) ∧
∀x(F x → ∃y(Gy ∧ Rxy) ∧
∀y(Gy → ∃x(F x ∧ Rxy)]
Here, ‘Nx : F x’ is to be read: the number of F s.

6 I shall also silently alter some of the translations from which I quote, to make them uniform
in their terminology.
i i
i i
i i

i i
Frege’s formulation of HP might initially seem very different, and not just
because his notation is so different.7 Translating Frege’s notation into ours, of
course, he would write HP as:
Nx : F x = Nx : Gx ≡ ∃R[Map(R)(F, G) ∧ Map(Conv(R))(G, F )]
Here, ‘Map(R)(F, G)’, which Frege would write as ‘f SgS⟩r’,8 may be read:
R maps the F s into the Gs. Conv(R), which Frege would write as ‘Ur’, is the
converse of R, defined the obvious way:
df
Conv(R)(a, b) ≡ Rba
So HP itself, as Frege would formulate it, says that the number of F s is the
same as the number of Gs if, and only if, there is a relation that maps the F s
into the Gs and whose converse maps the Gs into the F s.
The mapping relation itself is defined as follows (Grundgesetze I, §38):
df
Map(R)(F, G) ≡ Func(R) ∧ ∀x(F x → ∃y(Rxy ∧ Gy))
Here, ‘Func(R)’, which Frege would write as ‘Ir’, means that R is “single-
valued” or “functional”. It too is defined the obvious way (Grundgesetze I,
§37):
df
Func(R) ≡ ∀x∀y(Rxy → ∀z(Rxz → y = z))
So R maps the F s into the Gs just in case R is single-valued and each F is
related by Rξη to some G. Note carefully: into, not onto. That R maps the F s
into the Gs says, of itself, nothing whatsoever about the relative cardinalities
of the F s and the Gs: As long as there is at least one G, there will always be
a relation which maps the F s into the Gs, in Frege’s sense, whatever concept
F ξ may be.
To see the relation of Frege’s formulation of HP to the usual one, unpack
the right-hand side of his version HP using the definitions:
∃R[Map(R)(F, G) ∧
Map(Conv(R))(G, F )]
7 Theorem 32, which is the right-to-left direction, reads:

”u = ”v
uS(vS⟩q)
vS(uS⟩Uq)
Talk about different! Green, Rossberg, and Ebert (2015) discuss Frege’s notation in detail.
8 Here and below, I shall use uppercase letters for concepts and relations and the corresponding
lowercase letters for the extensions of those concepts and relations.
i i
i i
i i

i i
∃R[Func(R) ∧
∀x(F x → ∃y(Rxy ∧ Gy)) ∧
Func(Conv(R)) ∧
∀x(Gx → ∃y(Conv(R)(x, y) ∧ F y))]
∃R[∀x∀y(Rxy → ∀z(Rxz → y = z)) ∧

∀x(F x → ∃y(Rxy ∧ Gy)) ∧
∀x∀y(Conv(R)(x, y) → ∀z(Conv(R)(x, z) → y = z)) ∧
∀x(Gx → ∃y(Conv(R)(x, y) ∧ F y))]
∃R[∀x∀y(Rxy → ∀z(Rxz → y = z)) ∧

∀x(F x → ∃y(Rxy ∧ Gy)) ∧
∀x∀y(Ryx → ∀z(Rzx → y = z)) ∧
∀x(Gx → ∃y(Ryx ∧ F y))]
What Frege has done is group ‘∀x∀y(Rxy → ∀z(Rxz → y = z))’ and

‘∀x(F x → ∃y(Gy∧Rxy)’ in the first conjunct, ‘Map(R)(F, G)’, and to group
‘∀x∀y(Rxz → ∀z(Ryz → x = y))’ and ‘∀y(Gy → ∃x(F x ∧ Rxy)’ in the
second conjunct, ‘Map(Conv(R))(G, F )’. We are more inclined nowadays to
group the conjuncts ‘∀x∀y(Rxy → ∀z(Rxz → y = z))’ and ‘∀x∀y(Rxz →
∀z(Ryz → x = y))’ (R is a one–one function…) and ‘∀x(F x → ∃y(Gy ∧
Rxy)’ and ‘∀y(Gy → ∃x(F x ∧ Rxy)’ (… from the F s onto the Gs). So, in
the end, the difference between Frege’s formulation and ours is mostly one of
emphasis, though Frege’s formulation has some technical advantages over the
usual one (Heck, 2012, §6.3).
In modern presentations, ‘Nx : F x’ is treated as a primitive notion gov-
erned by HP, which is itself treated as an axiom. Frege, by constrast, means
to prove HP and so defines ‘Nx : F x’ in terms of extensions.9 Now, in Grund-
gesetze, Frege treats extensions as a kind of value-range, but his definition of
number in Grundgesetze is otherwise the same as the one given in §68 of Die
Grundlagen:10
The number of F s is the extension of the concept: is [the extension of ]
a concept that is equinumerous with F .
We can formalize this as:
df
Nx : F x ≡ x̂{∃G[(x = ŷ(Gy) ∧ Eq(F, G)]}
9 What most obviously corresponds to our ‘Nx : F x’ is Frege’s ‘”f ’. In fact, however, as
– ε’.
Gregory Landini pointed out to me, ‘Nx : F x’ is definable in Frege’s system as ‘” εF
10 Frege’s definition does not contain the bracketed occurrence of the phrase ‘the extension of ’.
I argue elsewhere (Heck, 2019, §18.1) that it is nonetheless what he means.
i i
i i
i i

i i
where ‘Eq(F, G)’ abbreviates: ∃R[Map(R)(F, G) ∧ Map(Conv(R))(G, F )].

Here ‘x̂(F x)’ means: the extension of the concept F , and the notion of ex-
tension is to be governed by a version of Basic Law V:11
x̂(F x) = x̂(Gx) ≡ ∀x(F x ≡ Gx)
The proof of HP then needs little more than the observation that Eq is an
equivalence relation.
In fact, however, as May and Wehmeier (2019) point out in their contri-
bution to this volume, Frege never actually proves HP as a biconditional: He
proves its two directions, but never bothers to put them together. The right-
to-left direction is Theorem 32, which is the goal of the first chapter of Part II,
Chapter Alpha.12 The left-to-right direction is Theorem 49, which is proven in
Chapter Beta. The proof of the former is quite straightforward, and it follows
the outline in §73 of Die Grundlagen closely. The proof needs only the trans-
itivity and symmetry of equinumerosity. The proof of (49) that Frege gives is
somewhat peculiar, because it uses the definition of number in a more essen-
tial way than it really should. There is, however, a simpler proof, which Frege
must have known, that needs only the reflexivity of equinumerosity (Heck,
2012, §6.8).
Another point worth noting about these proofs is that the proof of (32)
needs only the right-to-left direction of Law V, which Frege calls Law Va and
which is the “safe” direction, whereas (49) needs the left-to-right direction,
Law Vb, which is the “unsafe” direction, the one that gives rise to Russell’s
Paradox. On reflection, this should not be surprising, since (32) is the “safe”
direction of HP, which by itself has no significant ontological consequences,
since it is compatible with there being only one number, shared by all the con-
cepts there are. Theorem 49, on the other hand, is the ontologically profligate
direction of HP, which entails the existence of infinitely many numbers.
1.2 THE AXIOMS OF ARITHMETIC
After having proven HP, Frege turns his attention to the proofs of various fun-
damental principles concerning cardinal numbers, including what we now call
the Dedekind–Peano axioms, for which see Table 1.1. Here ‘Nξ ’ is a predicate
to be read ‘ξ is a natural number’, and ‘Pξη ’ is a predicate to be read as ‘ξ
immediately precedes η in the number-series’. To prove these axioms, Frege
must of course define the arithmetical notions that occur in them.
11 As said, Frege actually works with the more general notion of a value-range, but, surpris-
ingly, he never makes use of the more general notion. All the value-ranges in which he is actually
interested in Grundgesetze are extensions of concepts.
12 Frege does not call these divisions chapters, but it seems the obvious name for them.
i i
i i
i i

i i
Table 1.1. One Version of the Dedekind–Peano Axioms

1. N0
2. ∀x∀y(Nx ∧ Pxy → Ny)
3. ∀x(Nx → ∃y(Pxy))
4. ¬∃x(Nx ∧ Px0)
5. ∀x(Nx → ∀y∀z(Pxy ∧ Pxz → y = z))
6. ∀x∀y∀z(Nx ∧ Ny ∧ Pxz ∧ Pyz → x = y)
7. ∀F [F 0 ∧ ∀x(Nx ∧ F x → ∀y(Pxy → F y)) → ∀x(Nx → F x)]
The definitions Frege gives in Grundgesetze are the same as the ones given
in Die Grundlagen. Frege defines zero, which he writes as ‘0’, as the number of
objects that are not self-identical (Grundgesetze I, §41; see Frege, 1884, §74):
df
0 ≡ Nx : x ̸= x
His definition of predecession, which he writes as ‘mS(nSs)’, is as follows

(Grundgesetze I, §43):
df
Pmn ≡ ∃F ∃x[F x ∧ n = Nz : F z ∧ m = Nz : (F z ∧ z ̸= x)]
That is, m precedes n if, as Frege puts it in Die Grundlagen, ‘there exists a
concept F , and an object falling under it x, such that the Number which
belongs to the concept F is n and the Number which belongs to the concept
“falling under F but not identical with x” is m’ (Frege, 1884, §76). We shall
return to the definition of ‘Nξ ’.
Frege proves Axiom 5 in Chapter Beta as Theorem 71; Axiom 6 in Chapter
Γ as Theorem 89; and Axiom 4 in Chapter Epsilon as Theorem 108. The
proofs are straightforward, but there is a philosophical discussion that occurs
during Frege’s informal exposition of the proof of Theorem 71 that is of sub-
stantial interest. It concerns the proof of Theorem 66:
F c ∧ Gb ∧ Nz : (Gz ∧ z ̸= b) = Nz : (F z ∧ z ̸= c) →
Nz : F z = Nz : Gz (66)
which is the key lemma in the proof of (71).13 To prove (66), what we want
to show is that, if there is a one–one correlation between the Gs other than
13 Suppose that Pxy and Pxw. Then, by the definition of ‘P’, there are F and c such that:
F c ∧ Nz : F z = y ∧ Nz : (F z ∧ z ̸= c) = x
and there are G and b such that:
Gb ∧ Nz : Gz = w ∧ Nz : (Gz ∧ z ̸= b) = x
So Nz : (Gz ∧ z ̸= b) = x = Nz : (F z ∧ z ̸= c), and (66) now implies that Nz : Gz = Nz : F z ,
so w = y , and we are done.
i i
i i
i i

i i
b and the F s other than c, and if b is a G and c is an F , then there is also a

one–one correlation between the F s and the Gs. Frege writes:14
If one were to follow the usual practice of mathematicians, one might say something
like this: we correlate the objects, other than b, falling under the concept G, with the
objects, other than c, falling under the concept F by means of the known relation,
and we correlate b with c. In this way, we have mapped the concept G into the concept
F and, conversely, the latter into the former. So … the cardinal numbers that belong
to them are equal. This is indeed much briefer than the proof to follow which some,
misunderstanding my project, will deplore on account of its length. What is it that we
are doing when we correlate objects for the purpose of a proof? (Grundgesetze I, §66)
There is, of course, nothing unusual about the sort of reasoning Frege re-
hearses, but he has a question to raise about it. It is not, of course, that he
thinks such reasoning might be invalid. But he wants to know what justifies
it.
Frege first emphasizes that, when we establish a correlation in this sense,
we do not create anything but “merely bring to attention, apprehend, what is
already there” (compare Frege, 1884, §26). He then seizes the opportunity to
take a swipe at psychologism. Having slain that familiar foe, Frege considers
the question how we might formulate “a postulate, in the style of Euclid”, that
permits such correlations, answering:
[It] would have to be understood this way: ‘Any object is correlated with any object’
or ‘There is a correlation between any object and any object’. What then is such a
correlation if it is nothing subjective, created only by our making? However, a particular
correlation of an object to an object is not what can be at issue here…; rather we require
a genus of correlations, so to speak, what we have so far called, and will continue to
call, a relation. (Grundgesetze I, §66)
So what the act of correlating brings to attention is a relation, and Frege goes
on to discuss how the concept-script allows us to specify such relations pre-
cisely and to show that they have the various properties we need them to have.
In fact, in this case, Frege goes on to suggest, the usual reasoning encourages
us to overlook certain subtleties, which emerge in his careful, rigorous present-
ation of the proof.
What is at issue here is a question fundamental to Frege’s logicism. The
informal argument he mentions involves just the sort of toxic mix of reason
and intuition that tends to obscure the epistemological status of the result
proved. One crucial question, in particular, is how we know that the “correl-
ations” we need actually exist. And, so long as one thinks of correlating as
something we do, rather than of correlations as something we discover, this
will tend to make one suppose that the existence of correlations depends some-
how upon mental activity. Frege’s view is, of course, different. In his system,
the existence of concepts, relations, and the like is guaranteed by Rule 9 on
14 In his exposition, Frege speaks of such things as “the u-concept”, by which he means the
concept whose value-range u is. I have silently replaced such talk with direct references to con-
cepts.
i i
i i
i i

i i
the list given in §48. This rule allows for the uniform replacement of a free
variable, of arbitrary type, by any well-formed expression of the appropriate
type. Such a substitution principle is equivalent to the comprehension scheme
of second-order logic
∃R∀x∀y[Rxy ≡ A(x, y)]
which asserts the existence of a relation co-extensive with any given formula of
the language (so long as it does not contain the variable R free). So Frege’s view
is that logic itself commits us to the existence of any relation we can describe.
Since “correlations” are just relations, it is thus logic, and not psychology, that
affirms their existence.
To return to the proofs of the axioms, then, what remain to be proved at the
end of Chapter Epsilon are the axioms concerning the notion of a natural (or
finite) number. So we now need to consider Frege’s definition of that notion.
It is, again, essentially the same as that given in Die Grundlagen, and it uses
Frege’s definition of the so-called ancestral, which he introduces in §23 of
Begriffsschrift. Given a relation Q, we say that a concept F is hereditary in the
Q-series just in case, whenever x is F , each object to which Q relates it is F :
∀x∀y(F x ∧ Qxy → F y)
We now say that an object b follows an object a in the Q-series just in case b
falls under every concept that is hereditary in the Q-series and under which
each object to which Q relates a falls. Formally, writing ‘Q∗ ab’ for ‘b follows
a in the Q-series’, Frege’s definition of the strong ancestral,15 which he writes
as ‘aS(bSMq)’, is (Grundgesetze I, §45):
df
Q∗ ab ≡ ∀F [∀x(Qax → F x) ∧ ∀x∀y(F x ∧ Qxy → F y) → F b]
Frege then defines the weak ancestral, which he writes as ‘aS(bSRq)’, thus
(Grundgesetze I, §46):
df
Q∗= ab ≡ Q∗ ab ∨ a = b
The concept Nξ is then definable as P∗= 0ξ . So an object is a natural number

just in case it belongs to the P-series beginning with 0. Frege has no special
symbol for this. He does, however, regularly read ‘P∗= 0ξ ’ as ‘ξ is a finite num-
ber’ (e.g., in Grundgesetze I, §108).
Axioms 1 and 2 then follow from general facts about the ancestral, The-
orems 140 and 133, respectively. Famously, Axiom 7, the induction axiom,
also follows from Frege’s definition. Matters are more complicated than often
seems to be supposed, however. We can quite easily prove:
Q∗= ab → ∀F (F a ∧ ∀x∀y(F x ∧ Qxy → F y) → F b) (144)
15 The strong ancestral is so-called because we need not have Q∗ aa, whereas we always have
Q∗= aa.
i i
i i
i i

i i
from which
P∗= 0b → ∀F (F 0 ∧ ∀x∀y(F x ∧ Pxy → F y) → F b)
follows by substitution. So we have
∀F [F 0 ∧ ∀x∀y(F x ∧ Pxy → F y) → ∀x(Nx → F x)]
by simple logical manipulations and the definition of ‘N’. But this is weaker
than Axiom 7. The hypothesis of induction is not that, whenever x is F , its
successor is F ; it is only that, whenever x is a natural number that is F , its
successor is F . That is, induction is:
∀F [F 0 ∧ ∀x∀y(Nx ∧ F x ∧ Pxy → F y) → ∀x(Nx → F x)]
This is easy enough to prove—it follows by substitution from Frege’s Theorem

152—but the difference between it and what the definition of the ancestral
delivers immediately turns out to be historically significant, as we shall see
shortly.16
The only remaining axiom, then, is Axiom 3, which asserts that every num-
ber has a successor. Lying behind Frege’s proof of Axiom 3 is a picture of how
the natural numbers are generated. The generative process begins, of course,
with zero. Frege insists that zero exists, even if nothing else does, because zero
is the number of things that are non-self-identical, and the non-self-identical
things exist even if nothing at all does. But if zero exists, then there is a num-
ber that is the number of things that are less than or equal to zero, and that
number we may call ‘one’. By HP, one is not zero: There can be no one–one
map between the things that are non-self-identical and the things that are less
than or equal to zero, since there is at least one thing less than or equal to zero,
namely, zero. But now both zero and one exist, and so there is a number that
is the number of things that are less than or equal to one. Call that number
‘two’. By HP, two can be neither zero nor one. So zero, one, and two exist,
and there is a number that is the number of things less than or equal to two…
Formally, what we want to prove is:
P∗= 0b → P(b, Nx : P∗= xb) (155)
which is the central result of Chapter H (Eta) and which says, roughly, that
every natural number is succeeded by the number of numbers in the P-series
ending with it (roughly, the number of numbers less than or equal to it). The
proof proceeds by induction, where we substitute:
P(ξ, Nx : P∗= xξ)
for ‘F ξ ’. So we need to prove that zero falls under this concept:
16 The difference is also of some technical import, since the weaker principle is easily provable
even in predicative systems that do not allow us to prove induction (Heck, 2011).
i i
i i
i i

i i
P(0, Nx : P ∗= x0) (154)

and that it is hereditary in the P-series beginning with 0:
∀y[P∗= 0y ∧ P(y, Nx : P∗= xy) → ∀z(Pyz → P(z, Nx : P∗= xz))] (150)
The proof of (154) is easy.
The interest lies in the proof of (150). It follows by generalization from:
P∗= 0d ∧ P(d, Nx : P∗= xd) ∧ Pda → P(a, Nx : P∗= xa) (150ε)
To prove this, suppose that d is a natural number, that d precedes the number
of members of the P-series ending with d, and that d precedes a. We must
show that a precedes the number of members of the P-series ending with a.
To do so, we must, by the definition of P, find some concept F and some
object x falling under F such that a is the number of F s other than x and the
number of F s is the same as the number of members of the P-series ending
in a. That is, we must show that:
∃F ∃x[a = Nz : (F z ∧ z ̸= x) ∧ F x ∧ Nz : P∗= za = Nz : F z]
The concept in question is to be P∗= ξa; the object in question is to be a itself.
Hence, we must show that:
a = Nz : (P∗= za ∧ z ̸= a) ∧ P∗= aa ∧ Nz : P∗= za = Nz : P∗= za
The last two conjuncts are trivial. The first we may derive from:
a = Nx : P∗= xd
Nx : P∗= xd = Nx : (P∗= xa ∧ x ̸= a)
by the transitivity of identity. The former follows from the fact that P is single-
valued, since, by hypothesis, we have both Pda and P(d, Nx : P∗= xd). The
latter, in turn, is the consequent of:
P∗= 0a ∧ Pda → Nx : P∗= xd = Nx : (P∗= xa ∧ x ̸= a) (149)
which Frege derives from:
P∗= 0a ∧ Pda → ∀x{P∗= xd ≡ (P∗= xa ∧ x ̸= a)} (149α)
and the extensionality of the cardinality operator:
∀x(F x ≡ Gx) → Nx : F x = Nx : Gx (96)
He derives Theorem 149α from the following two results:
Pda → [(P∗= xa ∧ x ̸= a) → P∗= xd]
Pda ∧ P 0a → [P∗= xd → (P∗= xa ∧ x ̸= a)]
∗=
which are Theorems 148α and 148ζ . For the latter, we need the central result
of Chapter Zeta:
P∗= 0b → ¬P∗ bb (145)
This says that there are no “loops” in the natural series of numbers, and it too
is proved by induction.
i i
i i
i i

i i
The argument here will look familiar to anyone who has studied Frege’s
informal proof, in §§82–3 of Die Grundlagen, that every natural number has
a successor. The proof in Grundgesetze has much the same structure, and many
of the steps in the proof are also mentioned in Die Grundlagen. For example,
(145) is the last proposition mentioned by Frege in §83. But there are im-
portant differences between the two proofs, as well. A close reading of Frege’s
discussion of the proof in §114 of Grundgesetze shows, in fact, that Frege was
aware that the proof sketched in Die Grundlagen is actually incorrect (Boolos
and Heck, 2011; Heck, 2012, §6.7). The earlier proof purports to rely only
upon what, as mentioned before, trivially follows from the definition of nat-
ural number:
∀F [F 0 ∧ ∀x∀y(F x ∧ Pxy → F y) → ∀x(Nx → F x)]
rather than upon mathematical induction proper:

∀F [F 0 ∧ ∀x∀y(Nx ∧ F x ∧ Pxy → F y) → ∀x(Nx → F x)]
In particular, the proof in Die Grundlagen was supposed to go via:

P(d, Nx : P∗= xd) ∧ Pda → P(a, Nx : P∗= xa)
which is a direct formalization of the proposition marked ‘1.’ in §82 of Die
Grundlagen and which is also (150ε) minus its first conjunct, P∗= 0d. Frege
mentions this formula explicitly in §114—it is the formula labeled (α)—only
then to say, in a footnote, that it “is, it seems, unprovable…”. It is hard to
see why Frege would so much as have discussed this proposition if it did not
figure crucially in his earlier argument. And that, to me, is the most impressive
evidence that Frege knew his earlier proof was flawed.
It follows, presumably, that Frege cannot actually have had a fully worked
out, formal proof of the existence of successors when he wrote Die Grundlagen.
He simply could not have made such a mistake otherwise.17 More interest-
ingly, it makes it plain that Frege was aware that there might be propositions
of his formal language that he would regard as true but that were nonetheless
unprovable from the Basic Laws he was then prepared to accept. For there
is good reason to think Frege regarded the “unprovable” formula (α) as be-
ing true. There is a very simple argument for it that uses Dedekind’s result
that every infinite set is Dedekind infinite, a claim that Frege accepted as true,
though, like many mathematicians of his time, he regarded Dedekind’s proof
as insufficiently rigorous (Frege, 1892, 271). I’ll return to this point below.
Finally, careful analysis of Frege’s proof in Grundgesetze reveals that the
only facts about numbers to which it essentially appeals are the extensionality
17 And that, in turn, suggests that the manuscipt mentioned in the letter from 1882 cannot
have contained such a proof, which makes it an interesting question what it did contain. Frege
says in the letter that he has set out to prove “the first principles of computation”, and the existence
of successors is not naturally so described.
i i
i i
i i

i i
of the cardinality operator (96), the fact that P is one–one, and the fact that
zero has no predecessor. Let me say that again: Once P has been shown to
be one–one and zero has been shown to have no predecessor, the only further
appeal to HP that is needed is that required to prove Theorem 96. But as
Boolos (1998b, 278ff) has emphasized, we might plausibly regard (96) as a
logical truth. So, in a sense, HP is not really involved in the proof that every
number has a successor, and we might reasonably regard Frege as having shown
that the existence of successors is a logical consequence of the other axioms
concerning predecession, given how he defines that notion. Frege does not
present the proof in sufficient generality to make this obvious, but various
remarks he makes along the way make it plausible that he knew his proof had
this consequence (Heck, 2012, §6.6).
1.3 THE INFINITE
By the end of Chapter H (Eta), then, Frege has proven all of the Dedekind–
Peano axioms. As mentioned earlier, however, Frege’s investigation of the Basic
Laws of Cardinal Number is by no means complete at that point.
Chapter Iota is divided into four pieces. In the first, I(a), Frege proves the
existence of an infinite cardinal number, namely, the one we know as ℵ0 but
which Frege calls ‘Endlos’ and writes: i. It is defined as the number of natural
numbers (Grundgesetze I, §122):
i = Nx : P∗= 0x
df
The key result is that Endlos is not a natural number, i.e., that it is infinite:
¬P ∗= (0, i) (167)
Given (145), it is enough to show that Endlos follows itself in the natural
series of numbers:
P(i, i) (165)
To prove this, we need only find a concept F and an object a such that
i = Nx : F x ∧ F a ∧ i = Nx : (F x ∧ x ̸= a)
We may take F ξ to be: P∗= 0ξ and a to be 0. So we have to prove:
i = Nx : P∗= 0x ∧ P∗= 00 ∧ i = Nx : (P∗= 0x ∧ x ̸= 0)
The first two conjuncts are trivial. To prove the last, Frege shows, as Theorem
165β , that P itself correlates the natural numbers one–one with the natural
numbers other than zero. Frege thus makes good here on a promise he had
made in §§84–6 of Die Grundlagen to secure the existence not just of infinitely
many numbers but also of infinite numbers, purely logically.
i i
i i
i i

i i
The most interesting result of Chapter Iota, however, is the one proven in
I(d):
∃Q[Func(Q) ∧ ¬∃x(Q∗ xx) ∧

∀x(Gx → ∃y(Qxy)) ∧ ∃x∀y(Gy ≡ Q∗= xy)] →
N x : Gx = i (263)
Suppose that there is a relation Q satisfying the following conditions: First,

it is single-valued; second, no object follows after itself in the Q-series; third,
each G stands in the Q-relation to some object; and finally, the Gs are the
members of the Q-series beginning with some object. Then, says Theorem
263, the number of Gs is Endlos.
Frege explains the strategy of his proof in in §144:
We now prove … that Endlos is the cardinal number which belongs to a concept, if the
objects falling under that concept can be ordered in a series that starts with a certain
object and proceeds endlessly, without looping back into itself and without branching.
By an “unbranching” series, Frege means one whose determining relation is
single-valued; by a series that does not “loop back into itself ”, he means one
in which no object follows after itself; by a series that “proceeds endlessly”, he
means one every member of which is immediately followed by some object.
It turns on showing that Endlos is the cardinal number that belongs to the concept
member of such a series … We use proposition (32) for this and need to prove that
there is a relation which maps the cardinal number series into the Q-series starting
with x and whose converse maps the latter into the former. The obvious strategy is to
correlate 0 with x, 1 with the next member immediately following after x in the Q-
series, and, in this manner, to correlate each immediately following cardinal number
to the immediately following member of the Q-series. We always pair one member of
the cardinal number series with one member of the Q-series and form a series out of
these pairs. … If, then, the pair (n; y) belongs to our series that starts with the pair
(0; x), n stands to y in the mapping relation that is to be exhibited.
We have two series, which we may picture thus:
P P P P P
0 ! 1 ! ··· ! m ! n ! ···
x = x0 ! x1 ! · · · ! xm ! xn ! · · ·
Q Q Q Q Q
and the result is thus to be proven by defining, by recursion, a relation between

the natural numbers and the members of the Q-series beginning with x: The
number n which is the immediate successor of a given number, m, will be
related to the member of Q-series, call it xn , that follows immediately after
the member of the Q-series to which m is related, say, xm . The proof of the
theorem will require a proof of the validity of such definitions. Frege’s sugges-
tion is that we may define this relation by defining a series of ordered pairs,
namely, the series (0; x0 ), (1; x1 ), etc., where, in general, (m; x) will stand in
i i
i i
i i

i i
the “series-forming relation” to (n; y) just in case Pmn and Qxy . The members
of this series will then be the extension of the relation to be defined.
The main work in Frege’s proof thus consists in showing how to define
relations like this one recursively. There is a great deal of messiness here caused
by the way Frege uses ordered pairs in his proof. His reasons for doing so are
technical, and he seems to have been aware that this use was eliminable (Heck,
2012, §7.2). For present purposes, then, I shall ignore these complications.
One of the most important lemmas in the proof of (263) is:
Func(Q) ∧ ∀y(Q∗= ay → ∃z(Qyz)) →

Map((PPQ)h0,a )(P∗= 0ξ, Q∗= aξ) (256)
Here, (PPQ)h0,a is the relation defined in the way just explained. (PPQ) is the
“coupling” of the relations P and Q, defined so that
(PPQ)((a; b), (c; d)) ≡ Pac ∧ Qbd
(PPQ)h0,a is then defined in such a way that
(PPQ)h0,a (x, y) ≡ (PPQ)∗= ((0; a), (x; y))
So (256) says that, if Q is single-valued and the Q-series beginning with a is

endless—if every member of that series is followed by another—then (PPQ)h0,a
maps the natural numbers into the members of the Q-series beginning with
a. But if we eliminate ‘Map’ via its definition, then we have:
Func(Q) ∧ ∀y(Q∗= ay → ∃z(Qyz)) →

Func((PPQ)h0,a ) ∧ ∀x[P ∗= 0x → ∃y(Q∗= ay ∧ (PPQ)h0,a (x, y))]
And what this says is that, if Q is single-valued, and if the Q-series beginning
with a is endless, then (PPQ)h0,a is a single-valued relation and every natural
number is in its domain. That is, (PPQ)h0,a is a total function on the natural
numbers. Moreover, it is not difficult to see that its range contains only mem-
bers of the Q-series beginning with a. So (256) allows for the definition of
functions by recursion.
The usual set-theoretic statement of this sort of result is:
Suppose g : A → A; let a ∈ A. Then there is a unique function φ : N →
A such that φ(0) = a and φ(Sn) = g(φ(n)).
Assume the antecedent and define Qξη ≡ [η = g(ξ)]. Q is then single-valued,
since g(ξ) is a function, and the Q-series beginning with a is endless, since
g is totally defined on its range. Thus, the antecedent of (256) is satisfied, so
(PPQ)h0,a is single-valued and its domain contains all the natural numbers.
Uniqueness is proven by induction.
i i
i i
i i

i i
As an example, consider the recursion equations:

φ(0) = a
φ(Sm) = S(φ(m))
As above, we define Qξη as η = Sξ ; thus, Qξη is just Pξη . So (PPP)h0,a (ξ, η)

is single-valued and satisfies these equations. We may write them, in more
familiar form, as:
a+0=a
a + Sm = S(a + m)
So the moral of the story is that
(PPP)h0,ζ (ξ, η)
defines η = ζ + ξ for natural numbers. I’ll leave it as an exercise to show how
multiplication can now be defined in terms of addition.
Theorem 256 follows from a more general result:
Func(R) ∧ ¬∃y(R∗= my ∧ R∗ yy) ∧
Func(Q) ∧ ∀x[Q∗= ax → ∃y(Qxy)] →
Map((RPQ)hm,a )(R∗= mξ, Q∗= aξ) (254)
which licenses the recursive definition of functions not just on the natural
numbers but on any “simple” series: one that is single-valued and contains no
“loops”. And Theorem 254 in turn follows from an even more general result:
∀y[Q∗= ay → ∃z(Qyz)] →
∀x{R∗= mx → ∃y[(RPQ)hm,a (x, y) ∧ Q∗= ay]} (241ζ )
which licenses the recursive definition of relations that may or may not be
single-valued on any endless series whatsoever. There are natural applications
to be made of (241ζ ). For example, let R be any relation. Then it is easy to
see that (RPP)ha,0 (b, n) just in case it is possible to get from a to b via a series
of n R-steps, that is, if we have
a = x0 Rx1 R · · · Rxn = b
where the xi need not be distinct. We’ll see an application of this below.
Perhaps the most intriguing result here, though, is one Frege does not
explicitly mention, though it is an obvious consequence of (254). Exchanging
‘R’ with ‘Q’, and ‘m’ with ‘a’, we have
Func(Q) ∧ ¬∃y(Q∗= ay ∧ Q∗ yy) ∧
Func(R) ∧ ∀x[R∗= mx → ∃y(Rxy)] →
Map((QPR)ha,m )(Q∗= aξ, R∗= mξ)
i i
i i
i i

i i
But Theorem 259 tells us that (QPR)ha,m is the converse of (RPQ)hm,a , so:
Func(Q) ∧ ¬∃y(Q∗= ay ∧ Q∗ yy) ∧

Func(R) ∧ ∀x[R∗= mx → ∃y(Rxy)] →
Map(Conv((RPQ)hm,a ))(Q∗= aξ, R∗= mξ)
Putting this together with (254), then, we have:
Func(Q) ∧ ¬∃y(Q∗= ay ∧ Q∗ yy) ∧ ∀x[Q∗= ax → ∃y(Qxy)] ∧

Func(R) ∧ ¬∃y(R∗= my ∧ R∗ yy) ∧ ∀x[R∗= mx → ∃y(Rxy)] →
Map((RPQ)hm,a )(R∗= mξ, Q∗= aξ) ∧
Map(Conv((RPQ)hm,a ))(Q∗= aξ, R∗= mξ)
I call this result the Isomorphism Theorem, because it tells us that all “simply
endless” series are isomorphic.18 Theorem 263 follows easily from it: Just sub-
stitute ‘P’ for ‘R’ and ‘0’ for ‘m’ and apply the axioms of arithmetic. In effect,
then, Frege’s proof of (263) proceeds by first proving that all series satisfying
certain conditions are isomorphic and then concluding that, since the series
of natural numbers satisfies those conditions, every such series is isomorphic
to it and hence equinumerous with it.
The conditions in question are:
1. Func(Q)
2. ¬∃x(Q∗= ax ∧ Q∗ xx)
3. ∀x(Q∗= ax → ∃y(Qxy))
But if we write the conditions slightly differently, we can see more easily what
Frege has done here. Write ‘0’ instead of ‘a’ and ‘P’ instead of ‘Q’ and intro-
duce a predicate ‘Nξ ’ in place of ‘Q∗= aξ ’, as in the statement of (263). Then,
eliminating ‘Func(P)’ via its definition, we have:
1. ∀x∀y∀z(Pxy ∧ Pxz → y = z)
2. ¬∃x(Nx ∧ P∗ xx)
3. ∀x(Nx → ∃y(Pxy))
4. Nx ≡ P∗= 0x
Conditions (1)–(4) are axioms for arithmetic, and we may think of them as
recording Frege’s own preferred axiomatization. The first condition here is,
of course, stated in the form in which Frege proves it: Predecession is single-
valued period, and not just on the natural numbers. But if we weaken it to:
1′ . ∀x(Nx → ∀y∀z(Pxy ∧ Pxz → y = z))
18 Strictly speaking, we would also need to prove that the orderings given by Q∗ and R∗ are
isomorphic. Frege does not prove this part of the result. Nor does Dedekind. But it is not hard
to prove.
i i
i i
i i

i i
then the Dedekind–Peano axioms are easily derived from Frege’s, and con-
versely. What the proof of Theorem 263 shows is thus that any two structures
satisfying Frege’s axioms for arithmetic are isomorphic, just as Dedekind’s
proof of the corresponding result shows that any two structures satisfying his
axioms are isomorphic (Dedekind, 1902, Theorem 132).
Frege proves the converse of Theorem 263 as Theorem 207, in I(c). Putting
the two together, we have:
N x : Gx = i ≡
∃Q[Func(Q) ∧ ¬∃x(Q∗ xx) ∧
∀x(Gx → ∃y(Qxy)) ∧ ∃x∀y(Gy ≡ Q∗= xy)]
Since the right-hand side is purely second-order, Frege would have regarded
it as uncontroversially logical. Theorems 207 and 263 thus yield a purely lo-
gical characterization of countably infinite concepts: They are the concepts
the objects falling under which can be ordered as a simply endless series.
1.4 THE FINITE

In Chapters K and Λ, Frege proves analogous results for finitude. To state
these results, we need another definition:
df
Btw(Q; a, b)(x) ≡ Func(Q) ∧ ¬Q∗ bb ∧ Q∗= ax ∧ Q∗= xb
Frege reads ‘Btw(Q; a, b)(x)’, which he writes as ‘xS(a; b) q ’, as: x belongs
y
to the Q-series running from a to b (Grundgesetze I, §158). But we may read
it, more briefly, as ‘x is Q-between a and b’. So x is Q-between a and b if,
and only if: Q is single-valued; b does not follow itself in the Q-series; and x
belongs both to the Q-series beginning with a and to that ending with b.
The central result of Chapter K is then:
∃Q∃x∃y∀z[F z ≡ Btw(Q; x, y)(z)] → P∗= (0, Nx : F x) (327)
and the central result of Chapter Λ is:
P∗= (0, Nx : F x) → ∃Q∃x∃y∀z[F z ≡ Btw(Q; x, y)(z)] (348)
Together, of course, these imply:
∃Q∃x∃y∀z[F z ≡ Btw(Q; x, y)(z)] ≡ P∗= (0, Nx : F x)
which constitutes a purely logical characterization of finite sets: A set is finite
if it can be ordered as a simple series that ends.
To understand the point of Frege’s so characterizing finitude, it is essential
to look at his proofs of these two theorems. Frege derives Theorem 327 from:
P∗= (0, Nz : Btw(Q; a, b)(z)) (325)
There is a trivial case here. It may be that nothing is Q-between a and b, be-
cause Q isn’t single-valued, or because b follows after itself in the Q-series, or
i i
i i
i i

i i
because b doesn’t belong to the Q-series beginning with a. The trivial case is
covered by (325γ ), so the work goes into proving:
Q∗= ab ∧ Func(Q) ∧ ¬Q∗ bb → P∗= (0, Nz : Btw(Q; a, b)(z)) (321)
The obvious way to prove this is to assume that Func(Q) and ¬Q∗ bb and then
to prove
Q∗= ab → P∗= (0, Nz : Btw(Q; a, b)(z))
by “logical induction” (Heck, 2012, §8.1). But Frege’s proof is different, and
much more interesting, though far more complicated.
Assume the F s have been ordered as a simple series that ends. Frege’s proof
rests upon the insight that, given such an ordering of the F s, they can be
mapped one–one onto an initial segment of the natural numbers. This picture
P P P P
0 ! 1 ! ··· ! m ! n
"R "R "R "R
a = x0 ! x1 ! · · · ! xm ! xn = b
Q Q Q Q
should make the idea clear enough, and also indicate the close relation between
this argument and the proof of (263). All we need to do is to define the rela-
tion R, which we can do by recursion, and then to show that it has the right
properties.
Frege does not proceed in quite this way, however, and his not so proceed-
ing is our best indication of the role (327) and (348) were intended to play
in his philosophy of arithmetic. What Frege does instead is to produce a one–
one mapping between the F s and the numbers between one and some natural
number n. On reflection, it is clear why. One way of producing a simple or-
dering of the F s is to count them, that is, to associate each one of them, in
turn, with a number, beginning with one and ending with some number n,
which is then the number of F s. Indeed, the relation Frege shows to correlate
the members of the relevant series with the numbers between 1 and n is the
relation: ξ is the η th member of the series, defined in the way described earlier
as: (QPP)ha,1 . This very correlation itself amounts to a counting of the F s. The
intuitive content of (327) and (348) is thus that the number of F s is a natural
number just in case the F s can be counted and that, in that case, the number
of F s is the natural number one reaches by counting.
This is especially clear from the proof of (348). Like (321), it admits of a
direct and utterly uninteresting proof by induction. Again, however, Frege’s
proof is more complicated than it needs to be and more illuminating than it
might have been. He derives (348) from two lemmas, the first of which is:
Nz : Gz = Nz : Btw(R; a, b)(z) → (347)

∃Q∃x∃y∀z[Gx ≡ Btw(Q; x, y)(z)]
i i
i i
i i

i i
In words: If the number of Gs is the same as the number of objects R-between

a and b, for some R, a, and b, then the Gs can be ordered as a simple series
that ends. The other lemma is:
P∗= 0n → n = Nz : Btw(P; 1, n)(z) (314)
So, if the number of Gs is a natural number, then the number of Gs is the same
as the number of numbers between 1 and the number of Gs; substituting into
(347) completes the proof.
That is to say, the proof of (348) actually proceeds via the following:
Nz : Gz = Nz : Btw(P; 1, n)(z) → ∃Q∃x∃y∀z[Gx ≡ Btw(Q; x, y)(z)]
which (347) generalizes. It is this which is really Frege’s goal in his proof of
(348), but Frege rarely passes up an opportunity to prove the most general
result possible. The proof proceeds by showing that, if the numbers between
1 and n can be mapped one–one onto the Gs, then the Gs can be ordered as a
simple series that ends: The relation that so orders the Gs is the image of the
predecessor relation under the relevant mapping. (The same picture as above
works here. It is just that we have R this time, and we need to define Q.) To
correlate the Gs one–one with the numbers between 1 and n is basically just
to count them. So Theorem 348 amounts to this: If the number of Gs is finite,
then they can be counted.
No doubt, (327) and (348) have a purely technical point: They yield a nice
characterization of finite sets, one closely related to Zermelo’s (Heck, 2012,
§8.5). But they also have an epistemological point: They show that the concept
of finitude is a concept of logic. Frege would have regarded his characterization
of finitude, given as it is in purely second-order terms, as uncontroversially
logical. But these two theorems have another point, too, for they reveal that
the notion of finitude thus shown to be logical is a rigorous version of our
intuitive conception of finitude, the intuitive conception being: A finite set is
one that can be counted. That is why Frege proves these theorems as he does,
why both proofs go through Theorem 314.
1.5 THE FINITE AND THE INFINITE

As well as proving the mentioned results about the finite and the infinite, Frege
also proves several results that relate these two notions. The first of these is:
P ∗= (0, Nx : F x) ∧ Nx : Gx = i → Nx : (F x ∨ Gx) = i (172)
which is the central result of I(b) and which says that the “union” of a count-
ably infinite concept and a finite concept is countably infinite. But the most
interesting results of this type are in the portions of Part II that are contained
in the second volume of Grundgesetze, which was published ten years after the
first.
i i
i i
i i

i i
Chapter Mu, the first in Volume II, is devoted to the proof of:
Nx : Gx = i ∧ ∀x(F x → Gx) →
P∗= (0, Nx : F x) ∨ Nx : F x = i (428)
What this says is that every sub-concept of a countably infinite concept is
either countably infinite or finite. Frege’s proof of this claim is long and com-
plex, and few of the details seem to be of much philosophical interest. But one
of the results he proves along the way, the central result of sub-chapter M(a),
is of significant interest:19
Func(Q) ∧ ¬∃y(Q∗ yy) ∧ ∃y(Q∗= ay ∧ F y) →
∃z[Q∗= az ∧ F z ∧ ¬∃x(Q∗= ax ∧ F x ∧ Q∗ xz)] (359)
The antecedent affirms that the Q-series is simple and that there is an F in
the Q-series beginning with a. The consequent says that, if so, then there is a
first F in the Q-series beginning with a, that is, an F that is preceded, in that
series, by no other F . So (359) is a generalization of the least number principle.
To derive the least number principle, we need to weaken the antecedent by
replacing ‘¬∃y(Q∗ yy)’ with ‘¬∃y(Q∗= ay ∧ Q∗ yy)’ and then replace ‘Q’ with
‘P’ and ‘a’ with ‘0’ to reach:
Func(P) ∧ ¬∃y(P∗= 0y ∧ P∗ yy) ∧ ∃y(P∗= 0y ∧ F y) →
∃z[P∗= 0z ∧ F z ∧ ¬∃x(P∗= 0x ∧ F x ∧ P∗ xz)]
The first two conjuncts are Theorems 71 and 145, so we have:

∃y(P∗= 0y ∧ F y) →
∗=
∃z[P 0z ∧ F z ∧ ¬∃x(P∗= 0x ∧ F x ∧ P∗ xz)]
which just is the least number principle, as Frege would have understood it.
Why does Frege prove these results? Theorem 428 is, to be sure, of signi-
ficant interest simply as a fact about cardinal numbers: It tells us that there is
no cardinal number between the finite numbers and Endlos. Frege seems to
have had a particular interest in results of this kind. When he says, in the first
paragraph of the Foreword, that “propositions about the cardinal numbers are
not yet present with the completeness initially planned” (Grundgesetze I, v),
the one proposition he mentions explicitly is:
P∗= (0, Nx : Gx) ∧ ∀x(F x → Gx) → P∗= (0, Nx : F x) (443)
which he proves in Chapter Nu, and which is obviously in much the same
spirit as (428). But there is something else that (428) tells us that would have
been extremely important to Frege.
19 Frege states the result using a new definition that encapsulates what is in the consequent. But
we will not need that definition here.
i i
i i
i i

i i
Part II of Grundgesetze is concerned exclusively with the cardinal numbers,

but there is also a Part III, which is concerned with the real numbers.20 In order
to develop the theory of the reals, Frege will need to show that the reals exist,
and to do that he will need to show that there are continuum-many objects
in the domain. HP will not yield continuum-many objects: There is a model
of HP in which there are only countably many objects in the domain. But
something stronger is true: We can actually interpret the theory HP + ‘the
only objects that exist are the natural numbers and Endlos’ in HP itself. The
interpretation is simply a relativization to: P∗= 0x ∨ x = i, and the key result
needed for the argument is (428): What we need to know is that, if we have
a concept that is true only of natural numbers and Endlos, then its cardinal
number is either a natural number or Endlos.
Was Frege aware of this consequence of Theorem 428? He nowhere men-
tions it, but it is an argument of a sort that would have been familiar to him:
Arguments of this type were common in geometry at the time, and Frege uses
them in his own non-foundational work (Tappenden, 2000). And, as already
said, this result would have been of great interest to Frege, since it implies
that the theory of the reals will need resources beyond HP: If HP + ‘the only
objects that exist are the natural numbers and Endlos’ is interpretable in HP,
then so is HP + ‘the universe is countable’. So if HP is consistent, so is HP +
‘the universe is countable’. But then it cannot follow from HP that there are
continuum-many objects.
1.6 FURTHER UNTO THE INFINITE
At first glance, Chapter Omicron is an odd hodgepodge of results.21 The first

result Frege proves, in O(a), is:
∀x(Gx → F x) ∧ ∀x(Γx → Φx) ∧

Nx : Gx = Nx : Γx ∧ Nx : (F x ∧ ¬Gx) = Nx : (Φx ∧ ¬Γx) →
Nx : F x = Nx : Φx (472)
In words: If every G is an F and every Γ is a Φ, and the number of Gs is the

number of Γs, and the number of F s that are not G is the number of Φs that
are not Γ, then the number of F s is the number of Φs. Or: If one can take
the same number of objects out of the F s and the Φs and end up with the
same number of objects in each case, then one started with the same number
of objects.
20 I believe there was also to be a Part IV, concerned with the complex numbers (Heck, 2012,
3, esp. n. 6).
21 Chapter Xi contains a proof that cardinal addition is unique, but not a proof that sums
always exist. There is something of value, I think, to be said about the matter (Heck, 2012, ch.
10), but I shall not pursue it here, as it is somewhat tangential to the story I am trying to tell.
i i
i i
i i

i i
The goal of O(b) is:
∀x(Gx → F x) ∧ P(Nx : Gx, Nx : Gx) → P(Nx : F x, Nx : F x) (476)
In words: If every G is F , and the number of Gs follows itself in the number-

series, then the number of F s follows itself in the number-series, too. And
there are two results proven in O(c):
∃z[P(Nx : F x, z)] (480)

∗=
Nx : Gx = i ∧ ∀x(Gx → F x) → ¬P (0, Nx : F x) (484)
The former says that every number, not just every natural number, has a suc-
cessor; the latter, that no concept that has a countably infinite sub-concept is
finite.
So, as said, Chapter O looks like an odd collection of results. What might
be their purpose?
Sundholm (2001, 61, n. 17) reports that, in the summer of 1889, Frege
taught a seminar on Dedekind’s Was Sind und Was Sollen die Zahlen? A few
years later, he writes in his review of Cantor’s Contributions to the Theory of the
Transfinite:
Mr. Dedekind gives as the characteristic mark of the infinite that it is similar to a proper
part of itself…, after which the finite is defined as the non-infinite, whereas Mr. Cantor
tries to do what I have done: first to define the finite, after which the infinite appears
as the non-finite. Either plan can be carried through correctly, and it can be proved
that the infinite systems of Mr. Dedekind are not finite in my sense. This proposition
is convertible [that is, it has a true converse]; but the proof of [the converse] is rather
difficult, and it is hardly executed with sufficient rigour in Mr. Dedekind’s paper.
(Frege, 1892, 271/180)
What we have in Chapter O, I suggest, are the somewhat meager fruits of
Frege’s investigation of the relationship between the two notions of infinity
he mentions here: The notion of infinity given by ¬P∗= (0, Nx : F x), which
is further characterized by the results discussed in Sections 1.3 and 1.4 above,
and the notion we now know as Dedekind infinity: A concept is Dedekind
infinite if it can be put into one–one correspondence with one of its proper
sub-concepts. I think, in particular, that Chapter O contains the results of
Frege’s attempt to give a properly rigorous proof of the converse he mentions:
that every infinite set is Dedekind infinite.
Dedekind proves in Was Sind? (1902) that a concept is Dedekind infinite
if, and only if, it has a countably infinite sub-concept. Frege need have had
no quarrel with Dedekind’s proof of this fact, which it is easy to formalize in
Frege’s system using (263). And given this fact, (484) is what Frege says, in
the passage just quoted, “can be proved”, namely, that “the infinite systems
of Mr. Dedekind are not finite in [Frege’s] sense”. But, of course, we do not
find in Chapter O a proof of the converse that Frege mentions in the review
of Cantor: that Frege infinite sets are Dedekind infinite. That is unsurprising,
i i
i i
i i

i i
since we now know that this result cannot be proven without the axiom of
choice—more precisely, the axiom of countable choice—and there is no axiom
of choice in Frege’s formal system. But this converse would have been of great
interest to Frege.
On the one hand, Frege has shown that a concept is of finite number if,
and only if, the objects falling under it can be ordered as a simple series that
ends (327, 348); on the other, that a concept is of countably infinite number
just in case the objects falling under it can be ordered as a simple series that
does not end (207, 263). But then it seems natural to wonder whether—and
perhaps even more natural to conjecture that—every concept is either finite or
has a countably infinite subconcept, that is, that every concept is either finite
or Dedekind infinite. One might reason informally thus. Suppose we set out
to build a simple series from the objects falling under F , successively choosing
distinct F s and adding them to the series so far constructed. If, at some point,
we exhaust the F s, then the F s will have been ordered as a simple series that
ends. And if we never exhaust the F s, won’t some of the F s then have been
ordered as a simple series that does not end?
The informal proof just given is hardly one with which Frege would have
been prepared to settle: It contains another dangerous mixture of reason and
intuition, one it was a large part of his purpose to disentangle. But, as I men-
tioned earlier, Frege was no more happy with Dedekind’s proof. When he re-
marks that Dedekind’s proof “is hardly executed with sufficient rigour”, this
suggests to me not just that he was dissatisfied with Dedekind’s prose—he
does not make the same complaint about Dedekind’s proof that every Dede-
kind infinite set is infinite—but that he had attempted to reproduce the proof
in the concept-script and had failed.
In formalizing the proof—in attempting to make it “gapless”—Frege would
quickly have discovered the gap he could not fill, an inference upon whose
validity Dedekind was tacitly relying. In the course of this investigation, Frege
would naturally have considered the problematic inference in its general form.
And when we do that, what we find is that Dedekind needs the following:
∀n[P∗= 0n → ∃G(Mx (Gx, n))] →

∃R∀n[P∗= 0n → Mx (Rnx, n)]
Here, ‘M ’ is a third-order variable indicating relations between concepts and

objects. And the displayed formula is an axiom of countable choice for con-
cepts.
The supposition that Frege discovered the axiom of countable choice raises,
in a very powerful way, a difficulty for those who would deny him the re-
sources to discuss its truth or its epistemological status, that is, to discuss the
question whether it is a law of logic.22 Frege set out to derive the basic laws of
22 Views in this vincinity have been expressed by van Heijenoort (1967), Dreben and van
Heijenoort (1986), Ricketts (1986a; 1986b), Weiner (1990), Kemp (1995), and Goldfarb (2001),
i i
i i
i i

i i
number within a certain formal system, second-order logic plus Basic Law V.
But Frege was not interested in formalization only for its mathematical bene-
fits. He hoped, and taught, that formalization would shed light upon philo-
sophical questions, too. Indeed, Frege is admirably clear about the relation
between formalization and epistemology:
I became aware of the need for a concept-script when I was looking for the funda-
mental principles or axioms upon which the whole of mathematics rests. Only after
this question is answered can it be hoped to trace successfully the springs of knowledge
on which this science thrives. (Frege, 1969, 362)
The epistemological status of arithmetic is not decided by formalization alone;
it is only after its axioms have been isolated that the question of arithmetic’s
epistemological status becomes tractable. But then, if logicism is not estab-
lished by a “reduction” of arithmetic to the formal system of Grundgesetze, or
any other formal system, it simply must be an intelligible question whether
the axioms of that system are logical laws. What other question could remain
at that point?
This argument concentrates upon Frege’s attitude towards the thoughts
he accepts as logical axioms. I myself regard it as conclusive, but it seems not
to have carried conviction with quite everyone.23 But a more powerful argu-
ment emerges if we concentrate upon Frege’s attitude toward a different sort
of thought, a good example of which is the axiom of countable choice. In so
far as one sets out to derive all the truths of some “branch of learning” from
a fixed set of axioms, one is immediately confronted with the possibility that
one’s axioms might not be “complete”: There may be propositions of whose
truth we had previously been convinced by informal argument that cannot be
proven within the system as it stands (Frege, 1969, 362). More interestingly,
there may be a certain kind of inference which, though commonly made, can-
not be replicated within the system. And so Frege writes:
By [resolving inferences into their simple components] we shall arrive at just a few
modes of inference, with which we must then attempt to make do at all times. And if
at some point this attempt fails, then we shall have to ask whether we have hit upon a
truth issuing from a non-logical source of cognition, whether a new mode of inference
has to be acknowledged, or whether perhaps the intended step ought not to be taken
at all. (Frege, 1969, 363)
It is of course possible that Frege is here speaking completely hypothetically.
But I suspect he is instead speaking from experience. He may have in mind
something like the version of countable choice mentioned above. As men-
tioned on page 13 above, Frege all but says in §114 that there are arithmetical
truths that are unprovable within his system as it stands, and the very claim
among others. I am in general agreement, myself, with the criticisms of such interpretations made
by Stanley (1996), Tappenden (1997), and Sullivan (2005).
23 As Sullivan (2005, §3.2) makes clear, objections to the argument tend to conflate questions
about logic with questions about particular formulations of logic.
i i
i i
i i

i i
we have been discussing, that every infinite set is Dedekind infinite, lies just
below the surface of that discussion (Heck, 2012, §6.7).
The interest of Frege’s reflections does not depend upon the correctness
of this speculation, however, which merely serves to make vivid the problem
that concerns him: a problem about the epistemological status of principles
like the axiom of choice. Frege is saying, quite reasonably, that the question
whether the axiom of choice is true at all, and if so, whether it is a logical law,
is not only intelligible but important.
If our attempt to formalize Dedekind’s proof in the concept-script has
failed, then our first task must be to identify the apparently unprovable pro-
position whose truth that proof tacitly assumes. Once we have done this, we
will, according to Frege, have three options. First, we may reject the propos-
ition in question as not being true, and so reject the proof. Second, we may
accept the proposition’s truth, but regard it as a truth peculiar to some special
science, say, to a part of mathematics not reducible to logic. Third, we may ac-
cept the proposition as a truth of logic. The presence of the first option shows
that Frege did not regard himself merely as formalizing accepted mathemat-
ical practice: His project had, in his own view, potentially revisionary con-
sequences. The presence of the other two options show that Frege regarded it
as an intelligible question whether (say) the axiom of countable choice, assum-
ing it is a truth, is a truth of logic. I thus regard it as demonstrable that Frege
believed the question whether a given truth is a truth of logic to be intelligible.
1.7 CONCLUSION
George Boolos once wrote:
Perhaps the saddest effect of Russell’s paradox was to obscure from Frege and us the
value of Frege’s most important work. Frege stands to us as Kant stood to Frege’s con-
temporaries. The Basic Laws of Arithmetic was his magnum opus. Are you sure there’s
nothing of interest in those parts of the Basic Laws that aren’t in prose?
(Boolos, 1998a, 201)
I heard Boolos read those words in 1987, when he delivered ‘The Consistency
of Frege’s Foundations of Arithmetic’ in Oxford. As I recounted at the begin-
ning of this chapter, Boolos would ask me essentially the same question four
years later, and that time I was actually listening. It should be clear from the
foregoing what my answer was to be.
REFERENCES
Boolos, George (1998a). ‘The Consistency of Frege’s Foundations of Arith-
metic’, in Boolos (1998c), pages 183–202.
Boolos, George (1998b). ‘On the Proof of Frege’s Theorem’, in Boolos (1998c),
pages 275–91.
i i
i i
i i

i i
Boolos, George (1998c). Logic, Logic, and Logic, ed. Richard Jeffrey. Cam-
bridge, Mass.: Harvard University Press.
Boolos, George and Richard Kimberly Heck (2011). ‘Die Grundlagen der
Arithmetik §§82–83’, in Frege’s Theorem. Oxford: Clarendon Press, 69–87.
(Orig. publ. under the name “Richard G. Heck, Jr”.)
Burgess, John P. (1984). ‘Review of Frege’s Conception of Numbers as Objects’,
Philosophical Review 93:638–40.
Dedekind, Richard (1902). ‘The Nature and Meaning of Numbers’, trans.
Wooster W. Beman, in Essays on the Theory of Numbers. Chicago: Open
Court, pages 31–115.
Dreben, Burton and Jean van Heijenoort (1986). ‘Introductory Note to
1929, 1930, and 1930a’, in Solomon Feferman, John W. Dawson, Jr,
Stephen Kleene, Gregory H. Moore, Robert M. Solovay, and Jean van
Heijenoort (eds.), Kurt Gödel: Collected Works, volume 1, third edition.
Oxford: Oxford University Press, pages 44–59.
Dummett, Michael (1991). Frege: Philosophy of Mathematics. Cambridge,
Mass.: Harvard University Press.
Frege, Gottlob (1879). Begriffsschrift: Eine der arithmetischen nachgebildete
Formelsprache des reinen Denkens. Halle a. d. Saale: L. Nebert. Trans. Stefan
Bauer-Mengelberg as ‘Begriffsschrift: A Formula Language Modeled upon
that of Arithmetic, for Pure Thought’, in Jean van Heijenoort (ed.), From
Frege to Gödel: A Sourcebook in Mathematical Logic 1879–1931. Cam-
bridge Mass.: Harvard University Press, pages 5–82.
Frege, Gottlob (1884). Die Grundlagen der Arithmetik. Eine logisch mathema-
tische Untersuchung über den Begriff der Zahl. Breslau: Wilhelm Koebner.
Trans. The Foundations of Arithmetic, second, revised edition, trans. J. L.
Austin. Evanston, Ill.: Northwestern University Press, 1953.
Frege, Gottlob (1892). ‘Georg Cantor, Zur Lehre vom Transfiniten’, Zeit-
schrift für Philosophie und philosophische Kritik 100:269–72. Trans. by Hans
Kaal as ‘Review of Georg Cantor, Zur Lehre vom Transfiniten’, in Frege
(1984), pages 178–81.
Frege, Gottlob (1893/1903). Grundgesetze der Arithmetik. Band I und II.
Jena: Hermann Pohle. Trans. Frege (2013).
Frege, Gottlob (1969). ‘On Mr. Peano’s Conceptual Notation and My Own’,
trans. V. H. Dudman, Australasian Journal of Philosophy 47:1–14. Reprin-
ted in Frege (1984), 234–48.
Frege, Gottlob (1980). Philosophical and Mathematical Correspondence. Trans.
Hans Kaal; Ed. Gottfried Gabriel, Hans Hermes, Friedrich Kambartel,
Christian Thiel, Albert Veraart, and Brian McGuinness. Chicago: Univer-
sity of Chicago Press.
Frege, Gottlob (1984). Collected Papers on Mathematics, Logic, and Philosophy.
Ed. Brian McGuinness. Oxford: Basil Blackwell.
Frege, Gottlob (2013). Basic Laws of Arithmetic. Derived Using Concept-Script.
i i
i i
i i

i i
Volumes I and II. Ed. and trans. Philip A. Ebert and Marcus Rossberg.
Oxford: Oxford University Press.
Geach, Peter T. (1955). ‘Class and Concept’, Philosophical Review 64:561–
70.
Goldfarb, Warren (2001). ‘Frege’s Conception of Logic’, in Juliet Floyd and
Sanford Shieh (eds.), Future Pasts: The Analytic Tradition in Twentieth-
Century Philosophy. New York: Oxford University Press, pages 25–41.
Green, J. J., Marcus Rossberg, and Philip A. Ebert (2015). ‘The Convenience
of the Typesetter: Notation and Typography in Frege’s Grundgesetze der
Arithmetik’, Bulletin of Symbolic Logic 21:15–30.
Hazen, Allen (1985). ‘Review of Crispin Wright, Frege’s Conception of Num-
bers as Objects’, Australasian Journal of Philosophy 63:251–4.
Heck, Richard Kimberly (2011). ‘Ramified Frege Arithmetic’, Journal of
Philosophical Logic 40:715–35. (Orig. publ. under the name “Richard G.
Heck, Jr”.)
Heck, Richard Kimberly (2012). Reading Frege’s Grundgesetze. Oxford: Clar-
endon Press. (Orig. publ. under the name “Richard G. Heck, Jr”.)
Heck, Richard Kimberly (2019). ‘Formal Arithmetic Before Grundgesetze’,
in this volume.
Kemp, Gary (1995). ‘Truth in Frege’s “Law of Truth”’, Synthese 105:31–51.
May, Robert C. and Kai F. Wehmeier (2019). ‘The Proof of Hume’s Prin-
ciple’, in this volume.
Parsons, Charles (1965). ‘Frege’s Theory of Number’, in Max Black (ed.),
Philosophy in America. Ithaca: Cornell University Press, pages 180–203.
Reprinted in William Demopoulos (ed.), Frege’s Philosophy of Mathematics.
Cambridge Mass.: Harvard University Press, 1995, pages 182–210.
Ricketts, Thomas G. (1986a). ‘Generality, Sense, and Meaning in Frege’, Pa-
cific Philosophical Quarterly 67:172–95.
Ricketts, Thomas G. (1986b). ‘Objectivity and Objecthood: Frege’s Meta-
physics of Judgement’, in Leila Haaparanta and Jaakko Hintikka (eds.),
Frege Synthesized: Essays on the Philosophical and Foundational Work of
Gottlob Frege. Dordrecht: Reidel, pages 65–95.
Stanley, Jason (1996). ‘Truth and Metatheory in Frege’, Pacific Philosophical
Quarterly 77:45–70.
Sullivan, Peter M. (2005). ‘Metaperspectives and Internalism in Frege’, in
Michael Beaney and Erich H. Reck (eds.), Gottlob Frege: Critical Assess-
ments of Leading Philosophers, volume II. London: Routledge, pages 85–
105.
Sundholm, Göran (2001). ‘Frege, August Bebel, and the Return of Alsace-
Lorraine: The Dating of the Distinction between Sinn and Bedeutung’,
History and Philosophy of Logic 22:57–73.
Tappenden, Jamie (1997). ‘Metatheory and Mathematical Practice in Frege’,
Philosophical Topics 25:213–64.
i i
i i
i i

i i
Tappenden, Jamie (2000). ‘Frege on Axioms, Indirect Proof, and Independ-

ence Arguments in Geometry: Did Frege Reject Independence Argu-
ments?’, Notre Dame Journal of Formal Logic 41:271–315.
van Heijenoort, Jean (1967). ‘Logic as Calculus and Logic as Language’, Syn-
these 17:324–30.
Weiner, Joan (1990). Frege in Perspective. Ithaca: Cornell University Press.
Wright, Crispin (1983). Frege’s Conception of Numbers as Objects. Aberdeen:
Aberdeen University Press.
i i
i i
i i

i i
2
Axioms in Frege
Patricia A. Blanchette
As is generally well appreciated, Frege’s conception of axioms is an old-fash-

ioned one, a conception according to which each axiom is a determinate non-
linguistic proposition, one with a fixed subject-matter, and with respect to
which the notion of a “model” or an “interpretation” makes no sense. As
contrasted with the fruitful modern conception of mathematical axioms as
collectively providing implicit definitions of structure-types, a conception on
which the range of models of a set of axioms is of the essence of those axioms’
significance, Frege’s view is a dinosaur.
It is the purpose of this essay to investigate some of the philosophically-
important aspects of that dinosaur, in order to shed light on Frege’s under-
standing of the foundational role of axioms, and on some of the ways in which
our current conception of such axiomatic virtues as independence and categor-
icity have (and in some cases have not) been informed by a move away from
Frege’s understanding of the foundational role of axioms.
2.1 CONSISTENCY AND INDEPENDENCE OF AXIOMS
2.1.1 Conceptual Analysis and Logical Entailment

We begin with a brief excursion into Frege’s views about the connection be-
tween conceptual analysis and proof in arithmetic. Early in Grundlagen, Frege
explains the project of that book as follows:
… the fundamental propositions of arithmetic should be proved, if in any way possible,
with the utmost rigor; for only if every gap in the chain of deductions is eliminated
with the greatest care can we say with certainty upon what primitive truths the proof
depends …
Overlapping parts of the material in this chapter have been presented to audiences at the Vrije
Universiteit Amsterdam, at the Logicism Today conference in Besse-en-Chandesse, at the Univer-
sity of Bucharest, at UC Irvine, and at NYU. Many thanks to the organizers of these conferences,
and to audience members for helpful feedback. Thanks also to Philip Ebert, Marcus Rossberg,
and two anonymous referees for helpful remarks on an earlier draft.
i i
i i
i i

i i
32 Patricia A. Blanchette
If we now try to meet this demand, we very soon come to propositions which
cannot be proved so long as we do not succeed in analyzing concepts which occur in them
into simpler concepts or in reducing them to something of greater generality. Now here it is
above all Number which has to be either defined or recognized as indefinable. This is
the point which the present work is meant to settle. On the outcome of this task will
depend the decision as to the nature of the laws of arithmetic.
(Grundlagen §4, emphasis added)
The basic idea here is a familiar one: truths expressed using terms such as
‘prime number’ or ‘continuous function’ can often only be proved once the
relevant notions have been broken down into complexes of simpler ones, so
that the truths themselves are expressed by sentences whose terms stand for
the relative simples, and whose syntactic complexity is greater than that of the
original sentences. Frege’s idea is that his analysis and clarification of such fun-
damental concepts as that of cardinal number, en route to the proof of claims
about cardinal numbers, is of a piece with standard instances of conceptual
clarification in the history of mathematics:
[I]n mathematics a mere moral conviction, supported by a mass of successful applica-
tions, is not good enough. Proof is now demanded of many things that formerly passed
as self-evident. Again and again the limits to the validity of a proposition have been
in this way established for the first time. The concepts of function, of continuity, of
limit and of infinity have been shown to stand in need of sharper definition. Negative
and irrational numbers, which had long since been admitted into science, have had to
submit to a closer scrutiny of their credentials.
In all directions these same ideals can be seen at work—rigor of proof, precise
delimitation of extent of validity, and as a means to this, sharp definition of concepts.
§2. Proceeding along these lines, we are bound eventually to come to the concept
of Number and to the simplest propositions holding of positive whole numbers, which
form the foundation of the whole of mathematics. (Grundlagen §§1–2)
The idea of conceptual analysis as an essential preliminary to proof is main-

tained throughout Frege’s work. In the 1914 Logic in Mathematics manuscript,
we find a similar sentiment:
In the development of science it can indeed happen that one has used a word, a sign,
an expression, over a long period under the impression that its sense is simple until one
succeeds in analysing it into simpler logical constituents. By means of such an analysis,
we may hope to reduce the number of axioms; for it may not be possible to prove a
truth containing a complex constituent so long as that constituent remains unanalysed;
but it may be possible, given an analysis, to prove it from truths in which the elements
of the analysis occur. (Frege, 1914 (1983, 226; 1979, 209))
The general picture painted here, and the pattern Frege follows in Grundlagen
and Grundgesetze, is as follows:
(a) We begin with a thought expressed in a relatively-unanalyzed way, e.g.,
via a sentence of ordinary arithmetic.
i i
i i
Another random document with
no related content on Scribd:
leaf is longer but thinner than that of the Congou sorts; folded rather
than curled or twisted, but possessing somewhat similar drinking
qualities. They are classed in trade as Lapsing, Tong-quam, Padrae,
Pekoe, Oolong, and Canton Souchongs.
Lapsing—Prepared in the district of Foochow, is also known to trade

as “Foochow-Souchong,” is a large, handsome, crapy leaf, finely
made and lightly fired, possessing a rich, wine-colored liquor with
fragrant flavor, entirely peculiar to itself, described as “tarry flavor,”
which when not too pronounced adds rather than detracts from its
worth. The product of the later pickings are of less strength and
flavor, but are still very smooth and pleasant in liquor and flavor, and
generally shipped to the Russian market, where they are held in high
esteem for their intrinsic qualities.
Oolong-Souchong—Is another variety of the foregoing, prepared

from the leaves of a plant that cannot well be made into either sort,
the greatest care being taken in its manipulation. It is stylish in leaf,
closely approximating to Foochow Oolong in the dried state, very
clear, rich and translucent in the infusion, but though light in weight
and color is yet very deceptive, being full of snap and sparkle,
fragrant and aromatic.
Tong-quam—Is a long, flat, black-leaf Souchong tea carefully

folded, but little understood by the general trade, owing to the liquor
possessing nearly the same flavor and pungency as that of a Red-
leaf Congou, usually more round and fuller, the dry leaf being slightly
bolder and blacker in appearance.
Padrae-Souchong—Is a jet-black leaf, small and “crapy” in texture,

usually prepared from the youngest and tenderest leaves of the
Congou order, and which it closely resembles in general character
and flavor. The dry leaf is, however, much smaller, flatter and darker,
but greatly excelling them in the delicacy and fragrance of the
infusion.
Pekoe-Souchong—Is prepared from the leaves that have
developed too much to be converted into the former kind, which is
small in size. The dry leaf is medium-sized, very black and
moderately “tipped” at the ends with a whitish-downy substance
termed “pekoe.” In liquor they are strong, dark, pungent and fragrant
in flavor and aroma.
Canton Souchongs—Are prepared from old and exhausted leaves

collected in a careless manner, exposed in the sun to dry, and
packed in baskets until they reach that city, where they are refired,
colored and scented in order to disguise their bitter, rank and
astringent properties.
SCENTED TEAS
form a special class of the Chinese product comprising Capers,

Pekoes and Pouchong teas, being known to trade as Foochows,
Cantons and Macaos.
Caper—Known to the Chinese as He-choo-cha, “Black pearl,” or

Gunpowder, from its small, round or spherical appearance,
resembling capers. It is prepared from the largest but most succulent
leaves of the first pickings, and cured by a series of brisk firings and
rollings, after which it is placed in moulds, in order to make it retain
its globular shape. The dried leaf is small, round and “shotty” in
appearance, reddish-black in color, glossy and highly scented. The
infusion is wine-colored, piquant and aromatic, possessing what is
technically termed a rich “bouquet,” the infused leaf, when uncurled,
being very symmetrical in form and dark-brown in color.
Pekoes—From the Chinese Pai-ho, or Pak-ho, signifying “white

down,” is applied to a variety of tea having a whitish downy or
“silvery” tip at the end of the leaves. It is usually prepared from the
youngest and tenderest leaf-buds first expanding, and was at one
time claimed to be composed of the flower or blossom of the tea
plant, hence its French name, “fleur de thé,” an error long since
corrected, as the tea blossom possesses none of the properties of
the leaf, though frequently used for scenting purposes.
Orange-Pekoe—Recognized by its long, flat, even and artistically

folded leaf, jet-black color, and yellowish downy tips at the ends. It is
highly scented, yielding a rich wine-colored liquor, piquant, pungent
and aromatic in the cup, the infused leaf being small, bright and
closely resembling that of choicest Oolong variety.
Flowery-Pekoe—Is a smaller but more evenly folded leaf, greenish-

black or olive-colored, with ends ornamented by whitish, “velvety”
tips, being also very highly scented. The infusion is lighter in color
and body but piquant and aromatic in flavor, the infused leaf small,
dark and perfectly formed.
Hung-muey—Is still another variety of Pekoe rarely exported,

having a plain black leaf lightly tipped and lightly scented, and
yielding an infusion dark in color, thin in body, but very fragrant and
aromatic in flavor.
Pouchong—Derives its trade name from Paou-cheong, meaning

“wrapped sort.” The leaf is rough and bold in style, dull-black in color
and peculiar in scent. The latter being imparted by the admixture of
the seeds of the Lan-hoa, or Chulan flower, the finer grades of which
are deep red, rich and pleasing, but the lower ones are often
abominable.
Pouchong-Pekoe—Is usually prepared from the undeveloped

leaves or just expanding buds of the tea plant, and is a small, glossy-
black leaf with yellowish-golden tips, yielding an intensely rich liquor
very piquant and highly aromatic in flavor.
Padrae-Pouchong—Is a medium-sized leaf, exceedingly black in

color and well folded. The liquor is dark, full, round and aromatic in
flavor, but light and thin in body.
Canton Scented Teas—Known to trade as Congee—“Lic” or “made

teas,” to a large extent being purchased in the natural state,
converted into Capers and Pekoes at will, and doctored or scented
up to a certain standard by contract. They are much higher scented
than Foochows, but lacking in the properties of true tea, less
pungent in liquor and devoid of character or flavor.
Macao Scented Teas—Known also as “New district,” are closely

allied to Cantons in make, appearance and character of scent. The
dry leaf is somewhat larger and darker in color, the flavor being dull
and peculiar in the infusion.
The fragrance of Scented teas is not, as is generally supposed,
natural to them, but imparted by the admixture of the flowers,
blossoms, leaves, or oils extracted from the seeds or roots of other
plants, such as the Orange, Jessamine, Chlorantus, Gardenia, and
Oleo-fragrans. The leaves and blossoms of the Iris, Curcunia, and oil
of Bixa orelana being also extensively used. In some districts the
scenting material is added to the tea during the firing process, and
afterwards separated by sifting. It is, however, more generally
introduced into the tea after it is prepared and ready for packing; one
pound of leaves or blossoms being the usual proportion to each
hundred pounds of tea. They are spread over the top of the tea in
the chest and allowed to remain for at least a day, or until it becomes
strongly impregnated by absorbing their moisture, and then
removed, the duration depending on the character of the scenting
employed, the scent increasing after the tea is packed for export. But
though scenting in general is supposed to be confined to the choicer
grades of tea it is as often applied to the inferior sorts, with the object
of disguising or concealing their defective or damaged condition, and
imparting a pleasant odor, a much larger quantity being used in the
latter. The scenting greatly modifies and improves the flavor,
however, without adding any pernicious or deleterious substance to
the tea.
Consumers not accustomed to using these varieties erroneously
imagine, from the dark color of the leaf and liquor, that they are much
stronger and more exciting than that of the Oolong or Green tea
sorts. While the contrary is the case, it requiring one-third more leaf
of corresponding quality to yield an infusion of equal strength than of
Oolong or Green tea sorts. The “smoky” and “tarry” flavors
possessed by many of them, and for which this variety is so
remarkable, is due in a great measure to the use of ill-made charcoal
in firing and the use of soft woods containing tar or pitch, such as fir
and pine, in its preparation. The worst feature about which is that this
“smokiness” and “tarriness” does not develop until long after the teas
have left China, and are offered for sale. It is also a noticeable fact
that certain waters serve to bring out these peculiarities more
prominently than others, American waters in particular.
OTHER CHINESE VARIETIES.
Besides these numerous ordinary teas of commerce, there are

several other varieties cultivated in China, but principally for home
consumption and rarely if ever exported, among which may be
mentioned:—
Suen-cha—Or “Sweet tea,” made from the leaves of a slender shrub

growing in the western province of Sze-chuan, and peculiar only to
that section. The leaf is large, thick and odorless in the green or
natural state, but when cured exhales a rare and peculiar odor, and
possesses a sweet, liquorice-like taste in the infusion, not altogether
pleasant.
Peh-Yuen-cha—Or “White cloud tea,” prepared from another rare

species of the tea shrub found near the summit of Mount Ombei in
the same province and most dissimilar in character and flavor from
that of the regular teas of commerce. It yields an aromatic infusion,
peculiar but palatable, and is chiefly used by pilgrims and travelers in
that country.
Mandarin Tea—Is still another rare variety, seldom if ever exported,

its use being confined to the Mandarins and aristocracy of China.
The leaf is exceeding small, dark, crisp and tender, lightly fired and
highly scented, commanding as high as fifteen dollars per pound in
the home market.
Brick Tea—Is composed of the old leaves, stems, siftings and
sweepings of the Chinese tea hongs, ground fine, moistened and
compressed into shapes somewhat larger than regular building
bricks. It has nothing to recommend it as a tea, being sold chiefly to
the Mongols, Tartars and other tribes of Central Asia, among whom it
also serves as a currency.
Tablet Tea—Is a “new make” of tea recently introduced in China,

appearing for the first time in the trade returns last year. It is
prepared by machinery from the best quality of tea-dust, formed by
pressure alone into small cakes in the form of tablets perfectly hard
and solid, resembling chocolate in make and appearance. It is not,
like “brick tea,” moistened by steam before being compressed, and
the flavor is not in any way impaired by the process of manufacture.
One of the chief advantages claimed for this form of tea is that, being
subjected to heavy hydraulic pressure, all the cells are broken and
the properties of the tea are more easily and completely extracted by
the boiling water, thus effecting a considerable saving in the quantity
required for a given amount of the beverage. Its principal market is
Russia, which took from China last year over 500,000 pounds in the
form of tablets.
Medicine Tea—Is prepared from the coarse leaves and stems of the
ordinary tea plant, ground and mixed with medicinal herbs, packed in
bundles and used for medicinal purposes among Asiatic tribes.
Log-tea—Is also prepared from the ordinary teas of commerce. It is

a very inferior grade, prepared from the stalks, packed in the shape
of logs, weighing from 8 to 10 pounds, and wrapped in the leaves of
the bambusa, and packed in this manner from motives of economy
and freight.
The total production of tea in China is unknown, and can at best be
only roughly estimated, and while we have no certain means of
ascertaining the quantity consumed in that country itself, fair
conclusion may be drawn from the data at hand. Taking the
population at 400 millions and considering that the use of tea is
universal among its inhabitants, an average of five pounds per capita
would not be an overestimate, making a total of two billion pounds
alone for home consumption. Again averaging the product at 100
pounds of cured tea per acre and the total area under tea cultivation
at 20 million acres, if, therefore, we admit the home consumption of
tea in China to be two billion pounds, we cannot but be surprised at
the relatively small quantity which is exported from that country.
According to the latest statistics, we find that the total exports to all
countries from China does not exceed 200 million pounds, which is
less than one-tenth of the total production of that country.
JAPAN TEAS.
Tea is grown for commercial purposes all over the Japanese islands,
from Kiusiu in the south to Niphon in the north, but both in quantity
and quality of their product the central provinces of Hondo are the
finest, particularly that produced in the districts on the coast
provinces of the interior sea. The tea soil of Japan is described as
slate atmospherically dissolved with gypsum and phosphoric acid,
produced by manuring. The system of cultivation and methods of
preparation do not differ materially from those of the Chinese, the
first picking, which is the best, occurring about the beginning of May,
the second a month later, the third is often, however, omitted
altogether, in order not to injure the plants. In Japan the raw leaves
are generally sold to the exporters, by whom they are prepared and
converted into the several descriptions known to commerce.
When a sufficient quantity has been accumulated they are carried to
the hong or “drying house” and first placed in large bamboo baskets,
in which they are subjected to a steam bath for about a minute, after
which process they are spread out in the open air to cool and dry
thoroughly, previous to being fired and curled. Only about five
pounds of the leaves are put in the pans at a time for manipulation,
the process being identical with that of China, with the exception that
they are finally dried in bamboo baskets suspended over the
furnaces by cords from the ceiling for about fifteen minutes. During
this time they are gently agitated by the hands of the operators in
order to diffuse the heat and more thoroughly dry them. They are
then removed by a dextrous motion with fan-like scoops and tossed
in the air to free them from dust and stems, and afterwards picked
over by women and children before packing in the lead-lined chests
for export.
In color, flavor and character, Japan teas are totally distinct from any
and all other varieties, the finer grades being exceedingly delicate,
rich and peculiar to themselves. They yield a light-colored liquor,
very fragrant in flavor, but apt to deceive the casual drinker, as after
continued use they are found to possess greater strength and
pungency than most China teas, their effect on the nervous system
being very soon perceptible. They are classed commercially as
Yama-shiro, Uji, Kioto, Yedo, Eisyie, Suringar, Hatchoji, Nagahama,
Nagasaki, Tosia and Bancha, grading in value in the order named,
and converted into Pan-fired, Sun-dried, Basket-fired, Nibs and
Siftings, with occasionally small lots of Pekoe, Congou, Oolong,
Imperial, Gunpowder and Young Hyson makes.
Pan-fired—The finer grades have a long, well-curled, natural green

leaf, presenting an unbroken appearance, sinking immediately to the
bottom of the cup on infusion, uncurling rapidly and showing more or
less perfect leaves in the infused state. It yields a clear, bright liquor,
which remains unchanged in color and flavor until cold. The flavor is
delicate and fragrant in odor somewhat like that of new-mown hay.
The medium grades are correspondingly rougher in make, darker in
liquor and duller in flavor, while the commoner ones are coarse and
unsightly in style, varying from a greenish to a mottled blue in color,
and possessing a “brassy” or metallic taste, due to the cosmetic or
artificial coloring-matter used in their preparation.
Sun-dried—As the name implies, are steamed and dried in the sun
before firing, in order to fix their color permanently. The leaf is olive-
green, well fired, compactly curled and “toasty” in the cup, owing to
their thorough fermentation before firing, and although not as well
appreciated as the Pan-fired, are much superior in drinking
properties, their extra fermentation destroying the “grassy” flavor so
characteristic of many Japans. The lower grades range from a
yellowish to a dull-green, indifferently rolled and often “fishy” in
flavor, said to be contracted from the use of fish manure in the coast
districts.
Basket-fired—So named from being cured by the “basket process,”

and in contradistinction to those fired in pans. The finer grades are
long, dark and exceedingly well twisted or curled, entirely free from
stems, dust and other extraneous matter, clear and bright in liquor,
and mellow or “mealy” in flavor, the latter quality making them a very
valuable sort for blending purposes. The commoner grades are
rough, and uncouth in style, brownish-black in color, thick and heavy
in liquor, but lacking in “grip” and flavor.
Kumo—Or “Spider-leg” Japan, is in reality only a finer grade of

basket-fired; long, narrow, black, and “wirey” in leaf, and elastic in
texture. It is of the Pekoe order in make, but still retaining all the
properties of liquor and flavor of a Japan tea pure and simple.
Nibs—Are composed of the refuse of the foregoing kinds, bearing

the same relation to Japans that Twankays do to Green teas, many
of them drawing and drinking exceedingly well, according to the
grade separated from.
Up to 1856 China tea was the only tea used in the United States, but
during that year a small quantity of Japan teas, consisting of about
50 half-chests, was first received in this country. Being found pure
and free from coloring-matter, it soon became very popular with
consumers, a large number of whom had been prejudiced against
China green teas at the time, under the impression that they were
more or less artificially colored. The demand steadily increased, 400
half-chests were imported the following year, which was increased to
1,100 chests in 1859. About 1860 the Japanese changed their mode
of curing, adopting that of the Chinese as applied to Green teas, with
the result of altering the color from a dark to a light green, and of
imparting a high “toasty” or malty flavor, in lieu of the uncooked or
“grassy” taste which characterized the first importations, since which
period and change they have continued to grow in popular favor. But
the supply of Japan teas being at one time greatly in excess of the
demand and the price declining in many instances below the cost of
production, in connection with the fact that the teas as originally
prepared were used only in the American market, induced the
Japanese to convert their surplus leaf into other varieties, such as
Pekoes, Congous, Oolongs, Imperials, Gunpowders, and Young
Hysons, in imitation of the Chinese “makes,” with the futile
expectation of popularizing them in England and other countries,
where, heretofore, only very small quantities were consumed. With
this intention Chinese skilled labor was imported into the tea districts
to aid them in the experiment of preparing these makes of teas. The
result proved most unsatisfactory as was anticipated at the time by
experts and others interested in the project, only very small
quantities of the respective kinds being produced occasionally. It is
predicted, however, that all the different descriptions now received
from other countries will be eventually prepared in Japan, in
evidence of which a tea rivalling the finest Formosa in general
character is now produced in the Hondo district from a variety of the
Japan plant.
Japan Pekoe—Is a long, dark-green, flat leaf tea, usually “tipped,”

but as often not, approaching to that of the India variety in style and
appearance. But while looking remarkably well in the hand and up to
standard in drink, being smooth in liquor and “malty” in flavor, as a
general rule it is through overfiring lacking in the scent and aroma of
the China and even India prototype.
Japan Congou—Approximates in many of its leading features to

that of the India species, the cured leaf possessing similar properties
to many of the finer grades of the latter. The infusion is brighter in
color but thinner in body, and more acidulous in flavor, and the
reverse of palatable, owing to its imperfect fermentation and high or
overfiring.
Japan Oolongs—Although cured in identically the same manner as

the China variety, resemble them only in general contour. The leaf is
darker in color but finer in make, approaching more to the Souchong
order. The infusion is also darker in draw, but very “toasty,” that is,
“burnt” in flavor, owing to too high firing, retaining all the original
peculiarities of a regular Japan tea.
Japan Imperials, Gunpowders and Young Hysons—Differ only

from the ordinary Japan teas in form, make and color. Being
prepared from the same leaf, they naturally possess the same
general characteristics and cup qualities; the demand not justifying,
they are not produced in any appreciable quantities.
The production of tea in Japan is constantly increasing and its quality
improving, a wider area being devoted to its cultivation each year,
largely superseding sericulture in many districts. The total area now
under cultivation amounts to nearly 42,000 Cho, or about 100,000
acres. The total annual product is estimated at 100,000,000 pounds,
a gain of over 30,000,000 as compared with 1890, of which
40,000,000 pounds, or 44 per cent. of the total production was
consumed in the United States during the fiscal year of 1891. The
American taste for Japan teas continues to grow in proportion,
particularly in the Northwestern and Pacific States, their consumption
in this country nearly doubling that of Oolongs and Congous
combined, and trebling that of Green teas of all makes. This too,
notwithstanding the fact that only a very small proportion of really
choice Japan teas are ever exported, rarely exceeding one per cent.
of the entire crop, being principally retained for home consumption.
INDIA TEAS.
One of the most remarkable circumstances in connection with the
development of the Tea trade is the rapid increase in the production
and consumption of India Teas. Almost unknown to commerce thirty
years ago, they are fast becoming an important factor in the
business, particularly in the English and colonial markets, India being
already of such importance to them as a source of tea supply that it
is only a question of a very short time when the tea consumers of
these countries will no longer regard China as a tea-growing country
indispensable to them.
As far as can be ascertained, the first announcement of the
discovery of the tea-plant in India was made in 1833, but owing to
imperfect specimens being sent to botanists for inspection, it was not
at the time considered a true species. It was fully demonstrated,
however, in 1835, when a plant with perfect leaves, flowers and
seeds was obtained which proved on analysis to be a species of the
genus tea allied to, but not identical with that of China; Burmese and
Chinese experts, to whom the specimens were submitted,
concurring in the statement. The report being favorable, an
experimental plantation was immediately established under
government auspices with results not known. The first plantation for
its cultivation on a commercial scale was formed in Lukhimpore in
1836, from which the first samples were received in 1839, and the
first sales made in 1840. But, owing to the unfavorable reports given
on the first samples of the tea prepared from the India leaf, it was
rejected by the London brokers. The propriety of introducing the
China species was next suggested by some planters, and tons of
seed were at once imported from that country, large estates being
formed from the plants raised from it. Many of the plantations were
finally composed of hybrids or crosses between the China and India
species, which is now claimed to have been an error, as the nearer
each variety approaches to the indigenous the higher its excellence.
The tea-producing districts of India are widely scattered, the largest
—Assam—being situated in the extreme northeast of the country
bordering on the Burmese Empire, the others being located on the
northwestern boundary of Nepaul and the Punjaub, while Central
India appears to be entirely devoid of tea gardens up to the present.
There are numerous plantations, however, scattered over the
southwestern provinces of the peninsula, most notably in Wynaad,
Neilgherry and Travancore. In India, tea is grown on extensive
estates, often comprising thousands of acres, situated principally in
the alluvial valleys of large rivers, or formed on land reclaimed from
primeval jungle, possessing all the richness of virgin soil and
cultivated either by the individual owners or the agents of companies
commanding considerable capital. Every detail of cultivation and
preparation is conducted under close and careful European
supervisors. The plants are raised from seed sown in nurseries until
they are about 18 inches high, when they are transplanted to the
rows in the gardens in which they are to grow, the closest attention
being paid to weeding and irrigating. The young trees are carefully
pruned periodically and reduced to a bushy form, until they are from
two to three years old, when the first picking commences, the exact
time for picking being determined by the overseer. The leaves are
removed in such a manner as to cause no subsequent injury to the
plants, by which care the India planter is enabled to obtain from
twelve to sixteen pickings in a single season, the Chinese grower
being limited to three or four at the utmost.
Each separate picking in India is termed a “Flush,” a number of
flushes constituting a “Break” or “Chop,” as in China, which is rarely
more than 100 chests and frequently as low as 20, but generally
uniform in grade. There is another remarkable feature about India
teas; it is that while the first, second and third pickings of all other
teas are respectively inferior to each other there is nothing in the
India pickings to denote their relationship to any crop or gathering.
The number of pickings from the India plant also varies considerably
according to the soil, situation, garden and season. When all these
conditions are favorable, the plantation will yield as many as sixteen
“flushes,” while ordinarily and often under the most unfavorable
conditions five to six are obtained in a single season.
There is no radical difference between the Chinese and Indian
methods of preparation up to what is termed the “Rolling process;” it
being performed in the latter country very lightly and only by a
minimum of pressure by machinery. Each day’s collection is
immediately “withered” until thoroughly evaporated, when they are
as promptly cured and fired. The processes of fermenting and firing
are not as detailed or complete as in China, the India planter aiming
to secure the component properties of a strong tea at the expense of
flavor and keeping qualities. In India the tea is generally prepared
from the young shoots, two leaves only being picked at a time and
“withered” in the open air without any extraneous aid, much,
however, depending on the skill and knowledge of the operators in
arresting the process at the exact moment. When the proper point is
reached they are immediately removed to a “drying” room, and laid
out on trays until the excessive moisture has been dissipated, this
process being hastened by occasional blasts of hot air driven
through by a machine. When sufficient moisture has been extracted
they are placed in a heavy rolling machine and tossed about until all
the cellular tissues are broken, when they begin to curl up tightly, as
if by the action of the hand, after which they are placed in heaps on
tables for some hours to allow them to ferment; the color, meanwhile,
changing from green to a dark bronze during the process.
In the process of “firing” the leaves are spread out in a series of wire-
gauze trays, placed in layers in a hot-air machine, known as a
“Sirocco,” from the fact that the current of vapor arising from it is
suggestive of the hot winds of the desert, and in which the
temperature averages some 300 degrees. These screens are
operated either in a lateral or rotary direction also by steam, the tea
being thoroughly fired in from twenty to twenty-five minutes, and
separated into the different grades at the same time. But on some
plantations the tea is afterwards bulked in large tin-lined cases until a
considerable quantity is accumulated, when it is again lightly fired,
the operations of sorting and grading being again performed by
machinery previous to being packed in the teak-wood chests, in
which it is finally shipped. The curing and firing of tea by hot air and
machinery in India is fast superseding the primitive arrangements
and charcoal processes so long in use in China. Yet though much
more rapid and effective in its work, and certain not to taint the
leaves in any manner, it is still an open question whether the older
and slower methods of curing in pans over charcoal fires is not after
all the better one. That the teas are not properly cured or thoroughly
fired by this over-hasty method is evidenced by the fact that India
teas in general are noted for their great excess of tannin and peculiar
raw, “grassy,” uncooked or herby flavor. But labor and fuel-saving
machinery are effecting such economy in the cultivation and
preparation of tea in India as to yearly reduce the cost of its
production. So many improvements for drying, rolling, firing and
sorting are annually being recorded that it is difficult even to estimate
at what figure it may be produced there in the future.
India teas comprise Assams, Cachars, Darjeelings, Deradoons,
Kumaons, Dooars, Chittagongs, Juligoories, Rangworths and
Neilgherries, district terms, ranking in the order named, and are
converted into Pekoes, Souchongs, Pekoe-Souchongs, Congous,
Broken-leaf and Fannings. In make, style, color, flavor, and general
appearance, India teas resemble most the Congou sorts of China,
but many of them being produced from a combination of the China
and India plants are hybrid in character, differing essentially from
either originals. Most of them possess a sharp, acrid taste, not to be
found in any other variety, and a peculiar flavor rarely liked by
consumers, unless when tempered with the softer and more mellow
China growths, and to neutralize which peculiarity it is at all times
necessary to use only the best India grades. In make they are in
general longer and narrower in leaf, darker in color, more shapely,
better curled or twisted, and finer in texture than the corresponding
Chinese varieties.
Assams—Are greyish-black in color, the leaf of the finer grades

being “Pekoe-tipped” and evenly curled. The liquor is unusually
strong and pungent, in addition to being thick and heavy in body. The
infused leaf dark-brown, with a reddish tinge, and almost perfect in
form.
Cachars—Are blacker in color, but not as well curled or even in

appearance. The liquor is softer and occasionally “fruity,”
approaching a burnt flavor, while the infused leaf is larger, darker
and not as finely shaped.
Darjeeling—Is a hybrid variety, produced from a cross between the

China and India species, and partaking somewhat of the character of
both. It is still blacker in the dry leaf, but on an average not as finely
curled, and while full in body is not as pungent or flavory in the cup.
The infused leaf is more bright, tender, shapely and “salmony” red in
color.
Kangras—As a rule are dark and symmetrical in leaf, light in liquor,

but delicate and aromatic in flavor. The infused leaf is reddish-brown
in color, with dark or burnt edges, but perfect in shape and form.
Deradoon—Is a high-fired tea, loosely made and deteriorating

rapidly, becoming sour on exposure to the air. Occasionally the flavor
is “earthy,” analagous to that of Ankoi Oolong, for which reason they
are not much sought after.
Kumaon—Is generally converted into Green teas, including

Imperials, Gunpowders, and Hysons, all being prepared from the
same leaf. The chief difference lies in their make and color, as they
still retain all the characteristics of liquor and flavor of India teas.
Chittagongs—Are strong, thick and heavy in the cup; “nutty” in

flavor and considered good, useful teas for blending purposes, from
their great strength and positive character, for which qualities they
are always in good demand.
Dooars—Approximate to Cachars in color, make and general

appearance, strong, but rough in liquor, pungent and pleasing in
flavor, a valuable tea for blending, imparting tone and character to
any combination in which they may be used.
Neilgherry—Is a very inferior sort, bearing the same relation to India

teas that Ankois do to Oolongs and Pingsueys to Green teas. The
leaf in general is black, coarse, “tippy,” rough and unsightly in the
hand, while the liquor is thin, muddy and rank or “weedy” in flavor.
Travancore—Is a “new district” tea, which, like all new teas, is large
and coarse in leaf, heavy and dark in liquor, and strong and wild or
“grassy” in flavor.
Juligoorie and Rangworths—Are bold in style, rather rough in

make, but regular and well developed. The liquor is thick and rich in
color, rough or “rasping” in flavor, but occasionally smooth and
“toasty,” while the infused leaves are bright and well formed as a
rule.
SUB-VARIETIES OF INDIA TEA.
India Pekoes—Are ordinarily of a greyish-black hue, with a fair

sprinkling of grayish-yellow tips, downy in appearance, while the
liquor is very strong, brisk and pungent, varying in quality and flavor
according to the district of production.
Orange-Pekoe—Is a small, evenly-curled leaf, having a yellowish or

golden “tip” at the ends. In liquor and flavor it approximates close to
plain Pekoe, being devoid of scent, that many growers make no
distinction between them.
Flowery-Pekoe—Is not picked from the plant, but separated from

the other grades, only the buds and youngest leaves being selected.
The cured leaf is small, uniform and tender, silvery-green in color,
although highly-fired, pale but strong in liquor, approaching that of a
Moyune Green in flavor, being very deceptive in strength and
astringency. The infused leaf is symmetrical in form, small and light-
green in color, approaching that of a Foochow Oolong in appearance
in the cup.
Souchong—Forms the bulk of the India product and may be classed

as the “Standard grade;” the qualifications for being comprehended
under this rating are its even, straight, slightly curled leaf, dark color,
stylish appearance and greater quantity. Yet while its liquor does not
possess the deep strength and pungency of the Pekoe sorts, it is
generally full and round in body and mellow or “malty” in flavor.
Pekoe-Souchong—Is a term applied to Pekoe leaves devoid of tips,

as well as to Souchong containing a fair sprinkling of tipped leaves.
But, as a general rule, it is an unassorted tea, composed principally
of the larger and coarser leaves of both Pekoe and Souchong that
will not pass through the sieves, and possessing in the cup the
distinctive properties of the combination.
India Congou—Is a tea of the Souchong order too large to be made

into that kind or a smaller leaf unevenly prepared. In liquor and flavor
it is much the same as Souchong, but is not always as heavy, strong
or mellow in flavor.
Broken-leaf—As its name implies, is composed of a mixture of the

various kinds broken in manipulation, and is a term of great
comprehensiveness, as it may include all the lower grades or
approach the choicest kinds in character and value. It varies in color
from brown to blackish, its strength being seldom great, though the
flavor of the finer grades is, in general, good; that of the commoner
ones being poor, thin and coarse.
India Bohea—Consists chiefly of the old and coarser leaves which

do not attain a desirable black color in firing, being devoid of sap.
The leaf is generally brown, sometimes yellowish in color, the liquor
possessing scarcely any strength, usually coarse and rough in flavor,
and never of much value at any time.
Fannings—Are composed of the refuse, much broken leaves and

dust of all the preceding kinds, and bear the same relation to India
teas that Twankays do to Green and Nibs to Japan teas.
Namuna—In Hindostanee literally means “Sample,” being

accidentally applied to a class of India tea, possessing great strength
and high, peculiar flavor not confined to any particular district or
plantation. The dry leaf may have the regular grayish-black hue, or
be of a greenish-black color, the green leaves being intermixed and
distinct from the black ones. It invariably yields a pale, corn-yellow
colored liquor, resembling that of Oolong, heavier and stronger than
ordinary Pekoe, and in flavor like a Moyune, yet distinct from the
former and not as pungent as the latter. Frequently, however, it is
intermingled with a nasty black leaf, the flavor of which is destroyed
by over-firing, the green leaves being due to deficient or under-firing.
There are many serious objections to the general use of India teas,
one of which is the great excess of tannin (tannic acid) which they
contain, ranging from 13 to 18 per cent. in this variety, and to which
property tea owes its astringency, constipating effect on the bowels
and the ink-black color which it imparts to water containing salts of
iron. In England a crusade is being preached against their use by
medical authorities on this account, the marked increase in dyspeptic
and nervous diseases in that country being attributed to their general
consumption there. Some experts argue that by a shorter infusion—
sufficiently long to extract the theine with less of the tannin—this
serious defect may be eventually remedied. Such, however, is not
the case, as experiments made with it at three and five minute
infusions have still shown an excess of tannin, in addition to that of
making the liquor raw, herby, and entirely unsatisfactory in flavor.
The same time-tests resulting in favor of both China and Japan teas,
and which, judging by the bitterness and astringency, the amount of
tannin yielded by India teas in a five-minute draw is incredible. While
China teas, under the same conditions, possesses little or no trace
of tannic acid, or offending the most sensitive palate or constitution,
but on the contrary being both pleasing and refreshing to the most
sensitive natures. Another distinct and dubious feature of India teas
is the formation of a gummy or oily film which settles on top of the
infusion when drawn, and claimed to be very injurious to the nervous
system and digestive organs. When first infused this substance is
scarcely discernible, but just as soon as the liquor begins to cool this
opaque coating forms and develops on top. It is of an oily, creamy or
gummy nature, forming a thin layer of a dull, whitish-brown color,
more dense than the liquor and changing to a darker shade as it
cools. Its nature or effect has not yet been definitely determined, but
sufficient is known to prove that it is particularly unwholesome, for
their selection is also more difficult than that of any other variety
owing to their well-known tendency to early decay, becoming sour
and rancid on short exposure to the oxydizing influences of the
atmosphere, the greatest caution having to be exercised in avoiding
those that will not keep for any length of time owing to this most
objectionable peculiarity, losing flavor quicker and decaying faster
than any other kinds, not even excepting low-grade Japans. This
loss of flavor and rapid decay is greater in some sorts than in others,
the grades most easily affected in this manner being the highly-fired,
light-flavored and open-leaf makes.
The demand for India teas in this country is only limited, owing to the
present taste of consumers, and there appears little hope of any
increase in the future. What little is sold being used chiefly for
blending with the softer and more mellow-flavored teas of China; the
India grades supplying the absent quality of strength to the latter.
Strenuous efforts have and are being made to introduce them, but so
far with indifferent success. The character of the liquor after the
infusion is so entirely foreign in body, color, flavor and aroma from
that of the China and Japan sorts to which the people have been
accustomed, and which appears to be an inherited taste, so deeply
is it set, that little or no progress can be made in these attempts. The
great strength, pungency and pronounced flavor of the choicer
grades rendering them valuable only for blending purposes. Still it is
difficult to overestimate the importance of India as a source of tea
supply. Twenty years ago it furnished only about 10,000,000 pounds
to the world’s supply, but so rapidly has its production increased that
the crop for 1892 is estimated at 110,000,000 pounds. Its
consumption in England is annually increasing, the total deliverance
for that year being 103,000,000 pounds as against 99,000,000
pounds for 1890, while for 1889 the increase was upwards of
12,000,000 pounds over that of 1888. These enormous strides in the
consumption of India teas in England is only equalled by that of
Ceylon teas, the British public demanding strong, dark liquoring teas
irrespective of flavor, aroma or effect.
CEYLON TEAS.
The tea-plant, though claimed to have been first introduced into
Ceylon by the English, who, on principle, “claim everything,” was
originally carried by the Dutch from China to that island as early as
1800, notwithstanding that Percival maintains that it was first
discovered there in a wild state. But while it is admitted that a
species known as Matara was found in some parts of the island,
later investigation proved that it had no relation whatever to that of
the regular teas of commerce. Tennant, in 1842, was the first
Englishman to speak of Ceylon as a possible tea-growing country,
but the highly profitable cultivation of coffee at that time attracted so

Essays On Freges Basic Laws of Arithmetic First Edition Edition Ebert Full Chapter

Uploaded by

Copyright:

Available Formats

You might also like

Essays On Freges Basic Laws of Arithmetic First Edition Edition Ebert Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Essays On Freges Basic Laws of Arithmetic First Edition Edition Ebert Full Chapter

Uploaded by

Copyright:

Available Formats

Essays on Frege’s : basic laws of

arithmetic First Edition. Edition Ebert

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 1 — #1

Essays on Frege’s Basic Laws of Arithmetic

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 2 — #2

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page iii — #3

Essays on Frege’s Basic

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page xii — #12

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page v — #5

1 The Basic Laws of Cardinal Number 1

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page vi — #6

12 Frege on Creation 285

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page vii — #7

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page viii — #8

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page ix — #9

work by Michael Dummett. Yet a broader engagement with Frege’s magnum

Philip Ebert, Stirling, Scotland

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page x — #10

Michael Beaney is Professor of History of Analytic Philosophy at the Hum-

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page xi — #11

Stewart Shapiro is O’Donnell Professor of Philosophy at The Ohio State Uni-

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page xii — #12

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 1 — #13

Frege begins his Grundgesetze der Arithmetik as follows:

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 2 — #14

2 Richard Kimberly Heck

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 3 — #15

The Basic Laws of Cardinal Number 3

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 4 — #16

4 Richard Kimberly Heck

1.1 THE PROOF OF HP

Here, ‘Nx : F x’ is to be read: the number of F s.

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 5 — #17

The Basic Laws of Cardinal Number 5

7 Theorem 32, which is the right-to-left direction, reads:

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 6 — #18

6 Richard Kimberly Heck

∃R[∀x∀y(Rxy → ∀z(Rxz → y = z)) ∧

∃R[∀x∀y(Rxy → ∀z(Rxz → y = z)) ∧

What Frege has done is group ‘∀x∀y(Rxy → ∀z(Rxz → y = z))’ and

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 7 — #19

The Basic Laws of Cardinal Number 7

where ‘Eq(F, G)’ abbreviates: ∃R[Map(R)(F, G) ∧ Map(Conv(R))(G, F )].

x̂(F x) = x̂(Gx) ≡ ∀x(F x ≡ Gx)

1.2 THE AXIOMS OF ARITHMETIC

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 8 — #20

8 Richard Kimberly Heck

Table 1.1. One Version of the Dedekind–Peano Axioms

His definition of predecession, which he writes as ‘mS(nSs)’, is as follows

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 9 — #21

The Basic Laws of Cardinal Number 9

b and the F s other than c, and if b is a G and c is an F , then there is also a

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 10 — #22

10 Richard Kimberly Heck

The concept Nξ is then definable as P∗= 0ξ . So an object is a natural number

“Essays_on_Frege_Basic_Laws_OUP_final” — 2019/7/29 — 11:29 — page 11 — #23

The Basic Laws of Cardinal Number 11

This is easy enough to prove—it follows by substitution from Frege’s Theorem