Schemata: The Concept of Schema in the History of Logic

Author(s): John Corcoran

Source: The Bulletin of Symbolic Logic , Jun., 2006, Vol. 12, No. 2 (Jun., 2006), pp. 219-
240
Published by: Association for Symbolic Logic

Stable URL: https://www.jstor.org/stable/4617260

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Association for Symbolic Logic is collaborating with JSTOR to digitize, preserve and extend
access to The Bulletin of Symbolic Logic

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
THE BULLETIN OF SYMBOLIC LOGIC

Volume 12, Number 2, June 2006

SCHEMATA: THE CONCEPT OF SCHEMA IN THE HISTORY OF

LOGIC

JOHN CORCORAN

Abstract. Schemata have played important roles in logic since Aristotle's Prior Analytics.
The syllogistic figures and moods can be taken to be argument schemata as can the rules
of the Stoic propositional logic. Sentence schemata have been used in axiomatizations of
logic only since the landmark 1927 von Neumann paper [31]. Modern philosophers know
the role of schemata in explications of the semantic conception of truth through Tarski'
1933 Convention T [42]. Mathematical logicians recognize the role of schemata in first-orde
number theory where Peano's second-order Induction Axiom is approximated by Herbrand's
Induction-Axiom Schema [23]. Similarly, in first-order set theory, Zermelo's second-order
Separation Axiom is approximated by Fraenkel's first-order Separation Schema [17]. In some
of several closely related senses, a schema is a complex system having multiple components
one of which is a template-text or scheme-template, a syntactic string composed of one or
more "blanks" and also possibly significant words and/or symbols. In accordance with a side
condition the template-text of a schema is used as a "template" to specify a multitude, often
infinite, of linguistic expressions such as phrases, sentences, or argument-texts, called instances
of the schema. The side condition is a second component. The collection of instances may
but need not be regarded as a third component. The instances are almost always considered
to come from a previously identified language (whether formal or natural), which is often
considered to be another component. This article reviews the often-conflicting uses of the
expressions 'schema' and 'scheme' in the literature of logic. It discusses the different definitions
presupposed by those uses. And it examines the ontological and epistemic presupposition
circumvented or mooted by the use of schemata, as well as the ontological and epistemi
presuppositions engendered by their use. In short, this paper is an introduction to the history
and philosophy of schemata.

?1. Introduction. In some of several closely related senses, a schema

ral: schemata, or schemas), also known as a scheme (plural: schemes),
complex system having several components one of which is a template-t
or scheme-template, a syntactic string composed of one or more "bla
and usually also words and/or symbols to which meanings are assig
The template-text of a schema is used as a "template" to specify a multit
often infinite, of linguistic expressions such as phrases, sentences, argum
texts or proof-texts. The collection of specified expressions, called instan

Received July 30, 2005.

? 2006, Association for Symbolic Logic

1079-8986/06//1202-0003/$3.20

219

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
220 JOHN CORCORAN

of the schema, may but need not be re

The instances' are almost always consid
tified language (whether formal or nat
another integral component of the sch
Each schema is accompanied, implicitly
specifying how the "blanks" (dummi
are to be "filled" to obtain instances, and also, sometimes, how the other
words or symbols are to be understood [42, p. 55]. This side condition may
also be regarded as another component of the schema or schematic system, as
it may appropriately be called. Characters and character strings commonly
used as blanks include: a "space", a dash, a short line, three dots, three
dashes, dot-dash-dot, and so on. But there are many other strings used as
blanks: ordinal number expressions ('the first', 'the second', etc.), letters
of various alphabets, numerals, circled characters, and boxed characters,
to mention a few other possibilities. It is important to note that the side
condition does not assign a denotation to the blanks nor does it assign them
even the "meaning" normally attached to a variable. In short, blanks are
mere place-holders; they do not "range over a domain".
More precisely, the template-text of a given schema is a string of characters
each of which is in the alphabet of the underlying object language or is an
auxiliary character. As suggested above, the auxiliary characters are called
blanks, place-holders, dummies, or other similar term (cf. [49, p. 473]).
Regrettably, they are also misleadingly referred to as (syntactic or metalan-
guage) variables.
There are two limiting cases of such strings over an extended alphabet:
(1) no blanks, a string of characters of the object language, and (2) nothing
but blanks. The strings falling under the first limiting case are simply object
language expressions, not template-texts.2 The strings falling under the
second limiting case are rather strange - they are simply strings of blanks,
auxiliary characters. In this case it would be awkward to use a "space" for a

' The word 'instance' is ambiguous. With regard to schemata it is unambiguous: in

order for an expression from the underlying language to be an instance of a given schema it is
necessary and sufficient for that expression to be a member of the set specified by that schema.
In other connections, however, the word 'instance' is used as a relation noun for any one of
several other relations. For example, it is often used as a synonym for 'token' in Peirce's sense
of a concrete inscription. But here an instance of a schema is not a concrete string-token but
an abstract string-type, again in Peirce's sense. Peirce's type-token distinction dates from the
first few years of the 1900s. Although he made the distinction earlier, his use of the words
'type' and 'token' to express it dates from the 1906 Monist article on pragmaticism [35],
quoted in Ogden-Richards [33, Appendix D, Section 6, esp. pp. 280-281] and reprinted in
volume IV of Collected Papers [36, Paragraph 537]. See below for more on type-token
ambiguity.
'This point is just a convenient technical convention. Imagine using the sentence' '0 = 1'
is true iff 0 = 1' with the side condition that both occurrences of '0 = 1' are to be filled by
occurrences of one and the same English sentence.

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 221

blank. Surprisingly perhaps, many of these turn up as template-texts i

interesting and familiar historically important applications to be menti
below.
Schema T. Among the best-known schemas is Tarski's schema T,
known as the schema (T), whose template-text is the eight-word two
string: ' ... is a true sentence if and only if... '. The side condition r
that the second blank is to be filled in with a (declarative) sentence of
and the first blank is to be filled in by a name of that sentence [42, p.
The following two strings are instances - the first uses a quotes-na
the sentence filling the second blank, the second uses a phonetic descrip
name:

'zero is one' is a true sentence if and only if zero is one

zee-ee-ar-oh-space-igh-es-space-oh-en-ee is a true sentence if
zero is one

More revealing instances are obtained by using a sentence not known to

be true and not known to be false: 'every perfect number is even' is a true
sentence if and only if every perfect number is even. Another such instance
is: 'some perfect number is odd' is a true sentence if and only if some perfect
number is odd.3
Defining the notion. Usage varies and can be controversial, and there
is much that is not clear or not known. Moreover, no generally recognized
definition or characterization of a concept of schema is known to the author.
Consequently, it would be premature, and ultimately unhelpful, to attempt
a single, strict, formally and materially adequate definition of "schema" at
this point, not to mention an even more demanding analytically adequate
definition that would "capture some definite concept". Nevertheless, it may
be heuristically useful to consider possible tentative, informal definitions
that seem to be presupposed in the literature.
The literature seems to confirm that in many cases a schema, or schematic
system, may be construed as a four-part system (or ordered quadruple)
consisting of a scheme-template, a side condition, an underlying language,
and a set of instances. Other construals may be found to be convenient
for various purposes. For example, a schema may be identified with the
three-part system (ordered triple) composed of the scheme-template, the
side condition, and the underlying language. The latter construal may be
found especially attractive to those who wish to emphasize that the set of
instances is determined by the three-part system composed of the scheme-
template, the side condition and the underlying language. Some logicians

3The meticulous accuracy and discipline needed in this field is illustrated by the fact, often
ignored, that Tarski's side condition implicitly requires that the initial capitalization and final
period (full stop) not be considered part of the sentence. For example, the sentence is 'zero
is even', not 'Zero is even', not 'zero is even.' and certainly not 'Zero is even.'. Some people
never grasped this point (Quine [39, p. 12]).

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
222 JOHN CORCORAN

may sometimes seem to identify the

the set of instances alone, or with
scheme template and either the side
Of course, nothing precludes the side
of the underlying language.
The wording used by Tarski [42, p
he identifies the schema or scheme
A decade later, using the word 'sch
expression, a string, that he calls
"expression" here called a schema-t
schema of a sentence". Incidentally, t
'schema' in [45]. It is clear that 'schem
term but rather that it is being used a
This terminology was also adopted by
and it continues to be used (e.g., G
Church wording [6, p. 149] seems
of the schema with the schema-tem
paper [49] "Axiomatizability by a Sc
'is axiomatizable by a schema' without
his remarks using the word make per
this paper. However, none of the lo
know of, directly raise the issue of wh
simply as a string of characters. It is
be no objective criticism of identifyi
this is a matter of terminology and n
It is clear that one and the same s
of any number of different schem
on the underlying language or on b
write as though the schema is to b
alone also use one and the same sch
ferent underlying languages and thus
since different characters are used
one notational change produces a diffe
(Corcoran [12]), one and the same set
ferent scheme-template/side-conditio
It may be this fact that leads some au
to be identified with the set of insta
specified instances that is of primary

4In order for "two" strings to be identical (

for "them" to be character-by-character id
in the same order. This commonplace poin
Rosenbloom [40, p. 152] has a clear statemen
they have the same length and the same sig

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 223

what is involved in specifying it is considered a mere technicality. Each suc

construal or identification corresponds to a different sense of the ambiguo
word 'schema'.
Pre-logical usage. Speakers of English are well aware that the nouns
'schema', 'scheme' and their cognates are parts of normal English and that
non-logical uses of these words predated their incorporation as technical
terms into logic. Since stipulated technical usage is powerless to completely
cancel metaphorical implications, overtones and connotations accompany-
ing ordinary usage, it will be useful to review, however briefly, a contempo-
rary lexicographic treatment of them. According to the authoritative 2000
Merriam- Webster's Collegiate Dictionary [30], the noun 'scheme' entered
the language in 1610 almost 300 years before 'schema' which is dated "ca.
1890". Moreover the older word 'scheme' was borrowed from Latin where
it meant "arrangement, figure". In contrast, the newer word 'schema' was
borrowed from Greek where it was cognate with the verb "echein to have,
to hold, to be in (such) a condition". The older noun 'scheme' has been
used in connection with mathematical diagrams, graphic sketches and out-
lines, concise statements and tables, plans of action (especially those which
are crafty or secret) and systematic organized frameworks or designs. The
words 'design' and 'plan' are among the synonyms listed for 'scheme'. Use
of the younger noun 'schema', again according to this dictionary, has been
much more limited: aside from a technical use in cognitive psychology to-
tally irrelevant to present concerns, its only use listed is for "a diagrammatic
presentation; broadly: a structured framework or plan". The word 'outline',
literally "a line that marks the outer limits of an object or figure", is the only
synonym listed for 'schema'. References to other dictionaries will be made
below as relevant.
The rest of the paper is organized as indicated in the list below.
2. Examples of schemas.
3. The ontological status of schemas.
4. Schemas in the history of logic.
5. Argument-texts and argument-text schemas.
Bibliography.

?2. Examples of schemas. Excluded-Middle. The fourteen-word sentence

'either zero is even or it is not the case that zero is even' is an instance of
the excluded-middle sentence schema for English, which involves the eight-
word-two-blank string 'either ... or it is not the case that ... '. The side
condition is that the two blanks are to be filled by occurrences of the same
well-formed English declarative sentence, that the discontinuous expression
'either . . . or...' expresses classical non-exclusive "or" and that the six-word
sentence-prefix 'it is not the case that' expresses classical negation. Notice
that this scheme-template is not an English sentence and it would be strictly

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 224 JOHN CORCORAN

speaking incoherent to use it as a sen

another method to indicate blanks wo
For example, instead of using the th
correct to use a letter such as pee o
string:
either p or it is not the case that p
with the side condition that the two blanks are to be filled by occurrences of
the same well-formed English declarative sentence, that the discontinuous
expression 'either ... or ... ' expresses classical non-exclusive "or" and
that the six-word sentence-prefix 'it is not the case that' expresses classical
negation. The schema-template resembles in some ways the string that
Tarski [43, p. 38], [44, p. 33] uses to express what he calls the law of identity:
If p, then p.
But Tarski has already introduced an abbreviational convention according
to which initial universal quantifiers are omitted. Thus the string that Tarski
is using to express his law of identity is strictly speaking not the above but
rather:

For every sentence p, if p, then p.

It is clear that pee is not a blank but a "sentential variable" whose range is
the class of whatever sentences denote or have as "values".
However, the proper part of the schema-template devoid of blanks is
literally a discontinuous part of each of its instances and, in fact, what all
instances have in common is having this discontinuous string-like entity as a
part. It is only in the strictest and most rigorous contexts where a distinction
is to be made between, for example, the discontinuous two-word expression
'if... then ... ' and the continuous two-word two-blank scheme-template
indicated by the same string.
Naming. As another example, the seven-word sentence 'the expression
'zero' names the entity zero' is an instance of the naming schemaJ or English,
which involves the five-word-two-blank string 'the expression ... names the
entity ... ', where the second blank is to be filled by a proper name of an
entity and the first blank is to be filled by a proper name of a proper name of
the entity.5 Another instance of the naming schema is 'the expression zee-
ee-ar-oh names the entity 0', where 'zee-ee-ar-oh' is an orthographic proper
name of the four-letter word 'zero', which is a proper name of the entity 0,
i.e.. zero. The twelve-character string 'zee-ee-ar-oh' in this usage is similar
to the Tarskian structural-descriptive name of the four-letter word 'zero' as
in the 1933 truth-definition paper [42, pp. 157. 160, 172].

5Notice that the first blank may but need not be filled with a name of the expression
filling the second blank. The previous example illustrates the "may" and the next example
illustrates the "need not". These and related points were suggested by Ricardo Santos (per.
comm.).

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 225

Use and mention. As Tarski points out, phonetic or other similar descr
tive devices permit a much sharper distinction between "use" and "menti
than the quotes-name method. In perhaps the most direct senses of the e
pressions 'uses' and 'mentions', in every case the quotes-name method
an expression in order to mention it. As Frege said [19, p. 261]: "As
proper names of the sentences of the object-language I use these very
tences, but enclosed in quotation marks". In a phono-orthographic me
it is only in rather atypical cases that an expression is used in order to m
tion it. For example, zee-ee-ee names zee. The last letter of the alphab
the first letter of its own name. Normally the expression mentioned is n
used. In the instance 'the expression zee-ee-ar-oh names the entity zero' i
clear that 'zero' is used but not mentioned in the object position, whereas
is mentioned but not used in the subject position.
However, in order to make an analogous point about the instance
expression 'zero' names the entity zero' it is necessary to stipulate
natural sense of 'uses' or to make the claim, called metaphysical and
theological by non-logicians, that although the subject of the senten
the six-character string quote-zee-ee-ar-oh-quote, which the sentence
the sentence does not use the four-character string zee-ee-ar-oh (Tarski [
pp. 104-105]).6
However, if we use the expression 'word' in one of the familiar rich sen
in which a word is something that, for example, is inherently meaningful
that might have more than one spelling or more than one meaning, then t
is a lot more to using a word than just using the string that spells it. Wh
I use the English word spelled see-aitch-ay-tee, I am not thereby usin
English words aitch-ay-tee, ay-tee, and ay. And I am not thereby using t
French word spelled see-aitch-ay-tee, the German word aitch-ay-tee, and
Spanish word see-aitch-ay. Similarly, in a straight-forward sense of '
not every sentence using the eight-letter word 'recenter' uses the four-l
word 'cent'. But, it would invite criticism to say that one can use the str
word 'recenter' without using the string 'cent'.
Tarski's Original Template-text. The example given in the introduc
section above, Tarski's schema T, warrants further discussion. In the
mentioned above, the template-text for Tarski's schema T is the eight-w
two-blank string: ' ... is a true sentence if and only if... '. In its ori
1933 formulation Tarski used ecks for the first blank and pee for the se
x is a true sentence if and only if p
The side condition includes the requirement that the second blank is to
filled in with a sentence of English and the first blank is to be filled in
name of that sentence (Tarski [42, p. 155]). It is clear that ecks and p

6Frango Nabrasa (per. comm.) said: "This act of using would be like eating an e
hard-boiled egg without eating the yolk, something only slightly less miraculous than
Monotheistic Trinity".

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 226 JOHN CORCORAN

place-holders, that ecks is not a "sen

not a "sentential variable." Tarski's 19
string but it raises no new theoretical
uses the letter pee as the only blank.
"p" is a true sentence if and o
What is remarkable here is that the
quote is not being used as a quotes nam
formulation of the side condition cont
where "p" is to be replaced on
which does use quote-pee-quote as a
1944 template-text contains the follo
" ... " is a true sentence if an
whose first gap follows the first q
second occurrence of igh-eff. Later
ess-quote in two senses: for the quote
gap enclosed in quotes, rather for the
abbreviated by ess [47, p. 108].
Sentence. Although Tarski does no
'sentence' is taken in the sense of the clausal form of sentence which does not
end with a period (or full stop) and does not begin with an uppercase unless
its initial word is capitalized. For example, 'zero is one' is the clausal form of
the sentence whose assertoric form is 'Zero is one.' ending with a period or
"full stop". By an assertoric form of sentence is meant a string of characters
that is or could normally be used to make an assertion. By a clausal form
sentence is meant a string of characters that is or could normally be used as
a proper part of an assertoric form.7
Further, in the 1933 article Tarski [42] is not using the word 'sentence'
in the modern sense of an "uninterpreted sentence" or "closed formula"
(Boolos [4, p. 103]), rather he intends, in his words [42, p. 178], a "meaningful
sentence" that says something in virtue of which it has a truth-value. Tarski
is using the word 'sentence' in exactly the same sense that traditional post-
scholastic logicians attached to the word 'proposition', according to Church
[6, p. 26, and fn. 68].8 In Tarski 1933, a sentence is a string of characters
that is true or is false as it stands; in Boolos et al 2002 a sentence per se is

7Although Frege [19, p. 261] introduces no new terminology nor does he note the initial
capitalization and final period, he does note other differences between assertoric and clausal
forms in German. viz. word order. In word-for-word translation it would be 'Jan is a man'
and 'Jan is mortal', but 'If Jan a man is, is Jan mortal'.
8This is not the sense of 'proposition' that Church prefers. In his preferred sense, a
proposition per se has no wording but can be expressed in many different wordings in
different languages. Davis [13, p. xxii] wrote: "The fundamental property of a statement is
that it is asserts a proposition that must be true or false". Thus, he seems to follow Church's
preferred use of 'proposition' while using 'statement' as Tarski uses 'sentence'.

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 227

not something that is either true or false, rather it is either true under a
interpretation or false under that interpretation.
Moreover, 'if and only if' is to be taken in the truth-functional sense a
not as involving any modality or necessity.9 In order for an instan
Tarski's schema T to be modal it is necessary and sufficient for the sente
on the right of the "principal" occurrence of 'if and only if' to be modal
important point here is that an instance of Tarski's schema T does no
that anything is necessarily the case. For example the instance menti
above says that 'zero is one' is a true sentence if and only if zero is on
does not say that it is necessarily the case that 'zero is one' is a true sent
if and only if zero is one.
Circularity and Quotes-names. Tarski's schema T, which he calls a schem
not a schema, is of special philosophical significance since, accordin
Tarski and many others, each instance of the schema is a "partial definit
of truth", or rather of "true sentence". Accordingly, the naming sch
may be regarded as giving partial definitions of the concept of naming,
rather of the relational concept "names".
Phonetic naming of alphabetic and other characters, 'ay' for the first l
ter, 'bee' for the second, and so on, was a commonplace in ancient wr
Greek and Latin but is curiously rare in written English except for its pro
nence in the 1933 truth-definition paper and in the literature deriving f
it. Its advantages continue to be exploited in the best of contemporary lo
ical writing (cf. e.g., Boolos [4, pp. 102, 107]). It is common, but stric
speaking incorrect, to talk of the name of a sentence or other string;
string has many names. Tarski [42, pp. 156, 157], [47, pp. 104, 105] featur
two systems of string naming, both already used above: the quotation-m
name, as 'zero' and the structural-descriptive name, as zee-ee-ar-oh. B
other systems are widely used, e.g., the italic name, as the proper name
and the common noun thing (Church [6, p. 8]). The structural-descri
or phonetic-name method, which Tarski characterizes as "a letter-by-lett
description" [47, p. 105], is preferred by Tarski because it avoids the
pearance of a vicious circle" when an instance of Tarski's schema is regar
as a definition as mentioned above. Compare the following three inst
of the naming schema.
the expression 'zero' names the entity zero
the expression zee-ee-ar-oh names the entity zero
the expression zero names the entity zero
The first has two occurrences of the string zee-ee-ar-oh, i.e., of the st
'zero', one in the subject and one in the direct object, an apparent circula
or triviality. The second has only one, in the direct object; thus avoi

9I do not recall seeing this point in print by Tarski or by anyone else. However, T
uniformly avoids modal assertions. He even goes so far as to construe ' Q is a conseq
of P' as the non-modality "every model of P is a model of Q".

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 228 JOHN CORCORAN

any appearance of circularity or tr

reader may have either opinion on th
occurrence in its subject.

?3. Ontological status of schemas.

mixed ontological status of schemas. I
of words, symbols, etc., and blanks
the schema for emphasis, is a syntact
involves the same ontological presupp
and the like. As seen above the templ
'the expression ... names the entity
involving twenty-seven letter-occurr
seven occurrences of the period. In
of a schema is another syntactical o
expression 'ten' names the entity 10
for English. But the side conditions
propositions in the traditional abstrac
Nagel [7, pp. xxii, xxiii]) or abstract t
Demopoulos (per. comm.).
Type-token-occurrence. Further d
requires Peirce's type-token distinctio
Monist article [35, pp. 504-505] quo
281].
A common mode of estimating the amount of matter in a ...
printed book is to count the number of words. There will ordinar-
ily be about twenty 'thes' on a page, and, of course, they count as
twenty words. In another sense of the word 'word,' however, there
is but one word 'the' in the English language; and it is impossible
that this word should lie visibly on a page, or be heard in any voice
.... Such a ... Form, I propose to term a Type. A Single ...
Object ... such as this or that word on a single line of a single
page of a single copy of a book, I will venture to call a Token.
... In order that a Type may be used, it has to be embodied in a
Token which shall be a sign of the Type, and thereby of the object
the Type signifies. I propose to call such a Token of a Type an
Instance of the Type. Thus there may be twenty Instances of the
Type 'the' on a page.
I do not know whether Peirce ever supplemented his type-token distinction
with the concept of "occurrence" as needed to make the point that, although
the type 'the' has only one occurrence of the type 'e', the type 'e' occurs twice

~?As important as this passage has become in modern linguistics, analytic philosophy, and
logic I have been unable to find a single quotation of it besides the one by Ogden-Richards [33]
copied here.

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 229

in the type 'thee' and it is instantiated, betokened, or embodied" (

Peirce's term) twice in every token of the type tee-aitch-ee-ee. In order
two string occurrences in the same or different string-types to be occur
of one and the same string-type it is necessary and sufficient for them
"character-by-character identical", i.e., for them to have the same le
say L, and for the nth character occurrence in one to be an occurren
the same character-type that the nth character occurrence in the other
occurrence of, for n between 1 and L.
Some authors explicitly make the three-part type-token-occurrenc
tinction without introducing special terminology for the occurrence rel
Lyons [29, Vol. 1, pp. 13-18] has a section called Type and token whic
cusses the three-way distinction while using the same expressions for to
and occurrence. However, the present terminology is familiar to logician
pointed out in my "Meanings of word: type-occurrence-token [ 11, p. 11
At some point in the history of logic the type-token dichotomy gav
to the type-token-occurrence trichotomy. Given Peirce's penchant fo
chotomies and his logical creativity, one is led to speculate, even hope, t
it was Peirce who made this discovery.
Template-text and instance. A scheme-template is a string type ha
indefinitely many tokens in Peirce's sense (Corcoran et al. [12, fn. 5, p.
But none of the tokens of a scheme-template are instances of the sc
This is an ontological necessity and not a mere matter of convention
the fact mentioned above that no instance of a sentence schema is a schema-
template. In fact, every instance of a schema is a string type having its
own tokens. As mentioned above, the word 'instance' is being used here,
and throughout the modern literature on schemata, as a relation noun for a
relation certain string types bear to certain schemas, not in Peirce's sense. The
word 'token' is a relation noun for a relation certain macroscopic physical
objects bear to certain abstract objects. The ontological status of a scheme-
template is the same as that of an instance of a schema. Moreover neither
a schema nor a scheme-template is a common noun denoting the instances
and neither is a proper name of a set of instances.
Some philosophers emphasize the ontological economies possible by us-
ing schemas rather than second-order axioms (see below). But, as is often

1'Notice that Peirce says that the token embodies the type and is an instance of the type, not
that the token is in or is a member of the type. He says that the type is a word in one sense of
'word'. There is no suggestion however slight that the type is to be identified with the huge and
evolving class of its tokens. Of course, Peirce's theory-laden terminology can not be regarded
as in any way binding on later logicians. Nevertheless, it is a gross but common mistake,
perhaps an anachronism, to write as if Peirce and those who have adopted his terminology
take the type to be the class of its tokens. This mistake occurs in otherwise respected and
authoritative sources such as the 1993 New Shorter Oxford English Dictionary [32, p. 3441 of
Volume 2] and the 1999 Cambridge Dictionary of Philosophy [1, pp. 936-937].

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
230 JOHN CORCORAN

the case, economies have their own

phers present a full and objective disc
commitments" entailed by the use
ple, number theory per se presuppose
functions and numerical properties, b
of mathematical notation and it a for
of the vast, intricate notational syste
theory. Sometimes the use of schem
mitments of the object language whil
or at least not achieve any net saving
Alternatives to schemas. Normally, t
uses a given schema goes far beyon
example, Tarski's schema T has heur
thetic effectiveness that is not overlo
were specification of instances the so
without using a schema. For example,
be specified as the expression obtaine
fixing to it the result of prefixing on
string 'is a true sentence if and onl
ay-em is a true sentence if and only
prefixing adds a space - prefixing ay
not ay-bee, or 'ab'. In other contexts,
for the operation of concatenating the
ond to yield a third string whose len
is the sum of the lengths of the given
Truly subtle points. No discussion
would be complete unless it at least r
cerning languages, both ideographic
string-types, and the operation Tar
discussion of schemata presupposes th
selves or other strings, which begs th
denoting - the relation that a string
here referred to as the ontological sta
is similar to the issues going back to
versals". Tarski forthrightly addresse
paper where he says that it raises "tr
Tarski's treatment, mentioned in the
et al. [12, footnote 5]), attempts to re
objects by "identifying" the abstract
This class is taken to be an equivalenc
equiform relation which is defined

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 231

shape12. Tarski is well aware that this could undermine the plausibility o
theory of truth by calling into question some of the axioms of his metath
- axioms that are entirely plausible when interpreted as concerning abstr
tions but which imply the existence of arbitrarily large material objects w
taken as he suggests. For example, see Tarski's disclaimers [46, p. 31, f
p. 156, fn. 1, p. 174, fn. 2].

?4. Schemas in the history of logic. The Greek word 'schema' was u
in Plato's Academy for "[geometric] figure" and in Aristotle's Lyceum
"[syllogistic] figure". As mentioned above, the syllogistic figures fall und
an interesting limiting case of "schema" in the senses of this article, as w
be explained below. Each of the four traditional syllogistic figures
argument schema. Kant used the words 'schema' and 'schemata' in a
only remotely related to the senses of this article.
Connotations. The articles schema through scheme in the 1971 Ox
English Dictionary [34, Compact Edition, Volume II, pp. 204-206] and
1993 New Shorter Oxford English Dictionary [32, Volume 2, p. 2711
certainly interesting to anyone concerned with the conceptual framewor
of modern linguistics, logic and philosophy - despite the fact that ne
work recognizes the senses of the words 'schema', 'schemata', 'schema
and 'scheme' most relevant to this article.13 What emerges from reading t
articles is a vague connotation that must have been operative in the think
of the people who coined the current range of meanings and in the think
of those of us who found it easy, even natural, to accept the coinages. I s
to find the vague connotation of 'schema' and its cognates to be close
"incomplete" than to "complete", closer to "potential" than "actual", c
to "apparent" than "real", closer to "abstract" than "concrete", and cl
to "form" than to "matter". But I would not be surprised if others di
with me on such a subjective issue.
Other uses in logic. The above points are relevant to the concerns of th
paper for several reasons not the least of which is that modern logic
even those who are meticulous about using 'schema' in a technical se
continue to use the cognates as standard non-technical English. For examp
Tarski [43, p. 159] uses 'schema' in the context 'schema of the proof' i
sense of "style" or "form", Carnap [5, p. 1] says that an uninterpre
language "is not a language but a schema or skeleton of a language"
Goldfarb [21, p. 352] refers to ordinary non-logical constants as "sche

'2Even this requires modification in view of the fact that in some alphabets toke
different character types have the same shape and differ in some other way. For exam
some mathematical languages union, intersection, inclusion and subset are expressed by
characters whose tokens are all U-shaped.
13There are other uses of 'schema' and its cognates not recognized by the OED
example, 'schema' has been used in the sense of a draft document to be discussed and r
(Encyclopedia Britannica [15, Vol. 6, p. 591]).

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
232 JOHN CORCORAN

placeholders" to emphasize the fact th

the late 1890s, Hilbert said that a th
basic elements can be thought of any
Moreover, in the theory of recursiv
'schema' and its cognates that may s
with in this paper but are actually qu
ples are found in the early papers by
Schemas are also used in computer
Dr. E. Ray Lucas, formerly of IBM
The development of linguistics and
in a proliferation in the use of sc
grammars are schemas, a well-kn
Backus-Naur Form (BNF) product
languages. Another important use
is in the specification of the struct
As late as 1969, the North-Holland
of Symbols of Mathematical Logic
paragraph to this concept but it did n
or any of its cognates in the index
schema' as occurring in one paragra
only once without comment or explan
The 1933 truth-definition paper (Tar
first publications to prominently use
in a sense close to this sense, has be
[42, pp. 155, 156]. The word 'schem
less conspicuously, used by Tarski in
114, 310, 386, 423). One of the earl
is instructive and possibly of historic
word in a different, but related sense.
tion to Mathematical Philosophy [41
function ... may be taken to be a m
receptacle for meaning, not someth
use of 'schema', it is clear that Tarski
technical term nor is he presenting
the authors are simply using the com
thoughts. Boole [3, pp. 33, 42] had alr
of logic' in his discussion of ways of
Frege had used 'scheme' in connecti
as did G6del [20, p. 584]. Logicians
give a first-order formalized first-ord
we find no axiom schemas in Gode
second-order sentences with suppre
that would at first glance appear to

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 233

place-holders. It is clear that in none of these cases is a technical term b

introduced for what we have been calling schemas. On the other han
just as clear that Church 1956 [6] takes 'schema' to be a technical ter
logic. It would be interesting to determine when and how the word acqu
its standing as an element of the technical vocabulary of logic. Ther
other issues to be settled: for example, whether cognates of 'schema' are
only words used in these senses as technical terms today, why it is that
words became accepted so widely and so quickly, and whether these
were ever expressed using any other words.
In the past twenty-five years, the prominence and importance of sche
has become increasingly evident. For example, in Quine's Philosoph
Logic [39] schemas have a prominence not equaled in many previous
For many years, Quine was especially active in exploring various sch
and their uses in logic and philosophy of logic: "On the Logic of Qua
cation" (Quine [37]) was especially influential. The most recent exte
development and articulation of a Quinean approach may be found in Go
farb's 2003 Deductive Logic [21].
Aristotle should probably be credited with introduction of argum
schemas, or more explicitly, premise-conclusion argument-text schemas
though this point has not been settled in the minds of historians of
For example the two-premise argument text
Polygon belongs-to-every multilateral.
Multilateral belongs-to-every square.

Polygon belongs-to-every square.

is an instance of the argument-text schema
P belongs-to-every M
M belongs-to-every S.

P belongs-to-every S

called BARBARA, whose side condition includes requirements th

both occurrences of pee are to be filled with occurrences of one and the
common noun, (2) both occurrences of em are to be filled with occurren
of one and the same common noun other than the one used for pe
both occurrences of ess are to be filled with occurrences of one and the
same common noun other than the ones used for pee and em, and that (4)
the expression 'belongs-to-every' is taken to express universal affirmative
predication as in Prior analytics. There is no dispute that Aristotle used
argument-text schemas similar to the one called BARBARA above. But
it is disputable whether he was aware of the ontological presuppositions
involved, e.g., that we are warranted in speaking of alphabetic characters
as existent individuals, that the common nouns of a language form a well-
defined totality, or that a schema may seem to be a linguistic unity that

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 234 JOHN CORCORAN

ontologically transcends the plurality

that a Platonic form transcends the p
in it.
Every argument text that is an instance of BARBARA is in the first
syllogistic figure, i.e., is an instance of the argument-text schema
PxM
My S.
PzS

whose side condition includes requirements that (1) both occurr

are to be filled with occurrences of one and the same comm
both occurrences of em are to be filled with occurrences of one and the same
common noun other than the one used for pee, (3) both occurrences of ess
are to be filled with occurrences of one and the same common noun other
than the ones used for pee and em, and that (4) the occurrences of ecks,
why, and zee are to be filled with occurrences of 'belongs-to-every', 'belongs
to no', 'belongs to some' or 'does not belong to every' taken to express the
categorical predications as in Prior analytics. The other three figures also
correspond to schemas. Notice that the four figure schema-templates are
composed entirely of blanks and thus fall under a limiting case of schemas.14
Fraenkel [18] may have been more important than is usually thought as
far as the introduction of the use of axiom schemas in modern theories for
delimiting an infinity of axioms. But the full history has yet to be written.
Quine [38, p. 89] writes that "the practice of describing an infinity of axioms
dates only from von Neumann [31]". One year later Kleene [27, p. 140]
also credits von Neumann with the method of using axiom schemata in the
formalization of logic. A few years later Church [6, p. 158] also makes the
assertion that this use of axiom schemas was introduced by von Neumann
in 1927.

14The original 1999 Corcoran article [9] on schemata does not discuss the history of the
subject and consequently omits this theoretically and historically interesting if arcane point
that I was unaware of at the time. The first couple of published versions of the 2004 Corcoran
article [10] on schemata do indeed discuss the history of the subject, but nevertheless they
also omit this point. Even worse. although they cited the Aristotelian moods (tropoi) as
schemas, they explicitly denied that the four traditional figures (schemata) were schemas in
the modern sense. At the time of the publication of the very first of these, it did not occur
to me that there could be schema-templates composed entirely of blanks and containing no
characters from the underlying object language. But as soon as I saw it in print. I realized my
mistake. Instead of immediately making the correction, I decided to wait until more mistakes
were brought to my attention so that the corrections could be done all at once. In addition, I
was curious whether someone else would notice the mistake. Indeed, at least one person did
- my long-time friend and former colleague Mary Mulhern (per. comm.). Even if the line
is regarded as a meaningful character in the argument schema, the three-character sentence
schemas occurring in it are schemas composed entirely of blanks. This was first pointed out
by Joaquin Miller (per. comm.).

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 235

It is hard to date self-conscious use of the word 'schema' in the modern

sense. As mentioned above, Russell's 1919 Introduction to Mathematical
Philosophy [41] uses it casually on page 157 to describe "propositional func-
tions" or, roughly, interpreted open sentences in Quine's sense.
Since first-order number theory cannot be finitely axiomatized, the math-
ematical induction schema is used to effectively specify an infinite number of
axioms: 'If zero is such that ... and the successor of every number such that
... is also such that ... , then every number is such that ... ', where the four
blanks are to be filled in with one and the same arithmetic open sentence
such as 'it is zero or it is a successor', 'it precedes its own successor' or 'it is
finite'. Every instance of the mathematical induction schema is implied by
the second-order principle of mathematical induction: every property that
belongs to zero and that also belongs to the successor to every number to
which it belongs, also without exception belongs to every number. In other
words, the principle of mathematical induction is that in order for a given
property to belong to every number it is sufficient for it to belong to zero
and for it to belong to the successor of every number to which it belongs
(Russell [41, p. 5, 6]).
Contrary to what would at first be thought, the principle of mathematical
induction is a consequence of the set of all instances of the mathematical
induction schema, but of course, it is not a consequence of the set of all first-
order instances. In fact, the principle of mathematical induction is logically
equivalent to a certain single instance of the schema, where the open sentence
is taken to be: 'is such that it has every property that belongs to zero and
to the successor to every number to which it belongs'. To be sure, the set of
first-order instances is weaker than the second-order principle, i.e., it does
not logically imply the second-order principle, as is shown in any advanced
logic text, e.g., Boolos et al. [4].
Sometimes the blanks in a schema are marked by variables; thus, in the
minds of some, blurring the crucial distinction between, on one hand, an
open sentence, such as '(x + y) = (y + x)' whose object-language numerical
variables ecks and wye range over the numbers and, on the other a schema
such as the number-theoretic commutativity schema '(X + Y) = (Y + X)'
whose template-text involves the side condition that the two occurrences of
ecks are to be replaced by two occurrences of one and the same numeral
and likewise for the two occurrences of wye. One critical distinction is that
an open sentence is a formula of the object language whose variables range
over the universe of discourse and which is satisfied (or not) by sequences
of objects in that universe, whereas the variables occurring in the schema's
template-text are what Church calls syntactic variables [6, Sects. 08, 27, 30
and 33], variables that are not in the object language vocabulary but rather
belong to the metalanguage and which range over syntactic objects, strings
of characters, not entities such as numbers in the universe of discourse of

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
236 JOHN CORCORAN

the interpretation of the object-langu

related points. In the case of the va
it is not clear whether there is any n
"range over" syntactic objects. What
syntactic objects as substituents. Th
might be justifiable to call them su
that the expression 'dummy letter'
sometimes to place-holders.
As Church has emphasized (e.g., [6
treatment of schemas requires use of
and it requires an axiomatized theory
in the 1933 Tarski truth-definition
the history, philosophy, and mathem
neglected field, string theory or pur
(Corcoran et al. [12]).

?5. Argument-texts and argument

properly, an argument-text is a two
tences called the premises and a sin
the various ways of presenting an a
misinterpretation is the premises-li
listing the premises followed by a lin
are four examples.

A3 Every circle is a polygon.

Every triangle is a circle.
Every square is a triangle.

Every square is a polygon.

A2 Every circle that is a square th

Every circle that is a square that

Every square that is a circle is a t

'5A presentation of an argument-text is

argument-text. It does not include an asser
argument. Therefore, any sign indicating va
same thing) an implication, is inappropriat
an assertion of the truth of any propositions
misleading to attach the word 'therefore' or an
tions. These points have been made in variou
Structure of Classical Logic"[8]. In view of th
is a sentential connective, its use in a present
"argument" with "sentence" - not to mention

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 237

Al No square is a triangle.
No triangle that is a circle is a square.

AO

No triangle that is a square is not a triangle.

There is no limit to the number of premises in an argument-text. In f

zero-premise argument-texts like AO above have come to play a surprisin
important role in modern logic. Likewise important are infinite-pre
argument-texts, especially the so-called omega argument-texts illust
below.

A Zero is not the successor of zero.

Zero is not the successor of the successor of zero.
Zero is not the successor of the successor of the successor of zero.

Zero is not the successor of any number.

As usual, an argument is said to be valid if the conclusion is a logical

consequence of, or follows logically from, its premise-set. When it is clear
how the sentences are to be interpreted, the same terminology can be applied
to argument-texts, although strictly speaking it is only an argument expressed
by an argument-text under a given interpretation that can properly be said
to be valid or invalid, not the argument-text itself. This parallels the familiar
point that when it is clear how the sentences being discussed are to be
interpreted the sentences themselves can be said to be true or false, although
strictly speaking it not the sentence per se but the proposition it expresses
under an interpretation that is true or false.
An argument schema is said to be pan-valid if all of its instances are valid,
pan-invalid if all are invalid, and neutro-valid if some but not all of its in-
stances are valid, i.e., if it is neither pan-valid nor pan-invalid. See e.g.,
Cohen-Nagel [7, pp. xxxiv-xxxvii]. The advantages of using the same words
to discuss arguments and argument-texts are evident as are the disadvan-
tages. For example, it might lead to confusing argument-schemas with what
the 1993 edition of Cohen-Nagel [7, pp. xxxi-xxxvii] calls argument-patterns
and argument-forms. Also see [1, pp. 511-512].
Just as the schema-template of a sentence-schema may be thought of
as obtained from a sentence by replacing various constituents with place-
holders of one sort or another, the schema-template of argument-text-
schema may be thought of as obtained from an argument-text by replacing
various constituents with place-holders. Rules of inference are associated
with argument-text schemas and axiom schemas can be thought of as zero-
premise argument-text schemas.

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 238 JOHN CORCORAN

Acknowledgements. This article orig

of meetings of the Buffalo Logic Coll
Dictionary Project. It owes much
G. Boger, R. Dipert, P. Hare, K. Herre
J. McNabb, J. Sagiiillo, B. Smith and
it was discussed twice by the Buffa
again in 2005 after corrections and ot
The first publication arising from
in the 1999 second edition of The C
edited by Robert Audi. Readers of t
be told that I have learned a lot abo
deem it near-miraculous that no mista
light, unless it was a mistake that I di
text (a string containing at least one
abstract form or mold. The ambiguou
vague is used in the 1999 article for
species.
The template-text-versus-template distinction, only implicit in that article,
was made explicit by an anonymous referee-editor of the second publication
arising from this project, namely the 2004 article "Schema" in the Stan-
ford Encyclopedia of Philosophy [10]. Of the several scholars who suggested
improvements in previous drafts and versions, I want to mention for spe-
cial thanks those already mentioned above, especially J. Sagiiillo, and also
R. Grandy, D. Hitchcock, A. Kanamori, F. Nabrasa, A. Preus, J. Miller,
M. Mulhern, R. Santos, M. Scanlan and an anonymous referee for The
Bulletin of Symbolic Logic. Above all I thank my friend of more than forty
years Dr. E. Ray Lucas, who informed me of the role of schemata in com-
puter science and reminded me of several relevant points that we had learned
together in the early 1960's in Baltimore, Madison and Berkeley, but that I
had almost forgotten.

REFERENCES

[1] R. Audi (editor), The Cambridge Dictionary of Philosophy, 2nd ed., Cam
versity Press, Cambridge, 1995/1999.
[2] P. BERNAYS, Axiomatic Set Theory, Dover, New York, 1958/1991.
[3] G. BOOLE, The Mathematical Analysis of Logic, Macmillan, Cambridge,
[4] J. BOOLOS, G. BURGESS and R. JEFFREY, Computability and Logic, Camb
sity Press, Cambridge. 2002.
[5] R. CARNAP, Introduction to Symbolic Logic and its Applications, Dover,
1954/1958.
[6] A. CHURCH, Introduction to Mathematical Logic, Princeton University Press, Prince-
ton, 1956.
[7] M. COHEN and E. NAGEL, Introduction to Logic, 2nd ed., Hackett, Indianapolis, 1993,
1st ed. Harcourt, Brace and World, New York, 1962, originally published as Book I of An

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 THE CONCEPT OF SCHEMA IN THE HISTORY OF LOGIC 239

Introduction to Logic and Scientific Method, Harcourt, Brace, and Company, New Yo
1934.
[8] J. CORCORAN, Conceptual structure of classical logic, Philosophy and Phenomenolog
Research, vol. 33 (1972), pp. 25-47.
[9] , Scheme, In Audi [1], 1999.
[10] - , Schema, Stanford Encyclopedia of Philosophy, 2004.
[11] , Meanings of word: type-occurrence-token, this BULLETIN, vol. 11 (2005),
p. 117.
[12] J. CORCORAN, W. FRANK, and M. MALONEY, String theory, The Journal of Symbolic
Logic, vol. 39 (1974), pp. 625-637.
[13] M. DAVIS, Computability and Unsolvability, 2nd ed., McGraw-Hill, Dover, New York,
1958/1980.
[14] , The Undecidable, Raven Press, Hewlett, 1965.
[15] ENCYCLOPEDIA BRITANNICA, Chicago, London, Toronto, 1953.
[ 16] R. Feys and E Fitch (editors), Dictionary of Symbols of Mathematical Logic, North-
Holland Publishing Company, Amsterdam, 1969.
[17] A. FRAENKEL, Introduction, [2], 1958/1991.
[18] , The notion "definite" and the independence of the axiom of choice, In Hei-
jenoort [22], 1967, pp. 284-289.
[ 19] G. FREGE, Philosophical and Mathematical Correspondence, Chicago University Press,
Chicago, 1980.
[20] K. GODEL, The completeness of the axioms of the functional calculus of logic, In
Heijenoort [22], 1967, pp. 582-591.
[21] W. GOLDFARB, Deductive Logic, Hackett, Indianapolis, 2003.
[22] J. van Heijenoort (editor), From Frege to Gidel, Harvard University Press, Cam-
bridge, MA, 1967.
[23] J. HERBRAND, Logical Writings, (W Goldfarb, Tr. Goldfarb, and van J. Heijenoort,
ecitors), Harvard University Press, Cambridge, MA, 1971.
[24] R. HUGHES, Philosophical Companion to First-order Logic, Hackett, Indianapolis,
1993.
[25] S. KLEENE, General recursive functions of natural numbers, Mathematische Annalen,
vol. 112 (1936/1965), no. 5, pp. 727-742, also, in [14], pp. 237-254.
[26] , Recursive predicates and quantifiers, Transactions of the American Mathe-
matical Society, vol. 53 (1943/1965), pp. 41-73, reprinted in [14], pp. 255-287.
[27] - , Introduction to Metamathematics, Van Nostrand, Princeton, 1952.
[28] L. LINSKY, Semantics and the Philosophy of Language, University of Illinois Press,
Urbana, 1952.
[29] J. LYONS, Semantics, Cambridge University Press, Cambridge, 1977, 2 Vols.
[30] MERRIAM-WEBSTER'S COLLEGIATE DICTIONARY, Merriam-Webster, Springfield, MA,
2000.

[31] J. VON NEUMANN, Zur Hilbertschen Beweistheorie, Mathematische Zeitschrift, vol. 26

(1927), pp. 1-46.
[32] NEW SHORTER OXFORD ENGLISH DICTIONARY, Oxford University Press, Oxford, 1993.
[33] C. K. OGDEN and I. A. RICHARDS, The Meaning of Meaning, 1923, reprint of the
eighth edition, Harcourt, New York.
[34] OXFORD ENGLISH DICTIONARY, compact ed., Oxford University Press, Oxford, 1971.
[35] C. S. PEIRCE, Prolegomena to an apology for pragmaticism, Monist, vol. 16 (1906),
pp. 492-546.
[36] -, Collected papers of Charles Sanders Peirce, (C. Hartshorne and P. Weiss,
editors), Harvard University Press, Cambridge, 1933.

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms
 240 JOHN CORCORAN

[37] W. QUINE, On the logic of quantification,

pp. 1-12.
[38] - , Mathematical Logic, Harvard University Press, Cambridge, MA, 1951.
[39] , Philosophy ofLogic, Harvard University Press, Cambridge, MA, 1970/1986.
[40] P. ROSENBLOOM, Elements of Mathematical Logic, Dover, New York, 1950.
[41] B. RUSSELL, Introduction to Mathematical Philosophy, George Allen and Unwin,
London, 1919.
[42] A. TARSKI, The concept of truth in the languages of the deductive sciences, Prace
Towarzystwa Naukowego Warszawskiego, Wydzial III Nauk Matematyczno-Fizycznych, vol.
34 (1933), reprinted in [50], pp. 13-172: expanded English translation in [48], pp. 152-278.
[43] , Introduction to logic and to the methodology of deductive sciences, Dover,
New York, 1941/1946/1995, Trans. O. Helmer.
[44] , Introduction to logic and to the methodology of deductive sciences, Oxford
University Press, New York, 1941/1994, ed. with preface and biographical sketch of the
author by J. Tarski.
[45] ~ , The semantic conception of truth, Philosophy and Phenomenological Research,
vol. 4 (1944), pp. 341-376, in [28], pp. 13-49.
[46] - , Logic, Semantics, Metamathematics, papers from 1923 to 1938, Oxford Uni-
versity Press, Oxford, 1956.
[47] - , Truth andproof' Scientific American, (1969), also, in [24], pp. 101-125.
[48] , Logic, Semantics, Metamathematics, papers from 1923 to 1938, 2nd ed.,
Hackett, Indianapolis, 1983, edited with introduction and analytic index by J. Corcoran.
[49] R. VAUGHT, Axiomatizability by a schema, The Journal of Symbolic Logic, vol. 32
(1967), pp. 473-479.
[50] J. Zygmunt (editor), Alfred Tarski, Pisma Logiczno-Filozoficzne, 1 Prawda,
Wydawnictwo Naukowe PWN, Warsaw, 1995.

UNIVERSITY OF BUFFALO
DEPARTMENT OF PHILOSOPHY
BUFFALO, NY 14260-4150, USA

E-mail: [email protected]

This content downloaded from

137.189.171.235 on Thu, 04 Jan 2024 09:55:27 +00:00
All use subject to https://about.jstor.org/terms