The Problem of Language and Mental Repre John O'Callaghan

The Problem of Language and Mental Representation in Aristotle and St.
Thomas
Author(s): John P. O'Callaghan
Source: The Review of Metaphysics, Vol. 50, No. 3 (Mar., 1997), pp. 499-545
Published by: Philosophy Education Society Inc.
Stable URL: https://www.jstor.org/stable/20130071
Accessed: 27-03-2020 12:40 UTC
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 support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Philosophy Education Society Inc. is collaborating with JSTOR to digitize, preserve and
extend access to The Review of Metaphysics
This content downloaded from 129.74.250.206 on Fri, 27 Mar 2020 12:40:22 UTC
All use subject to https://about.jstor.org/terms
THE PROBLEM OF LANGUAGE AND MENTAL
REPRESENTATION IN ARISTOTLE AND ST. THOMAS
JOHN P. O'CALLAGHAN
Introduction. In the opening passages of his De Interpretationen

Aristotle provides a simple summary of how he thinks language re
lates to the mind and the mind to reality, a sketch which has often
been called his "semantic triangle." He writes:
Now spoken sounds are symbols of affections in the soul, and written
marks symbols of spoken sounds. And just as written marks are not the
same for all men, neither are spoken sounds. But what these are in the
first place signs of?affections of the soul?are the same for all; and
what these affections are likenesses of?actual things?are also the
same. These matters have been discussed in the work on the soul and
do not belong to the present subject.2
It has been called a triangle because of the three vertices, words,

affections of the soul, and actual things. It is semantic because it has
been interpreted to be providing a sketch of the meaning of words,
and how they relate to things. As Norman Kretzmann points out, in
the form of Boethius's sixth-century Latin translation, this passage
"constitute[s] the most influential text in the history of semantics,"3
having an enormous influence on the subsequent philosophical tradi
tion of reflection upon the interrelations of language, mind, and the
world, or as Hilary Putnam often puts it, "how language hooks onto
Correspondence to: Department of Philosophy, University of Notre

Dame, Notre Dame, IN 46556.
xThe Greek title is Peri hermeneias.
2Aristotle, Aristotle's Categories and De Interpretatione, trans, with
notes by John L. Ackrill (Oxford: Clarendon Press, 1963), 43.
3Norman Kretzmann, "Aristotle on Spoken Sound Significant by Conven
tion," in Ancient Logic and Its Modem Interpretations, ed. John Corcoran
(Dordrecht, Holland: D. Reidel Publishing Company, 1974), 3-21. Norman
Kretzmann, "Semantics, History of," in vol. 7 of The Encyclopedia of Philoso
phy, ed. Paul Edwards (New York: Macmillan Publishing Co., 1967), 367.
The Review of Metaphysics 50 (March 1997): 499-545. Copyright ? 1997 by The Review of
Metaphysics
500 JOHN P. O'CALLAGHAN
the world." This is particularly true of the Middle Ages, but also be
yond into modern philosophy. Indeed, among some contemporary
philosophers, there is a vision of this opening passage that one might
call a standard or received view, namely that Aristotle's reflections
upon language in the De interpretation planted a seed that grew rela
tively continuously in Western philosophy, flowered within British
empiricism, and continues to influence the philosophy of language to
this day.
In light of criticism in this century directed at this relatively con
tinuous tradition, however, these contemporary philosophers charac
terize the Aristotelian tradition as fundamentally flawed. Consider
Michael Dummett's remarks:
A continuous tradition, from Aristotle to Locke and beyond, had as

signed to individual words the power of expressing ideas', and to com
binations of words that of expressing complex ideas'; and this style of
talk had blurred, or at least failed to account for, the crucial distinction
between those combinations of words which constitute a sentence and
those which form mere phrases which could be part of a sentence.4
Perhaps the most important of all the contributions made by Grund

lagen to general philosophy is the attack on the imagist or associationist
theory of meaning. This is another of those ideas which, once fully di
gested, appear completely obvious: yet Frege was the first to make a
clean break with the tradition which had flourished among the British
empiricists and had its roots as far back as Aristotle. The attack that
was launched by Frege on the theory that the meaning of a word or ex
pression consists in its capacity to call up in the mind of the hearer an
associated mental image was rounded off by Wittgenstein in the early
part of the Investigations, and it is scarcely necessary to rehearse the
arguments in detail, the imagist theory now being dead without a hope
of revival.5
In fact, even recent translators and commentators on the De interpre

tation make oblique reference to the "notorious" problems with Aris
totle's remarks.6 Norman Kretzmann, on the other hand, hopes to pro
tect Aristotle from these kinds of criticism by distancing Aristotle's
text from its Latin tradition of interpretation. He believes that the "tra
ditional misreading of [these] passages" in the West is a product of
Boethius's unfortunate translation of the Greek words for "symbols"
4 Michael Dummett, Frege: Philosophy of Language, 2d ed. (Cambridge:

Harvard University Press, 1981), 3-4.
5 Ibid., 637-8.
6 ". . . the inadequacy of this as an account or explanation of thought is
notorious"; Ackrill, Aristotle's Categories and De Interpretatione, note to
16a3,p. 113.
LANGUAGE AND MENTAL REPRESENTATION 501
(auu?oXa) and "signs" (or||xeia) by the single Latin word "notae,"

thereby "obliterating the Aristotelian distinction between symbols and
symptoms."7 He believes that Boethius's translation, in conjunction
with the adverb "primarily" (jcq??tcoc) has led, or misled the tradition
into attributing the view to Aristotle that our words primarily or di
rectly involve mental impressions, and only secondarily or indirectly
involve things.8 This charge would certainly gain support from the
Blackfriars translation of question 13, article 1 in the Prima Pars of
the Summa Theologiae, which, citing the passage from the Peri
hermeneias, reads:
Aristotle says that words are signs for thoughts and thoughts are like
nesses of things, so words refer to things indirectly through thoughts.9
The first phrase ending in "of things" is the translator's paraphrase of

the quotation from the Peri hermeneias that appears in the Latin text.
The second phrase ending in "thoughts" is the translation of St. Tho
mas's explanation of the Aristotelian quote. The translator translates
mediante conceptione intellectus as "indirectly through thoughts." It
is no great leap to go on and conclude from this translation that St.
Thomas is committed to the thesis that words are directly related to
thoughts, perhaps that they even directly "refer" to thoughts. I be
< lieve, however, that this is a misunderstanding, and a misleading trans
lation. In addition, Kretzmann's charge is false, if it is leveled at St.
Thomas's interpretation of Aristotle. However, I cannot argue that
case here.10 For the purposes of this article, I merely want to indicate
that it is a plausible charge directed at the Aristotelian tradition.
Though he does not attribute responsibility for the problem to Ar
istotle, Robert Sokolowski, casting a critical eye over philosophical
accounts of the relation between words and concepts or ideas, writes:
7 Kretzmann, "Semantics," 367.

8 See Kretzmann, "Aristotle on Spoken Sound," 7.
9 "Dicendum quod secundum Philosophum, I Peri hermeneias, voces
sunt signa intellectum, et intellectus sunt rerum similitudines; et sic patet
quod voces referuntur ad res significandas mediante conceptione intellec
tus"; Thomas Aquinas, Summa Theologiae (hereafter, S.T), I, q. 13, a. 1,
trans. Herbert McCabe (Cambridge: Blackfriars, 1964), 49.
101 pursue this in a forthcoming paper, a paper that specifically ad
dresses how in their commentaries Ammonius, Boethius, and St. Thomas,
among others, interpret this passage in Aristotle. One can also examine John
Magee's fine treatment in John Magee, Boethius on Signification and Mind
(Leiden: E. J. Brill, 1989).
In a Lockean spirit we have allowed words to range only over the do

main of our ideas, and we have tacitly taken ideas to be some sort of in
ternal things. But philosophically this is terribly naive.11
This problem, the yoking of language to ideas or mental representa

tions conceived of as "internal things," animates many discussions of
language and the philosophy of mind in contemporary philosophy.
The contemporary critics often use mental representationalism as a
blanket term to refer to the philosophy of mind to which the account
of language is yoked. John Haldane provides the following paraphrase
of mental representationalism:
It is the view that the immediate objects of cognitive acts or states are
internal entities: species, ideas, images, sentential formulae and such
like, which may or may not stand in some further referential relation to
objects and features in the world; and that it is the former, inner, rela
tional attitudes which constitute the essential object-directed', or inten
tional character of cognitive states.12
Alexander Broadie points out that Thomas Reid, already in the eigh
teenth century, criticized Hume specifically, but with him the whole
tradition back to Plato, for holding theses such as Haldane describes.13
More recently, in Representation and Reality, Hilary Putnam re
fers to the theses from the De interpretatione as "a scheme that has
proved remarkably robust" for characterizing meaning and reference,
though he ultimately rejects it as fatally flawed.14 He refers to it
broadly as "the Aristotelian view," and associates its constitutive the
ses with a number of others purportedly advocated by Mill, Frege,
11 Robert Sokolowski, "Exorcising Concepts," The Review of Metaphys

ics 40 (March 1987): 456.
12 John Haldane, "Reid, Scholasticism and Current Philosophy of Mind,"
in The Philosophy of Thomas Reid, ed. Melvin Dalgarno and Eric Matthews
(Dordrecht: Kluwer Academic Publishers, 1989), 287.
13 Alexander Broadie, "Medieval Notions and the Theory of Ideas," Pro
ceedings of the Aristotelian Society 87 (1987): 153-67.
14 Hilary Putnam, "Meaning, Other People, and the World," in Represen
tation and Reality (Cambridge: MIT Press, 1988). For a more recent and
congenial approach by Putnam to Aristotelian treatments of the relation be
tween language and the mind, see his "Aristotle after Wittgenstein," in Mod
em Thinkers & Ancient Thinkers, ed. Robert W. Sharpies (Boulder: West
view Press, 1993), 117-37; and his recent Dewey Lectures, "Sense, Nonsense,
and the Senses: An Inquiry Into the Powers of the Human Mind," The Journal
of Philosophy 91, no. 9 (September 1994): 445-517. See also his joint article
with Martha Nussbaum, "Changing Aristotle's Mind", in Essays on Aristotle's
De Anima, ed. Martha C. Nussbaum and Am?lie Oksenberg Rorty (Oxford:
Clarendon Press, 1992), 27-56.
Russell, Carnap,15 and to a certain extent Chomsky, Fodor, and Searle.

Consider, for example, John Searle's programmatic comments about
the philosophy of language:
Since Frege, reference has been regarded as the central problem in the
philosophy of language; and by reference I mean not predication, or
truth, or extension but reference, the relation between such expressions
as definite descriptions and proper names on the one hand, and the
things they are used to refer to on the other. I now think it was a mis
take to take this as the central problem in the philosophy of language,
because we will not get an adequate theory of linguistic reference until
we can show how such a theory is part of a general theory of Intention
ality, a theory of how the mind is related to objects in the world in gen
eral.16
Any philosopher in the least favorable toward the passage from the De
interpretation cannot but look with approval at this passage from
Searle. As we have seen, however, Putnam is not alone in his charac
terization of this approach and its criticism. Consider again Michael
Dummett:
The British empiricists and Aristotle, whereby a word represents an
"idea," and a phrase or sentence accordingly represents a complex of
ideas, is simply too crude to serve even as a starting point; it virtually
forces us to adopt the conception whereby the meaning of a word is em
bodied in a mental image.17
The fundamental criticism of these contemporaries is that our lan

guage never actually succeeds in attaining the world, but remains
trapped in the murky internal depths of the mind, despite the explicit
desires of Aristotelians. On the supposedly Aristotelian account, in
our apparent talk about trees at best we succeed in bringing to mind
our mental representations of trees. There are serious difficulties
with the further thesis that language can "hook onto the trees" via
these mental representations. If these critics are right, it is only rela
tively recently, that is, in this century, that the philosophical tradition
has managed to emerge from this unfortunate account of mind and
15 Putnam, "Meaning, Other People, and the World," 19.

16 John Searle, Expression and Meaning: Studies in the Theory of
Speech Acts (Cambridge: Cambridge University Press, 1979), xi. From this
quote, it would appear that Searle differs from Putnam on the interpretation
of Frege. For an argument from Putnam on why Frege, despite his anti-psy
chologism, ought to be included in the list, see his "The Meaning of 'Mean
ing'," in Mind, Language and Reality (Cambridge: Cambridge University
Press, 1975), 222.
17 Michael Dummett, "Frege, Gottlob," in vol. 3 of The Encyclopedia of
Philosophy, 228.
language. We might refer to the characterization that these critics pro

vide as the "received view of Aristotelian representationalism," and
their criticism, to the extent that it displays a certain unity, the "re
ceived criticism."
However, casting an eye over the long tradition embodied in Aris
totle and his interpreters, one might wonder whether Sokolowski's ap
praisal, confining itself to Locke, isn't considerably more judicious
than Putnam's, or Dummett's. Though we cannot survey the whole of
the Aristotelian tradition, we can examine one of its longest lasting,
and most influential interpretations, namely that of St. Thomas
Aquinas. If we consider this particular interpretation of Aristotle's
theses, does it appear anachronistic to assign to Aristotle responsibil
ity for the problems of language and mental representationalism? If
we examine St. Thomas's interpretation of Aristotle in its own con
text, do we find that it suffers from the difficulties leveled at it in con
temporary philosophy? Is the received view accurate?
In fact, St. Thomas's account of Aristotle's semantic triangle
avoids the main thrust of the contemporary criticism. To show this, I
will focus upon Putnam's criticism. However, my purpose is not to
evaluate the soundness of Putnam's larger argument with mental rep
resentationalism, but rather to test its adequacy as an argument ad
dressed to Aristotle and St. Thomas. Consequently, I will do four
things. First, I will give a brief account of Aristotle's remarks in the De
interpretation, and of St. Thomas's understanding of them. Next, I
will provide a sketch of the historical context in which Putnam's criti
cism should be understood. I will then examine directly Putnam's crit
icism. Finally, I will indicate why this criticism does not, in fact, "cut
at the joints" of St. Thomas's interpretation of Aristotle.
II
Aristotle and St. Thomas. At the beginning of the article, I

quoted Ackrill's translation of Aristotle. I did this because in his notes
to the passage, Ackrill himself seems to share something like Putnam
and Dummett's view, calling the problems associated with the text
"notorious." Here, because I am going to examine St. Thomas's appro
priation of Aristotle, I will provide a translation of the text St. Thomas
was commenting on, namely Boethius's Latin translation of the Greek.
Boethius's Latin does not differ that much from the Greek except, as
Kretzmann notes, in the translation of both "symbols" (?u|i?oA,a) and
"signs" (?Tifxeia) by the single Latin word "notae." In addition, where
the Greek has the adverb "jtQ x ?," translated by Ackrill as "in the
first place," Boethius has the genitive plural adjective "primorum."18
Aristotle writes:
Thus, those which are in articulated sound are signs of those passions
which are in the soul: and those which are written are of those things
which are in articulated sound. And just as letters are not the same for
all, so neither are articulated sounds the same: but the passions of the
soul are the same for all, of which first passions these are signs, and the
things of which these are likenesses are also the same. But these things
were spoken of in those which were said of the soul, for that is another
work.19
The first point to notice is that "sign" is used to characterize the writ
ten word relative to the spoken word, and the spoken word relative to
the passions of the soul. However, when it comes to characterizing
passions of the soul relative to things, "sign" drops out of the discus
sion, to be replaced by "likeness." As we will see, the use of "likeness"
in this context adds weight to the contemporary criticism. On the
other hand, it suggests that the relation of passions of the soul to
things is not the same kind of relation as between words and passions
of the soul.
In any case, according to St. Thomas the subject matter of the
work is "significative vocal sound."20 Aristotle himself provides the
appropriate context for understanding his remarks, by his reference to
"[those] things [which] were spoken of in those which were said of the
soul, for that is another work." This is commonly taken to be a refer
ence to the De anima. Apostle and Edghill simply cite the De anima.
Ackrill is more specific in associating it with De anima III 3-8, start
ing where Aristotle distinguishes sensation, imagination, understand
ing, and opinion, and where he treats mostly of imagination, ending
where Aristotle recapitulates what he has just done. All this comes
18 A central presupposition of Kretzmann's claims about the errors of

Boethius's translation centers upon this difference between an adverb and an
adjective. As I note above, I intend to examine his claims in another paper.
19 "Sunt ergo ea quae sunt in voce earum quae sunt in anima passionum
notae. Et ea quae scribuntur eorum quae sunt in voce. Et quemadmodum
nece littere omnibus eaedem, sic nee eaedem voces. Quorum autem hae pri
morum notae sunt, eaedem omnibus passiones animae sunt, et quorum hae
similitudines, res etiam eaedem. De his quidem dictum est in his quae de an
ima dicta sunt; alterius est enim negocii"; text of Aristotle in 16a3-9 in Sancti
Thomae De Aquino, Eocpositio Libri Peryermenias (hereafter, CPH) (Roma:
Commisio Leonina, 1989).
20 St. Thomas cites Boethius. The Leonine editors attribute this citation
to Boethius's Commentarii in librum Aristotelis Peri hermeneias, proe
mium.
before chapter 9, where he begins to treat of the soul as moving princi

ple of the body. Ackrill, however, notes that the text may not have
been Aristotle's at all, but rather the addition of an editor. This textual
suspicion coheres well with Ackrill's general philosophical suspicions
concerning the text. The Leonine editors note that Aristotelian com
mentators are uncertain as to the place (de loco dubitant expositores)
in the De anima.
St. Thomas also attributes the citation to the De anima, writing:
[Aristotle] excuses himself from a more attentive consideration of these
things, because what nature the passions of the soul may be, and in
what way they may be likenesses of things, was spoken of in the book
De anima; for this work does not pertain to logic, but to the philosophy
of nature.21
His commentary suggests that he has roughly the same passages in

mind as Ackrill. St. Thomas picks up Aristotle's point and stresses the
account of the soul and its cognitive powers as background for the De
interpretation. By referring back to the De anima, the analysis of
the signification of words and statements is immediately associated
with a logically prior discussion of the means22 by which human be
ings understand things (res) existing beyond the soul (extra ani
mam).23 St. Thomas writes elsewhere, "any name would be aimlessly
21 ". . . excus?t [se] a diligentiori horum consideratione, quia quales sint

animae passiones et quomodo sint rerum similitudines, dictum est in libro De
anima; non enim hoc pertinet ad logicum negotium, sed ad naturalem";
Aquinas, CPH, L. 1, lectio 2, no. 22.
22 "Non enim se habent [species intelligibiles] ad intellectum sicut quod
intelligitur, sed sicut quo intelligit." Sancti Thomae Aquinatis, In Aristotelis
librum de anima commentarium (hereafter, CDA), L. 3, lectio 8, nos. 717-18
(Turin: Marietti, 1936).
23 Consider also Jerry Fodor: "The idea, to put it in a nutshell, is that it
might be possible to pull off a double reduction: First derive the semantic
properties of linguistic symbols from the intentional properties of mental
states; then postulate a population of mental symbols?mental representa
tions, . . . , and derive the intentional properties of beliefs and desires from
these"; "Mental Representation: An Introduction," in Scientific Inquiry in
Philosophical Perspective, ed. Nicholas Rescher (Washington: University
Press of America, 1987), 114. Consider1 also the passage from Searle quoted
above. I mention Searle and Fodor not to suggest that they are simply con
tinuing the Aristotelian account, but merely to gesture at the similarities. It is
not clear that when Aristotle and Aquinas relate terms with understanding,
their intention is a "reduction" in whatever sense Fodor might have in mind.
Further, it will become clear in this article that the latter's intention is not to
postulate a realm of "mental representations," in the modern and contempo
rary sense.
predicated of something unless through that name we might under

stand something concerning it."24
According to St. Thomas, Aristotle's reference to "passions of the
soul" (passions animae) involves an effort to avoid Platonism, in the
account of the signification of general terms in scientific discourse.
His text reads:
But here [his] saying concerns sounds signifying by human institution;

and so it is necessary that here passions of the soul be understood as
conceptions of intellect,2* which names, and verbs, and sentences sig
nify, according to the view of Aristotle. So it cannot be that they signify
things themselves without mediation, which appears from the mode of
signifying: for this name man signifies human nature in abstraction from
singulars. So it cannot be that it signifies a singular man without media
tion; hence the Platonists held that it signified the separate idea itself of
man. But because in the view of Aristotle, this, according to its abstrac
tion, does not really subsist, but is solely in the intellect, it was neces
sary for Aristotle to say that articulated sounds signify the conceptions
of the intellect without mediation and things by their mediation.26
General vocal terms signify the intellect's concepts, and, via the medi
ation of the concepts, signify res extra animam. St. Thomas's ac
count of Aristotle's thesis focuses on the manner in which general
terms signify, their modus significandi. General terms do not signify
general things beyond the soul (res extra animam) or Platonic ideas.
The generality of the term is taken from the human activity of know
ing singular things in a general concept. The extramental thing signi
fied by the noun "man" is Socrates, or Xanthippe, or Helen, or Plato,
or Aristotle, all singular beings, not a general Platonic idea.
"Thing signified" is used loosely and informally here; it should not
be taken for the more formal notion of the res significata, or simply
24 ?Frustra aliquod nomen de aliquo praedicatur nisi per illud nomen al

iquid de eo intelligamus"; S. Thomae De Aquino, Summa Contra Gentiles, L.
1, cap. 33 (Turin: Marietti, 1946).
25 He is ruling out the possibility in this context that "passions of the
soul" be taken as "affections of the sensitive appetite," since certain groans
may more or less naturally, not by institution, signify pain.
26 "Sed nunc est sermo de vocibus significativis ex institutione humana,
et ideo oportet passiones animae hie intelligere intellectus conceptiones quas
nomina et verba et orationes significant, secundum sententiam Aristotelis:
non enim potest esse quod significent immediate ipsas res, ut ex modo signif
icandi apparet: significat enim hoc nomen homo naturam humanam in ab
straction a singularibus, unde non potest esse quod significet immediate
hominem singularem. Unde Platonici posuerunt quod significaret ipsam
ideam hominis separatam; sed, quia haec secundum suam abstractionem non
subsistit realiter secundum sententiam Aristotelis, sed est in solo intellectu,
ideo necesse fuit Aristoteli dicere quod voces significant intellectus concep
tiones immediate, et eis mediantibus res"; Aquinas, CPH, L. 1, lectio 2, no. 15.
the significata of a general vocal utterance. Using the more formal

notion of the significata, a general vocal utterance does not take as
its signification a singular thing itself, say Aristotle, but rather the
form or nature of a singular thing, say human nature in Aristotle, or
Plato, or Xanthippe. St. Thomas writes:
. . . there are two things to be considered in any name whatsoever:
namely that from which the name is imposed, which is called the quality
of the name; and that to which the name is imposed, which is called the
substance of the name; and the name, properly speaking, is said to sig
nify the form or quality from which the name is imposed; on the other
hand it is said to stand in place of [to supposit for] that for which it is
imposed.27
Using this more formal notion of the significata of a general vocal ut

terance, we would then specify its suppositio in a particular utter
ance, where suppositio is much closer to the contemporary term "ref
erence." So, for example, in "the man who was a student of Plato,
teacher of Alexander, and founded the Lyceum wrote the De interpre
tation^ the significata of "man" is the human nature of Aristotle,
while its suppositio is the singular being Aristotle. In "the man who
taught Aristotle, failed in his mission to Syracuse, and founded the
Academy was not present at the death of Socrates," "man" signifies
the human nature of Plato and supposits for Plato. Because of the
common character of human nature in Aristotle and Plato, "man" has
a common signification in the two sentences, but its supposition dif
fers in the two contexts.
However, if we look closely at the passage from St. Thomas's
Commentary, it is clear that signification, in its application to things
beyond the soul, is not being used in this technical sense, but more in
formally. St. Thomas writes uman signifies human nature in abstrac
tion from singulars," which abstraction is then associated with the in
tellect's concept. Nowhere in the passage does he write that "man"
signifies human nature in singular things. Instead, he writes that the
general vocal utterance "man" does not "signif[y] a singular man with
27 ". . . in quolibet nomine duo est considerare: scilicet id a quo imponi

tur nomen, quod dicitur qualitas nominis; et id cui imponitur nomen, quod
dicitur substantia nominis: et nomen, proprie loquendo, dicitur significare
formam sive qualitatem a qua imponitur nomen; dicitur vero supponere pro
eo cui imponitur"; In 3 Sententia, d. 6, q. 1, a. 3 (my translation), from Gyula
Klima, "The Semantic Principles Underlying Saint Thomas Aquinas's Meta
physics of Being," Medieval Philosophy and Theology 5 (March 1996): 111. I
am indebted to Klima's analysis for the formal notion of signficata. There
are of course further distinctions to be made with respect to the supposition
and signification of concrete and abstract terms. See 15-20.
out mediation." However, having established that it signifies "without

mediation" a conception of the intellect, he returns in the end to tell us
that it does signify "things" by the mediation of the intellect's concept.
Though he does not use "singular" to characterize "things" in this last
line, it seems fairly clear that he has in mind the "singular m[e]n," that
he has just denied are signified without mediation, not human nature
in those men.
Later in his commentary, St. Thomas does make the requisite dis
tinction between signifying the nature in res extra animam, and the
singular res extra animam, when he asks how a general word is to be
"taken" in a particular enunciation.28 There it is clear that what is sig
nified by the nomen "homo" is human nature, which may be that of So
crates, or Xanthippe, or Plato, or Aristotle, or Helen, all singular be
ings. However, it is important to keep in mind that in analyzing the
passage of interest to us here, his use is informal.
Things beyond the soul can be signified by a general term be
cause the general term without mediation, though conventionally, sig
nifies a general concept. The general concept in turn is a natural like
ness of things beyond the soul that are singular beings, because it
shares their form, a form that, when considered absolutely, does not
differ among the diverse things. We can call the form, considered ab
solutely, an intelligible characteristic of the thing. In the intellect,
such a form constitutes a principle of knowing a thing beyond the
soul; it informs the act of understanding, and as a principle of the in
tellect's act it is called the intelligible species.29 The act so informed
28 See, for example, CPH, L. 1, lectiones 10-11.

29 ". . . nam species intelligibilis, qua fit intellectus in actu, consideratur
ut principium actionis intellectus, cum omne agens agat secundum quod est
in actu; actu autem fit per aliquam formam, quam oportet esse actionis prin
cipium"; S. Thomae Aquinatis, Quaestiones Disputatae dePotentia, q. 8, a. 1,
respondeo (Turin: Marietti, 1949).
See also: ". . . reducatur in actum specierum intelligibilium, quemadmo
dum sciens, idest habens scientiae habitum, habet species in actu, et tune
dicitur intellectus qui est secundum actum. Hoc autem accidit, statim, cum
aliquis potest per seipsum operari operatione intellectus, quae est ipsum in
telligere: sicut et quamlibet formam tune aliquis in actu habet, quando potest
operationem illius formae explere"; Aquinas, CDA, L. 3, lectio 8, no. 700.
I am well aware that the interpretation of the intelligible species in Me
dieval authors is complicated and involved. See Leen Spruit, Species Intelli
gibilis: From Perception to Knowledge (Leiden: E. J. Brill, 1994). See also
John Peifer, The Concept in Thomism (New York: Bookman Associates,
1952). Let me say that I am simply stipulating its use here?the form taken
precisely as informing the act of understanding.
is called the concept.30 In the thing beyond the soul that is known, the
form is the principle of being of the thing.31
Despite the conventional relationship between vocal word and
concept, the relationship between the concept and the thing is natural
and the same for all human beings.32 However, this is not a doctrine of
innate ideas, as if the concepts are naturally present in or emanate
from the nature of the intellect as likenesses of things. The concepts
signified by vocal terms are "passions of the soul" (passions ani
mae). St. Thomas writes that the term "passion" is used here to indi
cate receptivity from something else; they have their origin "from
things by means of a certain impression or passion," which things act
as agents upon the soul to cause them.33 It is natural for the intellect
to be structured and delimited by the intelligible characters or forms
that structure and delimit things, to the extent that it is "potentially all
things."34 That passions of the soul are actually likenesses of things
beyond the soul results from the natural principles of the things,
namely, from their origin in things beyond the soul, not from the natu
ral principles of the intellect that is modified by them.35
Without doubt there are many issues presupposed, and difficul
ties to be sorted through in this account of language and the intellect.
I have simply provided a summary. However, for my purposes here, I
30 Here again I am simply stipulating the use of "concept," which has

several different uses in St. Thomas's works.
31 This is not quite accurate, for the contrast does not consider cases in
which something is known from its effects, in which case the principle of
knowing is not formally the same as the principle of being. The full account
is as follows: "... dicuntur eadem essendi et cognoscendi esse principia, quia
quaecumque sunt principia essendi, sunt etiam principia cognoscendi; non
autem e converso, cum effectus interdum sint principia cognoscendi causas";
S. Thomae Aquinatis, Quaestiones Disputatae de Veritate (hereafter, QDV),
q. 3, a. 3, ad 7 (Turin: Marietti, 1949).
32 "In passionibus autem animae oportet attendi rationem similitudinis
ad exprimendas res, quia naturaliter eas d?signant, non ex institutione";
Aquinas, CPH, L. 1, lectio 2, no. 19.
33 For a recent discussion of the relation between concept and res, see
John Haldane, "The Life of Signs," The Review of Metaphysics 47 (March
1994): 469-70.
34 "Concludit autem ex hoc quod non contingit naturam intellectus esse
'?eque unam,' idest nullam determinatum, sed hanc solam naturam habet
quod est possibilis respectu omnium"; Aquinas, CDA, L. 3, lectio 7, no. 681.
35 For a discussion of how "likeness" or "image" is and is not being used
here, see "The Signification of Names" in Ralph Mclnerny, The Logic of Anal
ogy (The Hague: Martinus Nijhoff, 1961), 51-2: "Most properly, the image
originates from the imaged, as the son is the image of the father." Emphasis
added.
want to focus on this fundamental structure as interpreted by St. Tho

mas?words signify things, by signifying, without mediation, concepts
that constitute the understanding of those things.
Ill
Historical Setting. As I indicated at the start, in contemporary

philosophy these supposedly Aristotelian theses have not fared as
well in the hands of recent influential philosophers of language and
mind as they did in the hands of St. Thomas. However, as the passage
from Sokolowski that I quoted suggests, it is important to understand
the difficulties faced by accounts of language and mental representa
tionalism in contemporary philosophy within their historical setting in
modern philosophy. Consider the following comments of Bertrand
Russell:36
The view seems to be that there is some mental existent which may be
called the "idea" of something outside the mind of the person who has
the idea. ... in this view ideas become a veil between us and outside
things?we never really, in knowledge, attain to the things we are sup
posed to be knowing about, but only to the ideas of those things.37
There are two key elements to Russell's difficulty, first the existence
of an internal mental entity which is directly known and which medi
ates the knowledge of things outside the mind, and second the subse
quent skepticism about knowledge of the external world.
The first difficulty receives illumination in the light of themes
stemming from, among others, Descartes, Locke, Berkeley, and Hume.
Let us begin, however, with Locke, for whom it is immediately evident
to the consciousness of every man that "that which his mind is applied
about whilst thinking, being the Ideas, that are there."38 These ideas
function as pictures before the mind, and "represent to us Things un
der those appearances which they are fitted to produce in us."39 The
object of one's knowledge, what one is thinking about or what one
36 Russell speaks of ideas, a term I have not introduced. However, the

gist of the quotation stands as is for concepts. Further, he is not strictly dis
cussing the signification of terms. However, it will become apparent why his
comments bear upon the discussion nonetheless.
37 Bertrand Russell, "Knowledge by Acquaintance and by Description,"
in Proceedings of the Aristotelian Society 10 (1910-11): 119.
38 John Locke, An Essay Concerning Human Understanding (Oxford:
Clarendon Press, 1979), book 2, chap. 1, no. 1.
39 Locke, Essay, book 4, chap. 4, no. 4.
knows, is primarily the private stock of ideas internally present to his

or her own consciousness. He writes:
Every Man's Reasoning and Knowledge, is only about the Ideas existing
in his own Mind, ... ; and our Knowledge and Reasoning about other
Things, is only as they correspond with those our particular Ideas.40
The mind is a "closet wholly shut from light, with only some little
openings left, to let in external visible Resemblances, or Ideas of
things without; would the Pictures coming into such a dark Room but
stay there."41 The ideas in their simplest forms42 have their origin
from things external to the mind.43 However, the things external to
the mind are not present to the understanding.44 In Locke's account
an external thing is represented internally by an appearance that func
tions like a picture of it. Thought attends directly to this appearance.
The signification of words is then constrained to the only things
directly known and talked about, namely, the ideas that fill up our
minds. Because the contents of the mind are wholly internal and
private,45 words are "voluntary signs" having as their "primary or im
mediate signification"^ ideas, serving to communicate them to oth
ers.47 The suggestion is that the goal of language is not to communi
cate directly about extramental things, but to communicate our

42 With the exception of those that arise from the mind's reflection upon
its own activity.
43 "The Understanding seems to me, not to have the least glimmering of
any Ideas, which it doth not receive from one of these two. External Objects
furnish the Mind with the Ideas of sensible qualities, which are all those dif
ferent perceptions they produce in us: And the Mind furnishes the Under
standing with Ideas of its own operations"', Locke, Essay, book 2, chap. 1,
no. 5.
44 ?por since the Things, the Mind contemplates, are none of them, be
sides it self, present to the Understanding, 'tis necessary that something else,
as a Sign or Representation of the thing it considers, should be present to it;
and these are Ideas"-, Locke, Essay, book 4, chap. 21, no. 4.
45 "MAN, though he have great variety of Thoughts, and such, from
which others, as well as himself, might receive Profit and Delight; yet they
are all within his own Breast, invisible and hidden from others, nor can of
themselves be made to appear"; Locke, Essay, book 3, chap. 2., no. 1. Also,
"... the Scene of Ideas that makes one Man's Thoughts, cannot be laid open
to the immediate view of another..."; Locke, Essay, book 4, chap. 21, no. 4.
47 "... because the Scene of Ideas that makes one Man's Thoughts, can
not be laid open to the immediate view of another,... ; Therefore to commu
nicate our Thoughts to one another, as well as record them for our own use,
Signs of our Ideas are also necessary"; Locke, Essay, book 4, chap. 21, no. 4.
internal ideas about extramental things. If the contents of our minds

were accessible to other minds, there would be no need for language,
or so Locke seems to suggest.
Locke's use of "voluntary signs" and his reference to the "pri
mary or immediate signification," of words ought to remind us of
the De interpretation. It certainly suggests some similarity with the
discussion in Aristotle and St. Thomas. It would be a mistake, how
ever, to think he was directly commenting upon or employing the De
interpretation. There is no suggestion of that. Earline J. Ashworth
suggests that he had near contemporary scholastics at Oxford in
mind.48 Still, given the similarity, we might expect Locke to use the
phrase "by the mediation of ideas" to characterize the signification of
external things by words. But it presently appears how substantively
different his account is. According to St. Thomas, for Aristotle general
words signify concepts without mediation, but they can also be said
properly to signify things by the mediation of concepts. I think that
we have in St. Thomas's thought two analogous uses of "signification."
One use applies to the pair [words, passions of the soul]; this involves
a two-term relation. Another use applies to the triple [words, passions
of the soul, things beyond the soul]; this involves a three-term relation.
Since one relation is a two-term relation, while the other is a three
term relation, the application of "signification" and its cognates to
them cannot be univocal, but must be analogous.49
For Locke on the other hand, words properly speaking signify
only the ideas internal to the mind. He says that men give a "secret
reference" to words beyond the ideas of the mind, namely to the ideas
present in the minds of others, and to things external to the mind. So
some sort of signification of things occurs, but he suggests that it is il
licit, writing:
... it is a perverting the use of Words, and brings unavoidable Obscurity

and Confusion into their Signification, whenever we make them stand
for anything, but those Ideas we have in our Minds.50
48 See Earline J. Ashworth, "Locke on Language," Canadian Journal of

Philosophy 14, no. 1 (March 1984): 45-73; and Earline J. Ashworth, "Do
Words Signify Ideas or Things?: The Scholastic Sources of Locke's Theory of
Language," Journal of the History of Philosophy 19 (July 1981): 299.
49 Since the two uses are not unrelated, pure equivocation is not an op
tion.
Properly speaking, words should not signify external things, but only
the internal representations of external things. What is analogous in
St. Thomas is perverse in Locke.
Hume in large measure follows Locke in his account of ideas as
the internal objects of knowledge.51 He differs from him in drawing
out the implications more rigorously, casting doubt upon whether
these internal objects really are representations of external objects.
For him, ". . . no beings are ever present to the mind except percep
tions."52 He writes:
[t]he most vulgar philosophy informs us that no external object can

make itself known to the mind immediately, and without the interposi
tion of an image or perception. That table, which just now appears to
me, is only a perception, and all its qualities are qualities of a percep
tion.53
Our conviction that external objects resemble these internal represen

tations arises because the internal representations are a necessary
condition for describing external objects.54
However, Hume criticizes55 Locke's conviction that there are ex
ternal objects represented by our resembling perceptions. If all we
can actually know are internal sense impressions and less vibrant
more attenuated ideas, no sense can be made of the claim that objects
exist external to the mind that cause resembling and representative
51 However, Hume chided Locke for applying the term "idea" to all the
mind's perceptions, writing that Locke "perverted" the term from its original
sense. However the dispute is a terminological one, as Hume specifies that
ideas are a species of perception less vibrant and forceful though similar to
the original perceptions that are impressions of sensation.
52 David Hume, A Treatise of Human Nature, ed. Lewis A. Selbe-Bigge
(Oxford: Clarendon Press, 1978), book 1, part 4, sec. 2, p. 212.
53 Hume, Treatise, book 1, part 4, sec. 5, p. 239.
54 "We never can conceive anything but perceptions, and therefore must
make every thing resemble them"; Hume, Treatise, book 1, part 4, sec. 2, p.
216.
55 This is a criticism reminiscent of Berkeley. "But philosophers having
plainly seen that the immediate objects of perception do not exist without
the mind, they in some degree corrected the mistake of the vulgar, but at the
same time run into another which seems no less absurd, to wit, that there are
certain objects really existing without the mind, or having a subsistence dis
tinct from being perceived, of which our ideas are only images or resem
blances, imprinted by those objects on the mind"; George Berkeley, Princi
ples of Human Knowledge (London: Penguin Books, 1988), part 1, no. 56.
perceptions in minds.56 Further, the very notion of an external object

resembling our internal representations is the "monstrous offspring"57
of a philosophical system.58 Reminiscent of Locke's dark room,
though without the connotation of a luminous external world, he com
pares the mind to a stage: "The mind is a kind of theatre, where sev
eral perceptions successively make their appearance; pass, repass,
glide away, and mingle in an infinite variety of postures and situations.
.. ."59 It seems a mere fa?on de parler for Hume to continue to refer to
the contents of the mind as representations of external objects, or as
referring to external objects, for we have no good reason for thinking
that they represent or refer to anything external. It seems that at best
language is only about the perceptions that appear in the theater of
the mind.
So at least historically Russell's description finds real correlates,
and his worries about the inevitability of skepticism find justification
in Hume. To this point, however, little outside of Locke has been said
about the meaning and reference of words. Most of the discussion has
centered on the external objectivity of knowledge, not language. Yet
it is the yoking of language to this account of mind and its objects, as
well as its suspected patrimony in Aristotle, that our contemporaries
find so problematic. The turn to language in the last hundred years,
and what it can be understood in part to have been a turning from, has
to be understood in light of this so-called classical representational
ism.
It is important to recognize that the criticism of classical repre
sentationalism in this century, though in large measure influenced by
56 "... as no beings are ever present to the mind but perceptions, it fol
lows that we may observe a conjunction or a relation of cause and effect be
tween different perceptions, but can never observe it between perceptions
and objects. It is impossible, therefore, that from the existence or any of the
qualities of the former, we can ever form any conclusion concerning the ex
istence of the latter, or ever satisfy our reason in this particular"; Hume,
Treatise, book 1, part 4, sec. 2, p. 212.
"As to those impressions, which arise from the sense, their ultimate
cause is, ... , perfectly inexplicable by human reason, and it will always be
impossible to decide with certainty, whether they arise immediately from the
object, or are produced by the creative power of the mind, or are derived
from the Author of our being"; Hume, Treatise, book 1, part 3, sec. 5, p. 84.
58 The first book of the Treatise is in large measure devoted to carrying
through this critical analysis for the impressions and ideas of substance,
cause, self, mind, and personal identity.
Wittgenstein, did not begin with him. There are certain foreshadow
ings of his criticism in others, especially Husserl and Frege. Near the
turn of the century both of them, concerned for the objectivity of
truth, meaning, and the reference of language, reacted strongly against
classical representationalism and the identification of the meaning of
terms with internal mental representations. Besides the classical em
piricists, both of them had in mind their latter-day descendants, espe
cially the psychologism of John Stuart Mill in his Logic. Husserl
writes:
If I hear the name 'Bismark' it makes not the slightest difference to my
understanding of the word's unified meaning, whether I imagine the
great man in a felt hat or coat, or in a cuirassier's uniform, or whatever
pictorial representation I may adopt. It is not even of importance
whether any imagery serves to illustrate my consciousness of meaning,
or to enliven it less directly.60
He devotes a large section of Investigation 2 to the analysis and cri

tique of Locke, Berkeley, and Hume on these questions of the repre
sentational character of thought. He criticized them for making ideas
interior objects of perception. Noting that in Locke, such interior pre
sentations are usually intentional experiences, that is, "each idea is an
idea of something, it presents something," he criticizes him for con
fusing the "presentation with what is as such presented, appearance
with what appears, the act or act-phenomenon, as a really immanent
element in the stream of consciousness, with the object intended."61
One can see in this passage from Husserl a parallel with Russell's crit
icism. Something, an "act or act-phenomenon" is intruding itself in be
tween the mind and the object. Further, with regard to universal
terms, he writes, "naming by means of universal names does not,. .. ,
consist in picking out particular universal ideas from such complexes
of ideas, and attaching them to words as their 'meanings'."62 In a pas
60 Edmund Husserl, Logical Investigations, trans. John N. Findlay (New

York: Humanities Press, 1970), book 1, chap. 4, no. 30.
61 Husserl, Logical Investigations, book 2, chap. 2, no. 10.
62 Ibid. Also, "now we completely see the deceptive confusions in
Locke's train of thought. From the obvious truth that each universal name
has its own peculiar universal meaning, he passes on to assert that a general
idea corresponds to every general name, which idea is for him simply a sepa
rate intuitive presentation"-, and "we tend naturally to turn our gaze among
logical phenomena to whatever has primary intuitive palpability; we are then
misled into taking the inner pictures which are found to accompany our
names as the meanings of those names. If we become clear, however, that a
meaning is merely what we mean, or what we understand by an expression,
we cannot maintain such a conception"; Husserl, Logical Investigations,
book 2, chap. 3, no. 15.
sage that foreshadows the problems we will see in Wittgenstein, he

writes of Hume:
... the same individual idea fits into many circles of similarity, though in
each definite thought-context it only represents ideas from one such cir
cle. What circumstance picks out this circle of representation in a given
context, what limits the representative function of the individual idea in
this manner and so makes unity of sense possible?63
For Husserl, the fundamental problem in the "historical treatment of

abstraction," or general representation, is that "the sign did duty for
the thing signified." The sign stands in place of or substitutes for
what is intended because the latter is not present. We saw earlier that
contemporary Husserlian commentators like Robert Sokolowski con
tinue this criticism. Continuing the quote from Sokolowski that I cited
in the beginning, namely that "in a Lockean spirit we have allowed
words to range only over the domain of our ideas, and we have tacitly
taken ideas to be some sort of internal things," Sokolowski adds, "an
idea is not an internal entity and in an important sense it is really not
other to the thing of which it is the idea."64
When we turn to Frege, we find that he wishes "to separate
sharply the psychological from the logical, the subjective from the ob
jective."65 He writes:
What is a content of my consciousness, my idea, should be sharply dis
tinguished from what is an object of my thought. Therefore the thesis
that only what belongs to the content of my consciousness can be the
object of my awareness, of my thought, is false.66
It was in part the subjective and private character of mental states that
led Frege to posit his "third realm" of thoughts to account for the "ob
jective" meaning of language. These objections and the common con
cern with the skepticism perceived to be attendant upon mental
63 Husserl, Logical Investigations, book 2, chap. 5, no. 34.

64 Sokolowski, "Exorcising," 456. See also Barry Smith, "Frege and Hus
serl: The Ontology of Reference," Journal of the British Society for Phenom
enology 9, no. 2 (May 1978): 111-25.
65 Gottlob Frege, introduction to Foundations of Arithmetic, trans. John
L. Austin (Evanston: Northwestern University Press, 1980), p. x.
66 Gottlob Frege, "The Thought: A Logical Inquiry," Mind 65, no. 259
(July, 1956): 306. Trans. Anthony M. and Marcell Quinton.
representations, form a historical point of intersection between Hus

serl, Frege, and their heirs.67
Wittgenstein in the Philosophical Investigations submitted clas
sical representationalism68 to criticism for its picturing and its inter
nalism.69 He criticized the thesis that the meaningfulness of language
can be accounted for by internal representations conceived of as im
ages of the objects talked about, in conjunction with meaning-confer
ring acts that correlate or assign the representations to external
sounds or written symbols.70 The criticism in the Investigations of
the pictorial determination of reference finds its context in the Trac
tatus's picture theory of sentence truth,71 but also in the Fregean the
67 For a discussion of the original interaction and common concerns of

Husserl and Frege, see J. Mohanty, Husserl and Frege (Bloomington: Indiana
University Press, 1982); Herbert Spiegelberg, The Phenomenological Move
ment (The Hague: Martinus Nijhoff, 1982), 84-100; Barry Smith, "On the Ori
gins of Analytic Philosophy," Graezer Philosophische Studien 35 (1985),
153-73; Michael Dummett, Origins of Analytic Philosophy (Cambridge: Har
vard University Press, 1994); Robert Sokolowski, "Husserl and Frege," Jour
nal of Philosophy 84, no. 10 (1987): 521-8; Smith, "Frege and Husserl"; Rich
ard Cobb-Stevens, Husserl and Analytic Philosophy (Dordrecht: Kluwer
Academic Publishers, 1990). This last is particularly interesting for the stress
it places upon seeing Husserl as continuing the Aristotelian tradition.
68 He does not, however, address by name Descartes, Locke, Berkeley,
and Hume, and the Investigations are by no means exhausted by the treat
ment of representationalism.
69 In what follows I am partially indebted to Haldane's "Life of Signs."
70 ". . . nothing is more wrong-headed than calling meaning a mental ac
tivity! Unless, that is, one is setting out to produce confusion"; Ludwig Witt
genstein, Philosophical Investigations, trans. G. E. M. Anscombe (New York:
Macmillan Publishing Co., 1958), no. 693.
71 "At the first glance the proposition [Satz]?say as it stands printed on
paper?does not seem to be a picture of the reality of which it treats. But nor
does the musical score appear at first sight to be a picture of a musical piece;
nor does our phonetic spelling (letters) seem to be a picture of our spoken
language. And yet these symbolisms prove to be pictures?even in the ordi
nary sense of the word?of what they represent"; Ludwig Wittgenstein, Trac
tatus Logico-Philosophicus, trans. Charles K. Ogden (London: Routledge &
Kegan Paul Ltd., 1986), no. 4.011. Emphasis added.
It is important to recognize that according to Wittgenstein in the Tracta
tus, sentences are not pictures of objects, but rather of facts [no. 2.1], which
facts are themselves composed of states of affairs. Even more it is impor
tant to recognize that he did not conceive of the picturing relation generally
along sensual lines, despite his phrase "even in the ordinary sense of the
word [pictures]." A sentence pictures a fact because it shares the very same
logical structure as the fact, not because its sensual appearance is like the
sensual appearance of the fact it pictures?it pictures the fact's logical struc
ture. For more on the Tractatus and picturing, see Justus Hartnack, Wittgen
stein and Modern Philosophy (Notre Dame: University of Notre Dame Press,
1986); and G. E. M. Anscombe, An Introduction to Wittgenstein's Tractatus
(Philadelphia: University of Pennsylvania Press, 1971): chaps. 3-5.
sis that the sense of a term is "the mode of presentation" of its refer
ent,72 or as it is more commonly said these days, "sense determines
reference."73 If sense is to determine reference, and we identify the
sense or meaning of a term with an internal mental picture presenting
to the mind an external object, then what is it about pictorial represen
tation that is sufficient for performing this function? How is it that we
know an object is referred to from an examination of its internal men
tal representation? Wittgenstein writes:
What is essential is to see that the same thing can come before our
minds when we hear the word and the application still be different. Has
it the same meaning both times? I think we shall say not.74
Suppose a Renaissance painter has his wife and baby pose as models
for a painting of the Madonna and child. The painter's intention is that
it represent Mary and the child Jesus, but anyone who knows his wife
and child would likely take it to be representing the latter, not Mary
and Jesus. Nothing about the painting as such determines unambigu
ously what it represents. If it is replied that it naturally represents his
wife and child by visual resemblance, while conventionally and sym
bolically representing Mary and Jesus, then we can still ask whether
by an examination of the picture alone one could determine unambig
uously that it is of this particular woman and child, rather than say her
closely resembling sister and child. We need only consider Hume's
point about the universal character of ideas, namely that:
Abstract ideas are, therefore, in themselves individual, however they
may become general in their representation. The image in the mind is
only that of a particular object, though the application of it in our rea
soning be the same as if it were universal.75
Perhaps the picture is not intended to represent any particular mother

and child, but all of them. How can we determine this?
72 Gottlob Frege, "On Sense and Reference," in Translations from the

Philosophical Writings of Gottlob Frege, ed. Peter Geach and Max Black
(Oxford: Basil Blackwell, 1970), no. 26.
731 have not been able to find Frege actually using the phrase "sense de
termines reference," though admittedly I am making use of translations.
Those who use this phrase in the context of Frege typically cite the passage
in which he says that the sense of a term is the mode of presentation of its
reference. See for example Dummett, Frege, chap. 5, "Sense and Reference."
74 Wittgenstein, Investigations, no. 140.
75 Hume, Treatise, book 1, part 1, sec. 7, p. 20. Compare John Haldane,
who writes, "the representational properties of images are either particular,
in which case generality is left unexplained, or else they are general and sin
gular reference remains unaccounted for"; Haldane, "Life of Signs," 459.
It seems that the painter determines the object represented by the

painting, in some sense already knowing what he wants to represent;
or perhaps the observer determines what is represented in view of his
own context and concerns, and irrespective of the intent of the
painter?"only when one knows the story does one know the signifi
cance of the picture."76 Outside of the context of use, representations
are, in a sense, faceless icons. However, when the picture is made
into an internal mental image, the difficulties are compounded. Who
is the painter or observer establishing the referential properties of the
internal presentation, determining it to an object? Not the mind acting
on the image, for by supposition the mind does not have independent
knowledge of the reference by which to fix the representative charac
ter of the image, but seeks to determine the reference from the image.
In fact, one line of criticism of the thesis is that it leads to an infinite
regress of homunculi conferring reference on internal images.77
Wittgenstein's criticism centered upon recognizing that represen
tations likened to pictorial resemblances require a context of interpre
tation in order to represent determinate objects, which internal men
tal images must do, if they are going to determine reference. Yet "any
interpretation still hangs in the air along with what it interprets, and
cannot give it any support. Interpretations by themselves do not de

77 "We cannot explain Intentionality in terms of representation because
in order for there to be a representation there must be some agent who uses
some entity?a picture or a sentence or some other object?as a representa
tion. Thus if a belief is a representation it must be because some agent uses
the belief as a representation. But this offers us no account whatever of be
lief because we are not told what the agent does in order to use his belief as a
representation, and furthermore the theory requires a mysterious homuncu
lus with its own Intentionality in order that it can use the beliefs as represen
tations; and if we follow this through it will require an infinite regress of ho
munculi, since each homunculus has to have further Intentional states in
order to use the original Intentional states as representations or indeed in or
der to do anything"; Daniel Dennett, Brainstorms (Montgomery, Vermont:
Bradford Books, 1978), 122-5, as quoted in John Searle, Intentionality: An
Essay in the Philosophy of Mind (Cambridge: Cambridge University Press,
1988), 21. Searle remarks that "Dennett, who thinks this is a genuine prob
lem, calls this 'Hume's problem' and believes that the solution is to postulate
whole armies of progressively stupider homunculi!" See in general chap. 1:
"The Nature of Intentional States," esp. 13-14, 21-2. See also Searle's more
recent The Rediscovery of Mind (Cambridge: MIT Press, 1994), 212-14. Of
course Searle is speaking in the idiom of "intentionality." See, however,
Haidane's "Life of Signs," 459, using the idiom of reference, and also John Mc
Dowell, "Meaning and Intentionality in Wittgenstein's Later Philosophy,"
Midwest Studies in Philosophy 17 (1992): 40-52, esp. 42.
termine meaning."78 What is required is a custom and context of use79

to determine the meaning of linguistic characters, which meaning in
turn determines the interpretation of the images that may or may not
accompany speech. According to Wittgenstein, what is clear is that
"when I think in language, there aren't 'meanings' going through my
mind in addition to the verbal expressions: the language is itself the
vehicle of thought."80
IV
Putnam's Criticism. This historical setting provides a context

for understanding Putnam's criticism of the "Aristotelian view." By
now we find familiar his paraphrase:
the picture is that there is something in the mind that picks out the ob
jects in the environment that we talk about. When such a something
(call it a "concept") is associated with a sign, it becomes the meaning of
the sign.81
For Putnam it is clear that the central difficulty is raised by the

Fregean thesis adapted to representationalism: that the mental repre
sentation determines what the word refers to, if anything. In his arti
cle "The Meaning of 'Meaning'," which is the locus classicus of the
fundamental elements of his criticism, this element is key, along with
the account of what he calls methodological solipsism. Methodologi
cal solipsism is the thesis that "no psychological state, properly so
called, presupposes the existence of any individual other than the sub
ject to whom that state is ascribed."82 The Fregean thesis then re
quires that the representations that fall within the scope of method
ological solipsism must still function to "pick out" the objects in the
world thought and talked about. By a knowledge of the mental

79 For a discussion of the inadequacies of identifying "meaning" with ac
tual use in accord with established rules of use, and nothing more, see John
Haldane, "Life of Signs", esp. part 3: ". . . all such rules tell us is how to pro
ceed in [a language]; they advert to a usage but do not explain how it is possi
ble for the usage itself to exist. To explain that "cat," say, can refer to a cat
because people use "cat" to refer to cats, does not yet explain how this or any
other sign can be invested with significance"; 454.
81 Putnam, Representation and Reality, 19.
82 Putnam, "The Meaning of 'Meaning,'" 220.
representation identified with the meaning of a term, the Aristotelian

is supposed to know the referent of that term.
Putnam's criticism is so broadly conceived in scope that it can be
difficult to determine just what the specific target is. However, I think
several theses can be picked out as essential to it, theses that in vari
ous ways can be seen to be at play in the classical modern tradition,
and Wittgenstein's criticism of it. The first thesis is that there are
things or objects in the mind,83 which things or objects are pictures,
images, and appearances of things outside the mind. So in addition to
the mind and external things, there is a third realm of mental things
that are "interposed" between the mind and external things. I will call
this the third-thing thesis. The second thesis is that the mind in its
conscious activity of thinking attends to these internal objects as what
it primarily knows. Comparing so-called Platonistic versions with the
Aristotelian, Putnam indicates that what they share in common is that:
. . . speakers are supposed to be able to direct their mental attention to

concepts by means of something akin to perception, and, if A and B are
different concepts, then attending to A and attending to B are different
mental states. . . . The mental state of the speaker determines which
concept he is attending to, and thereby determines what it is he refers
to.84
The difference for Putnam between the Platonist and the Aristotelian
is that the Aristotelian's concepts are internal mental objects. Else
where, he characterizes a mental object as "something introspectible."
So I will refer to this as the introspectibility thesis. Finally, there is
the thesis that there is no intrinsic or necessary relation between the
so-called mental representations in the mind, and the represented
things outside it. He writes:
Even a large and complex system of representations, both verbal and vi
sual, still does not have an intrinsic, built-in,... connection with what
it represents?a connection independent of how it was caused and what
the dispositions of the speaker or thinker are.85
For Putnam, it betrays a "survival of magical thinking"86 to suggest

that there is some intrinsic relation between mental representations
and what they represent. Colin McGinn, in Mental Content?1 stresses
83 Putnam thinks that they are events or occurrences in the mind.

^Putnam, Representation and Reality, chap. 2, note 1, p. 129.
85 Hilary Putnam, Reason, Truth, and History (Cambridge: Cambridge
University Press, 1981), 5.
86 Putnam, Reason, Truth, and History, 2.
87 Colin McGinn, Mental Content (Oxford: Basil Blackwell, 1989).
the importance of this element of the background that Putnam as

sumes and wishes to criticize; the internal things are thought to be in
dividuated solely by facts internal to the mind. This specifically incor
porates the assumption of methodological solipsism. Following
McGinn, I will call this the internalist thesis.
Putnam does not often formally argue against these theses,88 but
provides counterexamples and imaginative thought experiments, pro
ceeding in Wittgensteinian fashion89 to argue that mental representa
tions of whatever sort do not determine reference. Perhaps his most
famous counter example is the idea of "Twin Earth."90 The burden of
the Twin Earth thought experiment is to indicate how two different
persons or communities of persons, one on earth and one on a ficti
tious Twin Earth, could have qualitatively identical mental representa
tions associated with a word, yet with that same word refer to distinct
natural kinds. So for instance on Twin Earth, substance XYZ may
look, smell, feel, and taste just like our earth's H20. When our Twin
Earth cousins use the term "water," however, they are not referring to
H20, as we do when we use the term, but to XYZ. So the mental repre
sentation, which ex hypothesi is identical on the two earths, cannot be
determining the reference of the term. As McGinn points out, if the
mental representation were determining the reference, then varying
88 This is not strictly accurate. He will make reference to his use of the
Lowenheim-Skolem theorem, which is intended to formally prove the impos
sibility of another thesis that he calls metaphysical realism. Of course, in or
der for his proof to have the kind of generality he wants, he assumes, and
does not show, that any system of mental representation is a formal theory in
the sense required by the theorem. It would remain open for an opponent
simply to deny that a particular system of representation is a theory in the
sense required by the theorem, if it can be thought to be a theory at all.
Though this point about the sense of "theory" required for use of the theorem
occurred to me independently and prior to my reading of the following arti
cle, see Michael Hallett, "Putnam and the Skolem Paradox," in Reading Put
nam, ed. Peter Clark and Bob Hale (Oxford: Basil Blackwell, 1994), 66-97.
Hallett provides a detailed and in-depth discussion, as well as a bibliography,
of Putnam's use of the Theorem.
89 He writes, "none of the methods of representation that we know about
has the property that the representations intrinsically refer to whatever it is
that they are used to refer to"; Putnam, Representation and Reality, 21.
90 ?Twin Earth" made its first appearance in Putnam's article "The Mean
ing of 'Meaning'." This article has recently been reprinted in The Twin Earth
Chronicles: Twenty Years of Reflection on Hilary Putnam's "The Meaning
of Meaning'," ed. Andrew Pessing and Sanford Goldberg (Armonk: M. E.
Sharpe, 1996). This is a collection of some of the most prominent discussions
of Putnam's piece that have appeared over the years.
everything in the environment except the mental representation

should not vary the reference.
Putnam provides variations on this example,91 but what is com
mon is that the qualitative identity of the mental representations is put
in terms of the identity of descriptions predominantly involving sense
experience, and the ability on the basis of sense experience to distin
guish distinct external objects. The sense experience so described,
however, is itself predominantly understood along the lines of appear
ances internal to the mind; we talk about water that exists external to
the mind, but our experience of water consists of appearances that ex
ist internal to the mind. His claim requires that on the Aristotelian
view, we can be talking about two distinct things or kinds of things,
while the internal appearances that determine what we are talking
about are qualitatively identical to our introspection. So, in our use of
words, consciously attending to these internal mental representations
cannot be determining what we are talking about. The qualitative
identity of the appearances embodies the assumption that the mental
representation, in its representational aspect, is not intrinsically re
lated to what it represents outside the mind, that is, the assumption of
methodological solipsism. For if it were so related, one could not vary
the environment in the manner necessary for the thought experiment,
and have the representation remain the same.
Putnam believes the representationalist is committed to this di
chotomy between internal appearances and external reality. The Aris
totelian pictures the language user as holding before his mind an inter
nal appearance; by knowing this internal appearance the language
user is supposed to know the external thing pictured by this appear
ance, and so know the thing talked about. There is some "third thing"
in addition to the language user and the external thing talked about.
The third thing, the mental representation, "intrudes itself... between
the mind and the object," as Russell says. In fact, it is difficult while
reading Putnam's analysis not to think of Locke's various treatments
of real and nominal essences in the Essay.
There are then three crucial elements to Putnam's argument by
counterexample, corresponding to the three theses that we outlined
above. First, corresponding to the third-thing thesis, there is a realm
91 The example in Twin Earth is the term "water," and the substances
H20 and XYZ, but as Putnam indicates, it could be real world examples in
volving gold and brass, and the term "gold." Also his "brain in a vat" can be
taken to be a variant. The latter is in a way an updating of Descartes' evil de
mon thought experiment.
of internal appearances in addition to the mind and external things.

Second, corresponding to the introspectibility thesis, we know what
we are talking about because of our internal and introspective knowl
edge of these appearances that represent the external objects. Third,
corresponding to the internalism thesis, the conditions of identity of
the appearances do not include the external things represented, but
only facts entirely internal to the mind, or in Putnam's terminology,
solely by "theoretical and operational constraints." Each of these the
ses can be seen to be at play in one way or another in Locke and
Hume, and their metaphors of the mind being a "dark closet" or "the
ater of appearances" are particularly apt. Further, Putnam's criticism
appears to be a fanciful updating of Wittgenstein's.
If we try to make the picture fit more explicitly St. Thomas's un
derstanding of Aristotle described at the beginning of the article, then
the third thing appears to be the concept, existing inside the soul. The
concept has no intrinsic and necessary relation to things outside the
soul. Furthermore, because words immediately signify concepts, and
mediately signify extramental things, it is by knowing the concepts as
primary objects of knowledge, holding them before the mind's con
scious attention, that the language user knows what extramental ob
jects are talked about. Now, in light of these three theses, we may
move on to determine whether St. Thomas's account is subject to the
criticism directed at the union of language and mental representation
alism in modern and contemporary contexts.
Cutting at the Joints of the Three Theses: The Third-Thing The

sis. Recall that the third-thing thesis asserts that there is some inter
nal mental thing (res) that is interposed between the mind and th
world. John Haldane, summarizing the difficulty, indicates that ther
isa:
common tendency to assume that the existence of a process implies the
existence of a product distinct from it. More particularly, to suppose
that talk of forming a thought, or of making a representation, implies the
manufacture of a third thing to which consciousness is then directed.92
92 John Haldane, "Brentano's Problem," Graezer Philosophische Studien

35 (1989): 25. Though not always agreeing with them, in what follows I am in
debted to Haldane and Searle in Intentionality for clearing away many obsta
cles in my own thought on this point.
Is there such a tertium quid manufactured and interposed between

intellect and world? For St. Thomas, is the concept a thing (res) other
than the intellect that knows,93 and other than the thing (res) extra an
imam that is known? Sokolowski, for the most part confining his crit
icism to modern trends, thinks that an affirmative answer to this ques
tion is philosophically naive; "thing" is not properly applied to
concepts. The answer to the question, how many things are essen
tially involved, ought to be two, knowing thing and known thing, not
three.
In the account I have given of St. Thomas's thought, a concept is
the informed activity of the intellect, as it grasps things beyond the
soul (res extra animam). The intellect is simply the human sub
stance's capacity to engage in just that act, that is, the concept, as I am
using it, is to the intellect, as act to potency. There is a process-prod
uct ambiguity when describing this informed activity of the intellect.
This ambiguity is not, however, peculiar to the intellect's act. Often
times in describing an act, we are correct in our estimation that, in ad
dition to the act, something other than it is produced. So for instance,
we normally think that the product of the pitcher's act of pitching re
sults in the ball having a certain trajectory, which we call the pitch.
On the other hand, there are activities we engage in that involve no
such additional product, though the language we use might suggest
otherwise. We might say, "Sebastian strolled in Bologna." Alterna
tively, we might say, "Sebastian took a stroll in Bologna." Do we sug
gest by the latter that there is some additional thing or object that Se
bastian took, to accompany him while strolling? If we do, how does it
relate to other predicates with modifying phrases that we normally
find appropriate in descriptions of things taken? Did he take the stroll
off the shelf? Did the stroll belong to someone else, so that he did not
take a stroll, but stole a stroll? Did he take the stroll in hand, perhaps
to pitch woo at it in the moonlight, under the starlight? No. Or sup
pose we say, "Marat took a fateful bath." Here there are other things
involved?the tub, the soap, the water, and so forth. Yet none of these
is the thing that presumably we would have in mind when we say, "Ma
93 Of course, strictly speaking it is the intellectual substance that knows,

by means of his intellect. See Aquinas, CD A, L. 3, lectio 7, no. 690: "Manifes
tum est enim quod hie homo intelligit_intellectus igitur possibilis est, quo
hie homo, formaliter loquendo, intelligit." See also Aquinas, S.T., I?II, q. 17, a.
5, ad 2.
rat took a fateful bath." It would be entirely inappropriate to respond,

"Oh, wasn't it too heavy for him to take alone?"
These examples suggest that something is awry. Clearly we are
not using "take" in any literal sense when we speak of "taking a stroll,"
or say, "Marat took a fateful bath." At best, "to take a stroll" or "to
take a bath" appear to be elliptical ways of saying "to stroll," or "to
bathe," which by their forms do not indicate additional things that are
"taken." One way of seeing this is to recognize that the conjunction,
"Sebastian strolled in Bologna and Sebastian took a stroll in Bologna,"
is no more informative than the simple "Sebastian strolled in Bolo
gna"; it is like asserting "p & p," rather than just "p." Nominalization of
verbs into substantives is a way of talking about our activities, not a
way of recognizing another realm of things in addition to our activi
ties. "We are up against one of the great sources of philosophical be
wilderment: a substantive makes us look for a thing that corresponds
to it."94
This point is not particularly surprising, or insightful. The inter
est in these examples lies in the context they provide when we turn to
consider concepts. In fact the difference between the examples is the
difference for St. Thomas between a transient act and an immanent
act. In a transient act, the act has its termination in some being other
than the agent, completing or perfecting that other being. In an imma
nent act, the act has its termination in some sense within the agent,
completing or perfecting it. When we employ words like "grasp," "ap
prehend," "employ," "produce," "develop," and so forth, as in "he
grasped the concept," or "by speaking with them, we develop in chil
dren the concepts necessary to function linguistically," should our use
of such words be analyzed along the lines of the example, "He pitched
the ball," or rather like the examples, "He took a stroll," and "He took a
fateful bath"?
I think that the answer is clearly along the lines of the stroll and
the bath. "Concept" is a nominalized form of talking about our act of
conceiving, not a way of referring to an additional class or category of
objects or things, in addition to our acts. Strictly speaking, if we retain
the tactile word "grasp,"95 we do not grasp concepts. We conceive,
94 Ludwig Wittgenstein, The Blue and Brown Books: the Blue Book (New
York: Harper & Row, 1965), 1.
95 "It follows that the soul is analogous to the hand; for as the hand is a
tool of tools, so the mind is form of forms and sense the form of sensible
things"; Aristotle, De anima, in The Basic Works of Aristotle, ed. Richard
McKeon (New York: Random House, 1941), 432al-3.
and in our conceiving we grasp things other than our conceiving. St.
Thomas writes, "someone has a form in act inasmuch as he is able to
execute an operation of that form."96 Our conceiving is informed by
the forms of things, and the activity differs thereby, just as Sebastian's
bodily movements may have the form of a stroll, or the form of a run,
or may be informed more or less by speed or grace. His activity will
differ thereby, as Marat's bath may be more or less warm, mundane,
medicinal, or fateful. Consider another analogy. If I grasp the pen in
my hand, then we can certainly say that my grasp is of a pen, and not a
ball. In the antecedent of this conditional, "grasp" functions as a verb,
while in the consequent "grasp" functions as a noun. Do I need, then,
to posit a third thing between the pen and my grasping hand, namely
the grasp itself, a third thing named by the use of "grasp" as a noun? If
I fail to grasp the pen, do I say that I succeeded in grasping the grasp,
but failed to grasp the pen? No. There is no third thing that exists be
tween hand and pen. Similarly, there is no third thing between the
conceiving intellect and res extra animam.
If I grasp with my hand a ball instead of a pen, the form of my
grasp will be different than when I grasp a pen, corresponding to what
is required to grasp a ball, and not a pen. Now my hand, grasping ei
ther a ball or a pen, does not look like either a ball or a pen; nonethe
less, in its relation to the pen that it grasps, it is perfectly appropriate
to call its form "the form of a pen-grasping hand," rather than say "the
form of a ball-grasping hand"; I grasp differently, according as the
form of my grasping must differ for the diversely informed things I
wish to grasp. Similarly, concepts are informed acts that we perform,
usually in order to understand things beyond the soul (res extra ani
mam). Concepts are informed diversely inasmuch as we conceive di
versely informed res extra animam. They are not, however, things
other than the intellect that grasps res extra animam. In Aristotelian
terms they are the acts of the power capable of grasping those things,
the acts of the intellect.
Once we recognize that a concept is related to the intellect as act
to potency, it makes very little sense to treat intellect and concept as if
they were not one, as if the intellect is one thing and a concept another
thing. St. Thomas, in another context, provides interesting insight on
96 "... quamlibet formam tune aliquis in actu habet quando potest Opera
tionen! illius formae explere"; Aquinas, CD A, L. 3, lectio 8, no. 700 (cap. 2,18
19).
the more general question of how act and potency are one and not di
verse. Discussing various erroneous views about what makes soul
and body one, he sets them in a more general setting concerning the
unity of act and potency. He writes:
. . . they held such [erroneous views] because they inquired about what
makes a potency and an act to be one thing, and they sought the differ
ences of these things, as if it were necessary to collect them as through
some single medium, just like things which are diverse according to act.
But as has been said, the ultimate matter, which is appropriate to a
form, and the form itself are the same thing (idem). For one of these is
as potency, while the other is as act... . Potency and act are in a certain
respect one. For that which is in potency is made to be in act. And so it
is not necessary for them to be united through some bond, like those
things which are entirely diverse. So no cause makes those things
which are composed from matter and form to be one, except that which
moves a potency to act.97
Of course the particular point that St. Thomas is pursuing here is the
union of soul and body. Yet, as he often does, he discusses that point
by bringing to bear a larger context, and that context is how act and
potency are not diverse things, but one thing. This is the thesis that in
terests us. There is no question that in his analysis of knowledge, St.
Thomas extends by analogy the pairs form-matter and act-potency,
from their original use in the discussion of change and motion. We
have discussed at length the form informing the intellect. The intel
lect, as such, is compared as matter to the form that informs it. If we
recognize that a concept is compared to the intellect as act to potency,
then the general point that St. Thomas makes in the passage above ap
plies straightforwardly. It is just misguided to treat the concept as if it
were a distinct thing from the intellect, only to ask how this tertium
quid is then related or united to the intellect. We should not be
97 ". . . quod causa quare talia posuerunt, est, quia inquirebant quid fa
ciens unum potentiam et actum, et inquirebant differentias eorum, quasi
oporteret eas colligari per aliquod unum medium, sicut ea quae sunt diversa
secundum actum. Sed sicut dictum est, ultima materia, quae scilicet est ap
propriata ad formam, et ipsa forma, sunt idem. Aliud enim eorum est sicut
potentia, aliud sicut actus. . . . Potentia et actus quodammodo unum sunt.
Quod enim est in potentia, fit in actu. Quod enim est in potentia, fit actu. Et
sic non oportet ea uniri per aliquod vinculum, sicut ea quae sunt penitus di
versa. Unde nulla causa est faciens unum ea quae sunt composita ex materia
et forma, nisi quod movet potentiam in actum"; Sancti Thomae Aquinatis, In
Duodecim Libros Metaphysicorum Aristotelis Expositio, L. 8, lectio 5, no.
1767 (Turin: Marietti, 1950).
looking, as St. Thomas writes, for something to relate concepts to the

intellect, but for the cause that moves the intellect to its act.
VI
The Introspectibility Thesis. Consider the second thesis that

concepts are introspectible primary objects of the mind, the knowl
edge of which secondarily provides knowledge of external objects. As
Putnam puts it:
. . . speakers are supposed to be able to direct their mental attention to
concepts by means of something akin to perception, and, if A and B are
different concepts, then attending to A and attending to B are different
mental states. . . . The mental state of the speaker determines which
concept he is attending to, and thereby determines what it is he refers
to.98
We know what we are talking about because our language is yoked to

this philosophy of mind. Does this thesis cut at the joints of St. Tho
mas's account? No.
We are blessed here because St. Thomas explicitly poses some
thing very much like this problem in several places. In the Summa,
for instance, he asks whether the "intelligible species [the intelligible
form in the intellect] ... is related to our intellect as that which is un
derstood?"99 There he explains that the form informing the intellect's
activity is a means of knowing something else. He relates the discus
sion to the more general account of transitive and immanent activ
ity,100 and the forms or principles by which such activities takes place.
He is adamant when he writes, "it ought to be said that the intelligible
species is related to the intellect as that by which the intellect under
stands."101
However, by itself this is not sufficient to establish the point. For
the introspectibility thesis holds something like this, namely, that in
ternal objects once known become means for knowing whatever ex
ternal objects they represent. Instead it is St. Thomas's denial that
these internal acts are primarily objects of knowledge that allows him
98 Putnam, Representation and Reality, chap. 2, note 1, p. 129.

99Aquinas, S.T., I, q. 85, a. 2.
100 For transitive and immanent activity, see the previous section.
101 "... dicendum est quod species intelligibilis se habet ad intellectum ut
quo intelligit intellectus"; Aquinas, S.T., I, q. 85, a. 2.
to escape the introspectibility thesis and its criticism. Commenting on

Aristotle's De anima, he writes very forcefully that:
it is clear that the intelligible species, by which the intellect is made to

be in act, are not the object of the intellect. For they are not related to
the intellect as what is understood, but as that by which it under
stands.102
He asserts that it is "manifestly false" that "such an [intelligible] spe

cies is that which is understood."
In the Summa Theologiae, he cites two reasons for denying what
I have been calling the introspectibility thesis. First, he appeals to the
objects of science which he takes for granted are things beyond the
soul, not the intellect's own contents. The introspectibility thesis
threatens the objectivity of knowledge, both scientific and everyday.
He uses this argument in the commentary on the De anima as well.
Second, he appeals to the experience of error. If our knowledge has
as its primary object internal appearances of things, then error would
be impossible, for everyone would always judge truly of how things
appear to him or her. Thus echoing Aristotle, he writes:
the error of the ancients would follow, who said that everything is true as
it appears..., and so ... contradictories would at the same time be true..
.. and so it would follow that every opinion will be equally true_103
Far from leading to skepticism in the sense of "nothing can be known

to be true," as we might expect when the objectivity of knowledge is
threatened, instead the thesis leads in the other direction, to a sort of
hyper-anti-skepticism?nothing at all can be known to be false.104 It is
clear in St. Thomas's account that the passiones animae referred to in
the De interpretation are not things known first, or, as Putnam puts
102 Aquinas, CDA, L. 3, lectio 8, no.718, esp., "Manifestum est etiam, quod
species intelligibiles, quibus intellectus possibilis fit in actu, non sunt objec
tion! intellectus. Non enim se habent ad intellectum sicut quod intelligitur,
sed sicut quo intelligit."
103 "Secundo, quia sequeretur error antiquorum dicentium quod omne
quod videtur est verum; et sic quod contradictoriae essent simul verae. Si
enim potentia non cognoscit nisi propriam passionem, de ea solum judicat.
Sic autem videtur aliquid secundum quod potentia cognoscitiva afficitur.
Semper ergo judicium potentiae cognoscitivae erit de eo quod judicat, scil
icet de propria passione, secundum quod est; et ita omne judicium erit
verum;" Aquinas, S. T., I, q. 85, a. 2.
104 Of course, one might argue that if nothing at all can be known to be
false, nothing at all can really be known to be true. St. Thomas, however,
chooses to stress the former.
it, objects to which conscious attention is directed, in order that res

extra animam may subsequently be known.
On the other hand, introspective self-knowledge of one's activity
of understanding is possible. The intellect is capable by introspection
of coming to understand the means by which it primarily knows exter
nal things. St. Thomas writes:
... because the intellect reflects upon itself, it understands according to

that reflection its own understanding and the species by which it under
stands. And so the intellectual species is secondarily that which is un
derstood. But that which is understood first is the [external] thing of
which the intelligible species is a likeness.105
This parallels very closely the discussion of the same issues in the
commentary on the De anima at L. 3, lectiones 8 and 9, where he
writes:
what the intellect understands is an essence which is in things, but not
the intelligible species, except insofar as the intellect itself reflects upon
itself,
and
the possible intellect, which is as such in potency in the order of intelli
gible things, neither understands nor is understood except through a
species taken into it.106
For St. Thomas, things beyond the soul are known, and only subse
quently are the passiones animae known. First the passiones ani
mae are means of knowing things, and only secondarily are they them
selves objects of knowledge. This accords well with St. Thomas's
often repeated axiom that a thing is known only insofar as it is in act.
The act of the passio animae in question here, a concept, is simply to
be the understanding of some object extra animam. The passio ani
mae can be known only insofar as it is a means of knowing something
else.
In the criticism of mental representation, however, the order is
reversed. In that criticism, it is presupposed that mental impressions
105 "Sed quia intellectus supra seipsum reflectitur, secundum eandem re

flexionem intelligit et suum intelligere et speciem qua intelligit. Et sic species
intellectiva secundario est id quod intelligitur. Sed id quod intelligitur primo
est res cuius species intelligibilis est similitudo"; Aquinas, S.T., q. 85, a. 2.
106 "... quod intellectus intelligit est quidditas quae est in rebus, non au
tem species intelligibilis, nisi in quantum intellectus super se ipsum reflecti
tur"; and "intellectus possibilis, qui est tantum in potentia in ordine intelligibi
lium, nec intelligit nec intelligitur nisi per speciem in eo susceptam"; Aquinas,
CDA, L. 3, lectio 8, no. 718; and lectio 9, no. 725.
or passiones animae are known first, and things (res) subsequently.

The problem attributed to the mental representationalist is that he be
lieves he must know the means for knowing things, before he can
know those things. St. Thomas denies this. One must know the things
before one can know the means for knowing them. Consider this anal
ogy. A healthy rotator cuff is necessary for good pitching. Yet most
good pitchers learn to throw good pitches without first consciously
knowing anything at all about their rotator cuffs. Indeed most of them
find any knowledge of their rotator cuff superfluous, until they injure
it. People can know things, and happily apply their knowledge in lan
guage or practical activities, without first knowing the means by
which such knowledge takes place. The means are in a sense trans
parent to the agent in pursuit of an end.
This is where evident disanalogies begin to appear with the sort
of examples I used in the discussion of the third-thing thesis. In the
example of the hand grasping the pen, it is possible to know the hand
prior to its grasping the pen. In the case of the intellect, however, St.
Thomas seems to claim that we have no independent prior access to
the intellect, apart from its actually grasping things beyond the soul.
Centuries later John of St. Thomas will make a useful terminolog
ical distinction between what he calls formal and instrumental signs:
a formal sign is an informed knowing, which of itself, not by the media

tion of another, represents. A sign is instrumental, which by itself being
known first as something, represents something other than itself, as for
example the footprint of the cow represents the cow.107
St. Thomas does not himself call concepts "signs." Indeed, I suspect
that applying "sign" to concepts may in the end prove to be unfortu
nate and misleading, to the extent that it proves to be more amenable
to modern accounts of mind, as well as to the criticism directed at
those latter accounts. That, however, is not an argument I am pre
pared to make here. In any case, the troublesome terminology may be
John of St. Thomas's; but the distinction he is trying to capture is St.
Thomas's, for it finds nearly exact parallels in St. Thomas's descrip
tions of how things are known:
107 "Signum formale est formalis notifia, quae seipsa, non mediante alio,
repraesentat. Signum instrumentale est, quod ex praeexistente cognitione
sui aliquid aliud a se repraesentat, sicut vestigium bovis repraesentat bovem";
John Poinsot, Tractatus De Signis, 10al3-17, ed. John Deely (Berkeley: Uni
versity of California Press, 1985).
. . . something is understood in two ways: in one way in itself, when

namely the power of the one observing (acies intuentis) is formed from
the thing itself understood or known; in another way when something is
seen in another, which other having been known, it is known.108
The "first way" of understanding something that St. Thomas mentions

involves what John of St. Thomas calls a "formal sign," the informed
power knowing a thing. The "second way" that St. Thomas mentions
involves what John of St. Thomas calls an "instrumental sign."
John of St. Thomas's terminology is useful because it clears the
way for a direct contrast to be drawn between St. Thomas and more
modern accounts. Using that terminology, concepts for St. Thomas
are formal signs, not instrumental; in St. Thomas's "first way" con
cepts are "the power of the one observing... formed from the thing it
self understood or known." What it is for a concept to be, its essential
character qua act of intellect, is for it to be a means for knowing some
thing else, not for it to be something known. Once something is
known by its means, only then can the concept itself become known.
In the understanding of the British empiricists, and as presupposed by
Putnam's criticism, concepts or ideas are instrumental signs, not for
mal. To the extent that they then exemplify the introspectibility the
sis, they become subject to Putnam's criticism. However, because in
St. Thomas's account the concept signified by an articulated sound is
not itself what is primarily known, but is simply a means by which an
extramental thing is known and signified, he avoids the introspectibil
ity thesis and the part played by it in Putnam's criticism.
Only by first coming to know something other than ourselves
does our capacity for knowing and its exercise come to be knowable
to us; our cognitive acts are introspectible, but they must first be acts.
"In the beginning was the deed." Further, this introspection does not
require a new special faculty, apart from the faculty in virtue of which
we understand res extra animam.
St. Thomas writes:
... that which is primarily cognized by the human intellect is [the nature
of material things]; and secondarily the act itself is cognized by which
the [the nature of material things] is cognized; and through the act the
intellect itself is cognized, of which to understand is the perfection.
108 ". . . dicendum, quod aliquid intelligitur dupliciter. uno modo in

seipso, quando scilicet ex ipsa re intellecta vel cogita formatur acies intuen
tis; alio modo videtur aliquid in altero, quo cognito et illud cognoscitur";
Aquinas, QDV, q. 2, a. 3, ad 5.
And so the Philosopher says that objects are cognized before acts, and
acts before powers.109
The point of the passage is that it is the very same power or faculty,
the human intellect, that knows res extra animam, knows its own
acts, and knows itself. There is no special introspective faculty, a
"mind's eye," nor can the intellect separate itself from itself to make
its own acts objects. It is true that the intellect understands its acts by
further acts: "hence the act by which the intellect understands a stone
is other than the act by which the intellect understands itself under
standing a stone, and so on."110 Still it is the very same faculty that un
derstands a stone and understands itself understanding a stone. Here
we receive a foretaste of what one might call St. Thomas's external
ism which we will examine in the next section. As a condition for
making this distinction between the two acts, understanding a stone
and understanding understanding a stone, St. Thomas can only iden
tify the acts by reference to a stone which is originally understood
and, as a material nature, falls within the primary object of the intel
lect's act.
By our cognitive awareness and attention, we are primarily and
by nature directed to others, not upon ourselves or our concepts. For
St. Thomas, an account of human knowing that begins with introspec
tion is off target. The human person in his or her cognitive capacity is
fundamentally and primarily directed and open to other beings, in par
ticular material beings. Only by the actual exercise of this capacity for
uniting with another does the human person by introspection come to
self-understanding and self-enlightenment. If for Socrates an unexam
ined life is not a life worth living, for St. Thomas an unlived life is not a
life worth examining.
109 "Et ideo id quod primo cogniscitur ab intellectu humano est huius
modi objectum; et secundario cognoscitur objectum; et per actum cognosci
tur ipse intellectus, cuius est perfectio ipsum intelligere. Et ideo Philosophus
dicit quod objecta praecognoscuntur actibus, et actus potentiis;" Aquinas,
S.T., I, q. 87, a. 3, respondeo. I have translated "the nature of material things"
for "objectum," because the passage immediately before the one quoted
makes it clear that "objectum" refers to this.
no "Tjnrje alius est actus quo intellectus intelligit lapidem, et alius est ac
tus quo intelligit se intelligere lapidem, et sic inde"; Aquinas, S.T., q. 87, a. 4,
ad 2.
VII
The Internalist Thesis. Finally, consider the internalist thesis.

According to it there is no intrinsic or necessary relation between the
concept and extramental things; the concept is supposed to be individ
uated by facts solely internal to the mind, embodying the assumption
of methodological solipsism. Does this thesis cut at the joints of St.
Thomas's account? No.
Recall that in discussing the likeness of the passiones animae to
res extra animam, I wrote that a general concept is a natural likeness
of things beyond the soul which are diverse singular beings, because it
shares their form, a form that when considered absolutely does not
differ among the diverse things. I wrote there that we can call the
form, considered absolutely, an intelligible characteristic of the
things. In the intellect such a form or intelligible character constitutes
a principle of knowing things beyond the soul (res extra animam)-, it
informs the act of understanding, and as a principle of the intellect's
act it is called the intelligible species. The act so informed is called
the concept.
Intelligible character and the absolute consideration of a form
hearken back to the discussion of natures absolutely considered (ab
soluta consideratio ipsius) in St. Thomas's early work De ente et es
sent?a,111 composed at Paris between 1252 and 1256. However, the
thesis of interest to us remains with St. Thomas throughout his career.
A nature can be considered absolutely, in which case one considers or
takes into account only those characteristics that pertain to it as such.
In the case of man, being rational and being animal are considered,
111 ". . . secundum naturam et rationem propriam. . . ." S. Thomae

Aquinatis, Opusculum De Ente et Essentia, cap. 4, no. 1 (Turin: Marietti,
1957). The De Ente is an early work, preceding St. Thomas becoming Master
at Paris, dated between 1252 and 1256. See James Weisheipl, Friar Thomas
D'Aquino (Washington: Catholic University of America Press, 1983), 78-9,
386-7. Consider also, "Dicendum, quod, secundum Avicennam in sua Meta
physica, triplex est alicuius naturae consideratio. Una, prout consideratur
secundum esse quod habet in singularibus;... Alia vero est consideratio ali
cuius naturae secundum esse suum intelligibile;... Tertia vero est consider
atio naturae absoluta, prout abstrahlt ab utroque esse; secundum quam con
siderationem consideratur natura lapidis, vel cuiuscumque alterius, quantum
ad ea tantum quae per se competunt tali naturae"; S. Thomae Aquinatis,
Quaestiones Quodlibetales, no. 8, q. 8, a. 1, respondeo (Turin: Marietti, 1949).
See also Aquinas, S.T., I, q. 85, a. 3, ad 1, 2 and 4; and Aquinas, CD A, L. 1, lec
tio 1, no. 13.
but not being here or being there, at this time or at that time, one or
predicable of many, in re or in anima. In the case of red being a
color is considered, but not being the color of an apple, and so forth.112
On the other hand, a nature can be considered with respect to two
ways or modes of existing, one way in singular things (in singulari
bus) as a principle of being, and another way in anima as a principle
of knowing. Some characteristics pertain to the nature in virtue of
these modes of existence, but not as such?being here or being there,
at this time or at that time, one or predicable of many, in re or in an
ima.
If we leave out of consideration the particular characteristics a
concept may possess in its mode of existence in anima, for example,
abstract, predicated universally, and in this intellect now, or in that in
tellect then, it should be clear that a concept is identified with an
other, or individuated and diversified from another by the nature it ex
presses?the intelligible character, which intelligible character is
what it is in virtue of the thing or things for which it is a principle of
being. This is the basis of the natural likeness or similitude between
passiones animae and res extra animam. St. Thomas writes:
the intention of Aristotle is not to assign identity to the conceptions of
the soul by a comparison to articulated sounds, as if for one articulated
sound there should be one concept; because articulated sounds are di
verse among diverse peoples; but he intends to assign the identity of
conceptions of the soul through a comparison to things, which things he
similarly says are the same [for all].113
In this respect, St. Thomas's analysis appears to be a form of external

ism, as described by Colin McGinn in his Mental Content. St. Thomas
112 See also ".. . per modum simplicis et absolutae considerationis; sicut
cum intelligimus unum, nihil considerando de alio. . . . Sed . . . modo abstra
here per intellectum quae non sunt abstracta secundum rem, non habet falsi
taem; ut in sensibilibus manifeste apparet. Si enim intelligamus vel dicamus
colorem non inesse corpori colorato, vel esse separatum ab eo, erit falsitas in
opinione vel in oratione. Si vero consideremus colorem et proprietates eius,
nihil considerantes de pomo colorato, vel quod sic intelligimus, etiam voce
exprimamus; erit absque falsitate opinionis et orationis. Pomum enim non
est de ratione coloris; et ideo nihil prohibet colorem intelligi, nihil intelli
gendo de pomo"; Aquinas, S.T., I, q. 85, a. 1, ad 1.
113 "... intentio Aristotelis non est asserere identitatem conceptionis ani
mae per comparationem ad vocem, ut scilicet unius vocis una sit conceptio:
quia voces sunt diversae apud diversos; sed intendit asserere identitatem
conceptionum animae per comparationem ad res, quas similiter dicit esse
easdem"; Aquinas, CPH, L. 1, lectio 2, no. 21.
"is individuating mental states (to put it intuitively) by reference to

something other than mental states."114 The concepts depend for the
conditions of their formal identity upon the existence of things be
yond the soul. McGinn contrasts externalism with what he calls,
straightforwardly enough, internalism, the thesis "that mental states
are determined by facts relating to the subject considered in isolation
from his environment,"115 that is, the assumption of methodological
solipsism. I mention McGinn's terminology and analysis, not as if it
provides a lens through which to view St. Thomas's thought more
clearly. After all, part of my thesis is that St. Thomas's appropriation
of Aristotle should be read in its own context. McGinn's analysis sim
ply provides a point of intersection where St. Thomas's account can
meet the contemporary objections.
This claim of identity of form or intelligible character between
passiones animae and res extra animam is a difficult thesis. To Put
nam, "it makes no sense." He objects:
even if we could somehow make sense of the claim that objects and
events have intrinsic form, there still remains the question of the rela
tion between that form and the form of whatever it is that thinks about
or represents the object. To say that the relation is identity, whether
"identity" be taken literally or metaphorically makes no sense.116
I suspect that Putnam believes this thesis makes no sense because he

thinks the Aristotelian is trying to provide an explanation of knowl
edge, when the latter claims that there is an identity of form between
knower and known, an explanation that is supposed fortuitously to
answer Putnam's difficulties. Here I am going to make a claim that to
some may appear very contentious. The discussion of identity in St.
Thomas is not an explanation of anything; rather, this claim makes ex
plicit what is presupposed in being able to talk about knowing. Once
we have made explicit what is presupposed in knowing, we can go on
to investigate what characteristics the human person must have in or
der to be capable ofthat knowing. In these contexts, St. Thomas often
echoes Aristotle with the principle that powers are known through
their acts, which acts are in turn known through their objects.
Recall from the previous section that we have no cognitive access
to the intellect, a human power, apart from reflection upon the exer
cise of its acts of understanding. One way to understand the identity
114 McGinn, Mental Content, 3.

115 Ibid., 2.
116 Putnam, "Aristotle After Wittgenstein," 126.
claim is to consider what remains to characterize an act of under

standing some object X, if we leave out of consideration the character
istics that pertain to it as an act existing in this intellect, at this time, at
this level of generality, and so forth. All that remains is for us to spec
ify of what it is an understanding. For example, suppose we ask "how
does our understanding of the substantiality of a dog differ from our
understanding of the substantiality of a man, leaving out of consider
ation all those attributes that accrue to the acts of understanding as
modes of being in anima?" We will not answer "one is an understand
ing now, the other yesterday," or "one is an understanding in your in
tellect, the other in mine," and so forth. In answering this question, all
that remains is to make reference to those characteristics that pertain
to a dog as such, and those that pertain to a man as such; we might an
swer that the one is an understanding of an animate rational sub
stance, while the other is of an animate but non-rational substance. If
we leave out the intentional phrase "understanding of," these are the
very same characteristics we would list if we were answering the ap
parently different question, "How does the substantiality of a dog dif
fer from that of a man?" We will end up specifying just those charac
teristics of the natures extra animam that pertain to their absolute
consideration. So absolutely considered an act of understanding an
object X does not differ from the object X understood, when the latter
is absolutely considered; absolutely considered, what it is for an act
of understanding to be of an object X, the act's essence or quod quid
est esse, does not differ from what it is for the object X to be, the ob
ject X's essence or quod quid est esse.
We have to be careful here that we do not inadvertently reintro
duce the third thing we denied in the previous section. A nature abso
lutely considered does not constitute an additional mode of being, in
addition to the two modes that St. Thomas discusses in the De ente.
St. Thomas's debt in the De ente et essentia to Avicenna's Metaphys
ics is well known.117 In fact, in his eighth quodlibetal question, orga
nizing the distinctions slightly differently, he gives credit explicitly to
Avicenna for having made them.118 However, one respect in which his
discussion differs from the Latin translation of Avicenna concerns this
117Weisheipl, Friar Thomas DAquino, 78-9, 386-7.

118 ". . . secundum Avicenna in sua Metaphysica, triplex est alicuius
naturae consideratio.... Una, prout consideratur secundum esse quod habet
in singularibus_Alia vero est consideratio alicuius naturae secundum esse
suum intelligibile, . . . Tertia vero est consideratio naturae absoluta. . . ";
Aquinas, Quaestiones Quodlibetales, no. 8, q. 1, a. 1.
very point, namely, the being attributed to the nature considered abso
lutely. Avicenna, like St. Thomas after him, does not attribute being
in re or being in anima to natures considered absolutely, or as such,
or per se. Using the example of the definition of horse, Avicenna says
that "from itself, it is neither many nor one, nor existing in these sensi
bles nor in the soul."119 However, taking another example, considered
apart from existing in sensibles or in the soul, he does attribute to a
nature as such a certain being (esse):
... for from this being (esse) it is neither a genus nor species nor an indi
vidual nor one nor many. But from this being it is animal as such and
manas such... .12?
According to Avicenna, there is a sense of "being" that is appropriate

to the nature considered as such or absolutely, apart from the modes
of existence it has either in re or in anima. Here, however, St. Tho
mas departs from Avicenna; there is no single thing that is the nature
absolutely considered that exists now in res extra animam, now in
anima. Absolute consideration is just that, a mode of considering a
nature, not a mode of being for a nature. As Joseph Owens describes
the De ente, "the text seems to be framed carefully in a way that
avoids any implication of being in the essence absolutely consid
ered."121 Presented with a nature in res extra animam, we can con
sider it absolutely. Presented with another nature in anima, we can
consider it absolutely. The result of the two considerations may not
differ. Yet it does not follow that there is a third nature, existing qua
absolutely considered that is both the nature in res extra, animam,
and the nature in anima.
Perhaps Putnam's objections to the identity of knower and
known arises from the suspicion that the Aristotelian is claiming that
there is just such a third nature, a nature that is identically the same in
res extra animam and in anima. It is clear, however, that this suspi
119 "Ipsa enim in se nec est multa nec unum nec est existens in his sensi
bilibus nec in anima"; Avicenna, Avicenna Latinus: Liber de philosophia
prima sive scientia divina (Louvain: E. Peeters, 1977?80), 228; quoted in
Joseph Owens, "Common Nature: A Point of Comparison Between Thomistic
and Scotistic Metaphysics," in Inquiries into Medieval Philosophy, ed.
James Ross (Greenwood Publishing, 1971), 187. My summary of Avicenna is
heavily dependent upon Owens' analysis.
120 "... ex hoc enim esse nec est genus nece species nece individuum nec
unum nec multa. Sed ex hoc esse est tantum animal et tantum homo. . . .";
Avicenna, Philosophia prima, 234; quoted in Owens, "Common Nature," 190.
121 Ibid., 192.
cion is off base. Extra animam, individual characteristics accrue to a

nature, while in anima characteristics of both individuality, namely,
as an accident of this or that particular intellect, and universality ac
crue to a nature. Yet the theses from the De Ente do not constitute a
further positive attribution of additional characteristics beyond those
associated with the two modes of being, additional characteristics
that would presumably pertain to a third mode of being. The theses
are essentially a negative characterization: there are distinct individ
ual principles of being and knowing, namely, natures, that do not dif
fer when considered absolutely.
Despite the language of formal identity here, there is an order of
priority between the things related. In the language of St. Thomas's
Quaestiones disputatae de veritate, the forms of res extra animam
are the measures against which the mind's activity is judged to be ade
quate or not, namely, true or not, because those res extra animam
provide the form of the intellect's activity. In fact, commenting on the
Peri hermeneias, St. Thomas explicitly repeats this when he discusses
the sense in which a simple act of understanding, a simple passio ani
mae, must be true. He writes:
In one manner [something is compared to the intellect] as the measure

to the measured, and so natural things are compared to the human spec
ulative intellect. And so the intellect is said to be true according as it is
conformed to the thing [rei], but false according as it is at variance from
the thing_and just as a thing [res] is called true through a comparison
to its measure, so also with sense and intellect, of which the measure is
a thing beyond the soul [res extra animam]. So a sense is called true
when through its form it is conformed to a thing existing beyond the
soul. And so it is understood that a sense is true of proper sensibles.
And in this way the intellect apprehending the essence of a thing [quod
quid est] without composition or division, is always true, as is said in III
De anima.122
122 "Uno modo sicut mensura ad mensuratum, et sic comparantur res

naturales ad intellectum speculativum humanum. Et ideo intellectus dicitur
verus secundum quod conformatur rei, faisus autem secundum quod discor
d?t a re_Et, sicut dicitur res vera per comparationem ad suam mensuram,
ita etiam et sensus vel intellectus, cuius mensura est res extra animam: unde
sensus dicitur verus quando per formam [suam] conformatur rei extra ani
mam existenti. Et sic intelligitur quod sensus proprii sensibilis sit verus. Et
hoc etiam modo intellectus apprehendens quod quid est absque composi
tione et divisione semper est verus, ut dicitur in III De anima"; Aquinas, CPH,
L. 1, lectio 3, nos. 29 and 31.
The concept is related to things beyond the soul as what is measured

to its measure. In practical knowledge, the order is reversed; the
mind's activity measures what is to be made or done. In speculative
knowledge, however, the intellect apprehending this "essence of a
thing without composition or division" is called true to the extent that
it conforms to res extra animam.
It is not, however, simply a matter of being identical in form, or
reflecting or being measured by res extra animam that determines
the externalism of St. Thomas. Innate concepts might also stand in
these sorts of relations to res extra animam. St. Thomas, analyzing
the opening passages of the De interpretation, stresses that the con
cept is in its formal characteristics a passion of the soul. He writes
that the term "passion" is used to indicate receptivity from something
else; passions of the soul have their origin "from things by means of a
certain impression or passion," which things act as agents upon the
soul to cause them. Form brings identity to things, causing them to be
units of a certain kind. This is no less true of concepts than it is of res
extra animam.123 Thus, as we have seen, the forms that bring unity
and identity to the intellect's act of understanding do not arise from
the intellect itself, but from the extramental things understood, in
which these forms are principles of being.
Criticizing current causal accounts of mental representation, Put
nam writes that external causes, as presently conceived by the philo
sophical community, do not have enough "form" to be "self-identify
ing."124 This is not the case in St. Thomas's Aristotelian account. In
fact Putnam now seems to recognize this as a virtue of the Aristotelian
account.125 But then his real problem comes into view. The final ob
jection is to the underlying metaphysics, not to the internalism of the
account of language and our knowledge of things, which we now
know to have been erroneously attributed to the Aristotelian. Having
given the Aristotelian credit for a robust enough form to overcome
this part of his objection, Putnam denies that there are unique es
123 This is not contrary to the well-known thesis in Aristotle and St. Tho
mas that matter is the principle of individuation. The context of that discus
sion makes it clear that what is being shown is how units of the same kind are
distinguished from one another, precisely because they are formally identi
cal, or of the same kind. The latter thesis does not involve the claim that the
identity of units of the same kind consists wholly in their material disposi
tion.
124 Putnam, "Aristotle After Wittgenstein," 124.
125 Ibid., 124-9.
sences. His denial, however, appears very weak, based as it is upon

the multiplicity of sciences. It just looks like an unquestioned meta
physical assumption on his part that because separate scientific inves
tigations consider different characteristics essential to their subject,
that therefore there must be, or at least may be, a multiplicity of es
sences for any one thing, studied by those distinct sciences. Of
course, St. Thomas was well aware of this, for he often reminds us not
to confuse the ways of knowing with the things known; a multiplicity
of ways of knowing does not imply a multiplicity of things known.
Putnam himself understands and accepts this point, though in
more contemporary terms. In "The Meaning of 'Meaning'," he clearly
explains for his readers that a multiplicity of senses does not imply a
multiplicity of referents; he is the one who reminds us that a differ
ence of intension does not imply a difference of extension. Here, in
committing this fallacy, Putnam has the best of company, since ac
cording to St. Thomas this was the characteristic mistake of the Pla
tonists. It is clear that the essential descriptions differ between a ge
neticist, an evolutionary biologist, and a biologist studying the
morphology of dogs. However, Putnam needs to give us some argu
ment that the multiplicity of descriptions map straightforwardly on to
a multiplicity of things described, a multiplicity of essences in dogs.
After all, the different scientists mentioned by Putnam all claim to be
scientifically investigating dogs, albeit in different ways. What is it in
virtue of which they can all claim to be investigating the same thing,
the class of dogs, if they are investigating many and distinct essences?
In his recent Dewey Lectures, "Sense, Nonsense, and the Senses:
An Inquiry Into the Powers of the Human Mind," Putnam suggests
that, despite its attraction, he cannot follow the Aristotelian down the
road to essences and their place in the account of knowledge. He sug
gests that he could have titled his lectures "Aristotelian Realism with
out Aristotelian Metaphysics," and at the same time titled them:
Deweyan Realism for Dewey, as I read him, was concerned to show that
we can retain something of the spirit of Aristotle's defense of the com
mon-sense world, against the excesses of both the metaphysicians and
the sophists, without thereby committing ourselves to any variant of the
metaphysical essentialism that Aristotle propounded.127
126 Putnam, "The Meaning of'Meaning'," 216-19.

127 Putnam, "Sense, Nonsense, and the Senses," 445-517.
Of course the problem immediately arises whether Aristotle's "de

fense of the common-sense world" can in fact be divorced from his
metaphysics. An Aristotelian account of knowledge, divorced from its
larger philosophical context, would seem to be Aristotelian in name
only. What motive is served by continuing to call it Aristotelian?
Where Putnam wishes to separate Aristotle's account of language and
knowledge from its largely metaphysical context, the virtue of St. Tho
mas's account is that he seeks to place it ever more firmly in that very
context. Here there may be a fundamental equivocation between the
sense of "metaphysical" as used by Putnam and as used by the Th
omist to describe St. Thomas. The two may be talking at cross pur
poses.
At the end of the day, it represents progress to recognize where
the real argument lies. Many contemporaries may not wish to hold on
to the underlying metaphysics involved in St. Thomas's externalist ac
count. Such a rejection, however, will have to be on grounds other
than that of internalism. As the discussion of the concept that I have
provided makes clear, res extra animam have enough form, and are
"self-identifying," even if they identify themselves to the intellect not
as "this or that thing" simply, but "a thing of this or that sort." There is
indeed an intrinsic and necessary connection between res extra ani
mam and concepts, and the internalist thesis does not cut at the
joints.
VIII
Conclusion. We have seen that objections to the Aristotelian tra

dition of reflection on language, mind, and the world are perhaps a lit
tle too quick when directed at one of its greatest exponents, St. Tho
mas Aquinas. Approaching St. Thomas from within his own context,
we have seen that with respect to at least three elements of Hilary Put
nam's picture, his criticism of "mental representationalism" finds no
target in St. Thomas's appropriation of Aristotle. First, the concepts
primarily signified by terms do not constitute a third realm of mental
things or objects interposed between the mind's activity and external
things; they are the informed activity of the mind in uniting itself to
those extramental things. Second, the mind's concepts are not prima
rily objects of knowledge, the introspective knowledge of which leads
to knowledge of extramental things. Their being is to be to and for
others, namely the primary objects of knowledge?extramental

things. Finally, there is an intrinsic relation between the external ob
jects understood and the mind's concepts, a relation by which those
extramental objects can be said to be self-identifying in a certain re
spect, and can be said to be formally identical with the intellect that
knows them. "We name as we know." What we name are res extra an
imam.
University of Notre Dame

The Problem of Language and Mental Repre John O&#39;Callaghan

Uploaded by

Copyright:

Available Formats

You might also like

The Problem of Language and Mental Repre John O&#39;Callaghan

Uploaded by

Document Information

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

The Problem of Language and Mental Repre John O&#39;Callaghan

Uploaded by

Copyright:

Available Formats

The Problem of Language and Mental Representation in Aristotle and St.

Introduction. In the opening passages of his De Interpretationen

It has been called a triangle because of the three vertices, words,

Correspondence to: Department of Philosophy, University of Notre

A continuous tradition, from Aristotle to Locke and beyond, had as

Perhaps the most important of all the contributions made by Grund

In fact, even recent translators and commentators on the De interpre

4 Michael Dummett, Frege: Philosophy of Language, 2d ed. (Cambridge:

(auu?oXa) and "signs" (or||xeia) by the single Latin word "notae,"

The first phrase ending in "of things" is the translator's paraphrase of

7 Kretzmann, "Semantics," 367.

In a Lockean spirit we have allowed words to range only over the do

This problem, the yoking of language to ideas or mental representa

11 Robert Sokolowski, "Exorcising Concepts," The Review of Metaphys

Russell, Carnap,15 and to a certain extent Chomsky, Fodor, and Searle.

The fundamental criticism of these contemporaries is that our lan

15 Putnam, "Meaning, Other People, and the World," 19.

language. We might refer to the characterization that these critics pro

Aristotle and St. Thomas. At the beginning of the article, I

18 A central presupposition of Kretzmann's claims about the errors of

before chapter 9, where he begins to treat of the soul as moving princi

His commentary suggests that he has roughly the same passages in

21 ". . . excus?t [se] a diligentiori horum consideratione, quia quales sint

predicated of something unless through that name we might under

But here [his] saying concerns sounds signifying by human institution;

24 ?Frustra aliquod nomen de aliquo praedicatur nisi per illud nomen al

the significata of a general vocal utterance. Using the more formal

Using this more formal notion of the significata of a general vocal ut

27 ". . . in quolibet nomine duo est considerare: scilicet id a quo imponi

out mediation." However, having established that it signifies "without

28 See, for example, CPH, L. 1, lectiones 10-11.

30 Here again I am simply stipulating the use of "concept," which has

want to focus on this fundamental structure as interpreted by St. Tho

Historical Setting. As I indicated at the start, in contemporary

36 Russell speaks of ideas, a term I have not introduced. However, the

knows, is primarily the private stock of ideas internally present to his

40 Locke, Essay, book 4, chap. 17, no. 8.

internal ideas about extramental things. If the contents of our minds

... it is a perverting the use of Words, and brings unavoidable Obscurity

48 See Earline J. Ashworth, "Locke on Language," Canadian Journal of

[t]he most vulgar philosophy informs us that no external object can

Our conviction that external objects resemble these internal represen

perceptions in minds.56 Further, the very notion of an external object

He devotes a large section of Investigation 2 to the analysis and cri

60 Edmund Husserl, Logical Investigations, trans. John N. Findlay (New

sage that foreshadows the problems we will see in Wittgenstein, he

For Husserl, the fundamental problem in the "historical treatment of

63 Husserl, Logical Investigations, book 2, chap. 5, no. 34.

representations, form a historical point of intersection between Hus

67 For a discussion of the original interaction and common concerns of

Perhaps the picture is not intended to represent any particular mother

72 Gottlob Frege, "On Sense and Reference," in Translations from the

It seems that the painter determines the object represented by the

76 Wittgenstein, Investigations, no. 663.

termine meaning."78 What is required is a custom and context of use79

Putnam's Criticism. This historical setting provides a context

For Putnam it is clear that the central difficulty is raised by the

78 Wittgenstein, Investigations, no. 198.

representation identified with the meaning of a term, the Aristotelian

The Problem of Language and Mental Repre John O'Callaghan

The Problem of Language and Mental Repre John O'Callaghan

The Problem of Language and Mental Repre John O'Callaghan