Hornsby on the Identity Theory of Truth

Author(s): Julian Dodd

Reviewed work(s):
Source: Proceedings of the Aristotelian Society, New Series, Vol. 99 (1999), pp. 225-232
Published by: Blackwell Publishing on behalf of The Aristotelian Society
Stable URL: http://www.jstor.org/stable/4545306 .
Accessed: 25/06/2012 21:36

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

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.

The Aristotelian Society and Blackwell Publishing are collaborating with JSTOR to digitize, preserve and
extend access to Proceedings of the Aristotelian Society.

http://www.jstor.org
 DISCUSSIONS

HORNSBY ON THE IDENTITYTHEORY

OF TRUTH

by JulianDodd

JenniferHornsby(1997) has recently made a case for a position which

she calls 'the identity theory of truth'. I believe that she is right in
thinkingthat a version of 'the identitytheory is worthconsideringto the
extent to which correspondencetheories are worth avoiding' (1997, p.
6). But it seems to me thatthe kind of identitytheorywe should consider
as an antidoteto the correspondencetheory has to be ratherdifferentto
that envisaged by Hornsby.
I
Let me begin by outliningHornsby'sposition. Heridentitytheoryof truth
is, she says (1997, p. 1), a developmentof the sort of view propounded
by John McDowell (1994).1 McDowell puts his case like this:
[T]here is no ontological gap between the sort of thing one can...
think, and the sort of thing that can be the case. When one thinks
truly,whatone thinksis whatis the case... [T]hereis no gap between
thought,as such, and the world. (1994, p. 27)
Hornsbytakes this to mean thattruepropositions(which both she (1997,
p. 2) andMcDowell (1994, p. 179) call 'thinkables')arethe same as facts.
And it is this identity thesis, along with its accompanyingconception of
truth,that she wishes to defend.
Of course, before we can go any further,we must know more about
the respective terms of the identity relation. Identity claims risk being
vacuous unless something substantial is said about the things being
identified.When it comes to facts, Hornsby,like McDowell, is clear.The
world, says Hornsby(quoting McDowell paraphrasingWittgenstein),is
'everythingthat is the case'; it is 'a constellation of facts' (1997, p. 2).
Facts, then, are as a correspondencetheoristconceives of them: entities
(with objects and propertiesas constituents,presumably)whose totality

1. Hornsby acknowledges that McDowell himself would hesitate before calling his
position an identitytheory(n. 2). His Wittgensteinianhostility to 'constructivephilosophy'
(1994, p. 95) would, I think, preclude him from deeming his remarks as theoretical.
Nonetheless, as we shall see below, Hormsbyis correctin viewing McDowell as agreeing
with what her kind of identity theoristhas to say.
226 JULIANDODD

makes up the world. They are occupants of what Frege calls the 'realm
of reference': the reality which containsthe entities relevantfor the truth
of what we say.2
As for thinkables,Hornsbyherself says nothingabouttheirontological
nature,but I interprether silence on the matterto be an endorsementof
McDowell's explicit claim that thinkables are located in the realm of
sense ratherthan the realm of reference:
Given the identitybetween what one thinks(when one's thoughtis
true) and what is the case, to conceive the world as everythingthat
is the case (as in Tractatus Logico-Philosophicus, ? 1) is to
incorporatethe world into what figures in Frege as the realm of
sense. The realm of sense (Sinn) contains thoughtsin the sense of
what can be thought(thinkables)as opposed to acts or episodes of
thinking. The identity displays facts, things that are the case, as
thoughtsin thatsense-the thinkablesthatarethe case. (McDowell
1994, p. 179)
I thus read Hornsby as supposing thinkables to be what Frege termed
'thoughts':entities with modes of presentationof objects,and not objects
themselves, as constituents. And, as McDowell suggests, then, the
claimed identity is between an item from the realm of reference(a fact)
and an item from the realm of sense (a thought).3
Having said this, Hornsby'sclaim, again following McDowell (1994,
p. 27), is that the identity between true thinkablesand facts is truistic.
Indeed,the simple statementof identitybetweentruethinkablesand facts
'is not supposed to tell us anything illuminating' (1997, p. 2) and
'embodiesnothingmetaphysicallycontentious'(1997, p. 9). Nonetheless,
the theory,as distinct from the simple identity claim, is said by Hornsby
to be substantial.Hornsby gives two reasons for this. First, her identity
theorytakes a standon what the vehicles of truthare(1997, p. 3). Second,
and more importantly,it rejectsthe pictureof mind/worldrelationswhich
she thinks is foisted upon us by correspondencetheorists: a picture in
which 'an ontological gap between thought and the world opens up'
(1997, p. 8). As we have noted, Hornsbyagrees with the correspondence
theoristthatfacts are worldly items. Whatshe objects to is the conception
of such facts as things which are located 'outsidethe realmof thinkables'
(1997, p. 7), where this means that a fact cannot enter the mind but only

2. This way of puttingit is derived from Michael Dummett(1973, p. 153).

3. Max de Gaynesford (1996, p. 506) denies that McDowell identifies facts with true
thinkables, and suggests that McDowell's explicit denial of any ontological gap between
thought and the world should be distinguishedfrom a commitmentto the identity thesis.
However, the quotation I have just cited illustrates that de Gaynesford has misread
McDowell. McDowell's view could not be expressed any clearerthan it is: '[t]he identity
displays facts, things thatarethe case, as thoughtsin thatsense-the thinkablesthatare the
case' (McDowell 1994, p. 178).
 HORNSBY ON THE IDENTITYTHEORYOF TRUTH 227

matchsomethingin the mind. This, essentially,is McDowell's diagnosis.

According to McDowell, the correspondencetheorist conceives of the
world as independent of our thought about it, as beyond the 'outer
boundaryof the space of concepts' (1994, p. 8). As a result,so McDowell
claims, the correspondencetheorist is committed to 'the myth of the
given': the view that the world is 'simply received in experience' (1994,
p. 6) withoutmediationby concepts.
If Hornsby is correct, her identity theory exists to provide us with a
way of avoiding this way of thinking, something which reveals it to
embody a conception of truthricher than that of the deflationist (1997,
pp. 20-22). From the vantage point affordedus by her identity theory,
we can explain to a correspondencetheoristthatfacts are not things that
the world takes sole responsibilityfor (1997, p. 8): the things which a
correspondence theorist takes to be 'outside the realm of thinkables'
(1997, p. 7) are nothing but thinkables which are true. As Hornsby
explains (1997, p. 2), quotingMcDowell, if her identitytheory is correct
'there is no gap between thought, as such, and the world' (McDowell
1994, p. 27). The world of facts is swallowed up by the realmof sense.
This, then, is Hornsby'sidentitytheory.In the remainderof this paper
I hope to show the following things: that Hornsby's identity theory is
incoherent; that this incoherence stems from a misdiagnosis of the
mistakemadeby the correspondencetheorist;andthatsome of Hornsby's
remarksare perhapsbest seen as suggesting a different identity theory
which, as it turns out, is quite compatible with the deflationaryattitude
towardstruth.
II
It is useful to distinguishbetween robustand modest identitytheories of
truth,4 theories which differ over the nature of facts (and hence
thinkables).According to a robustidentitytheorist,facts areentities from
the realmof referencewhose totalitymakesup the world.Truethinkables
are then identified with such things, thus inevitably prompting a
Russellian or (perhapsbetter)Moorean view of (true) thinkablesas the
constituents of reality.5A modest identity theorist, by contrast,follows
Frege (1918, p. 51) in identifying facts with true thoughts.On this latter
view, needless to say, facts occupy the realm of sense.
How should we classify Hornsby'sidentity theory?On the one hand,
Hornsby admits to a feeling of reluctanceto think of robusttheories as
identity theories at all (1997, n. 5); but, nonetheless,her theory turnsout
to be a robust one. The reason for this is simple. As we noted in ?I,

4. This distinction is introducedin my 1995.

5. For the attributionof whatI have termedthe robustidentitytheoryto Russell andMoore,
see Candlish 1995, Baldwin 1991 and Cartwright1987.
228 JULIANDODD

Hornsby follows McDowell (who, in turn, follows Wittgenstein) in

regarding facts as occupants of the realm of reference: the real world
which our thoughts are about.
Two furtherfeatures of Hornsby's view entail that it is robust. First,
there is Hornsby's conception of what is wrong with correspondence
theories: that feature of correspondencetheories which it is the raison
d'e^treof her identity theory to repudiate.We saw in ?1 that Hornsby
thinksthe definitivemistakemadeby a correspondencetheoristto be that
of committingherself to the view thatthe world (of facts) lies beyond the
realm of thought. For the moment, I do not intend to challenge this
diagnosis. What is interesting about it is that, if correct, it demands a
particular response. For if the correspondence theorist introduces a
philosophically suspect ontological gap between thoughtand the world,
the mistake can only be rectified by closing thatgap. The world must be
a (literally) thinkableworld. But this is not the sort of move that could
be made by a modest identity theorist.A modest identity theory cannot
close a gap between thoughtandthe worldbecause, if the modest identity
theoryis correct,facts are not to be found in the world.A modest identity
theory merely removes facts from the realm of referenceto the realm of
sense. It has nothingto say aboutmind/worldrelations.Consequently,if,
like Hornsby,we see thejob of ouridentitytheoryas being thatof sorting
out mind/worldrelations, our identity theory can only be robust.
The robustnessof Hornsby'sidentitytheoryis also revealedby her felt
need to respondto a chargeof idealism. Following McDowell (1994, pp.
27-29), Hornsbyarguesthather identitytheory avoids idealism because
'it circumscribesthe world using the notion of a thinkable' ratherthan
an act of thinking (1997, n. 15, my italics). Although the world is made
up of things we can think-items from the realm of sense-it does not
follow from this that there could not be a mindless world (1997, p. 2).
But that the charge of idealism is relevant at all (leaving aside the
question of its accuracy) goes to show that Hornsby'sidentity theory is
robust. A modest theorist, in holding that facts are true thoughts, does
not circumscribethe world using the notion of a thought.For the modest
identitytheorist,facts arenot in the world,so the spectreof idealismdoes
not threatenher. Idealism is only a potential threatto a robust identity
theory:a theory which takes facts to be worldly and then appearsto 'cut
the universedown to size'6 by identifyingsuch facts with truethinkables.

III

What follows from the fact thatHornsby'sidentitytheoryis robustrather

than modest? One thing is for certain:pace Hornsby,her identification

6. This phrase is borrowedfrom Thomas Nagel (1986, p. 109).

 HORNSBY ON THE IDENTITYTHEORYOF TRUTH 229

of facts with true thinkables cannot be truistic. Hornsby quotes, with

approval(1997, p. 2), the following remarksmade by McDowell:
[T]o say that there is no gap between thought, as such, and the
world, is just to dress up a truismin high-flown language. All the
point comes to is that one can think, for instance, that spring has
begun, and thatthe very same thing, thatspringhas begun, can be
the case. That is truistic, and it cannot embody something
metaphysicallycontentious...(McDowell 1994, p. 27)
But we should not be misled by this. What is truisticis thatwe can think
that spring has begun, and that springhas begun can be the case: we can
use the very same words to specify what is thought and what is a fact.
But it is a substantivephilosophical thesis to say that what is thought is
(when true) literally identical with a fact. Indeed, how could it be
anything but a substantial thesis. After all, the whole purpose of
propoundingthe theoryis to pitch in to the substantivedebateconcerning
the merits of the correspondence theory's conception of mind/world
relations. I do not understandhow Hornsby's identity theory could do
this if it were nothing but a truism.
Things, however, get worse. As it turnsout, Hornsby'sidentification
of facts with true thinkablesis incoherent.7For facts (as Hornsbythinks
of them) and thinkables (as Hornsby takes them to be) are of different
ontological categories:occupiers of the realm of referenceand the realm
of sense respectively.If the worldis to be everythingthatis the case, then
the things that are the case-facts-must have objects and propertiesas
constituents. Thinkables, meanwhile, if they are to be occupants of the
realm of sense, must have modes of presentationas constituents.They
must be thoughts. Consequently,Hornsby's identificationof (worldly)
facts with true (Fregean) thinkables cannot be made good. A mode of
presentationis of an object; it cannot be identified with it.

IV
At the root of Hornsby's difficulties lies a misunderstandingof the
correspondence theory of truth and a consequent misdiagnosis of the
trouble with it. Hornsby (and McDowell) portray the correspondence
theorist as a kind of extreme, or transcendental,realist: someone who
takes the world (of facts) to be a self-subsistent realm beyond our
concepts. Once this picture is seen as objectionable, and once it is
supposed to be the job of an account of truthto correct it, Hornsby is
bound for trouble.For if the objectionablefeatureof the correspondence
theory is its portrayal of facts as beyond the outer boundary of our

7. 1 press the same point against McDowell in my 1995.

230 JULIANDODD

concepts, then it is tempting to suppose (with McDowell and Hornsby)

thatthese facts shouldbe broughtwithin the realmof sense and identified
with (true)thoughts.This resultsin incoherence,if we assume (correctly,
in my view) that thinkablesare Fregean.
Actually, I do not believe thata correspondencetheoristneed commit
herself to the facts being beyond the reach of our concepts. It appearsto
me thatthe distinctivefeatureof a correspondencetheoryis not so much
a commitmentto this metaphysicalpicturebut to the separableidea that
every truthmust have a truthmaker.A truthmakerfor the truththat a is F
is something whose every existence entails that a is F. As D. M.
Armstrong puts it, 'the truthmakerfor a truth must necessitate that
truth.... [I]f a certain truthmakermakes a truth true, then there is no
alternativeworld where that truthmakerexists but the truth is a false
proposition' (1997, p. 115). This being so, I think we can say that the
distinctiveintuitionheld by a correspondencetheoristis expressedby the
following principle:
(TM) For a thinkableto be true, there must exist at least one item
whose existence entails that it is true.
One more move is requiredto reacha correspondencetheoryproper:the
observation that facts are best equipped to be truthmakers.8Once this
move is made, we arriveat the correspondencetheorist'ssingularvision
of facts: items, distinct from thinkables,which make thinkablestrue.
This vision of facts as truthmakers, the benchmark of a cor-
respondencetheoryof truth,is quiteseparablefroma commitmentto 'the
myth of the given'. In fact, an acceptance of 'the myth of the given' is
neither necessary nor sufficient for one to think that facts make truths
true.It is unnecessarybecause one could take the facts which occupy the
truthmakingrole to be things which were in some way conditioned by
our concepts. (A Kantianwho took phenomenato be arrangedin facts,
distinct from the vehicles of truth, would be such a philosopher.)It is
insufficient because someone could agree with Hornsby'sopponentthat
the world lies beyond the 'outer boundary of the space of concepts'
(McDowell 1994, p. 8) and yet take that world to be a world of objects
merely, and not facts. (Such would be the view of a philosopher who
coupled a Davidsonian antagonism towards facts with an 'Australian
Realist' view of mind/worldrelations.)

8. This claim is swiftly made by Armstrongafter he commits himself to (TM). 'We are
asking', he says, 'what in the worldwill ensure,maketrue,underlie,serve as the ontological
ground for, the truththat a is F. The obvious candidateseems to be the state of affairs of
a's being F. In this state of affairs(fact, circumstance)a and Fare broughttogether'(1997,
p.11I6).
 HORNSBYONTHEIDENTITY OFTRUTH
THEORY 231

The construalof facts as truthmakers(as opposed to truethoughts)is the

place at which correspondence theories are vulnerable. And, indeed,
things appearto look quitepromisingfor the would-becritic.Ontological
economy favoursthe identificationof facts with truethoughts,as long as
it is agreedthatthere are good theoreticalreasonsfor positing thoughts;9
and P.F. Strawsonhas arguedelegantly againstfacts being worldly items
(1950). In addition, one might question whether (TM) can be
satisfactorily motivated. Armstrong,for one, says merely that (TM) is
'fairly obvious once attention is drawn to it' (1989, p. 89), but we are
entitled to wonder what is so obvious about it. Even if we think that we
are bound to talk about particulars instantiating universals, thus
committing ourselves ontologically to particularsand universals,this is
still some way from admittingthat the instantiationof a universalby a
particularis itself an entity.10We are still a good distancefrom being able
to justify a conception of facts as worldly truthmakers.
I do not wantto prejudgewhetherthe case for facts being truethoughts
can be made good. All that need be said now is that if such a case could
be made, we would therebyend up holding the modest identity theory I
outlinedin ?11.Indeed,fromthe perspectiveof a modestidentitytheorist,
Hornsby repeats a mistake made by the correspondencetheorist: the
placement of facts in the world. She does not compound the error by
taking such things to be truthmakers,but only at the price of putting
forwarda theory which cannot be coherentlystated.
If we need to respondto correspondencetheoristsby puttingforward
an identitytheory,such an identitytheorycannotbe Hornsby's.Crucially,
however,were Hornsbyto retreatto the modestidentitytheory,she would
have to accept its compatibilitywith the sort of deflationaryaccount of
truth propounded by Paul Horwich (1990). For although the claimed
identity between true thoughts and facts is a substantialaccount of the
nature of facts, and although it amounts to a rejection of the
correspondencetheory,it says nothing with which the deflationist need
disagree. As we noted in ?11,to say that facts are true thoughts is to
venturenothing about the characterof mind/worldrelations;indeed, all
that is therebysaid is that truthcannot be a matterof correspondenceto
fact. Of philosophical interestthis may be, but it is not, as Hornsbytakes
her identity theory to be (1997, p. 16), an alternativeto deflationism.

9. In my 19971 arguethatthoughtsmustbe admittedinto ourontology, if we areto succeed

in giving a plausible account of oratio obliqua.
10. This point is well made by Terence Horgan in the context of a critique of Jaegwon
Kim's conception of events as attribute-exemplificationsat times (1978, p. 42).
232 JULIANDODD

Nonetheless, Hornsby occasionally makes comments which cohere

nicely with the modest identity theory.I have already noted her explicit
denial that her own theory is robust (?11).Besides making this remark,
Hornsbyalso observes of semanticanti-realists,that '[t]heirformulations
often appearto invoke a conception of a truth-makerwhich will suit a
correspondence theorist but which an identity theorist cannot allow'
(1997, p. 8). This fleeting interestin the notionof a truthmakeris a missed
opportunityto focus on the essence of the correspondencetheory.That
the opportunityis missed, and that Hornsbyeventuallyends up offering
an incoherent kind of identity theory, are facts explained by her
misconstrualof correspondenceas a sort of transcendentalrealism.
Philosophy Group
Bolton Institute
ChadwickStreet
Bolton BL2 JJW
E mail: jd4@bolton.ac.uk

REFERENCES

Armstrong, D. 1989: Universals: An Opinionated Introduction, Boulder:

Westview Press.
Armstrong, D. 1997: A World of States of Affairs, Cambridge: Cambridge
UniversityPress.
Baldwin, T. 1991: 'The IdentityTheory of Truth',Mind 100, pp. 35-52.
Candlish, S. 1995: 'Resurrectingthe IdentityTheory of Truth',Bradley Studies
l,pp. 116-124.
Cartwright,R. 1987: 'A Neglected Theory of Truth', in Philosophical Essays,
Cambridge,Mass.: M.I.T.Press.
de Gaynesford, M. 1996: 'Critical Notice: John McDowell, Mind and World',
AustralasianJournal of Philosophy 74, pp. 495-509.
Dodd, J. 1995: 'McDowell andIdentityTheoriesof Truth',Analysis 55, pp. 160-
165.
Dodd, J. 1997: 'Parataxis,IndirectSpeech andthe Natureof Things Said', Journal
of Philosophical Research 22, pp. 211-227.
Dodd, J, & Homsby, J. 1992: 'The IdentityTheory of Truth:reply to Baldwin',
Mind 101, pp. 319-322.
Dummett,M. 1973: Frege: Philosophy of Language,London:Duckworth.
Frege, G. 1918: 'Thoughts', in Nathan Salmon and Scott Soames (eds.)
Propositionsand Attitudes,Oxford:Oxford UniversityPress, 1988.
Horgan,T. 1978: 'The Case AgainstEvents', PhilosophicalReview87, pp. 28-47.
Homsby,J. 1996/7: 'Truth:The IdentityTheory', Proceedingsof the Aristotelian
Society 97, pp. 1-24.
Horwich,P. 1990: Truth,Oxford:Blackwell.
McDowell, J. 1994: Mind and World,Cambridge, Mass.: HarvardUniversity
Press.
Nagel, T. 1986: The ViewFromNowhere,Oxford:Oxford UniversityPress.
Strawson,P. 1950: 'Truth',reprintedin George Pitcher (ed.) Truth,Englewood
Cliffs: Prentice-Hall,1964.