1
Against truth pluralism
Pascal Engel
Université de Genève/ EHESS
1. Pluralism about truth
The concept of truth is, on the face of it, remarkably unified. As Wiggins
(1980), Wright (1993) and many others have pointed out, our ordinary concept of
truth is associated to various “marks” or “platitudes”:
i)
ii)
iii)
iv)
v)
vi)
vii)
viii)
To assert a statement is to present it as true ( assertion)
« P » is true if and only if P (equivalence)
A true statement corresponds to the facts ( correspondence)
A statement can be true without being justified and vice versa (justification)
Truth has no degree (absoluteness)
Truth is timeless (timelessness)
Truth is objective and implies at least convergence (convergence)
Truth is the norm of belief ( Normativity)
To say that these are platitudes is not meant to say that these are part of a
definition or an explanation of truth in terms of more basic notions, or that these
are analytic or a priori about the concept of truth. The idea is just that these
characterize prima facie the concept of truth. It is neither clear that the list is
complete, that it could not be shorter. But the natural temptation of philosophers is
to consider that these platitudes can be transformed into some explanatory
definition of truth. Thus some philosophers would like to define truth through the
third platitude as a substantive notion (truth as correspondence), whereas others
would like to reduce it to the second platitude only as a “thin” notion: there is
nothing more to truth than the fact that “P” is true if and only if P
(disquotationalism or deflationism). Some the platitudes may not actually be
platitudes. As Ramsey once said, “There is no platitude so obvious that some
philosophers have not denied it”. Some experimental philosopher might go ask
people in the street, most probably recommending the addition to the list of the
platitude that truth is relative to something else. Heideggerians might try to
convince us that “true” means something like “disclosure”. Nevertheless no one
denies that these platitudes characterize at least a central concept in our thinking,
and correspond to our core notion of truth.
The core notion, however, is subject to two threats. The first, just mentioned, is
the relativist menace, according to which truth is not absolute, but relative to some
parameter - place, time, person, language, culture, etc. - which can either, once
2
fixed, leave truth intact, or destroy its unity, making it essentially relative. Let us call
this the horizontal menace. I shall put it aside here. The second threat is vertical.
There are many kinds of truths: many true propositions, about many kinds of
things, about cities, persons, institutions, mountains, abstract things such as justice
or freedom. How could they all be true in the same way, given that they are apt to
be true? There are many domains where we can talk of truths: mathematics,
physics, psychology, history, sociology, ethics, aesthetics, comedy and even fiction,
etc. the pluralist intuition has it that the predicate “true” cannot keep its core
meaning and that a number of platitudes have to be abandoned. According to truth
pluralism truth cannot be uniformly faithful to the core concept in all domains. So
there must be variety of kinds of truths, and not one kind only, which could apply
to all domains. The pluralist about truth is not the one who says that truth does not
apply to diverse domains. On the contrary he insists that truth applies to a large
number of domains. But he denies that truth keeps the same meaning in all of
these. It is the threat which occupies me here.
What is exactly the nature of the pluralistic threat? A first aspect of it is the
following. Faced with the variety of truths couched in diverse vocabularies (the
vocabulary of mathematics, of physics, of ethics or of aesthetics, and so on) we can
ask: Are these compatible descriptions of the same world or not? Do they
contradict themselves? The existence of alternative compatible descriptions of the
same world in distinct vocabularies or frameworks poses no threat to the unity of
truth. What is problematic - indeed it is a contradiction – is the existence of
incompatible descriptions of the same world. Some philosophers claim that there
can be incompatible but equivalent (equally true) descriptions of the same world,
for instance of the world described in set-theoretic-terms and in mereological
terms1. But that is impossible. What is possible – and is illustrated by the example is that there are compatible but not equivalent descriptions of a world, descriptions
which use a different notion of an object. The real problem about pluralism – and
the second aspect of the threat - is whether there can be alternative descriptions of
the same world in different vocabularies. I can describe a world as containing three
objects, as containing three beautiful girls, as containing three thoughtful girls, as
containing three divinities, as containing the three Graces, as containing Charm,
Beauty and Creativity. The descriptions are in numerical terms, aesthetic terms,
psychological terms, mythological terms, or in n allegoric language of abstract ideas.
These are compatible, if I mean to describe a painting of the three Graces by
Botticelli or by Rubens. Do these various descriptions describe the same world, or
different worlds? In the first case, how can the world contain all these kinds of
entities, i.e is it a world which contains - in addition to ordinary physical objects and
properties - numbers and numerical properties, aesthetic objects and properties,
morals objects and properties? And how can all these properties be properties of
the same objects described in a basic vocabulary? If we hold that the world is not
only the totality of objects and facts but also the totality of physical objects or facts,
1
see Putnam 2006, and the discussion by Boghossian 2006, p. 37.
3
or of natural objects and facts, how can it contain, in addition to these,
psychological, moral, and aesthetic facts? How can the facts – hence the truths (if
we suppose that facts correspond to truths) about one domain – say, the natural
domain – be compatible with the facts of another domain – say the moral domain and how can they fit together? Shall we say that these are ontologically independent
layers of the world, which is plural in its essence, or that the layers are supervenient
upon, or can actually be reduced to, a more basic one, the natural world of physical
objects? This is what is sometimes called “the placement” problem: how can we
place of fit in the truth of ethics or psychology within the natural world (Jackson
1998)?
To make this issue vivid for the issue of the prima facie plural character of truths,
let us consider the familiar problem in meta-ethics of making sense – or not – of
the idea of moral truth. It can be formulated as a trilemma, in which the following
statements seem both plausible but are incompatible with each other:
(1) Some value judgments are true
(2) No value judgment is true
(3) All true judgments represent reality
Ethical cognitivists accept (1) and (3) but not (2), error theorists accept (2) and (3)
but reject (1). Expressivists accept (1) and (2) but reject (3). They answer the
placement problem by rejecting the moral truths from the fabric of the world. The
problem arises because there seems to be more truths than truth makers.
Confronted with a set of discourses which prima facie ontologically autonomous,
that is which purport to be true of a set of entities or of facts (about ordinary facts,
mathematical, ethical, literary, etc.), one can either (i) eliminate them or consider
them as useful fictions (eliminativism, fictionalism), (ii) reduce them to a more basic
kind of fact (reductionism), (iii) take them at face value add them to our admissible
set of facts. Thus, to keep with the case of ethics, an error theorist is an instance of
(i), expressivism and naturalistic reductionism instances of (ii), moral realism or
cognitivism a case of (iii). The last option is also the pluralistic one, which consists
in adding more facts to our basic naturalistic set of facts. To make the problem
more clearly one about the nature of truth, consider the sentences:
(a)Penguins waddle
(b)Snow is white
(c) Mandela is a hero
The first is true in virtue of an ornithological fact, the second in virtue of a (say)
physical fact, the third true in virtue of a certain kind of ethical fact. The problem is
that the predicate “true” applies to all three statements, although there is nothing in
common between the waddling of penguins, the whiteness of snow, and the
heroism of Mandela. If we set aside eliminativism and reductionism, there are two
options: the pluralist one, which says that here are as many truths as there are kinds
4
of domains (ornithology, physics, ethics, etc.), and the deflationist one, which says
that truth is nothing but a disquotational device captured by our second platitude
“P” is true iff P, and which does not entail the existence of facts to which our
statements could correspond (Horwich 1990). But it seems that this gives us
pluralism for cheap: there are as many “facts” as there are truths, just because there
are as many assertions. Can we combine deflationism and pluralism? I do not think
so. Although several philosophers (including Putnam 2006) call themselves
pluralists, I shall in what follows examine only two options. One is the kind of
alethic pluralism defended by Crispin Wright (1993) and Michael Lynch (2009)
among others. The other one is Price’s. I shall concentrate upon Price’s and
Lynch’s versions2. My diagnosis will be that they are either incoherent or collapse
into a form of deflationism.
2. Price’s pragmatic pluralism
In Naturalism without Mirrors (2011) and in Expressivism,Pragmatism and
representationalism (2013) Huw Price takes up a theme on which he has written a long
time ago in his Facts and the function of truth (1988) the impossibility and the
uselessness of the correspondence theory of truth and of the Tractarian image of a
“Great Mirror”. Since Rorty’s Philosophy and the Mirror of Nature (1979) and
Putnam’s Realism with a human face ( 1992) it has been customary to attack a view
described as “External Realism” , but which is nothing else than the kind of
correspondence view of truth associated for instance to logical atomism, according
to which the nature and function of Thought is to represent reality in some kind of
Great Mirror. Such a view is opposed by those philosophers who call themselves
“pragmatists”, and it is supposed to pervade almost all of contemporary thinking
about truth and metaphysics – from Frege and Russell to Midwestern Realism and
to Australian and Lewisian metaphysics - and it is utterly wrong. Pluralism and
pragmatism are the alternatives. Price wants to defend pragmatism in the form of a
kind of pluralism about truth and a “naturalism without mirrors”. His pluralism
does not consist in expanding our ontology, and adopting as many ontologies as
there are discourses, but in attributing the plurality of discourses to the diversity of
things that we do with language, hence in denying that the added styles of
discourses correspond to any new set of facts. As he notes, this is a kind of
“Wittgensteinian” proposal: plurality has to do with the variety of language games,
not with a plurality within the world.
It is important to distinguish Price’s pluralism from other versions of this thesis.
One version is what one may call simple alethic pluralism (Lynch 2009). It is the view
that « true » is ambiguous, depending on the domain of objects to which it applies,
in the way “bank” is ambiguous. There would be truth for ornithology, for ethics,
for mathematics, for comedy, etc., and these truths would have no more in
2
I have already dealt with Lynch’s version in Engel 2005 and 2012 . On pluralist theories, see Edwards 2013
5
common than penguins, snow and heroes have anything in common. But this
simple alethic pluralism is doomed by a very simple objection3 : if the meaning of
“true” differs depending upon the domain, then simple inferences which contain
“mixed” truths cannot be valid, for their premises are ambiguous. Thus the
inference:
(1) Waddling penguins are funny [comic truth]
(2) This penguin waddles [ornthological truth]
(3) Hence this penguin is funny [comic truth]
is invalid since one cannot infer a comic truth from a comic truth and an
ornithological truth. Alternatively we could not even say that the conjunction of (a),
(b) and (c) above is true, since each conjunct is distinct kind of truth. The problem,
however, is easily solved if one adopts deflationism: if truth is but a device of
assertion, reduced to the principle of equivalence ( platitude (ii)) there is no
problem with (a)-(c) and (1)-(3) inferences, for the content of the assertions, which
determine the domain, are irrelevant: truth is a mere assertoric device, and the fact
that we deal with various truths does in no way affect the “light” nature of the truth
predicate, which is just a logical device. Deflationism can be understood as a form
of pluralism, if we consider that there are as many truths, having nothing in
common- except the disquotational feature - as there are instances of the
equivalence schema ‘P’ is true iff P. Since to be true for a statement is just to assert
it, and that there are many (actually an infinity of) assertions about many sorts of
things, we could say that there as many truth as there are assertions. But this kind
of pluralism for cheap is not the one which deflationists defend. For we do not
simply have a list of individual sentences of the form P’ is true iff P. We can conjoin
and generalize these and say things like “All of what you said is true”, or “All
assertions of the kind K are true”, etc. The equivalence schema is a device of
disquotation which allows us to say that there is only one unique nature of truth,
which is precisely to function as such a device. There are many truths, on this view,
but only one concept, the logical concept consisting in the disquotational device. In
this sense disquotationalism is not a form of pluralism.
The problem with deflationism, however, is that it does not account for the role
of truth in assertion, which is not exhausted by the principle of equivalence. There
is more to truth than sampling asserting or presenting as true proposition. Truth is
a norm of assertion, and this norm is objective, prescribing to take not only as
subjectively true – relative to an asserter –, but also as objectively true, a given
statement. Wright (1993) gives to that effect his “inflationary” argument: truth is an
objective norm of assertion and so cannot be reduced to a mere assertive device.
Price (1998, 2011: 168 sq.) has a closely related, but more graphic argument. He
distinguishes three norms of assertibility, to weak ones (a norm of mere assertibility
of what one believes and a norm of justification and a strong one (a norm of truth):
3
First proposed by Timothy Williamson (1994) and taken up by Christine Tappolet (1997)
6
WEAK (i) Subjective (sincerity): it is prima facie correct to assert that p if one
believes that p (or : one is incorrect to assert that p if one does not
believe that p)
WEAK (ii) Objective (justification): P is objectively assertible if S’s belief that p is
Justified (or: one is incorrect to assert that p if though one believes that p
one does not have adequate grounds for believing that p)
STRONG (iii) Hyper objective (truth): if p is true one should assert that p
or: one is incorrect to assert that p if, in fact it is not the case that p
Price enjoins us to imagine a community of speakers (the MOAs) who criticize
assertions for flouting the principles of subjective assertibility and objective
assertibility but not for flouting that of hyper-objective assertibility. They express
merely opinionated assertions (hence MOAs).They criticize one another for making
insincere or inadequately justified assertions, but not for asserting what’s false. In
other words They can only express faultless disagreements (they are relativists of
sorts, in the sense of Mc Farlane 2005). We can also imagine these speakers being
fully competent in using a disquotational truth predicate, and so in applying the
deflationist truth concept. They fully understand the deflationist truth concept,
then, but not the concept of truth. The MOAs became extinct because they lacked
he capacity to express genuine disagreements. Indeed, the strong objectivist norm
of assertion is better conceived as a norm of knowledge : assert (or believe) only
what you know (Williamson 2000).
A deflationist, however, would say that Price’s distinction begs the question,
and would deny that the MOA lack the concept of truth. The deflationist would
object here that we have no reason to think that (iii), the Hyper-objective norm,
expresses the notion of truth. He would probably say that it expresses a stronger
concept, which it not truth (my suggestion: the concept of knowledge: see below).
Price disagrees. The MOA, according to him, have no means to express genuine
disagreements, and the expression of these is part and parcel of our ordinary
concept of truth. Nor is Rorty’s insistence that there is no distinction between truth
and justification for our peers enough to capture the ordinary concept of truth.
Price, nevertheless, does not want to grant the existence of the strong, hyperobjective norm of truth. This would, according to him, amount to accepting that a
realist concept of truth rules our discourse. If he did he would have to accept some
kind of correspondence or representationist conception of truth. But according to
him, we do not need to make such a supposition. We only need to say that the
MOAs do as if they had such a norm:
“Suppose there is no substantial, objective, property of this kind, which the
Mo’ans’ belief-like behavioural dispositions either have or lack. Nevertheless, it
might turn out to be very much to the Mo’ans’ advantage to behave as if there were
such a property. As it turns out, it isn’t difficult to adopt this pretence. The practice
Mo’ans need to adopt is exactly the same as that required by the previous
7
alternative. They simply need to ensure that when they believe that p, they be
prepared not only to assert (in the old MOA sense) that p, but also to ascribe fault
to anyone who asserts not-p, independently of any grounds for thinking that that
person fails one of the first two norms of assertibility (Price 1998:251).
This view, however, sounds like a form of fictionalism, although one with a
pragmatist twist, according to which our discourse practices need the objective
norm of truth. It also smacks of expressivism, by adopting a dual stance: on the
one hand, from the point of view of the theorists who stands outside the language
game, there is no objective norm of truth and our concept of truth could not be the
realist one, but on the other hand, from within the language game, when we are
involved in it, we do as if we has such an objective concept. In a similar way, Price
distinguishes two notions of representation, one which he calls “i-representation”,
which corresponds roughly to the conceptual roles of a speaker, and the other
which he calls e-representation” which is supposed to refer to something external to
speakers. But the latter kind of representation is not representation of an external
world, but purports to do so. Our e-representations have reference and truth, but
only in a deflationary or semantic sense: all there is to reference is “ x refers to t” and all
there is to truth is “ true(p) iff p” . “So we get “something like, not a plurality of worlds,
but a plurality of ways of worldmaking. Or rather, what we get is one way of
worldmaking, the only way – viz., the adoption of a practice of making factual
claims – but now recognized to have a plurality of potential applications, associated
with the plurality of possible assertoric language games.” (Price 2013: 49)
Unlike other forms of expressivism and quasi-realism, such as Blackburn’s
(1993), which, which is local, and restricted to ethical discourse, Price’s quasirealism or expressivism is meant to be global, hence to extend to all kinds of
discourses – ethical, mathematical, modal, aesthetic, etc. - which thus fail to be
descriptive or factual ( Price 2011: 239, 259). Blackburnian expressivism, by
contrast, entails that there is a contrast between some statements which are truthapt or factual (those about ordinary objects and physical entities) and those which
fail to be such (ethical, etc.). The expressivism is global because Price tells us that
all kinds of discourse are equally truth-apt, and can be accounted in functional
terms, i.e in terms of the specific role that they play in the language as a whole. All
assertive statements, are in the business of truth-stating. As noted above he
defends the deflationist view that these statements lend themselves to the truthschema (or some equivalent of it, depending on the version of deflationism that
one adopts) “the proposition that p is true if and only if p”. This looks at first sight
like an assertion-conditional or inferentialist conception of meaning, according to
which the meanings of statements is given by they “conceptual” or “inferential”
role, the truth-conditional role being derivative from it.
There are two common objections to this view. The first (Boghossian 1990) is
that truth minimalism is incompatible with expressivism, since an expressivist
cannot be a minimalist about the truth-aptness of those statements which are truthapt (in contrast to those, which , as in ethics, are not). But Price denies that
8
expressivism and minimalism about truth cannot go together: even the so-called
“truth apt” statements (for instance, on a Humean view, those that describe states
of mind of a speaker from which statements about moral properties are
“projected”) are not truth apt, and have to be minimally construed. But if this is
the view, how can it be a form of expressivism, given that expressivism entails that
there is a contrast between truth-aptitude and non-truth-aptitude? Price actually
denies that there is a clear frontier between factual and non-factual talk. Price
actually defends a form deflationism or quietism: all discourses are “ truth-apt” just
because they lend themselves to truth talk, but besides that talk there no substantive
essence, no “reality” which the language would have to represent.
The second kind of objection is that the view is as unstable as expressivism
about ethics. Let us concentrate on the idea that with each kind of assertion there
are specific standards or norms. For the expressivist about epistemic norms in
general (and about the norms of assertion and belief in particular) accepts a kind of
dual stance, one from which he does not accept, as a theoretician, the realist
commitment carried by those norms, the other from which, as a participant in the
discourse for which he adopts an expressivist perspective (ethics or intellectual
inquiry), he does accept this commitment. But he can’t have it both ways: to accept
the commitment is to take the norm as regulating our inquiries or conversations,
not to do as if one did accept it. Epistemic expressivism is tenable only if we can
reach an epistemically disengaged standpoint. But we can reach an epistemically
disengaged standpoint only if we can meaningfully abstract from our own epistemic
goals, while inquiring. However, in light of the truth norm or the truth goal, we
can’t meaningfully abstract from our own epistemic goals while inquiring. Our
commitment to the truth goal norm implies that one is not irrealist about this
norm. As Lynch says, in my view rightly:
« One can assess something epistemically only from the epistemically engaged
standpoint. To ask for a justification from the epistemically disengaged standpoint
is to ask for nothing. » (Lynch 2009a)
Price tells us that the source of his pluralism is the Wittgensteinian picture of
language as a city with many quarters and suburbs, all identified through the various
use one makes of various sentences, and with no representationalist “downtown”.
But to what extent are we entitled to move from pluralism about meaning
(“discourse pluralism”) - the view that there are many uses of language - to
metaphysical pluralism - the view that there are many kinds of things? Price accepts
that there are many kinds of facts, but only to the extent that there are many factstating discourses which are actually not in the business of describing any fact or of
referring to anything. But to what extent is this “pluralism” a metaphysical or
ontological view at all rather than a merely semantical view? Price clearly defends
the latter, along lines similar to Carnap’s (Price 2011, ch. 13) : there are as many
possible “ontologies” as there are possible semantic frameworks, but we cannot
state their ontological commitments from outside the frameworks. So the
9
frameworks do not have any ontological or metaphysical import. What Price
objects to contemporary truth-maker realism and to truth-conditional semantics in
general is that it smuggles in metaphysical issues under the guise of an inquiry into
semantics. But semantics must, in his view, refrain from the descriptive task of
providing ontological accounts of the relationships between our language and the
world. The entities we talk about are real or existing, but “real” and “exist” occur
only in the “formal” or “internal” rather than on the “material” or “external”
mode. If it is a “metaphysical pluralism”, as Price describes it (Price 2011, ch2) , it
is metaphysical only by courtesy. It does not say that we can add up to the natural
world as many kinds of worlds as there are discourses (what Price calls “additive
monism”). It should rather be called “discourse pluralism”, with no ontological
import. This is why Price calls it “minimal realism” (Price 2011: 49).
But what kind of realism is that? Since, as we saw , Price accepts the objection
to deflationism that it fails to account for the normative character of truth, and for
the fact that truth register a norm which is distinct from mere assertibility, a truly
objective one, which leads us toward realism rather than pragmatism about truth (Price
2011: 180). But as we saw above, he also insists that his account makes truth a
convenient “friction”, i.e a fiction, but one which has the function of a realistic
device? What kind of realism is that? I do not see how one can combine a
Carnapian neutralism, which is actually a form of metaphysical quietism, with a
form of realism, unless the minimal realism in question is but a kind of deflationism
about truth4. Price’s view is an attempt to keep deflationism about truth – which he
takes to be an essentially anti-representationist and anti-”mirror-like” view of truthwhile at the same time accepting that truth carries heavier commitments than those
which are usually accepted by the deflationist . He tells us that “discourse pluralism
is not an irrealist position, since it accepts that there are moral states of affairs,
possible worlds, numbers, etc. What he or she rejects is the additive monist
attempt to put a metaphysical gloss on such claims” (Price 2011: 49). In other
words, the existence of the states of affairs, worlds, numbers, etc. in question is
only an “internal” question, not an external one. But this is pure Carnapian
neutralism, or realism by courtesy5.
3. Lynch’s alethic fonctionalism
An alternative line that a pluralist about truth can take is to adopt a form of
functionalism not about assertion, but about truth itself. This is the line adopted
by Lynch’s alethic functionalism (Lynch 2009). According to alethic functionalism,
truth is not an ambiguous predicate like “bank” or “river”. The proper analogy is
rather with descriptions such as “the color of the sky” which can designate a blue
sky, a grey sky or a white sky, depending on the circumstances. Just as definite
4
In Engel (2002) I have defended a view which I called « minimal realism ». But it was meant to have commitment
extending far beyond deflationism or Wrightian minimalism, to the idea that we can actually know an independently
real world.
5 See, for a similar conclusion, Tiercelin 2013
10
descriptions indicate a role common to various distinct entities which instantiate
them, truth is a role word, or rather a functional one, which can be realized in
different ways, depending on the domains of discourses (mathematics, ethics,
ordinary antics, etc.). As Wright says :
« The concept of truth admits a uniform characterization wherever it is applied –
the characterization given by the minimal platitudes, which determine what is
essential to truth…The form of pluralism for which space is allowed by this
overarching uniformity is variable realization. What constitutes the existence of a
number is different from what constitutes the existence of a material object. »
(Wright 1996)
Our concept of truth is unified by the platitudes (i)-(viii) above, but the property of
being true is plural (Wright 2001: 752). In other words, according to alethic
functionalism:
(AF) (∀x) x is true if, and only if, x has a property that plays the truth-role)
The truth role can be understood as conjunction of functional properties, or their
Ramsey sentence6. In analogy with functionalism about mental properties, truth is
the higher-order property composed of the set of platitudes, which is realized in the
lower-order properties of each domain. As in the case of mental properties which
can be realized in distinct physical bases, the truth-properties can be realized
differently in various domains. Thus truth can be realized as correspondence for
physical entities, as coherence for moral properties, as super-warrant for
mathematics (depending on arguments to that effect). Like in the mental truth is a
disjunction of various truth properties (Tn)at the lower level for a specific domain
Di :
(T) (∀p) (TG(p) ↔ ((T1(p) D1(p)) … (Tn(p) Dn (p)))
The view is not a form of alethic pluralism (truth is characterized by one single
concept), but there is a “multiple manifestation of truth’s essence”:
“According to our definition of manifestation, a property manifests truth only if it
has the ‘‘truish’’ features in some particular way. Consequently, depending on what
property manifests truth for a particular proposition, we can say that what
makes it correct to believe that proposition is that it has the property of superwarrant, or correspondence ( Lynch 2009: 153-154)
6
Lynch 2009
11
(AF) seems to promise a reconciliation of our intuitions about the unity of our
concept of truth and the disunity of its applications. Notice also that it has no
pragmatist or expressivist consequences: Lynch accepts that correspondence truth
and realism are open options for certain domains. But is AF coherent? It
encounters a variety of problems (see the various essays in Pedersen and Wright
2013). I shall mention only three. A general problem with the view is the following.
Is truth is the role property at the higher level or is it the realized property? If truth
is the higher order property of having these features, we lose the idea that it is truth
itself which has the given platitudinous properties, since truth is realized at the
lower level (for a domain D). But if truth us the realized property (say
correspondence for a domain), then we lose the idea that it is truth which is the
functional property. A plausible answer might be to say that truth is the complex
constituted by the functional property and its realization
The second problem for AF is that, unlike with functionalism in the philosophy
of mind, there is a strong disunity between the various kinds of truths which can be
realized in various domains. In the mind case, if a state, say “pain” is realized in
various physical bases, these have a number of properties in common. But when
we says that, e.g truth is realized as correspondence in one domain and as coherence
in another, these kids of truth are utterly different: makes true a proposition in the
correspondence sense is a truth maker in the world, whereas what makes true a
proposition in the coherence sense are other propositions (indeed we can hardly
talk of a truth making relation in the second sense). In other words, it is not clear
that what is “realized” as truth in a domain is the same as what is realized in
another domain.
A third problem is: how distinct is AF from deflationism? For, on the one hand,
if truth has, in virtue of its role, certain normative features ( say being such that it is
a norm of belief and of assertion), we can expect that these features will be
manifested at the level of the realized property. But by definition if the realizer
properties of truth are diverse, they might be such that they do not instantiate
certain properties of the role. For instance the normative properties can be realized
for mathematical truth, but they may not be realized for ethical truth, or for
comical truth (how could comical discourse be regulated by an objective norm of
truth?). On the other hand, if we allow the role property of truth to be thin
properties, purely formal or structural, how are we to distinguish (AF) from
deflationism? The problem is highlighted by Lynch when he says:
“According to functionalism, both normative truisms about truth are integral to
what truth is. They are part of the core folk theory of truth that individuates the
truth-role. Consequently, any property that manifests truth must satisfy these
normative platitudes. So for example, any property that plays the truth-role for
propositions of a particular domain must be such that it is correct to believe
propositions that have that property. Crucially, however, this needn’t be because of any
intrinsic normative facts about the manifesting property itself. Such properties considered
independently of their role in manifesting truth, may be fully ‘‘descriptive’’. That is,
12
correspondence qua correspondence may have no normative features. It may only
be that correspondence qua manifestation of truth has such features.”(Lynch 2009:
154-155)
If truth’s essence is realised as correspondence, say in the domain of our talk about
physical or mathematical objects, how can it fail to inherit the normative features
that characterise the truth role in general? And if we do not take these normative
features as substantive , how can alethic functionalist be distinct from deflationism.
If the normativity of truth is but a platitude, how can it register a robust property
of truth? The view to be both pluralistic and coherent, the norm of truth must
exhibit different degrees of substantiality depending on the domains. Hence there
must be distinct norms of truth, and not one only. The unity part of truth is then
lost at the expense of the plurality one. On the alethic functionalist’s own terms,
the norm of truth for belief cannot be a feature which would be manifested in
different ways. It has to be univocal, and the same everywhere. AF cannot renounce
this commitment unless running the risk of bringing it dangerously close to
deflationism. But if he accepts this commitment, how can he maintain that there is
a commitment to the norm in every region of discourse (e.g in mathematics and in
comic discourse, in physics as well as in narrative fiction? Presumably, if we grant
that it is present there the norm of truth has to be distinct in fictional narratives
and in mathematics. A possible answer to this objection, on the part of the alethic
functionalist, would be to say that the norm of truth is realised in different ways in
various discourses. And here we could use Price’s distinction between the three
kinds of norms. Thus it could be the weak subjective norm for fiction or comedy,
the stronger norm of justification for ethics, and the hyper-strong norm for
physics. But this reply would have a disastrous effect for the functionalist theory.
It would mean that it is not the norm of truth which is realised, but only the weak
norm of assertibility. But if it is the case, it would not be the same norm when it is
realised. Compare with the case of mental properties. Pain is supposed, on the
functionalist picture, to be realised differently in mammals, reptiles and fish. But it
is supposed to be pain in each case, whatever the multiple realisations. If the norm
of truth is realised as a weak norm in one domain, and a as a strong one in another,
it will not be the same norm.
4. Pluralism about correspondence and knowledge?
Although I have not here given a full argument against truth pluralism (for more
see Engel 2009 and 2013), the conclusion which I am inclined to draw from these
objections is that we should rather be truth monists and reject deflationism. In
other words we should adopt the view that truth is one, and not many, and that it is
a substantive property of our judgements and statements, which has a real essence,
and a version of the correspondence theory seems to me the best option. But it is
not the place to argue for this. I just want to motivate it, on the basis of the
difficulties of truth pluralism.
13
One might actually start from correspondence truth, and argue that there can be
a variety of kinds of correspondence (Sher 2004, 2013). Sher (2013: 159-160)
proposes to take as a platitude about truth the correspondence intuition:
COR Truth is a matter of substantive and systematic connections between
language and the world. These connections hold in a particular case if and only of
that aspect of reality that a given sentence is about is, directly or indirectly, based
on the pertinent principles ( according to the type of case) as the sentence or theory
says it is.
The conditions in question are immanence (the cognitive stance that we take
from within a theory, when we aim at describing the world), transcendence (an
external perspective from which we can assess our immanent thoughts about the
world), and normativity (truth as a standard of belief). These general principles,
according to Sher, leave open what kind of correspondence is to found in each
domain, and can be combined with a degree of holism. On that basis she gives an
account of truth as correspondence in mathematics, as a theory of the formal
structures of reality.
Sher’s account has two virtues. The first is that she insists, rightly, on the fact
that truth has better be a kind of correspondence. As hard as it can be to specify
the appropriate correspondence relation, any theory of truth which would renounce
to give a central role to correspondence and which would fail to be realist would
fail to be a theory of truth. The second is that she insists on a cognitive constraint
on truth, emphasized by Benacerraf’s (1973) challenge: a theory of truth which
would not be constrained by our knowledge of truth in the appropriate way, would
not be a theory of truth. Truth is a “standard for immanent thought”.
Sher’s account, however, is problematic in several ways. First, if it meant to be
compatible with the diversity of truth, it has nothing to say about other kinds of
truths than mathematical and logical ones. If pluralist correspondentism is
supposed to hold, one needs to show how it can hold in the problematic case of
value statements about ethics and aesthetics, which motivates the expressivist
stance. If correspondence is to hold in these domains, one has to show how the
common features of immanence, transcendence and normativity hold in the
problematic domains, and if they do not why they do not. The whole business of
talking about moral truths or aesthetics truths lies there, and Sher does not give the
beginning of an account of correspondence form moral truth, for instance. If we
accept the notion of truthmaking, what would be the moral truthmakers for moral
truth? How would they differ from the truth makers of others? In the second place
it does not show how the correspondence relations are supposed to hold. However
unspecific and open ended they can be, we need an account of these relations and
of the entities which are in relation. Will they be facts, states of affairs, things and
their properties? And how can we here avoid doing some ontology? In the third
place, although Sher does not commit herself to truth functionalism, the same
problems as those raised above reproduces in Sher’s pluralistic correspondence:
14
with respect to normativity, in particular, we need to know how the norm is
instantiated variously in the different domains.
Sher’s modest correspondence pluralism, however, suggests another form of
pluralistic account, which would not be based upon the idea of a disunity of truth,
but on the idea of a relative disunity of knowledge as against the unity of truth.
Sher is right to stress the immanence requirement that a central constraint on truth
is knowledge : truths must be such that they can be known or cognitively accessible .
This indeed does not mean that knowledge or cognitive accessibility is part of the
definition of truth, or that truth has to be known, or is essentially known, as it is
implied by coherence, pragmatist, “internal realist” or super-assertibilist
conceptions of truth. Truth is not justification, not even ideal justification, and
there are truths which cannot be known. That something is true does not entail that
it is known. Truth is not identical to knowledge. But that something is known
entails that it is true, an truth is such that it is a possible object of knowledge. A
constraint on a theory of truth for a given domain (say mathematics), is that it
should explain how, if it were known, it could be such. In other words, it has to
meet what Peacocke (1999) calls the integration challenge: to be such that the
metaphysics of truth is compatible with its epistemology. As a number of writers
have argued the norm of truth for assertion and belief, is a norm of knowledge
(Williamson 2000, Engel 2005, Gibbons 2013).
My purpose here is not to argue for these claims, but to indicate how they can
be the basis of a more plausible kind of pluralism. Knowledge, like truth, has quasiformal core features, which epistemologists try to spell out in full generality ( I call
them quasi-formal because they are not formal in a logical sense). The most
plausible of these core features of knowledge are: it is factive, hence entails truth, it
entails belief, and it is belief which is sensitive (such that if a belief were not true it
would not be believed), safe (such that if one were to believe p, p would be true),
and not subject to epistemic luck or which can be Gettierized (Pritchard 2005).
Depending on the view of knowledge that one has, one could add to this core
features a justification and a reliability condition. There are, however, different
kinds of knowledge: depending on its sources - perceptual knowledge, inferential
knowledge, knowledge by testimony, a priori knowledge – and depending upon its
subject matter or domain – empirical mathematical, logical, moral, etc. Although the
general and formal features must be common to all kinds of knowledge, there is no
reason to suppose that the sources of knowledge are the same and uniform for
each domain. For instance unless one embraces a full-blown empiricism, there is
no reason to suppose that mathematical or moral knowledge are species of
empirical knowledge, and there is no reason to suppose that moral knowledge and
mathematical knowledge share the same features, apart from the core ones. So
there is a unity of knowledge in its core features, and a diversity of things known.
Knowledge (at least propositional knowledge) is knowledge of truths (or, according
to some philosophers, of facts). The truths are diverse depending on their domain
and in the way they are known, but they are uniform in their sharing the property
15
of being true. This suggests a form of pluralism about knowledge, in contrast with
a monism about truth, and a criterion for truth aptness, i.e the capacity, for a domain
to involve truth and falsity claims: a domain – say ethics, or aesthetics - is apt for
truth if there if it is susceptible to be an object of knowledge. Take for instance the
case of literature, in the form of fictional narratives or novels. One can defend a
“no-truth” view of this domain, according to which there are no literary truths, but
only various kinds of interests and social conventions which define a work as
literary (Lamarque and Olsen 1989), or a merely expressivist view of literature,
according to which it is the expression of feelings. But one can also defend a truth
view, and hold that there are literary truths, which can be the object of a specific
kind of knowledge. If one defends this cognitivist view of literature, however, one
needs to say in what sense fictional narratives and other literary forms can be
knowledge. Thus the condition for talking of literary truth is that one is able to
spell out what literary knowledge is. I propose to make it a general condition: for
any domain, we can say that it is liable to truth and falsity if it is liable to
knowledge. If, for a given domain, we can articulate how it can give rise to
propositions which can be known, then we can say that it is truth apt. As I said
above, there is, nevertheless, no reason to suppose that knowledge is uniform part
from its core features. Knowledge is diverse from domain to domain, but truth is
not. There is only one truth concept, but there are many truths, because there are
many ways of knowing these truths. Thus what we get is a pluralism about
knowledge, but a monism about truth.
16
REFERENCES
Benacerraf , P. 1965 “What Numbers could not be” Philosophical Review 74 (1):47-73 (1965)
Boghossian, P. 1990 “The Status of content”, Philosophical Review 99 (2):157-84
2006 Fear of Knowledge, Oxford, Oxford University Press
Edwards, D. “Pluralist theories of truth”, Stanford Encyclopedia of philosophy, 2012
Engel P. 2002, Truth, Bucks : Acumen
2005 « Truth and the Aim of Belief », dans D. Gillies (dir.), Laws and Models in science,
Londres, King’s College, 2005, pp. 77-97
2009 “Truth is One”, Philosophia Scientiae, 13, 1 , http://philosophiascientiae.revues.org/68
2013 “Alethic Functionalism and the Norm of Truth”, in Pedersen & Wright 2013, 69-86
2014 “Can Deflationism account for the norm of truth”, in D. Achouriotti,
Henri Galinon & José Martinez eds, Unifying Theories of Truth, Berlin, Springer
Gibbons, J. 2013 The norm of belief, Oxford: Oxford University Press
Horwich, P. 1990, Truth, Oxford: Blackwell
Jackson, F. 1998 From metaphysics to Ethics, Oxford: Oxford, Oxford University Press
Lamarque, P. & Olsen , S. Truth, fiction, and literature, Oxford: Oxford University Press
Lynch M. 2009 Truth as one and as many , Oxford: Oxford, Oxford University Press
Mc Farlane, 2005 “Making Sense of Relative Truth”, Proceedings of the Aristotelian
Society, 105, 321-39.
Peacocke, C. Being Known, Oxford, Oxford University Press
Pedersen, N.L.J and Wright, C.D. 2013 eds, Truth pluralism, Oxford : Oxford
University Press
Price H 1998 “Three Norms of Assertibility” Nous, 32, Issue S12, 241–254
2011 Naturalism without mirrors , Oxford: Oxford, Oxford University Press
2013 Expressivism , Pragmatism and representationalism, Cambridge: Cambridge University Press
Pritchard, D. 2005 Epistemic Luck, Oxford: Oxford University Press
Putnam, H. 1992, Realism with a Human Face, Harvard, Harvard University Press
2006 Ethics without ontology, Harvard: Harvard University Press
Rorty, R. 1979 Philosophy and the mirror of nature, Princeton, Princeton University Press
Sher, G. 2004“In Search In Search of a Substantive Theory of Truth, Journal of Philosophy 101:5-36
“Forms of Correspondence: The Intricate Route from Thought to Reality.” Truth &
Pluralism: Current Debates. Eds. N.J.L.L. Pedersen & C.D. Wright. Oxford University Press
2013: 157-179.
Tappolet, C. 1997, Mixed Inferences: A Problem for Pluralism about Truth Predicates, Analysis,
57 (3), 209–210.
Tiercelin, C. “No pragmatism without realism”, Review of Price 2011, Metascience Nov 2013, 22,
3, pp 659-665
Wiggins, D. 1980 "What Would be a Substantial Theory of Truth?", Philosophical Subjects - Essays
Presented to P. F. Strawson, Z. van Straten, ed., Oxford : Clarendon Press,189-221.
Williamson, T. 1994 “Review of Wright 1993”, International journal of philosophical studies 2, 10–44
2000, Knowledge and its limits, Oxford: Oxford University Press
Wright, C. 1993 Truth and Objectivity , Oxford: Oxford University Press
1996, “Précis to Truth and Objectivity” and “Response to commentators,” Philosophy and
Phenomenological Research, 56: 863–868, 911–941
2001 Minimalism, deflationism, pragmatism, pluralism,” in Michael Lynch (ed.) The
Nature of Truth: Classical and Contemporary Readings, Cambridge, MA: MIT Press, pp. 751–787.
2013 “A plurality of pluralisms”, in Pedersen and Wright 2013