Truth as One
The Correspondence Theory of Truth is held as the realist position because it takes seriously the claim that there is one objective world about which we can have objective knowledge. The Objectivity Truism is at its heart, whether it be Plato, Aristotle or contemporary theorists who are writing from that perspective and it is commonly objected to on the grounds that the theory is vacuous, merely restating a platitude and consequently adding nothing to truth as such. Today, Correspondence has grown or developed into Representationalism. It is manifest in disciplines beyond the boundaries of philosophy, such as cognitive neuroscience, which holds that the mind represents the world and that beliefs are the vehicles of representation.
Representationalism can be traced back to early Wittgenstein and Russell in the early twentieth century; for these philosophers, correctly represented beliefs are ‘true’ beliefs, a la correspondence theory. True beliefs represent facts, which are in turn constituted by objects and/or properties, thus facts, importantly, are not metaphysically distinct from objects or properties. What is the element of the belief which represents the object or property? Concepts: these are the ‘components’ of belief. And thus we have contemporary naturalistic representationalism.
Representationalism is essentially a two-part theory of truth. In the first element, the truth of a belief is defined in terms of the representational features of its component concepts. Hence the representationalist’s basic intuition that beliefs are true because their components stand in certain representational relations to reality, and that reality is a certain way. This basic intuition can then be applied to more complex propositions. The second element is a theory of how concept denote objects or properties. For some, such a theory is explanatorily trivial in that all a theory of denotation amounts to saying is:
<c> denotes x iff c = x
Contemporary philosophers, however, regard a theory of denotation as a substantive issue, claiming that denotation can be explained naturalistically, in the same way as psychology provides an explanation of perception. There are, briefly, two theories of denotation: causal and teleological. The former prioritises appropriate conditions whilst the latter prioritises the biological/evolutionary function. In essence, both can be thought of as a framing hypothesis for naturalistically investigating mental representation.
Both the causal and teleological theories can be combined with a model of representation to give a representational theory of truth that not only incorporates truisms as part of the theory, but also offers an explanation of those truisms:
(CC) Causal-correspondence: The belief that a is F is true iff the object causally mapped by <a> has the property causally mapped by <F>
(TC) Teleological correspondence: The belief that a is F is true iff the object functionally mapped by <a> has the property functionally mapped by <F>
Another criticism of the representationalist’s position is about the possibility of unbelieved truths; for example, is it possible that a proposition is true or false if there is no possibility of discovering warrant for the proposition? In answer to this objection, one could go the way of Donald Davidson and dismiss the significance of them altogether, for nothing would be true or false if there were no thinking creatures. On the other hand, if you want to take the objection seriously, representationalists can argue for a subjunctive bi-conditional (norm of belief), such as:
(UB): The proposition that p is true iff were the proposition that p to be believed, the belief would be true
Representationalism also helps us to answer other interesting questions such as: why is truth an aim of inquiry? Answer: a truth making property is regulative of any practice aimed at belief; inquiry aims at truth because true beliefs are those that correctly represent the way the world is. The representationalist’s theory of truth is also a component part of a relatively simple way to explain interesting phenomena like intentionality. On this theory, truth is reductively explained in terms of an internal connection with representation. As such, platitudes have to be combined with concepts like represent/causal in order to give an informative an explanatorily interesting account of the nature of truth. Overall, we can conclude that representationalism is a successor theory of truth to older correspondence theories of truth and, far from being a decrepit topic in metaphysics, representationalism is taken seriously beyond the boundaries of philosophical discourse including in the scientific realm.
There are, nonetheless, particular problems with representationalism. The first of these is the problem of scope. TC and CC are only plausible in the case of ‘middle sized dry goods’ [Lynch, p.32], that is where we can make a statement that is responsive to the action of <F>. Responsiveness is plausible if mental states with a certain content <G> are causally responsive to an external environment that contains this content <G>, hence the conclusion that truth or falsity rests on the correct assertion of the proposition and the content matching that proposition. Responsiveness is not so plausible if the states or content are not causally responsive, that is, if there isn’t enough <G> to make the content of the proposition correct. Thus, some other account of what makes these statements true comes into play.
As it turns out, Lynch argues, the representationalist is committed to two further conditions. First, that true beliefs map objects that exist and have mind-independent properties (realist position). Second, that objects and properties that are so mapped are capable of entering into at least indirect causal interaction with our minds (causation position). Thus, because of these additional commitments, the scope problem arises because it seems highly implausible that all of our true propositions can fulfil the conditions demanded. For example, in the case of mathematical or moral truths such as ‘2+2=4’ or ‘torture is wrong’, how do our minds interact with numbers? How is wrongness a natural property? Even if we reject (2), non-naturalist correspondence theorists are still committed to the concept of mind independence, but how is that legal facts, for example, are mind-independent when they are the paradigmatic mental construction?
Various ‘isms’ have been constructed in order to deal with the scope problem, such as expressivism, fictionalism and error theory (not an ism, fair enough, but you get the point). But for Lynch, the sheer fact that these are even necessary points to the seriousness of the criticism and, moreover, it isn’t clear that any of the modifications have sufficiently overcome it. Thus, ‘the more substantive the correspondence theory becomes – as when it is seen as part of a larger theory of representation – the more it is vulnerable to the scope problem, and the less plausible it is as a universal theory of truth’ [Lynch, p.35-6].
On the opposite side of the ring, we have antirepresentationalism and superwarrant. For early twentieth century thinkers such as C.S. Peirce, truth is defined as the End of Inquiry, that is: “The opinion which is fated to be ultimately agreed to be all who investigate is what we mean by truth”. For the representationalists, what is true is so because something makes it true. For Peirce and the pragmatists, what is true is so because we agree upon it. In contemporary philosophy, Putnam has expanded upon the early antirepresentationalist intuitions through a theory of internal realism that claims that a proposition that p is true if the proposition that p could be warranted to believe in ideal epistemic circumstances for assessing the proposition that p. Positive aspects of Putnam’s account include his use of the subjunctive to get around having to claim the actuality of ideal epistemic circumstances, and that circumstances are not global but tailormade for each particular beliefs.
There are, however, some negatives, in particular the ‘conditional fallacy’. Crispin Wright has posited ‘superassertability’ to overcome this problem, or the ‘superwarrant’:
(SW) Superwarrant: A belief is true iff it is superwarranted
The benefits of SW are that it negates entirely the concept of ‘ideal’ epistemic conditions, since a belief is warranted by the information available at that present time to ordinary inquirers. Thus to be superwarranted is to be continually warranted at each stage of inquiry without defeat. Disagreements still arise amongst defenders of superwarrant, for example: is inquiry strongly incomplete? What is the nature of warrant? In answer to the latter question, defenders might align themselves with a coherence theory of truth, thus SW becomes SC: A belief that p is true iff that belief is supercoherent.
Other positive aspects include being able to incorporate the objectivity truism into the theory, if one is accepting of two further platitudes: 1) when I believe that p, things are as I believe them to be iff p; 2) the metaphysical view of idealism: p iff the belief that p is superwarranted. Moreover, it entails an attractively simple theory of content, which both explains how we grasp content and how we manifest that grasp in our behaviour.
Antirepresentationalism falls prey to the scope problem, though obviously for different reasons than did representationalism, for (SW) both requires that all content is non-representational and implies that truth is globally epistemically constrained. Now, in the case of some normative truths such as what is deemed to be ‘funny’, or legal truths, epistemic constraint seems plausible, for why would there be a truth for something if no one could ever be warranted in believing it? On the other hand, epistemic constraint becomes implausible when we try to apply it across the board: there must be some truths for which no evidence will be available on principle. Lynch refers to this as ‘humility in the face of the size of the universe [which] seems to demand that’ [Lynch, p.43]. But SW would force us to deny this, therefore resulting in the ‘absurd consequence’ that one must argue that all truths are justifiably believed by someone, or face admitting that the theory is limited in scope.
A second criticism results in a similarly absurd consequence; the ‘many systems objection’, which was pre-empted by Russell, argues that there could be more than one supercoherent system but SW has no way of showing us why two instance of P could not be members of rival systems. Even if one were to argue that supercoherence is predicated on propositions that are undefeasible – even from challenges from rival systems – defenders are still only able to say of one system of belief that it is primary ‘just because it says of itself that is so’: an absurd consequence. The antirepresentationalist faces two options: she can either restate the claim that what make <F> true is that it is a member of S (which leads to absurdity through not answering the many systems criticism) OR she can accept that propositions are not true in virtue of being members of S (which means abandoning coherence as a theory of truth altogether). Thus, there is no way to give an account of warrant apart from one that goes ‘all the way down’ and results in these problems.
The net result is that Lynch can now cite the scope problem as ‘entirely general’: ‘for any sufficiently characterised truth property <F>, there appears to be some kind of propositions <K> which lack <F> but which are intuitively true (or capable of being true)’ [Lynch, p.49]. The scope problem paves the way for Lynch to bring in his third alternative to dealing with the problem. Whilst the traditionalists ‘go for broke’ and damn the counterexamples, and the deflationists dismiss the whole project of a metaphysical account of truth altogether, functionalists argues that there can be more than one property that makes beliefs true. This is the third way that Lynch argues for throughout the rest of his book.