Your complimentary articles
You’ve read one of your four complimentary articles for this month.
You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please
Mirroring without Metaphysics
Michael Philips on truth and the Correspondence Theory.
Most Americans believe that Osama bin Laden ordered the attack on the World Trade Center. Some Muslims believe the Israelis were behind it. Many Postmodernists write as if there is no truth of this or any other matter (only narratives). Pragmatists of Richard Rorty’s school claim that there is a truth but that truth really ultimately means ‘what’s good for us to believe’. Both Rorty and the Postmodernists defend their position by attacking the Correspondence Theory of Truth. They are not the first to do so. Many have argued that the Correspondence Theory presupposes a false or incoherent metaphysics.
The basic idea of the Correspondence Theory is simple enough: a belief (statement) is true if and only if it corresponds to the world. Taken by itself, this seems too obvious and insubstantial to be dignified by the word ‘theory’. It becomes a theory when we provide accounts of the constituent terms and relations. What exactly is a belief? What exactly are statements? In what ways do statements ‘hook up with’ the world? As we try to answer these and related questions we often find ourselves doing metaphysics. The critics of the Correspondence theory argue that the metaphysics that underlie that conception are indefensible.
Richard Rorty titles this metaphysics ‘mind as the mirror of nature’. He argues that it originates with Descartes. Descartes represents us as nonmaterial minds directly aware of nothing but the contents of our consciousness, which he describes as ‘ideas’. Our problem is to determine which of these ideas (mental representations) accurately represent the world. But since our awareness is limited to our ideas, since we have no direct access to the world itself, we have no way to compare our ideas to the world itself. Descartes ‘solves’ this problem by declaring that God is no deceiver. Accordingly, some of our ideas do accurately represent the world. But Godless contemporary philosophers who adopt the Cartesian starting point are in trouble. They have no guarantee that there even is a world independent of our ideas. Furthermore, they have no way to make sense of someone holding representations up to the world to determine whether or not they mirror the world. Since we can be aware of ideas and ideas only, since whatever we are conscious of is an idea, all we can do is compare one set of ideas with another.
Since Rorty and his followers believe that the Correspondence Theory of truth presupposes this Cartesian Metaphysics (or some close cousin), they argue we must give up the Correspondence Theory of Truth. We should just stop pretending that we have some way of comparing our inner representations of the world to the world itself. We should confess that all we have are certain contingent and variable cultural practices for justifying our beliefs, practices that can be justified on the ground that they produce ‘what’s good for us to believe’. If, by following their justified practices, Americans conclude that they know the identity of the World Trade Center hijackers, that is their truth. If certain Muslims, following their justified practices, believe the people on the American list have been framed, that is their truth. There is no such thing as the truth in the good old fashioned sense. We can’t ask who is right about this really.
The strange consequences and self-referential problems with this view are well known. But Rorty and his followers believe that the correspondence view of truth is even harder to live with because it presupposes a Cartesian metaphysics (as if that were stranger still). But why is the correspondence view of truth shackled to Descartes’ metaphysics? Why, in fact, is it shackled to any metaphysics at all? The correspondence view of truth is prosaic, straightforward and utterly sensible. As suggested, it may be converted into various correspondence theories of truth by the addition of certain metaphysical and/or scientific accounts of or replacements of its key terms (for example, replacing beliefs with descriptions of the brain processes that support them). But the fact that one such elaboration of the correspondence view – the Cartesian elaboration – is inadequate doesn’t undermine the correspondence view itself. To undermine that one needs a general argument that no such elaboration is even possible (and what could that be?). Or one needs a simple, old fashioned philosophical argument against the correspondence view itself (for example, that it is incoherent or that it has counterintuitive consequences).
No one, to my knowledge, has adopted the daunting former strategy but many philosophers have adopted the latter. One common complaint against the correspondence view is that the idea of correspondence is deeply mysterious or incoherent. And it is, if understood against the background of certain metaphysical pictures (like the Cartesian picture sketched above). But understood independently of such pictures, the idea is clear enough. Correspondence is accurate representation. It’s the relation that holds between a good map and a terrain, a good blueprint and a building, a good graph of changing values and the fluctuations of those values. All of these representations are accurate to the extent that certain lines on a page correspond to some feature of the world (for example, the locations of streets, the sizes and positions of rooms, the fluctuations of stock prices). Anyone who does not understand this relationship, anyone who does not know what it is for a map to represent a terrain, should seek government assistance for his curious mental handicap.
Some philosophers who say that the idea of correspondence is mysterious mistake the source of their perplexity. Their real problem is not with the idea of correspondence per se, but with understanding how sentences and beliefs can correspond to the world. Philosophers may be perplexed by this because they think they need a deep theory of sentences and/or beliefs understand it. In particular, they think they need a theory of meaning for sentences and a theory of mind for beliefs.
I don’t see why. We don’t need a deep theory of maps, blueprints and graphs to understand what it is for them to correspond (or fail to correspond) to what they purport to represent. So why do we need one for sentences or beliefs? We can convey the same information we get from maps, blueprints and graphs in sentences and vice versa. So if we know what it is for one to correspond to the world (as we do), then why not the other? Sentences, of course, are governed by various rules and conventions, require interpretation and so forth. But the same is true of maps, graphs and blueprints. Maps, graphs and blueprints are, in a certain way, more pictorial than sentences. Sentences don’t look anything at all like what they represent. But the colors representing population densities on my demograhic map of England don’t look anything like populations either nor do the dashes lines and spaces representing my kitchen on the blueprint of my house look anything like my kitchen walls.
It would be wonderful to have a deep theory of representation. But if we need one to know how sentences represent, we also need one to know how graphs, flow charts, musical scores, equations, blueprints, electrical diagrams, seating charts and the periodic table of the elements do. And as we’ve seen, though, we don’t need a theory of this kind to know that these forms of representations represent or to understand what it is for them to do so
Questions about how words (sentences) represent or otherwise relate to the world have been front stage center in Western philosophy since Plato. They have also been a central focus of English language philosophy in the 20th century. In recent years, our attempts to answer them have drawn on work in cognitive and neuroscience. Still, we don’t have anything like a consensus on these matters (not to mention anything we can reasonably expect people to accept a hundred years from now). That is a lot to ask of any philosophical theory. But the fact that we have no deep theory of this kind is no reason to deny the facts such theories attempt to explain. Like maps and graphs, sentences (at least many of them) represent the world. They can do so accurately or inaccurately.
Interestingly, although enormous attention has been lavished on the relation between sentences and the world by philosophers, relatively little has been devoted to other forms of representation. But certain features of the world are more economically and clearly represented by graphs, diagrams, maps and so forth than they are by words. Much could be learned by thinking about why this is so. But I digress.
I have argued so far that there is nothing mysterious about the idea of correspondence and that we don’t need a deep theory of representation to know that beliefs or sentences can correspond to the world. But what about the world to which sentences, pictures, and so forth are supposed to correspond? Is there something about it that makes it impossible for us to understand how sentences (maps, pictures, etc.) can represent it?
Eschewing subtleties, this is what Derrida and his many sympathizers in the English speaking world say. They think the Correspondence Theory requires that we have direct access to the world as it is in itself. But we can’t have such access. We are trapped within our own contingent linguistic or conceptual forms. There is no way to break out of the world of ‘signifiers’ to get to the world itself.
I confess I find Derrida’s writings extremely obscure and difficult. So I’m not entirely clear on his reasons for this view (though, if we take him seriously we must suppose that there is no such thing as what he really means and my guess is as good as any). Some of his followers, however, take the problem to be that there are an indefinitely large numbers of ways to categorize the ‘things’ in this world. Every culture (every individual) settles on its own set of divisions, its own list of sameness and difference. These selections are not in every sense arbitrary. They can be explained in relation to interests, power relations, traditions and the like. But they are arbitrary in one sense, namely, that they can’t be justified on rational grounds. We have no basis for saying that one way of doing it more accurately ‘cuts nature at the joints’ than any other. Since the categories we ‘throw at’ the world are in these ways expressions of our own contingent circumstances, they do not represent the world as it is in itself. Worse, the fact that we can’t escape our own contingent categories renders the very idea of the world as it is in itself incoherent.
This argument raises interesting and important issues that I can’t discuss here (I discuss some of the in ‘The World as it is in itself Revisited’, Philosophy Now, Issue 34). But four related points can be made. First, a contingent, optional form of representation, one that has no greater claim than some alternative to ‘cut nature at the joints’, can be used accurately to represent the world. We can choose to measure distance in feet, meters or cubits. The choice is optional and the one we’ve made is a reflection of our historical situation. Still, the sentence “Michael Jordan is 6’5” tall” accurately (or inaccurately) represents the height of Michael Jordan. Similarly, we can divide the animal kingdom into beings that weigh more than three pounds and can fly (‘fleugels’) and beings that either weigh less the three pounds or can’t fly (‘fleegals’). This taxonomy is hardly a candidate for ‘cutting nature at the joints’. But the sentence “There are more fleegals in the world than fleugals” accurately represents the world. Just go out there and count.
Second, the fact that we can divide things in the world into kinds implies that there are ways in which things in the world are alike and unlike one another. If there weren’t, we couldn’t divide them into classes in the first place. The resemblances and differences in the world exist before we encode them in language. To think otherwise is to believe in some curious form of verbal conjuring. Once we encode these resemblances and differences in language – once we have names for classes – we have a convenient way of making true or false sentences that we didn’t have before. This doesn’t mean that every noun in our language is grounded in some resemblance that exists in the world. Sometimes we falsely believe things exist (goblins, witches). These classes are also based on resemblances. But things that bear these resemblances to each other exist only in our imagination. Sentences that assert or imply otherwise do not correspond to reality.
Third, one can’t reject the correspondence view on the grounds that 1) accurate representation requires a language that ‘cuts nature at the joint’ and 2) we can’t have such a language. What follows from this is that we can’t accurately represent the world. That is, we can’t say true things about it. This conclusion is highly counterintuitive. It seems more reasonable to say that we don’t need a language that ‘cuts nature at the joints’ to make true statements. Questions of accurate representation arise in relation to whatever vocabulary we have.
Finally, the correspondence view of truth is not about The Ultimate Truth or The Final Answer. The Ultimate Truth (if we can make sense of such a thing) might require a language that ‘cut nature at the joints’. The correspondence view, on the other hand, is about what makes plain old ordinary sentences true. The fact that The Ultimate Truth about the universe may make no mention of hats, car alarms and zip lock bags doesn’t imply that we can’t say true (or false) things about them.
In sum, I know of no good direct arguments against the correspondence view of truth. Again, we get correspondence theories by fleshing out the correspondence view with metaphysical and/or scientific accounts of sentences, beliefs, representation and so forth. What is their ontological status? How do they fit into a theory of the mind? How are they stored and/or activated in the brain? What makes one kind of representation (e.g., graphs) a more economic or powerful way to represent aspects of the world than another? How does the theory of representations relate to information theory? Answers to these and related questions may draw on the work of linguists, cognitive scientists and neural scientists as well as philosophers.
But we don’t need definitive answers to these questions to hold the correspondence view of truth. If anything, the correspondence view of truth constrains our answers to them. We start with the obvious fact that maps, graphs, flow charts, sentences and the like can accurately or inaccurately represent the world. And, unless our investigations unearth some extremely powerful reasons to do otherwise, we should reject any deep theories that makes this representation problematic or impossible.
© Michael Philips 2002
Michael Philips is a professor of philosophy at Portland State University in Portland, Oregon. In his spare time he is a potographer and performance artist.