Propositions
What is a proposition? What indeed: no one really knows. There have been many proposals: a sentence, a statement, a thought, a meaning, a combination of concepts, a combination of objects, a set of possible worlds, a possible state of affairs, a picture, a model, an image, a name of a truth-value, a description of a fact, whatever is true or false, a family resemblance concept, a non-existent entity, a move in a language game, an ordered pair, the content of a belief, an argument of a truth-function, a sentence in the language of thought. The suggestions range from the mental and linguistic, on the one hand, to the modal and extralinguistic, on the other. Wittgenstein was obsessed with the question, and both his books are about it.[1] I know what a cat is and I know what a mat is, but I don’t know what the proposition that there’s a cat on the mat is; it has to do with cats and mats but it isn’t either or both. It hovers between the world and the mind, unsure where it belongs. It is metaphysically insecure. We might put the problem as follows: an ordinary object (a “substance”) consists of a plurality of properties (“accidents”) co-instantiated in a single thing, but a proposition doesn’t instantiate the properties it refers to—it attributes them. The proposition that the cat is on the mat isn’t a mat or a cat. It contains those properties in the mode of attribution but not instantiation. But what kind of entity does that? You can’t see it or touch it or store it in the attic. It doesn’t belong in the same box as the objects and properties it is about (and what is that?). It is ontologically peculiar; yet it is hardly a mystery. Propositions aren’t like consciousness—there is nothing it’s like to be one. They are ordinary but puzzlingly exceptional. Could the question be investigated empirically? Propositions aren’t individual substances, or accidents of substances, yet they seem to have an individual identity, a sort of cohesive unity. Do they even have parts? Hard to say—not in the way a car has parts. Somehow, we grasp them, know them, apprehend them; but what that is remains obscure. We believe them, but we don’t know what this belief relation amounts to—is it like uttering a sentence or singing a musical note? Do we construct them or do we find them ready-made? Neither answer seems obviously correct. Without them there would be neither language nor thought, but they aren’t themselves linguistic or mental—unless we stipulate as much. They aren’t really psychological, but they have a lot to do with psychology. They also form the subject matter of logic. To describe them as material is to court accusations of category confusion. The OED calls a proposition “a statement expressing a judgment or opinion”: but beliefs are not statements and they take propositions as objects, and what about propositions as they occur in conditionals? A proposition is not a speech act, despite the speech act of propositioning. The dictionary seems as lost as the rest of us. Perhaps they can only be shown but not said, sensed but not sensed. If Alice were to encounter one, what would it say to her? “Don’t you know it’s rude to stare?” “But you are so—funny looking!”
But let’s not wallow in ignorance. Let’s take the proposition by the horns. Here is what I think: a proposition is first cousin to a possibility. The proposition that the cat is on the mat is akin to the possibility that the cat is on the mat. But it isn’t identical to that possibility, because it is not neutral about it; propositions are committed beings—they take a stand. They say of a possibility that it is actual. Of course, they don’t strictly say anything—people do that—but it is as if they say things (the dictionary isn’t wrong about that). A proposition declares its own truth. The proposition that snow is white rejects the proposition that snow is not white. It performs an act of exclusion. A proposition predicates actuality of possibility: it identifies a possibility and then classifies it as actual—all without saying a word. It is as if it speaks. Indeed, if it did not speak, we would be unable to, because in asserting propositions we rely on them to put their foot down. We can only assert things because they do. It is no good trying to assert something without expressing a proposition, say by turning pink; you have to say something propositional (perform an act of saying that p). Propositions are like guns: they kill people not people and their fingers (you can’t kill someone by crooking your index finger a certain way). Propositions are what enable statements to be made—the sine qua non of communication. Animal signaling systems don’t (as far as we know) express propositions, but human language does; and this is what makes languages such powerful tools. But they do so only by virtue of the propositions they encode—they are the real agents of communication (knowledge transmission). We affirm because they affirm. We can’t make them affirm by a sheer act of will—as if we could assert a proposition just by throwing a stone in the air and shouting “Say that p!”. Unless you know how to identify and express propositions you will not get very far as a communicator. Propositions are the machinery of human communication, as of thought; and they work because they have attributive powers built into them. A proposition is not an agnostic but a believer—it is committed. It knows (or thinks it knows) what possibilities are real. It isn’t just a complex but an opinionated complex (“The cat is bloody well on the mat!”). Of course, this is very strange—propositions are not people! But people are also not propositions; they need them to make statements or say anything meaningful (not counting “ouch!” and the like). They don’t need us, but we need them. They give thought content.
How should we understand this property of propositions? As follows: a proposition selects a given possibility as real or actual. It is selective in what it is committed to, very selective; it won’t do to select a possibility coextensive with the possibility in question. It is very particular about what possibilities it selects as real. Again, this is agential talk, calculated to raise eyebrows, but it is hard to avoid it and intuitively appealing to indulge in it. The proposition selects a particular possibility and affirms it as actual: it performs a kind of double act—referring and endorsing. At least we naturally describe it so, borrowing from our vocabulary of human actions; the reality may well be better described otherwise. Our conception of propositions has them referring and endorsing (it’s a bit like how our conception of atoms has them attracting and repelling.) The proposition does something analogous to these human actions. That is, if we are realists about propositions, we end up talking this way—and realism seems the best bet, here as elsewhere. The workings of the mind depend on the workings of propositions not vice versa. We assert, deny, reason, and conclude with the aid of propositions; they don’t rely on us (what would that be?). We also rely on objects and properties to get our thoughts up and running; they don’t depend on us. Concepts without corresponding properties are empty and impotent; if reality contained no properties (“universals”), we would be unable to think about it. Propositions are the pre-existing material of thought and language not consequences of them—that is the realist doctrine. It may be puzzling, even totally confounding, but that’s down to us not them.[2]
I have described what propositions do but not what they are (their ontology). Here I think we draw a blank. We know they are not substances, or anything like substances (so worse off than numbers and geometrical figures). But we have no positive conception of their ontological category: they are not spatial occupants or cohesive bodies or perceptible particulars. They are creatures of their own devising—attributional beings, committed abstracta, opinionated oddballs. You can’t even imagine them. They belong in Wonderland like disembodied feline smiles and frumious bandersnatches. Yet here they are in the real world all around us, populating our minds, controlling our language, shaping our history. Propositions, you have a lot to answer for. Your politics is not always of the finest. You exist in the mouths and hearts of angels and demons.[3]
[1] It would be interesting to trace out this obsession in the two works; it dates back to his earliest philosophical preoccupations. Pictures or practices?
[2] The position I am defending, or presupposing, is clearly very similar to Frege’s conception of objective “thoughts”—propositional realism. I have always felt that he misnamed these entities, given his anti-psychologism, but perhaps he was motivated to do so by a recognition that propositions have an inbuilt commitment to their own truth, like thoughts in the psychological sense. Propositions are ascriptive: the cat is on the mat; it isn’t a mere possibility. It is as if they think.
[3] Personifying propositions seems unavoidable—as if they have opinions and will. They are, indeed, what makes us persons—thinkers, speakers, rational beings. We are propositional creatures. Our attitudes and emotions are propositional. And we share these propositions; they are not private property. There is a fund of them that we all tap into. If early modern philosophy discovered ideas, twentieth century philosophy discovered propositions. Human nature is (partly) constituted by our propositional competence. Our species success depends on this competence. When God created propositions, he made human thought possible—science, morality, art, conversation. No wonder we personify them; they make us the persons we are. But we don’t have a clear idea of their nature (and therefore of our own nature). We are both conscious and propositional, and neither of these traits is pellucid to us, immediate as they are.
