March 2023 - Colin McGinn

Saying and Showing

Wittgenstein famously introduced the distinction between saying and showing in the Tractatus. I won’t be concerned with his treatment of the distinction, either by way of interpretation or evaluation; but I will be using the terminology. I want to say that every speech act includes an act of showing, as well as an act of saying, and also that showing is not a type of saying. It is not (pace Wittgenstein) that what is shown cannot be said, but only that the showing that occurs in speech acts is not a type of saying in that very speech act. The kind of showing I have in mind is perfectly familiar and non-mystifying: it is the mere utterance of a sentence with its characteristic form. The speaker displays or exhibits or presents a sentence to the hearer’s senses (generally vision and hearing), thus showing him that sentence. The speaker doesn’t say he is showing the hearer a sentence; he simply does it—as he might show the audience a coin in his hand. But the hearer is then in a position to know what the speaker is saying: the speaker says something to the hearer by showing him a sentence. He might, for example, produce a written sentence from behind his back that the hearer is now in a position to interpret and make an ascription of saying. Showing in this sense is an act of proffering an item to the senses, and this is what enables the speaker to communicate by acts of speech. So two acts are performed in a given speech act: an act of showing and an act of saying, where the former enables the latter. In other words, the act of uttering (saying in the oratio recta sense) is an act of showing (displaying, exhibiting, etc.) in the performance of which something is said in the oratio obliquasense. Utterance is not merely sounds issuing from the speaker’s mouth—that might be just involuntary babble—but an intentional presentation to someone’s senses of a sentence (conceived as such) for a specific purpose, viz. to say something to that person. The speaker is showing something to the hearer in roughly the sense in which a tour guide might show you the way to a cathedral. The purpose of the showing is to perform an act of saying and it is a sine qua non of that.

In fact there is a bit more complexity here. First, there is not just the speech act of saying but also of commanding and questioning (and any other type of speech act you may believe in). The speaker shows the sentence “Shut the door!” in order to command the addressee to shut the door, or shows the sentence “What time is it?” in order to ask what the time is. Notice that the variety of speech acts performed is accompanied by uniformity in the act of showing: all speech acts involve showing, though not all involve saying. So there is something in common to all speech acts—they all involve an act of showing. Moreover, they all involve something not specifically linguistic, because showing occurs in a wide range of activities: we show sentences to each other in much the way we show things in general to each other (maybe the one capacity derives from the other). Second, there are two parts to the kind of showing that occurs in acts of communication: one part is the act of the speaker in showing a sentence to the hearer; the other is the sentence itself showing its form to the hearer (this is closer to Wittgenstein’s use of the concept). I show you a sentence S and S shows you its grammatical and logical form—without saying anything about this form. A conjunctive sentence doesn’t say it is a conjunction—it just is one. So strictly there are two acts of showing in any speech act: speaker showing and sentence showing. The speaker shows you a sentence and the sentence shows you its form (as well as its vocabulary). The speaker doesn’t say, “I am showing you this sentence” but simply does it, and the sentence doesn’t say, “I am an existentially quantified sentence” though it manifestly is one. This double showing is integral to the success of the speech act and is not to be viewed as mere acoustic production: it is part of speech as a rational purposive activity.

I am tempted to suggest that this way of talking comes naturally because we are a theatrical species. Our social interactions have a theatrical character (think Shakespeare and Erving Goffman). We are always “putting on a show”. Thus the idea that speech involves performance is a theatrical idea (we perform speech acts as actors perform their lines). Our speech comprises a display that is designed to be interpreted by an audience as an act of saying (etc.). If all the world’s a stage and we are merely players, then our speech will involve acts of theatrical showing—skilled presentations that reveal states of mind. We show other people things in order to get things across to them: we wave our hands, point our fingers, make urgent sounds when in extremis, and produce grammatical strings. We proffer things to other people’s senses in the hope that we will be understood. This requires skills akin to those of an actor; and isn’t speech often a kind of acting? We have to put in a goodperformance, a convincing verbal display (e.g. Winston Churchill giving a mesmerizing speech). So we naturally think of our speech performances in theatrical terms—as a type of show we put on. The great speaker or writer is exceptionally good at showing people sentences that produce the best audience effect. In any case, speaking involves showing—displaying, exhibiting. In speaking I reveal my thoughts, parade my desires, and exhibit my intentions—I show you sentences from which you make suitable inferences. Sign language is exemplary in this respect: here the speaker uses her hands to show the audience signs that communicate states of mind—and it lookstheatrical. The poet, too, shows you words and sentences that convey ideas and emotions (the love letter is similar). The language teacher may proceed by exhibiting sentences for the student to learn—while simultaneously saying something.

The point of my saying all this is to supplement the usual conceptual apparatus of speech act theory with another layer of concepts: it is not just about saying, commanding, etc. but also a sophisticated type of action aptly captured by the showing terminology. Nothing like this exists in the work of the later Wittgenstein or Austin or Searle or Strawson or Grice: we just have an etiolated notion of “utterance”. But the speech act as it exists in humans is a more complex and subtle phenomenon than this terminology suggests, possibly because of lingering behaviorist assumptions. The general form of a speech act consists of a double act of showing combined with an illocutionary act of saying (commanding, questioning, etc.). It isn’t just people making noises that other people interpret this way and that but an act of putting words on display with all that that implies.[1]

Colin McGinn

[1] Showing is not opposed to saying but the two form an indissoluble whole: in saying we show and in showing we say. I will also note that showing is a type of externalization: an inner process is externalized as we show sentences to interlocutors—thought is revealed in spoken words.

Time and Truth

Truth relates to time in an interesting way: once a fact obtains a corresponding proposition is instantaneously true. On the one hand is a fact, say the fact that it just started raining at a certain place, while on the other is a proposition (a belief or assertion), say that it is raining at said location, and the second thing acquires the property of being true at the very moment the fact begins to obtain. Generally, when an object acquires a property a propositional entity comes to possess the attribute of being true at exactly that time—there is no time lag—even though the two things may be far apart (even at the other end of the universe). It is customary to speak of facts as truth-makers, so we can say that facts make propositions true instantly. Facts and true propositions are not identical, yet facts can confer truth on propositions. The truth of the proposition is a consequence of the fact, though it is a consequence that takes no time (rather like logical consequence and unlike causal consequence). This can seem puzzling—like the “non-locality” spoken of in connection with quantum theory. How can the fact manage to reach across space and make a proposition true without any temporal delay? For it is not just that the fact obtains and the proposition is simultaneously true but that the proposition’s being true is a result of the fact obtaining. It would be different if the fact were simply identical to a true proposition, but that is evidently not the case, since facts are not true propositions and true propositions are not facts (they involve different ontologies). No, these are entities of different types, yet one can influence the other at arbitrarily large removes. (One might indeed see the puzzle as a motivation for regarding an identity theory of facts and truths with more favor.) In any case, it appears to be a fact that facts and truths are connected in this way—by a kind of instantaneous action at a distance. Notice that objects and singular terms are not likewise connected: objects don’t make terms refer to them—they are not “reference-makers”. We are more inclined to speak of terms as determining the object referred to, not of objects as determining the reference of terms. In the case of truth, by contrast, the entity in the world contrives to bestow the semantic property of truth on the extraneous propositional entity, which may be an utterance at some remote location (a different galaxy, say). The reason the proposition is true is simply that the fact obtains (an object has a certain property): that is the explanation of its truth.

I won’t say anything more about how to resolve this puzzle, or even whether it really is a puzzle; I will simply take it for granted that truth has the property in question, viz. that truth is conferred at the exact time that the reported fact comes to obtain. My purpose is to use this property of truth to undermine certain ideas about the nature of truth. The property is thus not a trivial property consistent with any theory of truth but rather has polemical teeth. Suppose we try to identify truth with verification: then truth will turn out not to be simultaneous with the fact stated. For verification takes time and is generally subsequent to the time of statement. Suppose that at time t I say that there are five oak trees in my garden; and suppose it takes five minutes to verify that this is true: then the statement will not be true until five minutes after t, according to the thesis that truth is verification. There were five oak trees in my garden at t and so my statement was true at t, but the statement was not verified to be true until five minutes later, which would make it true then according to the theory that truth consists in verification. The only way to avoid this is to claim that there were no oak trees in my garden at the time of utterance and that there only came to be five oak trees at t + 5 minutes. But surely we want to allow that facts can be verified after the time at which they actually obtain (now the fact, later the verification of the fact). Truth arrives at the time of facts not at the time of the verification of facts, so we can’t tie truth to verification by identifying the two. This is why it makes sense to ask how long it will take to verify a proposition, but it makes no sense to ask how long it takes for a proposition to be true. Verification takes time, but it takes no time for a proposition to be made true. Verification is an activity spread out in time, but making-true is not similarly spread out in time—it happens instantly. We can say that a proposition can be verified as true, but we can’t say that a proposition can be true bybeing verified; the two concepts are logically quite different, as is shown in their relation to time. Similarly, we can say that a procedure of verification is time-consuming or that it was executed slowly, but we can’t say anything like this about truth (“It took so long for his assertion to be true”). We might choose to replace the concept of truth with the concept of verification, but we can’t claim to analyze the former by the latter. Simply put, verification is a temporal concept but truth is not. Such ideas as that truth is an “epistemic concept” fail because of this obvious point: evidence gathering is essentially temporal but truth is essential atemporal (in the sense intended here).

The same goes for pragmatist theories of truth that seek to identify truth with something like “convergence of inquirers’ beliefs in the long run” or “what leads to human satisfaction”. These things also take time, sometimes a lot of time, but truth takes no time. Any theory that identifies truth with the effects of belief will fall foul of this point, since effects occur subsequently and are spread out in time. Statements are true or false at the time you make them, depending on the facts, not at some later time. Intuitively, the facts immediately stamp propositions as true or false at the time they obtain no matter what the future holds; so any theory of truth that ties it to future facts will fail. Of course, our judgments of truth are enmeshed in time, being dependent on verification procedures, but truth itself is not—truth itself depends entirely on the prevailing facts (and suitable truth bearers). Truth comes straight from the world in the blink of an eye (so to speak). It is a peculiar property in this respect, and maybe a puzzling one, but any theory of truth needs to accommodate it.[1]

[1] We could add to Tarski’s famous formula the following truth about truth: “snow is white” is true at the moment that snow is white. It would be different if snow had to send a signal to the sentence to inform it that snow is white, since that would take time. No, truth is conferred on the sentence at the very second the fact obtains.