Criteria of Meaningfulness

Criteria of Meaningfulness

The positivists created a question that had not existed before, viz. what is the criterion for whether a string of words is meaningful? Their proposal was that such a string is meaningful if and only if it is empirically verifiable (with due allowance made for analytic sentences). The intention was to place metaphysical sentences on the wrong side of the line—to declare sentences that seemed meaningful not to be meaningful at all. They were trying to expose illusions of meaning—sentences with pseudo meanings. As there is fake jewelry, so there is fake meaning. Verifiability was to be the test of authentic meaning. As is well known, their efforts came to naught for a whole series of reasons (mainly the criterion kept ruling out too much); but the enterprise itself was not cast into doubt. The positivists simply had the wrong criterion—some other criterion would do better. Accordingly, other criteria were mooted: falsifiability, truth conditions, use, inferential role, grammaticality, language-games, etc. Without going into detail all these proposals ran into problems of one kind or another: the existence of meaningful non-indicative sentences, excessive narrowness, vagueness of formulation, lack of selective bite, and so on. Perhaps most instructive is the grammaticality criterion: it sounds eminently reasonable at first but quickly comes to grief. A sentence is meaningful just if it is grammatical, i.e. is well formed according to the rules of grammar. This is generous enough to let in what intuitively counts as meaningful (including metaphysical sentences) but strict enough to rule out nonsense strings and things like rocks. But two questions can be pressed: (a) what makes a word meaningful and (b) by which grammatical rules exactly. Can’t there be meaningless words (“brillig” and the like)? And can’t grammars vary from language to language? What is grammatical in one language may not be grammatical in another. Even if there is a common basic grammar to all natural human languages, what about invented languages or languages spoken by aliens—can’t they be meaningful too? So this proposal also flops. The question seems hard to answer, like many philosophical questions, despite the plethora of attempted answers. Is this one more puzzle to be added to an already long list? The mind-body problem and the problem of meaningfulness—both are resistant to solution.

An array of possible responses to the difficulty suggests itself. Maybe the property of meaningfulness is a primitive property, like Moore’s simple indefinable good; maybe “meaningful” is a family resemblance term with no property shared by all instances of the meaningful; maybe we just haven’t found the answer yet but should keep trying; maybe it is a bona fide philosophical mystery like the mind-body problem; maybe the whole idea of meaning is an illusion and should be eliminated from our thought. Again, I will not discuss these options, noting merely that they all seem like using a sledgehammer to crack nuts. Instead I will defend my own answer: the question is not a sensible question—not as philosophers intend it anyway. It has no interesting philosophical answer. It is an illusory question. We can certainly distinguish between the meaningful and the meaningless: words and sentences are meaningful but rocks and mountains are meaningless. It would be odd to say of a mountain that it is meaningless—it isn’t even a candidate for being meaningful. A mountain isn’t made up of symbols or even of purported symbols: it’s just a big lump of stuff. It isn’t verifiable or falsifiable or grammatical simply because it isn’t symbolic at all. Most things aren’t—they are devoid of meaning, meaning-less. But this distinction is not what philosophers were getting at who sought a criterion of the meaningful—of course rocks and mountains are not meaningful! Whoever thought they were? They give no impression of meaning, no illusion of it; no one is tempted to regard them as meaningful. So the genuine distinction illustrated by the difference between words and rocks is irrelevant to the philosopher’s quest; the philosopher is interested in the distinction as it applies within the class of symbols. But that is exactly where the project runs aground: for anything that counts as a symbol is already meaningful. Recall Grice’s theory of speaker meaning: an utterance is meaningful if it is used to induce in an audience a belief by way of the audience’s recognition of an intention to do just that. But any symbol can be used to achieve that aim just by being a symbol: metaphysicians can certainly induce beliefs in others by using metaphysical strings of symbols. They could even do this by using a system of whistles or hand gestures or even heads on platters (to use Grice’s own example). It is only too easy to be meaningful given that there are speakers intent on communicating thoughts. That’s all there is to it really—what is meaningful is what people treat as meaningful, i.e. use to communicate. What kind of speech acts did the positivists think the metaphysicians were performing as they talked to each other? Obviously they were getting things across to each other, so they must have been speaking meaningfully. And the same for all the other interesting criteria proposed by philosophers. In so far as the question has any answer, it is entirely trivial. The sentences are meaningful because you can mean things by them, i.e. communicate thoughts.

But wait: can’t we reformulate our question to focus on beliefs and thoughts? What is the criterion for having a contentful thought? Now we have created a new concept (and a new word)—the concept of the contentful—and raised the question of when it is satisfied. When does a belief have content and when does it not? Certainly we can sensibly report that some things have content and some do not—again, states of mind do but rocks and mountains do not. But that doesn’t give us any sense in which the distinction applies within the class of beliefs and thoughts—as it might be, beliefs that give the illusion of having content but don’t really have it. And that distinction is barely intelligible: to lack content is to lack a condition necessary for being a belief. What did metaphysicians have according to the positivists—fake beliefs, non-beliefs masquerading as beliefs, illusions of introspection? That is absurd: of course they had beliefs with content just by having beliefs. They reasoned with them, disputed each other’s, and abandoned them under argumentative pressure. But then their sentences were meaningful by Gricean standards, since they served to communicate such beliefs to each other. The idea of a belief just is the idea of something that is contentful, as the idea of a symbol just is the idea of something meaningful (given its employment in an act of communication). So the notion of a criterion of meaningfulness, as sought by philosophers, is an incoherent notion, a kind of displacement of a genuine distinction into an area in which it can’t properly apply.

It is rather like trying to find a criterion of the humorous. It is true that some things are humorous and some things are not: jokes are but funerals are not (or rocks and mountains). But it would be folly to try to find a criterion within the class of jokes: jokes are funny qua jokes—some funnier than others perhaps—but it would be bizarre to suggest that a certain class of things that are regarded as jokes are not really jokes. A joke is a joke and a sentence is a sentence. A joke can lack in humor as a sentence can lack in significance, but that is not to say that either can belong to a class of the literally non-humorous or non-meaningful. It might be said that a joke can be in bad taste or tedious or crappy in some other way, as a sentence uttered by a metaphysician can be trivial or absurd or crappy in some other way; but that is not a matter of being an illusory joke or an illusory meaning—something that someone might be tempted to wrongly classify in these ways. What is humorous is what makes people laugh, as what is meaningful is what makes people believe (according to Grice). It would be strange to say that a metaphysician’s humor is not real humor by some tendentious standard, as a metaphysician’s meaning is not real meaning, or a metaphysician’s belief is not real belief. One might wish to say that such humor or meaning or belief is a waste of time, or is nothing like science, or is totally obscure, or is a mere game with words—but not that it isn’t even humor or meaning or belief at all.

Another way to see what is wrong with the search for a criterion of meaningfulness is to note that it treats meaningfulness as an evaluative notion: it is supposed to be good to be meaningful and bad to be meaningless. The positivists didn’t think that metaphysics is perfectly fine but entirely meaningless! But this attitude is clearly mistaken with regard to things that really are meaningless—like rocks and mountains. These things are not defective through lack of meaning; they simply are not the kind of thing for which we expect meaningfulness. What are supposed to be defective are symbols that lack meaning, since they purport to mean something. It is the illusionof meaning that they present that is deemed deplorable not the mere fact that they lack meaning. That would no doubt be a bad thing—a kind of deceptiveness—but it is not really possible, as I argued above. On the other hand, real lack of meaning is perfectly possible—most things not being symbols to start with—but this is not what the seeker after a criterion of meaningfulness is looking for. There is all the difference in the world between denouncing the metaphysician’s words as meaningless and remarking that his bowtie is meaningless (or the soles of his feet). The philosopher should never have taken the concept of meaningfulness as evaluative in the first place; that notion simply marks a natural division between two kinds of fact, semantic and non-semantic. If the metaphysician were simply practicing elocution by uttering his metaphysical sentences, he would not be disturbed by the allegation that his words are meaningless; what hurts is that he intends thereby to speak deep truths. But this whole idea is misguided because there cannot be a criterion of meaningfulness of the kind the philosopher seeks. A grammar book will tell you that a sentence is a part of speech that serves to convey a complete thought: that is quite correct, and the condition is easily satisfied. No criterion of meaningfulness can be acceptable that entails that sentences in this sense can fail to be meaningful. The concept of meaningfulness is not a philosophically interesting concept, at least as philosophers approached it in the first half of the twentieth century (perhaps the question of whether life is meaningful is a real question). This could be why it was not a topic of philosophical interest before that time (contrary to some projective history of philosophy).[1]

Colin McGinn

[1] For instance, it would be quite wrong to interpret Hume in this way, as some positivists were inclined to do. One might try to claim that Heidegger’s “Nothing noths” is literally meaningless, but (a) Heidegger offers a gloss on that sentence that gives it a kind of sense and (b) if it is meaningless that is because it is ungrammatical not because it is unverifiable. In any case, it is hardly representative of the class of metaphysical sentences (if indeed there is such a class). I note that it is hard to find anything in Wittgenstein’s Philosophical Investigations that can be construed as an attempt to give a criterion of meaningfulness, in contrast to the Tractatus. That whole project was pretty much dead by the latter half of the twentieth century, though the idea of it lingered and was never fully repudiated.