Notes on Nonsense

Notes on Nonsense

We don’t talk about nonsense enough. Let’s show it some respect. Nonsense belongs to language not reality: there are no nonsensical facts or objects or properties; there are only nonsensical words or strings thereof. Reality itself is completely…what? We have no word for the opposite of “nonsensical”—the word “sensical” exists neither in ordinary discourse nor in the OED.[1] I don’t know why this is; it ought to make perfect sense (be sensical). In any case, reality lacks the property of being nonsensical. With respect to language we can distinguish two kinds of nonsense: the grammatical and the ungrammatical. The grammatical kind is exemplified by “Colorless green ideas sleep furiously” and “Twas brillig and the slithy toves did gyre and gimble in the wabe”; the ungrammatical kind is produced by flouting the rules of grammar, as in “It up dog random” and “Rainbow the over”. The first thing to notice is that nonsense exists in close proximity to sense: it’s easy to get from sense to nonsense. The same mechanisms that generate sense can generate nonsense: either rules of grammar or simple word concatenation. The grammatical nonsensical strings obey normal grammatical rules and merely juxtapose clashing semantic units, or else employ nonsense words in a grammatical form. The ungrammatical cases simply join perfectly meaningful words that don’t clash with other words. Nonsense doesn’t arise by going completely outside the normal workings of language; it occurs within language. This is why it is wrong to describe nonsense as simply lack of meaning or sense: random squiggles or sounds are meaningless (like bricks and mortar) but they are not instances of nonsense. Nonsense presupposes functioning language. In fact nonsense is a type of meaning not a lack of meaning—the nonsensical type. There is a lot of meaning in the sentences I gave earlier; they are not semantically lifeless. They have, we might say, meaningless meaning—a second-class, degraded kind of meaning (“para-semantic meaning”). What the sentences express is neither true nor false—it is not “propositional”—but it is imbued with meaning of some sort. This makes them puzzling from a theoretical perspective: how do standard theories of meaning apply to them? How, say, do truth conditions theories of meaning apply to grammatical nonsensical sentences, or Gricean theories, or use theories, or verification theories? Here we seem to have a type of meaning that violates all such theories.

            The question becomes sharper when we ask whether nonsense possesses a logic. In the case of ungrammatical nonsense we can rule this out, since logical relations need at least the semblance of statement making; but it is clear enough that grammatical nonsense exhibits logical properties. Such sentences can be conjoined, disjoined, negated, and put into conditionals; and the normal logical rules will apply. For example, conjunction elimination is still valid, or modus ponens. From “All toves are slithy” and “Socrates is a tove” we can infer “Socrates is slithy”: the form of the sentences allow this inference irrespective of the content. But we can’t say that validity here is a matter of truth-preservation, because nonsensical sentences are never true (or false). The meanings (or quasi-meanings) entail one another, yet there is no truth-value to be preserved. We know that if green ideas sleep furiously then they sleep, but it is neither true nor false that green ideas sleep furiously. So is our usual approach to logic too narrow? Do we need a special logic of nonsense? Some have urged the need for a “para-consistent logic”; do we also need a “para-semantic logic”? We need a logic that can handle category mistakes (“The number 2 is cheerful”) and analytic falsehoods (“Janet is a happily married bachelor”); and it looks like we also need a logic that can handle outright nonsense. It is possible to reason with such sentences, so they ought to fall within the scope of logic. But if nonsense has a logic, it must be meaningful.

            Can nonsensical expressions refer? Or better: can speakers refer using nonsensical terms? Can we contrive a Donnellan case in which a speaker picks out an object for an audience even though the term used is pure nonsense? Sure we can: someone may remark at a party, “The slithy tove in the corner is a famous philosopher”, thereby picking out an individual of slippery but dapper appearance (or just a guy known to like the works of Lewis Carroll). The definite description “the square root of Paris” should receive the same semantic analysis as “the Queen of England”: semantically these are expressions of the same general category. The demonstrative “that colorless green idea” functions as a singular term subject to a Kaplan-style analysis despite its nonsensical status. Couldn’t we introduce a proper name “Zippy” by stipulating that it denotes whatever “the square root of Paris” denotes, viz. nothing? True, there are no nonsensical existing entities for such terms to refer to, but language doesn’t know that; it allows us to generate nonsense expressions of all semantic categories. Some of this nonsense may even have a use—a role in a language game—and may even be used to refer to ordinary things, as with that slippery dapper chap in the corner. Nonsense does not preclude acts of reference and other linguistic practices—whole books may be composed of it (Finnegan’s Wake). And nonsense poems are precisely poems.

            Now that we have a feel for the reality of nonsense (if I may put it so) we can pose some more adventurous questions. Could there be an ideal language that precluded the possibility of nonsense? It is hard to see how there could be: language allows for unlimited grammatical (and ungrammatical) combinations—that is part of its inherent creativity—and this will always permit the possibility of nonsensical combinations. Hence the proximity of sense and nonsense in the mechanisms of language production; the same thing is capable of producing both. Nonsense is as embedded in language (as a formal apparatus) as sense is. We don’t have much use for nonsense most of the time, but it is always latent in the linguistic system: “colorless green ideas” is as much part of language as “bright red flowers”. The notion of a logically perfect language incapable of producing nonsensical monsters is therefore an illusion. Grammar is nonsense-neutral (of course not with respect to ungrammatical nonsense). We could even say that meaning itself is nonsense-neutral, since nonsense sentences have a kind of meaning. Then here is a second (vertiginous) question: Is there room for a brand of skepticism that questions the meaningfulness of our discourse? I don’t mean the familiar positivist-Quine-Kripke-Wittgenstein types of semantic skepticism, but rather a type of skepticism that dangles the possibility that we are talking nonsense all the time, despite our belief that we are making sense. Some of Lewis Carroll’s characters talk nonsense without knowing it—might we be in like case? That is, is the belief that we are talking sense fallible? I say, “Little white lambs sleep peacefully”: have I said something analogous to “Colorless green ideas sleep furiously” without knowing it? Don’t I sometimes say things in my dreams that I later realize were nonsensical? Might I be dreaming all the time and hence possibly talking nonsense constantly without knowing it? It sounds impossible to maintain that my simple utterances might actually be nonsense, but the skeptic is a resourceful enemy: can I be certain that “It’s raining” isn’t nonsense? Is this as certain as the Cogito? Maybe my brain is generating nonsensical strings and then disguising them as making sense. The mind can play peculiar tricks. Haven’t some people been totally convinced that their utterances make sense and yet upon examination they turn out to be nonsense (the holy trinity, Newtonian absolute space and time, the unrestricted concept of a set)? Nonsense can be a sneaky thing. So maybe no one has ever said a “sensical” thing ever—all is nonsense. In the beginning was the nonsense word. Maybe “I think, therefore I am” is itself a piece of nonsense! Judgments of what makes sense don’t seem immune to skeptical doubt. The Wittgenstein of the Tractatus was fond of saying that many of our ordinary utterances are strictly nonsense—what about an ultra-Wittgenstein who thinks that all our utterances are really nonsense? It’s all “slithy toves” and “colorless green ideas”.

            It has been maintained that a principle of charity must govern our practice of interpretation: we must find the other largely logical and truth believing. Some have wished to weaken this to a principle of humanity: we must find the other rationally explicable, though not necessarily logical and truthful. But we can picture a further weakening to allow for the possibility of the nonsensical other—the alien who talks a lot of nonsense, perhaps complete and total nonsense. Couldn’t we be forced to conclude that the linguistic behavior of the alien consists mainly, or wholly, of sheer nonsense? Isn’t this what Alice concludes about some of the aliens she encounters through the looking glass? I don’t see why not: perhaps our target tribe has a malfunctioning brain (by our standards) that produces only nonsense; perhaps they utter nonsense all the time just to amuse themselves; perhaps there is a religious taboo prohibiting sensible utterance. In any case they spout nothing but nonsense from dawn till dusk: it’s all jabberwockies and bandersnatches, relieved only by the odd colorless green idea. There need be no assent to any of this nonsense on their part, no “holding true”, and hence no route from such data to an assignment of meaning; still, they do mean something by their utterances, even if we deem it a kind of substandard meaning. Their spoken language has meaning of a sort—it isn’t devoid of all semantic content—but it doesn’t map onto ours in the way envisaged by proponents of the principle of charity. As Wittgenstein would say, they play a particular language game, and within that game language has a use, a purpose. They may think perfectly sensible things, and know that others do too, but their actual speech is knowingly made up of nothing but nonsense; they may wonder at the pedestrian verbal ways of sense-making speakers such as ourselves.[2]

            Nonsense is a kind of robust semantic presence not a mere semantic absence. It isn’t the lack of meaning but a special type of meaning. Philosophers of language have gradually expanded out from what they conceived to be central cases of meaning (usually verifiably true sentences) to other types of meaning (imperative meaning, performative meaning, context-dependent meaning, emotive meaning, etc.); I am suggesting we expand out a stage further to include nonsensical meaning. Even Wittgenstein, with his inclusive notion of the language game, didn’t see fit to include nonsense as a legitimate form of meaning, but there are good reasons to bring nonsense into the semantic fold. Talking nonsense is one form of talking, one way that language manifests itself. And nonsense is as much part of language as sense; indeed it exemplifies the creativity that is the essence of language. Perhaps we should do more of it.

[1] Actually the word “sensical” is not unheard of, but it is not generally accepted as part of the English language.

[2] Could there be a completely nonsensical conceptual scheme? Now that is pushing it: how could the language of thought be composed of nothing but nonsense? Could all thought be inherently nonsensical? In our case there is always a bedrock of sense on which nonsense is parasitic, but in the case of the nonsensical conceptual scheme it is nonsense all the way down. This is hard to make sense of (perhaps it is a species of nonsense).