Rethinking Understanding: Wittgenstein versus Chomsky on Meaning

It is no coincidence that Wittgenstein’s investigation of a concept as philosophically perplexing as understanding is preceded by a lengthy and unconventional discussion of what counts as philosophical inquiry, of what the job of philosophy actually is. In fact, Wittgenstein gets his insights on understanding precisely by employing the therapeutic methods whose virtues he had just posited. So, to approach the debate most productively, the stretch of §143-184 will be critically evaluated as a test case for a grammatical investigation as such. This brings out two central questions: 1) Why does Wittgenstein believe understanding is not a mental event, process or state? 2) Is a grammatical investigation, on its own, able to render such commonplace (perhaps even intuitive) views incoherent? To get to the answers, I will first give a brief overview of Wittgenstein’s negative account by running it through three assumptions about the categorical nature of understanding (as experience, process, and conscious state). Then, in considering unconscious dispositional states, I will discuss where Noam Chomsky’s theory of meaning goes wrong. Ultimately, my argument in favor of Wittgenstein is just an attack on Chomsky’s theory by means of a dissolution of the apparently mistaken presuppositions that underlie them. I conclude with a brief discussion on where this leaves theories like Chomsky’s,and where it leaves philosophy in general.

It seems plausible to infer, based on intuition or our own experiences, that understanding happens in the mind. For one, the concept of understanding is clearly tied up with that of meaning; if we meant our words a particular way, and if an interlocutor did not understand them the way we intended, the consistent use of language would apparently be difficult, if not impossible. Since we use our language proficiently, this cannot be the case. After all, it seems like we imbue our cold, meaningless signs with their power, by meaning them a certain way when we say them, or understanding them a certain way when we hear them. This is mysterious, but it seems like it must take place in the mind, as some sort of mental happening. If we fail to understand a word or statement, it seems clear that this imbuing did not occur in us; we have the sign, but lack the meaning. Furthermore, we are familiar with the phenomenon of understanding something instantaneously. When we figure out, for instance, how to continue a series of numbers (or have similar Aha-moments), we naturally conclude that we can go on because we have understood, internally, how to do so. It just feels like an internal process; importantly, it seems that our future behavior flows from the fact that we have understood the series, but the understanding itself did indeed happen beforehand (§146).

Our natural reflections on meaning or understanding, Wittgenstein might say, give us a particular picture of the essence of these concepts, which is that they take place in the realm of the mind. And with this kernel in place, we move on without reservation to the plethora of questions and issues that come spewing out. The first set of issues comes as we attempt to categorize understanding, to explain what kind of mental happening it is. Roughly three significant categories crop up: understanding might be an experience, process, or a disposition (B&H, pp. 367). Philosophers and linguists have devised intricate theories of meaning on the basis of many of these suggestions, but Wittgenstein will attempt to undermine them all. The reasons he gives for rejecting each such categorization, he hopes, will become apparent upon a close inspection of how ‘understand’ is used/abused in language games that feature the categorization. The upshot is that we will come to see that we are putting the concept in the wrong category in grammar. As I see it, the author therefore hopes to disabuse us of those natural kernel ideas we initially had, bit by bit, slowly revealing to us the right category. The task at hand is to evaluate whether he succeeds.

To see understanding as an experience is quite natural, and the pervasiveness of feelings of sudden comprehension seems to support it. To dissolve this notion, Wittgenstein asks the reader to consider what really happens in such moments: if someone suddenly grasps the rule of a mathematical series and exclaims: “I can go on!”, this is no doubt a unique experience; but they could have grasped the wrong rule, said the same thing, had the same experience — and not understood the series. Perhaps more importantly, the right formula to continue the series might have come to mind, but they still did not understand the series (§152). Or they could have said nothing, perhaps found the task trivial, and just continued the series without recalling any special feeling after having done it (§151, §179). What Wittgenstein infers is that the occurrence of the right formula is not understanding, for it may have not been applied; and simply stating that one understands is not a criterion for actually having understood something. And indeed, it seems that we will judge whether the pupil truly understood the series by what happens after his exclamation or non-exclamation of understanding. Our propensity for thinking that the exclamation arises as a symptom of some mysterious comprehension, and that the symptom is a genuine report of this mental activity, is therefore mistaken. It seems right, then, that understanding as experience is a categorical mistake. What about understanding as a process in the mind? After all, the phenomenology of hearing our mother tongue, for instance, is quite distinct from listening to someone speaking a foreign language which we do not speak (B&H, pp. 370). But is this postulated process not just a sequence of experiences, none of which would help us determine understanding (§148)? Regardless, we would not be able to say that a particular experience, or the characteristic sequence thereof, constitutes understanding. Any process (in the right use of that word) has a duration, and can be interrupted. But we would not say, e.g., that understanding the multiplication table is somehow a process extended over time.

To use understanding in such a way is, once again, to categorize it incorrectly; understanding is rather like a gradually-learned skill. Wittgenstein reiterates: if the listener understood what was said, they would be able to recount, paraphrase, etc. after the fact. When we are interested i n understanding, therefore, we are not interested in some internal experience, but what a person is able to do.

We ought to stop and evaluate what Wittgenstein has accomplished so far. By laying out the grammar of ‘experience’ and ‘process,’ he believes that he has shown that ‘understanding’ is not appropriately placed in either category in our language. But if he has done so, which I am inclined to believe, is this tantamount to disproving that understanding exists as a mental experience or mental process? Here, our third category appears in the form of an objection: one might argue that a grammatical investigation cannot reach beyond the language itself, and categorical limitations can be escaped in practice. How so? Well, as a first attempt, we might come to know that the brain fires in a similar way every time we have an aha moment, for example. Why should we not categorize such mental or neural patterns as understanding? Here, Wittgenstein will rebuke: there are performances that exhibit understanding, and there are instances where the pattern obtains. But we can conceive of the pattern obtaining, us asking follow-up questions to the subject, and them clearly failing to exhibit understanding (§149, B&H pp. 369). The critic is not satisfied: he will suppose that the pattern obtains every time that understanding is exhibited. If so, Wittgenstein will retreat to the domain of language once more, and argue that observing the neural pattern itself — the one emergent when someone understands something — presupposes a commitment to a certain grammar of ‘understanding.’ But understanding something is not a state of the brain or a mental state, in the same way that the ability to read is not such a state (§156); put quite crudely, we cannot see the ability to read on an MRI scan. The supposition that we could see such a pattern obtaining is in fact a thesis about the nature of understanding, or rather a presupposition that it is traceable to a mental event in the brain; it is “this way of talking that confuses [us]” (§154), and so our qualms begin and end in language. This realization is quite significant; it seems to support the grammatical enterprise as one that is capable of deriving important insights into concepts on its own, without empirical knowledge.

By exposing the grammar, Wittgenstein is able to show that understanding does not belong to any of these three categories. Instead, we find that the right way to look at a manifestation of understanding is as the exercise of an ability. The processes of making the grammar of a category explicit, and of disavowing the reader of the idea that understanding is a mental happening, are concurrent. The grammatical method’s resistance to objections that purport to take understanding outside of the realm of language seems to confirm its virtue; namely, that it alone can tell us about the nature of concepts. The upshot is that the grammatical inquiry is sufficient to deny the existence of a mysterious mental mechanism which explains understanding or other ‘mental’ processes.

Yet we still find significant resistance to Wittgenstein’s grammatical inquiry, and indeed to the notion that understanding is not in the realm of the mind. Noam Chomsky’s theory of meaning (which might also be called Universal Grammar theory) trades on the idea of understanding manifested via unconscious dispositional states. I paraphrase his innateness hypothesis: there exists a system of mechanisms and principles that is active in the acquisition of grammar, which itself is a subsystem of rules that determines the formal and semantic properties of sentences; such a system is an innate property of the human mind. It forms a set of conditions and rules that are elements of all human languages by biological necessity (Chomsky, pp. 31, 77). Understanding, on his account, is an unconscious process in which sentence-constituents are analyzed, their contributions to the general meaning computed, and the general meaning thus derived. Chomsky is most interested in how humans can understand sentences they have never heard before. The fact that we can do this in spite of our limited number of training data gives Chomsky confidence that we must possess this sort of innate mechanism. He evidently took issue with the fact that Wittgenstein’s account failed to provide an answer to his central question. And we can indeed agree that Wittgenstein did not provide an explicit answer to how we understand sentences we have never heard before; but what we ought to have taken away from his investigation is that such a question is an abuse of our grammar. To see what is meant here, we only need to go back to Wittgenstein’s rejection of the three categories imposed on understanding. It’s clear, in the form of the question itself, that Chomsky presupposes some of those kernel beliefs Wittgenstein attempted to clear out through the grammatical investigation. In asking about how we can understand sentences we have never heard before, a certain view of the grammar of the word is already supposed. Namely, that it is either an experience, state, process or some sort of mental happening (Baker and Hacker elucidate this, LSN pp. 347). But it is clear from our prior discussion that this is subverting the grammar of the word. In reality, the criterion for understanding just is the publicly observable action, coupled with its relevant antecedent skills and later justification. No inner act, process or state, whether it is conscious or unconscious, is a necessary or sufficient requirement for understanding (B&H, LSN pp. 348). With this back in mind, the question dissolves into something else entirely.

To put Chomsky’s error another way, look at Wittgenstein’s example of the novice and expert reader: “What goes on in the [two readers] when they utter the word can’t be the same. And if there is no difference in what they are currently conscious of, there must be one in the unconscious workings of their minds, or, again, in the brain. So we’d like to say: There are, at any rate, two different mechanisms here! And what goes on in them must distinguish reading from not reading. But these mechanisms are only hypotheses, models to explain, to sum up, what you observe.” (§156, my emphasis) Chomsky, we might say, places far too much emphasis on the models that explain what he observes; but they are at most inferences, and they have no substantial support. The truth of the matter is unlocked by considering the grammar of the language: “There is no deep question of how it is possible for us to understand sentences we have never heard before, only deep confusions about the concept of understanding.” (B&H, pp. 356) Chomsky exhibits that propensity toward mentalistic accounts that the grammatical inquiry was deployed to remove, and turns the dogmatism into a theory of meaning. But then, if Wittgenstein is right, where does that leave Chomsky’s intricate account? Or better, where does it leave any theory of understanding/meaning? It seems that the grammatical investigation, if taken seriously (and I have shown that it ought to be), threatens as many accounts as conflate understanding with mental activity of some sort. It therefore seems to me that any attempt to attack Wittgenstein’s ideas on understanding might only be effective in the form of an attack on the very method of philosophizing. If our initial arguments against this philosophical method do not hold water, we are forced into submission, or moved to ingenuity.