Thought and Time

                                                            Thought and Time

A speech act occurs in time: it has a beginning, middle, and an end–it can be clocked. The more words a sentence contains the longer it takes to say it. Some people speak more quickly than others, thus taking up less time to utter the same sentence. Writing is the same: you start on one side of the page and scrawl your way across to the other side, all this taking a certain amount of time. Call this generative process sequential assembly: the utterance (by mouth or hand) proceeds by starting with an initial element and then adding elements until the complete act is performed. Evidently, this is because of the nature of the generative process—making sounds with the mouth and marks with the hands (or shapes in space for sign language). We can’t make two (or more) sounds at once and we can’t make two (or more) marks at once. Hence utterances unfold in time. Parts of speech acts occur at different times.

            But what about thought—does it result from sequential assembly? The answer is not immediately obvious, as it is for speech and writing. What about inner speech when words “go through the mind”? This does seem to take time and to be sequential: silently say the sentence, “Monogamous marriage is doomed to failure” to yourself, being careful to enunciate each syllable. The process begins with the syllable Mon and ends with the syllable ureand takes a definite amount of time (about 2 seconds in my case). It is tempting to see such silent speech as the internalization of noisy outer speech, so we might well suppose that the temporal properties of outer speech are preserved on the journey inward. The organ of speech articulation operates sequentially and it preserves this design feature when deployed inwardly. We certainly don’t want several auditory images, corresponding to different words, to co-occur in consciousness—the inner sounds need to be separated in time. It is a bit puzzling why inner speech should mimic outer speech in this way, given that we are not using the vocal organs to generate the sentence in question, but it seems to be the case.

What should we say about non-verbal thought? You have a thought and no sentence of your spoken language runs through your mind; the thought is purely conceptual. It does not appear that such a thought is assembled in the sequential manner described so far: the process does not proceed by first producing an initial element, then adding a further element, then repeating this operation till the thought is complete—all this occurring over a finite period of time. The thought seems to be produced by simultaneous assembly—all at once. Some generative process selects a set of concepts and then operates on them simultaneously to produce a complete thought; there is no time at which the thought is only half finished or close to being complete or just getting started. The thought may take some time to form, but it appears to come fully formed, not in dribs and drabs. It is as if the thought is projected all at once onto a screen from an array of concepts stored in the mind. Verbal utterances are likewise assembled from a store of elements, but they come into existence in a temporal sequence; thoughts, by contrast, appear to arise without temporal sequencing. The machinery of thought production must thus operate by different principles from the machinery of utterance production.

It might be supposed that this difference counts against the idea of a language of thought, but that presupposes that linguistic operations are always temporally sequential—that sentence construction must occur in linear order. However, sentences could be constructed objects and yet not constructed by an ordering from beginning to end: instead all the pieces are slotted into place simultaneously. Then thought will follow this process of production, apparently emerging without construction (we don’t introspect the opening segment of a thought and then wait for the later segments to appear). But we can infer that this property of thought counts against the idea that the language of thought mirrors outer language in its mode of production: assembling inner sentences to act as the medium of thought is not any kind of utterance—if that means some kind of temporally extended process. Thinking in language is thus not a form of speaking to oneself—it is not inner saying. It is inner symbol manipulation, to be sure, but it is not temporal concatenation—not adding one symbol to another over time. This also means that it is not like inner speech, properly so-called, even unconscious inner speech, which is sequential. To put the point differently, the brain produces sentences in the language of thought by means of a parallel process: it inserts symbols into sentences simultaneously—in parallel. For example, it might take the set consisting of “John” and “runs” and simultaneously insert these words into a sentence frame of the form Fx, where this represents the grammatical subject-predicate form. It doesn’t insert one and then the other, but both together—and similarly for longer sentences. Putting it biologically, we have an innate faculty of simultaneous linguistic assembly, which presumably evolved at some point, and this faculty is what enables us to think as we do.  [1] At a later point, when we acquired spoken language, we began to employ sequential assembly; but the language of thought itself operates by means of simultaneous assembly (which is not to say that the process is instantaneous).

Two questions may be raised about this picture, neither of them easy to answer. The first question is whether it is conceivable that thought should have a sequential mode of assembly: Are there possible beings that think in the way that they speak? Could there be fast and slow thinkers as there are fast and slow talkers? Could someone start a thought and be interrupted in the middle of it, like uttering only the beginning of a sentence? Could there be thoughts that take a full hour to put together? Would it possible to think John runs spread out over a period of thirty minutes? Can you intend to have a certain thought but then fail to have it through lack of time? These things certainly sound strange, but is that just because we are so used to our simultaneous mode of thought production? Maybe Martians evolved a sequential style of thought production, so that they think much as they speak.  

The second question runs counter to the first: am I not exaggerating the simultaneity of thought—is it perhaps just that thoughts form very quickly, so that we don’t notice their temporal character? After all, it may be said, there are sequences of thoughts, as in a stretch of logical reasoning, and these sequences are spread out in time—so why aren’t the individual thoughts that compose them also spread out in time? Also: what about extremely complex thoughts—don’t they take time to reach their conclusion? Here we have to ask whether such duration as there is to thoughts results from sequential assembly or from some other source. Thoughts may well take time to form (the brain needs time to do its thing), but are they assembled in the manner of spoken utterances? When someone has a conjunctive thought, say, do they first think the first conjunct, then insert conjunction, and finally think the second conjunct? That sounds wrong to me, as if the thought isn’t conjunctive until the pieces are all glued together over time—wasn’t it conjunctive from the start? Or rather, there is no start, just a fully formed conjunctive thought taking shape in the mind. As to reasoning, it is consistent to suppose that the arrangement of thoughts into an argument is a sequential temporal process without supposing also that thoughts themselves are sequentially assembled. True, we don’t engage in parallel reasoning, with premises and conclusion simultaneously before the mind, but thoughts always involve such simultaneous combination—the constituent concepts are presented as a unified contemporaneous whole. Our minds may range from one thought to another over time, more or less logically, but we don’t move from one concept to another as we cobble together a thought. We construct arguments over time, but we don’t construct thoughts over time. We don’t build a thought as we build a house—one brick after another. But we do build speech acts this way, by adding one component at a time. Thus speaking and thinking are different kinds of act, even though language is involved in both (assuming a language of thought). Utterances have beginnings, middles, and ends; but thoughts leap into existence all at once.

  [1] We could compare this ability to the way the brain constructs the visual field—not by painstaking construction from left to right but by massive parallel assembly. All the parts of the visual field are presented simultaneously (and produced simultaneously); it is not that we only produce one section of it at a time. So seeing is not like speaking, in which we are restricted to one sound at a time.