To: Michael W.
From: Geoffrey Klempner
Subject: Does thought entail possession of language?
Date: 29 April 2003 10:33

Dear Mike,

Thank you for your e-mail of 20 April, with your second essay for the Philosophy of Language program, in response to the question, 'Does thought entail the possession of language?'

You pose the question in the context of reflections about the impossibility of attaining a neutral standpoint. The idea of attempting to 'compensate' for your particular standpoint in the world - including, crucially, the fact that you are a language user rather than a non language user - seems a good one. The question is how one puts this into practice.

How are we to tackle the question about thought and language? One plausible approach is to consider what would be the minimum required to justify the attribution of thoughts to another subject. That is the appropriate methodological stance, if you are trying to 'compensate' for one's own subjective view.

Your strongest argument appears to be empirical rather than philosophical - Varley's study of the a-grammatic aphasic man. I don't remember if I told you that my father had a very similar language problem as the result of a stroke. For thirteen years, until his death in 1998, he continued to cope with things like accounts, remembering birthdays, giving advice (he had no difficulty in saying 'yes' and 'no'). And yet we were left speculating just how much he really knew or understood.

Could this be an empirical question? I have no difficulty with the thought that the acquisition of language wires the subject's brain up to cope with certain kinds of information processing tasks, which the subject is able to continue to do when the language function is impaired.

Connected with this, is your statement that 'At the physical level a thought is a state of the brain over a finite period of time...'. However, there is a difference between saying that thoughts are embodied or realized in brain states, and saying that a thought *is* a brain state. If a thought simply is a brain state, which sets the subject up to be able to perform certain tasks, then suitable experimental tests could settle the issue.

I would argue that this misses out the dimension of normativity. Judgements are true or false, or justified or unjustified. Anecdotal evidence bears this out. My father expressed views, you couldn't argue with him. You could reason out, from the yesses and nos what views he probably held, but you could never be sure. Nor was there any reliable indication of his reasoning.

Two things are linked here: the identification of a thought as being that particular thought, rather than some other particular thought, and the normative evaluation of the thought in question. As well as truth and falsity, and justification or the lack of it, when thoughts are expressed in language there is the possibility of raising the question whether words are being used in their correct sense, or indeed whether the subject really understands what they are saying.

In my argument against attributing thoughts to brute animals, the crucial step was the absence of a 'hook' with which to identify particular thoughts. It is not like hypothetico-deductive explanation where you posit an unobservable inner mechanism to explain given behaviour. In the case of a mechanism, we have a notion of what a particular mechanism is, what it would look like (e.g. under an electron microscope). Whereas the only handle that we have on a particular thought is through its linguistic expression. This is an argument, in principle, against the idea of a machine which could read 'brain writing', and translate it into, say, English.

The case of the aphasic man is especially challenging, because of the temptation to imagine 'what it would be like'. I could be writing this letter in my head even though I was totally paralysed. I would know what I was saying. I could be thinking just as I am thinking now. Realizing that such a move gets you nowhere, you have deliberately set out to assess the question objectively, and that is the correct stance.

We are both agreed that the fact that language is necessary for human development does not settle the issue.

It is interesting that you have brought in the language of the emotions, and body language.

Emotions reflect the natural life of human beings, but they are also anchored in our cultural life. You see that someone is upset, you read their emotion on their face. This looks like a way of identifying particular thoughts without need of conventional language. And in fact we do this all the time. The question one is led to ask is whether it could be done reliably, in the case of a subject who lacked language.

The problem is that we are dealing with an alloy of nature and culture, not the 'substance' in its pure state. It would be possible to imagine a disability (rather like tourette's syndrome) whereby a subject who was perfectly lucid in their ability to express themself verbally, gave the impression of being angry one minute, joyful the next, then despairing and so on. What we would say is that something had gone wrong with the 'wiring' to cause these natural signs of emotions to appear even though the emotions themselves were absent, as the subject was able to confirm to our satisfaction.

- I am impressed by the amount of work you have done for this essay. I do think that you could have assembled something resembling a conventional 'essay' more closely. But I accept that this is exploratory, rather than simply an attempt to 'argue a case'.

It would be helpful if you included references when mentioning sources (like Astington, Peterson and Siegal, Varley).

All the best,

Geoffrey