Wednesday, September 12, 2007

Five Private Language Arguments (International Journal of Philosophical Studies 12, no. 2 (2004))

My paper Five Private Language Arguments is available here.

Comments welcome.

It provides a fairly clear explanation of five different private language arguments that philosophers have supposed Wittgenstein offers in PI 258. I intended it to function as a good introduction to the whole private language argument topic.

If I ever get round to writing a book on the Philosophical Investigations, it will form part of that book.

Incidentally, the clearest intro to the Investigations is still Marie's McGinn's (though I don't agree with McGinn on e.g. the private language argument - see my paper). I have just ordered Block's How to Read Wittgenstein - will let you know what I think.

7 comments:

Timmo said...

Stephen,

If you don't mind the suggestion, I really liked David Stern's discussion of the private language argument toward the end of his book Wittgenstein on Mind and Language. He draws off texts other than the Philosophical Investigations to elucidate what Wittgenstein was getting at, such as Wittgenstein's 'Notes for a Philosophical Lecture'. Stern's close attention to the rest of Wittgenstein's corpus of notes definitely strengthens his reading. (Not to suggest, though, that other scholars just drop the ball on doing this...)

The Barefoot Bum said...

I have to differ: I think Wittgenstein is (relatively) clear, he's correct in general, and wrong in particular.

The problem seems relatively simple using the language of science: There's no such thing as a truly private language, in the sense that—to employ your analogy—the prisoner's association of the marks on his diary with seeing rats is not in principle unavailable: One could, with sufficiently advanced technology, image the prisoner's brain and discover the correspondence. The correspondence is ineluctably "written down" in his brain.

BNC said...

Hello Stephen,

I have just read your paper on PLA and I have a question: you say "

To this Pears may reply that in the prisoner example I help myself to the prisoner’s existing linguistic skills. The only reason the prisoner can immediately introduce a new sign into his vocabulary and then go on immediately and unerringly to apply it just like that is because the prisoner has already learnt the general skill of introducing and using signs in this way."

From this you conclude that (in this context) a PL is possible, but not as a first (that is initial) language.

Here is my question: Wittgenstein's Private language is, of course, a radically private language, it is a language which it is impossible IN PRINCIPLE for anyone else to understand.

Surely the so called "Private language" yet set up above is most definitely a non-readical private language, it is no more than a secret language with no implications at all. There is no logical barrier to the prisoner explaining to me what his language means.

Accordingly the argument misses the mark.

Or have I misunderstood your point?

BNC said...

Hi Stephen,

I made a typo in my post, I wrote "yet" instead of "you"

Here it is again with the typo fixed.

I would be really grateful for a response, especially if I have the wrong end of the stick.

Thanks.



Hello Stephen,

I have just read your paper on PLA and I have a question: you say:

"To this Pears may reply that in the prisoner example I help myself to the prisoner’s existing linguistic skills. The only reason the prisoner can immediately introduce a new sign into his vocabulary and then go on immediately and unerringly to apply it just like that is because the prisoner has already learnt the general skill of introducing and using signs in this way."

From this you conclude that (in this context) a PL is possible, but not as a first (that is initial) language.

Here is my question: Wittgenstein's Private language is, of course, a radically private language, it is a language which it is impossible IN PRINCIPLE for anyone else to understand.

Surely the so called "Private language" you set up above is most definitely a non-readical private language, it is no more than a secret language with no implications at all. There is no logical barrier to the prisoner explaining to me what his language means.

Accordingly the argument misses the mark.

Or have I misunderstood your point?

Stephen Law said...

No it doesn't miss it's mark if the defender of the possibility of a PL cannot explain why the introduction of essential privacy should be a problem for the linguist. What difference does it make that not only can the linguist not check, it is in principle impossible for such a check to have existed?

Apparently, none. The reason that the prisoner exercises a skill is the he is a reliable user of the sign (whether or not he can check). But then why can't the Wittgensteinian linguist be a reliable user?

Kim Tullar said...

Hello Stephen;

I'm taking philosophy at Cherwell school, and due to my interest, have engaged myself somewhat with Wittgenstein. In doing so I have read some of Anthony Kenny's book, and this article (your website being a highly useful source in general on philosophy). As a result, I believe I may have spotted a mistake in your approach to Kenny's argument, which I shall attempt to explain.

As you rightly say, Kenny's argument hinges upon the notion of recalling a sensation. Let us give an example: you are currently experiencing K and wish to check if K = S which was signified by you as "S" yesterday. At this point in time, you cannot remember what "S" signified exactly (cannot remember S), but can remember the notation "S". So, you are attempting to check whether K, which you know, equals something of which you have no clue! It is impossible! Now, it would be plausible to go through memories and check whether they were K or not. However, it is impossible to know whether these K's are the same as the intended S.

It is slightly unsatisfying for me to leave my argument like this, so I shall expand slightly. Firstly, although we do not know what "S" was intended to signify, we can guess what it was intended to stand for. Let us say we guess "S" to signify Q, where Q is a sensation which may or may not be S. Clearly in the act of guessing we are unsure of what "S" actually signifies. Therefore, at that point in time, if we asked ourselves "what does "S" signify?" we must admit that we cannot answer with any certainty. We do not know what "S" equals, but we may think we know, or guess. So what if we say that we are "reliable" and that the chance of us guessing right is very high? Even so, we, nor anyone else, will ever know whether our choice of what "S" signifies was right. Therefore the use is merely a matter of chance, and so clearly not a definition.

The wider point here is that reliability breaks down when no one can ever make a check on the thing which should be reliable. When no check can be made, the thing may or may not be true, as dictated by chance. It is nonsensical to call someone a reliable user if their use is dictated purely by chance. In the example of the prisoner, let us say he created the “!” notation to record rats yesterday. The next day he wakes up, and looks at the notation. He does not immediately know what it means. He consults his memory. He finds that he remembers seeing lots of rats, but also lots of prison guards. He may mistake the “!” notation to mean seeing a prison guard. Mistakes changing the meaning may happen many times. However, this example, and allegory in general, is slightly misleading, as some checks can be made by the prisoner. For instance, instead of using a “!”, the prisoner could draw a rat. Here, there is a similarity between the two different visual senses – the drawn rat and the real rat. Both look very similar and share the same form. But with sensations, there is no possible way of constructing any interior symbol which can share the same form or a similar one as a sensation. Therefore, no check can be made.

Kim Tullar said...

An attempted counter-argument could be as follows: A check can be made if it is assumed sensations have qualities which can be described in terms other than sensations. If so, these terms can be recorded, and therefore referred back to easily. Therefore, instead of using “S”, we may use “sharp, tingly”, and then refer back to memories which similarly exhibit these qualities.

This argument suffers due to its use describing sensations via things other than sensations – it begs the question.

Another attempted counter-argument: When we have a specific sensation S, mark it in our memory – “draw a ring round it in our memory”. Now, when we have a new sensation K, let us feel the current sensation K, and then feel the previous sensation S which can identify itself through the ring drawn around it in our memory. We can check if our sensations are the same.

If we can re-experience sensations via memory, this argument would appear sound; however, its problems arise due to line “mark it in our memory”. Of course this is a subtle way of proposing we signify it, and therefore the argument begs the question. It is possible to go back to a memory of a sensation U and check if K is the same as that sensation, but U is just some random sensation you have gone back to.

To conclude, an illustration. I experience a sensation S today, and record in a book the exact time, place and situation in which it occurred. I record all the details of the context possible. Tomorrow, I wish to remember what S was, perhaps so I can check a different sensation with S. I look back in my book, see the context, and attempt to go back in my memory to S. I find I cannot. Why? Because the memory of S is detached from the memories of the context. Knowing the context tells you nothing of the qualities of S, nor does it point you to any space in your mind around which a group of memories, including S and the context, exist. Therefore, Wittgenstein was right in saying that when we undertake such an exercise, all we result in finding is a term such as “happiness” which stems only from its use in a language-game.