Thursday, February 28, 2008

A counterexample to the Private Language Argument

Wittgenstein's Private Language Argument contends that it is impossible for one to form a language by oneself. Here's a counterexample. You live on an island that has no people other than yourself. You live for forty years there. Then you step in a time machine that was left there by someone else. You go back forty years. You do this 999 times. (Let's assume that the time machine also fixes up your body so you can live for a subjective length of 40000 years.) To an outsider, the island looks populated by a thousand people of remarkably similar appearance. There is a community there. But that community in fact includes only one person, you. So, it's possible to have a community with only one person. But there is no reason why such a community couldn't develop a language, since it functions just like all other communities do.

A fun question: Suppose Marcy and George join you on the island, but they don't do any time travel, and so there you are located in 1000 places on the island, and then there is Marcy and there is George. In elections, should you get 1000 votes, with each of them getting only one—there does, after all, seem to be a sense in which you have a lot more interests—or should each of the three people on the island get exactly one vote?

By the way, the clip below illustrates the wrong way of imagining the scenario. In my scenario, it is a mistake to think of first the island having you in one place, then of it having you in two places, and so on. Over the 40 year period in my scenario, you always are in 1000 places.


Heath White said...

A fun thought experiment ... but not really how I read the PLA. Somewhere Wittgenstein says (I am too lazy to look it up now) that there is no objection to having a private language as one might have a private sewing machine--a tool which only you use, but which in principle could be used by many. Rather, the argument is supposed to show the impossibility of a language which made reference only to in-principle private phenomena, e.g. sense-data, one's thoughts, pains, etc. (That is, it's an argument against phenomenalist empiricism.) Maybe there is some way your example could be fixed up to accommodate this--I frankly don't know what to say about the privacy of sense-data of multiple versions of myself at a time!

Alexander R Pruss said...

You may be right--I'm embarrassed.

But what if we enrich my story with telepathy?

Here's the idea. First, start with an island of a thousand distinct people who have telepathy. It's not implausible that they can have a public language just about sense-data, thoughts, pains, etc. For for them, sense-data, thoughts, pains, etc. are not private phenomena! So that's not a counterexample to the PLA.

But now vary the case to be like my original one. There is just one person on the island, but he's there a thousand times over, duplicated via time travel. Moreover, he has auto-telepathy: he can sense the private phenomena of his past and future selves whenever these past and future selves are contemporaneous with him. Linguistically, this island can function just like the preceding island, and develop a language for talking about these phenomena. But now the phenomena are once again private!

Maybe Wittgenstein would deny the possibility of the relevant kind of auto-telepathy, or say that whatever is subject to it is not really private. (Actually, thinking about the possibility of telepathy makes me wonder if anything is in principle private.0

Paul M. Rodriguez said...

Can the distinction between public and private language be made in any but a public language? That is to say: wouldn't your iterated telepath's language be private only so long as he was unaware of the existence of other people outside of his island? If he knew of them, his language would have a word for "other"; and if the language had a word for "other", then a stranger could in principle speak to him in his now-public language by prefixing every word with "other": "other-pain" &c. Conversely: wouldn't a person who lacked a "theory of mind" (say, if the islander were autistic--or even a dogmatic solipsist), even if he were to use a "public" language, be using it as a private language, with reference only to private experience?

Alexander R Pruss said...

Good point!

There are three variants of my scenario:
1. The iterated islander erroneously thinks there are 1000 people on the island. Then the language is non-private.
2. The iterated islander knows there is only one person on the island, but has the idea of people not on the island. It could, however, still be that he is unable to express the idea of a sensation had by anybody off the island. He may not know anybody else is capable of having sensations. His language may be such as not to allow any possessive qualification of sense-data attributions.
3. The iterated islander believes it is not possible for anybody other than himself to exist, and his language recognizes this.

Maybe the way Wittgenstein could now respond is this. In case (1), we have a public language. And it is not clear that language is at all possible in either case (2) or case (3). What makes us think that language can develop on this island is that we confuse (1), on the one hand, and (2) or (3), on the other. That seems to me to be a powerful rebuttal.

Alexander R Pruss said...

By the way, here's one possible answer to my question about island democracy. Your vote should be weighed more heavily than the votes of those who are not iterated, because you spend more time on the island, so you are more heavily affected how things are there.

But if this is right, then older people should have a vote that counts for less than the vote of younger people. That doesn't seem right.

Kyle said...

Hi Alexander,

I don't think what you have said would mean that younger people's votes should be given a greater weighting.

Suppose you are voting to elect a representative for one month. During that one month, you will experience 1000 months, but each of the new comers will only experience 1. The same is not true in the case of old and young people.

Another senario in which this issue might arise:

Suppose in the future it is possible to download a person's brain into a computer, and this person's mental life continues and we give them a vote. Now imagine that the computer that they are on is upgraded and their brain runs at double speed. They would experience two months of subject time, for each month of ours. Should their vote be given double weighting?