Re: Wittgenstein and Concepts

From: Harnad, Stevan (
Date: Sun May 25 1997 - 19:25:06 BST

PY104: You need not worry your heads about this; it will not be on ther
exam. It's a question someone in philosophy asked about one of the PY104
postings. Chrs, Stevan

> Date: Sat, 24 May 1997 22:21:45 GMT
> From: Counter, SRM <>
> I saw recently that you posted some stuff on Wittgenstein's interesting
> ideas on language. I noticed that you did not agree with most of what
> he said - could you possibly tell me your reasons for why you disagree
> with his work (especially in case no 1 on private language) as I am
> studying philosophy and it would be of much help to me.

What I wrote was this:

>>> (1) There can be no "private" language: a Humpty-Dumpty-like one, that
>> you invent, in which words mean whatever you mean them to mean:
>> Language arises from a community, where there are shared conventions
>> about what means what. Imagine if you had a word like
>> "auto-pleasance," which would mean "looks good to me right now": Not
>> only would no one else have any idea what is and isn't auto-pleasant,
>> but you yourself could not know whether or not something you saw was
>> really auto-pleasant, or only seemed auto-pleasant. There is no error
>> in a private language, and hence no correction for error.
>> (I think Wittgenstein's private language argument is wrong, by the way,
>> but you asked only for his interesting ideas...)

Wittgenstein was right about words that name private experiences,
because in order to use a word in a NON-arbitrary way (Humpty Dumpty's
way is arbitrary) there has to be something other than your private whim
against which to check whether or not you have used the word correctly.

Here it is in a nutshell: there have to be (1) things that your word --
let's call your word "X" -- correctly calls X, and there have to be
things that are not X, that you again correctly call not-X. To be
right, (2) you also have to be ABLE to be wrong. Normally this means that
you have to learn by trial and error (sometimes mistakenly calling X's
not-X, and Non-X's X), with corrective feedback after you have made a
mistake, until you get it right what X's and non-X's are.

Now Wittgenstein correctly says that we need a community of
language-users who all more or less agree what word means what, and
then they can correct you when you use the word wrongly. But if you
invented a private language, no one could correct you -- because no one
could read your mind and figure out whether what you had decided for
yourself was or and wasn't X [in my example I used "autopleasance"]
REALLY was or wasn't an X.

Now this has come to be called the "problem of error."

That much is correct -- for words describing private sensations like
"autopleasance". [And some Wittgentsteinians would say "Fine, that's
all Wittgenstein wanted to claim!" and perhaps they are right.] But
there are further questions we could naturally ask about private
language: This has come to be called "the problem of error": Supposedly
Only a public community of language users can provide this
error-correction service to guide our use of words.

So what if you were abandoned as a child on an island: Could you not
invent a language? Well, first, it's not clear why you would want to: I
mean, who would you talk to? But supposing words, at least, if not
sentences, are useful memory aids, so if you learned to call something
X and another Y and another Z, and you need to rehearse to remember
where those three things were (e.g., in locations called A, B and C) --
they might be special kinds of mushroom that you need to eat at certain
times of day, and their locations vary from day to day, so you need to
remember each day after foraging which kind of mushroom is located
where on that day -- then you could certainly name the mushrooms and
the places, and there WOULD be a source of corrective feedback, because
of course you go back to the wrong place or ate the wrong kind of
mushroom, there would be consequences, even though there was nobody
else on the island.

So Wittgenstein is right about a private language for private
experiences, but he's wrong that a community is needed to provide any
corrective feedback at all: Even a nonhuman environment can provide that
for you.

Now it's hard to imagine why someone on a desert island who had never
seen another speaking human would BOTHER to invent anything more than
this kind of verbal shopping list, but in principle he could -- all
the way down to the description (and discovery) of all the biological
species on the island, and their biological basis. In fact, with some
miraculous ingenuity, he could in principle deduce Newton's law and the
existence of gravity (after all, it was a private moment when the apple
fell from the tree that Newton wrote was the source of his idea), etc.;
he could invent mathematics and prove theorems. He could even imagine
other creatures and write imaginary novels!

This is all highly unlikely, of course, but it is not impossible -- and
it's the impossibility of a private language that's at issue. Am I
cheating or changing the subject if I speak about an external mushroom
word rather than an internal state word like "autopleasance"?
This begins to get more subtle and you might not want to venture further
into this, but (1) if I'm the only one on the island, and may daily food
is determined by how accurately I name, remember, and recovery my
mushrooms, how am I doing it if not by the sensations I get when I see
and recognise and "X" mushroom? I could be deceived: I could think it's
an X mushroom, yet it turns out to be a Y mushroom (and I find out,
from a stomach ache afterward, that my sensation -- my private sensation
-- must have been wrong). But I still got corrective feedback, and it
did not come from a language community jointly agreeing on how to use
words. It came from solitary foraging, seeing, naming, and getting it

I am of course naming the mushroom, not the jushroom sensation when I
call the mushroom an X. But it is nevertheless via a private sensation
-- in fact, many samples of sensations that are different with different
mushrooms -- that I learn which private sensations reliably signal
an X mushroom.

At this point philosophers are fond of imagining that my sensations
could be varying wildly, so that the one I had when I correctly
picked an X mushroom is not the same as the next time, when I think I'm
seeing it again; instead, it's the Y sensation that I am having, and
mistakenly calling the X sensation because it FEELS as if it was the X
sensation again.

But who cares about that? Sensations are only a means to an end:
it's getting the right mushroom that matters, not the right sensation.
And there's no problem of error with that, despite there being no
language community to provide the feedback.

Anyway, that's a flavour of the kinds of issues that are discussed in
connection with Wittgenstein's private language argument.

> I still also am unclear on how one can prove that we have minds. The
> philosopher Descartes did not actually prove that we have minds in his
> work as his Cogito Ergo Sum does not make sense to say the least. He
> made what has become known as a category mistake and thus the debate on
> whether we actually have minds is still open.
> You say we need minds in order to feel and witness emotions
> etc but surely a brain can do all of this?

I have now, on-line, run through the private language issue for you;
I did some of your homework because I was taking a break. Come back
after you've done some reading on this because just about everything
you say about it is incorrect. I'll make it easy for you. If you
read this on the Web, all you need to do is click on the links below. :simply doesn't make sense: This is Descartes' famous insight in "Cogito<br> :Descartes...<br> :Descartes has already done this for us, with his "Cogito Ergo Sum"<br> :what Descartes knows, which is that feelings are being felt by him.<br> :pay heed to Descartes and his cronies (who believe that the only reason<br> :definite limitation on what we can know for sure: Descartes, on the<br> :(1) The first of Descartes' certainties were the truths of mathematics<br> :Descartes found he could doubt all the truths of science; there was no<br> :(2) Descartes' second and most famous certainty was the one he<br> :In other words, Descartes saw he could not doubt that he had a mind<br> :It DID have a bearing on other creatures' lives, however, for Descartes<br> :than myself have minds, so Descartes may have been responsible for some<br> :Let's leave this unfortunate aspect of Descartes' legacy aside. Some<br> :people also think the Mind/Body Problem was Descartes' invention and<br> :been around for millenia before Descartes in the form of the belief in<br> :that Descartes only gave people and not animals the benefit of the<br> :made by a French philosopher named Rene Descartes.<br> :Descartes wanted to rid himself of the received wisdom of his Age, and to<br> :know. So how do I know that my life is nothing but a dream". Descartes<br> :existed, it was something he could conclude by reason -- Descartes was<br> :Descartes went on to say lots of dodgy stuff about the only other thing<br> :the "self." We know from Descartes that you can't be wrong about the<br> :an example of the kind of thing Descartes (correctly) said we could not<br> :Descartes 1640, cited in Putman 'Philosophy and Our Mental life'.<br> :<i>&gt; Descartes 1640, cited in Putnam 'Philosophy and Our Mental life'.</i><br> :Putnam's paraphrase of Descartes here -- and probably Descartes own<br> :And about that you can have no doubts. That is Descartes' "Cogito Ergo<br> :Descartes did. Descartes doubted everything he could doubt. And one of<br> :<i>&gt; observed material. Descartes perception of 'true knowledge' may, in</i><br> :A few of things are being mixed together here. Descartes was talking<br> :mind. This can be related to Descartes' "Meditations Metaphysiques", in<br> :studying the mind, although it is difficult to say that Descartes<br> :IS subjectivity. And Descartes was right that a subject cannot be wrong<br> :IS subjectivity. And Descartes was right that a subject cannot be wrong<br>

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:53 GMT