I like a look of agony,
Because I know it’s true;
Men do not sham convulsion
Nor simulate a throe.

The eyes glaze once, and that is death.
Impossible to feign
The beads upon the forehead
By homely anguish strung.

(Emily Dickinson)

An important part of seeing what Wittgenstein is up to is to recognize that his starting point when “doing philosophy” is that there are all of these phenomena of life:  we talk meaningfully about dreams, the future, sensations, chairs, music, numbers, good and bad; we follow rules; we recognize the feelings of others, though sometimes people are good at hiding them; and much, much more.  We succeed in doing all of these things.

We are tempted to say “How is all of that possible?”

We then hypothesize things like intentionality of the mind, hidden mechanisms underlying meaning, reference, and feeling, and much more—ways of trying to explain how all the other stuff is possible, how it works.

Wittgenstein wants to expose the limitations of many of these theories and their pictures.  He employs various methods, e.g., the method of §2 mentioned in §48 of PI (all section references are to PI unless otherwise noted).  Part of that method, I think, is the trying out of what it would mean if X were true.  So, e.g., what if sensations were really private in the super strong sense of §243?

Since our sensation talk is intimately connected with sensation/feeling behavior (groaning, sighing, smiling, laughing, wincing, etc.), and it is through such behavior that we often know others are in pain (or happy, sad, bored, etc.), then a private sensation language would have to be one without sensation behavior (as Wittgenstein says in §256).

Part of what is going on with the consideration of a private language is that Wittgenstein wants to reorient us so that we see the vital importance that our natural expressions of feeling/sensations play in language’s being meaningful.  What are some examples of natural expressions?

Think of the behaviors and the appearances of the body/face when we feel:  tired, angry, happy, surprised, joyous, nauseous, annoyed, ill, wired, distant, etc.

Now let us consider a case that might show the importance of the natural expressions of feeling/sensations.  What if there were beings who had no natural expressions of feeling, who didn’t do or have any of the behaviors and appearances characteristic of the feelings mentioned above?

Can we really imagine the lives of such beings?

What would their language be like?  Could they talk about their feelings and sensations?

We can perhaps imagine them talking about having physical objects, e.g., books, clothes, etc., in their possession and they could describe their properties.

Would they still avoid fire with quick movements backward?

Couldn’t one move away quickly from a fire and another ask why he is doing that?  What could the other one say?  “When my hand is in the fire I have something—something like I have when I smash my thumb under a rock.”  Couldn’t the other reply:  “Ah, I also have similar things when my hand is in the fire and when my thumb is smashed by a rock.”  “Well, let’s call what we are both having in these cases ‘pain’”?

Isn’t the above possible?  Well consider:  with what right can we say of two such beings that they really have the same sensation when their hands are in the fire?

Well, they both back away from the fire and avoid smashing their thumbs.

But what if one did those things because of pain and the other because they cause too intense a sensation of pleasure (worse than being tickled say)?

But wouldn’t the difference come out somewhere?  One remarks that the hand in the fire gives him too much of what he has during sex.  The other says “What do you mean?  They are nothing alike!”  (And we couldn’t say that the one who says sex and fire give different things might possibly be feeling pain instead of pleasure during sex, since he doesn’t avoid sex.)

But can they properly speak of “like” and “unlike” things here?

Well they presumably take these concepts of “like” and “unlike” from their experiences of talking about physical objects—so why not give them “like” and “unlike” for what they have “on the inside”?

Perhaps part of the problem with the above line of thought is that we didn’t count the behaviors of avoidance and fleeing, and preference and embracing as natural expressions of feeling/sensations, but we should have.

So what about beings who have absolutely no behaviors or appearances that express feelings/sensations or that could mark differences of feeling, sensations, etc.?  Even though none would ever see in another’s face any feeling, couldn’t one who feels pain when his hand is in fire and when smashing his thumb ask another if he also has something similar in both cases, and then ask about other similarities and differences?

What if they agreed about all such similarities and differences of “what they have” in all the different cases?  Couldn’t they agree to call the fire and thumb smashing things they have “pain” and the sex and good food things they have “pleasure”?

Well is it possible that they could agree on all the similarities and differences (thumb smashed is similar to a hand in fire, both are different from sex and good food, etc.) and yet still have very different sensations from each other (In the same way we might wonder if whether all the things I see as blue you see as green)?  In which case, it is not the sensations that are important but their relationships of similarity and difference to each other, and that fact that they agree about these.

But couldn’t they say that they want more of what they have during sex and less of what they have when their hands are in fire?

Remember that we said they couldn’t reflect these preferences and aversions in their behavior.  Given that, how much sense would it make to say they want more of that which they have during sex than that which they have from their hand in fire?

And we should also consider what their form of life would really be like.  How would they behave toward each other?  How would a parent know when to feed a baby?  Could such beings actually evolve as a species over time from simpler organisms?  Could they evolve as a social species capable of speaking any language at all?  (What are some of the reasons that groups of social creatures evolve into language users?)

In our imagined case of the beings who could express nothing through non-verbal behavior, have we perhaps encountered a form of life that is so different from our own that we cannot, with justification, draw implications from it to our own?  If that is so, what does it say about language, pain talk, etc., and the place of natural expressions, and a private language?

Well we said that natural expression would keep a language from being private.  So in order to try to make sense of the possibility of a super private language, we imagined beings with no natural expressions.  But two things resulted:

1) Insofar as they could talk about their sensations, they had to use the public language—so they didn’t create a private language of sensation talk.

2) Their form of life is so different from ours that it seems irrelevant to our own.

Now because of 1), we are left to make sense of a private language along the lines of §258, where there are no natural expressions of sensation, nor some other language that can be used to help “create” the private language.  And in §258 there is the problem that one cannot name anything since there is no way to disambiguate the concentration of attention.

But couldn’t one point out now that Wittgenstein was wrong when he insisted that natural expressions were the only way to “connect” words with sensations?

Well, first, we should note that Wittgenstein does not say that natural expressions are necessary for speaking meaningfully about sensations.  Rather, he says that it is one possibility for connecting them (§244).  Second, it happens to be the case that for us (humans), natural expressions play an important role. Wittgenstein need not be seen as offering up the necessary and sufficient conditions for language’s being meaningful.  And moreover, it is just not clear at all in what sense the beings without any natural expressions are a real possibility.  (If they aren’t possible, is it a causal or logical impossibility?  Why should it matter?  Perhaps only insofar as philosophers tend to think they deal only with the “logical must.”)

Lastly, as in §§288-290 (and elsewhere), Wittgenstein seems to say that without natural expressions for sensations we would need a criterion of identity for our sensations, for talking about them, identifying them as the same at different times.  If that is true, then our imaginary beings would face that problem as well.

Regarding this issue of criteria of identity, it seems Wittgenstein wants to suggest that it so happens that we use the natural expressions of sensations to connect the sensations up with language.  We are trained to do so as we learn the language.

Hence we get in §290 the claim:  “It is not, of course, that I identify my sensations by means of criteria; it is rather that I use the same expression.”  By “expression” he means “linguistic expression” I take it.  This passage makes sense when we bring to bear §238:  “The rule can only seem to me to produce all its consequences in advance if I draw them as a matter of course.  As much as it is a matter of course for me to call this colour ‘blue’.  (Criteria for ‘its being a matter of course’ for me.)”  We “identify” our sensations as a matter of course; we don’t need to identify them and then name them just as we don’t identify a color and then say its name:  we just see that it is blue; we just feel that this is pain.  But this relies on the idea he mentions in §244, where the child is taught new pain behavior by replacing the natural expression of pain with pain talk.  This training is what makes it a matter of course that this is pain.