Wittgenstein's Private Language Argument And Its Practical Applications

[Pulvinar]

That is not what Wittgenstein discussed, however. His answer to you would be, sure you can do all that. But you didn't create that language in the first place to write yourself the note. He would also say, sure you could create your own language just to spite him, but it would be based on a language that you already learned from others.

However we want to discuss this, language is a means of communication between individuals (however defined); that is simply what it is. As such, if there were only one person alive on the planet, ever, with no other companion, ever, what sense would there be in constructing a language? That being, one living in isolation, would be entirely different from anyone raised in a language community. I don't even see how the concept of creating a written language to leave oneself a note in the future could even arise in such a situation.

I don't see why I couldn't create a language just to write notes to myself at a later date. In fact I've done just that while designing computer languages-- I had to be able to read and understand them before I wrote the compiler or interpreter to allow the computer to understand them.

If I work out the details of the language in my head, how is that not a private language?

 

[Ichneumonwasp]

I don't see why I couldn't create a language just to write notes to myself at a later date. In fact I've done just that while designing computer languages-- I had to be able to read and understand them before I wrote the compiler or interpreter to allow the computer to understand them.

If I work out the details of the language in my head, how is that not a private language?

This is not an issue of transliterating or translating words from one language to a new one that you have just created, even if you are the only one who knows it. You already learned language within a language community, so you can create any new one based on the old ones you learned and not share it with anyone. It's a bit silly, but as I said above you could do it just to spite Witt. or just because you were bored, but that still does not touch his argument.

You did not start off creating words that name your internal states without reference to other language users.

 

[Pulvinar]

This is not an issue of transliterating or translating words from one language to a new one that you have just created, even if you are the only one who knows it. You already learned language within a language community, so you can create any new one based on the old ones you learned and not share it with anyone. It's a bit silly, but as I said above you could do it just to spite Witt. or just because you were bored, but that still does not touch his argument.

You did not start off creating words that name your internal states without reference to other language users.

We have a problem here with the concept of language-- if Wittgenstein is going to declare any language using parts of another language as not a *true* language, then there could only be one true language per person. And that language would have to be some baby talk. But when do those babblings turn into a true language? Trying to draw boundaries for this true language is hopeless.

 

[Ichneumonwasp]

We have a problem here with the concept of language-- if Wittgenstein is going to declare any language using parts of another language as not a *true* language, then there could only be one true language per person. And that language would have to be some baby talk. But when do those babblings turn into a true language? Trying to draw boundaries for this true language is hopeless.

He wouldn't make that argument. I'm not suggesting it either. It's just that once you have already acquired a language, then you can translate that language into another one that you alone use just to prove him wrong. But that doesn't prove him wrong because, at least from my reading of it, he was speaking about natural languages and how we come to talk about internal states. We don't create words for internal states (like pain) without a surrounding language community where a word such as pain can be defined in terms of what others experience.

And people do not create their own languages in order to speak to themselves generally.

This doesn't concern a right or true language. It does concern natural language creation, though.

 

[Pulvinar]

He wouldn't make that argument. I'm not suggesting it either. It's just that once you have already acquired a language, then you can translate that language into another one that you alone use just to prove him wrong. But that doesn't prove him wrong because, at least from my reading of it, he was speaking about natural languages and how we come to talk about internal states. We don't create words for internal states (like pain) without a surrounding language community where a word such as pain can be defined in terms of what others experience.

And people do not create their own languages in order to speak to themselves generally.

This doesn't concern a right or true language. It does concern natural language creation, though.

I understand how his argument is based on the natural development of natural language, but don't see how it can escape being falsified without limiting its use to the point of being useless for understanding the brain.

An example: I experience a new type of pain but don't tell anyone about it. Is my internal concept of that pain private or not? Is it only part of a language if I give that pain a unique name? Is a language that contains one private word private? How about one that only contains one public word? This seems to be all just taxonomy.

 

[Ichneumonwasp]

I understand how his argument is based on the natural development of natural language, but don't see how it can escape being falsified without limiting its use to the point of being useless for understanding the brain.

An example: I experience a new type of pain but don't tell anyone about it. Is my internal concept of that pain private or not? Is it only part of a language if I give that pain a unique name? Is a language that contains one private word private? How about one that only contains one public word? This seems to be all just taxonomy.

Your internal experience of the pain is private. As to the concept of it, that is up for discussion; we would need a clear sense of what that means and I'm not sure how we can separate it from language very easily. At a neuronal level the concept of that type of pain will need to be represented in mentalese, but mentalese, again, is about as universal a language as possible.


Would it be a part of language if you gave it a unique name? No. If you give it a unique name no one would understand the word or the concept. We only understand words, particularly new words, unless they are expressed in terms of other words that we already understand. So, you can certainly explain what you mean by this new pain by use of analogy to other types of pain. That is what he is really talking about -- not that this is all about taxonomy or about single new coinages.

 

[Pulvinar]

Your internal experience of the pain is private. As to the concept of it, that is up for discussion; we would need a clear sense of what that means and I'm not sure how we can separate it from language very easily. At a neuronal level the concept of that type of pain will need to be represented in mentalese, but mentalese, again, is about as universal a language as possible.

If you can call pain private because I can't sense your pain (yet I believe I can sense similar pains), you should also call seeing a chair a private experience because I can never see it just as you do. We can argue over just how much of the experience must be matchable for it to be classified as public-- which is why I call it taxonomy.

Would it be a part of language if you gave it a unique name? No. If you give it a unique name no one would understand the word or the concept.

You mean no one else would. *I* would understand it when I recalled it.

We only understand words, particularly new words, unless they are expressed in terms of other words that we already understand. So, you can certainly explain what you mean by this new pain by use of analogy to other types of pain. That is what he is really talking about -- not that this is all about taxonomy or about single new coinages.

That can't be right: how did the first words ever get understood? Context and feedback play a big role here.

 

[Ichneumonwasp]

If you can call pain private because I can't sense your pain (yet I believe I can sense similar pains), you should also call seeing a chair a private experience because I can never see it just as you do. We can argue over just how much of the experience must be matchable for it to be classified as public-- which is why I call it taxonomy.

OK. As to seeing a chair as a private experience I did just that earlier in the thread with 'blue'. It's all private experience; language is a way of bridging private experience so that we can speak of an 'objective' world, by which we mean inter-subjective.

You mean no one else would. *I* would understand it when I recalled it.

Right, it wouldn't be part of a language. It would be your own neologism within a language that already exists.

That can't be right: how did the first words ever get understood? Context and feedback play a big role here.

Yes, I misstated as it relates to all possible scenarios. First 'words' would have directly referred to physical objects. But that probable goes back to dinosaurs or possibly before.

 

[Minoosh]

Blue is no more broadcast than a stick might be. Imagine a world of sticks flying through the air, broadcast to every living thing just as light rays are in our world. Would we not speak of pain being broadcast in that world? Would we speak of these flying things as pain-sticks more readily, or possibly just pain?

The slings and arrows of outrageous fortune ... yes, arrows are pain sticks, that's pretty much the point. Hah! The point!

I could read some Wittgenstein and then it's possible I'd know what I was talking about.

 

[Minoosh]

For some reason this reminds me of something Jim Bouton said in "Ball Four" - that people couldn't grasp that it hurt him to pitch, because it didn't hurt them to watch him pitch.

But they could probably agree if the sky was blue.

Both perceptions happen within individual nervous systems, but they seem different to me. Somehow.

 

[Pulvinar]

OK. As to seeing a chair as a private experience I did just that earlier in the thread with 'blue'. It's all private experience; language is a way of bridging private experience so that we can speak of an 'objective' world, by which we mean inter-subjective.

Ok-- all private experience, and optional public attempt to share that experience through language. I don't see a solid line between the experiences that lead to the words "chair" and "pain". The experience of pain may be more private, but not absolutely private unless we allow that for all sensations. So why should its label be treated differently?

Right, it wouldn't be part of a language. It would be your own neologism within a language that already exists.

What if someone happens to overhear me say it without my knowledge? Or if I tell someone but they don't actually hear me? Anywhere you draw the line here I bet I could come up with a case that neatly straddles it.

 

[Ichneumonwasp]

Ok-- all private experience, and optional public attempt to share that experience through language. I don't see a solid line between the experiences that lead to the words "chair" and "pain". The experience of pain may be more private, but not absolutely private unless we allow that for all sensations. So why should its label be treated differently?

Well, that's where I disagree with him.

What if someone happens to overhear me say it without my knowledge? Or if I tell someone but they don't actually hear me? Anywhere you draw the line here I bet I could come up with a case that neatly straddles it.

I don't see how it would change the basic situation. A neologism makes no sense to anyone else. They would have to have it explained in terms of words they already understand.


Perhaps we could concentrate on how it cannot understand the functioning of the brain. I disagree with this idea. There can't be anything entirely private within the brain; the posterior thalamus doesn't speak its own private language with respect to the parietal-occipital junction or the superior collicus.

 

[Minoosh]

There can't be anything entirely private within the brain; the posterior thalamus doesn't speak its own private language with respect to the parietal-occipital junction or the superior collicus.

What?

Isn't part of the point that there is meaning within individual minds/brains, that doesn't carry over into the perception of other minds/brains? If the mind/brain is talking to itself, are we not to conclude that they are, at least, within the same skull?

I love these pointless discussions. Time to get a life.

 

[Ichneumonwasp]

What?

Isn't part of the point that there is meaning within individual minds/brains, that doesn't carry over into the perception of other minds/brains? If the mind/brain is talking to itself, are we not to conclude that they are, at least, within the same skull?

I love these pointless discussions. Time to get a life.

I'm speaking from a different perspective -- the neuronal perspective. The 'language' within the brain is identical for all areas or interaction wouldn't be possible.

Pulvinar brought up the issue of neuronal communication being 'personal' or subjective, but that is based on an equivocation over the word. From the perspective within the brain there are different regions that communicate using a common language that we often call 'mentalese'.

 

[HikakaGirl]

Interestingly, the whole question of what is a "chair", what is "blue" and what is "pain" is sort of what Popper was on about. Note that my comments below rest on the notion that we're kind of stuck using words we know to communicate concepts, whether novel or not. I am going to try and integrate the whole public/private language discussion here as well, but no promises that I'll do it well.

Soooo...Popper says that there are three worlds -- the concept, how we measure (or describe) that concept, and what is in the real world. In the case of the color we call "blue", we would take Popper's framework and apply it this way: In an effort to develop our *best* definition of the color "blue" (the concept), we would show a lot of people a lot of different colored items, say pieces of paper, or objects found in nature (how Aristotelian of us) which represent what is in the real world. Perhaps we'd start with a bunch of stuff that was sort of blue, maybe some greens and purples that are on the blue side. This would be an a priori selection of items, and sometimes we want to do this if we already have a good base to work from (a more public language already established for the concept?) and sometimes we don't because we want to really work it like Aristotle would want us (who cares about what you think; what are the facts in nature?).

Through an iterative process of showing these blue and bluish items to people and having them identify the items as blue or not blue, or even more or less blue, we can identify a prototypically blue object, one which a statistically significant portion of our group of people would consider blue. Note that there is error inherent in this process, we never assume it's perfect, we just want it to work well enough that we have confidence in it. This prototypical blue object then would serve as our measure for what "blue" is. Many people consider the color of the sky as this standard (although I've lived in places where the sky was more white than blue most of the time).

So now we have our prototypical object (reality), which represents blue (the concept) that we can then use as a standard (measure) for what is "blue". The cool thing about this process, is that it doesn't stop there. We continually validate the prototypical object as being representative of blue in reality each time we use it. We can even enhance it representativeness as we go along if we find a more prototypically blue object. The expectation is that we will always use the most prototypical representation of the concept, and if for some reason someone doesn't, the peer review process kicks in, and they're corrected.

Another caveat is that the process, clearly, works better for things that are more easily representative and observable by others, and/or things that we've established a greater amount of public language about. So for example, we are much more likely to have widespread agreement on what a chair is than what blue is, and in turn what blue is than what pain is.

I guess another way to connect all the concepts together would be to suggest that Popper's philosophizing lead to a system whereby language could be migrated from one that is more of a private language to one that is more of a public language. Or am I not understanding that argument well? Admittedly, I'm not as familiar with it as any other poster in this thread. By a long shot.

HG

 

[Pulvinar]

Well, that's where I disagree with him.

So I misunderstood your first post-- you find W's argument useful, but wrong.

I don't see how it would change the basic situation. A neologism makes no sense to anyone else. They would have to have it explained in terms of words they already understand.

You could be overheard saying the word while in a context that makes it clear what's being referred to. Or even explaining it to yourself. Unlikely but possible cases that can blur the privacy status of a language.

Perhaps we could concentrate on how it cannot understand the functioning of the brain. I disagree with this idea. There can't be anything entirely private within the brain; the posterior thalamus doesn't speak its own private language with respect to the parietal-occipital junction or the superior collicus.

(Assume you meant "cannot help us understand")

The answer depends on how we define "private" when talking about communication between subunits of a brain. The language in this case is the set of patterns and their meanings in the information flowing between them. Part or all of these patterns and meanings could be different for each subset of pathways-- unknown to those subunits that don't receive them. What I would call private.

That's certainly the case in electronic circuits. For example, a network interface chip will communicate in a particular language with the the CPU. That language will be quite different from the one the CPU uses to talk with the graphics unit. Within the network interface will be further subunits, each talking in a language consisting of just those symbols needed for the commands and responses between them.

The clearest example is when you subdivide down to the gate level, where a single gate with say 2 binary inputs is clearly not going to be able to understand more than 4 symbols, let alone understand the meanings assigned to (respond in an appropriate way to) the symbols of all other pathways in the device. So those other languages have to be considered private with respect to each gate.

 

[Ichneumonwasp]

So I misunderstood your first post-- you find W's argument useful, but wrong.

I think part of it is wrong. And, yes, I find it useful.

You could be overheard saying the word while in a context that makes it clear what's being referred to. Or even explaining it to yourself. Unlikely but possible cases that can blur the privacy status of a language.

I still don't see how that is an objection to his basic argument about natural language.

The answer depends on how we define "private" when talking about communication between subunits of a brain. The language in this case is the set of patterns and their meanings in the information flowing between them. Part or all of these patterns and meanings could be different for each subset of pathways-- unknown to those subunits that don't receive them. What I would call private.

That's certainly the case in electronic circuits. For example, a network interface chip will communicate in a particular language with the the CPU. That language will be quite different from the one the CPU uses to talk with the graphics unit. Within the network interface will be further subunits, each talking in a language consisting of just those symbols needed for the commands and responses between them.

The clearest example is when you subdivide down to the gate level, where a single gate with say 2 binary inputs is clearly not going to be able to understand more than 4 symbols, let alone understand the meanings assigned to (respond in an appropriate way to) the symbols of all other pathways in the device. So those other languages have to be considered private with respect to each gate.

Communication between different units is not possible unless those languages map in a consistent pattern that is 'translated' between the units. It can't happen by magic. You could speak to me in Chinese, and I'm only going to act on what you say if I somehow understand it.

 

[Ichneumonwasp]

Interestingly, the whole question of what is a "chair", what is "blue" and what is "pain" is sort of what Popper was on about. Note that my comments below rest on the notion that we're kind of stuck using words we know to communicate concepts, whether novel or not. I am going to try and integrate the whole public/private language discussion here as well, but no promises that I'll do it well.

Soooo...Popper says that there are three worlds -- the concept, how we measure (or describe) that concept, and what is in the real world. In the case of the color we call "blue", we would take Popper's framework and apply it this way: In an effort to develop our *best* definition of the color "blue" (the concept), we would show a lot of people a lot of different colored items, say pieces of paper, or objects found in nature (how Aristotelian of us) which represent what is in the real world. Perhaps we'd start with a bunch of stuff that was sort of blue, maybe some greens and purples that are on the blue side. This would be an a priori selection of items, and sometimes we want to do this if we already have a good base to work from (a more public language already established for the concept?) and sometimes we don't because we want to really work it like Aristotle would want us (who cares about what you think; what are the facts in nature?).

Through an iterative process of showing these blue and bluish items to people and having them identify the items as blue or not blue, or even more or less blue, we can identify a prototypically blue object, one which a statistically significant portion of our group of people would consider blue. Note that there is error inherent in this process, we never assume it's perfect, we just want it to work well enough that we have confidence in it. This prototypical blue object then would serve as our measure for what "blue" is. Many people consider the color of the sky as this standard (although I've lived in places where the sky was more white than blue most of the time).

So now we have our prototypical object (reality), which represents blue (the concept) that we can then use as a standard (measure) for what is "blue". The cool thing about this process, is that it doesn't stop there. We continually validate the prototypical object as being representative of blue in reality each time we use it. We can even enhance it representativeness as we go along if we find a more prototypically blue object. The expectation is that we will always use the most prototypical representation of the concept, and if for some reason someone doesn't, the peer review process kicks in, and they're corrected.

Another caveat is that the process, clearly, works better for things that are more easily representative and observable by others, and/or things that we've established a greater amount of public language about. So for example, we are much more likely to have widespread agreement on what a chair is than what blue is, and in turn what blue is than what pain is.

I guess another way to connect all the concepts together would be to suggest that Popper's philosophizing lead to a system whereby language could be migrated from one that is more of a private language to one that is more of a public language. Or am I not understanding that argument well? Admittedly, I'm not as familiar with it as any other poster in this thread. By a long shot.

HG

The way I take Wittgenstein's approach is that he arrived at the same place, but included internal states like pain. We learn words within a language community affixing terms onto our internal states (perception of blue, chair, pain) in relation to what others in the same language community do. That is how words get their meanings/definitions. Same is true for pain. We tend to think of pain as a different 'thing', as purely internal; but I don't see it that way. I don't see how, neurologically, we can separate the process by which we feel pain from the process by which we see a chair or blue (in broad strokes).

I differ from Witt in his analysis of chair as opposed to pain; but I think the concept of language as public is correct.

 

[Pulvinar]

I still don't see how that is an objection to his basic argument about natural language.

I was showing that the privacy status of a language has shades of gray, so you can't have a black-and-white argument about it.

Communication between different units is not possible unless those languages map in a consistent pattern that is 'translated' between the units. It can't happen by magic. You could speak to me in Chinese, and I'm only going to act on what you say if I somehow understand it.

You're thinking too high-level.

Two simple units could communicate with a 4-symbol language, where the meaning of each symbol is unique to those two units and was determined by evolution. Or it could be partially or completely learned.

The problem with this whole exercise is the near-infiite number of ways that a brain can be divided into units and information pathways grouped. For example you could consider all pathways together as one, with one very complex language among them all (which I believe is what you're doing). Or divide that in two, etc. To be useful you need to know the actual structure and workings of the brain, but I don't see how a consideration of privacy is much help in getting that.

 

[HikakaGirl]

The way I take Wittgenstein's approach is that he arrived at the same place, but included internal states like pain. We learn words within a language community affixing terms onto our internal states (perception of blue, chair, pain) in relation to what others in the same language community do. That is how words get their meanings/definitions. Same is true for pain. We tend to think of pain as a different 'thing', as purely internal; but I don't see it that way. I don't see how, neurologically, we can separate the process by which we feel pain from the process by which we see a chair or blue (in broad strokes).

I differ from Witt in his analysis of chair as opposed to pain; but I think the concept of language as public is correct.

I think the reason why pain is problematic to (and informative of) this discussion is that it's a more subjective term. In other words, there is greater variability among individuals' definitions of pain than their definitions of what a chair is. Some people find pinches to be very painful, others not so much, but it's much easier to agree on what a chair is. So perhaps it seems like pain is representative of a private language, but really it's public language interpreted with greater variability in what individuals understand the term to mean.

Now that I have a little better understanding of the issue, I tend to agree with you. I'm probably restating here, but the purpose of language is communication, so by definition I would think it would have to be at least somewhat public, otherwise the language is ineffective and would die out. Maybe I'm thinking about this from too much of a utilitarian standpoint.

HG