Alternatively, is the privacy clause in the argument relative? So to say, imagine two people A and B looking at a patch of blue together, while standing next to someone C who's blind. A says to B, "I'm currently visualizing a patch of blue that is the same shade of blue as the one we're both looking at. I'm copying that shade in my mind." Then B comes to copy the shade in their mind, too. Thus far, neither is using an I-private language, it would seem to me.

But now compare A and B to C. C has no direct knowledge of colors, as such, more or less at all.👁 Do A and B's references to patches of blue constitute a we-private language, here, relative to C? As if they are using a collectively private language, then.

👁Or do they? I often wonder if the brains of those who do not see with their eyes, never simulate sight anyway. Can't a blind person dream in color regardless? In other words, for reasons of apparent empirical possibilities, I am unsure about this subquestion (c.f. Molyneux's problem and modern approaches thereto).

Tangent: suppose a language L meant to have the meta-predicate "is closed under 'is a predicate in L." In other words, differentiate a closed from an open language. But now in fact is it possible for any language to be closed, or does the addition of more speakers always allow that some L can be opened up, and new predicates added thereto? Modulo the rest of this OP, the question is: would the concept of a private language be on a par with, or replaceable by, the concept of a strictly able-to-be-closed language, whereby if the able-to-be-closed type has nothing to do with I- vs. we-intentions as such, then we can omit that distinction in intentions when interpreting the character of Wittgenstein's distinction between public and private language?


1 Answer 1


The words of this this language are to refer to what only the speaker can know - to his immediate private sensations. So another person cannot understand the language.
- Section 243, Philosophical Investigations, Ludwig Wittgenstein

The Value of Logical Impossibility

You say:

At any rate, if a private language is "logically impossible," then I don't know what the point of using the phrase "private language" is. It should be trivial or empty, not useful as a point of contrast, shouldn't it, then?

First, let us recognize the linguistic pragmatism of the phrase. If I were to say to you, do you understand you are trying to square the circle, do you not know what I mean? Despite the logical impossibility of the activity, the words themselves carry meaning. As you are clever, you already know that the surd beings of various sorts have pragmatic truth even if they lack correspondent truth. Let's consider some cases you are likely familiar with.

  • Santa Claus is a surd being incapable of existing given his magical powers of giving all the children of the world gifts in one evening. Does he not have some meaning and value in society? Yes, to teach a lesson about gift giving. Here we lie to our children to teach them a truth. Uh, oh, another contradiction!

  • How is it possible to communicate a truth with a lie? Language has a literal and figurative aspect, and we can abstract from fictional examples and apply them to factual circumstances. Consider how you can watch a movie like Saving Private Ryan and learn some historical lessons. I suspect the how is complex, but surely you can see that the use of fictional scenarios to apply lessons to factual ones is pregnant with analogical reasoning.

  • Math is full of surd beings that serve pragmatic functions. How about a count that represents nothing to count? The zero is ridiculous they said. Fluxions and infintesimals are logically impossible, and yet it allows us to do calculus. Something with a measure of zero (see how math builds on absurdity) and still has extension? Ridiculous they said. Or what about the square root of a negative number? Ridiculous they said. Logical impossibility simply does not preclude meaningful communication in the spirit of the pragmatic.

So, even though "private language" is logically impossible, in linguistics it serves the purpose of communicating a meaning above and beyond a combination of words or even if it doesn't allowing us to learn by ostensive action. It's in the same class of linguistic artifact as "raining cats and dogs" (also impossible but happens all the time in Seattle, WA). There's a tidy term linguists use: idiom. From WP:

An idiom is a phrase or expression that typically presents a figurative, non-literal meaning attached to the phrase... in English alone there are an estimated twenty-five million idiomatic expressions.

Thus, philosophical idioms enrich our ability to have discourse in the same way any idiom does. The term "private language" is an idiom that facilitates philosophical discourse. That's the purpose it serves! Let's move to your actual question.

I- vs. We-Private Language

Yes, there absolutely is a difference between I-private and we-private. The private language argument and the beetle in the box give rise to two idioms meant to bring to the surface an apparent contradiction in language acquisition use. (Again, it is not problematic that there are no literal private languages nor are literal beetles involved.) On the one hand, utterances of a language are public affairs. You may listen to two French persons talking French, identify French by its phonological characteristics, and be completely clueless. Alternatively you may listen to the speaking in English about something you don't know about, such as growing up in a small rural village in a time before you were born. But you have no access to the sensations and arguably experiences they had or have. That is everyone's beetle.

How is it that we are able to have language at all if we have no access to each other's beetles? That is a philosophical problem that confronts us in our naivete about language acquisition and development. Of course, as Popper rightly noted, evolutionary epistemology gives us some insights into this. Cognitive science has given us a clearer picture of how it happens. Michael Tomasello's work (a gift from Philip Klöcking) has a number of works in social cognition that have tremendous philosophical gravitas in the philosophy of language.

It's all fascinating, but tangent to the question. At heart to your question is what sense should be manufactured from the erotetic impulse that gives us a question about differentiating I-privacy and we-privacy. And I would suggest the following:

I-private language is a philosophical problem that asks, given my private access to sensation, how is it that I can participate in a social institution like a language since I have access to no one else's first-person-ness? (I was tempted to say privates, but that would have been the wrong direction.)

I think the answer is provided by embodied cognition (SEP). If I have a similar visual system that informs my consciousness, then my visual system and your visual system must in some sense provide us similarities in qualia or tropes or whatever compositional atoms we observe to use and build language together. If I see blue and you see blue, and through our socialization 'blue' serves the same utility at the same time in the same way, then we negotiate in our language game uses for 'blue'. This would be collaboratively manufacturing a semantics from our purlieu.

Of course, the bulk of semantics for most people require no such efforts. All speakers inherit a language from millieu. They coin neologisms relatively infrequently unless they are cognitively enterprising sorts, let us call them philosophers, who are pushing the bounds of describing sensation and experience. For most people and most words and most experience, we simply inherit 'blue'. This is one truth that seems to provide for the ontological basis of extended intelligence which parenthetically is why the section is part of the SEP's article on embodied intelligence.

So, I-privacy answers the question of how does an individual move from his private head space to the public sound space. That is not what you are proposing by we-privacy. We-privacy, like our scenario with French speakers, means that you don't share the sensations and experience with others at all. It is not overcoming the subjective-objective barrier, but penetrating instead the in-group-out-group barrier. I-privacy challenges me to have language at all, but we-privacy is challenging me to having language to describe synesthesia when I am not a synesthete, or Olympic-like swimming skills, or growing up in a small French town in a time before my birth speaking a rustic dialect of French. We-privacy is just an exclusion from a range sensation and experience.


So, by collaborating and negotiating in a purlieu we build invent and adapt sense and reference, negotiate a phonology, stipulate definitions, and build a might edifice of grammar and corpus that allows us to use language for what is meant, to provide a serialized string of sounds that communicate our mental states between our ears to each other. Language has two aspects. The what, which is the language proper, and the why, which is language-game. And the language-game has both I-privacy and we-privacy, but they arise from two separate, but interrelated situations revolving around the dichotomy between first-person and third-person experiences. This is what Wittgenstein recognized. That since language is negotiated and has an agenda, it is not fully objective, and that creates a lot of tension if you don't accept that language expresses a certain amount of tension between the subjective and objective. Carnap and Quine simply went forward with their formal language and semantic ascent respectively starting here.

  • 1
    From the logical-pluralist vantage, it has become hard for me to track what I mean by "logically impossible" except as "logically impossible relative to some system," so I am unsure about classifying infinitesimals as logically impossible per se. Smooth infinitesimal analysis is LEM-unfriendly, but nonstandard analysis is consistent with the LEM, so the hyperreal-theoretic infinitesimals there would not be impossible modulo classical logic. I suspect, of course, that you would be more partial to SIA than NA, though ;) Commented Oct 8, 2023 at 17:07
  • I too read the phrase contextually, because it's not just a question of a theory of logic, but the context with which it is exercised. Metaphysical necessitation for instance is in a distinct domain of discourse than material conditional.
    – J D
    Commented Oct 8, 2023 at 17:28
  • A term that cannot be assigned a meaning in a given language (i.e.) has no use, can be given a use in an extended version of that language. So 0, 1, square root of 2 and negative numbers had no meaning in Greek mathematics. Later, in the Middle East, calculation techniques were developed in which those symbols were used, and bingo! they were not logically impossible after all. No meaning means no coherent use, that's all.
    – Ludwig V
    Commented Oct 8, 2023 at 18:24
  • 1
    So we're back to engineering: "Does it do anything?" isn't just an advertising slogan.
    – Scott Rowe
    Commented Oct 8, 2023 at 20:03
  • 2
    @ScottRowe In a way, that's exactly what evolutionary epistemology argues bluring, of course, the distinction between teleological and non-teleological language. Language should be seen as the result of solving the problems of communication.
    – J D
    Commented Oct 8, 2023 at 20:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .