Live data from Hacker News

Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

doi.org

31–40 of 58 posts

Re: Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

#32
post #3

Not to diminish Wittgenstein’s influence, quite the contrary, but I for one first heard about the idea of a “private language” while reading the works of Wilhelm von Humboldt [1], written around the 1830s (so about a century before Wittgenstein). I remember Humboldt saying something along the lines of “at the limit, we each speak our own, private language”. I read his work in translation, as I don’t speak German, so…

Wittgenstein didn’t come come up with the idea of a “private language” but in his Philosophical Investigations he famously argued that the concept itself is nonsense (not via a single logical “proof” but using several thought experiments). The idea that a private language exists (ie. ‘mentalese’) is taken for granted by a lot of authors. I recently read Stephen Pinker’s book The Language Instinct and he never seemed…

I have not read Pinker so I don’t have a good idea of what he meant by “mentalese”, but Humboldt was using the concept of “private language” more on the lines of two different people having different “understandings” (for a lack of a better word) for concepts bearing the sama name. For example when I think of a “red apple” I might think of a different thing compared to when you think of a “red apple” on your turn (leaving aside the issue that we still need to settle on what “thing” from the previous sentence actually means).

I must admit that I last read some Wittgenstein about 15 years ago, but even so, I don’t find that this understanding of what a “private language” really means is that far off from his writings. If anything, it completely invalidates the 2000-year old platonic “ideas”, as me and you thinking of different things when presented with the words “red apple” practically means that there is no underlying “red” or “apple” idea behind those concepts. That is something very valuable, Plato’s interpretation being debunked, that is.

Re: Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

#33
post #3

Not to diminish Wittgenstein’s influence, quite the contrary, but I for one first heard about the idea of a “private language” while reading the works of Wilhelm von Humboldt [1], written around the 1830s (so about a century before Wittgenstein). I remember Humboldt saying something along the lines of “at the limit, we each speak our own, private language”. I read his work in translation, as I don’t speak German, so…

Wittgenstein argues that there is no such thing as a private language.

Re: Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

#34
post #4

When I first read about Wittgenstein in some junior encyclopedia, it said that he discovered that many problems in philosophy were meaningless because words might mean different things. I imagined he was talking about words like "socialism" or "equality" -- words that are obviously slippery. But it turns out that words like "is" and "have" cause just as much trouble. Wittgenstein is famously hard to read. I have the…

> When I first read about Wittgenstein in some junior encyclopedia, it said that he discovered that many problems in philosophy were meaningless because words might mean different things. That's how I've always intuitively felt about the Ship of Theseus thought experiment. It only seems paradoxical when people try to apply imprecise language to a precise situation. The way I see it, it's technically a new ship with e…

So are you technically a different person when your every cell is replaced?

Re: Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

#35
post #26

My favorite self-quote: technology development is applied philosophy. Wittgenstein taught it; we're the people living it. When philosophy and science existed mostly as words handed from one generation to the next in written form, it was possible to amass enough of a web of meaning for any one general field for a body of practitioners to gain useful applicable knowledge for the rest of humanity. But as philosophers fo…

While natural languages and mathematics are two important formalisms by which technologists communicate, it seems to me that we also use a much broader set of visual symbolic systems. These include tables of data in various formats and figures that incorporate drawings, diagrams, charts, pictures and videos. These are now as important as the text and formulas. Problems with tables and figures are one of the main reas…

I'm using the term "mathematics" in a deliberately hand-wavy fashion because people keep wanting to poke at this from various angles and even though the underlying theory is the same, it can appear that you are receiving different answers simply because you are asking the question from different epistemological frameworks. There's a reason Wittgenstein was the man and I'm not.

If you like, we can assume that different numbers have a concrete, discrete meaning. We can then add commutation and association, maybe induction. At some point, however far down the rabbit hole we want to go, these numbers in a table represent some physical thing: six apples or whatever. These diagrams and other visuals represent some tangible, agreed-upon thing: a diagram of an enzyme, perhaps.

The mapping of various calculii to either physical things or commonly-held concepts is not on the table here. The point is that the concepts themselves can have subtly-different meanings for any two practitioners or observers. A doctor in the middle ages may check your humeurs and a modern nurse may check you blood pressure. Imagining for a second that both operations look similar, even though you may have the same number, you have a completely different understanding of what those numbers mean.

So whatever formal rules of mathematics you'd like to have, and whatever visuals or measurements you'd like to take or produce, the end goal is analysis or language creation around a shared interest. If I hold up a coconut and want to trade it, then you make some odd sound, it means nothing. But then if you hold up two bananas, we may be beginning to converse and exchange meaning. (In this case by way of commerce, but this is just one example of thousands) The visuals enable higher-bandwidth conversations. Do I then think that 1=2, since I had one coconut and you had 2 bananas? I might. I might not. If we're from completely different cultures we have a lot more work to do.

That's an obvious example, the deeper and much more profound truth is that the same thing can happen with a vapor trail in a particle accelerator. I'm reminded of John Wheeler's idea that maybe there's only one electron in the entire universe. Once the web of meaning reaches some degree of complexity, the human brain shuts down and stops evaluating all the possible alternative meaning paths; we are not aware of this. In our mind we've thought through everything and are sitting on top of thousands of years of received wisdom. It has always been like this.

So yes, they are much, much more important than text or formulas, but they're more important because they assist us in the drive for common language creation. Frankly, many times they do a much better job than the others. But the job in both cases is the same.

I will make another stupid analogy. Ever see the scintillating grid optical illusion?[1] It looks like you can see every other black dot except the one you're looking directly at. Meaning is similar in that whatever the concept under observation, it appears like some analysis can work out the problems. The other ones, farther away, don't need any work. They're all set in stone. But then, given time, when you look at those, you realize that no, actually there's some problems over here. That area you were looking at before? That's all fixed now.

The lesson here is that there is a cognitive limit to the things that you can simultaneously ponder about their meaning, relationships, and relevance to any one situation. After that limit, it's all just "common knowledge" or received wisdom. It has to be this way. (no time to go into why). But then you realize that it's all a web, we're all living in our own constructed grid that's different, and the goal is to align both the concepts under observation and the "given" concepts among several of us such that we can erect a language scaffold sturdy enough to make progress towards some common goal.

I know that sucks, but that's the best I've got in ten minutes. Hope it helped. There's an entirely other conversation about how we construct these grids, how we share them, and more importantly why this is a feature of sentience and not just a stupid primate trick. No time for that.

1. https://www.illusionsindex.org/i/scintillating-grid

Re: Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

#36

Earlier quoted context omitted.

> When I first read about Wittgenstein in some junior encyclopedia, it said that he discovered that many problems in philosophy were meaningless because words might mean different things. That's how I've always intuitively felt about the Ship of Theseus thought experiment. It only seems paradoxical when people try to apply imprecise language to a precise situation. The way I see it, it's technically a new ship with e…

So are you technically a different person when your every cell is replaced?

One may argue you are a different person even once some of your cells are replaced.

I imagine your next question might be whether you owe something to now-me if you were to take a loan from a previous-me.

This looks like a pair of mirror bugs. Yeah, the person is already a new one, but if we'd really treat them that way, we'd have to make our laws and morals even more complex. Instead, we just pretend that's the same person.

Re: Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

#37
post #32

Earlier quoted context omitted.

Wittgenstein didn’t come come up with the idea of a “private language” but in his Philosophical Investigations he famously argued that the concept itself is nonsense (not via a single logical “proof” but using several thought experiments). The idea that a private language exists (ie. ‘mentalese’) is taken for granted by a lot of authors. I recently read Stephen Pinker’s book The Language Instinct and he never seemed…

I have not read Pinker so I don’t have a good idea of what he meant by “mentalese”, but Humboldt was using the concept of “private language” more on the lines of two different people having different “understandings” (for a lack of a better word) for concepts bearing the sama name. For example when I think of a “red apple” I might think of a different thing compared to when you think of a “red apple” on your turn (le…

>practically means that there is no underlying “red” or “apple” idea behind those concepts. That is something very valuable, Plato’s interpretation being debunked, that is.

This is one of my favorite parts in a fictional book Anathem. The question of syntax and semantics is explored there, and even a solution is suggested.

Re: Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

#38

Earlier quoted context omitted.

Philosophers and programmers both take hazy cultural ideas and turn them into rulesets.

But to such different ends that the resemblance is very strained, surely? Programmers look to make an approximately useful tool, Philosophers look to discover fundamentals and insights.

But useful tools are built on fundamentals by way of insights.

Re: Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

#40
post #26

Earlier quoted context omitted.

While natural languages and mathematics are two important formalisms by which technologists communicate, it seems to me that we also use a much broader set of visual symbolic systems. These include tables of data in various formats and figures that incorporate drawings, diagrams, charts, pictures and videos. These are now as important as the text and formulas. Problems with tables and figures are one of the main reas…

I'm using the term "mathematics" in a deliberately hand-wavy fashion because people keep wanting to poke at this from various angles and even though the underlying theory is the same, it can appear that you are receiving different answers simply because you are asking the question from different epistemological frameworks. There's a reason Wittgenstein was the man and I'm not. If you like, we can assume that differen…

Apparently the ideas and mental concepts in one person's brain are roughly similar to those in another's, at least similar enough so that common sets of symbols, semantics, and syntax can be agreed such that communication is possible. The fact that various observers get the same impression of the scintillating grid is evidence for some commonality, although I think that in this case it arises from processing in the retina.[1]

I don't see how this obviates the possibility that individuals can have a private language that is in principle not understandable by others, per Wittgenstein. Consider an oenophile who has a most sensitive palate, and who can describe a wine using a whole set of adjectives that are meaningless to most and which must only vaguely represent the actual sensations being enjoyed by the expert. The expert may have a whole internal vocabulary, and due to imprecision of terms, one expert's internal vocabulary may be different from all others. You could say the same for perfumers, cheese mongers, color experts and others who have extraordinary powers of sensation, and which well might be unique to the individual. And the internal language of the individual may not be intelligible by any other because their olfactory or retinal apparatus may not be exactly alike.

I also suspect that this applies to conceptualization as well as sensation. Quite possibly Einstein's internal language was unique to him.

[1]There was an excellent MOOC on the visual system. "Light, Spike, & Sight: The Neuroscience of Vision" - https://courses.edx.org/courses/MITx/9.01x/3T2014/course/

Post reply on HN