This is amusing to me, because it nicely encapsulates some recent thoughts. Note: This is in its application to coding, not my life in general.
I've spent quite a bit of time recently branching out into new programming languages, writing significant works in each one. And each time I do so, I come back to my day job and realize just how limiting it is to only work in one "blessed" language. There are things which are can be eloquently expressed in Rust or Haskell which do not translate nicely over to Python. Sometimes I can do in a single expression what it would take an entire module of confusing metaprogramming to achieve in Python!
Sadly, that street runs both ways. Some of these new languages want to be so safe that they hobble you down with the chains of explicit types and memory protection; making you write pages of boilerplate to do some of the simplest things.
It's true, when writing code, your world is limited by your language. The flip side of that is also true: by choosing a language, you define the world in which you live.
We write with lots of DULs (Domain unspecific languages) in our "blessed" language of Unix, which even Windows speaks a dialect of. The unix language is the one which speaks of "files" and "processes", with a grammar known as "the shell", which is what we use to glue together our DULs (which happen to be files and processes). The reason we need "metaprogramming" in the first place is because there's this artificial boundary called "runtime" where all code must be put into a file, because once you warp the file into a process, it's too late to change the code. Fortunately, some clever people have found a way around this, by making processes which can read other files and simulate the operating system.
When writing code, your world is limited by your operating system too. Heck, we're even limited by the hardware, which hasn't changed conceptually since it was created in the 1940s. We're all under the spell of "instructions to the beat of a clock".
In a similar vein, showing up at my day job after an evening or weekend of playing music, I realize how limiting it is to only work with words, numbers, etc.
Sounds like the strong Sapir-Whorf Hypothesis[1]. When I studied linguistics, back in the 90ies, it was pretty discredited with strong evidence against.
"The Sapir-Whorf hypothesis is the theory that an individual's thoughts and actions are determined by the language or languages that individual speaks. The strong version of the hypothesis states that all human thoughts and actions are bound by the restraints of language, and is generally less accepted than the weaker version, which says that language only somewhat shapes our thinking and behavior." [2]
I think Wittgenstein's idea maybe more broadly applicable, meaning it isn't limited to natural languages like the Sapir-Whorf hypothesis. The author states:
"One way to construe this Wittgensteinism very broadly would be to think of it as the limits of my idiom are the limits of my world, with “idiom” construed broadly to include any way of talking about the world, and not merely a particular language."
So e.g. any kind of formal mathematics or anything that we can somehow represent syntactically and which we can understand (semantics) could be seen as an idiom.
Related to this, there was that recent article about a classical interpretation of the famous double-slit experiment (pilot wave theory) that helps bring quantum physics back closer to more intuitive descriptions
http://www.wired.com/2014/06/the-new-quantum-reality
And this can help for pedagogical purposes, too. Another example is infinitesimals in calculus, which were dismissed in favor of Weierstrass's epsilon-delta definitions over a hundred years ago, but found to be more understandable to students in more recent studies: http://academic.brcc.edu/johnson/Projects/Calculus%20with%20...
What does this say about the Chinese language, then? Chinese has approx 10,000 words in regular usage. Does the apparent low word count have an impact on thinking? It would appear to me that the relatively low word count (ideographs) has not impeded Chinese from thinking about the world around them in a negative way...
Chinese characters aren't what speakers of Western languages would consider "words". Like English words, they are symbols combining both meaning and pronunciation (one or slightly more of each), so sometimes they are like English words, but more often they are used in fixed combinations with each other making them more like roots and affixes (prefixes, infixes, suffixes).
Chinese characters are more like the four components of the word "nonbiological" (non, bio, logos, al) than like the word "nonbiological" itself but, unlike English, there are more cases where the components can also stand alone.
Ten thousand is on the high side for the number of characters an educated, native Chinese speaker would usually know (much less use), but the number of what Westerners might consider "words" known by such a Chinese speaker would typically be much higher than 10,000, with most of them being composites of multiple characters.
In practice, the Chinese language typically represents concepts in 2 or 4 ideograph segments. Not every possible combination is in regular usage, but the number of potential linguistic "words" is closer in order of magnitude to the 10000 choose 2 binomial coefficient, i.e. 49995000.
The saphir-worph hypothesis is thoroughly discredited, i.e. psychophysics shows that your ability to discriminate colors does not depend on the words you use for them. Language is dynamic and adjusts to reality even if people use words as weapons (ex. Illegal aliens)
As for interpretations of quantum mechanics that is a topic for 1970s popular books by drugged out authors and for the marginal fringe of physics journals.
The hypothesis as a whole is not thoroughly discredited; color is an extremely specific case: "The main strand of domain centered research has been the research on color terminology, although this domain according to Lucy and admitted by color terminology researchers such as Paul Kay, is not optimal for studying linguistic relativity, because color perception, unlike other semantic domains, is known to be hard wired into the neural system and as such subject to more universal restrictions than other semantic domains."[1]
That statement by Wittgenstein is far from being the same thing as the Sapir-Whorf hypothesis. For one thing, if you can discriminate colors then you can name the colors, if you want. That is, the limits of your language include being able to talk about those colors, assuming you can discriminate between them. In any case, it's jumping to probably mistaken conclusion to interpret "limits" of language and "limits" of the world that Wittgenstein was talking about as being about same thing as Sapir-Whorf hypothesis concerns.
I think you have a good point, but it is worth noting that being unable to communicate effectively significantly reduces your ability to change the world around you in a meaningful way. If you only have a few words for colors, then you're going to have a hard time exercising fine control over color-based problems.
However, one should also remember that language isn't just words. You can refer to colors by motion (pointing at a chart) or by proxy (using a code like #34F278). So we are limited by our language, but our language is not so limited as to make us fools for not having a word or two.
> [...]psychophysics shows that your ability to discriminate colors does not depend on the words you use for them.
I thought that some of the work on the Himba tribe's perception of colour was, tentatively, showing just that. Their lack of linguistic distinction between green and blue could explain that, in experiments, they took longer to differentiate between shades of blue and green than an English speaker would. It doesn't mean they don't see it though, just that it's a more complex task for them.
I've spent quite a bit of time recently branching out into new programming languages, writing significant works in each one. And each time I do so, I come back to my day job and realize just how limiting it is to only work in one "blessed" language. There are things which are can be eloquently expressed in Rust or Haskell which do not translate nicely over to Python. Sometimes I can do in a single expression what it would take an entire module of confusing metaprogramming to achieve in Python!
Sadly, that street runs both ways. Some of these new languages want to be so safe that they hobble you down with the chains of explicit types and memory protection; making you write pages of boilerplate to do some of the simplest things.
It's true, when writing code, your world is limited by your language. The flip side of that is also true: by choosing a language, you define the world in which you live.