You called “al-” a *phoneme* for crissakes, when it’s a definite article. It functions largely similarly to “the” and other definite articles, so it’s safe to call it that. Wikipedia even talks about its use:

ذكري: when the word being referred to has already been mentioned. An example is found in the word messenger in “We had sent to Pharaoh a messenger. But Pharaoh disobeyed the messenger…” (Qur’an 73:15-6).
ذهني: when the word being referred to is understood by the listener. An example is found in the word battle in “The battle is getting worse; I think we should retreat.”

To call it it’s own magical thing we can’t know doesn’t help us understand how it works in then language at all. Yes, there are other grammatical rules that affect its use, but what it basically does is understood and should not be ignored in favor of mysticalizing it as the unknown. If our linguistic capacity can handle it, certainly our intellectual one can.

Excellent way to put it Relja, I totally agree with you. True grammar is not the same as it’s abstraction (boring grammar), true-natural grammar is almost like a physical skill, something that you feel and live, something that just becomes part of you after being exposed to a language after obscene amounts of time.

… why didn’t I thought about it earlier? I’m jealous xD

I read hundreds of books in English before I finally got to high school English grammar drills. It was boring, and I liked it because I had all of the information already stored in the back of my head. The lessons further solidified my English into the back of my brain so I could continue reading and using English without even thinking about it.

This is where I plan to be with my Japanese, after I’ve read enough, after I’ve got my feet wet with a few hundred books. With a few hundred thousand sentences. With a few million kanji pairings…

Much of what passes for descriptive grammar in the .pdfs of academia isn’t even true. Arabic is not a “VSO” (or any particular word order for that matter) language: to squish and jam its stiff gelatin mass into the square hole that is the “VSO” paradigm, you have to basically find sentences that fit your preconception of the language. (The Wikipedia article on “VSO” even tacitly admits that few languages express this pattern with any predictable regularity). The “definite article” doesn’t work the same in all languages. In Arabic, “al-” is not really at all equivalent to “the”; it has its own meaning that’s impossible to translate into or describe in English without writing about 5,000 words on the subject (and even then you wouldn’t get the spirit of the phoneme).

This is not to toss our hands up and say that language is unknowable, that we should never seek to describe language in a scientific way. But a drastic change in scope is desperately required: we cannot continue studying language as though it were a kind of Sumerian ceramic style, because studying language is really to study ourselves. And just try to compress the full breadth of humanity into a few rules and descriptions.

Since then (or before then, since that’s just a well-known example), pretty much any time that the explanations ended up getting really cumbersome when trying to account for WHAT REALLY HAPPENS, inevitably would come a new epiphany that makes things simpler again if we only look at them in a different way. So yay. Happiness for science and especially the pure and theoretical world that is math.

Abstractions work great for anything that can be boiled down to something simple. Unfortunately languages cannot, thanks to the randomness of humanity. Even most of our deliberately created languages (C++ eh?) have inconsistencies in them, from some new change being applied only partially, because there was a critical point for laziness, or it just seemed weird in some situations. Humans do like order well enough, so they don’t go deliberately making things random and chaotic when determining WHAT REALLY HAPPENS in languages, but are bound by such semi-relevant factors as “this is awkward to pronounce” and “I’m too lazy to say the whole word so I’ll just use part,” which are all in some way related to convenience. (Plus where languages are widespread we have these changes going on in parallel and get dialects).

Humans like convenience better than order. Therefore language (something we use a lot and therefore want to be as lazy as possible with) grows more for the sake of convenience than for the simplistic beauty of pure order. All we “native” speakers know what we mean well enough, so it works for us. Sure, it might be easier to learn for other people if it acted in a way that made sense all the time, but that would take more work on our part, and en masse. Not gonna happen. (Especially in languages that are the bastard children of other languages, like English. Anyone who learned English first should know better. Why then are some of the most strict “grammarists” english-speaking? I blame British Imperial mentality.)

So any attempt to describe the human creation of language as if it were as sensible as objects obeying physics is pretty silly. On a high level we can get it to work, and it can indeed be very useful. It will point us in the general right direction and set up a basic intuition for how things work. But if we try to force it to accomodate the nitty gritty of all individual cases we set ourselves up for frustration and way more work than any insight from it is worth. Because unlike science, no real language has beautiful, underlying universal truths. Just vague generalities.

So I find grammar useful as long as I stay at the vague level of “it works sorta like that, eh?”

To summarize the summary of the summary: Laziness is more powerful than logic.