2018-08-20

Language Without Metalanguage 1:
Messages Over Deep Structure

To Learn Languages Faster, Abandon Theory

Imagine I were teaching you arithmetic how most language centers teach languages. This is how it would read:
Here's how you get 2 + 2 = 4: Define every natural number by a successor function S(n) on the empty set {}. Every successive natural number as a union (∪) of all of its preceding natural numbers. Since 0 is the first natural number, because it has no preceding natural number, it is equivalent to the empty set {}. We denote the values of the successor set up to 4 with the following values:
  • 0 = {}
  • 1 = {{}} = {0}
  • 2 = {0, 1}
  • 3 = {0, 1, 2}
  • 4 = {0, 1, 2, 3}
Under this model, addition (a + b) is a binary operation defined by the following scheme of the successor function on the successor set of natural numbers:
  • a + 0 = a
  • a + S(b) = S(a + b)
Therefore, 2 + 2 = 2 + S(1) = S(2 + 1) = S(3) = 4.
Now, here's the question: Did this explanation make you better at arithmetic? If I give you a harder calculation, will you faithfully follow these steps and get to the correct result? I bet most of you won't. If you try, it will probably take much more time. This is the problem of metalanguage. The explanation is correct, but it's impractical because the surrounding facts are tangential and more complicated than just saying this:
1 + 1 + 1 + 1 = 4, and 1 + 1 = 2, so 2 + 2 = 4.
That explanation would be incomplete if you were a professional number theorist. However, most people who learn arithmetic don't have that goal in mind. Similarly, people who focus on grammatical structure are not training themselves to be good at a language. Instead, they're training themselves to be good at analyzing it. But, this is not to say that language theory does not have its place. Linguistic metalanguage is useful for building language curricula or analyzing languages' features. The problem, though, is that too many theorists waste learners' time. They're teaching language theory instead of teaching language.

Fine, but how can we learn a language's grammar without being taught it?

Well, there's a problem with the question. The phrase "learn a language's grammar" is ambiguous. There are two modes of learning grammar.

The first is what I've already criticized. It's to learn rules, and then to build sentences from those rules. It's a top-down approach. It's structural and analytical. And, it taxes learners. It forces people to (1) memorize patterns with categorical slots, (2) memorize grammatical categories, (3) remember a list of terms for each category, and then (4) plug those terms into their categorically correct places. A basic syntax tree can reveal this much:

"If you want to master several (types of) languages, then you need not study language theory."
To understand this sentence's grammar in the above way, you'd have to understand at least these categories and concepts: complement phrases (CP), complementizers (C), sentences (S), determiner phrases (DP), inflectional phrases (IP), determiners (D), nulls (), noun phrases (NP), adverbial phrases (RP), inflections (I), pro-forms (e.g., pronouns [N-pro]), verb phrases (VP), valency (e.g., transitive verbs [V-t]), part-of-speech transformations (→), adverbs (R), negations (R-neg), and auxiliaries (also I).

Complicated? Check. Theoretical? Check. Universal? Not quite. Exhausting? Most def.

And, I haven't even begun to explain how debatable the tree is. I didn't explain how "的話" is also a relativizer and a noun phrase that combines with a preceding sentence to generate a noun phrase that indicates the topic of the overall sentence. Even if I did, how important is that to you as a learner? How many articles on this subject are you going to read to find out the truth? The answer: zero.

Oh, and one final question: Who's going to teach this heavy theory to you? You may not know this, but most language instructors don't know their shit. I've taught languages, designed curricula, consulted on projects, and trained developers for over a decade. I've met only a handful of people who knew the relevant theory. Most normal people don't know anything about it. Many of them move to foreign countries and work as foreign-language instructors. Even if they're certified, they almost never learn the linguistics. The result is worse than receiving grammar instruction. You receive incorrect grammar instruction. Their limited and wrong folk grammar becomes your limited and wrong folk grammar. The blind lead the blind to blindness.
Ah, folk grammar! A nice drink of warm bullshit!
So, you don't need the theory and can't learn it easily. But, not all hope is lost. We can learn a language's grammar. We just do so by attending more to languages' messages and less to their structure. This lets us focus on languages' logical form (as understood in both logic and linguistics) as opposed to their theoretical structure. The techniques behind this approach will be discussed in future posts. Right now, though, an example should show the value of logical form over syntactic analysis when we study languages. Look at that same Chinese sentence when it's "message-parsed" instead:

Theory belongs in the background.
With this tree, we only see a language's surface structure. The maximally literal translations then help us understand what the sentences are saying. That alone can give us all of the semantics and syntax that we need. It just needs to be read attentively and from bottom to top. To prove this, I'll offer no instruction. Instead, I challenge you to translate these sentences:
接受挑戰

  • "You want to study theory."
  • "You master several types of theories."
  • "You need not master several types of languages."
  • "If you study theory, then you want to study several languages."
  • "你想學會幾種理論。"
  • "你不用想這麼做。"
  • "你不用學語言。"
  • "如果你想這麼做的話,你就想學這種語言。"
This is what competent "grammar instruction" looks like. The theory is left to the experts (in this case, me). Normal people get a product that eliminates the need to learn complex theory. That's because learners learn from the bottom up, not the top down. Competent materials should have always reflected this.

Okay, fine, but what can I do about that?

You mean, besides not paying money for crap products? You mean, besides using a free app (wink, wink) that was developed with this approach in mind? Well, I guess you could wait for my next posts to learn how to make these message parse trees for yourself, even if you don't fully understand the language.

More immediately, however, you can change how you think about language instruction. You're not learning math or history here. Languages don't reduce to top-down rules. They don't reduce to facts that you memorize and repeat. Language proficiency is a skill. Acquiring syntax is a product of that proficiency. It's not a goal to be sought. You don't learn a new language to learn how to conjugate its verbs correctly, or whatever. You conjugate verbs and say stuff correctly only after you've acquired the syntax, only after you've pursued the language and not the theory. Techne isn't episteme, and it never will be. They each have their roles to play, but unless you're a language nerd for fun or for pay, pursue the former. Nugget, sent.

No comments:

Post a Comment