In Defense of Grammar

Here are some other language blogs worth checking out:

Steve Kaufmann at The Linguist

Benny at Fluent in 3 months (not sure about Benny's last name, although I have a theory that "Benny" is his last name and "Irrepressible" his first).

If you read both of these regularly (as I do) you will find some fundamental disagreements between them on the best ways to learn languages; on the other hand, they seem to agree on other points, specifically:

(1) Traditional language classes are a waste of time; and

(2) Studying grammar is a waste of time.

Today I'm out to refute hypothesis (2). I'll leave hypothesis (1) for the future.

As I understand it, the main arguments against grammar study run as follow:

1. Grammar is scary and frustrating. The terminology is unfamiliar and confusing.

2. Real fluency in a language demands speaking intuitively, without stopping to analyze what one is saying.

3. Listening and repeating (the way babies learn) is more "natural" and preferable to the "artifical" approach of learning rules and memorizing vocabulary.

Overall, I think the attitude of wanting to study a language but not wanting the study the grammar is misguided. It's like wanting to learn ones way around a foreign city but not wanting to use a map, because maps are covered with intimidating symbols, and someone who really knows the city wouldn't need a map, and babies don't use maps anyway. However, to address these arguments point-by-point:

1. Grammar is scary and frustrating. This sounds to me like the real issue is grammar doesn't yield immediate gratification. Someone looking to learn a new language ought to be the last to object to having to learn new words. And of course the terminology is unfamiliar, because the concepts are unfamiliar. This is an important part of what you get with a new language anyway--a new way of looking at the world. And this particular new way of thinking ultimately streamlines language learning.

I experienced this myself just recently, in Arabic class, when it comes to understanding why nouns end with the vowel a in some situations, u in others, and i in yet others. Having previously encountered noun cases in Russian and Sanskrit, or even Latin (after which Arabic noun cases are a day at the beach), the explanation made immediate sense. My younger, nimbler, but naive classmates, unfamiliar with the concepts of "nominative", "accusative", "genitive", had a vastly more difficult time understanding what's going on.

2. Real fluency demands speaking intuitively. It's true that a fluent speaker can't be stopping to think about "rules" in the course of formulating a sentence. But the ability to step back occasionally and consciously analyze a sentence is also exceedingly useful, and makes one a better communicator. Drivers instinctively keep to the right side of the road and stop at red lights, but when asked, all drivers can explain the rules underlying their behavior. This makes them better drivers, not worse.

For Sanskrit or Arabic, for example, the development of a formal grammar was an important cultural milestone, and indeed a major achievement of civilization--the realization that this "thing" (language) that everyone uses instinctively could be analyzed and codified. In the age of computers and software, the idea of "grammar" has become an essential ingredient of our technological civilization. Why take pride in one's ignorance of it?

3. Grammar study is "artificial." As pointed out by Khatzumoto (although perhaps in gentler terms) babies are actually lousy language learners. Who else can spend an entire year in a completely immersive environment and acquire only a handful of words and be unable to form even a simple sentence? Children do have a talent for mimickry and the ability to soak up large amounts of vocabulary, but adults can more than make up the difference with rational, analytical thinking.

What's more, what you hear in any language is only the surface layer of something that runs far deeper. Consider the English word "resign". The "g" is silent, so why not write the word as "resin"? Because the "g" still exists below the surface, as you see when you pronounce "resignation."

When I first started Romanian, I used the "child" method--just listening and repeating phrases. Meanwhile I tried to analyze what I was hearing (probably part of my personality, but it's a good idea for anyone). I noticed early on that nouns came in masculine and feminine, but a particular puzzle was that a particular thing could be masculine in one sentence and feminine in another. It was cleared up only when I started reading about grammar and learned that Romanian nouns also have a third gender, which is masculine in the singular and feminine in the plural. Instant clarity.

Could I have figured this out on my own? Maybe, but only by piecing together clues from many different sentences and gradually figuring out the pattern of masculine versus feminine. Why not take advantage of the pioneers who did the analysis before you got there?

A final argument, which may not apply to everyone: I like grammar. It's the same pleasure I get from watching a seagull soar or a dolphin swim--a naturally designed structure, beautifully adapted for the task at hand. The only difference is that seagulls and dolphins exist in physical reality, whereas grammar exists in abstract ideas. But then I'm a mathematician--I like abstract ideas.

(Image above from Gardiner's Egyptian Grammar.)

Practical Joke #8

Photo by TheLichfieldBlog

1. Go to the thrift store. Buy a used baby car seat and a life-sized baby doll.

2. Epoxy the seat to the roof of your car and strap the doll into the seat.

3. Drive around town. Even better in the rain and/or snow.

A Face Language

You know that strange nagging feeling you get sometimes when you meet someone new--that you've seen this person somewhere before? Eventually you might figure it out--the person has a resemblance to someone else you know, either a celebrity or a personal acquaintance. I had just this nagging feeling about one of my students last year. Eventually, to my great relief, I figured it out--he looked just like Dwayne Johnson's (hypothetical) younger, slimmer brother.

When you "get" the resemblance, you might be a little surprised--it may span different races, different ages, even different genders. My advisor said it once: "there are fewer possible faces than there are people walking around."

I thought of a game one could play to test his theory. Take a collection of passport photos. Process them to eliminate differences in skin tone, hair style, or size. Cut each into a top half and a bottom half. The goal is to match up the top and bottom halves. Who wants to bet that you could do much better than random matching?

Another variation on this game: make two sets of photos and have two players (or teams of two players). The game this time is that one member of the team picks a photo at random and describes the face, and the other member of the team needs to find the matching photo based solely on the description.

I suspect that the best strategy in this game is not to describe someone as "eyes close together, nose 35% larger than normal, etc." but rather as "Dwayne Johnson's slender younger brother," or "the missing love child of Walter Cronkite and Britney Spears," or so on.

Since one of the things I'm not very good at is remembering names and faces, I've often wished for a concise, yet precise way of verbally describing a face--again, not in terms of dimensions but in terms of the impression on the viewer: a face language, so to speak. Think of what a boon such a language would be, for example, to describe bank robbers.

I have no good idea how such a language could be constructed, but the proof that it should exist lies in the art of the caricature. A good caricaturist (for example, Mort Drucker's work above) can draw a face recognizable as a unique individual with a few deft strokes. Hirschfeld's work was positively cartoonlike and yet instantly recognizable. And maybe one way to identify such a language is to set a pair of caricaturists to play the face matching game and analyze what they say to each other.

A New Low

Photo by an0nym0n0us

So I sold my old car this weekend (not the one you see at the top of this post). With 185,000 miles, it was still in fairly good shape, but it had developed several small problems... the windshield wipers waved erratically, some of the dashboard lights were out, and so on. I had had the feeling for some time that any time when I was driving down the road all four wheels might pop off simultaneously like you see in the movies.

Among these problems was a slipping belt. This is not hard or expensive to fix--15 minutes with a wrench--but it makes the most awful noise, like the engine is about to tear itself apart. I didn't take the time to fix it because I knew I would be selling the car anyway. Friends who rode with my would hide their faces out of shame when the screeching started up.

Last week, I reached a new low. I was driving down a suburban street, when the screeching started up. No one was around, except a solitary dog trotting down the sidewalk. And even the dog stopped short and stared, clearly thinking "What is wrong with that car?"

Why learn languages?

Interesting perspective from Kató Lomb, Hungarian polyglot 1909-2003. Her book Polyglot: How I learn languages includes interesting anecdotes such as surreptiously studying Russian in WWII bomb shelter and running a tourist hospitality office that was mistaken for a bordello:

We should learn languages because language is the only thing worth knowing even poorly.

If someone knows how to play the violin only a little, he will find that the painful minutes he causes are not in proportion to the possible joy he gains from his playing. The amateur chemist spares himself ridicule only as long as he doesn’t aspire for professional laurels. The man somewhat skilled in medicine will not go far, and if he tries to trade on his knowledge without certification, he will be locked up as a quack doctor.

Solely in the world of languages is the amateur of value. Well-intentioned sentences full of mistakes can still build bridges between people. Asking in broken Italian which train we are supposed to board at the Venice railway station is far from useless. Indeed, it is better to do that than to remain uncertain and silent and end up back in Budapest rather than in Milan.

Launching Persian

Photo by alazaat

So I've embarked on the study of a new language: Persian, a.k.a. Farsi. I intend henceforth to report on my progress at intervals.

In once sense, I am starting from zero, since I know not a single word of Persian. In other senses, not quite. Having studied Arabic, I am already familiar with the alphabet (although Persian adds a few letters). Although Persian is fundamentally more closely related to English than Arabic, I would not be surprised to see some Arabic loanwords appear.

The language most closely related to Persian that I have previously studied is Sanskrit. I don't know how much to expect in the way of similarity.

To get the ball rolling, I am starting with the Pimsleur Persian program. If you are interested in studying a language, you owe it to yourself to try out a Pimsleur program, if one is available for your particular language. Even if I had a class to take, I would try to supplement it with Pimsleur.

With the Pimsleur method, you just listen to the recording and just follow the instructions. An example might go like this:

Tell her your son is a high-school student.

[You respond.]

How would she say her daughter is a college student?

[You respond.]

Ask her where the college is located.

[And so on...]

All the vocabulary, etc. is introduced gradually, in context. There are no explanations of grammar as such (not that I don't like grammar).

A "complete" course consists of 30 half-hour lessons. The most popular languages (French, Mandarin, etc.) have volumes I, II, III--90 lessons in all. You'd be amazed how much you can pick up in those 45 hours. Less popular languages (like Persian) only have Volume I (so far, that is). They are still working on new courses.

Advantages of the Pimsleur program:

(1) You get a lot of speaking practice. In a classroom, you spend most of your time listening, maybe reading and writing, and you speak only when it's your turn. With Pimsleur, it's all you. And dealing with an inanimate object frees you from self-consciousness (although this is not an issue for everyone). The extensive practice really lets you wrap your mouth around the language. Speaking a language is a physical activity, just like dancing or skiing, and just so, practice helps you avoid stumbling.

(2) If you spend a fair amount of time driving alone, as I do, it lets you make good use of otherwise dead time.

(3) You don't have to plan, or make many decisions. Just follow the program.

(4) The program is designed to reintroduce previous terms at a rate to keep you from forgetting them. Paul Pimsleur actually authored several articles on this approach, referring to research on the rate at which new terms are forgotten.


Disadvantages of the Pimsleur approach:

(1) Since it's strictly audio, you don't learn to read and write. (However, many courses include perfunctory "reading lessons." Since these can't be done while driving, I usually don't get around to them.

(2) Your accent may be so good that people will expect you to understand better than you actually do. I'm not kidding. I've heard several people report this.

(3) They don't come cheap. A full course runs more than $300. You can find abbreviated courses of 8 or 10 lessons cheaper, if you just want to try it out. You can also find the full courses cheaper if you search the Web, or rent them, or (best of all) maybe borrow them for free from a library.

Some people that I have introduced to the Pimsleur method are so enthusiastic after completing the first lesson that they immediately go on to the second lesson. This is a bad idea. It's like a second piece of cheesecake--just too much. As a matter of policy I won't go faster than one lesson per day. If I have extra driving time, I use it to go over the same lesson a second time.

Within somewhere between one and two months, I should have finished the Pimsleur Persian course--at which point I'll need to find something else to make progress with. Reports to follow.