November 28, 2006

Very very good pope

This morning I heard the voice of a Turkish woman, with a thick accent, being interviewed in an Istanbul street by an NPR reporter. The woman said:

Jampol very, very good.

Jampol is John Paul; she was speaking about the last Pope (who was popular and well received when he visited Turkey), and she was saying that he thought him a very, very good man.

And it made me realize that there was something amazing about what she said: her English was bad (note the lack of is in her utterance quoted above), but in one way it went beyond anything that had been described in English grammar textbooks by the end of the 20th century.

At least, I can say this much (and the part that follows has been slightly revised since I first posted it, to make it more accurate): I have never been able to find a published grammar which gives a full description of the following fact: pre-head modifiers of both adjective and adverb categories, in noun phrases and adjective phrases and adverb phrases, can be repeated to express intensification of the expressed quality, the number of repetitions being a signal of the degree of intensification (so that very, very good is better than very good; big, big, big problems are bigger problems than big, big problems, and so on).

Now, I am assuming that this is one generalization, not two. That is, I am taking very, very good and a good, good man to be instances of the same phenomehon. This could be wrong. But if it's right, I don't know of any prior grammar that points it out in full generality. The closest approach is in the best of the earlier large grammars, Randolph Quirk et al.'s A Comprehensive Grammar of the English Language (Longman, 1985), page 473: "Some intensifiers can be repeated for emphasis." (It's not really emphasis, it's intensification; I used the word "emphasis" in the first version of this post, but I shouldn't have.) An additional observation is made there: "the repetition is permissible only if the repeated items come first or follow so". That is, so very very nice is possible, and much much too kind, and very much too kind, but not *very much much too kind. Nice point.

What I don't find anywhere in Quirk et al.'s big book is the observation that attributive adjectives can repeat for intensificatory effect as well, as in a good, good man.

When I realized in 1999 that intensificatory reduplication (of both adjective modifiers in the noun phrase and adverb premodifiers in adjective phrases and adverb phrases) needed to be described in the Adjectives and Adverbs chapter of The Cambridge Grammar of the English Language, I rummaged around in all the earlier reference grammars I could find to see what they had said about it, and the answer was that the exact facts had apparently never been recorded. What Rodney Huddleston and I wrote for Chapter 6 of The Cambridge Grammar (pages 561-562) was apparently the first description that dealt with both adjectives and adverbs.

Yet the Turkish woman apparently knew how to do this kind of intensificatory repetition. At least, she knew the part of it that applies to adverbs like very. She has almost certainly never seen either the Quirk volume or The Cambridge Grammar (the latter has only been out four years and costs about $150). So how did she know you could repeat such words for intensificatory effect, when it is almost inconceivable that she could have learned it from a book?

You might say this is no big deal: it's not a difficult thing to learn, and hardly needs a description. I don't know about the first part of that, but the second part is not true. Here's why you need a description: not all adjectives can reduplicate, and which ones can does not follow from any basic principle known to me. This is not relevant for very, which really only occurs as a pre-head modifier; but there's an interesting general point about this not being a matter of mere basic common sense. Notice that in the following examples, the starred ones are not grammatical:

 Just then a huge huge huge spider appeared.
*The spider looked huge huge huge in comparison to the fly.
 Whether Airbus can overcome its major, major problems is not clear.
*Whether Airbus's problems are major, major is not clear.
 We need a good, good man to do this job.
*We need a man who is good good to do this job.

The generalization is simple: you can reduplicate an adjective for emphasis if it's in what The Cambridge Grammar calls attributive function, but not if the adjective is in predicative function. I don't see how you could have guessed that if you hadn't looked at the data and I hadn't told you what the answer was.

There are similar facts regarding adverbs. As pre-head modifiers they can reduplicate, but as (for example) verb phrase adjuncts they cannot:

 They did a really nice thing for me on my birthday.
 They didn't need to do anything for my birthday, really.
 They did a really, really nice thing for me on my birthday.
*They didn't need to do anything for my birthday, really really.
 He is totally awesome.
 The program wasn't eliminated totally.
 He is just totally, totally awesome.
*The program wasn't eliminated totally totally.

Well, the obvious answer to how the Turkish woman learned to reduplicate the modifier very is that she had heard people who spoke English saying very very, and she knew enough to imitate them in this regard.

Another possibility would be that reduplication of modifiers for emphasis happens to be a property of Turkish (I have been told that this statement is indeed true), and the woman tacitly knew just enough about English (namely that very was a pre-head modifier in an adjective phrase) that she was able to unthinkingly transfer it from Turkish to English, and by good luck she was right, because it is a feature of English too.

Somewhat less plausible, in my view, would be a Chomskyan line: that reduplication of modifiers for emphasis is a linguistic universal, held in common by all natural languages and built into human brains at birth or conception, so no one ever has to learn it.

I don't know which is right. But I have sometimes seen statements, by philosophers and other people who haven't done much close-up study of language acquisition process, suggesting that foreign adults learn the rules of the target language out of books, or are told what the rules are by their teachers. In the case at hand, such statements seem extraordinarily implausible. If it's right that no grammarian had written any account of this simple feature of English before 2002, we can be sure at least that any foreign speaker of English who has learned emphatic reduplication of pre-head adjective and adverb modifiers learned it in some other way than by reading about it in a grammar textbook, and any teacher who has ever given a lesson on it deserves to be congratulated for having done some original research.

[Update: I know I'm going to be flooded with mail (I won't be able to answer it all) from people who insist you can reduplicate predicative adjectives, and they'll send me examples like You were wrong, wrong, wrong!. Briefly, let me point out that you have to distinguish among different but superficially similar phenomena. Certainly, you can pause at the end of a sentence after an adjectival predication, and simply repeat the adjective, like this: It is disgraceful. Disgraceful. But in fact it's the whole adjective phrase that's repeated here: It is sad to see. Sad to see.

And you can do the same with any kind of phrase; it doesn't need to be an adjective phrase, it could be a preposition phrase: It is beyond belief. Simply beyond belief. You can even interrupt yourself and do a repetition in the middle of a sentence, as in the famous line from Casablanca: "I am shocked, shocked, to find that gambling is going on here." But these examples involve the intonation breaks associated with parenthetical additions or interruptions (and they are often written with dashes rather than commas). They do not have the quality the smoothly integrated arbitrary repetition for degrees of intensification that you find with attributive adjectives modifying nouns and adverbs modifying adjectives.

My examples above were carefully designed to be unsuitable for the sort of parenthetical restatement I'm referring to here. It's not that no one can find a predicative adjective being repeated; it's that close attention to the details reveals that attributive adjectives and pre-head adverb modifiers are being reduplicatively emphasized in ways that predicative adjectives and post-head adverb phrase adjuncts are not.

One other point: another undescribed feature of English I discovered in 1999, also described in The Cambridge Grammar (page 562) was tautologous use of synonymous but distinct adjectives for intensification, as in tiny little bird (or little tiny bird) or great big hole. And again, this is restricted to attributives: no one says *The bird was little tiny, or *The hole was great big.]

Posted by Geoffrey K. Pullum at November 28, 2006 01:30 AM