03 January 2014

Cross-language interference and failing to learn by exposure

Last month I was talking about the dangers of taking a hard-line view in your learning techniques. The specific example I used was spaced recognition software, and unfortunately my post focused too heavily on the example and not on the general principle I was trying to highlight.

Well, as it turns out, I stumbled across another example on a visit to a very wet and windy Edinburgh two days before Christmas. I waited for the worst of the rain to pass before leaving Haymarket station, then headed towards the centre of town. On my way, I passed a Turkish barber's with the following phrase in the window:
APPOINTMENTS NO-NECESSARY
Now the error here isn't as bad as you think, because this was in Scotland. In central and southern Scottish dialects, there is a distinction between the two functions that the English word not carries: negation of a preceding verb is carried out with nae and negation of a following adjective with no (as I understand it, in the north both situations use nae). The owner has obviously learned this use of no through spoken usage, and therefore has no explicit, formal knowledge of how it functions.

The use of the hyphen indicates that he views it as a prefix, closer bound to the adjective than it truly is. This is no surprise to me, as I know many Spanish people who pronounce phrases like no bad as though they are a single word. Given that this sort of colloquial speech is never taught, it is clear that the Turkish barber and my Spanish friends learned the structure by exposure.

Now I have always respected the role and importance of exposure in the learning of a language, but there are those who would exaggerate that importance and promote exposure to being the single determining factor of language learning, and that everything else -- study, teaching, practise -- is just window-dressing, a distraction to keep the learner motivated until such time as they accumulate enough exposure to "acquire" the language. but here we have an example of a feature that is learned in most cases with nothing but exposure, while all the study, teaching and practise is carried out using Standard English, and I have met precious few who have acquired the structure correctly.

Part of the problem is English's stress patterns. Spoken English has at least three stress levels: primary stress, secondary stress, and unstressed -- Spanish has only stressed and unstressed, as do many other languages. They just don't seem to recognise the three levels in English without conscious teaching. Worse: the opposite of necesario in Spanish is no necesario so they've got a pattern to match it to that misleads. Yes, innecesario also exists in dictionaries, but it's not the most common form (the presence of a double N marks it out as a pretty antiquated form), and there is a tendency for Spanish words to replace an in- prefix with no, and crucially, this no is spoken indistinguishably from a prefix -- it could just as easily be written nonecesario as no necesario. (The reason it isn't is probably just the usual case of orthography being a bit conservative and etymological, rather than a perfect model of the modern language.)

And Turkish, of course, is an agglutinative language, so prefers affixes to particles in almost every situation.

So as a soft-liner, I would say that a lack of conscious awareness or directed practise is to blame for the failure to learn this structure correctly, and that exposure, while a vital part of full language acquisition, couldn't correct for that, regardless of quantity and intensity.

A hard-liner would put it differently, claiming instead that these people simply aren't getting enough exposure. With my Spanish friends, I could accept that, but not with a barber. How many professions offer the same opportunities for exposure as this? Have you ever had a haircut without getting a long conversation thrown into the bargain?

The problem, as I've always said, is that all language has a high level of redundancy -- there's more information encoded in the language than we would need to understand the message. (It evolved this way in order to allow us to understand each other even with background noise. If you count up in English from 1, you'll be saying a different vowel every time up to and including 8 -- the numbers are so different that it would be very difficult to confuse one with another.) So if you don't need everything to understand the message, why would you even notice it? "Good enough is good enough," as they say, and the brain has no motivation to notice more when it has already got the message.

No comments: