Diacritiques

Bill Poser wonders on Language Log whether it is possible for someone who has learnt a foreign language from books only, without ever hearing it spoken or indeed any access to another speaker of the language, to achieve fluency while adopting an erroneous pronunciation that is so deeply ingrained that even subsequent daily exposure to native speakers can’t correct the idiosyncratic pronunciation. This happens to a character from the novel Shibumi by Trevanian.

Two remarks come to my mind.

The first one is about my personal history with English. I read both French and English with about the same speed and accuracy as my native language. Much of my speech is in French. Still, I write and think (and dream) in English more than in French and speak the language daily. But I haven’t always done so. There was a period of about ten years starting at the age of 16, when I dropped the subject in high school, during which I read voraciously in English and thereby improved my vocabulary and knowledge of English grammar, collocations and idiomatic expressions while having hardly any need to actually speak it. In addition, I am someone who — whether this happens to the same degree for everyone I do not know — hears what she read as if it was spoken into my “inner” ear. I therefore continually reinforced my own private version of English pronunciation.

Not that mine was a particularly horrible one (I like to think it wasn’t). I started from a pretty good level after all, and sometimes did speak the language, with visitors (only occasionally native speakers, though) or on trips to other countries. So I never got an entire phoneme wrong, like Trevanian’s fictional character did. But some persistent errors crept in none the less.

What reminded me of the phenomenon described but Bill Poser was that my idiosyncrasies proved relatively difficult to eradicate once I had started actively working on my English again (I also got some formal training in linguistics and grammar later).

The tough errors were of three types:

• Entirely wrongly learnt words, like the non-existent enlargen: I burdened the poor word with two causative affixes where one would have sufficed. Or maybe it was an overgeneralization triggered by embolden. I had always got lengthen, widen and heighten right, though. This one is not a pronunciation issue, of course. (I used enlargen for years in my usenet posts until an online friend was so kind to point it out to me. Whenever someone else wrote enlarge I would read, and my inner ear would hear, my own version instead.)
• Some vowel sounds carrying (primary or secondary) stress in specific words, none of which had come up during my high school English classes. Those were all cases where I erroneously thought I knew how they “had” to be pronounced. Maybe I was applying some elusive rule of my own pronunciation system, or, for transparent words, carried over German pronunciations: vicar (an important word when your reading diet contains big chunks of Agatha Christie and Dorothy L. Sayers), decipher, poll (which to this day I want to make rhyme with doll), sword (the vowel plus the unpronounced w) and cantaloupe (which regrettably doesn’t rhyme with soup) are examples of my former errors.
• The most interesting problem, maybe: stress, in particular in multi-syllable and compound adjectives. I was reluctant to place the primary stress earlier than the second-to-last syllable. But I have been able, with some effort, to change what my inner ear wanted to hear in attributive (I used to place the stress on the third syllable, lengthening it to [bju:]) or obligatory ([obli’geɪtəri] in my version before I corrected it). For compound adjectives (old-fashioned, hard-working…), having never heard of the rule that the stress usually falls on the second element, I carried over German stress rules. I also misapplied those to trade-union, for example.

Secondly, Bill Poser describes a case of what those who study second-language acquisition call fossilization. Timothy Mason has a chapter on the subject on his site. There is a slight ambiguity about what precisely the term refers to:

• The fact that some learners never progress beyond a sometimes surprisingly “broken” level of (grammatical) competence while acquiring good listening comprehension and fluency. This happens even though they may have daily exposure to the target language and certainly don’t lack “correct” input. Fossilization is usually studied from this angle with immigrants: people who have arrived in a new country with very little knowledge of the language, but need to be able function socially on an appropriate level.
• The fact that some advanced, even erudite learners sometimes can’t seem to master particular elements of the target language, even if these are repeatedly explained to them and cognitive barriers blocking their understanding of the rules are unlikely. In this case, theoretical explanation take the native language into account and stipulate that its structure interferes with the learner’s construction of a grammar of the target language.

Obviously, these examples deal mostly with syntax and morphology. It would be interesting to know if there is any research on fossilization from the phonological end, which is not the same as studying foreign accents. I’d be interested in difficulties to produce the correct sounds, intonation and stress patterns during genuine communication in cases where the speaker is able to perceive and produce the correct version when prompted.