OK, I just finished watching a segment on ABC Primetime about Foreign Accent Syndrome, a condition I’d heard of, but until now I hadn’t had the opportunity to hear speakers with it. Anecdotal evidence of FAS usually identifies an adult English speaker suddenly (usually following trauma) adopting a foreign accent, and being unable to speak using his or her natural accent. Interestingly, these patients are sometimes characterized as sounding like particular foreign accents or other English dialects. The textbook case is of a woman in England who suddenly sounded German after a head injury during the blitz. The two patients on the ABC segment were associated as sounding Russian in one case and French in another. Hopefully this link will work.
The first time I ever heard of this condition, I thought, how could somebody just “sound German” or “sound Russian” all the time? One implication is that the patient has a latent Russian phonology or German phonology neural structure which, post injury, becomes the only capable neural structure. Of course, this must be wrong, since most English speakers don’t get enough exposure to develop a reliable Russian or German phonological module. So, it’s a Mystery.
I think (part of) the real answer is, they don’t really “sound Russian”, but the people who hear them have no other way of describing it – basically they hear one striking phonological Shibboleth and associate the patient’s FAS with the nationality attached to that Shibboleth. I say this for several reasons.
First, the “Russian-sounding” woman had phonological features of a wide variety of accents, including what might be constured as Russian, Irish, AAVE, Southern American, and her own upper Midwest US. Likewise, the “French-sounding” man had odd prosody and what might have been a French -sounding uvular /r/, but otherwise was another mish mash of other accents.
Second, and crucially, it seems like the syntax and lexicon in these patients are untouched. No funny word order, no inappropriate omissions or insertions of articles, no lexical items from across the pond. Only the phonology is messed with.
Third, (maybe as crucially?), the phonology is really all over the place. Some unflapped /t/, some /ʌ/ coming out as [ʊ], some /e/ appearing as [i] (and I don’t mean e as in see, I mean [e] as in say, /se:/).
So my hunch is this: the patients are not really suddenly adopting a particular foreign accent. They have lost the mapping from each phoneme to its appropriate contextually-determined allophone (with perhaps some orthographic interference involved). In other words, for each segment, the patient produces some allophone of it, but not necessarily the appropriate one. To someone else’s ear, this sounds randomly foreign, and some of the features actually align to a known foreign accent. These identifiable feature are the Shibboleths that prompt listeners to think, “hey, he sounds French!”.
One last question is, wouldn’t all patients have the same exact traumatic foreign accent? No, unless they have the same exact native accent to begin with and the same exact brain injury.