Spectrogramorific!

[philena]

So I recorded Irina, whom I met through Joanne, this evening. She's very nice. She has a Russian accent even in English, which is added insurance that she will speak her Russian perfectly. And after we recorded things, I told her about what I was studying (namely, do the vowels ы and а merge or stay separate when they are unstressed after hard hushers, as in words like уничтожат, уничтожит), and we looked at some example spectrograms after we recorded, and guess what? In this very, very initial glance at the spectrograms, we couldn't tell the difference! Both vowels looked like schwa! The first three formants were evenly spaced, and it was almost impossible to guess which vowel belonged to which spectrogram! Of course, much, much more analysis will be necessary, but this is so encouraging!

The funny thing is that I thought I could hear a difference in the vowel when I played back the recording, which is fairly bizarre if the vowels themselves have merged completely. It reminds me of my meeting with Alan Yu at the beginning of this quarter, where we went over the possible outcomes of the experiment, and he said that if the instruments show a complete merger, while the listeners can still distinguish the words from each other, then something completely bizarre is happening. Oh, phonology, how I love thee! How much more wonderful thou art than syntax!*

Something else I have noticed about these experiments is how eager people are to look at the spectrograms what they sound like. First Sushu and now Irina have patiently sat by while I fumbled through praat, so that they could see little fuzzy lines which they have only my word to assure them represent their speech. And then when I show them what I'm looking for, their eyes widen and their mouths open, and they are completely fascinated. Irina told me that she even wanted to see my paper when I was done! How is that for interest? I love when people are curious about what I'm studying (my mother tends to run away with her hands over her ears), and this makes Daniel's admonishment to me all the more true. He told me that the experts in some esoteric field should not wield their esoterica around them like a magic cloak, expecting everyone else to gape and fall dumb at magic words like post-alvealor fricative or aspirated dental affricate. He said that instead we should explain our work simply enough that they can understand (something Valentina would say last year: if we can't explain an idea simply enough for a layment to comprehend it, then we don't really understand it ourselves.), and then there won't be this silly mystique surrounding academia, pushing out people who think they are not smart enough to understand anything beyond intro-level classes. I think he's right, but I'm also very aware that once you have finished up your cocktail party material of the difference between aspirated initial stops (pit) and unaspirated ones (spit), the only productive English infixes (abso-fucking-lutely is the canonical example, often bowdlerized to abso-bloomin'-lutely; although -ma- is catching up. ), and why Daniel's college-age little brother says some of his ls as ws but others as ls (and even there some people weren't so interested), there's not much you can do with phonology. Syntax is easier, because everyone can tell you whether a sentence is grammatical or not, but I don't like syntax. Phonetics seems to be the crowd-pleaser in all of this! Except I can't bring my computer with me, since I don't have a laptop. Maybe I should get one, in order to enable the spread of non-scary esoterica, via my USB microphone and praat.


*Gah! Syntax final tomorrow! Must study!

Comments

[farraige.livejournal.com]

I don't mean to intrude but (a) there are marked dialectal variations especially affecting the quality of the vocalic segments in Russian -- in fact there are certain Northern accents and South-Western accents that do precisely that, merging pre-nucleic [а] and [ы] around central close to mid, whereas the standard accents will have the two vowels in that position as quite distinct. (b) From the articulatory POV the pre-nucleic syllable is very different from the rest of the unstressed lot, while falling into the same descriptive scope; the precision of its articulation is closer to the typology of English secondary stress.

[farraige.livejournal.com]

The point of my sentiment was that the phonetics of somebody's 'perfect' Russian may be hugely tarred by their regional origins, just as if you were trying to form a general opinion of English diphthongs by using a Cockney speaker as your sole example.

[philena.livejournal.com]

By pre-nucleic do you mean occurring directly before the main stress? If so, then yes, I am aware that immediately pre-tonic (as I've heard them called) syllables are not as reduced as other reduced syllables. (Likewise for word-initial syllables.) What I was studying were the post-tonic syllables, which occur after the main stress, and are about as reduced as they can get.

As for dialects, yes indeed, I understand that. By "perfectly" I simply meant that she would not have acquired an English accent that so often happens to native speakers who have lived in the US for a long time. The regional variations were not so much a concern as the non-regional ones that resulted from her having lived here. She told me that she came from St. Petersburg, and I understand that the Petersburg dialect is one where in general reduced vowels are still differentiated, while in Moscow they all become a schwa.

[farraige.livejournal.com]

S.-Petersburg has a few phonological deviations affecting consonants (namely spelling-to-pronunciation affricates); its treatment of the vowels (the /a/ may be just a little more centralised than that of Moscow) is generally the same as that further south.
Besides I don't think it's possible to have one's articulatory phonetics influenced by their second language. First, the lexicon goes, then the syntax, then the intonation; while it may be possible to lose an accent to a degree in the second language, I'd like to argue that it's not possible to acquire one in the first. However, if you have evidence to the contrary I would be grateful for any literature you could point me towards.

[philena.livejournal.com]

I don't have any literature. I just have anecdotal evidence. My grandfather, for example, lost all German accent to his English, and acquired an English accent to his German, and he came to the US when he was 14, so his first language was pretty much set. But even intonation is important here. What I was going for was a speaker who could sound completely native in the recordings, and intonation will certainly affect that.