Solution to Last Month's Mystery Spectrogram

Solution for April 2008

"The good guys wear white hats."

[ð], IPA 131
Eth
Well, there's a nice little fricative here. It's really short, such that it looks like a release, but it ain't. It's a fricative. Kind of loud, but you can tell it's no sibilant--it has too much formant-like shaping to the spectrum. Like aspiration. Which is what I'd probably have guessed what it is if I didn't know any better. The transitions into the vowel (and the spectral change in the frication itself) looks like it's moving from something alveolar. Or at least coronal. So we're looking for a non-sibilant coronal. Vaguely voiced, I thought at the time. Now I'm less certain.

[ɨ], IPA 317
Barred I
Short little vowel, with mostly transitional looking movement during. Reduced. THe F1 looks decidedly low to me, hence barred-i, although that hardly matters.

[k], IPA 109
Lower-case K
The transitions into this look vaguely labial. The F2 seems to be coming down. The F3 is definitely coming down, but the F4 is definitely not. And there's no particular reason for all the formants except F4 to be responding to labialization. So if you thought this was labial, I hear you. Something weird is definitely happeing in F2. Turns out it's probably backing or rounding rather than moving to a labial closure. But how are you suppsed to know? Search me. On the other hand, the messy release looks sort of like a double burst. While not striclty pinchy, the F2 is clearly loweriing into the beginning of the voicing in the vowel, so maybe that's really velar pinch.

[ʊ], IPA 321
Upsilon
Highish vowel (low F1), very back and round (as my back and round vowels go), but transitions towards schwa rather than being steady or moving back. So this is sort of [uə], which is a pretty stereotypical realization of /ʊ/. Well, I'm trying.

[d], IPA 104
Lower-case D
The F3 is sort of sitting there, the F2 is sharply rising to about 1700 Hz. Gotta be an alveolar. Voiced. The clunk I think is the alveolar closure releasing--it doesn't look [s] shaped because there's no air moving from behind it, due to the ...

[k], IPA 109
Lower-case K
... closure back here. Voiceless. Again with a mushy, sort of double-looking release. Fairly clear pinch in the following transitions. So we have to posit a second closure, and it pretty much has to be velar.

[ɑɪ], IPA 305 + 319
Script A + Small Capital I
Ah, diphthongs. Or triphthongs. Or VISC. Or whatever. Look at that movement. F2 hits its minimum just ahead of 600 msec, around 1200 Hz. So quite back. Not at all central. At that same moment, the F1 is rising, hitting its peak just after 600 msec, at which the F1 is between 800 and 900 Hz. So about as low as my vowels ever get. Ordinarily, I'd locate a 'moment' at the F2 minimum, and another at the F1 maximum (since they don't coincide, they must be different moments) and consider this some kind of weird low-low diphthong. Since that's not really a likely option, I'll decide to ignore the two moments, and decide this is a low vowel. But then there's the the long, slow offglide, relatively speaking. Which is a fronting diphthong. Apparently.

[z̥], IPA 133 + 402
Lower-case Z + Under-ring
So here we have a weak fricative. It is [s] shaped, being broad band and strongest in the higher (above 4000 Hz) frequencies. But it's not voiceless. Let this be a lesson--/z/ can be realized as voiceless--probably passively, due to airflow issues rather than vocal abduction. But anyway, the correlates of /z/ involve a shorter, weaker fricative than a corresponding [s].

[w], IPA 170
Lower Case W
Nice little moment of approximant. Loks like a nasal (sorry, no comparisons in the spectrogram) with a strong voicing bar and nothing much in the way of resonances bove. But the transitions indicate something other than 'just' an oral stop articulation. Just look at that swooping F2. Wow. What could cause an F2 like that? Only something seriously back and seriously round. Which aren't really abundant in my dialect. Really, only [w] is really that back and round. And close enough that it can seriously damp the high frequencies. Seriously. Remember that, in case it comes up again....

[ɛ], IPA 303
Epsilon
Okay, this is not the clearest vowel. One the F3 is coming down, probably pushing the F2 down with it as it does. But the F1 clearly separates from the voicing bar, edging up past 500 Hz, if only just barely. Gotta be 'mid' at least. The F2 starts low because of the [w] but zooms up as far as it can unti the F3 starts to push it down again. So let's call this one 'clearly heading to someplace front'. Mid and front. Only a couple of possibilities, and one (in my dialect) typically has an unexpectedly low F1 (i.e. is really a high vowel).

[ɹ], IPA 151
Turned R
On the other hand, whatever else might be going on, the F3 is dropping like a rock. So here we have a coda /r/. But notice what happens, the F2 flattens out and the F3 starts to rsie again. And then ...

[w], IPA 170
Lower Case W
... something else happens. The F2 stays low low low, the F3 appears to at least start to head back to neutral, and the F1 drops again. So we've got something very close, and outrageoulsy round and back. With damped higher frequencies. Look familiar? Good.

[ɑɪ], IPA 305 + 319
Script A + Small Capital I
So here we go again. This one has a clearer offglide, and the F2 doesn't have a moment of its own separate from the F1 moment. Look at that. Who'd'a' guessed these two things were related.

[t], IPA 103
Lower-case T
The plosive here is weak, almost flap-like. Chew on that one for a while as you look around. But in the mean time, there's clearly an [s]-shaped/acute release burst just before 1400 msec, so there has to be some kind of alveolar release here. Unless we're really unlucky and it's just a clunk. It ain't. But I suppose it could ahve been.

[h], IPA 146
Lower-case H
So given that there's a plosive release, is this aspiration, or a separate thing (bonus points for anyone who can clearly explain why this isn't really an interesting question)? Well, notice that the release noise, if that's what it is, isn't 'continued' in the following frication. That is, it doesn't just sort of keep going, as aspiration/release nosie might. The noise is 'clustered' in the formants of the following vowel, so we're clearly dealing with something glottal, and probably not just 'aspiration' of the preceding plosive.

[æ], IPA 325
Ash
Very high F1. Very low vowel. But the F2 is sort of, well, starts high and doesn't really fall to 'low'. So this is not a back vowel. Which leaves a few possibilities, I admit, but if you're putting together words at this point, only one makes much sense.

[t], IPA 103
Lower-case T
Long gap. Gotta be a plosive. The last few pulses of the vowel look like the F2 is transitioning to a labial, but the F3 isn't moving. At all. So possibly, but not probably labial. Probably not velar (but see earlier). SO split the difference. It's a guess. At least this sentence is semantically predictable at this point. It wasn't going to be.

[s], IPA 132
Lower-case S
Or possibly a devoiced [z]. So think about that for a second. How will you choose? (Bonus points available.)