Solution for March 2002


"There was no mystery at all."


Eth + Raising Sign, IPA 131 + 429
[], [D], [ð̝]
There's some voicing here, so something is up. It looks a lot like a stop, clearly voiced and with a very short VOT. But underlying voiced stops are usually voiceless initially, and you don't get much more initial than initial in utterance. (Well, you can, but there's very little work on the segmental stuff of higher than utterance prosodic constituents.) The F2 and F3 transitions following start high, suggesting alveolar, or at least coronal. Could be a /d/, but will turn out not to be once you get the whole utterance together. What's the thing that's most likely to forticize to [d]? And then mark it raised. Moving on.

Epsilon + Rhoticity Sign, IPA 303 + 327
[] (no diac.), [E], [ɛ˞]
I really go back and forth on those rhoticity signs. But here we go. Middish vowel, actually might be a trifle high for no apparent reason (check the F1). Starts quite front, which narrows down the possibilities. The diving F3 (hence the rhoticity sign on the vowel) is clearly heading toward an /r/. Following general convention, I picked the 'lax' version of a mid-front vowel, with r-colo(u)ring.

Turned R, IPA 151
[], [], [ɹ]
F3 very low, in this case prototypically below 2000 Hz, although not nearly as low as it might be. Please note, all of you lip-rounding fanatics, that the following segment has an incredibly low F2, suggesting [w] more than anything else, and the F3 actually rises across it. I never want to hear another complaint about lip-rounding actually lowering F3. It doesn't. At least not enough to matter.

Lower Case W, IPA 170
[w], [w], [w]
There isn't much information here, since all the energy above about 900 Hz dies. But note the transitions into and out of this stretch (centered around 300 msec) tell us that the F1 is very very low, as is the F2. Must be very very high, and very very back/round. Note also that if you interpolate or extrapolate or whatever the course of the F3, the extreme rounding/backing going on here has *no* useful effect on F3. None. Zip. Zero. See above.

Turned V, IPA 314
[], [], [ʌ]
Ignoring the transitions from the preceding /w/, we've got a mid-to-low, moderately back or round vowel. There's only so many vowels back there, and I, being from the Western US have fewer than most. I must say, I would have transcribed this with a Script A, IPA 305, except I know what the word was. Actually, when I'm speaking in my fake professional voice, this vowel is very Script-A-ish. Somebody remind me to listen to this again.

Lower Case Z, IPA 133
[z], [z], [z]
Nice high frequency noise, but clearly voiced (note the striations at the bottom). Clearly sibilant, and alveolar to boot (due to the noise), and voiced. Onward.

Lower Case N, IPA 116
[n], [n], [n]
Well voiced, certainly. Sonorant probably. Apparent zeroes in between the resonances, suggesting not only weakness but actual zeroes. Hence probably nasal. Hard to tell what place, since the following transitions are, um, unhelpful, to say the least. Mark it as "N" and go on.

Lower Case O, Top Ligature, Upsilon, IPA 307 + 433 + 321
[ou] (no diac.), [oU], [] (no diac.)
Middish vowel, going from not particularly back to quite round. Get out your IPA vowel chart, and work it out for yourself. (I'm in a hurry this month, okay?)

Lower Case M, IPA 114
[m], [m], [m]
Another sonorant. Less obviously zeroed than the previous one, which might mislead you, except the edges are *really* sharp and clean, which you just don't get with your average oral sonorant. The resonances are different than the preceding one, so it has to be at a different place, and if you're really imaginative you can see that the second resonance in the previous one is at about 1500 and the resonance in this one is definitely closer to 1000 Hz. Hence, of the two, this one must have the longer side cavity, i.e. the oral closure is further forward.

Small Capital I, IPA 319
[I], [I], [ɪ]
Low F1, therefore high. Rather high F2 (though not as high as for /i/), so mostly front. Only so many choices in English, folks.

Lower Case S, IPA 132
[s], [s], [s]
This one is obviously voiceless (unless you are misled by the sound of my refrigerator in the background of this whole thing). And the noise is centered very high and mostly unfiltered at lower frequencies. If you don't recognize this as /s/, you need to go back and practice some more.

Lower Case T + Right Superscript H, IPA 103 +404
[t], [tH], []
In keeping with official IPA position, I'm trying to make sure I transcribe this as a unit, even if I really only segment off the closure portion. There's a very very very short closure portion, unless you think it isn't closed, in which case there isn't. Following a fricative, there's probably enough airflow that it doesn't take much time to build up any pressure behind it, hence it can be short and yet have a sharp burst and high-volume aspiration. Little useful transitional information, so default to alveolar, especially given the short duration. Please also notice the shape of the aspiration noise. There's this pattern where /tr/ gets rendered with something like a retroflex affricated release. That's not what's going on here. Here, the noise is just being pulled down and post-alveolarized by the bunching of the tongue. but you can see how it might be confused with an T-ESH affricate.

Turned R, IPA 151
[], [], [ɹ]
The very high F2 (required by the following vowel) is pushing the F3 up out of the usual range. But this is typical. Just cuz the F3 isn't below 2000 Hz doesn't mean it isn't an /r/, any more than a random F3 just below 2000 Hz must be an /r/. Just ain't so. But you can still tell this is an /r/ because the F3 is low (locally), and in fact about as low as it can get without having to push the F2 out of the way (which it would do, if there weren't another competing demand on the F2. Hence my belief that the F2 and F3 or /r/ are probably perturbations to the second and third resonance of the main tube, rather than side-cavity resonances. But proper acousticians disagree.

Lower Case I, IPA 301
[i], [i], [i]
Very low F1, very, very high F2. Must be [i]. Explain.

Glottal Stop, IPA 113
[], [?], [ʔ]
This is the stoppiest glottal stop I've ever produced. Note the weird voicing quality in the /i/, and the last couple of pulses which are just a little off. If you thought it was an oral stop, look at that 'release'. No transient, the high-amplitude spots of the first pulse/release thing is pretty evenly distributed across the frequency range below 4000 Hz, and exactly contiguous with the formants following. No transitions in the first few 'pulses'. If there were anything oral going on, you'd expect *something*. So this is glottal.

ASH, IPA 325
[], [Q], [æ]

Nice high F1, meaning a low vowel. Not particularly Front, but not particular back either. Lowest not-obviously-back vowel available is /ae/.

Fish-Hook R, IPA 124
[], [R], [ɾ]
Classic flap. Briefest of interruptions to the sonorance and pulsing of the surrounding vowels. SLight bursty transition thing, but that shouldn't trouble you too much.

Script A IPA 305
[], [A], [ɑ]
Very high F1, so very low vowel. Very very low F2, so as back as it gets, given the height of the F1. Low back vowel. Not many choices for me.

Lower Case L + Superimposed Tilde, IPA 209 (155 + 428)
[l] (no diac.), [l], [ɫ (l̴)]
On the other hand, what the heck is going on here? This looks like weird, phrase-ending-voice-quality variant of the preceding vowel, but how many words end with [] (or if you're SIL Doulosing or Unicoding, [A], [ɑ]). Well, lots actually, but notice that there's a moment, at about 1775 msec where there's an abrupt change in amplitude, or otherwise some kind of transient moment. So something changed at that moment. So how many consonants could it be? Must be quite back, may not be high, but definitely involves some kind of oral closure. Well, this one is tough, but once you consider /l/, you're home.