To properly view the phonetic symbols in the text below, you must have installed either SILDoulos IPA93 or Lucida Sans Unicode. If you are desperate to see phonetic symbols in SIL Sophia or SIL Manuscript, or some other kind of Unicode, you may volunteer to teach me how to do 'family' font calls, in the (now deprecated) FONT tag. Or teach me how to do it with a cascading style sheet.

Solution for July 2002


"It's an odd place to find it."


Eth + Raising Sign, IPA 131 + 429
[D3], [ð̝]
We can debate the extent of this segment until the cows come home, but here's my thinking. The left edge (about 125 msec) suggest a burst. The segment is voiced, and it looks a little like a nasal, in thtitseems to have a little bit of a resonance at about 1400 Hz and zeroes. But, if you can make them out, there's a little bit a noise in the upper 'poles', particularly as they move into the following vowel and the resonance of the upper formants (which is discontinuous with the apparent upper poles). So if you think this is a nasal, hold on to that. It's really a dental affricate, which is typical for initial Eth.

Schwa, IPA 322
[], [ə]
Teeny short vowel, especially for an utterance-initial syllable. This suggests a stressless vowel, which suggests reduction. Following the usual conventions, this is transcribed as Schwa.

Lower-case S, IPA 132
[s], [s]
Fairly typical sibilant, nice dark noisy energy, getting strongest in the highest frequencies. I'm tempted to suggest that you can tell this is syllable initial (or rather that it forms a syllable-initial cluster with the following stop) because the noise increases in amplitude until the stop closure. Somebody who works on syllable affiliation and fricatives should tell me whether that's really true or not, or if it's just that pressure always rises as you approach a following stop. Stay tuned.

Lower-case T, IPA 103
[t], [t]
Stop (ignore the reverberant noise in the upper frequencies), and voiceless. The transitions following the release look a little velar, in that F3 seems to be rising and F2 seems to be falling (i.e. like they start in a velar-pinch configuration). Note though that the burst/release thing is a) really loud, b) not at all double-looking, considering its apparent loudness, and c) not centered in the F2-F3 area (as would be typical of a velar release) but in F3 and F4 (and if you work at it, higher). Also, the closure itself is a little short (if we're still thinking velar), and the highest visible frequencies of the burst are too strong. So limited positive evidence either in the velar or bilabial direction, so pick alveolar. Note that this gap is clearly voiceless, but the stop is not aspirated.

Open O + Rhoticity Sign, IPA 151 + 419
[], [ɔ˞]
This is the backest vowel I've ever produced. Note that the F1 is fairly plainly mid, around 500 Hz. but the F2 is, depending on where you measure it, just below 1000 Hz. That's just outrageously back and round. The /r/ colo(u)ring (the lowering of F3, in anticipation of, well, what's coming up next) doesn't start until almost 100 msec in. The result is spectacular--there's actually a steady state where the F2 and the F3 are flat at the same time. Okay, it occurs very early (from about 490 to 525 msec), but there it is. Mid, back and round. That's only /o/ for me, but given that it's /r/-coloured, I suppose open-o is better.

Turned R, IPA 151
[], [ɹ]
No steady state, and I like to point out that the F3 minimum, which is the point at which I tend to measure these things, is not accompanied by any kind of indication that F2 is doing anything in particular, beyond moving from where it was to where it's going. F3 is nice and low, 1650 or something like that. That ought to be enough.

Lower-case I, IPA 301
[i], [i]
Okay, there are some vowels that don't reduce quite as drastically as others. /o/ and /i/, mostly, although /e/ particularly in apparent compounds (yesterday, monday, etc.) sometimes pops up this way. Then there's the whole if-it-doesn't-reduce-it-must-bear-some-stress argument which I don't believe. Anyway, this is an /i/, but reduced. Note the F2 goes way high (although not as high as it might be). The F1 looks like it stays mid, but that's partly an illusion having to do with the bandwidth and the underlying harmonics. It doesn't lower much tho. Hence, reduced. The proximity of F2 and F3, while it might be due to F3 being relatively low at the point the F2 reaches its max, suggests /i/ over /I/. Depending on who you believe, this could be /I/ underlyingly or even allophonically, but it doesn't sound like an /I/ to me.

Barred I, IPA 317
[], [ɨ]
The F3 is moving throughout, suggesting that, while a vowel, this vowel doesn't have a particular F2 target (following Keating, among others). If redution involves the relaxing or deletion of targets, this is what you would get. Also, given that this is a sequence of vowels, one of them has to be reduced, if not outright non-syllabic. This is too long to really be non-syllabic, and it doesn't look much like one of the usual off-glides. So this probably reduced. Following convention, since the F2 is closer to the F3 than the F1, I've transcribed it barred-i.

Lower-case Z, IPA 133
[z], [z]
What noise there is is in the high frequencies. It's not particularly loud, so you might miss it. But there it is. Nice striations at the bottom, indicating voicing.

Schwa, IPA 322
[], [ə]
Short, obviously reduced vowel. Pick one.

Lower-case B, IPA 102
[b], [b]
Nice little stop, fully voiced. The F2 and F3 transitions, such as they are, suggest bilabial, over anything else, although they're subtle. The release burst (at about 990 msec--the first 'pulse' of the following vowel) is not strong in the high frequencies or middle frequencies as it would be if it were alveolar or velar. As one would hope. Also, bilabial stops tend to support voicing in a way that alveolar and velar stops can't (extra points if you can explain why).

Lower-case A + Upsilon, IPA304 + 321
[aU], [aʊ]
This is a diphthong. Trust me. I used to always use Tie Lines with diphthongs, but it's just too much work in Unicode. Anyway, The F2 goes from something vaguely neutral to something clearly more back. The F1 isn't helpful, in that it's sort of just inverted-U shaped. But it clearly hits an extremum (maximum, in this case) at about 1075 msec, indicating that something in here is low. So low vowel, backing diphthong. Review the history of English, and there you go.

Fish-hook R, IPA 124
[R], [ɾ]
Nice little flap. A gap in the spectrogram, but incredibly short. Too short to be a 'real' stop. Nice little flap. Which means that this is probably a /t/ or /d/.

Schwa, IPA 322
[], [ə]
Another short little vowel, and given that the previous segment is a flap, this vowel is unlikely to be stressed. Get on with your life.

Lower-case M, IPA 114
[m], [m]
Now this is a nasal. Fully voiced. Nice little zero. Actually more than one. Without any other nasals in this spectrogram to compare it to, you'd have to know my voice pretty well to see that the pole (F2) is just a little low (for my nasals). Which is typical of my /m/s.

Barred I, IPA 317
[], [ɨ]
Lots of unstressed vowels in this spectrogram, but that's why it can be long and still readable. See above.

Lower-case K + Right Superscript H, IPA 106 + 404
[kH], [kʰ]
Following strict IPA conventions, aspiration is a diacritic mark on a (stop) segment. UCLABET, following ARPABET and a few others, suggest marking closure phases indpendently from release (plus aspiration) phases, which would make segmenting easier, but whatever. There's frontish vowels on both sides, so this is fronted (i.e. front velar), so the pinch, such as it is, is high rather than mid-frequency. Note that the burst is double, and centered around the F2/F3 (high) pinch area. Dead giveaway for velar. Voiceless during the closure, and strongly aspirated, with loooong aspiration duration. Velars like long aspirations. Extra points for explaining why.

Ash, IPA 325
[Q], [æ]
An incredibly high F1, indicating frankly the lowest vowel I think I've ever produced. The F2 looks like it's moving throughout, and it looks pretty neutral. But for most of it, it's just a hair high. So vaguely front, and very low. This is English, and this is me. Must be Ash.

Fish-hook R + Tilde, IPA 124 + 424
[R)], [ɾ̃]
Nasalized flap. I love this. How the heck are you supposed to know? Well, there's this very short thing. If it weren't so short, it would look like a nasal. Hence nasal flap.

Barred I, IPA 317
[], [ɨ]
Okay, this really looks like an /I/. If you want it to be an /I/, fine. But this vowel turns out to be radically unstressed, so I've transcribed it with a barred-i. Argue with me if you want to.

Lower-case K + Right Superscript H, IPA 106 + 404
[kH], [kʰ]
This is a nice example of a stop with velar pinch on either side. Released, definitely. Aspirated, let's argue about it.

Support Free Speech

Robert Hagiwara, Ph.D.

Linguistics Department
University of Manitoba
Winnipeg, Manitoba
CANADA  R3T 5V5

]