"Snow can dampen sound."
This is actually an observation I make every year after the first serious snowfall. It's just weird how different the world sounds for those first hours after a snow. First of all, all the usual sounds--footfalls, roadsounds, etc.--just don't have the same sound. Secondly, the usual reverby-echoey stuff that we take for granted disappear. The world gets correspondingly small and close. It's just odd. Anyway, that was on my mind when I was putting this together.
[s], IPA 132
Lower-case S
What we have here is a good 200 msec of serious noise. Broadband, and centered above 4000 Hz, this has to be an [s] of some kind.
[n], IPA 116
Lower-case N
Well, aside from that very low-frequency bursty thing at about 275 msec, we've got something that is clearly a nasal. Nice full voicing bar, clear zero(es) above, but evidence of nice resonances. Especially that one just below 1500 Hz. Which is right where I'd hope one would be for an alveolar [n].
[oʊ], IPA 307 + 321
Lower-case O + Upsilon
So starting from 325-350 msec or so and continuing until 550 msec or os there's one very mobile vowel. My new word this year is VISC (which stands for vowel-inherent spectral change) which refers to what I would call 'category-dependent formant dynamism'. But VISC is easier to say and to type, so there wo go. The point is this vowel doesn't seem to have a target and transitions but is intrinsically 'moving'. So let's look at th emovement. F1 starts a little high and ends a little mid-to-low. So this starts sort of lower-mid and rises to mid or higher-mid. The F2 starts just a hair low of central and travels lower. So this starts backish and moves very backer and rounder. So, this being my voice, we have limited choices in the mid-back range, and most of them don't move this much.
[kʰ], IPA 109 + 404
Lower-case K + Right Superscript H
So there's no velar pinch heading into this closure, but that's presumably because of the strong prosodic boundary here. TMSAISTI. Fortunately, the transitions into the following vowel, even through the aspiration. BTW, I've mentioned before that Someone Important once complained about my timescale. It's here that I can see the point. I try to get a comfortable 2.5 seconds in 800 pixels, which is about half the 'traditional' timescale, which is something like 2.5 seconds in 14 inches. Just work with me. Anyway, at this timescale, ti does sort of obscure the fuzzy line between 'long' VOTs and 'short' VOTs. But this is clearly a long VOT, in the sense that you can definitely see the gap between the (mushy) release and the first clear pulse. WIth shorter VOTs, that can be harder to see. But since this is clearly some VOT, we're almost definitely working with something aspirated. So what we have here is an aspirated velar plosive.
[ɨ], IPA 317
Barred I
On the other hand, this vowel is just short. The F2 is closer to the F3, so following Keating et al (1994) I transcribe it as a barred-i and move on.
[n], IPA 116
Lower-case N
So here's another nice nasal. Long. Probably means something, but I don't know what. Anyway, there's that nice full voicing bar, the zeroes, and a nice little pole around 1500 Hz. Sounds familiar.
[d], IPA 104
Lower-case D
This is the first time in a while that This looked like a plausible plosive and not just an oral release. But that's debatable. The voicing seems to die down for a few mpulses around 850 msec, and there's a nice burst and a pulse or so of irregular voicing before the regular pulsing sets in. So this has to be a plosive release of some kind. The F3 transition isn't doing much. THe F2 is sort of high, but isn't really reaching up to the F3, so this is an alveolar.
[æ], IPA 325
Ash
I'm getting really tired of my low front vowels. I'm going to have to do something about them. The problem is I don't know how much of them is the way I represent them and the way I let the coarticulate with other things. Maybe it's the following nasal. Oops, I shouldn't have spoiled that. But anyway, we've got a weird vowel here. F1 starts quite low but rises hitting a steady(ish) state in the last third or so of the vowel. At that point, it's fairly high. So by usual stuff, taking the steady state, we'd say this was a low vowel. But 'tensed', perhaps by the nasal. Although now that I think about it, there's no reason to raise a low vowel before a nasal. Anyway, F2 starts quite high, indicating something front. But then about one third of the way through, it starts to fall to a back position. Weird. So taking the steady state, we'd say this is front. So let's just say low and front and gloss over the difficulties. Like the F1 and the F2 are steady in two different and non-overlapping parts of the vowel. Argh.
[m], IPA 114
Lower-case M
So we've got a nice little nasal. Weak, and the voicing falls apart. I should proobably have treated this and the following release as a presnasalized stop, or an orally released nasal, or whatever we call these [mp nd] things. But the nasal portion of this thing starts with a pole just above 1000 Hz, which is rught where I want it to be for an [m]. Also the F2 transition in the vowel is just too low to be anything but labial..
[p], IPA 101
Lower-case P
Well, in the absence of any other evidence, guess homorganic and move on. Really. There's just nothing to this thing except the wonky release which doesn't tell us much.
[ə], IPA 322
Schwa
This is a classic schwa--(near) evenly spaced formants, short, low pitch (see the distance between the pulses), all of which points to deaccented, reduced, etc.
[n], IPA 116
Lower-case N
And here's another nasal. It's hard to see, but there is apole, more or less in line with the F2 in the previous vowel, and clearly higher than that hint of a pole in the [m]. So this is another alveolar.
Okay, at this point, some of you will be trying to fit in another vowel. Which is fair. But I swear I didn't produce one, and I don't really hear one. I think that's just a transient closure approaching the fricative, overlapped with the nasal. So there's a moment here where there's simultaneous nasal and fricative going on, and the point is it's all transitional so I'm going to ignore it. TMSAISTI.
[s], IPA 132
Lower-case S
So we've got a fricatative. Tilted to the high frequencies, one broad band, etc. Must be [s].
[aɪ], IPA 304 + 321
Lower-case A + Upsilon
So this vowel, and I'm looking from just after 1600 msec all the way to about 1950 msec, the F1 is very flat, amazingly so, and very high. So we have something very, very low. At the beginning, the F2 is steady and fairly neutral, so for something this low we've got something fairly central-to-front depending on how you think of the lower space. But then it starts to move, toward a back, round vowel. So I've transcribed it as a standard diphthong, even though the F1 doesn't seem to move at all. Which is why I don't like digraphs for diphthongs, which make it look like the glide portion is of equal importance to the nucleus, which it ain't. Conventionally we've got a nucleus and a movement, not two targets. I guess. I mean, I think there's a difference, at least phonetically.
[n], IPA 116
Lower-case N
Nasal. See the full voicing? No evidence of anything placewise--the F2 transition is very labial looking, but the F3 is rising if anything, so that could be alveoar. Or both--tongue moving one direction, lips moving one direction, and both having their effect on the output acoustics. Too bad the pole isn't very visible. Can't be velar. so must be bilabal or alveolar, unless it's coarticulating with the following plosive. Which it porbably is.
[d], IPA 104
Lower-case D
In keeping with this spectrogram treating homorganic oral releases to nsasals as separate elements (I don't know where my head was). The release noise is definitley tilted into the higher frequencies, even though it isn't classically [s] shaped. There's enough voicing in the release that you can see the F2, and it's nowhere near the F3--no pinch. So we're dealing with something that is either bilabial or alveolar. The sharpness of the release and its broad frequency range suggests alveolar. And there's not a lot pointing towards labial. So all things considered alveolar is a better guess--and I'm betting an [nd] will make a better word with [saɪ] than [mb].
| Robert Hagiwara, Ph.D. Dept. of Linguistics University of Manitoba Winnipeg, Manitoba CANADA R3T 5V5 |
How To - Research - Courses To the Lab - To the Department - To the University |
 |