To properly view the phonetic symbols on this page, you must have one of the following fonts installed on your system.

All are available freeware. Depending on which font(s) you have installed, one or the other symbol in paragraph headings may not display correctly.

Solution for June 2003


"Prairie folk are hardy folk."

Lower-case P + Right Superscript H, IPA 101 + 404,
[pH], [pʰ]
For what it's worth, I thought starting with a nice little plosive for a change. It's always hard to spot an initial voiceless plosive because, well, it's a gap. Nothing. Silence. So there we go. On the other hand, this one has a nice little burst, a transient at the moment of release, followed by some nice aspiration. The burst is unfiltered, it's not really stronger in any frequency range than any other, certainly not the higher frequenciies like we might expect for a /t/ or the middle frequencies (which for me in this context is about 1500 to 2500 or so) like we might expect for a /k/). So good bet this is a /p/. Well, good bet it's bilabial. It's clearly voiceless and aspirated, so it's a good bet it's a /p/.

Turned R, IPA 151,
[¨], [ɹ]
F1 is in the mid-range, about 500 or so. F2 starts out just low of mid, but not so low as to be that interesting, but check out that F3. Starts right about 1750. Which is 70% of 2500, which is (about) the neutral F3. Which is right where I said F3 of /r/ is in my dissertation.

Epsilon + Rhoticity Sign, IPA 303 + 419,
[EĠ], [ɛ˞]
On the other hand, while the F3 stays nice and low, it isn't flat. It rises from the onset of voicing (about 225 msec) to a peak about 50 msec later. At that point, the F1 is still resolutely mid, the F2 is resolutely central, and the F3 is still low. So you might be tempted to label this an r-colo(u)red schwa (mid-central), but then you'd have to explain what the difference between an r-colo(u)red schwa and a syllabic /r/ is. So we need another mid vowel, and I nominate front--the F2 here isn't strictly front, but it's being shoved down by the F3--see how close the F2 and F3 are? They don't have to be that close, they just are in this vowel. So I'm thinking this vowel is as front (in the sense of having a relatively high F2) as it can be, considering the F2 has to be lower than the F3. Those of you who believe F3 of /r/ is the second resonance of the vocal tract and the third resonance is wiped out by the side cavity (note the apparent zero in the 2000-3000 range, which is pretty typical, although I think it's not so much a zero as a general depression of frequencies above F3) need to explain to me where this F2 comes from.

Turned R, IPA 151,
[¨], [ɹ]
Well, here's another one. Get used to them. I really didn't intend this spectrogram to be so /r/ loaded, but this utterance actually came up in conversation when I needed something for a spectrogram. Anyway, once again, here's that low F3. Notice the also the amplitude here is a little lower, which to me says this is high-constriction, although the acoustic modelling camp believes it's just a zero.

Lower-case I, IPA 301,
[i], [i]
Okay, I think this is an [i]. But acoustically, the F1 just isn't low enough to be high. At least it's lower than it was for the preceding mid vowel, so maybe this should be small-cap I. But I don't think I get small cap I in this environment. But I could be wrong. Anyway, The F2 is too high to be anything but in the high front area, so either the F1 is weird in this environment or I'm just wrong about the F2.

Lower-case F, IPA 128,
[f], [f]
Well, this is pretty definitely voiceless--nothing at all going down where the voicing bar should be. Even though I high-pass filtered out the noise at the bottom, if you compare the clearly voiced things, this cleary isn't. It's pretty clearly fricative. Now to back up just a second, there are only five voiceless fricatives in English--the glottal, the two sibilants, and the two dentals (labio- and inter-). This fricative is pretty strong, but for its duration, which is over 100 msec, it's not as strong as a sibilant. If this were sibilant, it could only be the post-alveolar--the energy doesn't get stronger in the high frequencies. But it also goes on a little too far into the low frequencies. Esh usually has a zero below the mid-F2 or so, so there's rarely much energy below 2000 and almost none below 1500. And the energy here goes straight down (with a bit of an interruption at 1000 Hz) down to 500. If this were glottal (/h/), it would show resonances excited by noise. This one seems to, I can talk myself into that energy at 500 being the F1 continuing through, and the energy above that seems to connect the F2 on either side. But the F3 range just doesn't look continuous at all, and in generally the energy is just too diffuse, or broad band, to be excited by resonances of the vocal tract. So this must be anterior, either [f] or theta. Exactly which I'm not sure I could tell if I didn't know. So call this an anterior nonsibilant fricatve and move on.

Lower-case O, IPA 307,
[o], [o]
Okay, so look at that F1, right around 500, i.e. mid. Look at that F2, well below the midrange for F2. So this must be fairly back and/or round. The F3 is right about where my neutral F3 is, maybe just a little low, which suggests a bit of rounding, if anything. So what vowel is mid, and back and round? Pretty much only /o/ for me. I note with interest that if you really want to believe this is a diphthong, you'd have to do some hand waving. Looks pretty flat to me.

Lower-case K, IPA 109,
[k], [k]
Well, there's not a lot of transition in the preceding vowel, and nothing after, so all we arely ahve to identify what is going on here is the burst at 800 msec. So let's look at that burst. It's definitely got some shape to it, in the sense that it looks filtered, and so not labial. If you had to call it a fricative, you wouldn't call it /s/, too narrow band and too concentrated in the middle frequencies. I'd be hapier if this were clearly doubled, but it just doesn't look like a clear double burst. But what energy there is in this burst is in the F2 range (and F4/F5), and the low F2 range at that. So that makes it seem velar, if we've eliminated coronal. That's My Story And I'm Sticking To It.

Glottal Stop, IPA 113,
[?], [ʔ]
On the other hand, there's just too much gap between the release of the /k/ and the onset of voicing, so there must be something else in here, from 825 to 950 msec or so. There's something od about the first coule glottal pulses, but they're not so odd that I can convince myself they represent a burst. So what kind of plosive doesn't have a burst? Well, glottal stop springs immediately to mind. Typically the first few pulses of voicing after a glottal stop are irregular (i.e. creaky), but not always, particularly combined with relatively high pitch, which this is (look at the spacing of the pulses. The highest pitch in this utterance seems to be the syllable/vowel between 1150 and 1300 msec, the lowest that last vowel from 1500 to almost 1700 msec.)

Script A + Rhoticity Sign, IPA 305 + 419,
[AĠ], [ɑ˞]
The F1, which would be hard to discern without the formant track, is quite high, indicating a fairly low vowel. The F2 is very low, even lower, if anything, than the /o/. So this is back and quite low. The F3 is dropping over the course of this vowel, which is the hallmark of r-colo(u)ring, at least if you're not surrounded by /r/s.

Turned R, IPA 151,
[¨], [ɹ]
So if the previous vowel is r-colo(u)red, there must be an /r/ here somewhere...

Lower-case H, IPA 101,
[h], [h]
Another fricative. This one looks voiced, which will throw you if you don't know that intervocalic /h/s are almost always voiced. The formant track goes insane through this thing, but I think you can see wht I mean by 'well-shaped' resonances here, at least compared with the other fricatives. The noise here is clearly being supported by cavigu resonances, more or less exactly continuous (that's my story) with the surrounding vowels. Which is /h/ all over the place.

Script A + Rhoticity Sign, IPA 305 + 419,
[AĠ], [ɑ˞]
This should look familiar, but backwards. I want to point out that the F2 moves here suggesting two targets. Even though the F3 starts out in the extremely low, probably /r/, range, the fact that the F2 has one target at the beginning of this vowel and another at the end tell sus that there's two things here. The second one must be the /r/, and the first one must have the low F2. And open-o isn't really an option in my dialect.

Turned R, IPA 151,
[¨], [ɹ]
Once again, there must be one here somewhere.

Fish-Hook R, IPA 124,
[R], [ɾ]
This looks voiced, and although there's some noise slushing around in it, I think we can agree this looks like a gap. But really really really short. Which is how we describe the English flap/tap thing. So that's what this is.

Lower-case I, IPA 301,
[i], [i]
Another vowel. This one with a mid-to-high vowel's F1, and an extremely front looking F2.

Lower-case F, IPA 128,
[f], [f]
Again this should look familiar. I amn concerned about the obvious F2-shaping of the noise, but at least it's consistent. Whatever the first one was, this is the same thing.

Lower-case O, IPA 307,
[o], [o]
At this point, the patterns should be looking very familiar. And except for the pitch, this one looks pretty much the same as the previous /o/.

Lower-case K, IPA 109,
[k], [k]
And once again, you couldn't ask for a better match.