full-fledged pidgin


Exercise 3: Hampster Dance

Posted in lsa317, phonetics by clunis on the July 18th, 2007

(this is the last of the LSA 317 exercises I’ll be posting here. the post date/time should reflect when I started working on this exercise, but I didn’t actually end up finishing it until late night on the last night of the institute.)

question 1

No, there is no excrescent stop in “ham strips” in hamster1. The nasal air flow, nasal air pressure, and oral air pressure all release within 4ms of one another just prior to the onset of the fricative /s/. I also can’t hear anything that sounds like a /p/ :)

question 2

The oral pressure in the supposed /ms/ of hamster1 continues to climb well after the nasal air flow and pressure have both dropped off — suggesting an oral stop closure with the velopharyngeal port fully closed.

hamster, ham, dumpster, and dump

  hamster vowel dur pressure ham vowel dur pressure dumpster vowel dur pressure dump vowel dur pressure
1,2 122 69 114 43 108 63 84 65
3,4 127 66 129 61 102 71 78 66
5,6 126 64 115 58 98 72 75 74
7,8 131 60 118 58 95 58 68 46
9,10 127 54 128 52 90 72 66 60
11,12 124 61 120 57 98 60 78 60
13,14 106 64 89 39 119 54 61 60
15,16 112 51 93 40 89 66 77 64
17,18 125 58 96 49 94 65 84 55
19,20 118 54 101 51 110 62 88 41
21,22 124 60 125 32 101 61 77 63

hamster and ham

question 3

It honestly sounds to me like every production of `hamster’ has a [p] in it for this speaker. I’d say the best way to find points on the graph with weak or no stops at all is to look at the points for `ham’. Everything else has a stop either phonologically (dump, dumpster) or emergently (hamster). I don’t see a correlation between vowel duration and the stop.

question 4

Given that most (if not all) of the productions of `hamster’ in these sound files contain an `epenthetic’ [p] and the durations of `hamster’ and `ham’ are essentially equal, I’d say there’s no way to discount O’hala’s conclusion regarding timing. There just isn’t any evidence to refute the claim that a surface [p] will have a vowel duration similar to a word like `sumster’ or `clamster’ with no [p] (or /p/). There may, on the other hand, be enough evidence to conclude that this speaker has a /p/ in his underlying representation of `hamster’.

question 5

Yes, these vowels are nasalized. After an initial decrease in nasal airflow and pressure (indeed, after negative airflow and pressure), they are both above zero and climbing for the duration of the vowel. The mandible is lowering to make room for that open mid back unrounded vowel; this will pull the nasal passages further open. This accounts for the negativity seen in nasal air flow and pressure at the onset of the vowel (expanding the nasal cavity rarifies the air held within it).

question 6

I’m a little baffled by this. In hamster5, if the nasal airflow and pressure are calibrated properly, then the /æ/ in `hamster’ is, surprisingly, not nasalized. Did your speaker maybe have a slight cold or something? Looking at the spectrogram in praat and listening to the audio I’d sure say I hear nasalization and see a faint nasal formant, but the nasal airflow and pressure both seem to read 0 during this vowel. The situation seems identical in `ham’. I believe I hear nasalization and see a nasal formant in the spectrogram, but nasal airflow and pressure seem to be 0.

The sentential context here differs from that of dumpster/dump in that ham and hamster are each preceded relatively closely by another nasal consonant.

As an aside, I’ve noticed that people say that folks sound “nasal” when their nasal passages are so swollen and stuffed that no air can pass through them. Seems like it’d be more accurate to describe these people as “oral”.

question 7

In hamster21, yes. I do believe that the [t] has assimilated to the place of articulation of the [m]. With an alveolar closure after the [kʌ] I’d expect to see the F2 formant transition higher — instead it remains level. The difference between this [p] and the [p] in strips is that this one has an unaspirated release (can a word-medial stop technically be ‘unreleased’? I don’t have any books here to check). It looks to me (from the spectrogram) as if voicing continues into (and maybe through) the closure. I wouldn’t be especially surprised to see this transcribed as [b] rather than [p].

Comments are closed.