Re: What is gemination? What are geminates?

Roger Mills wrote:

>Kristian Jensen wrote:
>>The only one I know is Pattani Malay spoken in southern Thailand with long
>>consonants occuring in syllable-initial position; [bulE] "moon" vs [b:ulE]
>>"months". But I'm sure there are more out there, also with a hypothetical
>>contrast like /eb/ vs /ebb/.
>
>Interesting ex.   Any idea how [b:ulE] is pronounced?  Not easy to produce a
>voiced gem. stop without some kind of voiced onset, [@...] or maybe [m...];
>or possibly preglottalization?    I'd be willing to bet this plural form
>originated as a reduplicated form.... bVbulE.   (First thought was, this is
>*bulan with an odd, but not impossible, sound change; is [-an] > E seen
>elsewhere?   There's also *bulaj (= US usage *bulay) 'white, pale', much
>easier phonologically.
Here is what Ladefoged writes in "The Sounds of the World's Languages":

"Word-initial long stops are rare, but they exist, for example, in Pattani
Malay, which has a length distinction for initial consonants. Illustrative
examples are given...

     SHORT                          LONG
     [bulE]   "moon"                [b:ulE]   "many months"
     [kato?]  "to strike"           [k:ato?]  "frog"
     [labO]   "to make a profit"    [l:abO]   "spider"
     [makE]   "to eat"              [m:akE]   "to be eaten"
     [siku]   "elbow"               [s:iku]   "hand tool"

Abramson reports that the mean duration of the closures for the long stops
in word-initial position measured in a carrier phrase is three times longer
than that for the short consonants. Of course, the onset of closure of an
_utterance_-initial voiceless stop has no acoustic signature, and hence the
duration difference cannot be readily percieved in this position.
Nonetheless, Pattani Malay speakers can reliably recognize words
contrasting in initial consonant length in isolation. Abramson suggested
that the perceptual cues that compensate for the lack of information about
closure duration in initial voiceless unaspirated stops might include
intensity of the stop burst, rate of formant transitions, fundamental
frequency perturbations, and relative amplitude of the following vowel. In
a more recent report, he has shown that Pattani Malay listeners are indeed
sensitive to amplitude differences in the initial syllable in forming their
judgements about the category of utterance-initial stops."

-kristian- 8)