consonants of Jarai

The following charts are tentative. To create the tables I have borrowed HTML coding from Wikipedia's article on IPA.

The voiced plosives /b/, /d/, and /g/, along with the voiced affricate /dʒ/, condition breathiness on the following vowels. In two-syllable words in which the pre-syllable consonant is one of these four and the main-syllable consonant is not, the breathiness sometimes extends to the main vowel. (It doesn't always happen, and I'm not sure what the rule is yet.)

The consonants that I have labeled "pre-glottalized" might be something different. I was analyzing them as imploded at one time, but I think that the only two that technically could be be labeled as imploded are /ʔb/ (hence /ɓ/) and /ʔd/ (hence /ɗ/). I don't have enough experience to say for sure what these are, but they're certainly voiced consonants that stand in contrast to their non-preglottalized counterparts.

The presence of /c/ suggests that I should be seeing a /ɟ/ rather than a /dʒ/ as the voiced equivalent. I might reanalyze later, but right now I'm pretty sure that I'm hearing an affricate. And I'm fairly certain the /c/ is right instead of /tʃ/ because it occurs syllable-finally and results in palatalization (off-glide, if I'm remembering my terminology) on the preceding vowel (hence the words [boʲc] 'ground termite' and [bieʲc] 'to squeeze'). (The palatalization preceding the /c/ is also a feature of Khmer.)

The voiced velar fricative /ɣ/ is a bit of an eniɡma. It has decended from /ɾ/, and the Jarai spoken in Vietnam seems to always have a /ɾ/ where Jarai in Cambodia has a /ɣ/. (Interestingly, Jarai in Cambodia also has a /ɾ/, and it is not allophonic with /ɣ/ as far as I can tell.)

Naturally, there are probably more consonant clusters than these. But I have reasonable confidence that any additional consonant clusters would have /-l/, /-ɾ/, /-ɣ/, or /-ŋ/ as the second consonant.

Most of these clusters are found in words that were probably 2-syllable words sometime in Jarai or Chamic history. Eventually the vowel of the pre-syllable would have reduced to /ə/, and then it disappeared altogether. This would be easy enough to check against Thurgood's proto-Chamic lexicon, and I plan to do that at some time in the future.

The /-l/ clusters are straighforward. I analyzed /hl-/ as I did (instead of as /Ll-/) because, even if the phonetic reality is /Ll-/, the phonemic reality is probably /hl-/. I can test this by seeing whether any words starting with /hl-/ ever interchange with the form /həl-/. There may be other tests I can devise, as well. I suspect that there are several more consonants that can cluster with the /-l/ (/pʰ-/, for example).

There are fewer examples of /-ɾ/ clusters because this is a less-common consonant in the Jarai spoken in Cambodia. Typically, /ɾ/ has become /ɣ/ in the Jarai spoken in Cambodia.

I had (still have, in fact) many doubts about whether /-ɣ/ is legitimately clustering. Because it is voiced, it's hard to tell whether I'm hearing /Cəɣ-/ or simply /Cɣ-/. But there are so many of these -- with some being quite clear to me -- that I'm going to assume they exist for now. From an articulatory standpoint, this cluster group doesn't fit with the alveolars /-l/ and /-ɾ/. But because /ɣ/ is a reflex of /ɾ/, I suspect that it has inherited its ability to cluster.

I have serious doubts that /-ŋ/ is a genuinely clustering consonant. If it's not, though, /hŋ-/ will have to be added to the voiceless-voiced consonant chart (I can't type the IPA here), because I'm pretty sure there's no vowel intruding between /h/ and /ŋ/.

The consonants /l/, /n/, and /ŋ/ in syllable-final position are sometimes produced with a sort of echo. This echo is sometimes almost syllabic (to my ear), and I haven't yet fingured out how to represent the sound in transcription. (I should point out that phonemically, the sound certainly is never syllabic.) Some speakers produced this echo regularly, some almost never produce it.

I have used capital letters as Lafont does in his preface. He does not have a representation for /ɣ/, a phoneme that probably did not exist amonɡ the Jarai in Vietnam. He also does not have an equivalent for /Nn/. He represents the glottal stop (/ʔ/) in various ways -- sometimes with a diacritical over a vowel, sometimes with an apostrophe.

Lafont adds two preglottalized consonants that I never heard: 'Nh' /ʔɲ/ and 'L' /ʔl/.

I don't fully understand Lafont's 'Dj' or his 'J'.

vowels of Jarai