Datapoint for feature 1A and language wals_code_mzc

Discuss WALS Datapoint for feature 1A and language wals_code_mzc.

3 Responses to “Datapoint for feature 1A and language wals_code_mzc”

  1. Magnus Pharao Hansen Says:

    Silverman, Blankenship, Kirk and LAdefoged (AL:1995) analyze the phonology of Mazatec of Jalapa de Díaz and they analyze 35 consonant phonemes – this is because where Jamieson analyze clusters of nasal and consonant, and consonant and h they analyze as prenasalized and aspirated series tripling the number of consonants and making the syllable structure much simpler with fewer consonant clusters. It seems likely that this analysis would also be better for Chiquihuitlán Mazatec. The same phenomenon is found in the Otomí languages where Palancar’s analysis has about four times as many cononant phonemes as the older SIL linguists analyses – but much fewer clusters. The fact that this kind of disparity between different sources are not clear from the way data is presented greatly hinders WALS’ usefulness and is part of the problem with studies such as Atkinson’s recent highly published study of consonant inventories worldwide based on WALS data. It seems that it would be a good idea to remove some of the entries that are based on outdated linguistic data and include newer and more reliable analyses of many of these languages.

  2. Ian Maddieson Says:

    Hansen’s point about reinterpreting older materials applies equally to any analysis whether based on recent or older materials — the objective must be to find a consistent way to describe each language in a sample so that apparent differences are not due to different analytic traditions. The analysis of Chiquihuitlan Mazatec in WALS does reinterpret Jamieson’s 1977 analysis of nasal + obstruent and h + β and h + ʝ as sequences in favor of an analysis in which there are unitary prenasalized obstruents and two additional voiceless fricatives, which increases the number of contrasting consonants to 21 from the 15 that Jamieson reports. It may well be appropriate to revisit this description some more and evaluate whether his C + h should be analyzed as an aspirated series, and whether his ʔ + C sequences should be interpreted as laryngealized unitary consonants.

  3. Magnus Pharao Hansen Says:

    It is of course correct that it applies equally to any analysis, it seems clear howver at least for Mesoamerican languages that recent studies based on acoustic evidence are much more likely to posit phonemes with complex articulation, wheras older structural accounts tended to aim for the highest degree of parsimony in the inventory. I think it will be difficult to find a consistent way of describing each language while still being faithful to the data that exists and which only ever reaches us in an analysed form. For example in the chiquihuitlan case, it would be equally wrong to assume that the acoustic evidence would support an analysis like that in Jalapa Mazatec, as it would be to dismiss Blankenship et als analysis in favor for one with a smaller inventory. My main point was that there is no way of seeing which tradition the cited source follows, and there is no field in the presentation of data where comments about possible alternate analyses can be presented. Each entry should include comments about the nature of the sources used, what kind of evidence they rely on and theoretical approach. I think this would be a big help, for the linguistically naive reader. Especially when dealing with closely related languages like Jalapa and Chiquihuitlan Mazatec that are presented as having widely differing inventories.

Leave a Reply