I have brought our Kua consultant to Gaborone for two weeks. The most general goal of work on Kua now is to finish the grammar (co-authored with Andy Chebanne) so we can get it published. The only section that needs to be added is a short introductory section on the phonological inventory of the language (consonants including clicks, vowels and tones).
I have been sitting on this grammar for three years, so it is time to get it out there.
As usual, I have worked on things backwards: from syntax to phonology, instead of the other way around. In fact, my previous two grammars (N|uu and =Hoan) have no phonological section at all, but some readers complained about that, so I am now making an effort to add a phonology section to the Kua grammar.
The specific goals for the work in the next few weeks are:
a. Work on lexicon. Currently the number of words is around 700. I would like to get it to over 1000. Why? Because I think it is very hard to establish the phonological inventory of a Khoisan language with less than 1000 words.
b. Make sure we have corresponding words for Kua-Cua-Tsua, to make comparisons easier between the three languages. This allows us to look for tonal and segmental correspondences (and look at things like tonal depression and click loss). This is the kind of micro-comparative methodology that underlies Tim Mathes very interesting NYU thesis on tonal depression.
Because of (b), one subgoal is to work backward from the Cua notebooks which we (Zach, Andy and I) created in July and early August 2019. Every word elicited in Cua, I check for in Kua. This has often elicited new words for Kua and also in some cases second forms which are useful (that is, a single concept having two different words).
The trick with this kind of elicitation is to make sure you are not teaching the consultant a new word. So the elicitation can proceed either from meaning/concept or from phonetic transcription, but not both at the same time. For example, suppose you wanted to elicit Xam "lion". Then either you start from "lion" and try to elicit Xam, or you start from Xam (by saying Xam) and try to elicit "lion" as the meaning. But you should not give both at the same time, since then you risk eliciting a word that does not exist in the language (essentially teaching the consultant a word).
After I finish the Cua notebooks, I will look at the Tsua word list for words that have been elicited in Tsua but not Kua. That way, I will have the same set of words for all three languages.
I will also use the Sasi lexicon to find words to elicit in Kua. This will facilitate Sasi-=Hoan vs. Kua-Cua-Tsua comparison in the future (Northern Khoisan versus Central Khoisan).
Finally, I will make sure that all the words (including function words) that appear in the Kua grammar, also appear in the Kua lexicon. And, I will also make sure that the transcriptions and glosses in the lexicon match the transcriptions in the dictionary.
c. Clean up the FLEx dictionary. There are lots of doubles and also weird issues of font arising from the way it was created (merging a FLEx dictionary, based on texts, with an EXEL word list). Ultimately, the FLEx dictionary should be good enough to print off and send to publisher, or to post on Webonary.
For example, in the oral texts to write an "i" with a tone on it, I used a dotless "i". But in the EXCEL spreadsheet, I used a dotted "i" (with a tone mark obscuring the dot). These words now show up as doubles in FLEx. Lots of similar hard to spot font issues have come up.
Lesson learned: If you are going to use FLEx, use it from the very beginning of your project to avoid headaches. Transferring from other formats (EXCEL or Word) can cause problems that take time to resolve.
d. Resolve issues with the transcription of ejected click sounds: CX'V, Cq'V and C'V. It may be that CX'V and Cq'V are merged (as they are in Sasi). Also, it seems that CX'V is realized phonetically differently from other Khoisan languages I have worked with, involving a uvular instead of a glottal release into the vowel. I also need to establish that C'V exists and that it is clearly differentiated from CX'V.
Apparently, some of the same issues are relevant to Cua, so I am killing two birds with one stone here.
e. Establish phonological inventory of Kua for the grammar, and write up introductory section (consonants, clicks, vowels, tones) of grammar. Try to limit this to 10 pages.
f. Visit consultant's village, and make some videos of people speaking Kua. These videos will then be transcribed in later work. They will also open the way for working with more people on Kua in the future, and checking transcriptions with other people.
g. Take Kua consultant to Diphuduhudu (where Cua is spoken) and test the mutual intelligibility of Kua and Cua by having consultant try to speak with Cua speakers. Why? To get some intuitive idea of how divergent the languages are.
I have other goals for Kua, including writing theoretical linguistic papers on the internal structure of pronouns and the nominal linker. But I will only work on these other things after making headway on the goals listed above.
I have been sitting on this grammar for three years, so it is time to get it out there.
As usual, I have worked on things backwards: from syntax to phonology, instead of the other way around. In fact, my previous two grammars (N|uu and =Hoan) have no phonological section at all, but some readers complained about that, so I am now making an effort to add a phonology section to the Kua grammar.
The specific goals for the work in the next few weeks are:
a. Work on lexicon. Currently the number of words is around 700. I would like to get it to over 1000. Why? Because I think it is very hard to establish the phonological inventory of a Khoisan language with less than 1000 words.
b. Make sure we have corresponding words for Kua-Cua-Tsua, to make comparisons easier between the three languages. This allows us to look for tonal and segmental correspondences (and look at things like tonal depression and click loss). This is the kind of micro-comparative methodology that underlies Tim Mathes very interesting NYU thesis on tonal depression.
Because of (b), one subgoal is to work backward from the Cua notebooks which we (Zach, Andy and I) created in July and early August 2019. Every word elicited in Cua, I check for in Kua. This has often elicited new words for Kua and also in some cases second forms which are useful (that is, a single concept having two different words).
The trick with this kind of elicitation is to make sure you are not teaching the consultant a new word. So the elicitation can proceed either from meaning/concept or from phonetic transcription, but not both at the same time. For example, suppose you wanted to elicit Xam "lion". Then either you start from "lion" and try to elicit Xam, or you start from Xam (by saying Xam) and try to elicit "lion" as the meaning. But you should not give both at the same time, since then you risk eliciting a word that does not exist in the language (essentially teaching the consultant a word).
After I finish the Cua notebooks, I will look at the Tsua word list for words that have been elicited in Tsua but not Kua. That way, I will have the same set of words for all three languages.
I will also use the Sasi lexicon to find words to elicit in Kua. This will facilitate Sasi-=Hoan vs. Kua-Cua-Tsua comparison in the future (Northern Khoisan versus Central Khoisan).
Finally, I will make sure that all the words (including function words) that appear in the Kua grammar, also appear in the Kua lexicon. And, I will also make sure that the transcriptions and glosses in the lexicon match the transcriptions in the dictionary.
c. Clean up the FLEx dictionary. There are lots of doubles and also weird issues of font arising from the way it was created (merging a FLEx dictionary, based on texts, with an EXEL word list). Ultimately, the FLEx dictionary should be good enough to print off and send to publisher, or to post on Webonary.
For example, in the oral texts to write an "i" with a tone on it, I used a dotless "i". But in the EXCEL spreadsheet, I used a dotted "i" (with a tone mark obscuring the dot). These words now show up as doubles in FLEx. Lots of similar hard to spot font issues have come up.
Lesson learned: If you are going to use FLEx, use it from the very beginning of your project to avoid headaches. Transferring from other formats (EXCEL or Word) can cause problems that take time to resolve.
d. Resolve issues with the transcription of ejected click sounds: CX'V, Cq'V and C'V. It may be that CX'V and Cq'V are merged (as they are in Sasi). Also, it seems that CX'V is realized phonetically differently from other Khoisan languages I have worked with, involving a uvular instead of a glottal release into the vowel. I also need to establish that C'V exists and that it is clearly differentiated from CX'V.
Apparently, some of the same issues are relevant to Cua, so I am killing two birds with one stone here.
e. Establish phonological inventory of Kua for the grammar, and write up introductory section (consonants, clicks, vowels, tones) of grammar. Try to limit this to 10 pages.
f. Visit consultant's village, and make some videos of people speaking Kua. These videos will then be transcribed in later work. They will also open the way for working with more people on Kua in the future, and checking transcriptions with other people.
g. Take Kua consultant to Diphuduhudu (where Cua is spoken) and test the mutual intelligibility of Kua and Cua by having consultant try to speak with Cua speakers. Why? To get some intuitive idea of how divergent the languages are.
I have other goals for Kua, including writing theoretical linguistic papers on the internal structure of pronouns and the nominal linker. But I will only work on these other things after making headway on the goals listed above.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.