Monday, December 2, 2019

Kua Fieldwork 2019-12-02: Goals and Report

The goals and report of the following blog entry will give you a somewhat gritty low-level view of the kinds of things that I am doing nowadays in my fieldwork and why I am doing them. It will also allow you to see the kinds of issues that face me in working with consultants.

Starting today Monday December 2 2019, I am working with two Kua consultants (one male, one female) in Gaborone. They are siblings, call them O (male) and TG (female). We will work for at least one week, maybe two depending on their schedules. Most of our work will focus on filling out the dictionary in various ways. Below are my specific goals for the work during this session. If I can accomplish most of these goals, I will be satisfied.

Comment: For a syntactician, the existence of a solid dictionary is very useful. It makes it much easier to transcribe oral texts, and to come up with example sentences for elicitation purposes. The dictionary can also be used to come up with sets of words to investigate certain phenomena (e.g., double object constructions, mass nouns, gender marking, etc.). Most of the languages that I work with (e.g., Kua, Sasi, Cua) do not have dictionaries. So I am obliged to create them.

All the lexical work is now done in FLEx, which makes it much easier to do certain tasks (that is, easier than using either EXCEL or Microsoft word for the lexicon). For example, when adding a new word, FLEx automatically supplies you with potentially related words. Also, FLEx connects the lexicon to the oral texts. So any changes made in the lexicon will show up in the glosses for the oral texts. FLEx also easily connects with Webonary for online publication of dictionaries.

Report: I ended up working with four people, two in the first week (O and TG) and two in the second week (S and SD). Two are men (O and S) and two are women (TG and SD). All are from the same small village, Serinane. The change in consultants was due to scheduling constraints of O, who had to leave to plow after it rained. So I decided to pick up two more people for the second week. S and SD are new consultants, who I have not worked with before. So some of the second week was just teaching them basic consultant skills:

As it turns out, SD has a kind of mixed dialect, so she often produces different words from the others. It took me a while to figure this out, but the smoking gun was the use of the 3SG pronoun, which in Kua is abe (MASC)/ase (FEM), but she said ebe/ese, which is what is found in Diphuduhudu (in Cua). However, she does not use any of the very distinctive tonal patterns from Diphuduhudu (involving tonal depression). So I am not entirely sure what the origin of her dialect is. It could be a mix between Kua and Cua, or it could be some entirely different dialect. Based on answers to questions about her past, I favor the Kua/Cua mixing theory.

One of the consultants during this two week period was a blind person. I was a bit worried about bringing him to Gaborone to work, but as it turns out there was no problem at all. He quickly adapted to the house and the work schedule, and the other consultant helped him to get around a great deal.

Elicit missing Setswana glosses for Kua words.

Report: Done, no problem.

Fill in the tonal contour field for all Kua words.

Report: Done, no problem. The tonal contours are marked in two places in the dictionary. As diacritics on the words themselves, and also in a tone field (e.g., HH, HM, etc.).

Get recordings for TG for whole lexicon.

Currently I have recordings for all lexical items for O. I would like to get recordings of multiple speakers, at least two males and two females, for each word in the lexicon. It is turning out to be harder to get two women speakers than it is to get two men speakers, which is somewhat unusual for me (usually women speakers are easier to locate).

Report: I recorded 379 words for TG out of 1220 (some of these I had recorded for TG before this two week period). I have also recorded 302 words for S and SD.

This kind of recording across a number of different consultants is very useful in seeing what the range of possible pronunciations are (e.g., for [r] or for high tone), ultimately helping to put together a phonological analysis. Recording words with several consultants also often leads to small corrections in transcription and glossing.

Establish genders for all inanimate nouns.

Nouns in Kua are optionally marked with gender (with the exception of proper nouns, which have obligatory gender marking). I have not been very conscientious about eliciting grammatical gender. The practical consequence of this is that I have little data for generalizations about gender marking.

Report: I have verified the genders of 59 words (during the first week). I did this using elicitation. I elicited a sentence with the relevant noun and them modified the gender marking (MASC, FEM or neither) and ask the consultant to judge the resulting sentences. The consultant had no problem at all with this task. Of course, it remains to confirm this data with data from oral texts.

The generalizations that I established in the grammatical sketch all hold. But I have gained some new insights. For example, it is common for a noun to allow either gender, and also common for nouns to take one unique gender. Another generalization is that when gender is optional it sometimes varies with the shape of the object: tall is MASC, and short is FEM. These are significant and interesting generalizations that I will now include in the grammatical sketch.

For the upcoming Botswana NSF grant on Cua (different from Kua), I will probably assign the task of looking at gender marking to one of the students. They could systematically investigate generalizations and write a paper on it. I will also make it clear to all the participants that we should be establishing gender for all nouns right away.

Resolve CHECKs

For the most part, I can do lexical elicitation without a translator, even though my Setswana is still rudimentary. But for certain issues, I need a Setswana translator. I mark those issues as CHECK in the lexicon and resolve them when a translator is available.

Report: Done, no problem.

Lexical elicitation.

Right now, I have 1145 lexical entries. As a modest goal, I would like to get to 1200 during this session. However, I do not want lexical elicitation to be my main goal this session. Rather, my main goal is to fill in various gaps in the dictionary.

Report: I have now reached 1220 words. Given my experience with the consultants, I feel it should be possible to reach a much higher number of words, e.g., 2000 words. There are other Khoisan languages for which I feel reaching 2000 words would be very difficult (due to attrition in lexical knowledge).

Work through Sasi dictionary

One way to elicit further words is to use the Sasi dictionary and make sure we have the same concepts in the Kua dictionary. This will ultimately allow a comparison of Sasi (Northern Khoisan) with Kua/Cua (Central Khoisan), and help to determine what is borrowed vocabulary in Sasi and what is not.

Report: With the first group, O and TG, I continued to work through the Sasi lexicon. But there is still more than half left to go through.

Verify and record new words

Another way of eliciting new words is to go over all the existing vocabulary with TG, who often offers new words (synonyms) that I have not recorded before. Then, I need to verify and record these new words with OG.

Report: I have been stumbling across synonyms at a slow rate, both in week 1 and week 2. They should all be verified with O, who has the status of the main/central consultant on the project.

Resolve inconsistencies in transcription in dictionary.

This includes doublets (words that are identical, but which have slightly different inconsistent transcriptions), inconsistencies in the transcription of vowel sequences, inconsistencies in the transcription of tone (e.g., there should be LM, but not LH contours), inconsistencies in the transcription of the glottal stop, etc.

Report: I stumbled across a few cases of doublets, but I did not look at it systemantically. The big inconsistency in vowel sequences is [oa] vs. [ua]. I am pretty sure these are non-contrastive, but the sequences often sound distinct to me. So I need to decide how I will handle that.

Finish transcribing short oral text (in ELAN and FLEx) by TG.

I have recorded a number of oral texts with OG, but only one with TG. I now need to finish transcribing it.

Report: The short life history of TG has now been completed, at least in rough form. I have also recorded a number additional texts from S and SD. S turns out to have a rich knowledge of folktales, which is a bit surprising to me (based on what I found with Sasi where folktales have been hard to elicit). He told us the following six folktales that he says he learned from his parents: Baboon and Python, Hare and Hyena, Porcupine and Hare, Strange Animals, Ostrich and Ogre, Woman and Ogre. The story "Strange Animals" was a different version a story O told me. I have recorded all of these folktales on video (some with a lavalier mic and some with a boom pole mounted mic).

The life history of TG and "Hare and Hyena" are both now written out in a notebook (transcribed, glossed and translated). I will now enter them into ELAN, and then after that FLEx. I prefer to work in a paper notebook at first because of the sheer number of corrections and notes I make as I am working. Also, I am a bit old-fashioned. It may be that younger people will prefer just to enter the texts directly into ELAN without doing any work in a notebook first.

In our brief experiences with Cua in summer 2019, we found that there were lots of people who had very deep knowledge of Cua folktales, which is probably related to the fact that there are many more people who are fluent speakers of Cua than either Kua or Sasi. One of the goals of the Botswana NSF Cua project will be to record as many folktales as possible (even if we are not able to transcribe and translate them during the grant period). We should be able to get over one hundred such folktales.

Reread entire grammatical sketch, correcting small errors.

The grammatical sketch is basically ready for publication, and has been for three years. I really want to get it published by the end of this year. It is almost ready for the world. I need to overcome inertia and make the final push to get it out the door.

Report: I have been doing this slowly everyday, reading most of the grammatical sketch during the two week period. I have found many smaller errors, but I am also more and more convinced that the grammatical sketch is close to publication.

A few epiphanies occurred to me as I was rereading it. One has to do with the causative suffix -kaxo (LL). I have been transcribing this somewhat uncertainly as LL. The problem is that the second word in compounds and suffixes are often pronounced with a phonetically very compressed pitch contour. I never thought to look at the causative suffix with a following perfective suffix. The form of the juncture morpheme with the perfective suffix is sensitive to the tone of the verb, so that might help to clarify the tones of the causative suffix. I will get this data next time I work with O.

I have also realized that I need to go through the entire grammatical sketch and make sure all the lexical items from it are in the dictionary and that there are no inconsistencies in spelling. Unlike the connection between the dictionary and the oral texts in FLEx, there is no automatic connection between the dictionary and the grammatical sketch. So this has to be done by hand.

Lastly, I did a bit of work confirming grammatical patterns in the grammatical sketch (for possession and relative clauses) with the new consultants during this two week period. There are no inconsistencies at all between O (the main consultant) and the others, even for very complicated possession patterns. So I was happy about that!

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.