Turkish sign language: elicited vs. natural data collection

[Jörg Keller]

Dear Asli,
  basically both methods have their proper place in research. In your special case
>...  for describing a sign language for the first time
the method of data collection I strongly favour is the one you seem to have adopted already - eliciting.
Why? Simply because there is no better way to figure out what are most likely to be full-fledged grammatical utterances of TSL.

Of course, such an approach misses many of
>the characteristics of their "real" language
But regarding your aim, it is the first step in the uncovering of TSL-grammar and its constraints. If I understand you correctly, it is the grammar of space you want to start with, right? You are not - at least at this point - interested in "real language", are you? Real language includes grammar and processing and discourse and so on...

In other words, the richness of "real data" requires the equivalent richness of the methodological means applied to its analysis to make sense. Therefore, by collecting natural data for the purpose of a grammatical desciption, you would - later on - still have to sort out a multitude of influencing factors whose nature is not determined by grammar as such. These include plain errors, local or individual variants, SL proficiency of the individual, slips of the hand, extralinguistic situational/contextual/social/discourse cues present in conversation, elipses - just to name a few.  Indeed, natural data will not show in most cases what is grammatically disallowed.

In order to find out what the core of SL looks like and what is grammatically allowed and in order to delimit the differences between gesture, pantomime, TSL, polyglossia, pidgin, code-mixes, signed Turkish etc., I therefore suggest a theory based approach and elicited data.

I would like to add a final remark of some importance: On a theory based approach, if the adopted theory turns out to be wrong, other investigators may still use the elicited data and reinterpret it in the same spirit. In a more data based ethnomethodological approach, all natural data may be lost to you/others in case you find out - at some later point - that there was some specific, yet important variable that you missed in collecting the data.

Of course, there is much more to say on this matter.....

Regards
Joerg Keller


By the way: Is TSL the proper abbreviaton for Turkish SL? I recall that TSL is  reserved for Tunisian SL.

Who knows????


##################
Dr. Joerg Keller
please use the following e-mail-address for replies:
keller at lingua.uni-frankfurt.de