2012年5月17日木曜日

Project on Quantitative Description of Languages

Project on Quantitative Description of Languages in connexion with its possible use in historical linguistics

I would like to offer everybody to take participation in our project on tasting certain typological parameters which can be useful in genetic linguistics.
Essence of the idea is the following:
Each language is first of all a structure but not just a heap of lexemes.
Because of it we have to pay most attention to structures in any linguistic question but especially while we deal with questions of genetic relationship.
What parameters should be the matter of research is a separated question but the fact that structures/grammar is much more important than lexics is the basement.
After the detecting of structural similarities we should of course find material similarities of exponents.
These two points i.e. structural similarities and material similarities of exponents seem to be necessary and enough to make a conclusion about genetic relationship.
Comparison of lexicon can't be proof of genetic relationship of languages. Still Swadesh himself warned against it. Moreover the term of  "basic lexicon" is very questionable 'cause the whole of lexicon is culturally determined as it was shown by Hoyer.
All structural parameters can be subdivided into two groups: technical parameters and contentive parameters.
technical parameters describe technics of structurising used in a certain language (technical parameters are the following: prefixation development index, stems regularity index, afiтехes regularity index, mophemes joints regularity index and so on)
contentive parameters describe grammatical meanings which exist in a certain language and describe by what means exactly these grammatical meanings are expressed. (contentive parameters are the following: aspect, tense, modality, mood andso on)
And then the procedure of searching some possible relatives of a certain language can be subdevided into the following two parts:
1) analysis of different technics of structurising relevant for a certain language;
2) analysis of grammatical meaning which are relevant for this language: what grammatical meaning are relevant and by what technical means they are expressed.
Moreover, it's very important not just state that a certain feature exists or is absent in a certain language but estimate the degree of manifestation of this feature then estimation will be much more exact and conclusion won't be speculative, other words: all structural parameters should be presented as coefficients.
Now we are engaged in tate of prefixation development coefficient/prefixation development index correlation among languages which are firmly detected as genetically relatives. By this moment we have received the following results on prefixation development index:

And we are interested in data on another langauges and firmly collected groups 9especially in Korean, Sinitic languages, different groups of SEA and Indo-European languages).
To count the menaning of the coefficient one should the number count total amount of affixes of a certain language then cout the number of prefixes and devide the number of prefixes to the total number of affixes.