[Corpora-List] phonetic corpora as another means to measure language distances

[J Washtell]

Dear Yuri,

I may be showing my ignorance here, but isn't it likely that phonetic  
distance varies as much between the origin of the speaker, as between  
languages themselves? English spoken in various regions of Great  
Britain (which barely differ when written) is a case in point.

I have, on occasion, listening to some people from Newcastle and not  
understanding the words that they were speaking (a common enough  
occurrence for me, coming from Cambridge), thought that they sounded  
rather like native speakers of Dutch, or perhaps one of the  
Scandinavian languages. As I don't understand these languages, I would  
assume that a large part of this association was to do with the  
phonetic content. Similarly, on rarer occasions when I have heard  
people from my own part of the country speaking and for a moment not  
understood what they were saying, I have registered it as something  
more like German.

I don't think it is even an issue of regional dialects: it seems to me  
that the phonetic signature of any language spoken by a non-native  
speaker, or spoken as a second language, is generally skewed heavily  
towards the phonetic content of their native region and/or first  
language.

Can I ask how these issues factor in your research? For example, are  
you taking spoken corpora for each language from a very broad  
cross-section of speaker-regions, or are you also measuring phonetic  
distance between speaker-regions for a broad cross-section of  
languages and using this as some kind of contrast?

Justin Washtell
University of Leeds


Quoting "Tambovtsev: Yuri, Alina and Yuliana" <yutamb at mail.ru>:

> Dear Corpora colleagues, some linguists ask, why we should collect   
> phonetic corpora. Surely, one can find many ways to use phonetic   
> corpora. Our task is to use phonetic corpora as another means to   
> compare languages and to measure language distances. We are looking   
> forward from linguists who'd like to join our group. It may be one   
> more method which can add some additional information to classify   
> world languages into different taxa: subgroups, groups, families,   
> units, unions, etc. Looking forward to hearing from you to   
> yutamb at mail.ru  Yours sincerely Yuri Tambovtsev



----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.


_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora