No subject

[John Myhill]

I don't think that using texts necessarily solves the problems. Some texts
are better than others in this respect. For example, in studying Indonesian
texts it is obvious that academic Indonesian is much more influenced by
European languages than is the language of short stories. And some short
stories are better than others, depending upon the background of the
speaker. So I do not think that using texts is necessarily going to produce
a better result. But I DO know that using bilingual informants and feeding
them questions in, e.g. Spanish or Indonesian or Arabic because I cannot
myself speak e.g. Tzotzil or Balinese or Nubian is going to result in a
pretty peculiar description of these languages.
In terms of what 'better' is, I do not regard the Tzotzil of someone who
knows Spanish well to be generally 'worse' than someone who does not, but
as a typologically oriented linguist I'm interested in diversity, so I
think it's more interesting, and more revealing in terms of linguistic
universals, to study
languages with less interference from others. And I agree with Susanna's point
that there aren't so many people with no influence from other languages,
but there still are degrees.   John Myhill