cross-linguistic categorization

[Esa Itkonen]

Dear Funknetters: Many if not most people would subscribe to the following claims:

1) Meaning and form are two different things, and neither should be ignored.
2) Most if not all of the time we are dealing with cluster concepts, i.e. concepts based on several types of criteria
3) Most if not all of the psychological/social concepts are of non-discrete nature, and the description should reflect this fact, however imperfectly.
4) Any given entity may be viewed from several (non-arbitrary) points of view, which entails that, depending on the point of view (or level of abstraction) two entities A and B may be different, partly similar, or identical.
5) It is just as wrong to claim all categories to me psychologically real and to claim them to be psychologically non-real. This is something that cannot be non a priori.
6) It is better to know many languages than few languages.
7) Science is an on-going process, with a ('dialectical') feedback relation between data and theory, which makes it impossible for theory to become totally detached from data.
8) When forwarding several claims, one should make sure that they are mutually consistent.

If one accepts these claims and tries to consistently apply them to cross-linguistic data, it would be difficult (though perhaps not impossible) to arrive at a result very different from what was given in my 'Concerning the Role of Induction in Typological Linguistics' (cf. my homepage).

Still, things should be seen in perspective. The differences at issue cannot go very deep, since everybody seems to agree - grosso modo - on following the lead of such trail-blazers as H. Paul, N.S, Trubetzkoy, J. Greenberg, and T. Givón.

Esa


Homepage: http://users.utu.fi/eitkonen