theory-neutral grammars

Rob Pensalfini rjpensal at MIT.EDU
Tue Feb 4 16:23:03 UTC 1997

I like Mark's guidelines for writing ACCESSIBLE grammars, but I've never
seen a grammar where theory 'doesn't intrude on description'. Orthography
is theory intruding on description. Classification of sounds into phonemes
is theory intruding on description. Calling something a vowel, a glide, a
morpheme, a word, a verb etc. is theory intruding on description.
The only way to prevent theory from intruding on description, in my
opinion, is to present hours of randomly recorded (unelicted, of course)
sound, untranscribed.

Of course, I realise that there are different levels/degrees of theory
intruding on description, and that a description can be neutral between
Formal (named) theories (eg. neither Minimalist nor Phrase-structure nor
Systemic), but I do not see how there can be a truly theory-neutral

Of course, all this should be taken with handfuls of salt, since my own
work seeks to separate (theory-laden) description of Jingulu facts from
hard-core formal analysis thereof. I originally sought to write separate
sections: description and theoretical analysis, but I soon realised that
the theoretical analysis of the descriptive 'facts' was entirely dependent
on how the 'facts' were presented. This led me to my current position on
the impossibility of theory-neutral description.


Endangered-Languages-L Forum: endangered-languages-l at
Web pages
Subscribe/unsubscribe and other commands: majordomo at

More information about the Endangered-languages-l mailing list