Discussion Details
| Title: | Re: A Challenge to the Minimalist Community |
| Submitter: | Oren Sadeh-Leicht |
| Description: | Peter Svenonius (LINGUIST List 16.1491) wrote that:
Then I would use that corpus as the training ground for testing my simulacrum, because P&P (Principles and Parameters) theory is not trying to describe a Language Acquisition Device that can learn a language from the Wall Street Journal (with or without labelled brackets), but a Language Acquisition Device that can learn a language from a learning environment like the one described in the preceding paragraph. If my concerns are well-founded, then building a parser of the kind described by Sproat & Lappin would not even be a milestone on the road to a workable model of language; it would be a detour. And: As a theoretical linguist, I remain unconvinced from the discussion so far that building a parser of the kind proposed by Sproat & Lappin (LINGUIST 16-1156) would be as important as they suggest. These words remind me of the age-old debate between empiricism and rationalism in that they imply that theoretical linguistics is 'the real scientific study': trying to work out how natural language is acquired under natural conditions, whereas corpus parsing is studying just that: how to annotate a given corpus. Therefore, corpus parsing is irrelevant to the task at hand which is answering the question 'how language is acquired'. It appears that the linguistic community should wait for the right theoretical linguist who will put all aspects of language together (intonation, context, syntax and what not) and by that will solve once and for all this problem. Surely this is not the way to go about things. The way science progresses is by simplification and by testing narrow hypotheses. Later on, these might be expanded to become more general. In the study of human motion, scientists build up robots to test the movement of the human hand, for instance, in the hope to apply this practical knowledge to the movement of other limbs. Surely a human being does not move a hand like a robot, and other factors need to be considered (planning a movement, eye-hand coordination, balance, etc.), but just modelling a certain aspect of movement is essential to the understanding of human movement. It is therefore essential to model parsing of a limited corpus, even if it does not contain intonation contours or context like Peter Svenonius would like it, since this is just one little step forward in a scientific study. I think what Sproat & Lappin had in mind, in contrast to what Peter Svenonius wrote, was to show that given the innateness hypothesis, a given corpus will be quicker and easier to 'learn'. That is the main point: if a child possesses a-priori knowledge, the task of learning a language as complex and intricate as human language, becomes easier. They are asking to test this aspect of innateness (or UG), and do not claim that this would have a bearing on real human language acquisition. It would just be a little step forward in a scientific study. |
| Date Posted: | 11-May-2005 |
| Linguistic Field(s): |
Computational Linguistics
Linguistic Theories Language Acquisition Discipline of Linguistics |
| LL Issue: | 16.1506 |
| Posted: | 11-May-2005 |
