After spending considerable time modeling the structure of schemas and emergent combinatorial systems, I decided that my education needed a more pretentious and overly ambitious direction (which I’ll write about at some later time). I became disillusioned in a lecture that was surveying grammatical properties in some language area with lots of diversity and no consensus on whether it was a coherent topic except on grounds of historical contingency, I think. For each grammatical property surveyed, enough data was presented to allow the students to give 1,000 possible wrong analyses, and one right one, the latter clearly being the worst on grounds of economy and empirical coverage.
Before Chomsky, we all know, linguists were fond of using “discovery procedures” which were bad because they relied on bizarro positivist ideology or something like that. People still think discovery procedures exist. I once sat in a field methods class with 20 students from eight different countries. The first three classes were spent attempting to transcribe one monosyllabic word
Anyway, since Chomsky, you know, there has been much said about methodology. Here is my favorite pedagogical technique for students entering the field, that was once posed to me as a problem for some reason.
Question: How many Xs does it take to screw in a light bulb?
Answer: Two to hold the light bulb straight, while another turns the building around the light bulb.
If there are six things I learnt from these lectures, they are the following: (i) a morphological word can have more than one prosodic word inside it, (ii) a prosodic word can have more than one morphological word inside it, (iii) the prosodic word is the stress domain (unless it isn’t), (iv) the morphological word is the thing produced by the morphological component, (v) tautologies are tautological, and (vi) no one understands Tagalog anyway, so please shut up now.