>In spite of this disagreement, most seem to agree that a guidelines >document (and therefore a curriculum built on the basis of same) must >include the text- or discourse-level function of clauses/sentences as an >area of study. Anyone dissenting? Shall we debate it? I agree that a guidelines document will be helpful, but I would hope it to be a minimalist document, given the differences among schools of linguistics. It might just have to be that way until more huge data sets get analysed every which way to demonstrate what is truly useful & for what, and what can be left by the wayside. In any case, I think the strongest contribution that a guidelines document can make to rendering the field of language study both exciting and accessible to teachers and students would be to name some of the tensions between perspectives that have been brought up on this list -- to let teachers know that there are different descriptions of language, that the description they choose should be useful for their pedagogical purposes (it should provide the meta-language teachers want to share with their students for talking about the subject of the course, which could be language itself). In other words, give teachers some options. Let them elaborate the guidelines with extra materials as they see fit. The guidelines document could include references to good materials of various kinds. It would posit a common if rudimentary metalanguage that would grow through the investment of teachers' own interests and concerns -- bottom-up, not top-down. my two cents Judy Judith Diamondstone (732) 932-7496 Ext. 352 Graduate School of Education Rutgers, the State University of New Jersey 10 Seminary Place New Brunswick, NJ 08901-1183 Eternity is in love with the productions of time - Wm Blake