One of the purposes (and benefits) of the ATEG listserv is that it allows discussion of diferent perspectivies on grammar, which can only be of help to those who teach grammar. The exchange over the last few days on "constructions" and "consruction grammar" has been very interesting, but I suspect that I am not the only one who is confused on two counts. 1. What is a construction? 2. How would a learner recognize that what appears to be a construction in one case is not a construction in another? In other words, how does construction grammar address the issue of learnablility? On 8/3, Johanna described a construction as a "specialization of meaning" and "a sort of fossilization of phrase structure." Johanna went on to describe constructions as similar to idioms but also says that the meaning of a construction is more closely tied to the meaning of its parts than is the case with idioms. It seems that constructions have properties of both unanalyzed wholes and expressions whose interpretation depends on some sort of formal analysis (which all things being equal, construction grammamr would prefer not to do.) Obviously, a construct with these conflicting properties would be hard to identify. At least for me, both of the candidate expression types discussed so far have not made a case for how "construction grammar" would be of interest to grammar teachers. The "go Xing" structure doesn't seem to be a good candidate because its interpretation depends on an analysis into different semantic cagetories (leisure activities vs. other types of activity - go canoeing vs *go housecleaning). Also, it seems to me that the intepretation of these reflexive structures that Bob Yates presented is absolutely dependent on some kind of formal analysis. It doesn't appear that either the "go Xing" structure or reflexives count as "constructions" according to the definition provided by Johanna. Even Johanna's discussion of the interpretation of reflexives makes use of considerable formal analysis. In addition, apart from the debate over the methodology used by grammarians is the bigger question of how learners would recognize constructions without recourse to liberal use of some kind of formal analysis - which Construction Grammar apparently resists. This issue of learnability seems to me to be really important - especially for discussions on this list. I appreciate that Johanna is not an "expert" in Construction Grammar but brought it up on the list as an interesting idea to consider. Perhaps we could see some other examples of structures which would be more transparent to learners if presented as (relatively) unanalyzed wholes. Jim Kenkel, Eastern Kentucky University To join or leave this LISTSERV list, please visit the list's web interface at: http://listserv.muohio.edu/archives/ateg.html and select "Join or leave the list" Visit ATEG's web site at http://ateg.org/