This discussion is getting a lot better than it has been, but it still
has a long way to go. One of the major problems is that even the few
people who are contributing have MAJOR unstated assumptions. Those
assumptions, by the way, go much deeper than labels. I don't care if
Connie wants to call herself a "Constructivist," but I am not a
"behaviorist."
     I agree that "when," "because," and even "if" can be evoked in the
writing of primary school children, but that does not mean that they
have mastered those words as true subordinate conjunctions. O'Donnell's
suggestion of "formulas" is very relevant here. Vygotsky also notes that
forms (in the case, these words) often develop before the child masters
the underlying cognitive capabilities. As usual, what we need are
samples of these students writing, samples available on the web so that
various research can be effectively examined.
      On the subject on pre-school language development. I understand
what Johanna is saying, but my problem with it is that all the research
that I have read on the topic goes on to show that students NATURALLY
develop the relevant constructions -- they do not need to be taught.
What every child teaches him/herself is amazing, but, I still suggest,
largely irrelevant. If we want, as Johanna suggests, benchmarks for
where instruction should start, then we should have transcripts (again)
of the writing (or speech) of children in K-3. (Like Johanna, I believe
that serious instruction in grammar should begin in third grade.)
     By the way, the reference to Loban's studied reminded me that I
wanted to suggest to Johanna (and to anyone else involved in such
research) that she (they) might want to try to get the student's scores
on standardized read/language tests so that each sample could be
attached to such a score. Loban divided the students into three groups,
based on such scores and on teachers' recommendations, and he
effectively demonstrated differences between the groups. Any study that
does not address this question, particularly if it is a study that does
not present all the data, suffers from the effect of averaging the
group. Once again might I suggest that what we need is a longitudinal
database of samples of students' writing, i.e., year-by-year samples
from the same students across several years, ideally from grade 3 to
grade 12. Collecting this data would be a major task, but it would free
individual researchers from the difficult task of collecting and
transcribing data. It would also eliminate the problem of the researcher
affecting the data collection. (If I remember correctly, it was O'Hare
who had the students' writing transcribed such that all errors were
eliminated. [How's that for research!])
     And again, I'll state what I think is a major difference between
the KISS approach and most ( if not all)  of those being advocated by
others -- the focus of the KISS approach is to enable students to be
able to intelligently discuss THEIR OWN writing (and reading).
Non-standard speakers (and writers) still use prepositional phrases,
clauses, etc. If the instructional focus is put on using the grammatical
concepts to analyze real writing (their own as well as others), the
objection about learning SAE should not arise.
     I appreciate, by the way, the discussion of textbook publishers.
Having recently signed a contract with one, I've noticed the same thing
myself. (The term of my contract is not over, but I've noted that,
although I thought that we would at least be involved in some meaningful
discussion of, for example, what a sentence is, such discussion has not
happened.) Might I suggest that by using the web, we might eliminate
their textbooks? Several of us are already putting lots of free
instructional material on the web. Here again I would suggest that we
should band together in groups, based on similar underlying assumptions.
For one person to put an entire curriculum (with lots of instructional
material for each grade level) on the web would be impossible. But once
members of ATEG who are not committed to a specific approach begin to
see the outlines of several different approaches, they could choose one
they like and contribute to its development. (I am already beginning to
get some of that kind of help with the KISS approach.)  In a previous
post, I noted my objection to grammars based on five, seven, ten, etc.,
basic sentence types. One of those objections is that, unless the
grammar is developed into a curriculum, students get bounced from one
set of patterns to another. (One year, they use a book based on seven
patterns; the next year, one on ten?) A major problem with instruction
in grammar is that it is not systematic. No one can teach "grammar"
effectively in a single class or single year. I still think that ATEG
should be an umbrella for a number of different groups, each working on
different approaches to grammar in the curriculum. These groups, with
brief descriptions, and contact people for them, could all be listed on
the Grammar Curriculum Challenge page:
http://www2.pct.edu/courses/evavra/ATEG/GrCurr.htm

     Several people, both on and off list, have mentioned seeing me at
the conference. Although I submitted a proposal for a presentation
(just) before the original deadline, and I sent a follow-up query, I
received no response. The deadline for hotel rooms having expired, I
have decided not to attend, but rather to spend that time in developing
my web sites. I do wish the conference success.
Ed V.