Craig, Thank you for posting your response. It is important in a discussion about what theory of language is valuable for teaching that we actually examine what students do and the responses we would make. Here is the sentence, written by a real student, that I claim is problematic for a usage based theory of grammar. (1) By taking time out of your day to get something for someone else just really shows that you really care about them. As I consider what Craig wrote, remember that Craig subscribes to the following view of language from Langacker, among others. “The thrust of the content requirement is that the linguistic knowledge we ascribe to speakers should be limited to elements of form and meaning found in actually occurring expressions, or which derive from such elements via the basic psychological phenomena listed in 1.31: association, automatization, schematization, and categorization. By keeping our feet on the ground, this restriction assures both naturalness and theoretical austerity.” Craig writes: Like you, I'm sure, I wouldn't respond to the sentence without looking at its context in the paper and without some sense of the student. Was it just a slip on their part? The two really's also don't work for me. But if it is worth paying attention to, it is worth playing with. The student owns the sentence. If our (teacher) knowledge of language is "limited to form and meaning found in actually occurring expressions," on what basis is Craig able to ask "Is (1) just a slip"? More importantly, on what basis can we conclude it is a "slip" if we our knowledge is limited to actually occurring expressions? This is a naturally occurring expression. **** Craig provides the following explanation for why a writer might produce (1). That aside, it seems to me to half way follow a common construct: By X (participle head), X (noun phrase) Y's (finite verb). By sleeping in class, you missed half the lecture. By getting angry, Charlie lost all chance for the job. Our expectations depend on the familiarity of the construct, not on some innate grammar that predated our interactions with the world. Since it's not an entirely fixed construction (it has variable slots), construction grammar would call it schematic. Of course, how does he know this is the construct if his knowledge is limited to "actually occurring expressions"? I would suggest his examples are not really close to (1). Is the understood subject of "taking time out of the day to get something for someone. . . ." the same as the understood subject of "shows that . . . . . ."? I don't think so. As a consequence, Craig's explanation doesn't seem very explanatory. The following passage by Craig seems closer to why (1) is problematic. If the student seems to be comfortable with prescriptive grammar, I might point out that a "by" phrase isn't supposed to act as subject. But that is a different frame of reference. The problem is that gerund "taking time. . . ." is both the object of the preposition "by" and the subject of "shows." This double case-marking of the gerund -- object of the preposition and subject of a tensed verb -- is what makes (1) problematic. The following really doesn't explain where (1) comes from if our knowledge of language is based on "actually occurring expressions" If the grammar is innate, shouldn't the student know it already? If it's not, then it helps to have someone mentor the student along. Either they are already comfortable with the construction (and just lapsed in attention), or we can take our time to model it out. The analysis I offer to (1) is that the writer already has principles (that are innate) about how she is to order information in a sentence to meet her textual needs. In this case, it looks like she want to make a comment about the topic "taking time . . . " and that comment is that it "shows you care about them. Heavy subject noun phrases (in other words a lot of words in the subject position of a sentence) are one of the markers of maturity in writing. One of the ways for the writer to reduce the cognitive overload of (1) is to make the noun phrase an object of a preposition. Jim Kenkel just sent me the following examples of mixed constructions from various handbooks. 2) By wearing bell-bottom pants, love beads, long hair, and tye-died T-shirts, many young people expressed their opposition to mainstream values. (Longman Handbook, 2000, p. 380) 3) Because of the rebellious atmosphere generated by protests against the Vietnam war helps explain the often outrageous fashions of the time. (Longman Handbook, 2000, p. 379) 4) By designing the questionnaire carefully made Valerie's psychology study a success.(Longman Handbook, 2000, p. 381) 5) In the world created by movies and television makes fiction seem like reality. (New Century Handbook, 2002, 658) 6) For most drivers who have a blood alcohol content of >05 percent double their risk of causing an accident. (A Pocket Style Manuel, Diane Hacker, 1993, p. 9) Notice how all of them begin with a preposition like (1). Jim notes: "We know from Perera that complex subject NPs are rare in adult speech and rarely appear in writing before age 11-12. We want to claim that these innovative constructions make more salient the information units and allow for easier, local grammatical processing." In other words, Jim and I see mixed constructions as being principled from the writer's perspective. More importantly, they are NOT the result of "form and meaning found in actually occurring expressions." Of course, there may be other issues in the writing of students that a theory of language based on actual usage and without appeal to innate principles can help us with. And, I hope we will read some specific examples. However, quoting linguists on their theories without applying their theories to actual student writing is not very persuasive for me. Bob Yates, University of Central Missouri 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/