Today (in 2001), getting samples of students' writing which can be publicly  shared is difficult. For that matter, it was difficult even back in 1986. I therefore thank the school system which provided me with these samples. I do not wish to identify that public school system, except to say that it was predominately in a middle class area. The essays therefore probably reflect what middle school teachers then (and probably now) see in the writing of their students. I have said before, and I say again, that some national organization (such as NCTE or the Scholastic Aptitude Service) should undertake the collection of similar samples and simply make the samples permanently available on the web. A brief notice on the ATEG list has already resulted in several requests from teachers of teachers for permission to link to and use these essays with their future teachers. Permission to do so, by the way, is not needed -- the theme of this site is that in deciding what grammar to teach, and how to teach it, we need to look at the linguistic development of the students whom we are teaching. That is what these essays are here for.
     Back in 1986, when I was desperately searching for examples of students' writing,  I simply asked for samples, and, if possible, scores on national standardized tests to match with the samples. I couldn't be picky about the modes of writing, about the prompts the students were given, or about  more information on what the students had or had not studied. Such information would certainly be helpful. We know, for example, that syntactic performance differs across modes -- narrative,. expository, etc. We also know that some types of instruction, for example, sentence combining, has at least a short-term effect of the syntactic performance of the students. I had, however, to take what I could get. Ideally, discussions of these essays (both on and off this site) will prompt better and bigger collections of samples.
     Roughly speaking, the results of this study come close to the results reported by Hunt, O'Donnell, and Loban for seventh and eighth graders. There are major problems in comparing such studies because different researchers have defined procedures and constructions in different ways, and because the original samples of previous researchers are not available for reexamination (as these are).  In the 31 papers presented here, the students averaged 9.4 words per main clause. Hunt did not study the writing of seventh graders, but 9.4 is almost exactly half-way between the 8.9 reported by Loban and the 10.0 reported by O'Donnell.  The figures for subordinate clauses per main clause are somewhat different. The students represented her averaged 42 subordinate clauses for every 100 main clauses. That compares to the 28 reported by Loban, and the 30 reported by O'Donnell. It is interesting to note, however, that Hunt's eighth graders also averaged 42. These differences may be the result of differences in definitions of subordinate clauses, and readers interested in these questions are welcome to explore the various discussions of these issues on this site. Beyond words per main clause and subordinate clauses per main clause, the differences among researchers become even greater, so it is very difficult to compare the results here with the results of others. I simply wish to suggest that the essays presented here appear to be representative of seventh graders' writing. 
     This study is, of course, tentative. For one thing, I need to analyze examples of writing from later grades. I would, however, suggest that this study already implies that syntactic development takes place naturally -- without any specific instruction from teachers. The seventh graders average 9.4 words per main clause, compared to 7.7 for the fourth graders. The average of their longest main clauses jumps from 19 to 21. They average 42 subordinate clauses per main clause, compared to 22 for the fourth graders. The average almost twice the rate of infinitives (14.0 to 7.2.) [See.]  They begin almost twice as many of their sentences with prepositional phrases (5.7% to 2.7%). Likewise, there is an increase (from 2.8% to 4.3%) in sentences that begin with subordinate clauses. Their rate for embedding subordinate clauses within subordinate clauses is three times that of fourth graders. Other measurable changes in their writing also suggest their natural syntactic development. (Those interested should see "An Overview of KISS Research Projects.")
     As KISS theory suggests, however, there is no development in their use of  gerundives or appositives. The rate of gerundives per 100 main clauses is identical to that of fourth graders -- one.  Their use of appositives actually decreases, from 4.4 per main clause to 3.9.  This decrease reflects their getting beyond the fourth graders' tendency to list -- There are seven people in our house, my mother, my father, my two brothers, my two sisters, and me. It will not surprise me to see this rate in appositives continue to decrease, perhaps until ninth or tenth grades. KISS theory suggests that the true appositive (as opposed to the list or formula) develops through (not by bypassing) the subordinate clause. Even many of the appositives in the seventh graders' writing are probably lists. As these writers develop, the lists will probably break apart as the writers learn to supply more details about each member -- My two brothers are older and like to play sports.  If the preceding suggestions are correct, they imply that teachers who attempt to force gerundives and appositives on younger students are probably doing more harm than good. It's not nice to fool Mother Nature.
     If this study suggests that teachers who attempt to force advanced constructions on younger students are misguided, it also suggests what, perhaps, they should be doing.  One of the most common complaints of middle school teachers is that their students write  fragments, comma-splices, and run-ons. The papers presented here may verify that -- they average three fragments, five comma-splices, and nine run-ons for every 100 main clauses. These numbers, and the teachers' concerns, led me to explore the problem in a separate essay. As you will see if you read it, my conclusions are that the students' "errors" usually reflect growth and intelligence, and that our problem (not theirs) is that we are not teaching students in the way that we ought to be. In essence, we should be teaching seventh graders how to analyze their own writing and how to punctuate main clauses that express amplification or contrast. 
     The preceding suggestions are all based on the thirty-one essays presented here and on my own biases and perspectives. I may be wrong. More important than any of the preceding, therefore, is the idea that what we teach should be based on careful, statistical analysis of actual students' writing. Anecdotal evidence is not sufficient. (Teachers who say that they have seen appositives and gerundives in the writing of younger students often have no idea of what lists or formulas are.) Statistical analysis forces us to look closely and carefully at the actual texts. It helps us both to determine the nature of natural syntactic development and to see not only what stages our students are at, but also what we might do to help them.
     I cannot end this essay without noting the pleasure that working with these thirty-one essays has given me. Some of the essays are lackluster, but who does not at times create lackluster writing when asked to produce on demand.  Many of these essays, however, are vivid, imaginative, a couple even thought provoking. I have had to change the names of people and places, although some of these writers would recognize their own work, even after fifteen years.  I considered eliminating those essays the authors of which might possibly be identified, either by the writers themselves or by people who know them.  I decided not to do so however, because often these essays were the most interesting, both in terms of syntax in content. None of these student writers has anything to be ashamed of; I'm just wondering if we, as teachers, can say the same thing?