From E.D. Hirsch, The Philosophy of Composition (Chicago: University of Chicago Press, 1977)
I list the following maxims of commentary with some diffidence. Only the importance of the subject impels me to take a stand on matters about which we must all feel rather uncertain in the absence of good research.If the comments are to effect an improvement, they must be taken to heart and kept in mind by the student. And if the student is to keep the comments in mind when he writes again, they must be few in number, because his capacities are already heavily taxed by the act of writing. He cannot think of ten new points all at once. The worst vice of the schoolmarm is to correct everything. Of course, in some cases, it might be effective to correct more faults than one comments upon.
The chief talent of the teacher resides in the diagnostic ability that enables him to make this kind of selection. In the passage on Burgess, just quoted, the main fault demanding commentary was the deficiency of the stated theme in integrating the sentences of the paragraph. Instead of the topic, "not very organized," the student needed a theme like "ambiguous" and needed to repeat this work as an explicit leitmotif. A second criticism, admonishing the student to vary sentence form and length, would have sufficed as commentary.
Since the point of commentary is future, not past, performance, it should be put in a form that makes it available for future use by the student. The most useful form is probably a short paragraph referring to particular examples of a general point. Students and teachers will do well to conquer the assumption that the red ink on a paper is the defense of a grade. Most of the red ink ought to be in the summary.
This will save time and ink and will make the remarks more pointed and effective.
This spells out the principle for following maxim 2, namely, select only the most important matters for comment.
While a scolding tone may be just what is needed on occasion, the best results are likely to be produced by encouragement. Since this seems to be true in all the crafts, it is probably true in composition. It is also true that occasional expressions of dissatisfaction are also effective since they mean that one expects serious efforts.
If the teacher's analysis is accurate, repetition of a comment is going to be an effective teaching device. A lot of rehearsal is needed to change a student's productive schemata.
The teacher can accomplish this by keeping an abbreviated record in his gradebook. This record is useful also for student conferences. Obviously, the student should keep his own papers and should be instructed to review the comments on them.
Students of the subject usually believe that they are catering to their teacher's whims. E.B. White thought and apparently still thinks that his teacher, Strunk, was expressing his own definite and forthright personality when he ordered his students to put statements in a positive form. students, on their side, believe that tampering with their styles is tampering with their personalities. This makes the whole subject seem a serious invasion of privacy. that is a tangled half-truth which the teacher needs to untangle for his students and himself.
After comments, the most important device for teaching effective writing is probably directed revision. Of course, the principles for directing a revision will reside ultimately within the student himself. To put them there is the goal of teaching --the goal of the teacher's comments. any writer's first draft is already the product of a revision process in which some forms have been rejected in favor of others on the basis of editorial principles which the writer has internalized. But even a practiced writer will need to make further revisions in his first draft if he has directed a lot of his attention to the substance rather than the form of his discourse. But whether the editing/revising process occurs chiefly in the first draft or in a later one, revision principles are still central to the writer's craft. To learn writing is to learn principles of revision.