What I didn't do in that post was give a concrete account of my idea of how one should respond to writing. I'll attempt to outline it briefly below. These statements apply to writing by peers and by students alike.
- Writing is an act of creation. Properly conducted, so is responding to writing. A good response is the product of a serious intellectual engagement with a document, usually a document in an unpolished state, which takes a bit more effort to read than a fully articulated, proofread, edited document. It's your job to figure out what the document is accomplishing, what it aims to accomplish, and what it could accomplish. You're partaking of the creative act and projecting for the writer what you think the piece will be once fully realized. Consequently, a response should be positive in both senses. I don't mean this in a warm and fuzzy way. I mean that a response should offer suggestions, point out successful moves (that should be extended or repeated, perhaps), and ask questions, because these are the things that will help a writer proceed. It is easier to work from models than from interdictions. You can say what not to do, but the task of the writer is not to not do but to do -- something. The writer needs to know what you think that something ought to be, even if she or he will ultimately reject your suggestion.
- A response is not the same thing as an evaluation. When a piece of writing is truly good, it's important to say that, because handing over your work to somebody else is an act of trust. But the chief aim of a response is not to evaluate but to analyze. In particular: telling somebody that a piece of writing is bad gives the writer nothing to go on as far as revision is concerned. What should the writer do, upon being told the piece is bad? Chuck the piece and start over? Start over how? A response should reveal dimensions of the piece that the writer did not previously perceive. Honesty is necessary, of course, and on occasion a reader will be called to save a colleague from the grips of a truly terrible idea for a project. The Queer Temporality of Lolcats: A Bergsonian Analysis in Limericks was never meant to be. But these occasions are rare.
- A response should summarize what you see the writing accomplishing. Writing does not entail having an idea fully formed and then typing it out. Sophisticated writing is a recursive process of articulating, revising, and nuancing ideas that are at first only incipient. A good response summarizes what the reader has understood to be the aims and accomplishments of the piece of writing, usually with questions when unclarity arises. The reader helps the writer to identify the ideas that have developed in the writing, and the further implications that the writer may wish to draw out.
- A response should be appropriate to one's relationship with the writer and to the task at hand. A response must always be respectful, period. Apart from that, audiences vary. You can be colloquial with a friend from grad school. It's helpful to repeat composition terminology (like that old classic, "topic sentence") with a student. It's important to be gentle (not dishonest: gentle) with someone whom you know is having a mid-grad-school crisis. Likewise, comments that will prompt substantial rewriting are appropriate for an early draft of a dissertation chapter, and not appropriate for a draft of a proposal due tomorrow. Think of the person, the time frame, the situation.
- A few targeted comments are more helpful than a comprehensive account of every possible thing upon which the writer could improve. Time is finite, and while we all cherish hopes of writing the perfect essay, it's not reasonable to suppose that anyone ever will. Far more harmful to anyone's writing life than an essay circulating with a few flaws is the awkward, insular prose of the sitter-and-polisher. Each of us has an inner sitter-and-polisher, and that's not a bad thing (it's also known as our inner critic), but there must be limits. A tragic affliction often seen among advanced graduate students is the inability to show anyone work in progress and an utter unwillingness to submit essays before they are "ready." That way lies incompletes, a fifteen-year time-to-degree, and abject misery. Meanwhile, the writing becomes more and more stilted and anxious, because it never sees the light of day or benefits from the refreshing reality-checks of scholarly communication. Inundating a writer with a flood of comments only encourages the inner sitter-and-polisher who never circulates a draft with which she is not happy--and she is never happy. If there really are a lot of problems, then at minimum impose a hierarchy on your suggestions. When you give a response, you're suggesting revisions, and those revisions should be possible and finite. The goal, after all, is to end up with a good piece of writing, and part of being good is being done.