Writings Are Made of Sentences

by Peter Connelly

When I find myself making an inordinate number of marginal comments on the first page or two of a short paper, I stop reading and go back to evaluate the sentences I have read so far. If they aren't right, I am not really reading a paper anyway, I am trying to read the writer's mind. There is little to be gained by fighting my way to the end. I used to scribble on, hoping that the paper would become intelligible, but I now know how unlikely it is that the writer has either suddenly learned how to or suddenly decided to write proper sentences somewhere in the middle of the paper. If the first page is really awful, the rest will not be good.

In order to explain the problem to the student, I have found it helpful to have a little procedure for grading sentences. I give a numerical grade, two, one, or zero, to each one on the basis of its adequacy as a sentence, not as truth or insight. In other words, I am not evaluating the sentence as content, but as a grammatical and rhetorical entity. This way it is clear that the issue here is not "ideas," but sentences.

The good ones are two's. I'm not overly picky: superficially flawed sentences can get a two.

Those with genuine mistakes get a one, and I have to be able to identify the mistake, There may be errors in punctuation serious enough to create ambiguity, or syntax so complicated that I can't parse it. These sentences may be wordy; they may have mechanical errors, spelling mistakes, incorrect diction, faulty parallels, references to nonexistent antecedents (including "seconds" without "firsts" and any other similar incoherencies.) They may have errors in case, repetitions of noun phrases where pronouns should be, or any other error of similar magnitude. Sentences get a zero when they have multiple errors, or errors so devastating that I can't sort them out. I add up the sentence scores and divide by the number of sentences. (This seems elaborate, but remember it is done for only a very few papers.) If the quotient is less than 1.5, I have a paper in which more than half the sentences are not proper sentences. That by my standards is a paper I should not have to read. I would say a paper should never have less than a 1.7 sentence score: three bad sentences out of every ten. If that seems too high a standard, check it against your practice. Score the sentences on the next "A" paper you have. Different instructors have different tolerances for bad sentences. For me, however, this is a firm rule: if the quotient is less than 1.0, the paper fails; such a score means either that all the sentences have something wrong with them or that more than half of them are unreadable-- or something in between.

 

So I give the stinkers back partially unread--but with a few sentences precisely marked. That's better for the student, for whom unfavorable commentary from beginning to end is more depressing than instructive. And anyway, I am usually willing to allow the writer to have another go. Obviously he or she had misunderstood the assignment: it was to be a writing of some kind, and writings are made of (proper) sentences.

 

 

 

 

 

Back to Writing Forum index

 


last modified by Judy Hunter on 6/16/00