For most writers revision is a necessary part of production. I want to emphasize, however, the importance of revision as a way of learning. New writers produce flawed work not because they are slovenly, and sometimes not even because they lack knowledge of writing, but most often because they are unable to perceive the flaws. And while they see nothing wrong with their work, they are unwilling to change a word. Some new writers think they have revised a work when they have checked it for spelling errors and typos.
I can, and will, provide several examples of revisions, but no one can learn what revision is without practicing it on his or her own work. Learning revision is the process of learning to perceive flaws in your work. At first it is difficult to find anything wrong. The writer knew what he meant, put it on paper as best he could, and since it was the best he could do, he sees no way it could be done better. In this early stage, if he is to learn revision at all, the writer must look through his work to see where he has written things that are like the bad examples in the book. Then he must make changes like those suggested in the book in a mechanical way, whether he really understands the reasons for the changes or not, and regardless of his feeling that his work was pretty much okay to begin with.
This process alone can only advance a writer so far. The writer will not see errors unless they are very like the bad examples in the book, and no book can provide enough examples to cover all the variations of error. With practice however, the writer will come to see errors in his work which are not precisely the same as those illustrated by bad examples in the book, but which violate the principles the examples in the book are meant to illustrate. At this point the writer should begin to compose a book of his own bad examples. As the weaknesses of writers differ, the writer must become a specialist in overcoming the weaknesses which are most pronounced in his own writing.
After so many revisions, the writer may learn to stop himself as he is putting a flawed construction on paper, and then perhaps after much more revision, he will not have to stop himself because the flawed construction will no longer occur to him. In truth, very few writers reach the point that they need no revision at all. But most do learn to avoid their more serious blunders altogether.
Many readers of this book are not quite so green as that hypothetical new writer. Most students can revise if only they will do so. Of course not every revision results in a publishable manuscript. But every revision ought to result in learning to revise better. For this reason it is worth practicing revision even when publication of the material is not contemplated. I advised writers to practice writing with trivial pieces, and trivial pieces are ideal for practicing revision. If you have not learned the benefits of revision from practicing on pieces that you are relatively indifferent to, you may not be able to apply the knife to a work about which you have strong feelings.
I suggest that writers who use computers practice their first revisions on double- or triple-spaced printouts. I think it is useful to see the original wording along with the revision and to see whole pages of corrections at once. When you begin to make revisions without paper I think it is instructive to save the uncorrected original files for comparison with the revisions (as well as for backup). Of course, those with powerful word processors and proficiency in using them can use strike-out and insertion features which provide a distinctive appearance for parts that have been struck and have been inserted, but these features may be to cumbersome for many.
As a goal of revision, most beginners should aim to reduce the word count of their work by about 30% without altering or sacrificing sense. That is, you should be able to say everything you said in about 30% fewer words. Sometimes, of course, stronger expressions require more words than weaker ones, and brevity of itself is not the goal. But most weaknesses in beginners' writings are wordy, and a 30% reduction is a realistic and desirable goal for almost all beginners.
Some writers work by expanding outlines or sketches so that successive drafts contain more words. But these are not revisions. The revision is what is done to the story once all of its elements are in place. Something is wrong with a revision that does not reduce the word count.
In the hopes that writers will be motivated to attempt the performance of similar acts, I offer the following example of a word-by-word revision. A paragraph of a work of nonfiction as drafted:
A good editor can save an author considerable embarrassment. One of the editor's jobs is to correct errors in the author's manuscript without doing damage to variations in usage which could reasonably be considered to be literary style and without harming the special voices of narrator and the characters. Since some editors are not so good as others, authors must make every effort to make the manuscript as consistent and as nearly perfect as can be. If an error in print reproduces and error in the manuscript, the author has no one to blame but himself.
(96 words.)
In the first sentence we notice the adjective "good." "Good" belongs to a suspect group of adjectives. It must prove necessary or be eliminated. In this sentence it is not necessary. Any editor can save. Good ones do. We could simply strike out the "good." However, the most important thing we can do with any sentence is to improve its verb. Verbs with auxiliaries are, in general, weaker that verbs without auxiliaries. In this example, "saves" would be stronger than "can save." If we leave the "good" in, we can substitute "saves" for "can save," and this what we choose to do. The "considerable" is also suspect, but we move on.
The second sentence is horrid. Clearly "correct errors" is redundant. A good editor would not correct anything except errors. We should strike "errors in." In "doing damage to" we recognize "do" which is another verb to watch out for like "make" and "take." "Do" is coupled with the abstract noun "damage." Clearly this sentence would be improved if we could incorporate the sense of this abstract noun into the verb. This is easily done because "damage" is itself also a verb. We substitute "damaging" for "doing damage to." "Could reasonably be considered to be" is a weaselly construction for which we substitute "are considered." "Narrator" requires an article, and we always try to use the definite article "the" when we can. Insert "the." "Special" is another suspect adjective. The voices in a work of fiction are necessarily special. We delete "special." As we have decided that we are writing about a good editor we do not suspect him of doing harm, so "harming" is changed to "altering." The verb of this sentences is "is." We cannot make it "crashes" or "explodes," but we can make it "includes" which is a little stronger than "is." But to do so we must drop the "One of." This is just as well. "One of" is a phrase that can and should be revised out every time it occurs. That makes "jobs" become "job." We cannot do without the adjective "literary" of "literary style," because we must distinguish from "editorial style."
In the third sentence, we see immediately that "make every effort to" has to go. We have written "must" which in itself already means "make every effort to" and then some. The sentence is in two clauses of the form "since this, as a consequence that." Delete "since" and change the comma to a semicolon. This works better if not done too often. Now the problem is that "authors" are working on "the manuscript." "An author" is the solution. Then it should be "he can" at the end which is stronger than "can be."
We relax a bit for the final sentence. It is the topic sentence and it is in one of the two places (first and last) a topic sentence can occur. "Has" is too weak a verb to remain. "May blame" is not very strong because it includes the auxiliary verb "may," but it is certainly stronger than "has." So we write "may blame no one," rather than "has no one to blame."
The paragraph as revised:
A good editor saves an author considerable embarrassment. The editor's job includes correcting the author's manuscript without damaging variations in usage which are considered literary style and without altering the voices of the narrator or the characters. Some editors are not so good as others; an author must make the manuscript as consistent and as nearly perfect as he can. If an error in print reproduces and error in the manuscript, the author may blame no one but himself. (79 words.)
Ninety-six words have been reduced to seventy-nine. That is a reduction of only about 18%, but the author in this case is not a beginner.
Real writers really do revise. A practice writer thinks hard to remember the reasons for making changes such as have been demonstrated. He revises more quickly than he can explained. Seasoned writers do not commit such atrocities to paper in the first place. They have internalized many of the principles of strong writing. New writers gain this ability by practicing revision.
Some revisions: