Using the information in this chapter takes one very important kind of sensitivity—namely, sensitivity to the difference between illegitimate nonargumentative persuasion and the appropriate use of lively or emotively charged language. Without a nose for this difference, you may condemn a claim when it doesn't deserve condemnation; or you may censor your own writing so scrupulously that you wind up with astringent, unreadable prose.

Some rhetorical techniques have no appropriate uses, or almost none. Innuendo might add humor when it alludes to innocuous items, such as the colorfulness of someone's clothing. But true innuendo busies itself with things salacious and has no proper uses: Either spell out the accusation or leave it unsuggested.

It's just as hard to imagine good reasons for flinging stereotypes about. Even where people should not get offended—for example, the image of the impractical professor or the shrewd lawyer—stereotypes are to thinking what a cliché is to writing: probably unavoidable in some form, but never welcome, and a sign of laziness.

Most of the other categories in this chapter have their good uses. But what are they? Two general principles, while not exhausting the subject, help sort the sheep from the goats.

In the first place, you may evaluate the emotive force in language to see if it constitutes real slanting. Emotive force alone does not make a word or phrase a slanter. Consider "mother"— what connotations that has! But we would not call "female parent" a better choice of words in a sentence like "I promised my mother not to try cigarettes." "Mother" becomes a slanter in other contexts. "Recycling: Do It for Mother Earth." "The state is responsible for your existence as much as your mother is, so obey it as you would your mother." Or even, "Support Medicare reform today and help take care of your mother tomorrow."

Similar things happen with the word "suicide." "More teenagers commit suicide today than ever before" is a claim with emotive punch, the same punch that strikes in "His immigration policies will lead to widespread suicide" and "This mouthwash is social suicide." But the first claim, unlike the next two, contains no slanters.

This is partly a matter of the literal use of words rather than their metaphorical use, but not entirely. It is more illuminating to say that the words with emotive force become slanters when some other word or phrase would be more natural in their place.

Not that identifying the more natural word can always happen so easily. The debate over abortion includes many accusations from both sides of slanters on the other side. Is "fetus" a euphemism for "unborn baby," or the most natural description? Does the phrase "right to life" imply a rhetorical definition—the embryo is a human being with the right not to be killed—that unfairly characterizes the significance of the debate? In such cases the accusation of slanter-use is as controversial as the use of the slanters.

These cases, though they are important, do not make it impossible to identify the natural use of most words. Ask whether a claim would have a stronger rhetorical effect if you replaced its emotively charged word with a more neutral synonym. What's wrong with "More teenagers commit the self-induced cessation of bodily functions today than ever before"? That claim, much more than the equivalent claim with the word "suicide" in it, is trying to do something with its words. Here, and not in the other case, we have a slanter.

You may think of the second general principle as a broader version of the first. Rhetoric is appropriate when it has already been justified. Say someone asserts: "Alcoholism is a disease like any other." You may take this for a persuasive definition. But suppose the person has already argued that some people have a genetic predisposition toward alcoholic behavior or are otherwise powerless to will it away; that their alcoholism follows a progressive and debilitating growth; that it responds to treatment, including medical treatment. Now that the definition is not being made to do the work of persuading—now that it functions as a conclusion, however controversial, to other claims—it cannot be accused of slanting.

Or take this sentence: "Tonight they'll present their so-called original idea." On its own it downplays the idea. But imagine that the sentence before it said, "Our school board spent the last two months preparing a list of books to be removed from the high school library." "So-called" is just right. The downplayer turned into an appropriate expression as soon as we saw a good reason for it.

In short, don't let the persuasive effects of language, by themselves, become the goal of your critical evaluation. Use them as a first sign that a further justification is called for. Then look for it.