Book Review: The Accidents of Style
This is a difficult review to write. I disagree at length with a number of the author’s conclusions, but in actuality I quite like it and can, in certain contexts, heartily recommend it. My major disagreement with the book is not so much directed at the individual book, but at the genre, which I believe needs major restructuring. So please keep this in mind as you read the following.
The book in question is Charles Elster’s The Accidents of Style: Good Advice on How Not to Write Badly, St. Martin’s Griffin, 2010, paperback, $14.99
In the book, Elster details 350 “accidents” that careless writers get into. (Not going for 365 and the opportunity to put out an “accident-a-day” calendar as well seems to be a missed marketing opportunity.) Elster’s advice is sound; the book is well-written and engaging; and Elster maintains the central metaphor of not-bad writing as defensive driving throughout without it becoming tiresome or old. Below I cite ten of the accidents as problematic, but that’s 10 out of 350, less than 3%. That’s a good track record by anyone’s account. Accidents of Style makes a worthy addition to the shelves of those who like books about style.
My problem with the book is not with the book itself, which I rather like. My issue is with the entire genre of style advice books: they are usually rule based. The typical book giving advice on style lists a bunch of rules that writers should follow. Most of these rules are negative—writers should avoid particular usages or constructions. The most famous examples, and the ones that undoubtedly define the genre for modern stylists, are Strunk and White’s The Elements of Style, and Orwell’s “Politics and the English Language,” with its infamous six rules for good writing.
I want to separate style guides and usage manuals from writing advice, here. A style guide, usually written for a particular publisher, must be rule based. Prime examples are The Chicago Manual of Style and The Associated Press Stylebook. The point of these books is to create a consistent style among multiple writers producing works for a particular publisher—often other publishers will adopt them as their preferred style guide too; style guides are not about how to write well, but rather how to be consistent with a particular house style. Usage manuals are usually organized like dictionaries and comprehensively cover the most common problematic usages. Prime examples are Garner’s Modern American Usage and Merriam-Webster’s Dictionary of English Usage. Like style guides, usage manuals are reference books, intended to sit on a shelf and be consulted when necessary, but unlike style guides they are not intended for a particular publisher and do exist to promote good writing.
The genre of writing advice, on the other hand is usually a mishmash of the style guides and usage manuals. The organization of advice books tends not to lend itself to reference—Elster’s book, for example, contains an index, but the “accidents” are organized haphazardly with no apparent logic to their order of appearance. Writing-advice books usually have a “hook,” some central metaphor that differentiates it from all the others. In Accidents of Style it is the defensive driving metaphor. Advice books tend to be less comprehensive. With 350 “accidents,” Elster’s book is one of the more comprehensive of the genre, but in comparison, Merriam Webster’s Dictionary of English Usage has over 2,300 entries.
Writing-advice books are ostensibly written to help aspiring writers. And while I don’t know who actually buys the books, I suspect that they are mostly purchased by people who like to read books about writing advice rather than by those that are trying to improve their writing. I say this because rule-based writing advice books are actually not all that helpful.
I don’t know of any studies about improving writing in adults through use of writing-advice books, but the evidence is quite clear that formal grammar instruction has no effect on children’s and teenager’s ability to compose coherent and logical essays. In fact, grammar instruction may actually be deleterious in that it “turns off” students to the joys of writing. A 1963 review of research in the instruction of writing composition concluded that, “The teaching of formal grammar has a negligible or, because it usually displaces some instruction and practice in actual composition, even a harmful effect on the improvement of writing.” Furthermore, a comprehensive study of 248 New Zealand students over four years found no difference in the writing abilities of different students who had studied traditional school grammar, generative or transformational grammar, and no grammar at all.1 The point is that following rule-based writing advice may make you better at following the rules, but it won’t make you a better writer.
Writing, be it good or bad, cannot be boiled down to a few, or a few hundred, simple rules. Language is just too complex. And good writing is more about composing series of coherent and logical paragraphs than it is about the grammar or usage of a particular clause or phrase. Such writing-advice books ignore the elephant in the room, and as a result are largely useless. Sure, a rules-based manual can help you avoid problematic usages and the “accidents.” But just as learning defensive driving won’t help you become a NASCAR driver, accident avoidance in writing won’t make you a great writer. You can avoid every single problematic usage in Elster (or in Garner or MWDEU, for that matter) and still produce an incomprehensible, illogical piece of trash.
The best writing advice books are not lists of rules of what to avoid. They actually give instruction in how to compose coherent and enjoyable prose. Two examples of excellent books on how to write well are Joseph Williams’s Style: Toward Clarity and Grace and William Zinsser’s On Writing Well.
The following are ten examples of Elster’s “accidents.” I have selected them because they are all examples of what I consider to be bad advice—although I reiterate that such bad advice is the exception in Elster’s book. But in each case, I try to provide an explanation of how the advice could have been turned from a simple proscription into a lesson on how to write well.
Accident 34 is “avoid the lazy, mechanical use of basically,” and accident 72 is “avoid the phrases general consensus and consensus of opinion. These are good pieces of advice in and of themselves, but in both cases, Elster includes a sidebar that gives problematic advice.
Basically, and the raft of adverbs like it, are usually—but not always—meaningless filler. But in a sidebar, under the title “When You See an Adverb, Kill It,” Elster goes too far. He writes, “Good tight writing has no unnecessary words. The aspiring stylist
quickly learns to mercilessly cut them out, and the first ones to go are adverbs.” This is true if you want to affect a detached and emotionless style, but adverbs are often useful if you want to convey an emotional tone or intensity. And often a bare verb means something quite different than one qualified by an adverb. “Learning” is not the same as “quickly learning,” and “cut out” is not the same as “mercilessly cut out.” You can often cut words and still maintain the basic denotation, but just as often you also change the tone and style of the sentence in doing so. Sure it may denote the same thing, but the quality of how you have said it has changed. And good writing, as opposed to not writing badly, is all about that quality.
When it comes to adverbial overuse, what is really called for is not killing, but judicious pruning. Another of Elster’s examples, with adverbs highlighted so they can, presumably, be cut is, “The report clearly states that the only thoroughly and completely effective method for increasing sales rapidly is to competitively engineer and efficiently market our products.” Some of these adverbs can be pruned. Clearly, thoroughly, and completely are unnecessary and can be struck out without any significant loss. The others are not; they are essential to the meaning of the sentence, although the use of competitively is awkward. The “only effective method” is a very different thing than “an effective method,” as is “increasing rapidly” vs. “increasing” and “efficiently market” vs. “market.” Do you really need to say, “we can increase sales by marketing our products.” Rapidity and efficiency are vital components of this sentence, and adverbs are an excellent way to convey this. In the end, a well constructed version of this sentence would read something like, “The report states that the only effective method for increasing sales rapidly is to engineer our products to be competitive and then to efficiently market them.” (You could also substitute design for engineer, although there is a slight difference in meaning between the two.)
In the sidebar on consensus, Elster goes on a self-described “rant” against redundancy. Now, I don’t disagree with the fact that a lot of bad writing is redundant, but not all redundancy is bad writing. Elster criticizes use of fellow colleague and free gift, indicating that they should always be cut down to simply colleague or gift.
Regarding the first, there is a difference in tone between fellow colleague and simply colleague. A fellow colleague is not simply someone you work with, but someone for whom you have a measure of respect and perhaps even affection. Depending on what the writer wishes to convey, the inclusion of fellow may be appropriate. You may not always want to cut it.
This brings us to free gift. The phrase is a bugaboo among prescriptivists. Yes, it is redundant, but it is also an incredibly powerful and effective phrase. There is a reason why free gift appears so often in advertising—it works. (Perhaps this is because ads are meant to be seen, not read critically. It is the appearance of the word free that is important, not the sense of the phrase.) If good writing creates an emotional response and spurs action on the part of the reader, then free gift is good writing, at least in the context of advertising and marketing.
Good writers write differently for various audiences and genres, and there is no one-size-fits-all. Always ask the questions, what message am I trying to convey and do the words I’m using effectively convey that message to my intended audience? Do not mindlessly follow Strunk and White’s dictum to “avoid needless words,” but consider the audience and the desired tone and register before striking out words.
Accident 55 is “don’t use literally to mean ‘figuratively’ or as part of hyperbole.” All I can say regarding this is to point to the opening line of Joyce’s The Dead, “Lily, the caretaker’s daughter, was literally run off her feet.” And Nabokov in his Invitation to a Beheading writes, “And with his eyes he literally scoured the corners of the cell.” With all due respect to Elster, I prefer to take my writing cues from Joyce and Nabokov. The figurative use of literally is long established. (The OED dates it to the mid-nineteenth century.) Had Elster advised to avoid using the figurative literally because it is overused and hackneyed, that would be another matter, but he didn’t.
There are two lessons here. The first is that the literal sense of a word or phrase is a very poor guide to its actual range of senses and connotations. Frequently a term can mean more than just its literal sense, and it is not uncommon for a term to never carry a literal sense. The second has to do with overuse. In current casual writing and speech, the figurative sense of literally is hackneyed. While one will often want to avoid using hackneyed phrases, this is not always the case. But a careful writer is aware of such hackneyed uses and only uses them deliberately to create a certain effect.
Accident 65 is “don’t confuse hanged and hung.” Elster repeats the canard that objects are hung and people being executed are hanged. This distinction was unknown prior to the twentieth century and has been routinely ignored since, even by writers such as William Faulkner and Flannery O’Connor. It is purely a question of style. If you want to maintain the distinction, fine, but there is nothing wrong with choosing not to. Merriam-Webster’s Dictionary of English Usage gives a better version of Elster’s advice, saying it may be wise to maintain the distinction in order to “spare yourself the annoyance of being corrected for having done something that is not wrong.”
Good writers are aware that certain constructions, while grammatically and stylistically unassailable, can be unnecessarily distracting to the reader. Avoiding them is not simply an exercise in sparing oneself annoyance, but it is the discipline of keeping the reader focused on the message and the prose, not the grammar.
Accident 150 is “don’t confuse convince and persuade.” Evidently, convince is supposed to mean to make someone believe, and persuade is supposed to mean to spur someone to action.
The use of persuade to mean to make someone believe is a very old and very well established usage. Both Shakespeare and the King James Bible are rife with examples. The use of convince meaning to spur to action is, however, much newer. It does not start appearing until the mid-twentieth century. But once it appeared, it did so with a vengeance and is now utterly idiomatic. Elster’s attempt to maintain the distinction is an exercise in futility. Even Bryan Garner rates this as Stage 4 on his scale of language change. Stage 4 is “virtually universal but is opposed on cogent grounds by a few linguistic stalwarts”—which differs from Stage 5, “universally adopted except by a few eccentrics,” only by the fact that Garner considers himself to be a “linguistic stalwart” and not an “eccentric.”
If you want to maintain the distinction between convince and persuade in your own writing, by all means go ahead. Good writers know how to recognize which distinctions are their own pets and which are generally observed, but they do not expect their readers to appreciate the former.
Accident 162 is “it’s could not care less, not could care less.” Logic is on Elster’s side here, but language usage is not governed by logic. The phrase could care less is idiomatic, and idioms don’t have to make sense. Why people object to this usage but don’t object to others like head over heels is beyond me. Elster even acknowledges that his own searches of Google News turns up the fact that could care less is three to four times more common than couldn’t care less.
The true shame is that Elster has missed an excellent opportunity to make an important point, using the illogic of the phrase to segue into a discussion of the importance of logic in good writing. For while the construction of a particular phrase may not be logical, good writing is. I point to the entry on Illogic in Garner’s Modern American Usage. There Garner crafts an excellent lesson in how to construct a logical argument and use grammar and style to undergird and support the writer’s message. Use of the phrase could care less is not an accident; it is not wrong. But it may be a signal that the writer has not taken the necessary care in constructing her argument. But to simply and unthinkingly proscribe the phrase’s use is just an exercise in prescriptive futility.
Accident 210 is the use of alumni to refer to a singular person. Last I checked, we were speaking English, not Latin. The Latin singular and plural endings are not particularly relevant when it comes to English. It is not uncommon, and perfectly acceptable, for words to shift in meaning and take on English inflections when they are adopted into the language. In recent decades, alumni has changed and come to encompass both the singular and plural (like moose, sheep, or deer—maybe something about the university experience turns people into herd creatures).
My belief is that alumni ran straight into a confluence of the dropping of Latin from the curriculum, coeducation, and the demand for non-sexist language. First, few people learn Latin in school any longer and can correctly identify the correct Latin singular and plural endings. Unconstrained by knowledge of Latin, people were free to use the word as they saw fit. And since the plural predominates—there are a lot more opportunities to talk about the alumni in general that an alumnus in particular—the singular began to be dropped. The 1960s saw American universities opening their doors to women in increasing numbers. Since most of the universities were traditionally male institutions, the masculine forms, alumnus/alumni, dominated over alumna/alumnae, and the masculine form was used to encompass both men and women. (I note that Vassar College, traditionally a women’s school that is now coeducational, uses both alumni and alumnae in its official materials, but the Wikipedia page on the college counts (at least until someone changes it) Edna St. Vincent Millay and Meryl Streep among its “alumni.") Finally, like the use of the singular they, the plural alumni began to be used as a non-sexist alternative to alumnus/alumna.
In short, if you are writing a style guide for a university, you might want to maintain the original Latin endings, but for general writing there is no need.
Accident 225 is “close proximity is redundant.” The phrase may be a bit affected and bureaucratic, but it is not necessarily redundant. To be in close proximity is to be very near. Close here can act as an intensifier, modifying the proximity. To say that something is in the proximity of something else is rather vague; it is in the vicinity, but the distance can vary from a few hundred yards to several miles. To say that something is in close proximity means that it is right next door. In many cases you are probably better off substituting very near, but that’s a different complaint.
Accident 327 is “improper use of beg the question.” This entry is another example of Elster trying to hold back the tide. Almost no one uses the phrase to refer to the logical fallacy of petitio principii anymore, and to stand one’s ground on that usage is futile. The old usage has become what Bryan Garner refers to as a skunked term. Better would have been to condemn the new usage on the grounds that it is overused and hackneyed. It has become one of those terms used by people who staple phrases together in an attempt to make an argument. Just avoid the phrase altogether.
Accident 329 is “don’t use enormity to mean enormousness.” Here Elster, citing one of Barbara Wallraff’s rare errors, repeats the mistaken belief that enormity can only be properly used to refer to monstrous evil. We can probably lay the ultimate blame for this one on Henry Bradley, who edited the E volume of the OED (which is widely considered the weakest section of that dictionary). Apparently with no basis whatsoever, Bradley labeled the sense of sheer size as “now regarded as incorrect.” That note was retained in the second edition. (In fact, it’s not clear the entry was updated at all.) Strunk and White, in their absolutely wretched The Elements of Style, repeat Bradley’s fiat and the false idea that enormity could only mean monstrous evil became carved in stone.
But writers have never restricted themselves to these hamfisted dictates, and have used the word with great subtlety and inventiveness throughout the centuries. (My favorite is by Beverly Nichols, who writes in Punch in 1973 of “the enormity of the female behind.”) It is true that enormity does carry a negative connotation, and when not used to describe mere physical size, there is an element of badness in the description. But its use has never been restricted only to the monstrous or to extreme evil. Writers can properly use enormity to denote sheer physical size, but they need to be aware that the word often carries a negative, albeit not necessarily monstrous, connotation, and it may not always be the right choice.
The alternative, enormousness, has never been a favorite of writers, probably because it is inelegant and clumsy (the concatenation of /s/ sounds is jarring). It is a serviceable word, but that’s about the best that can be said of it.
1Charles A. MacArthur, Steve Graham, and Jill Fitzgerald, Handbook of Writing Research (New York: Guilford Press, 2006), 265–66.
Copyright 1997-2013, by David Wilton