Two Bad Papers on Language Usage

By coincidence, our two leading upper-middlebrow journals, Harper’s Magazine and The Atlantic Monthly, have almost simultaneously run long pieces about language usage and books that offer to guide us on it. In April 2001 Harper’s published “Tense Present: Democracy, English, and the Wars Over Usage” by David Foster Wallace; in May The Atlantic ran “Word Imperfect” by Simon Winchester. It’s reassuring to see that questions of usage are still live ones, and that editors can be induced to publish articles on them; it’s dismaying to see how poor even educated thinking on such matters is, and how little editors demand in the way of quality in what they print on the subject.

They are both, in very different ways, bad papers. Wallace’s makes some sound points in support of traditional good English, but is itself so badly written as to make the thought cross one’s mind, at least for a moment, that it was written by a clever enemy of good English in order to discredit the notion. Winchester’s is a little—not much—better written, but makes no sound points at all; its thesis is so nonsensical as to suggest that it was written by his bitterest enemy, out to ruin him professionally. Each piece is, at least nominally, an extended review of a particular language reference book: Wallace’s of Bryan A. Garner’s A Dictionary of Modern English Usage (Oxford, 1998), Winchester’s of Roget’s Thesaurus. Wallace has effusive praise for Garner’s book, Winchester little but scorn for Roget’s; both are wildly overheated in their reactions. (If this critique strikes you as possibly deserving the same charge, I ask only that you read the two pieces it deals with before passing judgement.)

Wallace on Garner

Wallace indirectly calls Garner’s book a work of genius; it is not. (I would guess that Garner himself is hardly gratified by such fulsome praise, especially when it becomes clear that the gush is due not so much to Wallace’s admiration for the book as it is to the use he wants to make of it in pursuing his own purposes. Wallace has very little to say in his 20-page piece about the contents of Garner’s book; he is concerned overwhelmingly with its tactics, as he understands them—more on this later.) If not a work of genius, the book is good ; Garner is a great admirer of Fowler (is, indeed, founder of The Fowler Society), and it shows. His book, like many published since Fowler, deliberately echoes Fowler’s title, and it is one of the two or three that are not ridiculously unworthy to pay that tribute and invite that comparison. He is a sensible prescriptivist; knows generally when to yield and when to stick to his guns; explains his principles clearly, and follows them; and is very thorough in coverage—his book is 723 large-format pages long.

Too much of that length, though, is wasted on illustrating points of usage with examples culled from his very wide reading in periodicals of all types, which he offers in excessive numbers (around 5,500, Garner tells us1), at excessive length, and with complete citations to author, journal, date, and page. He defends this practice in his introductory remarks, but in my view his arguments are weak; there are far too many examples, quoted at unnecessary length, and supplied with a degree of documentation that would be in order if they were hitherto unknown source material of profound historical importance—or if they were exhibits a lawyer wished to enter into a trial record.

Cutting these examples by at least half, and streamlining the scholarly apparatus—his readers would be, I think, quite ready to trust that he has not invented or misrepresented his examples—would have resulted in a handier, shorter, and less expensive (at least to produce) book; I estimate that it could be 100 pages shorter with no loss of value, and positive gain in usability. And Garner’s own prose is decent and clear, but not at the Olympian height of the crystalline, bone-dry, sharp-edged prose of Fowler, nor capable of Fowler’s occasional miniature masterpieces, like the celebrated entry on “Superiority” (as printed in the original edition of 1926; Gowers for some reason decapitated it when he revised Modern English Usage in 1965). In short, Garner’s is a workmanlike, solid job that I own and use, and recommend to others, but to call it a work of genius is to be either a schoolgirl with a crush, or a man who talks up a parcel of property because it adjoins his own.

Then, Wallace’s piece is badly written in a way that only an academic intellectual could manage: it is the most self-indulgent, narcissistic piece I can recall seeing in any mainstream publication. His manner is nauseatingly cute and condescending; he uses words imprecisely, offers bad analogies (whose point he then explains at length—the sure sign of a bad analogy), says everything several times, and intrudes much purely personal and otherwise irrelevant matter. A particularly objectionable feature of his piece is his alternation between parading erudition (often irrelevant) and cosying up to us as just another Regular Guy. One minute he's overawing us with an epigraph in Latin (untranslated) from St Augustine, the next he's speaking of a statement’s "biting Gove’s whole argument in the ass" he's Mr All-Things-to-All-Men. This wanting to have it both ways results in a tone that is by turns ingratiating, patronizing, and facetious, and repellent in all its moods.

His faults are best seen in the fifty-two “footnotes” that the text is encumbered with; I use distancing quotation marks because they are not, except in form, really footnotes, but just little asides in which Wallace can tip us a wink, or preen, or take the edge off something in the text that he’s afraid might put us off. One of them reads “(I’m not a total idiot.)”; another, “Please just don’t even say it”; a third, “Did you think I was kidding?” But not all the footnotes are like these; sometimes a Wallace footnote will actually attempt to instruct us, as does number 21, in which he explains the New Criticism as follows: “New Criticism refers to T. S. Eliot and I. A. Richards and F. R. Leavis and Cleanth Brooks and Wimsatt & Beardsley and the whole ‘close reading’ school that dominated literary criticism from WWI into well into the seventies.” This note expects that the reader who does not know what the New Criticism was will nevertheless be familiar with the work of at least some of the men who practiced it, and will be able to infer from this knowledge of their practice what the label Wallace has assigned to them means. (The note also conveys misinformation: F. R. Leavis, for example, was no New Critic, and Eliot and Richards are at least highly questionable.)

Note 33, though, is probably his masterpiece; in it Wallace performs the hat trick, combining in one sentence pedantry, irrelevance, and error. Having told us in the text that early English prescriptivists were so enthralled with Latin 2 that they sometimes wrote their English-usage guides in that language, he appends this note:

(Q.v. for example Sir Thomas Smith’s cortex-withering De Recta et Emendata Linguae Anglicae Scriptione Dialogus of 1568.)

Pedantic in referring to a book his readers will not have heard of nor have access to, irrelevant because a reader who doubts the statement it is there to support would hardly find his doubts thereby cleared up, and finally, in one of those rare events that actually deserve to be called ironic, a clear give-away of the thinness of his Latinity, Augustinian epigraph notwithstanding—“q. v.” stands for quod vide , “which see”; what Wallace meant to say was vide , “see.” And why use a Latin abbreviation at all, even if it were correct, rather than plain “see”?—the English word is actually shorter than the Latin, as well as being impossible to misunderstand. That this is itself a pedantic detail hardly excuses Wallace, it aggravates his offense—he chose to introduce the pointless Latin, then failed even to get it right.

There is no way to convey fully how distasteful Wallace’s style is without extensive quotation, which is impractical here; I can only direct the reader who wants substantiation of my claim to his full text. Nor is it easy to deal with his substantive argument, buried as it is in verbiage, and disorderly in presentation. The best I can do is to offer a précis of his piece, and comment on that (the opening text of each of the following bulleted items is my paraphrase of Wallace; my comments are so labeled, in bold):

•  Garner’s book is very good, but you need to know a lot about the Usage Wars before you can appreciate how good it is; I will now tell you as much about those Wars as you need to know.

•  The book is written for the small minority of those who care about the finer points of English; I admit half proudly, half ruefully that I am one of this elite; I poke a little fun at myself to show I’m a Regular Guy, too, but I remain one of the elite nevertheless.

•  The issue underlying the Usage Wars is political; therefore anyone wanting to be taken seriously as a player must show that he’s on the right side politically, and observes the Democratic Spirit—that is, eschews claims to authority, professes respect for everyone else’s views, and is self-critical. Garner handles this test beautifully.

•  Here’s a thumbnail sketch of the Usage Wars, starting with the publication in 1961 of Webster’s Third New International Dictionary (W3), which brought to the surface the question, who has the right to tell us how to use our language? —a question that all would-be guides and authorities, like Garner, have to answer. Garner does this by being modest, low-key, and diligent in research.

•  Other would-be guides and authorities have blown it by being sarcastic and superior—a sure-fire way of alienating most American readers. Comment : Wallace refers to the “plutocratic tone and styptic wit” of some of these would-be authorities, making one wonder if he knows the meaning of either adjective; what is the tone of the plutocrat—snide? haughty? languorous? wistful? —and as for the wit, it would be just the opposite of styptic—it is meant to draw blood, not stanch it. I suppose Wallace was thinking that when you apply a styptic pencil to a shaving cut, it smarts a little—hence “styptic wit” is sharp, burning wit.

•  Garner says that descriptivists and prescriptivists are really addressing different problems, and hence should not be in conflict, but here he’s being “disingenuous in the extreme”; descriptivist attitudes and ideas have dominated American education for decades, and pervade all the “progressive” movements of our day, while prescriptivists are typically conservative. Comment : How this makes Garner’s statement wrong, much less dishonest, I cannot see. Garner is thinking of the explicitly announced programs of the two parties—descriptivists want to study language, and prescriptivists to guide it—and pointing out, correctly, that they are very different sorts of interest in language. It is also true that the people who staff the descriptive and prescriptive armies, respectively, are likely to differ sharply on many other issues, but those disagreements spring from differences of temperament and outlook that are extra-linguistic; the purely linguistic programs of the two parties could in principle proceed simultaneously, moving on parallel if reciprocal courses, without even meeting, let alone coming into conflict.

•  The claim by some descriptivists, like Gove, the editor of W3, that they are simply dealing scientifically with language, is based on an outmoded, discredited idea of science, one so naïve as to assume that observation has no effect on the thing observed. And language is essentially normative, hence prescriptive; it is meant for communication, and if one way of using language communicates better than another, that way is better. And we have to do some things in certain ways just because it’s conventional, and we would look weird or silly doing them otherwise—e.g., boys in our culture wear pants, not skirts. Comment : this is one of the points on which I can largely agree with Wallace, but post-modern, pseudo-Heisenbergian notions about science are not the grounds for rejecting descriptivists’ claims that their views on usage are scientific. The reason for that rejection is that descriptivists’ views on usage have no logical connection to their linguistic studies, whether scientific or not.

•  Prescriptivists do dumb things, too, such as trying to make rules for English based on Latin; the rule against splitting infinitives, for example, is an attempt to make English follow the ways of Latin, in which infinitives cannot be split because they’re all just single words. Comment : But there is a critical difference between the two errors: Descriptivists are still claiming to be scientific; Prescriptivists are no longer trying (if they ever did; see note 2) to stuff English into a Latin suit of armor. Descriptivists have to keep digging up Bishop Lowth (1710-1787) to support their gibes at Prescriptivists, but we need only open almost any book by a contemporary linguist to see the descriptivist error in action.

•  Kids have to learn how to speak to various groups, in various circumstances—in effect, to learn several dialects of English. If all they can speak is Standard Written English (SWE), they’re going to be in a lot of trouble. SWE may be “correct” or “good” English, but it’s not always suitable. Comment: It’s well known that many gangs, cliques, and tribes punish those who don’t speak the way they do, but what does this sociological or anthropological fact have to do with linguistics or with the issue of the merits of one language or dialect as against another? We are always being told by linguists that so-called correct language is simply the language of the upper class, and of no special linguistic merit; are the dialects of other groups any more deserving of respect? And is dialect the right term for the collection of catchwords and mannerisms that one must adopt to pass muster with another stratum or sector of society than one’s own? 3

•  SWE has largely failed in the schools because it has been taught mainly by pedants and intellectuals, people that kids don’t admire or like. And the official arguments for SWE are understandably rejected by kids, who see them as dishonest or irrelevant to their own interests. Comment : The failure of the schools to turn out students who know SWE is due primarily to the fact that many teachers don’t believe in it—or know it—themselves; by now, many teachers couldn’t teach SWE if they wanted to. Incidentally, Wallace seems to think that the various “dialects” of English are equally good intrinsically, with SWE enjoying its “standard” position only because it’s the dialect of the ruling class. He is to some extent reversing cause and effect; one reason why the ruling class rules is that it is better educated, and has better linguistic tools.

•  Being candid with students about the reasons they should learn SWE (in addition to other dialects) means eschewing Politically Correct English (PCE), and that even more debased dialect, Academic English. Comment : I applaud Wallace’s rejection of PCE and the ethos it springs from. I’m also happy to see him attack Academic English, by which he means evasive, turgid, pompous, bureaucratic writing—bad writing, certainly, but not the most common kind of bad academic writing. More academics write the kind of bad English that he himself writes here, with that coyness and dreadful mechanical facetiousness that doesn’t even rise to the sad level of Donnish Wit.

•  Garner’s genius lies in presenting his prescriptive rules with a rhetorical sugar coating that makes the medicine go down painlessly. He deals with the question of authority—who is this guy to tell us how to speak our language? —by being very Democratic [see explication in the third of these bulleted items] Specifically, he keeps showing his readers how following the conventions he urges will make their writing more effective, and help their readers—and this is something usage guides haven’t done, or haven’t done anywhere near as well as Garner. Comment : Garner does handle this problem well, but he is simply following in the footsteps of several earlier usage manuals, such as Follett and Evans & Evans; it has long been customary for such guides to explain their principles, repudiate any claim to infallibility or even high authority, and otherwise tip their hat to Democracy; Fowler himself stated his principles, although not in the form of explicit, numbered rules listed in organized manner in the preliminaries, but rather dispersed among his entries. He did not bother to disclaim omniscience, since it would not have occurred to him that readers might otherwise think he was claiming it; he was writing for readers who were glad to find a book that would improve their writing, not for petulant anti-literates quick to resent as condescension any suggestion that they needed instruction. Garner’s tactical innovation in winning the sympathy of his readers, if there is any, lies in his disarmingly open, even boyish tone, which has him playing the part of the young provincial, almost of the country bumpkin, despite a curriculum vitae that includes Oxford, Cambridge, a law degree, and the presidency of his own company.

I think I have done Wallace a service by boiling his argument down to a number of bulleted steps; this précis preserves what is best in his paper, and suppresses what is intolerable, his fulsome style. And despite the comments I’ve made in the précis, I find I agree with some of his points, after they’ve been dug out of his sometimes turgid, often precious prose. I’m unhappy at having to assent to badly expressed points; I cannot quite give up the idea that good thinking must go hand-in-hand with good writing, especially when the writing is about the very subject of writing. But in this case at least I have to suppress my hopes or beliefs or prejudices, whichever they are, and say that I find some of his points sound, however repellent the style in which they’re presented.

It is another fault in Wallace that he is unwilling or unable to face the implications of his own experience as a teacher, which he tells us about at length. As an experienced teacher, he knows that many of his students are simply misplaced in nominally collegiate- or even university-level academic programs; they would be happier, and society would be better off, if they were in vocational training programs or apprenticeships. Wallace must know this, and know that a great deal of the usage war is due to the conflict between those whose first care is for the integrity of the language and the intellectual coherence that depends on it, and those whose first care is social peace at almost any cost. He refers scathingly throughout his piece to the hypocrisy of our society about social status, race relations, and so on, but he himself cannot face the fact that the only reason why American students can pass through college in the numbers they do is that academic standards have been lowered to the point where the work done by the average Bachelor of Something of today would hardly have earned him a high school diploma a few generations ago.

Unfortunately, American politicians do not keep their jobs by saying no to their constituents, and since academia is now the indentured servant of the federal and state governments, admission policies and standards of achievement have been adjusted to meet political requirements. And many even of those who are not simply the paid-for creatures of the political system are suffering from something like the Stockholm Syndrome; they have yielded to the demagogues without even trying to resist, and turn savagely on anyone who fails to do likewise.

The root of Wallace’s falsity of tone is his false relation both to his subject and his audience. He is not really interested in Garner’s book, despite his repeated use of the term genius —as I noted earlier, he says practically nothing about its actual contents—his real interest lies in its strategy for getting easily offended readers to accept instruction. This strategy is in fact the eminently sensible, but hardly original, one of speaking respectfully to the reader, allowing as to there being two or more sides to many questions, and collecting a multitude of instances showing that the problems one is discussing are real ones. That this seems a brilliant break-through to Wallace is perhaps explained by an anecdote he relates about a difficult counseling session he had with a young, female, African American student, who resented his manner or choice of words enough to make a formal complaint about him. Having gone through so career-threatening and generally unpleasant an experience may have sensitized Wallace on this topic to the point where someone who sidesteps it successfully may seem to him a genius, rather than merely a man with common sense.

Perhaps because the nominal subject of the piece is not the real subject, perhaps for some other reason, the piece is written for an impossible reader—one who is intelligent and cultivated enough to be ready to read a long piece on English language usage, but who at the same time is unaware of the long-standing war between the Prescriptivists and the Descriptivists, and even needs an explanation of those labels; who will recognize the names of the principal New Critics, but has no idea of what the New Criticism was; who is learned enough to appreciate the epigraph, but who has to be told who H.W. Fowler was—in short, a reader who does not and could not exist; an intellectual chimera. The moral is that if you write for a mythical monster, you may create something that is itself monstrous.

Winchester on Roget

The thesis of Simon Winchester’s article is that Roget’s Thesaurus has outlived its usefulness, is now for the most part being used wrongly, and should be abolished. It is “a crucial part of the engine work that has transported us to our current state of linguistic and intellectual mediocrity.” Its misuse is at the hands of people who imagine that its lists of words will supply the inspiration or ideas or polish they lack; if it were abolished, these people would somehow do better (it’s not quite clear whether they would start writing better, or stop writing altogether).

Winchester wastes no time: he utters his first glaring piece of nonsense in his opening words: “The writing world may at last be having second thoughts about Peter Mark Roget, Esquire—….” These second thoughts are ones that demote Roget from his previous exalted status, and Winchester’s evidence that such a development is taking place is:



•  The latest version of the Encyclopedia Britannica gives Roget only twenty lines, and forces him to share the pages his entry occurs on with some people of whom you will not have heard—or at least, of whom Winchester has not heard, or supposes you will not have heard. Comment: what we miss here is some indication of how many pages previous editions gave him, so that we might measure precisely how much Roget has fallen in the writing world’s estimation. If, for example, the preceding edition had given him forty lines, we could conclude that he had lost half his reputation since then; but Winchester gives us no baseline from which to measure his decline, so we cannot reach a scientific conclusion. Even more worryingly, the great Eleventh edition gave Roget no mention at all, suggesting that his star has risen rather than fallen in recent years.

•  An even more damning indication: the on-line dictionary integrated into such computer programs as Microsoft’s Word does not recognize “Roget”; if you type his name into a Word file, the spell-checker will suggest that you mean “rogue”! Comment : my own copy of Word 97 suggests, less amusingly, that “Roget” is a misspelling of either “Roger” or “rote,” but Winchester may have a wittier copy.

These two points constitute the entire body of evidence Winchester offers for his assertion that Roget is no longer the admired figure he used to be: the Encyclopedia Britannica gives him only twenty lines, and Bill Gates has never heard of him. There is no sign that Winchester is trying to be funny.

(A curious side issue raised by Winchester’s opening statement: if it were true, it would argue against his own major thesis. Throughout most of his piece, he clearly wants us to think that in charging Roget with doing the cause of good writing a disservice, he is boldly going where few or none have gone before—but if the “writing world” has already had second thoughts about Roget, where is Winchester’s boldness and originality?) As for Microsoft’s non-recognition of Roget, those who have suffered along with that company’s main products may well think that omission more indicative of Microsoft’s ignorance than of Roget’s dwindling status.)

Winchester then offers us a very hasty and confusing history of the publication of Roget, winding up with the three questions that, he tells us, we must focus on. Of these, the last is, “And what did [Roget] in fact achieve?” The answer is “superficially at least, self-evident”: it is that “ Roget’s Thesaurus is a stylish and comprehensive list of synonyms.”

At this point I abandon the attempt to follow Winchester paragraph by paragraph; the cost of doing so is too great. Having given us a “superficially at least, self-evident” characterization of the Thesaurus, he embarks on a major digression in which he describes usage manuals and language helps that appeared before Roget, makes sweeping generalizations about English society (“The style of the times [the 16 th and 17 th centuries] was all about glitter and reflection…”), tells us a great deal about Roget’s life that has nothing to do with the Thesaurus , and generally goes on heaping up page after page of dubious claims and irrelevant detail.

Winchester’s characterization of Roget is false. If by “stylish” he means that Roget sacrifices utility to elegance, he is wrong. But probably, like “comprehensive” and much else in the piece, “stylish” means nothing very specific, but is simply a kind of grace note or verbal doodling. It is quite clear, however, what he means in calling Roget a list of synonyms, and equally clear that on this fundamental point he is wrong . Roget explicitly recognizes that there are virtually no exact synonyms in the English language; he states in the Introduction that “it is hardly possible to find two words having in all respects the same meaning, and being therefore interchangeable….” 4 If in fact there were significant numbers of synonyms in English, we would hardly need Roget; a writer would need only to think of any one of the synonymous terms for the idea he was trying to express, and he’d be done—any one of them would be as good as another. On the contrary, Roget’s raison d’être is that no two terms in the English language mean exactly the same thing. (Curiously, Winchester says nothing about Roget’s modern competition, such as J. L. Rodale’s The Synonym Finder ; here is a book which, as its title indicates, actually is guilty of the crime Winchester charges Roget with, that of offering its user lists of supposed synonyms.)

Roget is meant for the writer who knows that there's a word that has exactly the meaning he wants, but who can't think of it at the moment. He turns to the thesaurus, finds via the index a section that contains a word close in meaning to the one he's trying to think of, and there, usually, finds the term he's seeking. He doesn't use Roget to find an elegant replacement for a term that’s correct, but which he’s grown tired of, or fears is not fancy enough, but for exactly the opposite reason: because he wants the one right word, not some loose pseudo-synonym. (If you have to use a dictionary in conjunction with a thesaurus, you're probably misusing the thesaurus; you should already know the meaning of the word you're trying to find—it's just the word itself you can't think of for the moment.)

To put it formally, the thesaurus, or converse dictionary, is for the writer who already knows the definiens, but needs the definiendum. This isn’t just the way I think the thesaurus is to be used, it’s the way Roget intended it to be used: the first paragraph of his Introduction states that “The object aimed at in the present undertaking is exactly the converse of [the dictionary’s]: namely, — The idea being given, to find the word, or words, by which that idea may be most fitly and aptly expressed.”

Winchester’s next misconception is that Roget had the “naïve” notion that since all users of his book would be as sensitive to language as he was, he needn’t define the terms listed in it. But as my last quotation from him shows, Roget meant his book for those who already knew the meaning of the word they sought, so definitions would be otiose. And although Winchester is the author of a book on the genesis of the OED, he has apparently forgotten that there existed even in Roget’s day reference books called dictionaries, which provide just such helps, and that even if Roget had ever thought of offering definitions, he might have considered that adding to his already large thesaurus the contents of a dictionary would produce a work that was monstrous in size, and needlessly reduplicative of a resource already available elsewhere.

To examine all of Winchester’s incidental blunders would be tiresome and unprofitable; I will content myself with just a few, selected for their amenability to quick handling. Discussing Roget’s attempt, where possible, to put words of opposite meanings in adjacent columns, he observes that “Agent and Workshop … do not have obvious opposites, of course.” Wrong — principal is generally taken as the opposite of agent . Still on the subject of antonyms, he tells us that “So set against Haste is (no, not Less Speed; there is precious little wit in Roget) Leisure.”  Winchester is even more confused here than usual; “Less Speed” is in no sense, witty or otherwise, the opposite of “Haste”; it is, if the proverb is to be trusted, the consequence of haste.

Winchester writes,

No, the central shortcoming in Roget’s Thesaurus, as I see it, stems not from the book’s troublesome structure but from something quite different—from Peter Mark Roget’s Panglossian regard for the intellectual merit of his likely readership. Roget never imagined, for instance, that … a barely literate board chairman bound for Liverpool would have his secretary’s volume by his side as he was writing his report to shareholders on the morning express from Euston.

Winchester may well be correct in suggesting that Roget did not foresee that by the twenty-first century all classes of society, even people as low as the board chairmen of corporations, would have to write or pretend to write, and that his thesaurus would be put by some to uses that he not only did not intend, but explicitly rejected. Unfortunately that’s what’s happened, and his book is therefore frequently misused today; what shall we think and do about that? Shall we condemn Roget and proscribe his book because, like many other resources, it is often abused? In trying to decide these questions, I suggest we begin by trying to think of any resource, intellectual or material, that cannot be abused. If we decide to abolish Roget , we might want to do so just after we abolish the automobile, which kills some 40,000 people a year in the United States alone, and the telephone, that medium of threatening and obscene messages, bad news, and dinner-interrupting sales pitches. Or if it’s literary abuses you’re chiefly concerned about, let’s abolish plagiarism by burning all existing literature; then the perpetrators will have nothing to plagiarize from.

Winchester generously recognizes that he is not the very first to criticize Roget on these grounds, and treats us to a couple of paragraphs from one Edwin P. Whipple, who also considers Roget a book that tries to spare writers the work of hard thought by putting unearned words at their disposal. Whipple writes, “In our opinion, the work mistakes the whole process by which living thought makes its way into living words, and it might be thoroughly mastered without conveying any real power or facility of expression.” But Roget makes no assumptions whatever about “the process”; it simply observes that we sometimes forget a word, and need to be reminded. (If Whipple’s charge against Roget were sound, how much more culpable the dictionary would be, which not only gives the lazy writer all the words in the language, but actually provides him, without effort on his part, with their meanings.) And just how might Roget be “thoroughly mastered”—by being memorized?

Winchester then tells us that when he began the piece, he considered writing or phoning some well-known writers to see how they used Roget , if at all. He offers us a list of a dozen or more potential respondents, sketches the questions he was planning to ask them—and then tells us that he didn’t bother to get in touch with any of them, since he already knew the answer he would get. This is too bad; it would have been interesting to see the answers he got from Anthony Burgess, Cole Porter, and Robert Lowell, all on his list—it is hard enough to get living writers to answer questionnaires; it would be truly sensational to collect answers by séance. (This may be unfair; perhaps Winchester began writing his piece when all these men were alive.)

But although he does not offer us messages from the beyond, he at least tells us the truths he would have heard if he’d taken his poll: “Everyone has the book. Occasionally one makes use of it. But one never, never relies on it to help with the making of good writing. It may be used once in a while, to jog the memory, to unstall a synaptic moment. But it should never be trawled through or mined; its offerings should never be taken and transfused into a paragraph as relief for emptiness of thought.” To which one can only say, “Right! And never try to eat a fire hydrant, or arm-wrestle with a grizzly bear.” It was for talking as Winchester does here that Polonius came to a bad end; one begins to sympathize with Hamlet’s impetuous thrust.

And, Winchester assures us, he practices what he preaches: “Whatever merit all these paragraphs may have as writing, it seems appropriate to mention that Roget was not once employed in the selection of les mots justes . Such words as these paragraphs contain came, as they should do for all writers, from within—from memory, experience, conversation, reading, imperfectly recalled strands of knowledge.” If Winchester’s prose is what results from denying oneself recourse to Roget, I can think of no better argument for leaning heavily on that book.

Building toward his peroration, Winchester quotes Shakespeare to lend a high tone to his concluding words. Strangely, though, he does not quote directly, but at second hand, drawing on a passage from The Complete Plain Words by Sir Ernest Gowers (whom he twice calls “Gower”). The quotation consists of the first quatrain of sonnet XXXIII, presented in two parts: the first pair of lines, then the words, “and continues, as Gower cites,” followed by the rest of the quatrain. I don’t recall ever seeing a piece of poetry so butchered, nor can I imagine what the purpose of dragging Gowers in might be—perhaps to vouch for Shakespeare? Then Winchester springs his surprise: he informs us that there was no Roget when Shakespeare wrote those lines, “and yet the writing is perfect….” (If it occurs to some curmudgeonly reader that neither was there a Winchester in Shakespeare’s time, no inference is to be drawn from that fact.) At any rate, with this coup, Winchester has really put paid to the claim, so often heard in literary circles, that it’s impossible to write well without Roget . But perhaps he wants us to infer more: to conclude that if Roget were abolished, we too could write like Shakespeare? If this is going too far, what is the point of telling us that Shakespeare wrote without benefit of Roget ?

The piece winds up by banishing Roget as a book whose use is shameful, even damaging to the brain, in a style that made me think of Doré’s etching of St Michael expelling Adam and Eve from the Garden. Winchester gets rather wound up and grandiloquent about the evils of Roget , perhaps forgetting (it is a long piece) that he earlier assigned the book at least a small role in the literary world; some pages back he reported the result of his imaginary poll: “[Every writer] has the book. Occasionally one makes use of it. But one never, never relies on it to help with the making of good writing. It may be used once in a while, to jog the memory…” How often it may in good conscience be so used, Winchester does not say; nor does he explain just how one avoids using it “to help with the making of good writing,” or what objective one should have in using it. But if used correctly—to jog the memory—why should there be any limit to the number of times a writer may have recourse to it? Winchester is unwilling to conclude by acknowledging that there is a proper use for Roget , and his refusal puts him in the silly position of proposing something like a quota system for consulting the book— and without specifying the quota.

Just as Wallace is not really interested in the contents of Garner’s book, Winchester is not really interested in the contents of Roget’s. The Thesaurus’ real offense in Winchester’s eyes is that it is a resource that, like almost all the resources of civilization, is of help chiefly to the educated and cultivated, and may even be harmful to those who are neither. Like all advanced tools, it multiplies both the advantages of the skilled and the disadvantages of the others, just as power tools let a good carpenter be even more productive, and a clumsy one ruin even more wood . This makes Roget wicked in the eyes of strict egalitarians, who would sooner that a resource not exist at all than possibly widen the gap between the favored and the non-favored. And the offense, if real, cannot be mitigated by combining the thesaurus with a dictionary, as Winchester seems to imply; if Roget is guilty as charged, is not the dictionary itself just another instrument of the elite, whereby they seek to impose their hegemonic authority on us? Aren’t all reference books, including the Encyclopedia Britannica that Winchester relies on for much of the information in his piece, just more tools that they use to keep us in bondage?

Making fun of Winchester is easy, and I intend to continue doing it; one must not shrink from performing a duty just because it’s easy. That Roget —like every other resource—is often abused is unquestionable. Abused, it enables bad writers to tart up their bad writing with words they don’t fully understand; this is too bad, but the idea that it does more harm than the book’s legitimate use does good is ludicrous; that the book so abused should be abolished (how? by being publicly burnt?) is simply nutty; the sort of notion that you hear from those who claim they are controlled by the CIA via messages received on the fillings in their teeth. And not only is this, its main thesis, silly, but virtually all the incidental points Winchester tries to make along the way are at best dubious; at their frequent worst, demonstrably false. (Improbable as it may seem, I have refrained from noting many of the faults in Winchester’s piece, as I did in Wallace’s; my energy, this journal’s space, and your patience are all limited.)

The question may occur to me only because of my special interest in English usage, but I wonder if there’s any other subject on which so bad a pair of articles as the Wallace and Winchester pieces could be published in quality magazines?

Article Footnotes

1 Bryan A. Garner, “A Texan Fowler?” English Today 64 (Cambridge U.P., October 2000), pp. 3-10, at 4.

2 Wallace is on shaky ground here; see, for example, Michael Dummett, Grammar & Style (London: Duckworth, 1993), 60-61; a forthcoming book by Peter Corey will treat the issue at greater length.

3 Those whose dialect one fails to fall in with are not always cruel. Some years ago, a fellow employee of a computer software company made me an offer that I declined with the words, “No, thank you.” He looked at me in a kindly manner, then gently helped me: “No way, José!” he said, teaching me the formula I should have used.

4 C. O. Sylvester Mawson, Roget’s International Thesaurus of English Words and Phrases (New York: Thomas Y. Crowell Company, 1911), Roget’s “Introduction,” page xiv.