Why Linguists Are Not to be Trusted on Language Usage

Mark Halpern

Writing a book on computer programming a few years ago, I had occasion to mention the fact, as I then supposed it, that Eskimos had special terms for a great many varieties of snow.  I was reluctant to trot out this old, well-worn story—next, I said to myself, you'll be quoting Santayana on repeating the past—but it was the perfect support for the point I was trying to make at the moment, so, taking heart from Fowler on clichés, I used it.  My discomfort at using so overworked an illustration became real chagrin, though, when I learned that it was not only hackneyed but false.

In a collection of short pieces with the provocative title The Great Eskimo Vocabulary Hoax I found an essay of the same title that exploded the notion that Eskimos have especially extensive snow vocabularies.  Pullum, professor of  linguistics and dean of graduate studies and research at the University of California, Santa Cruz, described how the error had originated and how it had passed from book to book until it became one of the things that every schoolboy knows.  So far so good; the embarrassment I felt on finding myself among the many dupes who had been taken in by this piece of misinformation, and propagated it, was easily overbalanced by the satisfaction of being finally enlightened, and by the interest of the story Pullum had to tell.  In the ordinary course, I would be grateful to Pullum for correcting my error, but the manner in which he does it makes it hard to feel gratitude.

First, he seems to feel that the error is a shameful one, illustrative of human credulity in the face of the absurd and irrational, and of "falling standards in academia."  But why we should have known better than to believe that Eskimos have an especially rich and nuanced vocabulary for describing their natural environment, he does not explain.  Nor does he explicitly suggest that readers should check against original sources every assertion made by every writer, no matter how plausible the assertion or how credible the writer; yet such a course would be the only way to preclude such errors.

Second, he is puzzled, and a bit angry, that the error was not quashed by a paper delivered by Professor Laura Martin at the 1982 annual meeting of the American Anthropological Association; referring to her presentation, he says, "that ought to have been enough for the news to get out."  Perhaps professors, even deans like Pullum, are still unworldly; this one appears to think that an obscure paper delivered at an obscure conference by a scholar whose name is hardly a household word should have made the world sit up, take notice, and mend its ways.  This strikes me as far harder to believe than that Eskimos have many words for snow.

Third—and by far worst—he and Professor Martin think our readiness to believe the canard is, in part at least, due to our racism.  He quotes her and adds a comment:

"We are prepared to believe almost anything about such an unfamiliar and peculiar group," says Martin, in a gentle reminder of our buried racist tendencies.

It is good of Professor Martin to be gentle, and to say "we" when she clearly means "you fools" or worse.  But she, too, fails to explain why it is racist of us to credit the Eskimos with words for many varieties of snow.  Speaking for my possibly bigoted self, I felt closer to the Eskimos when I supposed them to be walking Rogets on the subject of frozen water—that's just what I would be, I think, if I lived near the Arctic Circle.  But it may be that to understand my feelings, Professor Martin would have to learn something of the ways of my unfamiliar and peculiar group, the non-academics.

Pullum lays into us for our readiness to believe the worst of our primitive cousins:

Never does a month (or in all probability a week) go by without yet another publication of the familiar claim about the wondrous richness of the Eskimo conceptual scheme: hundreds of words for different grades and types of snow, a lexicographical winter wonderland, the quintessential demonstration of how primitive minds categorize the world so differently from us.

He is clearly enjoying himself, and goes on in this vein for some pages, but does not stop to show us why the belief that the Eskimos have many words for the realities they live with should amount to patronizing them as "primitive."  On the contrary, he tells us it would be "utterly boring, even if true," since it would be only what one would have expected.  The insult, he implies, lies in our thinking it necessary to make such a statement about Eskimos, when we wouldn't bother to remark about Texas horse breeders, say, that they had words for many types and conditions of horses.  So the insult lies in our making explicit observations, when describing a little-known people remote from us geographically and perhaps in other ways, that we would not feel we needed to make about our next-door neighbors.  A rather subtle insult, this; not an occasion for a great deal of indignation and charges of racism, one would have thought.

But if a rich vocabulary for their daily concerns is something one should in common decency take for granted in one's fellow humans, near or remote, what shall we think of the Eskimos now that Pullum tells us that they do not in fact possess such a thing?  Perhaps aware that he has talked himself into a dilemma—it isn't easy to tell, since he doesn't present his thoughts in a consecutive order—he later tells us that Eskimos probably aren't much concerned with snow anyway:

And actually, when you come to think of it, Eskimos aren't really that likely to be interested in snow.  Snow in the traditional Eskimo hunter's life must be a kind of constantly assumed background, like sand on the beach.  And even beach bums have only one word for sand.  But there you are: the more you think about the Eskimo vocabulary hoax, the more stupid it gets.

Strange to find a research scientist and upholder of strict scholarly standards telling us what snow "must be" in the life of an Eskimo hunter, rather than presenting the results of his investigation into the question, but perhaps Pullum deserves a breather from the scientific rigor he has been observing up to this point.  One does wonder, though, about the validity of an analogy between a beach bum's concern with sand, whose worst effect might be to get into his sneakers, and the Eskimo hunter's concern with snow, which if misjudged might kill him.

In my own book, I had reached for the Eskimo illustration (in the course of a discussion of mechanical translation and its difficulties) to support the point that all human languages have vocabularies that are especially rich in those topics that are important to their speakers; if I'd thought this were true only of the languages of "primitive" peoples, it would have been useless as evidence for an assertion about human language in general.  And although I have seen this illustration cited so often in print that, as I've said, I groaned at the idea of invoking it yet again, I have never seen a use that was in the least contemptuous or patronizing; the message is always, this is how we humans speak and think.

Pullum is so pleased with the title "The Great Eskimo Vocabulary Hoax" that he uses it for both the essay and the book, but the twice-made promise is not kept; his text charges no one with perpetrating a hoax.  What he describes is merely an ordinary and innocent error, copied by one writer from another because it is plausible; because it illustrates the truth that vocabularies generally reflect the interests of their users; and because, contrary to the suspicions of Pullum and Martin, it tends to create an affinity between temperate zone-dwellers and the Eskimos.  To present it in the excited and accusatory tones of a checkout-counter tabloid is to mislead, to generate factitious scandal—indeed, to hoax.

We all, of necessity, take as true anything that sounds reasonable, and that we have heard from a dozen respectable sources, until it is shown to be false.  This practice sometimes leads us into error, but to drop it would be to abandon any possibility of accumulating and transmitting knowledge at all.  And Pullum would do well to be a little more forgiving of the practice, since he himself believes (Hoax, p. 189) that Acton said "all power corrupts."  I hope his repetition of this popular canard is just an ordinary error, and not a sign of decline in academic standards, or of buried anti-Catholic tendencies.


The following letter appeared on page 8 of New York magazine for June 13, 1994:

In the "Fast Track" piece "The Very, Very Tiresome Season of Storms" [by Steven J. Dubner, February 28], Dr. Steven Pinker of MIT is quoted as saying that Eskimos don't really have hundreds of different words for snow.  He states, "They have exactly as many or perhaps two or three more words than English speakers."  I have spent the past 22 years living in an Inupiat Eskimo community.  According to the North Slope Borough's Inupiat History, Language, and Culture division, the Inupiats have more than 30 words for snow, and more than 70 for ice.  In the Arctic, the specific conditions of snow and ice are critical to hunting and survival; two or three words would hardly cover our needs.

     Elise Sereni Patkotak
     Public Information Officer
     North Slope Borough
     Barrow, Alaska
The Credo of a Czarist Émigré

At several places in his book, Pullum speaks of the pain professional linguists suffer in witnessing the blundering efforts of laymen to regulate the English language.  He wonders at their combination of unquenchable interest and benighted ignorance, their refusal either to stop meddling or to fully inform themselves on questions of usage. In the course of this discussion, he speaks with approval of a paper written some ten years earlier by his friend Geoffrey Nunberg, professor of linguistics at Stanford University, on language usage and the war between the descriptivists and the prescriptivists.  His reference to Nunberg's article (he gets its title wrong, calling it "The Grammar Wars," but as he points out, academic standards are everywhere in decline) suggested that the correct scientific attitude to language usage would be found fully displayed there, so I looked it up with the highest expectations.  And I was not disappointed; it does indeed display that attitude at length, and reveals why that attitude is unwarranted, pernicious, and to be resisted.

My section title is drawn from a remark of Nunberg's.  One of the chief whipping boys in his paper is John Simon, cultural critic and outspoken prescriptivist; Nunberg, after quoting what he thinks a particularly outrageous statement by Simon, says, "That is the credo of a czarist émigré, not an English grammarian."  I have chosen to throw Nunberg's words back at him for two reasons.  First, because it amuses me: I am largely in agreement with Simon, and my grandfather, a refugee from pogroms that were at least tolerated by the czar of Russia, would have been fascinated to learn that his grandson held views that caused him to be lumped together with the czarist supporters who had to flee after the Revolution.
Second, the simile Nunberg has chosen to express his outrage is revealing of his typically academic-cum-progressive assumptions.  It is now generally accepted by students of Russian history that under the last czars Russia was making progress—slow, painful, and far from unbroken progress, but progress nevertheless—toward modernity.  It was the regime that succeeded the Romanovs, and ended just a few years ago, that turned backwards toward despotism.  Would that the czars had remained in power to lead Russia out of the darkness, and the émigrés not been forced into the bitterness of exile and Nunberg's disapproval!  And as the czarist émigrés turned out to have been the real political progressives, compared with the regime that exiled them, they may be right as well on language usage.
But before engaging Nunberg, I must pay him tribute.  Only someone who has been closely following the debate—or better, the alternation of largely unrelated monologues that has passed for a debate—between the prescriptive and the descriptive grammarians can fully appreciate Nunberg's contribution.  It exhibits relatively little of the bad temper, personal scorn and hauteur that nearly always characterize the tone of the professional linguist addressing those who lack professional credentials but persist in trying to lay down the law about language usage.  It is witty, good-humored, and almost always fair, and Nunberg appears to have actually read some of his adversaries' writings, and even to have found some occasional merit in them.  All this is refreshing and encouraging, and before taking issue with him, I want to salute him as an adversary with whom it is a pleasure to do battle, and hope that in contesting his views I can leave the reader in as sunny a mood as he does.
Nunberg, although a university teacher of linguistics, presents himself not as the usual hardshelled descriptivist, but as a moderate and a seeker after compromise or synthesis (the subtitle of his piece, which I would guess was supplied by an editor without his knowledge, is "An Argument for a middle way between permissiveness and traditionalism").  And he almost succeeds in this imposture, claiming with some plausibility to be one of us at heart: he charmingly 'confesses' that he "can't overcome the feeling that it is wrong for me to use [gift and impact as verbs]," even while knowing that they may in future come into general use as verbs despite his feelings, just as contact has already done.  Then he quotes a famous passage of Fowler's, and not only praises it, but tells us that it caused him to modify one of his own writings.  In short, Nunberg very nearly disarms us by showing that he appreciates fine distinctions, and shares the literary tastes underlying the prescriptive position.
But his display of neutrality or even collegiality is misleading; Nunberg is indeed not the run-of-the-mill academic descriptivist, but something much more dangerous, a Trojan horse from the descriptivist camp.  His demonstration by means of well-chosen quotations and discriminating name dropping that he is as alive to the finest literary nuances and logical distinctions as any admirer of Beerbohm or Fowler is only a diversionary tactic.  In taking issue with us prescriptivists, he is gentler than are most descriptivists, and readier to grant some of our minor points, but in all doctrinal essentials Nunberg is the classic academic linguist, chastising us in the joint names of science and democracy.
He makes two chief objections against prescriptivism.  The first is the scientific objection: laws of nature are involved here, and those trying to influence linguistic events without knowledge of linguistic laws are simply demonstrating their ignorance, and making public fools of themselves—he likens them to landscape gardeners trying to stop or modify the processes of plate tectonics.  The second is the democratic objection: the prescriptivists are attempting to foist their own linguistic practices, which are usually the practices of the educated, affluent, fortunate members of society, on the less educated and affluent members—and since there is a strong positive correlation between these two classes and the white and non-white portions of the population, respectively, there is some suspicion of racism at work here as well.
One of the points that Nunberg, like all his school, is most eager to make is that the 'rules' of grammar, and of good usage generally, have no scientific basis; they are just someone's idea of what is proper, and that idea changes from generation to generation.  The descriptivists are so eager, indeed, to make sure this point has registered that they seldom stop making it long enough to hear the reply: "Yes, we know this; we do not contend that the rules we would like to see observed were handed down from on high.  They are ordinary man-made rules, and neither divine commandments nor scientific laws (although many have some claim to scholarly support), and we agree that they, like all man-made things, will need continual review and revision.  But these facts are no more arguments against laws governing language usage than they are against laws governing vehicular traffic.  Arbitrary laws—conventions—are just the ones that need enforcement, not the natural laws; the law of gravity can take care of itself, the law that you go on green and stop on red needs all the help it can get."
Nunberg objects to rules about grammar because, among other things, they interfere with our "working [grammatical issues] out" for ourselves.  One wonders what population Nunberg has in mind here; is he imagining that American high school or college students, or many people anywhere, would be puzzling out grammatical issues for themselves if prescriptive grammarians like John Simon had not, by laying down the law to them, stifled their healthy curiosity and hunger for learning?  He also objects to the right wing ideology that he sees behind the grammatical strictures of people like Simon, Safire and Buckley, telling us that earlier grammarians like Samuel Johnson would "probably be distressed to learn that their standard had been taken up by the right."  (This is interesting as evidence that the current non-specialists' misconception about Johnson is that he was a liberal; the traditional error was to think him merely a curmudgeonly Tory).  But he can be pretty insulting to the common people himself when he gets his dander up: we are told that William Safire brings a certain class of readers "out of the woodwork," whereupon they exchange with him "schoolmarm maxims and scraps of linguistic folklore."

Linguistic science is not the basis of Nunberg's views
There are two great flaws in Nunberg's thesis, either of them sufficient to destroy that thesis completely.  The first is his claim that linguistics is a science, and a science relevant to usage issues, so that those versed in it must have the last word—if not the only word—about such matters.
If one had to select a single sentence of Nunberg's to serve as a précis of his paper, it would be this: "But it is impossible to talk intelligently about the language nowadays without having an idea of what the program of modern linguistics is all about..."  And this key statement is false.  I contend—and I believe I have some idea of "the program of modern linguistics"—that what linguistic scientists have been doing in this century, whatever value it may have for other purposes, has absolutely no relevance to the constellation of literary-philosophical-social-moral issues that we are talking about when we discuss usage.  And Nunberg, in his own practice, confirms my point: he gives reasons throughout the paper for his own specific judgments in matters of usage, and nowhere are the findings of linguistic science among them.
Nunberg wants us to understand that the study of language is "complex and technically demanding," and that many questions that seem simple to the "pop grammarians" like Simon and Safire can really be handled intelligently only by the professional linguists.  He gives as an instance the fact that it is acceptable to say Everyone has taken his coat but not Everyone left, didn't he?, and tells us that "Linguists have puzzled a great deal over examples like these, in an effort to understand the basic workings of English syntax; it is small wonder that they should have little patience with popular grammarians who blithely proclaim that the whole problem is simple...."  But Nunberg makes no attempt to show the relevance of such puzzlings, or any other aspect of linguistic science, to the issues that concern prescriptivists.
Linguists think they are practicing a scientific discipline whose object is to learn the rules that govern a department of nature, linguistic behavior.  An example (to call it an anomaly, as Nunberg does, is to beg the question of whether there exist laws to be broken, but this point can be waived here) such as that just quoted from Nunberg is their meat and drink; to them, it is like the anomaly in the orbit of Mercury whose discovery and study led to such great advances in physics.  But whether the linguists are right or not in supposing themselves to be scientists, and in thinking that anomalies such as this are opportunities for the discovery of real linguistic laws, the phenomena they are studying have nothing to do with the interests of the prescriptivists.  Nunberg, speaking for "most of my fellow linguists," ridicules prescriptivists as landscape gardeners setting their faces against continental drift, and the analogy is better than Nunberg knows—the relationship between the linguistic scientists and the prescriptivists is indeed very like that between geologists and landscape gardeners, and can profitably be extended beyond Nunberg's use of it.  If the "frantic efforts" of the gardeners "to keep Alaska from bumping into Asia" are ridiculous, is it not equally silly for geologists to tell landscape gardeners that they must not presume to pollard a lime tree, or put in a fishpond, without deferring to the experts on plate tectonics?
It is in fact the most striking feature of Nunberg's paper that he makes no use of the "findings" of linguistic science at any critical point, and little use anywhere.  After telling us emphatically and at length that only those versed in the mysteries of the science he professes are really qualified to pronounce on issues of language usage, he nowhere calls on his own esoteric learning; his views on specific issues are based simply on his conception of manners, decency, courage, clarity, and other virtues with which one is often glad to agree, but which do not seem to be the product of linguistic science.  Many of Nunberg's specific judgments on points of usage are so tactful and sensible that we may easily, in our pleasure at being able to agree with him, overlook the fact that he makes them simply as a tactful, cultivated and sensible fellow, not as a scientist.  So Nunberg, in this paper, is not a scientist; but what if he were?

Language is not part of nature, hence has no laws or natural destiny
Here enters his second great error: to suppose, like all the descriptive grammarians, that language is an entity with its own laws of development, or natural destiny, and that prescriptive grammarians are trying to interfere with the course of that natural destiny.  He objects to this attempt on two grounds: it is futile, since language will follow its natural destiny despite all the efforts of the prescriptivists; and it is somehow wrong—immoral? unethical?—to try to interfere, even though the attempt must be futile.  But neither Nunberg nor any other linguist has offered any evidence for either of these points.
An acorn, left to itself, becomes an oak, and a geneticist altering its DNA to make it grow into an elm, or a fish, may justly be said to have interfered with its natural course.  But what does language, undisturbed, become?  What course do we know it would take, if only outsiders would cease to meddle with it?  No one has ever shown that language has such a natural course, much less that it would be in any sense wrong, if such a course did exist, to alter it.
Of the many attempts that have been made to regulate language usage one way or another, some have succeeded, some failed.  And we do not know, even now, why some succeeded and others did not; are we to take it on faith that the ones that succeeded were somehow in accordance with language's inherent nature, and the others were somehow not?  Nunberg and his allies have no scientific standing in their quarrel with a Simon or a Safire; if they disagree with such prescriptivists, they do so not as scientists observing from above the fray, distinguished by superior knowledge and disinterestedness, but simply as fellow gladiators down in the arena, as biased and opinionated as their adversaries.  How the battle will turn out, how it should turn out, no one can say with any authority; and since there is no natural course with which to interfere, there is nothing that can properly be called an interference; 'interferences' are simply what linguistic scientists call those events of linguistic history that they cannot accept.  The practitioners of linguistic science cannot reasonably quarrel with prescriptivists any more than they do with ghetto youths; like those youths, we are their data.
It is instructive that those who pride themselves on being pure descriptivists, standing neutral above all battles, recording the facts they find without bias, and searching for the laws of linguistic nature, should take issue with the prescriptivists when they do so with no one else.  Why do the grossest offenses against cultivated speech or writing arouse in the linguists no emotion, but only scientific interest, when other linguistic phenomena—the strictures of a John Simon, for example—arouse their passions to the point where even so civilized a man as Nunberg sounds as if he'd like to have Simon whipped?  Nunberg's view, I think, is this:
Simon is sophisticated, and a man who has thought long and hard (however mistakenly) about language usage; the uneducated offender against traditional correctness has given such matters no thought whatever.  Hence Simon is interfering with nature, where the ignorant offender is simply exemplifying it.  If Simon is allowed to influence matters, the natural laws of linguistic development will be obscured or distorted; he is interfering with the experimental animals and corrupting the experimental results.
Nunberg and his allies have not absorbed Burke's dictum: Art is Man's Nature.  And the nature under investigation by linguistic science is man's nature, of which a desire to erect standards, and use them to correct practice, is an essential element, not an aberration.
Perhaps because he thinks of himself as a scientist, and listens to ordinary speech as if observing the curious behavior of some primitive tribe, Nunberg's comments and questions on specific points of usage are sometimes thuddingly unperceptive.  Thinking that he has spied another of those contradictions in the views of "modern critics" that a scientist can use against them, he asks, "Why is it all right for a politician to use capital-intensive, once a technical term of economics, but not revenue enhancement?"  One has to think of oneself as a scientist, perhaps, to miss the simple answer that the former is merely a non-tendentious technical term now found useful by non-economists, while the latter is an attempt at deception; at concealing a tax increase from the public.  He defends borrowings from technical usage against the charge of barbarism, saying, "For the most part, the borrowing is natural and inevitable—what else would you call a minicomputer or a quasar?"  Again, only an expert could fail to see the answer: one would call them nothing else, because those are their proper and only names.  Technical terms are perfectly in place when applied to technical matters; it is only when they are borrowed (actually, stolen; they are never returned) to inflate some triviality or mask some mendacity that we "modern critics" get riled.
After expressing admiration for Fowler's little gem on using quote marks to show one's superiority to slang even while availing oneself of it, Nunberg says "It obviously has done little to stem the mania for quotation marks (We are "closed," I saw in the window of a shoe-repair shop the other day)..."[32/2].  No, Fowler's discussion of quote marks as devices for simultaneously using slang and showing ourselves superior to it did nothing to prevent the ignorant of a later time from misusing quote marks as general intensifiers, and for the excellent reason that it did not deal with that problem.
Nunberg's observation, although irrelevant to the argument in which it occurs, was nevertheless an important one for him to make and for us to notice. He has caught himself praising an effort made by an educated man to influence usage, and fears he may in doing so have conceded an important point to the opposition.  So despite its strictly logical irrelevance, he hastens to repair the damage with an anecdote that is supposed to show that individual effort, even that of a Fowler, must be futile in matters linguistic.  The laws of that realm, he implies in pointing to the shoemaker's sign, will work their will, whatever we may do or say.
Speaking of the conflation in popular speech of disinterested and uninterested, he says:
But there is no point making a fuss about [it], because it was forgone that disinterested would lose its older sense once interested lost the sense of "having a stake in," which we retain only in the fixed phrase interested party.  Even if disinterested had survived intact, therefore, it would eventually have become one of those curious asymmetric negatives like untoward and disgrace, whose senses are not recoverable as the sum of their parts.
Nunberg's reasoning is faulty in several respects here.  First, he contradicts himself in telling us that the disappearance of disinterested in the sense under discussion became inevitable once interested had ceased, for most purposes, to carry the opposite meaning; in the very next sentence he describes a whole class of terms—"those curious asymmetric negatives"—that have suffered the same loss, but survived very nicely.  Second, he makes a question-begging assumption in stating that disinterested must lose the sense he mentions—notice that he speaks of it in the past tense to convince us that it has already done so.  Finally, he suffers from the erroneous notion that disgrace is "one of those curious asymmetric negatives ... whose senses are not recoverable as the sum of their parts"; apparently the word has fallen from grace in linguists' eyes.
But as to Nunberg's main point—his claim that disinterested, if it had survived intact, would have joined the (presumably disreputable) company of untoward and its kin—its definitive, conclusive refutation is, So what?  What is undesirable about such "curious asymmetric negatives"?  Their lack of a positive form presents no difficulty to anyone, and they positively increase the gaiety of nations.  They lend themselves to a form of verbal playfulness based on writing and talking as if they do have such forms: there are those who derive harmless amusement from using the officially non-existent *ept, *choate, *kempt, *sheveled, *trepid, and the like; Nunberg may have misgivings, even be nonplused by the practice, but most of us have ample *givings, and remain coolly *plused. To such considerations Nunberg may be indifferent, but we czarist émigrés are quite *different.
And so susceptible is the language to unsanctioned amateur initiatives that such simple joking may eventually bring some or all of those non-words into general use and respectability (as with burgle, for example), whereupon Nunberg and his colleagues will write heavily footnoted papers explaining how symmetrising back-formation is part of the ineluctable march of linguistic laws, and further evidence that amateurs are out of their depth in talking about usage.

Scientists versus Mavens

The latest salvo in the bombardment of the amateurs by the professionals occurs in a recent book by Steven Pinker, who is described on the dust jacket as "professor and director of the Center for Cognitive Neuroscience" at MIT.  The novelty and main value of Pinker's book lies in his weaving together the recent findings of brain researchers, geneticists, and developmental psychologists to formulate an admittedly primitive but suggestive picture of how humans learn to speak.  It includes a presentation of the work of Noam Chomsky that is the clearest I have yet seen; a balanced and sober analysis of the efforts over the last decade or two to get chimpanzees and gorillas to use language as humans do; and an interesting attempt to show that the development of language can be accounted for by standard Darwinian natural selection, despite Chomsky's doubts.  He also tells with delight (pages 64-65) the story of how Pullum and Martin shot down the Eskimo canard, or Canada goose: "No discussion of language and thought would be complete without the Great Eskimo Vocabulary Hoax.  Contrary to popular belief, the Eskimos do not have more words for snow than do speakers of English."  And so on for a page and a half, unwittingly showing by example how silly stories get transmitted from book to book.

But in the twelfth of his thirteen chapters, "The Language Mavens," Pinker takes a little holiday from science, and attempts to show how ridiculous most non-linguist writers on language usage are—his chief target is William Safire, although glancing blows are aimed at John Simon and Jacques Barzun in passing.  He wastes little time with either of the last two: Simon is simply a "malicious know-nothing," while Barzun "earned an 'F' when he called a possessive noun like Wellingon's an adjective"—if we wonder where Barzun did this, and what he meant by it, we are referred in the endnotes to a book by Dwight Bolinger; Pinker offers no further explanation for Barzun's failing grade.

Most of the chapter consists of analyses by Pinker of specific usage arguments offered by the 'mavens'—sometimes attributed to a specific maven (usually Safire), sometimes to unnamed "defenders of the standard" (Pinker nowhere mentions Fowler).  And in most of these encounters Pinker, I think, comes out ahead.  He wins partly because he has carefully hand-picked the battles he fights, and has little trouble finding cases in which various mavens have taken indefensible positions; partly because he has chosen straw men to attack—is anyone still worrying about ain't, or the splitting of infinitives?  But the chief reason for his triumphs is that many of the mavens make the fatal mistake of trying to justify what I earlier called their literary-philosophical-social-moral views with historical facts, or supposed facts, about the development of the language—in short, with citations and arguments about which the professional linguist will usually be better informed, or at least able to sound more authoritative.

In acting as amateur linguists, the mavens not only commit the tactical error of playing their adversary's game, but profoundly falsify their own position; they often become so engrossed in finding what they suppose to be good logical or historical reasons for their preferences that they forget that they are not linguists, but moralists and literary critics.  There seems a kind of poetic justice in the mavens' losing so many of their battles this way; if it's ignorant and silly of the linguists to pretend to occupy the scientific high ground, it's unforgivable in the mavens—they should know better. They are as much to blame here as are the linguists when—as we have seen and shall see again—they try to foist their personal tastes in usage on us, while professing to be speaking for science.

The difference between the linguists and the mavens—or the descriptivists and prescriptivists, to use the more usual terms—is captured neatly in the dual meaning that "grammar" bears in such books as Pinker's.  To the descriptivist, "grammar" is usually short for "generative grammar": the hypothetical mechanism, embodied in the brain, that produces sentences.  To call any recorded utterance ungrammatical, given this sense of the term, is to make a strange, almost meaningless statement; it is like criticizing the way the stomach produces digestive juices.  To the prescriptivist—and the majority of the educated public—grammar means the mechanism, embodied in books and teachers, that decides whether what you've said was well said, and the charge of being ungrammatical is one that we often find justified.  Now Pinker is well aware of the dual meaning of "grammar" and its derivative forms—he spends the first pages of chapter twelve in making the distinction I have just summarized—and he draws a correct conclusion: "They [the descriptive and prescriptive meaning of rules, grammar, etc.] are completely different things....One can choose to obsess over prescriptive rules, but they have no more to do with human language than the criteria for judging cats at a cat show have to do with mammalian biology."

Pinker's phrase "obsess over" does not sound like the language of a scientist (unless he is seriously asserting that prescriptivists suffer from a clinically recognized variety of psychopathology), and his statement that rules have nothing to do with language needs qualification—he means that they have nothing to do with those aspects of language he is professionally interested in—but apart from these characteristic stigmata, his statement is unexceptionable.  But one wonders why Pinker is unable to proceed to the next step in the argument: if the two viewpoints are so radically different, on what grounds does the descriptivist criticize the prescriptivist?  Pinker provides in his biologist-versus-cat fancier analogy another argument, like Nunberg's analogy of geologists-versus-landscape gardeners, that deserves to be pushed a step beyond where its author leaves it.  Cat fanciers clearly have no grounds for telling mammalian biologists how to go about their business; have the biologists any more grounds for telling the fanciers that this shorthair is too cobby, or that Siamese's points too dark?

This utter dissimilarity in interests and goals makes it impossible for the descriptivists to criticize the prescriptivists with scientific authority; surely they may comment, but when they do, they do so merely as rival prescriptivists.  Pinker quotes from someone he identifies only as "a poet" (why he is reluctant to say "Auden" isn't clear) this dictum: "when [language] is corrupted, people lose faith in what they hear, and that leads to violence," then remarks

The linguist Dwight Bolinger, gently urging this man to get a grip, had to point out that "the same number of muggers would leap out of the dark if everyone conformed overnight to every prescriptive rule ever written."

One wants to ask Bolinger, with equal gentleness, "How do you know?"  A world in which all usage rules were followed would certainly be very different from the present one; perhaps the crime rate in so law-abiding a world would be lower.  An academic post in linguistics seems to qualify one to make authoritative statements on all sorts of matters.

Again, he examines an analysis by Safire of an utterance of Barbra Streisand's, and defends Streisand against Safire's charges of bad usage.  But at one point in his defense he says "Now, I cannot be sure that this is what Streisand had in mind...," which rather gives the game away; if we assume that Streisand wanted to be understood, then her utterance has failed—even her defender cannot be sure he knows what she meant.  Pinker's usual defense of vogue words, solecisms, and other problematic examples of language usage is just the opposite, though. In examining an assortment of utterances that have been attacked by prescriptivists, he defends them by saying in effect what every English teacher hears every day from students shocked at the grades their compositions have received: "You knew what I meant!"

To which the proper answer (but one which the teacher seldom gives, out of the kindness of his heart) is, "Of course I knew what you meant—but not because of the way you said it; rather, in spite of the way you said it.  I understand only because I'm older and wiser, and have a great deal of experience in reading students' compositions.  What I'm trying to do in this class is to teach you to write so as to be understood even by those who aren't that much more experienced than you are, and who aren't already aware of what you're trying to say, and who may not be ready to make the many accommodations and imaginative leaps I'm ready to make in reading your work."  In virtually every case that Pinker examines, the utterance under examination can be understood, at least by a reader who's willing to meet the author at least half way.  But equally, in virtually every case the impropriety the prescriptivist has complained of is one that will prevent understanding on another occasion, or allow the reader to think he has understood when he has not.

Another fallacy that is a standard part of the descriptivist campaign against the prescriptivists is the Irrelevant Historical Fact, and Pinker does not neglect it.  Talking about disinterested, he reveals that in the eighteenth century it had just the meaning that today's prescriptivists object to: uninterested, apathetic.  And this he seems to regard as a decisive consideration, rather than the antiquarian curiosity that it really is.  Careful writers today use disinterested to mean "having no personal stake in the matter"; what it meant in the eighteenth century is neither here nor there.  But Pinker builds another fallacy atop this one; the fact that the word meant one thing then and another now leads him into the "language is always changing, so there's no point in resisting any particular change" fallacy.  And in support of this fallacy he thinks he can enlist no less an authority than Samuel Johnson.

His chapter on "The Language Mavens" ends with what Pinker must consider a masterstroke: a quotation from the preface to Johnson's Dictionary that he evidently thinks a thorough vindication of his own position, all the more imposing for being from a great writer, lexicographer, and moralist:

Those who have been persuaded to think well of my design, require that it should fix our language, and put a stop to those alterations which time and chance have hitherto been suffered to make in it without opposition.  With this consequence I will confess that I have flattered myself for a while; but now begin to fear that I have indulged expectations which neither reason nor experience can justify.  When we see men grow old and die at a certain time one after another, from century to century, we laugh at the elixir that promises to prolong life to a thousand years; and with equal justice may the lexicographer be derided, who being able to produce no example of a nation that has preserved their words and phrases from mutability, shall imagine that his dictionary can embalm his language, and secure it from corruption and decay, that it is in his power to change sublunary nature, and clear the world at once from folly, vanity, and affectation.  With this hope, however, academies have been instituted, to guard the avenues of their languages, to retain fugitives, and to repulse intruders; but their vigilance and activity have hitherto been vain; sounds are too volatile and subtle for legal restraints; to enchain syllables, and to lash the wind, are equally the undertakings of pride, unwilling to measure its desires by its strength.

Only a linguistic scientist, perhaps, could read this as meaning "Isn't it just great that language changes!  And aren't people who dream of stabilizing it foolish!"  What, one wonders, does Pinker make of Johnson's statement that he did originally hope that his Dictionary might have the effect of fixing the language?  And what of Johnson's using the aging and death of men as a parallel to the processes by which languages change?  And what of his using terms like folly, vanity, affectation, corruption, and decay in reviewing the causes and consequences of those processes?

So Pinker closes his campaign against the prescriptivists by quoting at length from perhaps the greatest and wisest of prescriptivists, and utterly mistaking the man.  And in thus demonstrating an inability to grasp the tenor of a great piece of rhetorical prose, he gives us further reason to distrust linguists when they try to legislate on matters beyond their professional competence, such as how language should be used.  Had he read a little further in Johnson's Preface, Pinker would have come across words that he could not have misunderstood:

...tongues, like governments, have a natural tendency to degeneration; we have long preserved our constitution, let us make some struggle for our language.