by Bill Bryson
Nothing illustrates the scope for prejudice in English better than the issue of the split infinitive. Some people feel ridiculously strongly about it. When the British Conservative politician Jock Bruce-Gardyne was economic secretary to the Treasury in the early 1980s, he returned unread any departmental correspondence containing a split infinitive. (It should perhaps be pointed out that a split infinitive is one in which an adverb comes between to and a verb, as in to quickly look.) I can think of two very good reasons for not splitting an infinitive.
1. Because you feel that the rules of English ought to conform to the grammatical precepts of a language that died a thousand years ago.
2. Because you wish to cling to a pointless affectation of usage that is without the support of any recognized authority of the last 200 years, even at the cost of composing sentences that are ambiguous, inelegant, and patently contorted.
It is exceedingly difficult to find any authority who condemns the split infinitive—Theodore Bernstein, H. W. Fowler, Ernest Gowers, Eric Partridge, Rudolph Flesch, Wilson Follett, Roy H. Copperud, and others too tedious to enumerate here all agree that there is no logical reason not to split an infinitive. Otto Jespersen even suggests that, strictly speaking, it isn’t actually possible to split an infinitive. As he puts it: “ ‘To’ . . . is no more an essential part of an infinitive than the definite article is an essential part of a nominative, and no one would think of calling ‘the good man’ a split nominative” [Growth and Structure of the English Language, page 222].
Lacking an academy as we do, we might expect dictionaries to take up the banner of defenders of the language, but in recent years they have increasingly shied away from the role. A perennial argument with dictionary makers is whether they should be prescriptive (that is, whether they should prescribe how language should be used) or descriptive (that is, merely describe how it is used without taking a position). The most notorious example of the descriptive school was the 1961 Webster’s Third New International Dictionary (popularly called Webster’s Unabridged), whose editor, Philip Gove, believed that distinctions of usage were elitist and artificial. As a result, usages such as imply as a synonym for infer and flout being used in the sense of flaunt were included without comment. The dictionary provoked further antagonism, particularly among members of the U.S. Trademark Association, by refusing to capitalize trademarked words. But what really excited outrage was its remarkable contention that ain’t was “used orally in most parts of the U.S. by many cultivated speakers.”
So disgusted was The New York Times with the new dictionary that it announced it would not use it but would continue with the 1934 edition, prompting the language authority Bergen Evans to write: “Anyone who solemnly announces in the year 1962 that he will be guided in matters of English usage by a dictionary published in 1934 is talking ignorant and pretentious nonsense,” and he pointed out that the issue of the Times announcing the decision contained nineteen words condemned by the Second International.
Since then, other dictionaries have been divided on the matter. The American Heritage Dictionary, first published in 1969, instituted a usage panel of distinguished commentators to rule on contentious points of usage, which are discussed, often at some length, in the text. But others have been more equivocal (or prudent or spineless depending on how you view it). The revised Random House Dictionary of the English Language, published in 1987, accepts the looser meaning for most words, though often noting that the newer usage is frowned on “by many”—a curiously timid approach that at once acknowledges the existence of expert opinion and yet constantly places it at a distance. Among the looser meanings it accepts are disinterested to mean uninterested and infer to mean imply. It even accepts the existence of kudo as a singular—prompting a reviewer from Time magazine to ask if one instance of pathos should now be a patho.
It’s a fine issue. One of the undoubted virtues of English is that it is a fluid and democratic language in which meanings shift and change in response to the pressures of common usage rather than the dictates of committees. It is a natural process that has been going on for centuries. To interfere with that process is arguably both arrogant and futile, since clearly the weight of usage will push new meanings into currency no matter how many authorities hurl themselves into the path of change.
But at the same time, it seems to me, there is a case for resisting change—at least slapdash change. Even the most liberal descriptivist would accept that there must be some conventions of usage. We must agree to spell cat c-a-t and not e-l-e-p-h-a-n-t, and we must agree that by that word we mean a small furry quadruped that goes meow and sits comfortably on one’s lap and not a large lumbering beast that grows tusks and is exceedingly difficult to housebreak. In precisely the same way, clarity is generally better served if we agree to observe a distinction between imply and infer, forego and forgo, fortuitous and fortunate, uninterested and disinterested, and many others. As John Ciardi observed, resistance may in the end prove futile, but at least it tests the changes and makes them prove their worth.
Perhaps for our last words on the subject of usage we should turn to the last words of the venerable French grammarian Dominique Bonhours, who proved on his deathbed that a grammarian’s work is never done when he turned to those gathered loyally around him and whispered: “I am about to—or I am going to—die; either expression is used.”
10.
Order out of Chaos
How big is the English language? That’s not an easy question. Samuel Johnson’s dictionary contained 43,000 words. The unabridged Random House of 1987 has 315,000. Webster’s Third New International of 1961 contains 450,000. And the revised Oxford English Dictionary of 1989 has 615,000 entries. But in fact this only begins to hint at the total.
For one thing, meanings in English are much more various than a bald count of entry words would indicate. The mouse that scurries across your kitchen floor and the mouse that activates your personal computer clearly are two quite separate entities. Shouldn’t they then be counted as two words? And then what about related forms like mousy, mouselike, and mice? Shouldn’t they also count as separate words? Surely there is a large difference between something that is a mouse and something that is merely mousy.
And then of course there are all the names of flora and fauna, medical conditions, chemical substances,* laws of physics, and all the other scientific and technical terms that don’t make it into ordinary dictionaries. Of insects alone, there are 1.4 million named species. Total all these together and you have—well, no one knows. But certainly not less than three million.
So how many of these words do we know? Again, there is no simple answer. Many scholars have taken the trouble (or more probably compelled their graduate students to take the trouble) of counting the number of words used by various authors, on the assumption, one supposes, that that tells us something about human vocabulary. Mostly what it tells us is that academics aren’t very good at counting. Shakespeare, according to Pei and McCrum, had a vocabulary of 30,000 words, though Pei acknowledges seeing estimates putting the figure as low as 16,000. Lincoln Barnett puts it at 20,000 to 25,000. But most other authorities—Shipley, Baugh and Cable, Howard—put the number at a reassuringly precise 17,677. The King James Bible, according to Laird, contains 8,000 words, but Shipley puts the number at 7,000, while Barnett confidently zeroes in on a figure of 10,442. Who knows who’s right?
One glaring problem with even the most scrupulous tabulation is that the total number of words used by an author doesn’t begin to tell us the true size of his vocabulary. I know the meanings of frangible, spiffing, and cutesy-poo, but have never had occasion to write them before now. A man of Shakespeare’s linguistic versatility must have possessed thousands of words that he never used because he didn’t like or require them. Not once in his plays can you find the words Bible, Trinity, or Holy Ghost, and yet that is not to suggest that he was not familiar with them.
Estimates of the size of the average person’s vocabula
ry are even more contentious. Max Müller, a leading German philologist at the turn of the century, thought the average farm laborer had an everyday vocabulary of no more than 300 words. Pei cites an English study of fruit pickers, which put the number at no more than 500, though he himself thought that the figure was probably closer to 30,000. Stuart Berg Flexner, the noted American lexicographer, suggests that the average well-read person has a vocabulary of about 20,000 words and probably uses about 1,500 to 2,000 in a normal week’s conversations. McCrum puts an educated person’s vocabulary at about 15,000.
There are endless difficulties attached to adjudging how many words a person knows. Consider just one. If I ask you what incongruent means and you say, “It means not congruent,” you are correct. That is the first definition given in most dictionaries, but that isn’t to say that you have the faintest idea what the word means. Every page of the dictionary contains words we may not have encountered before—inflationist, forbiddance, moosewood, pulsative—and yet whose meanings we could very probably guess.
At the same time there are many words that we use every day and clearly know and yet might have difficulty proving. How would you define the or what or am or very? Imagine trying to explain to a Martian in a concise way just what is is. And then what about all those words with a variety of meanings? Take step. The American Heritage Dictionary lists a dozen common meanings for the word, ranging from the act of putting one foot in front of the other to the name for part of a staircase. We all know all these meanings, yet if I gave you a pencil and a blank sheet of paper could you list them? Almost certainly not. The simple fact is that it is hard to remember what we remember, so to speak. Put another way, our memory is a highly fickle thing. Dr. Alan Baddeley, a British authority on memory, cites a study in which people were asked to name the capital cities of several countries. Most had trouble with the capitals of countries like Uruguay and Bulgaria, but when they were told the initial letter of the capital city, they often suddenly remembered and their success rate soared. In another study people were shown long lists of random words and then asked to write down as many of them as they could remember. A few hours later, without being shown the list again, they were asked to write down as many of the words as they could remember then. Almost always the number of words would be nearly identical, but the actual words recalled from one test to another would vary by 50 percent or more. In other words, there is vastly more verbal information locked away in our craniums than we can get out at any one time. So the problem of trying to assess accurately just how much verbal material we possess in total is fraught with difficulties.
For this reason educational psychologists have tended to shy away from such studies, and such information as exists is often decades old. One of the most famous studies was conducted in 1940. In it, two American researchers, R. H. Seashore and L. D. Eckerson, selected a random word from each left-hand page of a Funk & Wagnalls standard desktop dictionary and asked a sampling of college students to define those words or use them in a sentence. By extrapolating those results onto the number of entries in the dictionary, they concluded that the average student had a vocabulary of about 150,000 words—obviously very much larger than previously supposed. A similar study carried out by K. C. Diller in 1978, cited by Aitchison in Words in the Mind, put the vocabulary level even higher—at about 250,000 words. On the other hand, Jespersen cites the case of a certain Professor E. S. Holden who early in the century laboriously tested himself on every single word in Webster’s Dictionary and arrived at a total of just 33,456 known words. It is clearly unlikely that a university professor’s vocabulary would be four to six times smaller than that of the average student. So such studies would seem to tell us more about the difficulties of framing tests than about the size of our vocabularies.
What is certain is that the number of words we use is very much smaller than the number of words we know. In 1923 a lexicographer named G. H. McKnight did a comprehensive study of how words are used and found that just forty-three words account for fully half of all the words in common use, and that just nine account for fully one-quarter of all the words in almost any sample of written English. Those nine are: and, be, have, it, of, the, to, will, and you.
By virtue of their brevity, dictionary definitions often fail to convey the nuances of English. Rank and rancid mean roughly the same thing, but, as Aitchison notes, we would never talk about eating rank butter or wearing rancid socks. A dictionary will tell you that tall and high mean much the same thing, but it won’t explain to you that while you can apply either term to a building you can apply only tall to a person. On the strength of dictionary definitions alone a foreign visitor to your home could be excused for telling you that you have an abnormal child, that your wife’s cooking is exceedingly odorous, and that your speech at a recent sales conference was laughable, and intend nothing but the warmest praise.
The fact is that the real meanings are often far more complex than the simple dictionary definitions would lead us to suppose. In 1985, the department of English at the University of Birmingham in England ran a computer analysis of words as they are actually used in English and came up with some surprising results. The primary dictionary meaning of words was often far adrift from the sense in which they were actually used. Keep, for instance, is usually defined as to retain, but in fact the word is much more often employed in the sense of continuing, as in “keep cool” and “keep smiling.” See is only rarely required in the sense of utilizing one’s eyes, but much more often used to express the idea of knowing, as in “I see what you mean.” Give, even more interestingly, is most often used, to quote the researchers, as “mere verbal padding,” as in “give it a look” or “give a report” [London Sunday Times, March 31, 1985].
In short, dictionaries may be said to contain a certain number of definitions, but the true number of meanings contained in those definitions will always be much higher. As the lexicographer J. Ayto put it: “The world’s largest data bank of examples in context is dwarfed by the collection we all carry around subconsciously in our heads.”
English is changing all the time and at an increasingly dizzy pace. At the turn of the century words were being added at the rate of about 1,000 a year. Now, according to a report in The New York Times [April 3, 1989], the increase is closer to 15,000 to 20,000 a year. In 1987, when Random House produced the second edition of its masterly twelve-pound unabridged dictionary, it included over 50,000 words that had not existed twenty-one years earlier and 75,000 new definitions of old words. Of its 315,000 entries, 210,000 had to be revised. That is a phenomenal amount of change in just two decades. The new entries included preppy, quark, flexitime, chairperson, sunblocker, and the names of 800 foods that had not existed or been generally heard of in 1966—tofu, piña colada, chapati, sushi, and even crêpes.
Unabridged dictionaries have about them a stern, immutable air, as if here the language has been captured once and for all, and yet from the day of publication they are inescapably out of date. Samuel Johnson recognized this when he wrote: “No dictionary of a living tongue can ever be perfect, since while it is hastening to publication, some words are budding, and some are fading away.” That, however, has never stopped anyone from trying, not least Johnson himself.
The English-speaking world has the finest dictionaries, a somewhat curious fact when you consider that we have never formalized the business of compiling them. From the seventeenth century when Cardinal Richelieu founded the Académie Française, dictionary making has been earnest work indeed. In the English-speaking world, the early dictionaries were almost always the work of one man rather than a ponderous committee of academics, as was the pattern on the Continent. In a kind of instinctive recognition of the mongrel, independent, idiosyncratic genius of the English tongue, these dictionaries were often entrusted to people bearing those very characteristics themselves. Nowhere was this more gloriously true than in the person of the greatest lexicographer of them all, Samuel Johnson.
Johnson,
who lived from 1709 to 1784, was an odd candidate for genius. Blind in one eye, corpulent, incompletely educated, by all accounts coarse in manner, he was an obscure scribbler from an impoverished provincial background when he was given a contract by the London publisher Robert Dodsley to compile a dictionary of English.
Johnson’s was by no means the first dictionary in English. From Cawdrey’s Table Alphabeticall in 1604 to his opus a century and a half later there were at least a dozen popular dictionaries, though many of these were either highly specialized or slight (Cawdrey’s Table Alphabeticall contained just 3,000 words and ran to barely a hundred pages). Many also had little claim to scholarship. Cawdrey’s, for all the credit it gets as the first dictionary, was a fairly sloppy enterprise. It gave the definition of aberration twice and failed to alphabetize correctly on other words.
The first dictionary to aim for anything like comprehensiveness was the Universal Etymological Dictionary by Nathaniel Bailey, published in 1721, which anticipated Johnson’s classic volume by thirty-four years and actually defined more words. So why is it that Johnson’s dictionary is the one we remember? That’s harder to answer than you might think.
His dictionary was full of shortcomings. He allowed many spelling inconsistencies to be perpetuated—deceit but receipt, deign but disdain, hark but hearken, convey but inveigh, moveable but immovable. He wrote downhil with one l, but uphill with two; install with two l’s, but reinstal with one; fancy with an f, but phantom with a ph. Generally he was aware of these inconsistencies, but felt that in many cases the inconsistent spellings were already too well established to tamper with. He did try to make spelling somewhat more sensible, institutionalizing the differences between flower and flour and between metal and mettle—but essentially he saw his job as recording English spelling as it stood in his day, not changing it. This was in sharp contrast to the attitude taken by the revisers of the Académie Française dictionary a decade or so later, who would revise almost a quarter of French spellings.