which we use to communicate thought and emotion to we know not whom nor how many, we must carefully employ only such forms as good use, in its broadest sense, sanctions.

We are now in a position to answer the question we asked ourselves a little while ago. Why is it, we asked, that a certain number of apparently arbitrary black marks on white pages should convey to us all the infinitely varied impressions — intellectual, emotional, æsthetic that we find in literature? Why is it that style-whose visible body is never anything more or less than these black marks should impress us primarily as something that possesses or lacks Clearness, Force, and Elegance? Simply and solely because the tacit agreement, the good use of many generations of human beings, who at least linguistically are our ancestors, has consented in the first place that certain articulate sounds shall be fixed as symbols for certain distinct ideas, and in the second place that certain arbitrary marks shall be fixed as symbols for certain distinct sounds. Good use, and good use alone, is the basis on which all style rests. A knowledge of good use so familiar as to be practically instinctive is the basis on which any writer who would be certain to write with clearness, force, and elegance must ultimately rest his own style. The limits of good use are wide and flexible; but finally they grow rigid. Whoever strays beyond them errs; whoever keeps within them may write, for various reasons, ineffec tively, but cannot be convicted of positive error.

Every question of positive right and wrong in style is a question concerning nothing whatever but good

use.

Good use, then, must be the basis of all good style. The next thing to ask ourselves is how to recognize good use. And here we are met by a fact that, more than any other I know of, confuses most people who begin seriously to consider the matter in question. For various reasons, the chief of which is that five centuries ago pretty much everything worth reading was comprised in what survived of the literatures of Greece and Rome, the education of civilized Europeans and Americans is still based on a prolonged and not always very fruitful study of classical Latin and Greek. Now, what makes Latin or Greek letters stand for Latin or Greek words, and what makes Latin or Greek words stand for the thoughts and emotions which are not only Latin and Greek, but broadly human too, is precisely what makes English letters and words stand for the thoughts and emotions that make up our conscious lives; namely, that many thousands of human beings tacitly agreed what this double system of symbols should symbolize, and so that good use arose. But between the classical languages, which we call dead, and the modern languages, whose life is more vigorous than the life of any human being, there is a broad distinction, not very often kept in mind. Good use, like all other vital things, not only comes into being and flourishes, but it passes out of being too; and Latin use and

Greek passed out of being with the nations whose political and intellectual lives they expressed. So completely are they things of the past, indeed, that so far as I can learn from friends who have given their lives to the classics, nobody to-day on earth has any real knowledge of how Latin or Greek was pronounced. At Harvard College, and elsewhere, to be sure, they have supplanted the unquestionably barbarous English pronunciation by one which they call probably ancient; but whether Pericles or Cicero could understand the most punctiliously learned nineteenth-century professor is a question not to be settled this side of Elysium. In short, though we know pretty accurately what words classical letters symbolize, and what thoughts and emotions are symbolized by classical words, one part of the classical languages — the sound, the thing that made them true languages or tongues is as dead as Alexander or Cæsar. And along with the sound has perished the vital principle of the languages, - the constantly changing use which brought them from the rude jargons in which they began into the exquisitely finished forms in which their literatures preserve them. In other words, the classical languages, like other things that have passed out of this world, are complete. Nothing but the occasional discovery of a manuscript or an inscription can add a syllable to them; nothing but the demonstration of a corruption or a forgery can take a syllable away. Nothing, in all human probability, can supply the place of that troublesome caret which used to

bother us so much in the old Latin grammars. Here lies the distinction between the classical languages and the modern, the dead and the living. Latin and Greek are complete; dictionaries and grammars can codify them with final authority. English, on the other hand, like every living tongue, must remain incomplete so long as it retains life enough to be spoken and written by living men; and so dictionaries and grammars can at most be mile-stones in its progress through this world.

Now, of course the unlearned in matters of style look for authority to the learned. And the learned, brought up from childhood on the authority, in matters of classical style, of Latin and Greek dictionaries and grammars, are accustomed to display what little human frailty survives the process of culture by attaching to dictionaries and grammars themselves an importance second only to that which good men attach to Holy Writ. They do not stop to remember, or at all events to remind us, that what makes Latin and Greek books of reference so finally authoritative is not that they are books of reference, but that the languages therein codified have long since ceased to grow; and so that these tongues can be codified with something which approaches perfection.

To be certain of what good use is in a living language, then, we must have other things to rest our case on than the fact that some maker of dictionaries or grammars has registered and given chapter

and verse for the words or phrases we would defend. There are other tests of good use to which we must turn. The most notable, I think, are that it must be Reputable, National, and Present, Reputable as distinguished from vulgar, slangy, eccentric ; National as distinguished from local or technical; Present as distinguished from obsolete or transient. In view of the fact that every question of right or wrong in style must ultimately be referred to good use, these three phases of good use are worth separate attention.

Reputable use is the use of no single writer, however eminent; it is the common consent of the great body of writers whose works, taken together, make up what we mean when we seriously use the term English Literature, a term which of course includes any literature written in the English language, Scotch, Irish, American, Australian. The fact that Shakspere uses a word, or Sir Walter Scott, or Burke, or Washington Irving, or whoever happens to be writing earnestly in Melbourne or Sidney, does not make it reputable. The fact that all five of these authorities use the word in the same sense would go very far to establish the usage. On the other hand, the fact that any number of newspaper reporters agree in usage does not make the usage reputable. The style of newspaper reporters is not without merit; it is very rarely unreadable; but for all its virtue it is rarely a well of English undefiled. And just here, I may say, lies perhaps the most crying fault of con