and find all its peculiarities in every State of the Union, and many of them in that national literature which we are promised.

What we are to expect, and if possible avert, is the corruption of the literary language by every-day use on the tongues of a hundred millions of people scattered over half the surface of the globe, and mostly knowing a large part of their words at first only from books or newspapers.*

English orthoëpy is already subjected to severe pressure from the democracy, and it is doubtful whether the etymological canons will not have to yield to analogy established as a universal rule. So many people know how to read without knowing how to pronounce, our orthoëpical principles are so entangling to moderate intelligence, and the ignorant or half-educated may exercise such a constant pressure upon the educated few, that we are disposed to prophesy the victory of the people whenever they agree in opposing the scholars. If you wish to say roman'ce or demon'strate, you must daily fortify yourself against the influence of ro'mance and demonstrate hurtling into your ears from the popular tongue. Especially when words of foreign origin become an instrument of daily life, the pronunciation founded upon analogy must prevail. Finance will be accented on the first syllable by ninety-nine in every hundred Americans in spite of all the dictionaries and scholars, and before they are aware the scholars will be heard imitating the people.

It is worth considering whether these new popular conditions do not demand of scholars some effort to render our orthoëpy more simple, to come at first gracefully to concessions of things. which will in the end be won even over their most stubborn resistance. There is no statute of limitations beyond which a word cannot wear a foreign accent. On one side, it parts with its brogue the next day after it gets into the newspapers; on the other, it keeps up a pretense of foreign flavor for an indefinite period. Since we have no rule ourselves, why not adopt the popular one?

In most cases the foreign air is sadly parodied in our speech. What Frenchman would recognize our finan'ce? It is neither French nor English; we have hung it up in an orthoëpical limbo,

*We refer here to the whole Anglican-tongued population of the world.

and will never have the courage to take it down. The people have naturalized it, and will force us to accept it as a citizen of the tongue. It is difficult to see why one of these changes from foreign to English pronunciation is any more safe or proper for extending through an indefinite period, and being accomplished only after a generation of scholars have wasted their protoplasm in nervousness and indignation at popular stupidity.

We have not yet fully comprehended that we are in a new world, which is not, and cannot be, governed by the few best speakers; certainly not upon the old system. Pure models in actual speech and the dictionary will both fail to reach the multitude, educated in the common schools, rushing into mercantile and political life, and becoming in turn models for admiring friends and constituencies. How few of them will ever doubt the application of the principle of analogy, or, if they doubt, will have or take the time to consult a dictionary. Orthography is in the hands of one or two hundred publishers, and these can successfully resist the ravages of democracy in the written forms; but the phonetic forms are under no such vassalage to any aristocracy.

The most that is now possible is to call attention to the influence of popular pronunciation, the difficulties of maintaining a heterogeneous orthoëpy, and the apparent necessity of modifying our canons, for accent at least, so as to give a wider application to the law of analogy, especially for new or recently-introduced foreign words.

A larger field is presented by the play of the laws of phonetic decay and renewal in the English spoken in America.

There is some vagueness in the use of the term phonetic decay, resulting from its application to three quite distinct linguistic phenomena. If all change is decay, the resolution of a consonant into its elements, or, more strictly, its separation into two or more letters, may be called decay; but whether it be a phonetic change is not absolutely settled, though it is probably such. In the Greek dis, and the Latin bis, (both from the Sanscrit dvis,) we probably have a labial and a dental developed from an older and more vague, or more complex, sound.* In the same way, a few early-formed consonants have furnished

*See Garnett's Philological Essays, p. 241.

by their expansion and division the consonantal wealth of modern tongues. This species of growth is so far removed from any proper decay that no case of it ought ever to be so designated.

Another set of phenomena involve local phonetic loss without affecting the sound-volume of the language. When rondus becomes round, local decay carries off the syllable us; but the sounds thus dropped from a word are retained in the speech, and local loss in number is attended with an increment in force. Scholars have attached great importance to these local losses, and not without reason, for they are the pivot of growth and decay in language. What they have sometimes failed to do is to mark the distinction between phonetic change affecting the sound-shell, and change in the intelligent contents of this shell. Max Müller's use of the French adverbial particle ment (Latin, mente) shows such a confusion of widely different things, the change noticed having but slightly affected the phonesis of the word in French, and not at all in Latin and Spanish. What has occurred is a change of meaning and use, arising from forgetfulness of the original meaning.*

There is still another set of facts, of wider application and more difficult of treatment, relating to the loss and renewal of the sounds of a language. Changes beginning in the attenuation of sounds are carried forward by the unconscious speakers until the phonetic elements concerned are pushed altogether out of the living tongue. These are true losses of phonetic wealth, even though they extend no further than the subtilizing of our vocal implements.

Unlettered dialects are probably protected by their poverty from being robbed by their speakers, who are believed to expend more energy upon their narrow phonesis than is used by cultivated people upon the same elements. The ear is so easily deceived in this kind of induction that one might challenge the claim for barbaric phonetic force if it did not seem to rest on well-supported principles. Since only so much strength will

*The attractive style of Max Müller disguises the logical blunder which runs all through his treatment of this subject. Phonetic decay riots in unwritten dialects, but is retarded or totally arrested in literary languages. Just the reverse statement is the key of Müller's argument. See "Science of Language," first series, pp. 54-79. All the changes referred to occurred in the dialectic stage.

be used for speaking, it is reasonable to assert that this force expended on a few sounds will give them greater volume than can be obtained when the same force is spread over a larger number. Besides, the barbarian or provincial peasant is believed to employ more energy in his vocal exertions than is usual in the elocution of his educated neighbors. It is to be desired that linguistic acoustics were enough studied, and sufficiently fortified by trustworthy observations, to justify implicit confidence in these general principles, for upon them rests the theory that dialects regenerate the phonesis of cultivated languages. That the ruder dialects renew the waste of words in the more advanced, but still unwritten ones, is established upon a very wide induction; but whether a quite modern form of waste, resulting from the preponderance of reading over speaking, and from other causes, is compensated by dialectic additions to vocabularies, and dialectic practice by their speakers, is a new question, dependent for its solution upon the relative strength, fullness, and volume of barbaric and provincial enunciation.

The question for this discussion is, Do we find reasons for believing that our spoken English is undergoing a process of gradual attenuation of its phonetic elements, and how far is this decay compensated by regenerating influences?

It may be that the opposing forces keep matters in even balance, but ours will not for that reason be a profitless inquiry. If there be any pleasure in witnessing a contest it is when the combatants are well-matched, and, if we are concerned for either contestant, we shall probably seek to lend him some assistance. Tendencies in language are often arrested by attracting public attention to what is going on.

Some general statements of facts neglected by our grammars will help us to estimate the problem. The first in order is the deficiency of our written notation. Our twelve English vowels are rather clusters of sounds than atomic elements. We see this in the a group, concerning which it has been debated whether there are four, five, six, or seven versions, and it is pretty clear that the evidence for more than four would make a good case for seventy while vindicating seven. There are two phonetic phenomena at the root of these diverging theories, one of which is that every new consonantal accompani

ment slightly modifies the sound of the vowel, and the other is what the astronomers call the personal equation.

The same principles involve consonantal phonesis in a perpetual diversity. New combinations modify the several elements, and each man has his special rendering of all sounds, just as in a rural congregation every man may have a private notation for Mear or Dundee.* Of course the range of diversity is very narrow, and seldom passes the line beyond which it would be observed. It is only insisted that the variation exists, and establishes instability in what may be figuratively I called the molecular constitution of our words. Dr. Latham speculates upon the fate of sounds which have become unstable; what destiny awaits a language all of whose phonal factors are already affected with instability? One thing seems clear, that is to say, that the slighter, feebler, more subtile forms may readily take precedence, and even monopolize the speech, if there be in operation causes tending to the diminution of the volume of sounds. The way is open for change; no implacable lines of law hem in and protect the phonetic atoms; each sound has already a great growth of resembling forms, and natural or artificial selection must do the rest. That such a result might occur, or rather that some change is to be expected, is sufficiently shown by the expansion of three or seven consonants into twenty or thirty in linguistic growth; and that a selection may take place in the elements of a sound really compound, is proved by the case of bis and dis, from dvis, already cited.

It is rather suggested than asserted that another general principle may be applied to our phonesis with valuable results. We know that barbarians usually have few sounds. The Polynesian dialects are given only seven or eight by the grammars, and so long as the people remain in a low intellectual condition we should not expect them to develop new consonants; but, looking at the history of language, we should expect civilization to enlarge their phonesis. Well, then, does culture lead people to discriminate in sounds? If it does induce division

* Instances of variation which are conceded may be cited. The Greek spiritus lenis is an aspirate generally omitted in notation. It occurs in ache, in sounding which a bubble of breath precedes the a. Our w and y in wet and yet are ooet and iet, to which, when sounded rapidly, an aspirate is added. This slight aspirate we mark; but in union the io takes on a stronger aspiration, which does not appear in our notation.