which are remarkably harsh and unharmonious, and not absolutely necessary, may justly be judged worthy of this fate. I call a word or phrase absolutely necessary when we have no synonymous words, in the event of a dismission, to supply its place, or no way of conveying properly the same idea without the aid of circumlocution. The rule, with this limitation, will, I believe, be generally assented to. The only difficulty is to fix the criteria by which we may discriminate the obnoxious words from all others.

It may well be reckoned that we have lighted on one criterion, when we have found a decompound or term composed of words already compounded, whereof the several parts are not easily, and, therefore, not closely united. Such are the words bare-faced-ness, shame-faced-ness, un-success-ful-ness, disinterest-ed-ness, wrong-head-ed-ness, tender-heart-ed-ness. They are so heavy and drawling, and, withal, so ill-compacted, that they have not more vivacity than a periphrasis to compensate for the defect of harmony.

Another criterion is, when a word is so formed and accented as to render it of difficult utterance to the speaker, and, consequently, disagreeable in sound to the hearer. This happens in two cases: first, when the syllables which immediately follow the accented syllable are so crowded with consonants as, of necessity, to retard the pronunciation. The words questionless, chroniclers, convénticlers, concupiscence, remémbrancer, are examples of this. The accent in all these is on the antepenultimate, for which reason the last two syllables ought to be pronounced quick; a thing scarcely practicable, on account of the number of consonants which occur in these syllables. The attempt to quicken the pronunciation, though familiar to Englishmen, exhibits to strangers the appearance of awkward hurry, instead of that easy fluency to be found in those words wherein the unaccented syllables are naturally short. Such are lévity, vánity, avídity, all accented in like manner on the antepenultimate. The second case in which a similar dissonance is found is when too many syllables follow the accented syllable; for, though these be naturally short, their number, if they exceed two, makes a disagreeable pronunciation. Examples of this are the words primarily, cursorily, súmmarily, perémptorily, perémptoriness, víndicative; all of which are accented on the fourth syllable from the end. It is to be wished that the use which now prevails in regard to the manner of accenting some words would alter, as we cannot afford to part with every term that is liable to exception in this respect. Nor is a change here to be despaired of, since we find it hath happened to several words already, as the places which they occupy in ancient poetry sufficiently evince.

A third criterion is when a short or unaccented syllable is

repeated, or followed by another short or unaccented syllable very much resembling. This always gives the appearance of stammering to the pronunciation. Such were the words hólily, fárriering, sillily. We have not many words chargeable with this fault; nay, so early have the people been sensible of the disagreeable sound occasioned by such recurrences, that it would appear they have added the adverbial termination to very few of our adjectives ending in ly. I believe there are no examples extant of heavenlily, godlily, timelily, dailily. Johnson hath given us in his Dictionary the word lowlily, which is as bad as any of them, but without quoting authorities. In these and such like, the simple forms, as heavenly, godly, timely, daily, homely, courtly, comely, seem always to have served both for adjective and adverb, though this too hath its inconvenience. It deserves our notice, that the repetition of a syllable is never offensive when either one or both are long, as in papa, mamma, murmur, tartar, barbarous, lily.

Besides the cases aforesaid, I know of none that ought to dispose us to the total disuse of words really significant. A little harshness by the collision of consonants, which, nevertheless, our organs find no difficulty in articulating, and which do not suggest to the hearer the disagreeable idea either of precipitation or of stammering, are by no means a sufficient reason for the suppression of a useful term. The monosyllables judg'd, drudg'd, grudg'd, which some have thought very offensive, appear not in the least exceptionable, compared with the words above mentioned. It would not do well to introduce such hard and strong sounds too frequently; but when they are used sparingly and properly, they have even a good effect. Variety in sound is advantageous to a language; and it is convenient that we should have some sounds that are rough and masculine, as well as some that are liquid and feminine.

. I observe this the rather, because I think there is at present a greater risk of going too far in refining than of not going far enough. The ears of some critics are immoderately delicate. A late essayist,* one who seems to possess a considerable share of ingenuity and taste, proposes the utter extirpation of encroach, encroachment, inculcate, purport, methinks, and some others, the precise meaning of which we have no single words in English that perfectly express. An ear so nice as to be hurt by these, appears to me in the same light as a stomach so squeamish as to nauseate our beef and beer, the ordinary food of the country. Such ears, I should say, are not adapted to our speech, nor such stomachs to our climate. This humour, were it to become general, would give * Sketches by Launcelot Temple, Esq., of late republished and owned by Dr. Armstrong.

a very unfavourable aspect to the language; and it might admit a question whether, on such principles, if an expurgation of the vocabulary were attempted, there would remain one third of the whole stock that would not be deemed worthy of excision. This would be particularly inconvenient, if everybody were as much an enemy as this gentleman seems to be to all new fashioned terms and phrases. We should hardly have words enough left for necessary purposes.*

CANON THE SEVENTH.

End

The second canon on this subject is, When etymology plainly points to a signification different from that which the word commonly bears, propriety and simplicity both require its dismission.

I use the word plainly, because, when the etymology is from an ancient or foreign language, or from obsolete roots in our own language, or when it is obscure or doubtful, no regard should be had to it. The case is different when the roots either are, or strongly appear to be, English, are in present use, and clearly suggest another meaning. Of this kind is the word beholden for obliged or indebted. It should

There

* I shall only observe here by the way, that those languages which are allowed to be the most susceptible of all the graces of harmony, have admitted many ill-sounding words. Such are in Greek σπλαγχνίζεσθαι, προσφθεγξασθαι, εγχριμφθεις, κεκακοκα, μεμιμημενον. In the last two one finds a dissonant recurrence of the same letter to a degree quite unexampled with us. is, however, such a mixture of long and short syllables, as prevents that difficulty of utterance which was remarked in some English words. Such are also, in Latin, dixisses, spississimus, percrebrescebantque. The last of these words is very rough, and the first two have as much of the hissing letters as any English word whatever. The Italian is considered, and I believe justly, as the most musical of all languages, yet there are in it some sounds which even to us, accustomed to a dialect boisterous like our weather, appear harsh and jarring. Such are incrocicchiare, sdruccioloso, spregiat rice. There is a great difference between words which sound harshly, but are of easy pronunciation to the natives, and those words which even to natives occasion difficulty in the utterance, and, consequently, convey some idea of awkwardness to the hearer, which is prejudicial to the design. There are, in the languages of all countries, many words which foreigners will find a difficulty in pronouncing that the natives have no conception of. The Greeks could not easily articulate the Latin terminations in ans and ens. On the other hand, there were many sounds in Greek which appeared intolerable to the Latins, such as words beginning with v, 40, 4, πT, KT, and many others. No people have so studiously avoided the collision of consonants as the Italians. To their delicate ears, pt, ct, and cs or x, though belonging to different syllables, and interposed between vowels, are offensive, nor can they easily pronounce them. Instead of apto, and lecto, and Alexandro, they must say atto, and letto, and Allessandro. Yet these very people begin some of their words with the three consonants sdr, which to our ears are perfectly shocking. It is not, therefore, so much harshness of sound as difficulty of utterance that should make some words be rejected altogether. The latter tends to divert our attention, and, consequently, to obstruct the effect. The former hath not this tendency, unless they be obtruded on us too frequently.

regularly be the passive participle of the verb to behold, which would convey a sense totally different. Not that I consider the term as equivocal, for in the last acceptation it hath long since been disused, having been supplanted by beheld. But the formation of the word is so analogical as to make it have at least the appearance of impropriety when used in a sense that seems naturally foreign to it. The word beholding, to express the same thing, is still more exceptionable than the other, and includes a real impropriety, being an active form with a passive signification. To vouchsafe, as denoting to condescend, is liable to a similar exception, and for that reason, more than for its harshness, may be dispensed with. Coaction and coactive, as signifying compulsion and compulsive, though regularly deduced from the Latin coactum, have so much the appearance of being compounded of the English words action and active, with the inseparable preposition co, which would give them a meaning quite different, that one can scarcely hear them without some tendency to mistake the sense. The verb to unloose should analogically signify to tie, in like manner as to untie signifies to loose. To what purpose is it, then, to retain a term, without any necessity, in a signification the reverse of that which its etymology manifestly suggests? In the same way, to annul and to disannul ought by analogy to be contraries, though irregularly used as synonymous. The verb to unravel, commonly, indeed, as well as analogically, signifies to disentangle, to extricate; sometimes, however, it is absurdly employed to denote the contrary, to disorder, to entangle, as in these lines in the address to the goddess of Dulness,

"Or quite unravel all the reasoning thread,

And hang some curious cobweb in its stead."*

All considerations of analogy, propriety, perspicuity, unite in persuading us to repudiate this preposterous application altogether.

CANON THE EIGHTH.

The third canon is, When any words become obsolete, or, at least, are never used, except as constituting part of particular phrases, it is better to dispense with their service entirely, and give up the phrases.

The reasons are, first, because the disuse in ordinary cases renders the term somewhat indefinite, and occasions a degree of obscurity; secondly, because the introduction of words which never appear but with the same attendants, gives the style an air of vulgarity and cant. Examples of this we have in the words lief, dint, whit, moot, pro, and con, as, "I had as lief go myself,” for “ I should like as well to go

66

* Dunciad, b. i.

66

myself." "He convinced his antagonist by dint of argument," that is, "by strength of argument." "He made them yield by dint of arms”—“" by force of arms." "He is not a whit better"-" no better." "The case you mention is a moot point" -"a disputable point." "The question was strenuously debated pro and con"-" on both sides."

CANON THE NINTH.

The fourth and last canon I propose is, All those phrases which, when analyzed grammatically, include a solecism, and all those to which use hath affixed a particular sense, but which, when explained by the general and established rules of the language, are susceptible either of a different sense or of no sense, ought to be discarded altogether.

It is this kind of phraseology which is distinguished by the epithet idiomatical, and hath been originally the spawn partly rodu ter of ignorance and partly of affectation. Of the first sort, which includes a solecism, is the phrase, "I had rather do such a thing," for "I would rather do it." The auxiliary had, joined to the infinitive active do, is a gross violation of the rules of conjugation in our language, and though good use may be considered as protecting this expression from being branded with the name of a blunder, yet, as it is both irregular and unnecessary, I can foresee no inconvenience that will arise from dropping it. I have seen this idiom criticised in some essay, whose name I cannot now remember, and its origin very naturally accounted for, by supposing it to have sprung from the contraction I'd, which supplies the place both of I had and of I would, and which had been at first ignorantly resolved into I had when it ought to have been I would. The phrase, thus frequently mistaken, hath come at length to establish itself and to stand on its own foot.*

Of the second sort, which, when explained grammatically, leads to a different sense from what the words in conjunction commonly bear, is, " He sings a good song," for "he sings well." The plain meaning of the words as they stand connected is very different, for who sees not that a good song may be ill sung? Of the same stamp is, " He plays a good fiddle," for "he plays well on the fiddle." This seems also

* Whether, with Johnson and Lowth, we should consider the phrases by this means, by that means, it is a means, as liable to the same exception, is perhaps more doubtful. Priestley considers the word means as of both numbers, and of such nouns we have several examples in the language. But it may be objected, that as the singular form mean is still frequently to be met with, this must inevitably give to the above phrases an appearance of solecism in the judgment of those who are accustomed to attend to the rules of syntax. But, however this may induce such critics to avoid the expression in question, no person of taste, I presume, will venture so far to violate the present usage, and, consequently, to shock the ears of the gencrality of readers, as to say "By this mean" or "By that mean."