159 to ought, which is one of our defective verbs: by consequence, meaning consequently, is preferable to of consequence; as this expression is often employed to denote momentous or important. In the preposition toward and towards, and the adverbs forward and forwards, backward and backwards, the two forms are used indiscriminately. But as the first form in all these is also an adjective, it is better to confine the particles to the second. Custom, too, seems at present to lean this way. Besides and beside serve both as conjunctions and as prepositions. There appears some tendency at present to assign to each a separate province. This tendency ought to be humoured by employing only the former as the conjunction, the latter as the preposition.

This principle likewise leads me to prefer extemporary as an adjective to extemporé, which is properly an adverb, and ought, for the sake of precision, to be restrained to that use. It is only of late that this last term begins to be employed adjectively. Thus we say, with equal propriety, an extemporary prayer, an extemporary sermon, and, he prays extempore, he preaches extemporé. I know not how Dr. Priestley hath happened to mention the term extemporary, in a way which would make one think he considered it as a word peculiar to Mr. Hume. The word hath evidently been in good use for a longer time than one thinks of searching back in quest of authorities, and remains in good use to this day. By the same rule we ought to prefer scarcely as an adverb, to scarce, which is an adjective; and exceedingly, as an adverb, to exceeding, which is a participle. For the same reason also I am inclined to prefer that use, which makes ye invariably the nominative plural of the personal pronoun thou, and you the accusative, when applied to an actual plurality. When used for the singular number, custom hath determined that it shall be you in both cases. This renders the distinction rather more important, as for the most part it would shew directly, whether one or more were addressed; a point in which we are often liable to mistake in all modern languages. From the like principle, in those verbs which have for the participle passive both the preterite form and one peculiar, the peculiar form ought to have the preference. Thus, I have gotten, I have hidden, I have spoken, are better than I have got, I have hid, I have spoke. From the same principle I think ate is preferable in the preterite tense, and eaten in the participle, to eat, which is the constant form

These nearly correspond to the conjunction præterea, and the preposition præter in Latin.

Yet I should prefer, "I have held, helped, melted," to "I have holden, holpen, molten," these last participles being now obsolete. Holden is indeed still used when we speak formally of courts or public meetings.

of the present, though sometimes also used for both the others.

But though in this judgment concerning the participles, I agree entirely with all our approved modern grammarians, I can by no means concur with some of them in their manner of supporting it. "We should be immediately shocked," says one of the best of them", at I have knew, I have saw, I have gave, &c. but our ears are grown familiar with I have wrote, I have drank, I have bore, &c. which are altogether as barbarous." Nothing can be more inconsistent, in my opinion, with the very first principles of grammar, than what is here advanced. This ingenious gentleman surely will not pretend, that there is a barbarism in every word which serves for preterite and participle both, else the far greater part of the preterites and participles of our tongue are barbarous. If not, what renders many of them, such as loved, hated, sent, brought, good English when employed either way? I know no answer that can be given, but custom; that is, in other words, our ears are familiarised to them by frequent use. And what was ever meant by a barbarism in speech, but that which shocks us by violating the constant usage in speaking or in writing? If so, to be equally barbarous, and to be equally shocking are synonymous; whereas to be barbarous, and to be in familiar use, are a contradiction in terms. in this manner does our author often express himself. "No authority," says he in another place, "is sufficient to justify so manifest a solecism." No man needed less to be informed, that authority is every thing in language, and that it is the want of it alone that constitutes both the barbarism and the solecism.

Yet

Canon the second.

The second canon is, In doubtful cases regard ought to be had in our decisions to the analogy of the language.

For this reason I prefer contemporary to cotemporary. The general use in words compounded with the inseparable preposition con, is to retain the (n) before a consonant, and to expunge it before a vowel or an (h) mute. Thus we say condisciple, conjuncture, concomitant; but co-equal, coeternal, co-incide, co-heir. I know but one exception, which is co-partner. But in dubious cases we ought to follow the rule, and not the exception. If by the former canon the adverbs backwards and forwards are preferable to backward and forward; by this canon, from the principle of analogy, afterwards and homewards should be preferred to afterward

Lowth's Introduction to English Grammar.

and homeward. Of the two adverbs thereabout and thereabouts, compounded of the particle there and the preposition, the former alone is analogical, there being no such word in the language as abouts. The same holds of hereabout and whereabout. In the verbs to dare, and to need, many say, in the third person present singular, dare and need, as "he need not go; he dare not do it.” Others say, dares and needs. As the first usage is exceedingly irregular, hardly any thing less than uniform practice could authorise it. This rule supplies us with another reason for preferring scarcely and exceedingly as adverbs to scarce and exceeding. The phrases Would to God, and Would God, can both plead the authority of custom; but the latter is strictly analogical, the former is not. It is an established idiom in the English tongue, that any of the auxiliaries might, could, would, should, did, and had, with the nominative subjoined, should express sometimes a supposition, sometimes a wish: which of the two it expresses in any instance, is easily discovered from the context. Thus the expression, "Would he but ask it of me," denotes either, "If he would, or I wish that he would but ask it of me." Would God, then, is properly, I wish that God would, or, O that God would. The other expression it is impossible to reconcile to analogy in any way. For a like reason the phrase ever so, as when we say, "though he were ever so good," is preferable to never so. In both these decisions I subscribe to the judgment of Dr. Johnson. Of the two phrases in no wise in three words, and nowise in one, the last only is conformable to the present genius of the tongue. The noun wise, signifying manner, is quite obsolete. It remains now only in composition, in which, along with an adjective or other substantive, it forms an adverb or conjunction. Such are sidewise, lengthwise, coastwise, contrariwise, likewise, otherwise. These always preserve the compound form, and never admit a preposition; consequently nowise, which is an adverb of the same order, ought analogically to be written in one word, and not to be preceded by in. In every ancient style all these words were uncompounded, and had the preposition. They said in like wise, and in other wise. And even if custom at present were uniform, as it is

What has given rise to it is evidently the French Plút à Dieu, of the same import. But it has not been adverted to (so servile commonly are imitators), that the verb plaire is impersonal, and regularly construed with the preposition à ; neither of which is the case with the English will and would.

y In proof of this I shall produce a passage taken from the Prologue of the English translation of the Legenda Aurea, which seems to have been made towards the end of the fifteenth century. "I haue submysed my selfe to translate into Englysshe the legende of sayntes whyche is called legenda aurea in latyn; That is to saye, the golden legende. For in lyke wyse as golde is moost noble aboue all other metallys; in lyke wyse is thys legende holden moost noble aboue all other werkes."

M

divided, in admitting in before nowise, it ought to be followed, though anomalous. In these matters it is foolish to attempt to struggle against the stream. All that I here plead for is, that when custom varies, analogy should decide the question. In the determination of this particular instance I differ from Dr. Priestley. Sometimes whether is followed by no, sometimes by not. For instance, some would say, "Whether he will or no," others, "Whether he will or not." Of these it is the latter only that is analogical. There is an ellipsis of the verb in the last clause, which when you supply, you find it necessary to use the adverb not," Whether he will or will not." I shall only add, that by both the preceding canons we ought always to say rend in the present of the indicative and of the infinitive, and never rent, as is sometimes done. The latter term hath been pre-occupied by the preterite and the participle passive, besides that it is only in this application that it can be said to be used analogically. For this reason, the active participle ought always to be rending and not renting.

Canon the third.

The third canon is, When the terms or expressions are in other respects equal, that ought to be preferred which is most agreeable to the ear.

This rule hath perhaps a greater chance of being observed than any other, it having been the general bent for some time to avoid harsh sounds and unmusical periods. Of this we have many examples. Delicateness hath very properly given way to delicacy; and for a like reason authenticity will probably soon displace authenticalness, and vindictive dispossess vindicative altogether. Nay, a regard to sound hath, in some instances, had an influence on the public choice, to the prejudice of both the former canons, which one would think ought to be regarded as of more importance. Thus the term ingenuity hath obtained a preference to ingeniousness, though the former cannot be deduced analogically from ingenious, and had besides been pre-occupied, and consequently would be equivocal, being a regular derivative from the term ingenuous, if the newer acceptation had not before now supplanted the other altogether.

About the time that our present version of the Scriptures was made, the old usage was wearing out. The phrase in like wise occurs but once, (Matt. xxi. 24.) whereas the compound term likewise occurs frequently.-We find in several places, on this wise, in any wise, and in no wise. The two first phrases are now obsolete, and the third seems to be in the state which Dr. Johnson calls obsolescent.

Canon the fourth.

The fourth canon is, In cases wherein none of the foregoing rules gives either side a ground of preference, a regard to simplicity (in which I include etymology when manifest) ought to determine our choice.

Under the name simplicity, I must be understood to comprehend also brevity; for that expression is always the simplest which, with equal purity and perspicuity, is the briefest. We have, for instance, several active verbs which are used either with or without a preposition indiscriminately. Thus we say either accept or accept of, admit or admit of, approve or approve of; in like manner address or address to, attain or attain to. In such instances it will hold, I suppose, pretty generally, that the simple form is preferable. This appears particularly in the passive voice, in which every one must see the difference. "His present was accepted of by his friend."-"His excuse was admitted of by his master." "The magistrates were addressed to by the townsmen,' are evidently much worse than, "His present was accepted by his friend."—"His excuse was admitted by his master.""The magistrates were addressed by the townsmen." We have but too many of this awkward, disjointed sort of compounds, and therefore ought not to multiply them without necessity. Now if once the preposition should obtain in the active voice, the rules of syntax will absolutely require it in the passive. Sometimes indeed the verb hath two regimens, and then the preposition is necessary to one of them, as, "I address myself to my judges."-"They addressed their vows to Apollo." But of such cases I am not here speaking.

Both etymology and analogy, as well as euphony and simplicity, determine us in preferring subtract to substract, and consequently subtraction to substraction".

Canon the fifth.

The fifth and only other canon that occurs to me on the subject of divided use is, In the few cases wherein neither

Subtract is regularly deduced from the supine subtractum of the Latin verb subtraho, in the same way as act from actum, the supine of ago, and translate from translatum, the supine of transfero. But it would be quite unexampled to derive the English verb from the French soustraire. Besides, there is not another instance in the language of a word beginning with the Latin preposition sub, where the sub is followed by an s, unless when the original word compounded with the preposition, begins with an s. Thus we say subscribe from sub and scribo, subsist from sub and sisto, substitute from sub and statuo. But we cannot say substract from sub and straho, there being no such word. There can be no doubt, therefore, that a mistaken etymology, arising from an affinity to the French term, not in the verb, but in the verbal noun, has given rise to this harsh anomaly.