leave his father, his father would die."* The word his father are in this short verse thrice repeated, and yet are not disagreeable, as they contribute to perspicuity. Had the last part of the sentence run thus, "If he should leave his father, he would die," it would not have appeared from the expression, whether it was the child or the parent that would die. Some have imagined, that the pronoun ought always regularly to refer to the nearest preceding noun of the same gender and number. But this notion is founded in a mistake, and doth not suit the idiom of any language, ancient or modern. From the rank that some words maintain in the sentence, if I may be allowed that expression, a reader will have a natural tendency to consider the pronoun as referring to them, without regard to their situation. In support of this observation, I shall produce two examples. The first shall be of the neuter singular of the third personal pronoun: "But I shall leave this subject to your management, and question not but you will throw it into such lights, as shall at once improve and entertain your reader." There is no ambiguity here, nor would it, on the most cursory reading, enter into the head of any person of common sense, that the pronoun it relates to management, which is nearer, and not to subject, which is more remote. Nor is it the sense only that directs us in this preference. There is another principle by which we are influenced. The accusative of the active verb is one chief object of attention in a sentence; the regimen of that accusative hath but a secondary value; it is regarded only as explanatory of the former, or at most as an appendage to it. This consideration doth not affect those only who understand grammar, but all who understand the language. The different parts of speech, through the power of custom, produce their effect on those who are ignorant of their very names, as much as on the grammarian himself: though it is the grammarian alone who can give a rational account of these effects. The other example I promised to give shall be of the masculine of the same number and person, in the noted complaint of Cardinal Wolsey immediately after his disgrace:

Had I but serv'd my God, with half the zeal
I serv'd my king; he would not in mine age
Have left me naked to mine enemies.‡

Here though the word king is adjoining, and the word God at some distance, the pronoun he cannot so regularly refer to that noun as to this. The reason is, the whole of the second clause beginning with these words, "with half the zeal," maintains but a subordinate rank in the sentence, as it is introduced in explication of the first, and might be omitted, not indeed without impairing, but without destroying, the sense. Yet neither the rank in the sentence, nor the nearness of position, will invariably determine the import of the relative. Sometimes, indeed, as was observed by the French author last quoted, the sense of the words connected is sufficient to remove the ambiguity, though the reader should have no previous knowledge of the

*Gen. xliv. 22.

Shakspeare. Henry VIII.

† Spect. No. 628.

subject. And, doubtless, it is equally reasonable to admit a construction which, though naturally equivocal, is fixed by the connexion, as to admit an equivocal term, the sense whereof is in this manner ascertained. Of an ambiguity thus removed the following will serve for an example: "Alexander, having conquered Darius, made himself master of his dominions." His may refer grammatically either to Alexander or to Darius, but as no man is said to make himself master of what was previously his own, the words connected prevent the false sense from presenting itself to the reader.

But it is not the pronouns only that are liable to be used ambiguously. There is in adjectives, particularly, a great risk of ambiguity, when they are not adjoined to the substantives to which they belong. This hazard, it must be owned, is greater in our language than in most others, our adjectives having no declension, whereby case, number, and gender, are distinguished. Their relation, therefore, for the most part, is not otherwise to be ascertained but by their place. The following sentence will serve for an example: "God heapeth favours on his servants ever liberal and faithful." Is it God or his servants that are liberal and faithful? If the former, say, "God, ever liberal and faithful, heapeth favours on his servants." If the latter, say, either "God heapeth favours on his ever liberal and faithful servants," or -————————— "his servants who are ever iiberal and faithful." There is another frequent cause of ambiguity in the use of adjectives, which hath been as yet, in our language, very little attended to. Two or more are sometimes made to refer to the same substantive, when, in fact, they do not belong to the same thing, but to different things which, being of the same kind, are expressed by the same generic name. I explain myself by an example: "Both the ecclesiastic and secular powers concurred in those measures." Here the two adjectives ecclesiastic and secular relate to the same substantive powers, but do not relate to the same individual things; for the powers denominated ecclesiastic are totally different from those denominated secular. Indeed, the reader's perfect knowledge of the difference may prevent his attending to this ambiguity, or rather impropriety of speech. But this mode of expression ought to be avoided, because, if admitted in one instance where the meaning perhaps is clear to the generality of readers, a writer will be apt inadvertently to fall into it in other instances, where the meaning is not clear, nay, where most readers will be misled. This too common idiom may be avoided either by repeating the substantive, or by subjoining the substantive to the first adjective, and prefixing the article to the second as well as to the first. Say, either "Both the ecclesiastic powers and the secular powers concurred in those measures;" or, which is perhaps preferable, "Both the ecclesiastic powers and the secular concurred in those measures." The substantive being posterior to the first adjective, and anterior to the second, the second, though it refers, cannot, according to grammatical order, belong to it. The substantive is therefore understood as repeated; besides, the repetition of the article has the force to denote that this is not an additional epithet to the same subject, but belongs to a subject totally distinct, though coming under the same denomination. There is,

indeed, one phrase liable to the aforesaid objection, which use hath so firmly established, that, I fear, it would savour of affectation to alter. The phrase I mean is, "The lords spiritual and temporal in parliament assembled." Nevertheless, when it is not expected that we should express ourselves in the style of the law, and when we are not quoting either a decision of the house of peers, or an act of parliament, I imagine it would be better to say, "The spiritual lords and the temporal." On the contrary, wherever the two adjectives are expressive of qualities belonging to a subject, not only specifically, but individually the same, the other mode of speech is preferable, which makes them belong also to the same noun. Thus we say properly, "The high and mighty states of Holland," because it is not some of the states that are denominated high, and others of them mighty, but both epithets are given alike to all. It would therefore be equally faulty here to adopt such an arrangement as would make a reader conceive them to be different. In cases wherein the article is not used, the place of the substantive ought to show whether both adjectives belong to the same thing, or to different things having the same name. In the first case, the substantive ought either to precede both adjectives, or to follow both in the second, it ought to follow the first adjective, and may be repeated after the second, or understood, as will best suit the harmony of the sentence, or the nature of the composition; for the second adjective cannot grammatically belong to the noun which follows the first, though that noun may properly suggest to the reader the word to be supplied. Thus I should say rightly, "It is the opinion of all good and wise men, that a vicious person cannot enjoy true happiness;" because I mean to signify that this is the opinion of those to whom both qualities, goodness and wisdom, are justly attributed. But the following passage in our version of the sacred text is not so proper, Every scribe "instructed into the kingdom of heaven, is like an householder, who bringeth out of his treasures things new and old."* Both epithets cannot belong to the same things. Make but a small alteration in the order, and say new things and old, and you will add greatly both to the perspicuity and to the propriety of the expression. In cases similar to the example last quoted, if a preposition be necessary to the construction of the sentence, it ought to be repeated before the second adjective. Thus, "Death is the common lot of all, of good men and bad." But when both adjectives express the qualities of an identical subject, it is better not to repeat the preposition. "The prince gave encouragement to all honest and industrious artificers of neighbouring nations to come and settle amongst his subjects." Here both qualities, honesty and industry, are required in every artificer encouraged by the prince. I shall observe lastly on this article, that though the adjectives relate to different things, if no substantive be expressed, it is not necessary to repeat the preposition. The reason is, that in such cases the adjectives are used substantively, or, to speak more properly, are real substantives. Thus we may say, either, "Death is the inevitable fate of good and bad, rich and poor, wise and foolish," or

*Matthew xiii. 52.

"of good and of bad, of rich and of poor.". When the definite article is prefixed to the first adjective, it ought to be repeated before the second, if the adjectives are expressive of qualities belonging to different subjects; but not if they refer to the same subject. Thus we say rightly, "How immense the difference between the pious and the profane." "I address myself only to the intelligent and attentive." In the former, the subjects referred to are manifestly different; in the latter they coincide as both qualities are required in every hearer. The following passage is by consequence justly censurable. The exceptionable phrases are distinguished by the character: "Wisdom and folly, the virtuous and the vile, the learned and ignorant, the temperate and debauched, all give and return the jest."* For the same reason, and it is a sufficient reason, that he said, "the virtuous and the vile," he ought to have said "the learned and the ignorant, the temperate and the debauched.”

I proceed to give examples in some of the other parts of speech. The construction of substantive nouns is sometimes ambiguous. Take the following instance: "You shall seldom find a dull fellow of good education, but (if he happen to have any leisure upon his hands,) will turn his head to one of those two amusements for all fools of eminence, politics or poetry." The position of the words politics or poetry makes one at first imagine, that along with the term eminence, they are affected by the preposition of, and construed with fools. The repetition of the to after eminence would have totally removed the ambiguity. A frequent cause of this fault in the construction of substantives, especially in verse, is, when both what we call the nominative case and the accusative are put before the verb. As in nouns those cases are not distinguished either by inflection or by prepositions, so neither can they be distinguished in such instances by arrangement.

The rising tomb a lofty column bore.‡

Did the tomb bear the column, or the column the tomb ?

And thus the son the fervent sire addrest.§

This, though liable to the same objection, may be more easily rectified, at least in a considerable measure. As the possessive pronoun is supposed to refer to some preceding noun, which, for distinction's sake, I have here called the antecedent, though the term is not often used in so great latitude, it is always better to be construed with the accusative of the verb, and to refer to the nominative as its antecedent. The reason is, the nominative, to which it most naturally refers, whether actually preceding or not, is always conceived in the order of things to precede. If then it was the son who spoke, say,

And thus the son his fervent sire addrest.

* Brown on the Characteristics, Ess. i. Sect. 5.

† Spectator, No. 43.

Pope's Odyssey, Book 12.

§ Ibid. Book 19.

If the father,

And thus his son the fervent sire addrest.

In confirmation of this, let us consider the way in which we should express ourselves in plain prose, without any transposition of words. For the first, "Thus the son addressed his father;" for the second, "Thus the father addressed his son;" are undoubtedly good : whereas, to say in lieu of the first, "Thus his son addressed the father;" and in lieu of the second, "Thus his father addressed the son," are not English. By the English idiom, therefore, the possessive pronoun is, in such instances, more properly joined to the regimen of the verb than to the nominative. If this practice were universal, as it is both natural and suitable to the genius of our tongue, it would always indicate the construction wherever the possessive pronoun could be properly introduced. For this reason I consider the two following lines as much clearer of the charge of ambiguity than the former quotation from the same work:

Young Itylus, his parent's darling joy,

Whom chance misled the mother to destroy.*

For though the words whom and the mother are both in the accusative, the one as the regimen of the active verb misled, the other as the regimen of the active verb destroy, yet the destroyer or agent is conceived in the natural order as preceding the destroyed or patient. If, therefore, the last line had been,

Whom chance misled his mother to destroy,

it would have more naturalty imported, that the son destroyed his mother; as it stands, it more naturally imports, agreeable to the poet's design, that the mother destroyed her son; there being in this last case no access for the possessive pronoun. I acknowledge, however, that uniform usage cannot (though both analogy and utility may) be pleaded in favour of the distinction now made. I therefore submit entirely to the candid and judicious the propriety of observing it for the future.

The following is an example of ambiguity in using conjunctions: "At least my own private letters leave room for a politician, well versed in matters of this nature, to suspect as much, as a penetrating friend of mine tells me." The particle as which in this sentence immediately precedes the word a penetrating friend, makes frequently a part of these compound conjunctions as much as, as well as, as far as. It will therefore naturally appear at first to belong to the words as much, which immediately precede it. But as this is not really the case, it ought to have been otherwise situated; for it is not enough that it is separated by a comma, these small distinctions in the pointing being but too frequently overlooked. Alter the arrangement then, and the expression will be no longer ambiguous: "At † Spect. No. 43.

* Pope's Odyssey, Book 19.