that one word bears to another being in a great measure determined by their pofition, we are often confined to one particular arrangement; and, when we depart from that, and attempt thofe deviations from the grammatical order which are fo graceful in antient authors, are apt to write obfcurely and affectedly.-In this refpect, however, the English tongue is more fufceptible of variety than the French, and English verfe than English profe. Indeed, almost all arrangements of words, that do not perplex the sense, are permitted in our poetry, efpecially in our blank verfe a privilege, whereof Milton availing himfelf in its full latitude, difplays in the Paradife Loft a variety and elegance of compofition, which have never been equalled in any other modern tongue, and may bear to be compared with the moft elaborate performances of antiquity. Our want of inflection in our nouns, adjectives, and participles, makes us, in our written language, more dependent upon punctuation, than the antients were. Indeed, of punctuation, as we understand it, they had no idea and it does not appear, that they fuffered any inconvenience from the want of it. Whereas, in modern language, the mifplacing or omiflion of a point will often alter the fenfe; and, if we had no points, we fhould find it difficult to write fo as to be understood; to write elegantly, and yet intelligibly, would be impoffible. There is a paffage in Cato, which, from being generally, if not always, mifpointed, is, I think, generally mifunderflood: The The ways of heaven are dark and intricate, Puzzled in mazes, and perplex'd with errors Thus the lines are printed in all the editions I have feen. And yet, it can hardly be fuppofed, that Addison's piety would have permitted him to fay, or to make Cato fay, that "the ways of "heaven are perplexed with errors;" or that his tafte would have warranted fuch an expreffion as, "the ways of heaven are puzzled." I therefore prefume, that the first line is a sentence by itself, and ought to end in a point or colon; and that the fequel, ranged in the grammatical order, amounts to this; Our understanding, puzzled "in mazes, and perplexed with errors, traces "the ways of heaven in vain :" which is both elegant and true. Now this ambiguity could not have taken place in Latin or Greek, nor indeed in French or Italian, even though there had not been one point in the fentence: because the participles puzzled and perplexed would have been made to agree with the fingular noun understanding; in which cafe they could not alfo agree with the plural noun ways. In explaining the feveral cafes, and fhowing, why there are neither more nor fewer, and why fo many, and what is the nature of each, fome authors have been more particular, and displayed greater fubtlety, than in my opinion was requifite. As to the number of cafes, grammarians have always differed in their fentiments, and are are not reconciled to this day. Many explode the ablative, because the Greeks could do without it; and fome will not allow the vocative to be a cafe, because it is often, both in Latin and in Greek, the fame with the nominative. Arif totle and the Peripateticks maintained, that the nominative is not a cafe; and the Stoicks were equally pofitive, that it is. In the Armenian language, the number of cafes is faid to be ten: and I fhould not wonder, if a grammarian, much given to novelty and paradox, were to affirm, that there are in English as many cafes almost as there are prepofitions. While opinions are fo different in regard to the precife num. ber, it is vain to inquire, why there are neither more nor fewer, and why fo many. The nature of each particular cafe may be better understood by examples, than by logical definition. Indeed, all the definitions I have feen of the feveral cafes, are liable to objection; except, perhaps, that of the nominative, which is given by Mr. Harris, who calls it, "That cafe, without which there is no regular " and perfect fentence." "The Accufative," fays the fame author, is that cafe, which to an efficient nominative, and a verb of action, fubjoins, either the ef"fect, or the paffive fubject:"-the effect, as when I fay, Lyfippus fecit ftatuas, Lyfippus made ftatues; the fubjcct, as in this example, Achilles vulneravit Hectora, Achilles wounded Hector. -But this, though frequently, is not univerfally true. When it is faid, Antonius lefit Ciceronem, the first word is an efficient nominative, the fecond an active verb, and the third an accufative, according according to the definition: but when I fay, Antonius nocuit Ciceroni, the efficient nominative and active verb are followed, not by an accufative, but by a dative. And there are other verbs of active fignification, as Potior, for example, which take after them, rarely an accufative, fometimes a genitive, and frequently an ablative. And what fhall we fay of accufatives governed by prepofitions; as habitat juxta montem, he dwells near the mountain? For neither is habitat, he dwells, an active verb; nor is the mountain, in any fenfe of the words, either the fubject or the effect of his dwelling; and yet montem, the mountain, is the accufative. The Genitive, according to the fame learned writer, expreffes all relations commencing from itfelf; and the Dative, all relations tending to itself. Yet, when I fay, editus regibus, defcended of kings, I exprefs a relation commencing from the kings, who are, notwithstanding, of the ablative cafe, in the Latin: and eripuit morti, he rescued from death, is in Latin dative, and expreffes, for all that, a relation tending, not to death, but from it. One may fay, indeed, that thefe are refinements in the language, and deviations from the primitive fyntax. But I know not, how we are to judge of cafes, except from the purpofes to which they are applied in the languages that have them; nor on what authority we have a right to fuppofe, that the primitive fyntax of Greek and Latin was different from that which we find in Greek and Latin authors. In a word, every cafe, almoft, is applied to fo many purposes in fyntax, that to defcribe its ufe ufe in a fingle definition, feems to be impoffible, or at least fo difficult, and withal fo unneceffary, that it is not worth while to attempt it. None of the antient grammarians, fo far as I know, has ever made the attempt: and I believe it will be allowed, that in this fort of fubtlety they are not inferiour to their brethren of modern times. § 2. Of Conjunctions. I divided Connectives into two claffes; Prepofitions, which connect words, and Conjunctions, which connect fentences. A Conjunction may be thus defined: "A part "of fpeech, void itfelf of fignification, but of "fuch efficacy, as to join fentences together, "and fhow their dependence upon one another." The Conjunction, fays Ariftotle, makes many one and Ammonius compares the words of this class to those pegs and nails by which the feveral parts of a machine are united. Perhaps it may be thought, that Conjunctions, as well as prepofitions, do fometimes connect words; as when we fay, He is a learned and a wife and a good man. But this fentence, when analyfed, will be found to confift of three diftinct fentences;-he is a learned man;-he is a wife man;-he is a good man; or,-he is learned, he is wife, he is good: which three would for ever remain diftinct and feparate, if we had no connecting words to unite them in one fentence; even as the feveral parts of |