OBSERVATIONS.
OBS. 1.--The relations of things to things in nature, or of words to words in discourse, are infinite in number, if not also in variety. But just classification may make even infinites the subjects of sure science. Every relation of course implies more objects, and more terms, than one; for any one thing, considered merely in itself, is taken independently, abstractly, irrelatively, as if it had no relation or dependence. In all correct language, the grammatical relation of the words corresponds exactly to the relation of the things or ideas expressed; for the relation of words, is their dependence, or connexion, according to the sense. This relation is oftentimes immediate, as of one word to an other, without the intervention of a preposition; but it is seldom, if ever, reciprocally equal; because dependence implies subordination; and mere adjunction is a sort of inferiority.
OBS. 2.--To a preposition, the prior or antecedent term may be a noun, an adjective, a pronoun, a verb, a participle, or an adverb; and the subsequent or governed term may be a noun, a pronoun, a pronominal adjective, an infinitive verb, or a participle. In some instances, also, as in the phrases, in vain, on high, at once, till now, for ever, by how much, until then, from thence, from above, we find adjectives used elliptically, and adverbs substantively, after the preposition. But, in phrases of an adverbial character, what is elsewhere a preposition often becomes an adverb. Now, if prepositions are concerned in expressing the various relations of so many of the different parts of speech, multiplied, as these relations must be, by that endless variety of combinations which may be given to the terms; and if the sense of the writer or speaker is necessarily mistaken, as often as any of these relations are misunderstood, or their terms misconceived; how shall we estimate the importance of a right explanation, and a right use, of this part of speech?
OBS. 3.--The grammarian whom Lowth compliments, as excelling all others, in "acuteness of investigation, perspicuity of explication, and elegance of method;" and as surpassing all but Aristotle, in the beauty and perfectness of his philological analysis; commences his chapter on conjunctions in the following manner: "Connectives are the subject of what follows; which, according as they connect either Sentences or Words, are called by the different Names of Conjunctions OR Prepositions. Of these Names, that of the Preposition is taken from a mere accident, as it commonly stands in connection before the Part, which it connects. The name of the Conjunction, as is evident, has reference to its essential character. Of these two we shall consider the Conjunction first, because it connects, not Words, but Sentences."--Harris's Hermes, p. 237.
OBS. 4.--In point of order, it is not amiss to treat conjunctions before prepositions; though this is not the method of Lowth, or of Murray. But, to any one who is well acquainted with these two parts of speech, the foregoing passage cannot but appear, in three sentences out of the four, both defective in style and erroneous in doctrine. It is true, that conjunctions generally connect sentences, and that prepositions as generally express relations between particular words: but it is true also, that conjunctions often connect words only; and that prepositions, by governing antecedents, relatives, or even personal pronouns, may serve to subjoin sentences to sentences, as well as to determine the relation and construction of the particular words which they govern. Example: "The path seems now plain and even, but there are asperities and pitfalls, over which Religion only can conduct you."--Dr. Johnson. Here are three simple sentences, which are made members of one compound sentence, by means of but and over which; while two of these members, clauses, or subdivisions, contain particular words connected by and.
OBS. 5.--In one respect, the preposition is the simplest of all the parts of speech: in our common schemes of grammar, it has neither classes nor modifications. Every connective word that governs an object after it, is called a preposition, because it does so; and in etymological parsing, to name the preposition as such, and define the name, is, perhaps, all that is necessary. But in syntactical parsing, in which we are to omit the definitions, and state the construction, we ought to explain what terms the preposition connects, and to give a rule adapted to this office of the particle. It is a palpable defect in nearly all our grammars, that their syntax contains NO SUCH RULE. "Prepositions govern the objective case," is a rule for the objective case, and not for the syntax of prepositions. "Prepositions show the relations of words, and of the things or thoughts expressed by them," is the principle for the latter; a principle which we cannot neglect, without a shameful lameness in our interpretation;--that is, when we pretend to parse syntactically.
OBS. 6.--Prepositions and their objects very often precede the words on which they depend, and sometimes at a great distance. Of this we have an example, at the opening of Milton's Paradise Lost; where "Of," the first word, depends upon "Sing," in the sixth line below; for the meaning is--"Sing of man's first disobedience," &c. To find the terms of the relation, is to find the meaning of the passage; a very useful exercise, provided the words have a meaning which is worth knowing. The following text has for centuries afforded ground of dispute, because it is doubtful in the original, as well as in many of the versions, whether the preposition in (i. e., "in the regeneration") refers back to have followed, or forward to the last verb shall sit: "Verily I say unto you that ye who have followed me, in the regeneration, when the Son of man shall sit in the throne of his glory, ye also shall sit upon twelve thrones, judging the twelve tribes of Israel."--Matt., xix, 28. The second in is manifestly wrong: the Greek word is [Greek: epi], on or upon; i. e., "upon the throne of his glory."
OBS. 7.--The prepositions have, from their own nature, or from custom, such an adaptation to particular terms and relations, that they can seldom be used one for an other wi
