CHAPTER XI.

OF PREPOSITIONS.

66

Connectives. 310. FROM the consideration of the Article, which Harris ranks among the Definitives, I proceed to the Prepositions and Conjunctions, which together form his class of Connectives. His reasons for adopting such a class are these. As in nature a substantive coalesces at once with its attribute, an action with its agent, a passion with its patient, and even a primary attribute with a secondary, so in grammar, the substantive may coalesce at once with its adjective, as a wise man," a "fierce lion;" the verb transitive may coalesce at once with its nominative and accusative, as "Alexander vanquished Darius ; and the adverb with the verb or adjective which it modifies, as "he fought bravely," "he was completely victorious." But when it is necessary to make any other union of conceptions, it can only be done either by a combination of words; by a change in the word which requires to be modified; or by a separate word, which, as it serves to connect the others, may be called a connective. Omitting for the present the two first methods, let us observe how connectives may be used. If in addition to the assertion that Alexander vanquished Darius, I wish to assert that he also vanquished Porus, I can effect this purpose by the connective "and," as "Alexander vanquished Darius and Porus." If I wish to state the motive of Alexander's fighting, I may say "he fought for fame." The word "and" is commonly called a conjunction; the word "for," a preposition: and it is true that they are both employed to connect words which would otherwise remain unconnected; but there is this important difference between them-the conjunction connects, and does nothing more; the preposition introduces a further conception, namely that of the particular relation in which the connected conceptions stand to each other. In the example given, I do not merely connect, in the mind of the hearer, the conceptions of Alexander, or of fighting, with the conception of fame; for they would be equally connected if fame had been the unexpected and unthought-of consequence of his fighting; but I show that fame stood towards the action in the particular relation of a motive. I therefore consider that the word which thus shows a distinct relation between two conceptions may be justly deemed a separate part of speech.

Preposition.

311. This part of speech has been called a Preposition, because in the Greek and Latin languages the words so employed were commonly (though with some exceptions) præposita, placed immediately

before the substantives to which they referred. In those languages,
too, the words in question were subject to few variations in point of
form. These circumstances, though merely accidental, were unfor-
tunately selected by some grammarians as essential properties of the
part of speech under consideration; and hence originated the well-
known definition, Præpositio est pars orationis invariabilis, quæ præpo-
nitur aliis dictionibus. The Greek grammarians, whom Harris followed,
ranked both the preposition and conjunction under the common head
of Zúvdeoμos, or the connective; and the Stoics, adding this cir-
cumstance to the ordinary position of the preposition in a sentence,
called this part of speech Σύνδεσμος Προθετικός, the " prepositive
connective.” Another accidental peculiarity of most of the words
which were used as prepositions in Greek and Latin, as well as in
some modern languages, was that their original and peculiar meaning
had, in process of time, become obscure; and from hence some persons
were led to think that these words had no signification of their own.
The learned HARRIS gives the following definition, "A preposition is a
part of speech devoid itself of signification, but so formed as to unite two
words that are significant, and that refuse to coalesce or unite of them-
selves." CAMPANELLA also says of the preposition, Per se non significat;
and HOOGEVEEN says, "Per se posita et solitaria nihil significat." Under
the same impression, the Port Royal grammarians say,
recours, dans toutes les langues, à une autre invention, qui a été d'inventer
de petits mots pour être mis avant les noms, ce qui les a fait appeller pré-
positions." And M. de BROSSES says, "Je n'ai pas trouvé qu'il fut
possible d'assigner la cause de leur origine; tellement que j'en crois la
formation purement arbitraire."

66

" On a eu

rejected.

312. Now in all this there was much inaccuracy of reasoning, as Errors applied to Universal Grammar. The position of this sort of words in a sentence, had the fact been so in all known languages, must have been owing to accidental causes; bút the fact is otherwise. Even in Latin the preposition tenus was always placed after the noun which it governed; so Plautus uses erga after a pronoun, as in mederga, for erga me; and cum is employed in like manner in the common expressions mecum, tecum, nobiscum, vobiscum. These and other examples of a like kind induced some authors to make a class of postpositive prepositions. "Dantur etiam," says CARAMUEL, Postpositiones, quæ præpositiones postpositiva solent dici;" but I shall elsewhere show that there are languages in which all the prepositions, so to speak, are postpositive. Some writers, who for this and similar reasons reject the word preposition, have adopted in its stead that of adnomen, adnoun; but as their example has been seldom followed, and as it is my object to change as little as possible received modes of expression, I shall adhere to the ordinary grammatical term, preposition, only reminding the reader that it is not to be taken as expressing an essential property of the part of speech in question. That prepositions are indeclinable may be the case in most languages,

Delinition.

Examples.

Sentence

complex.

That they sig

but is certainly no necessary part of their definition.
nify nothing of themselves, if it were true in any degree, would be
only part of their history, and would throw no light whatever on the
grammatical principles which regulate their use. It is not surprising
that Mr. TooKE should ridicule these postpositive prepositions, and
nonsignificant words which communicate signification to other words;
but unfortunately he only substitutes worse errors of his own, when
he asserts that prepositions are always names of real objects, and do
not show different operations of the mind.

313. The real character and office of the preposition have been
stated with a nearer approach to accuracy by Bishop WILKINS and
VOSSIUS; but neither of them seems to have given a full and satis-
factory definition of this part of speech. WILKINS says, "Prepositions
are such particles whose proper office it is to join integral with integral
on the same side of the copula, signifying some respect of cause, place,
time, or other circumstance, either positively or privatively." VOSSIUS
says, Præpositio est vox per quam adjungitur verbo nomen, locum,
tempus, aut caussam significans, seu positive seu privative.” It
suited Wilkins's scheme of universal grammar to call the preposition
a particle; but however appropriate this may be to a theoretical view
of language, such as it never did, and probably never will exist, it is
inconsistent with those philosophical principles on which the actual
use of speech among men depends; neither is it material on which side
of the copula a preposition may be placed by the idiom of any par-
ticular language. On the other hand, as Wilkins includes under the
term integral both the noun and the verb, he is in this respect more
accurate than Vossius, for the preposition does not merely join a noun
to a verb, but sometimes to another noun. I therefore, with that
diffidence which becomes all persons who endeavour in any degree to
clear the path of science, shall propose the following definition :—
A preposition is a part of speech employed in a complex sentence, and
serving to express the relation in which the conception named by a noun
substantive stands to that named by another noun substantive, or asserted
by a verb.
314. Thus, if I say,
"he hired a house with a garden,"
""Solomon
was the son of David," the words with and of are prepositions, the
former expressing the relation of contiguity between the substantives
"house" and " garden," and the latter expressing the relation of filial
descent between the substantives" son
"David." Again, if

and

I say, "he spoke concerning the law," "he marched from Capua to Rome," the words concerning, from, and to are prepositions, the first expressing the relation of subjectivity in which the noun "law" stands to the verb "spoke," and the two others expressing the different relations of locality in which the nouns "Capua" and "Rome" stand to the verb "marched."

315. In developing the above definition, I first observe that the sentence in which a preposition is employed must be a complex one:

and this is evident; for, in addition to the assertion of a connection
between a subject and its attribute (which together forms a simple
sentence, as "John walks," or "John is walking"), the preposition
expresses a conception of relation, which conception, if added to the
attribute and assertion in the verb, forms another simple sentence.
If I say,
"John walks before Peter," I, in effect, make two assertions,
first, that John is walking, and, secondly, that the walking is before
Peter. In the language of lawyers, I present two issues; for it may
be admitted that John walks, and denied that the walking is before
Peter; and this latter may chance to become an important question
affecting rights not only of precedence and station in society, but also
of property, and not only between individuals or families, but between
nations. In the secondary question, the relation of locality is ex-
pressed by the preposition before, which is necessary to connect the
assertion "walks" with the name "Peter;" for if it were omitted,
and I should say,
"John walks Peter," the sentence would be unin-
telligible. In like manner, if the conception of relation be added to
one of two connected substantives, as "Solomon was the son of
David," the sentence involves two assertions, viz., that Solomon stood
in the relation of a son, and that that relation connected him with
David; and the word expressing the connection is the preposition
"of."

316. It follows, from the nature of connectives, as stated by Verb neuter. Mr. Harris, that where a verb is neuter it may be connected immediately with a following substantive by means of a preposition. Thus the neuter verb "walks" is immediately connected with the following substantive "Peter" by means of the preposition "before;" but if the verb be transitive it cannot be immediately connected with a substantive by means of a preposition, but must first be followed by its proper accusative, that is to say, by the substantive expressing the recipient of the action, ex. gr. :—

Now with strong pray'r, and now with stern reproach,

He stirs their valour.

Here the sense would have been wholly lost if the accusative "valour" had been omitted: and the same rule applies where the relation is marked by an inflection of the substantive, as in the original of the passage just quoted

Nunc prece, nunc verbis virtutem accendit amaris,

where the ablatives prece and verbis amaris show the relation of instrumentality, in which the conceptions expressed by them stand to the verb accendit; but those ablatives would have been unmeaning had not the verb been followed by its proper accusative, virtutem.

317. In languages which admit of compounding a verb with a pre- Compound position, there may be differences of idiom. The verb, if neuter,

* Virg. Æn. 10, 368.

verb.

Relation.

Its foundation.

usually assumes a transitive character, as when Satan, who is described as forcing his way into Paradise,

at one slight bound, high overleap'd all bound. *

If the verb be transitive, then (according to the idiom of the language) the related substantive may be either inflected in accordance with the preposition in the verb or else accompanied with a separate preposition. When inflected, it adopts a case which is said by grammarians to be governed by the preposition in composition, as

Nam tibi, Thymbre,caput Evandrius abstulit ensis; †

where the preposition abs (though governing an ablative when alone) may be said, as forming part of the verb abstulit, to govern the dative tibi; and where both the preposition and the dative inflection express the relation of objectivity, in which the person (Thymbrus) stood to the act signified by the verb abstulit and its accusative caput, as if the phrase had been "abstulit caput abs te."

318. The next point to be considered in the definition of a preposition above given is the nature of the relations which it serves to express. Now, Relation, which is the fourth of the logical predicaments, supposes three things, the subject, or thing related, the object, or correlative, and the relation itself, or circumstance existing in the subject by means of which it is related to the object, and which logicians call the foundation. When we say, "John is before Peter," John" is the subject, "Peter" is the correlative, and "before" is the foundation, or, as I have been accustomed to speak, the conception of a particular relation, expressed prepositionally.

319. It is manifest, that the circumstance, whatever it be, that forms the foundation of a logical relation, or (which is the same thing) that constitutes (when expressed in language together with its subject and object) a preposition, may either be common to the two terms (as they are called) of the relation, or it may belong to one of them exclusively. If I say, "John is with Peter," the relation expressed by the preposition with belongs equally to Peter and to John; but if I say John is before Peter, the relation expressed by the preposition before belongs exclusively to John. In the first case it is perfectly indifferent whether I say "John is with Peter," or "Peter is with John;" it is perfectly indifferent which I make the subject and which the object of the relation: but in the other case, if I were to say "Peter is before John," I should not only vary the assertion, but I should directly contradict it. Still the foundation of the relation would be the same; for, as a great philologist has observed, "at the bottom of every preposition, in its original sense, there exists a relation between two opposite conceptions." Thus before implies behind, and over implies under. We may illustrate this with the trivial comparison of two children playing at see-saw. If John and † Virg. En. 10, 394.

* Milton, P. L. 4, 181.

Bopp, Comp. Gram. 1. 377.