Ch. 5. of them. Mr Harris has very properly divided them into præpofitive and sub

junctive, according to their order in the

fentence; and he has, with his ufual accuracy and elegance, explained the nature of that fubjunctive pronoun commonly called the relative, fuch as qui in Latin, who or which in English. And I think it is not improperly called the relative by way of eminence, because it marks not only that relation which all the pronouns of the third perfon, except the demonArative, have to the object mentioned before, but alfo the relation that it has with the fyntax or conftruction of the fpeech, which it joins together, and as Mr Harris expreffes it, renders more compact *.

From this account of the pronoun, the following definition of it may be extracted: A pronoun is a word denoting a fubftance, not directly, but by reference either to fomething prefent, or fomething mentioned in the preceding part of the difcourfe.

Before I conclude this chapter, I must obferve, that this part of fpeech is fo neceffary, that the moft barbarous langua

Hermes, pag. 79.

ges

ges have it, even the Huron, as I have Ch. 5. obferved. Thofe favages indeed have not the power of abstraction fo much as to form a feparate idea of it, and exprefs it by a diftinct word; but they always throw it in with the fignification of other words, particularly of the verb: and yet even fo expreffed, it fhows that they have been fo far philofophers, as to make in fome fort the analysis above mentioned of the fubjects of difcourfe, into the fpeaker, the hearer, and fome third perfon or thing. But neceffity will make philofophers even of favages.

Of the article, and the various ufes of it.

This of well

part of speech very well deferves Ch. 6. a chapter by itself; for, if I miftake not, it is of as fubtle fpeculation as perhaps any thing belonging to language, particularly as it is used in Greek. It is not a neceffary part of fpeech, for it is very feldom ufed by Ho

mer

Ch. 6. mer

*

; and it is not at all ufed in the most antient dialect of Greek that is preferved to us, I mean the Latin. And in the Ionic dialect it is used indifcriminately, either as an article or a relative pronoun. The appropriating of it therefore, for the purpose of an article, as is done by all the Attic writers, appears to be a refinement of the language in later times. But wherein this refinement confifts, has not, I think, hitherto been fufficiently explained, nor any fatisfying account given of certain ufes of it.

The Stoics, as we are informed by Prifciant, reckoned the article among the pronouns; and both Apollonius and Theodorus Gaza speak of it as a relative ргоnoun, diftinguifhed only from the common relative by its pofition in the difcourfe; and therefore they call the one the prapofitive article, and the other the subjunctive ‡. But I hope to be able to fhew, that its office is different from that of a pronoun of

*,, rò, is frequently ufed by Homer, in place of the relatives,,, but very feldom as an article.

Lib. 1. pag. 574. See alfo Hermes, pag. 74.

* Υποτακτικόν καὶ προτακτικόν άρθρον. See Hermes, pag. 18.

any

any kind, and that it deferves very well Ch. 6. to be ranked by itself among the parts of speech.

All the words of a language are either the names of individual things, or general terms; that is, in the language of grammarians, either proper names or appellatives. The article in Greek is applied to both; for they fay Ewxparns, as well as ὁ άνθρωπος. But they muft be both the name of fubftances of one kind or another; for the use of the article, as well as of the pronoun, is to fingle out and distinguish subftances from one another, though it does it, as I fhall fhew, in a different manner. We will begin with confidering it as applied to proper names.

The application of it in this way, may appear, at first fight, altogether unneceffary; for a thing feems to be fufficiently defined and distinguished, by being marked by a name. And accordingly, Mr Harris thinks, that the article added to the name of Socrates is a mere pleonaẩm, or that it can be of no ufe, unless perhaps to diftinguifh fexes*. And it would be fo,

* Hermes, pag. 226.

Ch. 6. if there had never had been but one Socrates in the world: for then it would have been as unneceffary, and as infignificant a pleonafm, to add the article to Socrates, as to add it to the pronouns of the first and fecond perfon, which point out particular perfons that cannot poffibly be confounded with any other. But we all know, that among the Greeks, as well as among us, the fame name was common to many individuals; nor indeed is it poffible, by the nature of things, that there fhould be a feparate name for every individual. And in this very inftance, there have been more of the name of Socrates than one; and particularly, as I remember, there is an ecclefiaftical hiftorian of that name; and, even while Socrates lived, there was another Socrates, who is introduced in one of Plato's dialogues, and distinguished by the name of Socrates How then is this Socrates to be younger.

diftinguished from any other? It is, I fay, by the addition of the article; and that in two different ways.

In the first place, if the name was mentioned before in the difcourfe or writing, the article denotes a reference to that for

mer