Ch. 4. be fhewn afterwards. This idiom too we have in English; for we fay the good, and the fair. In the fame In the fame way the Greeks form nouns of their participles, as the τὸ τρέχον, and the τὸ ποιον. We have the fame form of a noun in English; for we fay, the running, and the doing: but the meaning is different; for in English it denotes, as I have already observed, the action of the verb; whereas, in Greek, it fignifies the agent.

All these three kinds of nouns I call by the common name of fubftantives, distinguishing the first and second by the names of primary and fecondary fubftantives, according to the nature of the fubftances they exprefs. The last may be called ideal or fictitious fubftantives, being entirely of the mind's own creation; but I chufe to call them by their common name of abf

tract nouns,

СНАР.

CHAP. V.

Of pronouns.-The neceffity of inventing them.-The nature and different kinds of them.

Α'.

LL the objects of human knowledge, Ch. 5. and confequently of discourse, are either generals or particulars. The knowledge of generals, as I have already had occafion to obferve, is by far the more valuable knowledge, as by it we know even individuals; for we know nothing of Peter, James, or John, by hearing them named, or even by feeing them, unless we know the species to which they belong. But the knowledge of individuals is alfo abfolutely neceffary for human life, and in common life the greateft part of our converfation is concerning individuals. Now the number of individuals is infinite, at leaft with refpect to our capacities; yet the purposes of life require, that in the ufe of fpeech they fhould be fingled out, and diftinguifhed one from another. Here

Ch. 5. is one of the great difficulties that the inventors of language had to ftruggle with: let us fee how they got over it.

It may be thought that proper names for the feveral individuals, will ferve to distinguish them. But, in the first place, it is impoffible that all the individuals which may be the fubject of discourse, fhould have particular names, at least fuch as are known to the speakers and hearers; even the perfons who have occafion to converfe together may not know one another's names. 2dly, Suppofe that the fubjects of the converfation have all names, and that those names are known to the parties, the fame name may be common to feveral individuals, and indeed it is impoffible that every individual fhould have a different name; there must therefore be fome way of marking, that the name ufed by the fpeaker is the name of the individual whom the hearer knows, and of no other. And lastly, Suppofe this difficulty got over, and that the parties were agreed about the name, as applicable to the fame individual known to them both, it would be tedious, and a great incumbrance to the discourse, if the name was to be repeated

as

as often as the object was mentioned; and Ch. 5. accordingly we obferve it as a defect in the language of children, that instead of ufing the pronoun I, they name themselves *.

Names therefore will not folve the difficulty, and fome other way must be devised. The only way that feems poffible is, to divide the fubjects of conversation into certain claffes. But into what claffes ? The common divifion into specieses, by which the infinity of things is limited and circumfcribed, will not ferve the purpose; for the thing here to be done, is to dif tinguish the individuals of the several specieses, not the specieses themselves. We muft therefore try fome other fome other way of claffing the fubjects of difcourfe; and fuppofe we should divide them into fuch as are prefent during the difcourfe, and fuch as are not. The divifion is fufficiently comprehenfive; for every fubject of converfation must either be present or not prefent. But I doubt it will not ferve the purpose neither. The objects present indeed might be pointed out by the speaker to the hearer; but we are inquiring at present how they are to be distinguished

This is an obfervation of Dr Smith in his Differtation on the formation of Languages.

by

Ch.

5. by words, not by figns or geftures. Now

though the diftinction in general, might, no
doubt, be marked by words betwixt objects
prefent and objects not prefent, how are the
feveral particular objects present or absent,
to be distinguished from one another? for
there
may be many objects present during
the converfation, and the number of thofe
that are not present is without bounds.

But this divifion, though it do not folve the difficulty, leads to another diftinction that may perhaps do the business: for of the fubjects of converfation prefent, there are two which must neceffarily be present, and which, by their natures, are limited and determined; I mean the fpeaker, and the hearer, or the person to whom the difcourfe is addreffed. And every fubject of discourse must of neceffity be either the speaker, the hearer, or fome third object different from both. Here then is another divifion, equally comprehensive as the former let us try whether it will not answer the purpose better.

If either the speaker or hearer be the fubject of the difcourfe, there is no more ado but to invent two words to defign and diftinguifh them from one another. And these

words