He that would see more on this subject, may consult Ammonius the Peripatetic, in his Commentary on the treatise De Interpretatione, where the subject is treated at large with respect to the Greek tongue. We shall only observe, that as all such speculations are at best but conjectures, they should therefore be received with candour, rather than scrutinized with rigour. Varro's words, on a subject near akin, are for their aptness and elegance well worth attending. Non mediocres enim tenebræ in silva, ubi hæc captanda; neque eo, quo pervenire volumus, semitæ tritæ; neque non in tramitibus quædam objecta, quæ euntem retinere possunt.

To conclude this chapter. We may collect from what has been said, that both number and gender appertain to words, because, in the first place, they appertain to things; that is to say, because substances are many, and have either sex or no sex; therefore substantives have number, and are masculine, feminine, or neuter. There is, however, this difference between the two attributes: number in strictness descends no lower than to the last rank of species: gender, on the contrary, stops not

fortune, virtue, &c. in Greek, Latin, French,
and most modern languages, though they
are diversified with genders in the manner
described, yet never vary the gender which
they have once acquired, except in a few
instances where the gender is doubtful.
We cannot say aperǹ or å åperǹ, “hæc vir-
δι
tus," or "hic virtus," "la vertu," or "le ver-
tu," and so of the rest. But it is otherwise in
English. We in our own language say,
Virtue is its own reward, or Virtue is her
own reward; Time maintains its wonted
pace, or Time maintains his wonted pace.

There is a singular advantage in this liberty, as it enables us to mark, with a peculiar force, the distinction between the severe or logical style, and the ornamental or rhetorical. For thus, when we speak of the above words, and of all others naturally devoid of sex, as neuters, we speak of them as they are, and as becomes a logical inquiry. When we give them sex, by making them masculine or feminine, they are from thenceforth personified; are a kind of intelligent beings, and become, as such, the proper ornaments either of rhe toric or of poetry.

Thus Milton:

The thunder, Wing'd with red light'ning and impetuous rage, Perhaps hath spent his shafts.

Par. Lost, i. 174. The poet, having just before called the hail and thunder, "God's ministers of vengeance," and so personified them, had he afterwards said its shafts for his shafts, would have destroyed his own image, and

approached withal so much nearer to
prose.

The following passage is from the same poem:
Should intermitted vengeance arm again
His red right hand. Par. Lost, ii. 174.

In this place his hand is clearly preferable either to her's or it's, by immediately referring us to God himself, the avenger.

I shall only give one instance more, and quit this subject.

At his command th' up-rooted hills retir'd
Each to his place: they heard his voice and went
Obsequious: heav'n his wonted face renew'd,
And with fresh flow'rets hill and valley smil`d.
Par. Lost, b. vi.

See also ver. 54, 55, of the same book.

Here all things are personified; the hills hear, the valleys smile, and the face of heaven is renewed. Suppose, then, the poet had been necessitated by the laws of his language to have said, Each hill retir'd to its place, Heaven renew'd its wonted face; how prosaic and lifeless would these neuters have appeared; how detrimental to the prosopopeia which he was aiming to establish! In this, therefore, he was happy, that the language in which he wrote imposed no such necessity; and he was too wise a writer to impose it on himself. It were to be wished his correctors had been as wise on their parts.

* De Ling. Lat. 1. iv.

The reason why number goes no lower is, that it does not naturally appertain to individuals; the cause of which see before, p. 128.

here, but descends to every individual, however diversified. And so much for substantives, properly so called.

CHAPTER V.

CONCERNING SUBSTANTIVES OF THE SECONDARY ORDER.

We are now to proceed to a secondary race of substantives, a race quite different from any already mentioned, and whose nature may be explained in the following manner.

Every object which presents itself to the senses or the intellect, is either then perceived for the first time, or else is recognized as having been perceived before. In the former case it is called an object, Tys πрwτns yvwoews, "of the first knowledge,' or acquaintance;" in the latter it is called an object, Tŷs devTépas yvoσews, "of the second knowledge," or acquaintance.

Now as all conversation passes between particulars or individuals, these will often happen to be reciprocally objects Ts TWINS YOεws, that is to say, "till that instant unacquainted with each other." What then is to be done? How shall the speaker address the other, when he knows not his name? or how explain himself by his own name, of which the other is wholly ignorant? Nouns, as they have been described, cannot answer the purpose. The first expedient upon this occasion seems to have been 4eîgis, that is, "pointing, or indication by the finger or hand," some traces of which are still to be observed, as a part of that action which naturally attends our speaking. But the authors of language were not content with this. They invented a race of words to supply this pointing; which words, as they always stood for substantives or nouns, were characterized by the name of avтwvvμiai, or "pronouns." These, also, they distinguished by three several sorts, calling them pronouns of the first, the second, and the third person, with a view to certain distinctions, which may be explained as follows.

Suppose the parties conversing to be wholly unacquainted, neither name nor countenance on either side known, and the

See Apoll. de Syntaxi, 1. i. c. 16. p. 49; L. ii. c. 3. p. 103. Thus Priscian: Interest autem inter demonstrationem et relationem hoc; quod demonstratio, interrogationi reddita, primam cognitionem ostendit; qnis facit? Ego: relatio vero secundum cognitionem significat, ut, Is, de quo jam dixit. Lib. xii. p. 936. edit. Putschii.

* Ἐκεῖνο οὖν ἀντωνυμία, τὸ μετὰ δεῖξεως ἢ ἀναφορᾶς ἀντονομαζομένων Apoll. de

Synt. l. ii. c. 5. p. 106. Priscian seems to consider them so peculiarly destined to the expression of individuals, that he does not say they supply the place of any noun, but that of the proper name only. And this undoubtedly was their original, and still is their true and natural use. Pronomen est pars orationis, quæ pro nomine proprio uniuscujusque accipitur. Prisc. 1. xii. See also Apoll. l. ii. c. 9. p. 117, 118.

subject of the conversation to be the speaker himself. Here, to supply the place of pointing by a word of equal power, they furnished the speaker with the pronoun I. I write, I say, I desire, &c. and as the speaker is always principal with respect to his own discourse, this they called, for that reason, the pronoun of the first person.

Again, suppose the subject of the conversation to be the party addressed. Here, for similar reasons, they invented the pronoun thou. Thou writest, thou walkest, &c.: and as the party addressed is next in dignity to the speaker, or at least comes next with reference to the discourse, this pronoun they therefore called the pronoun of the second person.

Lastly, suppose the subject of conversation neither the speaker nor the party addressed, but some third object, different from both. Here they provided another pronoun. He, she, or it; which, in distinction to the two former, was called the pronoun of the third person.

And thus it was that pronouns came to be distinguished by their respective persons."

As to number, the pronoun of each person has it: I has the plural we, because there may be many speakers at once of the same sentiment; as well as one, who, including himself, speaks the sentiment of many. Thou has the plural you, because a

w The description of the different persons here given is taken from Priscian, who took it from Apollonius. Personæ pronominum sunt tres; prima, secunda, tertia. Prima est, cum ipsa, quæ loquitur, de se pronuntiat; secunda, cum de ea pronunciat, ad quam directo sermone loquitur; tertia, cum de ea, quæ nec loquitur, nec ad se directum accipit sermonem. L. xii. p. 940. Theodore Gaza gives the same distinctions. Πρῶτον (πρόσωπον sc.) ᾧ περὶ ἑαυτοῦ φράζει ὁ λέγων· δεύτερον, ᾧ περὶ τοῦ, πρὸς ἦν ὁ λόγος· τρίτον, ᾧ περὶ ἑτέρου. Gaz. Gram. 1. iv. p. 152.

This account of persons is far preferable to the common one, which makes the first the speaker, the second the party addressed, and the third the subject. For though the first and second be as commonly described, one the speaker, the other the party addressed; yet till they become subjects of the discourse they have no existence. Again, as to the third person's being the subject, this is a character which it shares in common with both the other persons, and which can never, therefore, be called a peculiarity of its own. To explain by an instance or two. When Encas begins the narrative of his adventures, the second person immediately appears, because he makes Dido, whom he addresses, the immediate subject of his discourse.

Infandum, regina, jubes, renovare do

lorem.

From henceforward, for fifteen hundred
verses, (though she be all that time the party
addressed,) we hear nothing further of this
second person, a variety of other subjects
filling up the narrative.

In the mean time, the first person may
be seen everywhere, because the speaker
everywhere is himself the subject. They
were indeed events, as he says himself,

Quæque ipse miserrima vidi,
Et quorum pars magna fui.
Not that the second person does not often
occur in the course of this narrative; but
then it is always by a figure of speech,
when those, who by their absence are in
fact so many third persons, are converted
into second persons by being introduced as
present. The real second person (Dido) is
never once hinted.

Thus far as to Virgil. But when we
read Euclid, we find neither first person nor
second in any part of the whole work. The
reason is, that neither speaker nor party
addressed (in which light we may always
view the writer and his reader) can possibly
become the subject of pure mathematics,
nor indeed can any thing else, except ab-
stract quantity, which neither speaks itself,
nor is spoken to by another.

speech may be spoken to many, as well as to one. plural they, because the subject of discourse is

at once.

He has the

often many

But though all these pronouns have number, it does not appear either in Greek, or Latin, or any modern language, that those of the first and second person carry the distinctions of sex. The reason seems to be, that the speaker and hearer being generally present to each other, it would have been superfluous to have marked a distinction by art, which from nature and even dress was commonly apparent on both sides. But this does not hold with respect to the third person, of whose character and distinctions (including sex among the rest) we often know no more than what we learn from the discourse. And hence it is that in most languages the third person has its genders, and that even English (which allows its adjectives no genders at all) has in this pronoun the triple distinction of he, she, and it.

Hence, too, we see the reason why a single pronoun to each person, an I to the first, and a thou to the second, are abundantly sufficient to all the purposes of speech. But it is not so with respect to the third person. The various relations of the various objects exhibited by this (I mean relations of near and distant, present and absent, same and different, definite and indefinite, &c.) made it necessary that here there should not be one, but many pronouns, such as he, this, that, other, any, some, &c.

It must be confessed, indeed, that all these words do not always appear as pronouns. When they stand by themselves, and represent some noun, (as when we say, This is virtue, or SELKTIKOS, "give me that,") then are they pronouns. But when they are associated to some noun, (as when we say, this habit is virtue; or SEIKTIK@s, "that man defrauded me,") then as they supply not the place of a noun, but only serve to ascertain one,

* Demonstratio ipsa secum genus ostendit. Priscian. l. xii. p. 942. See Apoll. de Syntax. I. ii. c. 7. p. 109.

The utility of this distinction may be better found in supposing it away. Suppose, for example, we should read in history these words: "He caused him to destroy him," and that we were to be informed the he, which is here thrice repeated, stood each time for something different; that is to say, for a man, for a woman, and for a city, whose names were Alexander, Thais, and Persepolis. Taking the pronoun in this manner, divested of its genders, how would it appear which was destroyed, which was the destroyer, and which the cause that moved to the destruction? But there are not such doubts, when we hear the genders distinguished; when, instead of the am

biguous sentence, he caused him to destroy him, we are told, with the proper distinctions, that she caused him to destroy it. Then we know with certainty what before we could not: that the promoter was the woman; that her instrument was the hero; and that the subject of their cruelty was the unfortunate city.

* Quæritur tamen cur prima quidem persona et secunda singula pronomina habeant, tertiam vero sex diversæ indicent voces ? Ad quod respondendum est, quod prima quidem et secunda persona ideo non egent diversis vocibus, quod semper præsentes inter se sunt, et demonstrativæ ; tertia vero persona modo demonstrativa est, ut, hic, iste; modo relativa, ut, is, ipse, &c. Priscian. 1. xii. p. 933.

they fall rather into the species of definitives or articles. That there is, indeed, a near relation between pronouns and articles, the old grammarians have all acknowledged, and some words it has been doubtful to which class to refer. The best rule to distinguish them is this: the genuine pronoun always stands by itself, assuming the power of a noun, and supplying its place; the genuine article never stands by itself, but appears at all times associated to something else, requiring a noun for its support, as much as attributives or adjectives.

As to the coalescence of these pronouns, it is as follows. The first or second will, either of them, by themselves, coalesce with the third, but not with each other. For example, it is good sense, as well as good grammar, to say in any language, I am he, Thou art he; but we cannot say, I am thou, nor Thou art I. The reason is, there is no absurdity for the speaker to be the subject also of the discourse, as when we say, I am he; or for the person addressed, as when we say, Thou art he. But for the same person, in the same circumstances, to be at once the speaker and the party addressed, this is impossible; and so, therefore, is the coalescence of the first and second person.

And now, perhaps, we have seen enough of pronouns, to perceive how they differ from other substantives. The others are primary, these are their substitutes; a kind of secondary race, which were taken in aid, when, for reasons already mentioned,"

* Τὸ ἄρθρον μετὰ ὀνόματος, καὶ ἡ ἀντωvvμía àvτ' òvóμaTos: "the article stands with a noun, but the pronoun stands for a noun. Apol. 1. i. c. 3. p. 22. Auтà obv τὰ ἄρθρα, τῆς πρὸς τὰ ὀνόματα συναρτήσews anоσTávra, eis Thy úñоTeтayuévny avтwvvμlav μeтañíπTE; "now articles themselves, when they quit their connexion with nouns, pass into such pronoun as is proper upon the occasion." Ibid. Again, "Отаν то áрероν μn μer' ovóμatos Tapaλαμβάνηται, ποιήσηται δὲ σύνταξιν ὀνόματος ἣν προεκτεθείμεθα, ἐκ πάσης ἀνάγκης εἰς ἀντωνυμίαν μεταληφθήσεται, εἴγε οὐκ ἐγγινόμενον μετ ̓ ὀνόματος δυνάμει ἀντὶ ovóμaтos πарeλpon: when the article is assumed without the noun, and has (as we explained before) the same syntax which the noun has, it must of absolute necessity be admitted for a pronoun, because it appears without a noun, and yet is in power assumed for one." Ejusd. 1. ii. c. 8. p. 113; 1. i. c. 45. p. 96. Inter pronomina et articulos hoc interest, quod pronomina ea putantur, quæ, cum sola sint, vicem nominis complent, ut quis, ille, iste: articuli vero cum pronominibus, aut nominibus, aut participiis adjunguntur. Donat. Gram. p. 1753.

merantes, finitos ea articulos appellabant ; ipsos autem articulos, quibus nos caremus, infinitos articulos dicebant. Vel, ut alii dicunt, articulos connumerabant pronominibus, et articularia eos pronomina vocabant, &c. Pris. l. i. p. 574. Varro, speaking of quisque and hic, calls them both articles, the first indefinite, the second definite. De Ling. Lat. 1. vii. See also I. ix. p. 132. Vossius, indeed, in his Analogia, (1. i. c. 1.) opposes this doctrine, because hic has not the same power with the Greek article, δ. But he did not enough attend to the ancient writers on this subject, who considered all words as articles, which being associated to nouns (and not standing in their place) served in any manner to ascertain and determine their signification.

b See these reasons at the beginning of this chapter, of which reasons the principal one is, that "no noun, properly so called, implies its own presence. It is therefore to ascertain such presence, that the pronoun is taken in aid; and hence it is it becomes equivalent to deîis, that is, to pointing or indication by the finger." It is worth remarking in that verse of Persius, Sed pulchrum est digito monstrari, et dicier, hic est,

Priscian, speaking of the Stoics, says as follows: Articulis autem pronomina connu- how the deîis and the pronoun are intro