he that has a desire to be informed. Who is it that commands? he that has a will, which he would have obeyed. What are those beings who either wish or pray? those who feel certain wants, either for themselves or others.

If, then, the soul's leading powers be the two above mentioned, and it be true that all speech is a publication of these powers, it will follow that every sentence will be either a sentence of assertion, or a sentence of volition. And thus, by referring all of them to one of these two classes, have we found an expedient to reduce their infinitude.*

The extensions of speech are quite indefinite, as may be seen if we compare the Eneid to an Epigram of Martial. But the longest extension with which grammar has to do, is the extension here considered, that is to say, a sentence. The greater extensions (such as syllogisms, paragraphs, sections, and complete works) belong not to grammar, but to arts of higher order; not to mention that all of them are but sentences repeated.

66

Now a sentence' may be sketched in the following description: a compound quantity of sound significant, of which certain parts are themselves also significant."

Thus when I say "the sun shineth," not only the whole quantity of sound has a meaning, but certain parts also, such as sun" and "shineth."

66

κ Ῥητέον οὖν ὅτι τῆς ψυχῆς τῆς ἡμετέρας διττὰς ἐχούσης δυνάμεις, τὰς μὲν γνωστικὰς, τὰς δὲ ζωτικὰς, τὰς καὶ ὀρεκτικὰς λεγομένας" (λέγω δὲ γνωστικὰς, μὲν, καθ' ἂς γινώσκομεν ἕκαστον τῶν ὄντων, οἷον νοῦν, διάνοιαν, δόξαν, φατασίαν καὶ αἴσθησιν· ὀρεκτικὰς δὲ, καθ ̓ ἃς ὀρεγόμεθα τῶν ἀγαθῶν, ἢ τῶν ὄντων, ἢ τῶν δοκούντων, οἷον βούλησιν λέγω, προαίρεσιν, θυμὸν, καὶ ἐπιθυμίαν) τὰ μὲν τέτταρα εἴδη τοῦ λόγου (τὰ παρὰ τὸν ἀποφαντικὸν) ἀπὸ τῶν ὀρεκτικῶν δυνάμεων προέρχονται τῆς ψυχῆς, οὐκ αὐτῆς καθ ̓ αὑτὴν ἐνεργούσης, ἀλλὰ πρὸς ἕτερον ἀποτεινομένης (τὸν συμβάλλεσθαι δοκούντα πρὸς τὸ τυχεῖν τῆς ὀρέξεως) καὶ ἤτοι λόγον παρ ̓ αὐτοῦ ζητούσης, καθάπερ ἐπὶ τοῦ πυσματικοῦ καὶ ἐρωτηματικοῦ καλουμένου λόγου, ἢ πρᾶγμα, καὶ εἰ πρᾶγμα, ἤτοι αὐτοῦ ἐκείνου τυχεῖν ἐφιεμένης, πρὸς ὃν ὁ λόγος, ὥσπερ ἐπὶ τοῦ κλητικοῦ, ἢ τινὸς παρ' αὐτοῦ πράξεως· καὶ ταύτης, ἢ ὡς παρὰ κρείττονος, ὡς ἐπὶ τῆς εὐχῆς, ἢ ὡς παρὰ χείρονος, ὡς ἐπὶ τοῦ κυρίως καλουμένης προστάξεως μόνον δὲ τὸ ἀποφαντικὸν ἀπὸ τῶν γνωστικῶν, καὶ ἔστι τοῦτο ἐξαγγελτικὸν τῆς γενομένης ἐν ἡμῖν γνώσεως τῶν πραγμάτων ἀληθῶς, ἢ φαινομένως, διὸ καὶ μόνον τοῦτο δεκτικόν ἐστιν ἀληθείας ἢ ψεύδους, τῶν δὲ ἄλλων οὐδέν. The meaning of the above passage being implied in the text, we take its translation from the Latin interpreter. Dicendum

igitur est, cum anima nostra duplicem potestatem habeat, cognitionis, et vitæ, quæ etiam appetitionis ac cupiditatis appellatur, quæ vero cognitionis est, vis est, qua res singulas cognoscimus, ut mens, cogitatio, opinio, phantasia, sensus: appetitus vero facultas est, qua bona, vel quæ sunt, vel quæ videntur, concupiscimus, ut sunt vo luntas, consilium, ira, cupiditas: quatuor orationis species, præter enunciantem, a partibus animi proficiscuntur, quæ concupiscunt; non cum animus ipse per se agit, sed cum ad alium se convertit, qui ei ad consequendum id, quod cupit, conducere posse videatur; atque etiam vel rationem ab eo exquirit, ut in oratione, quam percunctantem, aut interrogantem vocant ; vel rem: sique rem, vel cum ipsum consequi cupit, quicum loquitur, ut in optante oratione, vel aliquam ejus actionem: atque in hac, vel ut a præstantiore, ut in deprecatione ; vel ut ab inferiore, ut in eo, qui proprie jussus nominatur. Sola autem enuncians a cognoscendi facultate proficiscitur : hacque nunciat rerum cognitionem, quæ in nobis est, aut veram, aut simulatam. Itaque hæc sola verum falsumque capit: praeterea vero nuila. Ammon. in Libr. de Interpretatione.

Λόγος δὲ φωνὴ συνθετὴ σημαντική, ἧς ἔνια μέρη καθ ̓ αὑτὰ σημαίνει τι. Arist. See also De Interpret. c. 4.

Poet. c. 20.

But what shall we say? Have these parts again other parts, which are in like manner significant, and so may the progress be pursued to infinite? Can we suppose all meaning, like body, to be divisible, and to include within itself other meanings without end? If this be absurd, then must we necessarily admit that there is such a thing as a sound significant, of which no part is of itself significant. And this is what we call the proper character of a word. For thus, though the words sun and shineth have each a meaning, yet is there certainly no meaning in any of their parts, neither in the syllables of the one, nor in the letters of the other.

m

If, therefore, all speech, whether in prose or verse, every whole, every section, every paragraph, every sentence, imply a certain meaning, divisible into other meanings, but words imply a meaning which is not so divisible; it follows that words will be the smallest parts of speech, inasmuch as nothing less has any meaning at all.

To know, therefore, the species of words, must needs contribute to the knowledge of speech, as it implies a knowledge of its minutest parts.

This, therefore, must become our next inquiry.

CHAPTER III.

CONCERNING THE SPECIES OF WORDS, THE SMALLEST PARTS OF SPEECH.

LET us first search for the species of words among those parts of speech commonly received by grammarians. For example, in one of the passages above cited.

The man that hath no music in himself,

And is not mov'd with concord of sweet sounds,

Is fit for treasons.

Here the word the is an article; man, no, music, concord, sweet, sounds, fit, treasons, are all nouns, some substantive and some adjective; that and himself are pronouns; hath and is are verbs; moved, a participle; not, an adverb; and, a conjunction; in, with, and for, are prepositions. In one sentence we have all

- φωνὴ σημαντική,ἧς μέρος οὐδέν ἐστι Kað avrò onμAVTIKÓV. De Poetic. c. 20. De Interpret. c. 2 and 3. Priscian's definition of a word (lib. ii.) is as follows: Dictio est pars minima orationis constructæ, id est, in ordine compositæ. Pars autem, quantum ad totum intelligendum, id est, ad totius sensus intellectum. Hoc autem

ideo dictum est, nequis conetur vires in duas partes dividere, hoc est, in vi et res; non enim ad totum intelligendum hæc fit divisio. To Priscian we may add Theodore Gaza:— Λέξις δὲ, μέρος ἐλάχιστον κατὰ σύνταξιν Aóyou. Introd. Gram. 1. iv. Plato shewed them this characteristic of a word. See Cratylus, p. 385. edit. Serr.

those parts of speech which the Greek grammarians are found to acknowledge. The Latins only differ in having no article, and in separating the interjection, as a part of itself, which the Greeks include among the species of adverbs.

What then shall we determine? why are there not more species of words? why so many? or if neither more nor fewer, why these and not others?

To resolve, if possible, these several queries, let us examine any sentence that comes in our way, and see what differences we can discover in its parts. For example, the same sentence above,

The man that hath no music, &c.

One difference soon occurs, that some words are variable, and others invariable. Thus the word man may be varied into man's and men; hath, into have, hast, had, &c. Sweet into sweeter and sweetest; fit into fitter and fittest. On the contrary, the words the, in, and, and some others, remain as they are, and cannot be altered.

And yet it may be questioned, how far this difference is essential. For, in the first place, there are variations which can be hardly called necessary, because only some languages have them, and others have them not. Thus the Greeks have the dual variation, which is unknown both to the moderns and to the ancient Latins. Thus the Greeks and Latins vary their, adjectives by the triple variation of gender, case, and number; whereas the English never vary them in any of those ways, but through all kinds of concord preserve them still the same. Nay, even those very variations, which appear most necessary, may have their places supplied by other methods; some by auxiliars, as when for Bruti, or Bruto, we say "of Brutus," "to Brutus;" some by mere position, as when for Brutum amavit Cassius, we say, "Cassius loved Brutus." For here the accusative, which in Latin is known any where from its variation, is in English only known from its position or place.

If, then, the distinction of variable and invariable will not answer our purpose, let us look further, for some other more essential.

Suppose, then, we should dissolve the sentence above cited, and view its several parts as they stand separate and detached. Some, it is plain, still preserve a meaning, (such as man, music, sweet, &c.) others, on the contrary, immediately lose it, (such as and, the, with, &c.) Not that these last have no meaning at all, but in fact they never have it, but when in company or associated.

Now it should seem that this distinction, if any, was essential. For all words are significant, or else they would not be words; and if every thing not absolute is of course relative, then will all words be significant either absolutely or relatively.

With respect, therefore, to this distinction, the first sort of words may be called significant by themselves; the latter may be called significant by relation; or if we like it better, the first sort may be called principals, the latter accessories. The first are like those stones in the basis of an arch, which are able to support themselves, even when the arch is destroyed; the latter are like those stones in its summit or curve, which can no longer stand, than while the whole subsists."

This distinction being admitted, we thus pursue our speculations. All things whatever either exist as the energies or affections of some other thing, or without being the energies or affections of some other thing. If they exist as the energies or affections of something else, then are they called attributes. Thus to think is the attribute of a man; to be white, of a swan; to fly, of an eagle; to be four-footed, of a horse. If they exist not after this manner, then are they called substances.o Thus man, swan, eagle, and horse, are none of them attributes, but all substances, because however they may exist in time and place, yet neither of these, nor of any thing else, do they exist as energies or affections.

And thus all things whatsoever, being either substances or attributes, it follows of course that all words which are significant as principals, must needs be significant of either the one or the other. If they are significant of substances, they are called substantives; if of attributes, they are called attributives. So that all words whatever, significant as principals, are either substantives or attributives.

Again, as to words, which are only significant as accessories,

Apollonius of Alexandria (one of the acutest authors that ever wrote on the subject of grammar) illustrates the different power of words, by the different power of letters. Ετι, ὃν τρόπον τῶν στοιχείων τὰ μέν ἐστι φωνήεντα, ἃ καὶ καθ ̓ ἑαυτὰ φωνὴν ἀποτελεῖ τὰ δὲ σύμφωνα, ἅπερ ἄνευ τῶν φωνηέντων οὐκ ἔχει ῥητὴν τὴν ἐκφώνησιν. τὸν αὐτὸν τρόπον ἐστὶν ἐπινοῆσαι καὶ ἔπὶ τῶν λέξεων. αἱ μὲν γὰρ αὐτῶν, τρόπον τινὰ τῶν φωνηέντων, ῥηταί εἰσι· καθάπερ ἐτὶ τῶν ῥημάτων, ὀνομάτων, ἀντωνυμιῶν, ἐπιβῥημάτων-αἱ δὲ, ὥσπερεί σύμφωνα, ἀναμένουσι τὰ φωνήεντα, οὐ δυνάμενα, κατ' ἰδίαν ῥητὰ εἶναι—καθάπερ ἐπὶ τῶν προθέσεων, τῶν ἄρθρων, τῶν συνδέσμων· τὰ γὰρ τοιαῦτα ἀεὶ τῶν μορίων συσσημαίνει. “ In the same manner, as of the elements or letters, some are vowels, which of themselves complete a sound; others are consonants, which, without the help of vowels, have no express vocality; so likewise may we conceive as to the nature of words. Some of them, like vowels, are of themselves expressive, as is the case of verbs, nouns, pronouns, and adverbs; others, like con

sonants, wait for their vowels, being unable to become expressive by their own proper strength, as is the case of prepositions, articles, and conjunctions; for these parts of speech are always consignificant, that is, are only significant when associated to something else.” Apollon. de Syntaxi, l. 1. c. 3. Itaque quibusdam philosophis placuit nomen et verbum solas esse partes orationis ; cætera vero, adminicula vel juncturas earum: quomodo navium partes sunt tabulæ et trabes, caetera autem (id est, cera, stuppa, et clavi et similia) vincula et conglutinationes partium navis (hoc est, tabularum et trabium) non partes navis dicuntur. Prisc. l. xi. 913.

• Thus Aristotle: Νῦν μὲν οὖν τύπῳ εἴρηται, τί ποτ' ἐστὶν ἡ οὐσία, ὅτι τὸ μὴ καθ ̓ ὑποκειμένου, ἀλλὰ καθ ̓ οὗ τὰ ἄλλα. Metaph. Z. y. p. 106. edit. Sylb.

See

P This division of things into substance and attribute seems to have been admitted by philosophers of all sects and ages. Categor. c. 2. Metaphys. 1. vii. c. 1. De Cœlo, 1. iii. c. 1.

they acquire a signification either from being associated to one word, or else to many. If to one word alone, then, as they can do no more than in some manner define or determine, they may justly for that reason be called definitives. If to many words at once, then, as they serve to no other purpose than to connect, they are called for that reason by the name of connectives.

And thus it is that all words whatever are either principals or accessories; or under other names, either significant from themselves, or significant by relation. If significant from themselves, they are either substantives or attributives; if significant by relation, they are either definitives or connectives. So that under one of these four species, substantives, attributives, definitives, and connectives, are all words, however different, in a manner included.

If any of these names seem new and unusual, we may introduce others more usual, by calling the substantives, nouns; the attributives, verbs; the definitives, articles; and the connectives, conjunctions.

Should it be asked, what then becomes of pronouns, adverbs, prepositions, and interjections? the answer is, either they must be found included within the species above mentioned, or else must be admitted for so many species by themselves.

There were various opinions in ancient days, as to the number of these parts, or elements of speech.

Plato, in his Sophist, mentions only two, the noun and the verb. Áristotle mentions no more, where he treats of prepositions. Not that those acute philosophers were ignorant of the other parts, but they spoke with reference to logic or dialectic,' considering the essence of speech as contained in these two, because these alone combined make a perfect assertive sentence, which none of the rest without them are able to effect. Hence, therefore, Aristotle, in his treatise of Poetry,' (where he was to lay down the elements of a more variegated speech,) adds the article and conjunction to the noun and verb, and so adopts the same parts with those established in this treatise. To Aristotle's

4 Vol. i. p. 261. edit. Scr. r De Interpr. c. 2, 3.

Partes igitur orationis sunt secundum dialecticos duæ, nomen et verbum ; quia hæ solæ etiam per se conjunctæ plenam faciunt orationem; alias autem partes σvyKатпуорhμатα, hoc est, consignificantia appellabant. Priscian. 1. ii. p. 574. edit. Putschii. Existit hic quædam quæstio, cur duo tantum, nomen et verbum, se (Aristoteles sc.) determinare promittat, cum plures partes orationis esse videantur. Quibus hoc dicendum est, tantum Aristotelem hoc libro diffinisse, quantum illi ad id, quod instituerat tractare, suffecit. Tractat namque de

simplici enuntiativa oratione, quæ scilicet hujusmodi est, ut junctis tantum verbis et nominibus componatur.-Quare superfluum est quærere, cur alias quoque, quæ videntur orationis partes, non proposuerit, qui non totius simpliciter orationis, sed tantum simplicis orationis instituit elementa partiri. Boetius in Libr. de Interpretat. p. 295. Apollonius, from the above principles, ele gantly calls the noun and verb, Tà éμYux Tara μépn Toù λóyou," the most animated parts of speech." De Syntaxi, l. i. c. 3. p. 24. See also Plutarch. Quæst. Platon. p. 1009.

Poet. cap. 20.

66