the mind alone. Whence, then, is our power of recalling them with such marvellous precision and facility? How is it that we can keep them safely apart in the mind, instead of being obliged to look for them mingled and confused, in the objects from which we first disentangled them by reflection? By virtue of the name we have attached to each of them; which, like the labels upon the chemist's jars or the gardener's flowerpots, enable us at once to identify and secure the property we seek. Names then are the means of fixing and recording the result of trains of thought, which without them must be repeated frequently, with all the pain of the first effort.*

§ 25. (iii.) Leibnitz was the first, so far as I know, to call attention to the fact that words are sometimes more than signs of thought; that they may become thoughts. His distinction between symbolical and intuitive [notative] conceptions † conducts us to the third function of language, that it abbreviates the processes of thought. Where our notion of any object or objects consists of a clear insight into all the attributes, or at least the essential ones, he would call it intuitive. But where the notion is complex, and

* Upon this, consult Damiron, Logique, p. 200. seq. and Dural-Jouve, Logique, p. 199. seq.; Mill on the Human Mind, vol. i. p. 86.

† Erdmann's Ed. p. 79. Acta Erudit. an. 1684.

its properties numerous, we do not commonly realise all that it conveys; the process of thinking would be needlessly retarded by such a review. We make use of the name commonly given to the notion as a symbol, even for ourselves, of all the properties it possesses. A name then, employed in thought, is called a symbolical cognition; and the names we employ in speech are not always symbols to another of what is explicitly understood by us, but quite as often are symbols both to speaker and hearer, the full and exact meaning of which neither of them stop to unfold, any more than they regularly reflect that every sovereign which passes through their hands is equivalent to 240 pence. Such words as the state, happiness, liberty, creation, are too pregnant with meaning for us to suppose that we realise their full sense every time we read or pronounce them. If we attend to the working of our minds we shall find that each word may be used, and in its proper place and sense, though perhaps few or none of its attributes are present to us at the moment. A very simple notion is always intuitive; we cannot make our notion of brown or red simpler than it is, by any symbol. On the other hand a highly complex notion, like those named above, is seldom fully realised-seldom other than symbolical. Here then is a farther use of names ; they serve to abbreviate the process of thought, as we have seen that they are useful in recording its re

sults. And it may be noticed here that this distinction of cognitions throws a new light on the nature of definitions, or explanatory propositions, which are not, as they are often regarded, mere explanations to others of a meaning which we ourselves duly apprehend, but are real acts of thought, which by unfolding before us some marks of our conception, partially or wholly unseen by us, have all the power of new truths even for ourselves.

§ 26. (iv.) That language hath a fourth use, the most obvious of all, as the medium of communication between mind and mind, needs no explanation. We might dispense with articulate speech for certain purposes, and might make gestures and changes of the countenance, which are the language of action, supply its place. But actions and the play of features, whilst they serve to express love or hatred for some present object, need of food or rest, joy or sorrow, can but express a very small and confined list of thoughts if we would indicate our feelings towards some absent person, or our wish for something at a distance, or direct attention to some inward state or sentiment; we cannot guide the thoughts of the spectator to the object present to our own mind, with any precision and certainty. Hence, it is necessary to appropriate to every object a signal, always available, which all men by a tacit convention accept as a substitute for the object, and which therefore recalls

1

the object to the fancy whenever it is employed; and such a signal is a noun or name, defined by Aristotle to be "a sound which by convention is significant, but does not determine time."* The convention or agreement by which a whole nation confines a noun to one object or class of objects, is of course merely tacit; whatever theory of the origin of language we adopt, we cannot suppose that a nation ever formally met and agreed upon the several names that should thenceforward express their various notions. Language is based upon general agreement, if we give our assent to its use every day by hearing and answering it, just as truly as if the view of Maupertuis were correct, that language was originally formed by a session of learned societies. Names however are representatives of things; and the different states of things must find an expression likewise; hence the need of adjectives and verbs. The verb has the power of assigning to the thing at a particular time the condition of being, doing, or

* Ονομα μὲν οὖν ἐστὶ φωνὴ σημαντικὴ κατὰ συνθήκην ἄνευ χρόνου, ἧς μηδὲν μέρος ἐστὶ σημαντικὸν κεχωρισμένον. — On Enouncement, ch. 2. (The last words express that it divides into syllables only, and not words, otherwise it would be a sentence.) 'Pîμa (verb) dé éσti td πpos onμaîvov xpóvov.—Ch. 3. J. C. Scaliger traced the distinction between the noun and the verb to a difference of time, for the noun represented a permanent thing, the verb a temporary and transitory state.

undergoing something; but as every verb may be resolved into an adjective-notion, and one particular word simply expressive of past or present or future state, as for example, "he loved" is explained by "he was-loving," "he hopes" by "he is-hoping," we are justified in regarding all verbs as fundamentally one, the verb to be, with its three times or tenses of is, was, shall be, and their variety as arising from the incorporation of various adjective-notions with this simple verbal element. When two or more names come together, it is frequently necessary to express the mutual relation in which they stand; a thing may be to, from, by, in, near, above, or below another, and prepositions are invented to determine this. Here then are the four principal parts of speech, substantives, or names to express substances, adjectives to stand for attributes, prepositions to denote relations, and a single verb to assign attributes or relations to substantives at a determinate time.*

§ 26. Aristotle's mode of arranging the classes of words admits of a brief, and (it may be hoped) intelligible statement. Words are conventional signs

* See Condillac, Grammaire, ch. viii. The more advanced student will not fail to notice that as the ten Categories of Aristotle answer to the parts of speech, so the simpler division of categories adopted by many later writers, into substance, attribute and relation, answers to three parts of speech. See below, the Section on Categories.