All primitive and uncompounded languages seem to have a dual, as well as a plural number. This is the case of the Greek, and I am told of the Hebrew, of the Gothic, and of many other languages. In the rude beginnings of society, one, two, and more, might possibly be all the numeral distinctions which mankind would have any occasion to take notice of. These they would find it more natural to express, by a variation upon every particular noun substantive, than by such general and abstract words as one, two, three, four, &c. These words, though custom has rendered them familiar to us, express, perhaps, the most subtile and refined abstractions, which the mind of man is capable of forming. Let any one consider within himself, for example, what he means by the word three, which signifies neither three shillings, nor three pence, nor three men, nor three horses, but three in general; and he will easily satisfy himself that a word, which denotes so very metaphysical an abstraction, could not be either a very obvious or a very early invention. I have read of some savage nations, whose language was capable of expressing no more than the three first numeral distinctions. But whether it expressed those distinctions by three general words, or by variations upon the nouns substantive, denoting the things numbered, I do not remember to have met with any thing which could determine.
As all the same relations which subsist between single, may likewise subsist between numerous objects, it is evident there would be occasion for the same number of cases in the dual and in the plural, as in the singular number. Hence the intricacy and complexness of the declensions in all the ancient languages. In the Greek there are five cases in each of the three numbers, consequently fifteen in all.
As nouns adjective, in the ancient languages, varied their terminations according to the gender of the substantive to which they were applied, so did they likewise, according to the case and the number. Every noun adjective in the Greek language, therefore, having three genders, and three numbers, and five cases in each number, may be considered as having five and forty different variations. The first formers of language seem to have varied the termination of the adjective, according to the case and the number of the substantive, for the same reason which made them vary according to the gender; the love of analogy, and of a certain regularity of sound. In the signification of adjectives there is neither case nor number, and the meaning of such words is always precisely the same, notwithstanding all the variety of termination under which they appear. Magnus vir, magni viri, magnorum virorum; a great man, of a great man, of great men in all these expressions the words magnus, magni, magnorum, as well as the word great, have precisely one and the same signification, though the substantives to which they are applied have not. The difference of termination in the noun adjective is accompanied with no sort of difference in the meaning. An adjective denotes the qualification of a noun substantive. But the different relations in which that noun substantive may occasionally stand, can make no sort of difference upon its qualification.
If the declensions of the ancient languages are so very complex, their conjugations are infinitely more so. And the complexness of the one is founded upon the same principle with that of the other, the difficulty of forming, in the beginnings of language, abstract and general terms.
Verbs must necessarily have been coeval with the very first attempts towards the formation of language. No affirmation can be expressed without the assistance of some verb. We never speak but in order to express our opinion that something either is or is not. But the word denoting this event, or this matter of fact, which is the subject of our affirmation, must always be a verb.
Impersonal verbs, which express in one word a complete event, which preserve in the expression that perfect simplicity and unity, which there always is in the object and in the idea, and which suppose no abstraction, or metaphysical division of the event into its several constituent members of subject and attribute, would, in all probability, be the species of verbs first invented. The verbs pluit, it rains; ningit, it snows; tonat, it thunders; lucet, it is day; turbatur, there is a confusion, &c. each of them express a complete affirmation, the whole of an event, with that perfect simplicity and unity with which the mind conceives it in nature. On the contrary, the phrases, Alexander ambulat, Alexander walks; Petrus sedet, Peter sits, divide the event, as it were, into two parts, the person or subject, and the attribute, or matter of fact, affirmed of that subject. But in nature, the idea or conception of Alexander walking, is as perfectly and completely one single conception, as that of Alexander not walking. The division of this event, therefore, into two parts, is altogether artificial, and is the effect of the imperfection of language, which, upon this, as upon many other occasions, supplies, by a number of words, the want of one, which could express at once the whole matter of fact that was meant to be affirmed. Every body must observe how much more simplicity there is in the natural expression, pluit, than in the more artificial expressions, imber decidit, the rain falls; or, tempestas est pluvia, the weather is rainy. In these two last expressions, the simple event, or matter of fact, is artificially split and divided, in the one, into two; in the other, into three parts. In each of them it is expressed by a sort of grammatical circumlocution, of which the significancy is founded upon a certain metaphysical analysis of the component parts of the idea expressed by the word pluit. The first verbs, therefore, perhaps even the first words, made use of in the beginnings of language, would in all probability be such impersonal verbs. It is observed accordingly, I am told, by the Hebrew Grammarians, that the radical words of their language, from which all the others are derived, are all of them verbs, and impersonal verbs.
It is easy to conceive how, in the progress of language, those impersonal verbs should become personal. Let us suppose, for example, that the word venit, it comes, was originally an impersonal verb, and that it denoted, not the coming of something in general, as at present, but the coming of a particular object, such as the Lion. The first savage inventors of language, we shall suppose, when they observed the approach of this terrible animal, were accustomed to cry out to one another, venit, that is, the lion comes; and that this word thus expressed a complete event, without the assistance of any other. Afterwards, when, on the further progress of language, they had begun to give names to particular substances, whenever they observed the approach of any other terrible object, they would naturally join the name of that object to the word venit, and cry out, venit ursus, venit lupus. By degrees the word venit would thus come to signify the coming of any terrible object, and not merely the coming of the lion. It would now therefore, express, not the coming of a particular object, but the coming of an object of a particular kind. Having become more general in its signification, it could no longer represent any particular distinct event by itself, and without the assistance of a noun substantive, which might serve to ascertain and determine its signification. It would now, therefore, have become a personal, instead of an impersonal verb. We may easily conceive how, in the further progress of society, it might still grow more general in its signification, and come to signify, as at present, the approach of any thing whatever, whether good, bad, or indifferent.
It is probably in some such manner as this, that almost all verbs have become personal, and that mankind have learned by degrees to split and divide almost every event into a great number of metaphysical parts, expressed by the different parts of speech, variously combined in the different members of every phrase and sentence.[[29]] The same sort of progress seems to have been made in the art of speaking as in the art of writing. When mankind first began to attempt to express their ideas by writing, every character represented a whole word. But the number of words being almost infinite, the memory found itself quite loaded and oppressed by the multitude of characters which it was obliged to retain. Necessity taught them, therefore, to divide words into their elements, and to invent characters which should represent, not the words themselves, but the elements of which they were composed. In consequence of this invention, every particular word came to be represented, not by one character, but by a multitude of characters; and the expression of it in writing became much more intricate and complex than before. But though particular words were thus represented by a greater number of characters, the whole language was expressed by a much smaller, and about four and twenty letters were found capable of supplying the place of that immense multitude of characters, which were requisite before. In the same manner, in the beginnings of language, men seem to have attempted to express every particular event, which they had occasion to take notice of, by a particular word, which expressed at once the whole of that event. But as the number of words must, in this case, have become really infinite, in consequence of the really infinite variety of events, men found themselves partly compelled by necessity, and partly conducted by nature, to divide every event into what may be called its metaphysical elements, and to institute words, which should denote not so much the events, as the elements of which they were composed. The expression of every particular event, became in this manner more intricate and complex, but the whole system of the language became more coherent, more connected, more easily retained and comprehended.
[29]. As the far greater part of Verbs express, at present, not an event, but the attribute of an event, and, consequently, require a subject, or nominative case, to complete their signification, some grammarians, not having attended to this progress of nature, and being desirous to make their common rules quite universal, and without any exception, have insisted that all verbs required a nominative, either expressed or understood; and have, accordingly put themselves to the torture to find some awkward nominatives to those few verbs, which still expressing a complete event, plainly admit of none. Pluit, for example, according to Sanctius, means pluvia pluit, in English, the rain rains. See Sanctii Minerva, l. 3. c. 1.
When verbs, from being originally impersonal had thus, by the division of the event into its metaphysical elements, become personal, it is natural to suppose that they would first be made use of in the third person singular. No verb is ever used impersonally in our language, nor, so far as I know, in any other modern tongue. But in the ancient languages, whenever any verb is used impersonally, it is always in the third person singular. The termination of those verbs, which are still always impersonal, is constantly the same with that of the third person singular of personal verbs. The consideration of these circumstances, joined to the naturalness of the thing itself, may serve to convince us that verbs first became personal in what is now called the third person singular.