must necessarily be found in all languages. The most simple and comprehensive division of the parts of speech is, into substantives, attributives, and connectives.* Substantives, are all the words which express the names of objects, or the subjects of discourse; attributes, are all the words which express any attribute, property or action of the former; connectives, are what express the connections, relations, and dependeneies, which take place among them. The common grammatical division of speech into eight parts; nouns, pronouns, verbs, participles, adverbs, prepositions, interjections, and conjunctions, is not very logical, as might be easily shewn; as it comprehends, under the general term of nouns, both substantives and adjectives, which are parts of speech generically and essentially distinct; while it makes a separate part of speech of participles, which are no other than verbal adjectives. However, as these are the terms to which our ears have been most familiarized, and, as an exact logical division is of no great consequence to our present purpose, it will be better to make use of these known terms than of any

other.

We are naturally led to begin with the consideration of substantive nouns, which are the foundation of all grammar, and may be considered as the most ancient part of speech. For, assuredly, as soon as men had got beyond simple interjections, or exclamations of passion, and began to communicate themselves by discourse, they would be under a necessity of assigning names to the objects they saw around them, which in grammatical language, is called the invention of substantive nouns.*~~ And here, at our first setting out, somewhat curious occurs. The individual objects which surround us, are infinite in number. A savage, wherever he looked, beheld forests and trees. To give separate names to every

Quintilian informs us, that this was the most ancient division. “Tum videbit quot et quæ sunt partes orationis. Quanquam de numero parum convenit. Veteres enim, quo"rum fuerant Aristoteles atque Theodictes, verba modo, et nomina, et convinetiones tra"diderunt. Videlicet, quod in verbis vim sermonis, in nominibus materiam, (quia alterum "est quod loquimur, alterum de quo loquimur) in convinctionibus autem complexum “eoram esse judicarant ; quas conjunctiones a plerisque dici scio; sed hæc videtur ex "ouiderua magis propria translatio. Paulatim a philosophicis ac maximè a stoicis, auctus "est numerus; ac primum convinctionibus articuli adjecti, post præpositiones; nominibus, "appellatio, deinde pronomen; deinde mistum verbo participium; ipsis verbis, adverbia.” Lib. i. cap. iv.

I do not mean to assert, that among all nations, the first invented words were simple and regular substantive nouns. Nothing is more difficult and uncertain, than to ascertain the precise steps by which men proceeded in the formation of language. Names for objects must, doubtless, have arisen in the most early stages of speech. But, it is probable, as the learned author of the Treatise on the Origin and Progress of Language, has shown (vol. i. p. 371. 395) that, among several satage tribes, some of the first articulate sounds that were formed, denoted a whole sentence, rather than the name of a particular object; conveying some information, or expressing some desires or fears suited to the circumstances in which that tribe was placed, or relating to the business they had most frequent occasion to carry on; as, the lion is coming, the river is swelling, &c. Many of their first words, it is likewise probable, were not simple substantive nouns, but substantives, accompanied with some of those attributes, in conjunction with which they were most frequently accustomed to behold them; as, the great bear, the little hut, the wound made by the hatchet. &c. Of all which, the author produces instances from several of the American languages; and it is, undoubtedly, suitable to the natural course of the operations of the human mind, thus to begin with particulars the most obvious to sense, and to proceed, from these, to more general expressions. He likewise observes, that the words of those primitive tongues are far from being, as we might suppose them, rude and short, and crowded with consonants; but, on the contrary, are, for the most part, long words, and full of vowels.

This is the consequence of their being formed upon the natural sound which the voice utters with most ease, a little varied and distinguished by articulation; and he shews this to hold, in fact, among most of the barbarous languages which are known.

one of those trees, would have been an endless and impracticable undertaking. His first object was to give a name to that particular tree, whose fruit relieved his hunger, or whose shade protected him from the sun. But observing, that though other trees were distinguished from this by peculiar qualities of size or appearance, yet that they also agreed and resembled one another, in certain common qualities, such as springing from a root, and bearing branches and leaves, he formed in his mind some general idea of those common qualities, and ranging all that possessed them under one class of objects, he called that whole class, a tree. Longer experience taught him to subdivide this genus into the several species of oak, pine, ash, and the rest, according as his observation extended to the several qualities in which these trees agreed or differed.

But, still, he made use only of general terms in speech. For the oak, the pine, and the ash, were names of whole classes of objects; each of which included an immense number of undistinguished individuals. Here then it appears, that though the formation of abstract, or general conceptions, is supposed to be a difficult operation of the mind; such conceptions must have entered into the very first formation of language. For, if we except only the proper names of persons, such as Cæsar, John, Peter, all the other substantive nouns which we employ in discourse, are the names, not of individual objects, but of very extensive genera, or species of objects; as man, lion, house, river, &c. We are not, however, to imagine that this invention of general, or abstract terms, requires any great exertion of metaphysical capacity: for, by whatever steps the mind proceeds in it, it is certain that, when men have once observed resemblances among objects, they are naturally inclined to call all those which resembles one another, by one common name; and, of course, to class them under one species. We may daily observe this practised by children in their first attempts towards acquiring language.

But now, after language had proceeded as far as I have described, the notification which it made of objects was still very imperfect: for, when one mentioned to another in discourse, any substantive noun; such as, man, lion, or tree, how was it to be known which man, which lion, or which tree he meant, among the many comprehended under one name? Here occurs a very curious, and a very useful contrivance for specifying the individual object intended, by means of that part of speech called the article.

The force of the article consists in pointing or singling out from the common mass, the individual of which we mean to speak. In English we have two articles, a and the; a is more general and unlimited; the more definite and special. A is much the same with one, and marks only any one individual of a species; that individual being either unknown or left undetermined; as, a lion, a king.-The, which possesses more properly the force of the article, ascertains some known or determined individual of the species; as, the lion, the king.

Articles are words of great use in speech. In some languages, however, they are not found. The Greeks have but one article, ¿T, which answers to our definite, or proper article, the. They have no word which answers to our article a, but they supply its place by the absence of their article: Thus, Barius signifies a king; Barius, the king. The Latins have no article. In the room of it, they employ pronouns ; as, hic, ille, iste, for pointing out the objects which they want to distinguish. "Noster sermo," says Quintilian, "articulos non desiderat,

K

" ideoque in alias partes orationis sparguntur." This however, appears to me a defect in the Latin tongue as articles contribute much to the clearness and precision of language.

In order to illustrate this, remark what difference there is in the meaning of the following expressions in English, depending wholly on the different employment of the articles; "the son of a king. The son of "the king. A son of the king's." Each of these three phrases has an entirely different meaning, which I need not explain, because any one who understands the language, conceives it clearly at first hearing, through the different application of the articles a and the. Whereas, in Latin, "filius regis," is wholly undetermined; and to explain, in which of these three senses it is to be understood, for it may bear any of them, a circumlocution of several words must be used, In the same manner, "are you a king ?" "are you the king ?" are questions of quite different import: which, however, are confounded together in the Latin phrase, "esne tu rex ?" "thou art a man," is a very general and harmless position; but, thou art the man,' is an assertion capable, we know, of striking terror and remorse into the heart. These observations illustrate the force and importance of articles; and at the same time, I gladly lay hold of any opportunity of shewing the advantages of our own language. Besides this quality of being particularised by the article, three affections belong to substantive nouns, number, gender, and case, which require our consideration.

Number distinguishes them as one, or many, of the same kind, called the singular and plural; a distinction found in all languages, and which must, indeed, have been coeval with the very infancy of language; as there were few things which men had more frequent occasion to express, than the difference between one and many. For the greater facility of expressing it, it has, in all languages, been marked by some variation made upon the substantive noun; as we see, in English, our plural is commonly formed by the addition of the letter S. In the Hebrew, Greek, and some other ancient languages, we find not only a plural, but a dual number; the rise of which may very naturally be accounted for, from separate terms of numbering not being yet invented, and one, two, and many, being all, or at least, the chief numeral distinctions which men, at first, had any occasion to take notice of.

Gender, is an affection of substantive nouns, which will lead us inte more discussion than number. Gender, being founded on the distinction of the two sexes, it is plain, that in a proper sense, it can only find place in the names of living creatures, which admit the distinction of male and female; and, therefore, can be ranged under the masculine or feminine genders. All other substantive nouns ought to belong to what grammarians call, the neuter gender, which is meant to imply the negation of either sex. But with respect to this distribution, somewhat singular hath obtained in the structure of language. For, in correspondence to that distinction of male and female sex, which runs through all the classes of animals, men have, in most languages, ranked a great number of inanimate objects also, under the hike distinctions of masculine and feminine. Thus we find it, both in the Greek and Latin tongues. Gladius, a sword, for instance, is masculine; sagitta, an arrow, is feminine; and this assignation of sex to inanimate objects, this distinction of them into masculine and feminine, appears often to be entirely capricious; derived from no other principle than the casual structure of the language, which re

fers to a certain gender, words of a certain termination. In the Greek and Latin, however, all inanimate objects are not distributed into masculine and feminine; but many of them are also classed, where all of them ought to have been, under the neuter gender; as, templum, a church; sedile, a seat.

But the genius of the French and Italian tongues differs, in this respect, from the Greek and Latin. In the French and Italian, from whatever cause it has happened, so it is, that the neuter gender is wholly unknown, and that all their names of inanimate objects are put upon the same footing with living creatures; and distributed, without exception, into masculine and feminine. The French have two articles, the mascufine le, and the feminine la; and one or other of these is prefixed to all substantive nouns in the language, to denote their gender. The Italians make the same universal use of their articles il, and lo, for the masculine; and la, for the feminine.

In the English language, it is remarkable that there obtains a peculiarity quite opposite. In the French and Italian, there is no neuter gender. In the English, when we use common discourse, all substantive nouns, that are not names of living creatures, are neuter without exception. He, she, and it, are the marks of the three genders; and we always use it, in speaking of any object where there is no sex, or where the sex is not known. The English is, perhaps, the only language in the known world (except the Chinese, which is said to agree with it in this particular) where the distinction of gender is properly and philosophically applied in the use of words, and confined as it ought to be, to mark the real distinctions of male and female.

Hence arises a very great and signal advantage of the English tongue, which it is of consequence to remark.* Though in common discourse, as I have already observed, we employ only the proper and literal distinction of sexes; yet the genius of the language permits us, whenever it will add beauty to our discourse, to make the names of inanimate objects masculine or feminine in a metaphorical sense; and when we do so, we are understood to quit the literal style, and to use one of the figures of discourse.

For instance; if I am speaking of virtue, in the course of ordinary conversation, or of strict reasoning, I refer the word to no sex or gender; I say, "virtue is its own reward;" or, "it is the law of our nature." But if I choose to rise into a higher tone; if I seek to embellish and animate my discourse, I give a sex to virtue; I say, "she descends from "heaven; "she alone confers true honor upon man ;"" her gifts are the only durable rewards." By this means we have it in our power to to vary our style at pleasure. By making a very slight alteration, we can personify any object that we choose to introduce with dignity; and by this change of manner, we give warning that we are passing from the strict and logical, to the ornamented and rhetorical style.

This is an advantage which, not only every poet, but every good writer and speaker in prose, is, on many occasions, glad to lay hold of, and improve; and it is an advantage peculiar to our tongue; no other language possesses it. For, in other languages, every word has one fixed gender, masculine, feminine, or neuter, which can, upon no occasion, be changed; agern, for instance, in Greek, virtus, in Latin, and la vertu

The following observations on the metaphorical use of genders, in the English language, are taken from Mr. Harris's Hermes.

in French, are uniformly feminine., She, must always be the pronoun answering to the word, whether you be writing in poetry or in prose, whether you be using the style of reasoning, or that of declamation : whereas, in English, we can either express ourselves with the philosophical accuracy of giving no gender to things inanimate; or by giving them gender, and transforming them into persons, we adapt them to the style of poetry, and, when it is proper, we enliven prose.

It deserves to be further remarked on this subject, that when we employ that liberty which our language allows, of ascribing sex to any inanimate object, we have not, however, the liberty of making it of what gender we please, masculine or feminine; but are, in general, subjected to some rule of gender which the currency of language has fixed to that object. The foundation of that rule is imagined, by Mr. Harris, in his Philosophical Inquiry into the Principles of Grammar," to be laid in a certain distant resemblance, or analogy, to the natural distinction of the two sexes.

66

Thus, according to him, we commonly give the masculine gender to those substantive nouns used figuratively, which are conspicuous for the attributes of imparting, or communicating; which are by nature strong and efficacious, either to good or evil; or which have a claim to some eminence, whether laudable or not. Those again, he imagines, to be generally made femininę, which are conspicuous for the attributes of containing, and of bringing forth; which have more of the passive in their nature, than of the active; which are peculiarly beautiful, or amiable; or which have respect to such excesses as are rather feminine than masculine. Upon these principles he takes notice, that the sun is always put in the masculine gender with us, the moon in the feminine, as being the receptacle of the sun's light. The earth is, universally, feminine. A ship, a country, a city, are likewise made feminine, as receivers, or containers. God, in all languages, is masculine. Time, we make masculine on account of its mighty efficacy; virtue, feminine, from its beauty and its being the object of love. Fortune is always feminine. Mr. Harris imagines, that the reasons which determine the gender of such capital words as these, hold in most other languages, as well as the English. This, however, appears doubtful. A variety of circumstances, which seem casual to us, because we cannot reduce them to principles, must, unquestionably, have influenced the original formation of languages: and in no article whatever does language appear to have been more capricious, and to have proceeded less according to fixed rule, than in the imposition of gender upon things inanimate; especially among such nations as have applied the distinction of masculine and feminine to all substantive nouns.

Having discussed gender, I proceed, next, to another remarkable peculiarity of substantive nouns, which in the style of grammar, is called their declension by cases. Let us, first, consider what cases signify. In order to understand this, it is necessary to observe, that, after men had given names to external objects, had particularized them by means of the article, and distinguished them by number and gender, still their language remained extremely imperfect, till they had devised some method of expressing the relations which those objects bore, one towards another. They would find it of little uso to have a name for man, lion, tree, river, without being able, at the same time, to signify how these stood with respect to each other: whether, as approaching to, receding from, joined