s the art of speaking or writing any language with correctness and propriety; and the purpose of language is to communicate our thoughts.

Grammar, considered as an art, necessarily supposes the previous existence of language; and as its design is to teach any language to those who are ignorant of it, it must be adapted to the genius of that particular language of which it treats. A just method of grammar, therefore, without attempting any alterations in a language already introduced, furnishes certain observations called rules, to which the methods of speaking used in that language may be reduced; and this collection of rules is called the grammar of that particular language. For the greater distinctness with regard to these rules, grammarians have usually divided this subject into four distinct heads: Orthography, or the art of combining letters into syllables, and syllables into words; Etymology, or the art of deducing one word from another, and the various modifications by which the sense of any one word can be diversified consistently with its original meaning or its relation to the theme whence it is derived; Syntax, or what relates to the construction or due disposition of the words of a language into sentences or phrases; and Prosody, or that which treats of the quantities and accents of syllables, and the art of making verses.

But grammar, considered as a science, views language only as significant of thought. Neglecting particular and arbitrary modifications introduced for the sake of beauty or elegance, it examines the analogy and relation between words and ideas; distinguishes between those particulars which are essential to language and those which are only accidental; and thus furnishes a certain standard, by which different languages may be compared, and their several excellencies or defects pointed out. This is what is called Philosophical or

**Universal Grammar**

The origin of language is a subject which has employed much learned investigation, and about which there is still a diversity of opinion. The design of speech is to communicate to others the thoughts and perceptions of the mind of the speaker; but it is obvious, that between an internal idea and any external sound there is no natural relation; that the word fire, for instance, might have denominated the substance which we call ice, and that the word ice might have signified fire. Some of the most acute feelings of man, as well as of every other animal, are indeed expressed by simple inarticulate sounds, which, as they tend to the preservation of the individual, or the continuance of the species, and invariably indicate either pain or pleasure, are universally understood; but these inarticulate and significant sounds are very few in number; and if they can with any propriety be said to constitute a natural and universal language, it is a language of which man as a mere sensitive being partakes in common with the other animals.

Man is endowed not only with sensation, but also with the faculty of reasoning; and simple inarticulate sounds are insufficient for expressing all the various modifications thought, for communicating to others a chain of argumentation, or even for distinguishing between the different sensations either of pain or of pleasure; a man scorched with fire or unexpectedly plunged amongst ice, might utter a cry naturally indicative of sudden and violent pain; the cry would be the same, or nearly the same, but the sensations of cold and heat are widely different. Articulation, by which those simple sounds are modified, and a particular meaning fixed to each modification, is therefore absolutely necessary to such a being as man, and forms the language which distinguishes him from all other animals, and enables him to communicate with facility all that diversity of ideas with which his mind is stored, to make known his particular wants, and to distinguish with accuracy all his various sensations. Those sounds thus modified are called words; and as words have confessedly no natural relation to the ideas and perceptions of which they are significant, the use of them must either have been the result of human sagacity, or have been suggested to the first man by the author of nature.

Whether language be of divine or human origin, is a question upon which, though it might perhaps be soon resolved, it is not necessary to enter here. Upon either supposition, the first language, compared with those which succeeded it, or even with itself as afterwards enlarged, must have been extremely narrow and rude. If it was of human contrivance, this will be readily granted; for what it was ever invented and brought to a state of perfection in illiterate savages? If it was taught by God, which is at least the more probable supposition, we cannot imagine that it would be more comprehensive than the ideas of those for whose immediate use it was intended; that the first men should have been taught to express pains or pleasures which they never felt, or to utter sounds that should afterwards be significant of ideas which at the time of utterance had not occurred to the mind of the speaker. Man once taught the elements of language would be able of himself to improve and enlarge it as his future occasions required.

As all language is composed of significant words variously combined, a knowledge of them is necessary previous to our acquiring an adequate idea of language as conducted into sentences and phrases. But as it is by words that we express the various ideas which occur to the mind, it is necessary to examine how ideas themselves are suggested, before we can ascertain the various classes into which words may be distributed. It is the province of logic to trace our ideas from their origin, as well as to unfold the process of reasoning; but it is necessary at present to observe, that our earliest ideas are all ideas of sensation, excited by the impressions that are made upon our organs of sense by the various objects with which we are surrounded. Let us therefore suppose a reasonable being, devoid of every possible prepossession, placed upon this globe; and it is obvious that his attention would in the first place be directed to the various objects which he now existing around him. These he would naturally endeavour to distinguish from one another; and if he were rather learning or inventing a language, his first effort would be to give them names, by means of which the ideas of them might be recalled when the objects themselves should be absent. This is one copious source of words, and forms a natural class which must be common to every language, and which is distinguished by the name of nouns; and as these nouns are the names of the several substances which exist, they have likewise been called substantives.

It would likewise be early discovered that every one of these substances was endowed with certain qualities or attributes, to express which another class of words would be requisite, since it is only by their qualities that substances themselves can attract our attention. Thus, to be weighty, is a quality of matter; to think, is an attribute of man. Therefore in every language words have been invented to express the known qualities or attributes of the several objects which exist. These may all be comprehended under the general denomination of attributives.

Nouns and attributives must comprehend all that is essential to language; for every thing which exists, or of which we can form an idea, must be either a substance or the attribute of some substance; and therefore the two classes which we denominate substances and attributes must comprehend all the words that are necessary to communicate to the hearer the ideas which are present to the mind of the speaker. If any other words occur, they can only have been invented for the sake of dispatch, or introduced for the purposes of ease and ornament, to avoid tedious circumlocutions or disagreeable tautologies. There are indeed grammarians of great name, who have considered as essential to language an order of words, of which the use is to connect the nouns and attributes, and which are said to have no signification of themselves, but to become significant by relation. Hence all words which can possibly be invented are by these men divided into two general classes; those which are significant of themselves, and those which are not. Words significant of themselves are either expressive of the names of substances, and therefore called substantives; or of attributes, and therefore called attributives. Words which are not significant of themselves must acquire a meaning either as defining or connecting others, and are therefore arranged under the two classes of definitives and connectives.

That in any language there can be words which of themselves have no signification, is a supposition which a man free from prejudice will not readily admit. For to what purpose, we ask, should they have been invented? As they are significant of no ideas, they cannot facilitate the communication of thought, and must therefore be only an incumbrance to the language in which they are found. But in answer to this it has been said, that these words, though devoid of signification themselves, acquire a sort of meaning when joined with others, and that they are as necessary to the structure of a sentence as mortar is to the structure of an edifice; for as stones cannot be arranged into a regular building without a cement to bind and connect them, so the original words significant of substances and attributes cannot be made to express all the variety of our ideas without being defined and connected by those words which of themselves signify nothing. It is wonderful that he who first suggested this simile did not perceive that it tends to overthrow the doctrine which it is meant to illustrate. For surely the cement is as much the matter of the building as the stones themselves are; it is equally solid and equally extended. By being united with the stones, it neither acquires nor loses any one of the qualities essential to matter; it neither communicates its own softness nor acquires their hardness. By this mode of reasoning, therefore, it

This is the doctrine of many writers on the theory of language; yet it is not easy to conceive mankind so far advanced in the art of abstraction as to view attributives by themselves independently of particular substances, and to give one general name to each attribute wheresoever it may be found, without having at the same time words expressive of affirmation. We never talk of any attribute, a colour for instance, without affirming something concerning it; as, either that it is bright or faint, or that it is the colour of some substance. It will be seen afterwards, that to denote affirmation is the proper office of what is called the substantive verb; "Milk is white." That verb therefore appears to be as necessary to the communication of thought as any species of words whatever; and if we must range words under a few general classes, we should be inclined to say that nouns, attributes, and affirmatives, comprehend all that is really essential to language. would appear that the words called definitives and connectives, so far from having of themselves no signification, are equally essential to language, and equally significant with those which are denominated substantives and attributives; and upon investigation it will be found that this is the truth. For whatever is meant by the definition or connection of the words which all men confess to be significant, that meaning must be the sense of the words of which the purpose is to define and connect; and as there can be no meaning where there are no ideas, every one of these definitives and connectives must be significant of some idea, although it may not always be easy or even possible to express that idea by another word.

These different modes of dividing the parts of speech we have just mentioned because they have been largely treated of by grammarians of high fame. But it does not appear to us that any man can feel himself much the wiser for having learned that all words are either substantives or attributives, definitives or connectives. The division of words into those which are significant of themselves, and those which are significant by relation, is absolute nonsense, and has been productive of much error and much mystery in some of the most celebrated treatises on grammar. It is indeed probable that any attempt to establish a different classification of the parts of speech from that which is commonly received, will be found of little utility either in practice or in speculation. As far as the former is concerned, the vulgar division seems sufficiently commodious; for every man who knows anything, knows when he uses a noun and when a verb. With respect to the latter, not to mention that all the grammarians from Aristotle to Horne Tooke have differed on the subject, it should seem to be of more importance, after having ascertained with precision the nature of each species of words, to determine in what circumstances they differ than in what they agree.

In most languages, probably in all cultivated languages, grammarians distinguish the following parts of speech, viz. noun, pronoun, verb, participle, adverb, preposition, conjunction. The Latin and English grammarians admit the interjection among the parts of speech, although it is confessedly not necessary to the construction of the sentence, being only thrown in to express the affection of the speaker; and in the Greek and English tongues there is the article prefixed to nouns when they signify the common names of things, to point them out, and to show how far their signification extends. In the method of arrangement commonly followed in grammars, adjectives are classed with substantives, and both are denominated nouns. But it is certain that, when examined philosophically, an essential difference is discovered between the substantive and the adjective; and therefore some philologists of eminence, when treating of this subject, have given the following classification of words, which we shall adopt: The ARTICLE, NOUN, PRONOUN, VERB, PARTICIPLE, ADJECTIVE, ADVERB, PREPOSITION, CONJUNCTION, INTERJECTION. All these words are to be found in the English language, and therefore we shall examine each class, endeavour to ascertain its precise import, and show in what respects it differs from every other class. It is impossible to investigate the principles of grammar without confining the investigation in a great measure to some particular language from which the illustrations must be produced; and that we should prefer the English language for this purpose can excite no wonder, as it is a preference which is due to every tongue from those by whom it is spoken. The principles, however, which we shall here expound and elucidate, being in substance the same with those so clearly and ingeniously established by Dr John Hunter of St Andrews, in his oral prelections on philology, will be found to apply universally; and our inquiry, though principally illustrated from the English language, will be an inquiry into Philosophical and Universal Grammar.

I. Of the Noun or Substance.

Nouns are all those words by which objects or substances are denominated, and which distinguish them from one another, without marking either quantity, quality, action, or relation. The substantive or noun is the name of the thing spoken of, and in Greek and Latin is called *name*; for it is *nomen* in the one, and *nomen* in the other; and if in English we had called it the *name* rather than the *noun*, the appellation would perhaps have been more proper, as this last word, being used only in grammar, is more liable to be misunderstood than the other, which is in constant and familiar use. That nouns or the names of things must make a part of every language, and that they must have been the words first suggested to the human mind, will not be disputed. Men could not speak of themselves, or of anything else, without having names for themselves and the various objects with which they are surrounded. Now, as all the objects which exist must be either in the same state in which they were produced by nature, or changed from their original state by art, or abstracted from substances by the powers of imagination, and conceived by the mind as having at least the capacity of being characterised by qualities; this naturally suggests a division of nouns into natural, as man, vegetable, tree; artificial, as house, ship, watch; and abstract, as whiteness, motion, temperance.

But the diversity of objects is so great, that had each individual a distinct and proper name, it would be impossible for the most tenacious memory, during the course of the longest life, to retain even the nouns of the narrowest language. It has therefore been found expedient, when a number of things resemble each other in some important particulars, to arrange them all under one species; to which is given a name that belongs equally to the whole species, and to each individual comprehended under it. Thus the word *man* denotes a species of animals, and is equally applicable to every human being; the word *horse* denotes another species of animals, and is equally applicable to every individual of that species of quadrupeds; but it cannot be applied to the species of men, or to any individual comprehended under that species. We find, however, that there are some qualities in which several species resemble each other; and therefore we refer them to a higher order called a genus, to which we give a name that is equally applicable to every species and every individual comprehended under it. Thus, men and horses and all living things on earth resemble each other in this respect, that they have life. We refer them therefore to the genus called animal; and this word belongs to every species of animals, and to each individual animal. The same classification is made both of artificial and abstract substances; of each of which there are genera, species, and individuals. Thus, in natural substances, animal, vegetable, and fossil, denote genera; man, horse, tree, metal, species; and Alexander, Bucephalus, oak, gold, individuals. In artificial substances, again, edifice is a genus; house, church, tower, are species; and the Vatican, St Paul's, and the Tower of London, are individuals. Lastly, in abstract substances, motion and virtue are genera; flight and temperance are species; the flight of Mahomet, and temperance in wine, are individuals. By arranging substances in this manner, and giving a name to each genus and species, the nouns necessary to any language are comparatively few and easily acquired; and when we meet with an object unknown to us, we have only to examine it with attention, and, comparing it with other objects, to refer it to the genus or species which it most nearly resembles. By this contrivance we supply the want of a proper name for the individual; and, as far as the resemblance is complete between it and the species to which it is referred, and of which we have given it the name, we may converse and reason about it without danger of error; whereas had each individual in nature a distinct and proper name, words would be innumerable and incomprehensible; and to employ our labours in language would be as useless as that study of numberless written symbols which has been attributed to the Chinese.

Although nouns are thus adapted to express, not the individuals, but the genera or species into which substances are classified; yet, in speaking of these substances, whether natural, artificial, or abstract, all men must have occasion to mention sometimes one of a kind, and sometimes more than one. In every language, therefore, nouns must admit of some variation in their form, to denote unity and plurality; and this variation is called number. Thus in the English language, when we speak of a single place of habitation, we call it a house; but if of more, we call them houses. In the first of these cases the noun is said to be in the singular; in the last case it is in the plural number. Greek nouns have also a dual number to express two individuals, as we likewise some Hebrew nouns; but this variation is evidently not essential to language; and it is perhaps doubtful whether it ought to be considered as an elegance or as a deformity.

But although number be a natural accident of nouns, it can only be considered as essential to those which denote genera or species. Thus we may have occasion to speak of one animal or of many animals, of one man or of many men; and therefore the nouns animal and man must be capable of expressing plurality as well as unity. But this is not the case with respect to the proper names of individuals; for we can only say Xenophon, Aristotle, Plato, in the singular; as, were any one of these names to assume a plural form, it would cease to be the proper name of an individual, and become the common name of a species. Of this, indeed, we have some examples in every language. When a proper name is considered as a general appellative under which many others are arranged, it is then no longer the name of an individual, but of a species, and as such admits of a plural; as the Cassiers, the Howards, the Pelikes, the Montagues; but Socrates can never become plural, among us we know of no more than one man of that name. To reason of all this will be obvious, if we consider that every genus may be found whole and entire in each of its species; for man, horse, and dog, are each of them an entire and complete animal; and every species may be found whole and entire in each of its individuals, Socrates, Plato, or Xenophon, being each of them completely and entirely man. Hence it is, that every genus, though one, is multiplied into many; and every species, though one, is multiplied into many, by reference to those beings which are their subordinates; but as no individual has such subordinates, it can never in strictness be considered as many; and so, as well in nature as in name, is every individual which cannot admit of number.

Besides number, another characteristic, visible in substances, is that of sex. Every substance is either male or female, or both male and female, or neither the one nor the other. So that with respect to sexes and their negation, all substances conceivable are comprehended under the fourfold consideration, which language would be very imperfect if it could not express. Now the existence of hemaphrodites being rare, if not doubtful, and language being framed to answer the ordinary occasions of life, no provision is made, in any of the tongues with which we are acquainted, for expressing otherwise than by a name the sex on purpose, or by a periphrasis, duplicity of sex. Wh regard to this great natural characteristic, grammarians have made only a threefold distinction of nouns: those which denote males are said to be of the masculine gender; those which denote females, of the feminine; and those which denote substances that admit not of sex, are said to be neuter, or of neither gender. All animals have sex, and therefore the names of all animals should have gender. But the sex of all is not equally obvious, nor equally worthy of attention. In those species which are most common, or of which the male and female are, by their size, form, colour, or other outward circumstances, eminently distinguished, the male is sometimes called by one name, which is masculine, and the female by a different name, which is feminine. Thus in English we say, husband, wife; king, queen; father, mother; son, daughter; and so on. In others of similar distinction, the name of the male is applied to the female only by prefixing a syllable or by altering the termination; as man, woman; lion, lioness; emperor, empress, anciently emperess; master, mistress, anciently misteress; and so in other similar instances. When the sex of any animal is not obvious, or not material to be known, the same name, in some languages, is applied, without variation, to all the species, and that name is said to be of the common gender. Thus in Latin bos albus is a white ox, and bos alba a white cow. Diminutive insects, though they are doubtless male and female, seem to be considered in the English language as if they were really creeping things. No man, speaking of a worm, would say he creeps, but it creeps upon the ground. But although the origin of genders is thus clear and obvious, yet the English is the only language with which we are acquainted that deviates not, except in a very few instances, from the order of nature. Greek and Latin, and many of the modern tongues, have nouns, some masculine, some feminine, which denote substances where sex never had existence. Nay, some languages are so particularly defective in this respect, as to class every object, inanimate as well as animate, under either the masculine or the feminine gender, as they have no neuter gender for those which are of neither sex. This is the case with the Hebrew, French, Italian, and Spanish. But the English, strictly following the order of nature, puts every noun which denotes a male animal, and no other, in the masculine gender; every name of a female animal, in the feminine; and every animal whose sex is not obvious, or known, as well as every inanimate object whatsoever, in the neuter gender. And this gives our language an advantage above most others in the poetical and rhetorical style; for when nouns naturally neuter are converted into masculine and feminine, the personification is more distinctly and more forcibly marked. Some very learned and ingenious men have endeavoured, by what they call a more subtile kind of reasoning, to discern even in things without a sex a distant analogy to that natural distinction, and to account for the names of inanimate substances being, in Greek and Latin, masculine and feminine. But such speculations are wholly fanciful, and the principles upon which they proceed are overthrown by an appeal to facts. Many of the substances that, in one language, have masculine names, have in others names that are feminine; which could not be the case were this matter regulated by reason or nature. Indeed, for this, as well as many other anomalies in language, no other reason can be assigned than that custom,

Quem penes arbitrium est, et jus, et norma, loquendi.

It has been already observed that most nouns are the names, not of individuals, but of whole classes of objects termed genera and species. In classing a number of individuals under one species, we contemplate only those

It is almost unnecessary to observe, that the words genus and species, and the phrases higher genus and lower species, are taken here in the logical sense; and not as the words genus, species, order, class, are often employed by naturalists. qualities which appear to be important, and in which the several individuals are found to agree, abstracting the mind from the consideration of all those which appear to be less essential, and which in one individual may be such as have nothing exactly similar in any other individual upon earth. Thus, in classing the individuals which are comprehended under the species denominated horse, we pay no regard to their colour or their size; because experience teaches us that no particular colour or size is essential to that individual living creature, and that there are not perhaps upon earth two horses whose colour and size are exactly alike. But the qualities which in this process we take into view, are the general shape, the symmetry, and proportion of the parts; and in short, every thing which appears evidently essential to the life of the individual and the propagation of the race. All these qualities are strikingly similar in all the individuals which we call horses; and as strikingly dissimilar from the corresponding qualities of every other individual animal. The colour of a horse is often the same with that of an ox; but the shape of the one animal, the symmetry and proportion of his parts, are totally different from those of the other; nor could any man be led to class the two individuals under the same species. It is by a similar process that we ascend from one species to another, and through all the species to the highest genus. In each species or genus in the ascending series fewer particular qualities are attended to than were considered as essential to the genus or species immediately below it; and our conceptions become more and more general as the particular qualities which are the objects of them become fewer in number. The use of a general term, therefore, can recall to the mind only the common qualities of the class, the genus, or species, which it represents. But we have frequent occasion to speak of individual objects. In doing this, we annex to the general term certain words significant of particular qualities, which discriminate the object of which we speak from every other individual of the class to which it belongs, and of which the general term is the common name. For instance, in advertising a thief, we are obliged to mention his height, complexion, gait, and whatever may serve to distinguish him from all other men.

The process of the mind in rendering its conceptions particular, is indeed exactly the reverse of that by which it generalizes them. For, as in the process of generalization, it abstracts from the ideas of any number of species certain qualities in which they differ from each other, and of the remaining qualities in which they agree constitutes the first genus in the ascending series; so, when it wishes to make its conceptions more particular, it annexes to the idea of any genus those qualities or circumstances which were before abstracted from it; and the genus, with this annexation, constitutes the first species in the descending series. In like manner, when it wishes to descend from any species to an individual, it has only to annex to the idea of the species those particular qualities which discriminate the individual intended from the other individuals of the same kind.

This particularizing operation of the mind points out the manner of applying the general terms of language for the purpose of expressing particular ideas. For, as the mind, to limit a general idea, connects that idea with the idea of some particular circumstance, so language, as we have already observed, in order to limit a general term, connects that term with the word denoting the particular circumstance. Thus, in order to particularize the idea of horse, the mind connects that general idea with the circumstance, suppose, of whiteness; and in order to particularize the word horse, language connects that word with the term white; and so in other instances. Annexation, therefore, or the connecting of general words or terms in language, fits it for expressing particular conceptions; and this must hold good alike in all languages. But the methods of denoting this annexation are various in various tongues. In English and most modern languages we commonly use for this purpose little words, which we have chosen to style particles; and in the Greek and Latin languages, the cases of nouns answer the same end.

Cases, therefore, though they are accidents of nouns not absolutely necessary, have been often considered as such; and they are certainly worthy of our examination, since there is perhaps no language in which some cases are not to be found, as indeed without them or their various powers no language could readily answer the purposes for which it was formed.

All the oblique cases of nouns, if we except the vocative, are merely marks of annexation; but as the connections or relations subsisting among objects are very various, some cases denote one kind of relation, and some another. We shall endeavour to investigate the connection which each case denotes, beginning with the genitive.

This is the most general of all the cases, and gives notice that some connection indeed subsists between two objects, but does not point out the particular kind of connection. That we must infer, not from our nature or termination of the genitive itself, but from our previous knowledge of the objects connected. That the genitive denotes merely relation in general, might be proved by adding innumerable examples, in which the relations expressed by this case are different; but we shall content ourselves with one observation, from which the truth of our opinion will appear beyond dispute. If an expression be used in which are connected by the genitive case, two words significant of objects between which a twofold relation may subsist, it will be found impossible, from the expression, to determine which of these two relations is the true one, which must be gathered wholly from the context. Thus, for example, from the phrase injuria regis, no man can know whether the injury mentioned be an injury suffered or an injury inflicted by the king; but if the genitive case notified any particular relation, no such ambiguity could exist. This case therefore gives notice, that two objects are somehow connected; but it marks not the particular sort of connection. Hence it may be translated by our particle of, which will be seen afterwards to be of a signification equally general.

The dative and accusative cases appear to have nearly the same meaning; each of them denoting apposition, or the junction of one object with another. Thus, when any one says, Comparo Virgilium Homero, Homer and Virgil are conceived to be placed beside one another, in order to their being compared; and this sort of connection is denoted by the dative case. In like manner, when it is said latus huieros, breadth is conceived as joined to or connected in apposition with shoulders; and the expression may be translated "broad at the shoulders."

This apposition of two objects may happen either without previous motion, or in consequence of it. In the foregoing instances no motion is presupposed; but if one say, Misit aliquos subsidio corum, the apposition is there in consequence of motion. In like manner, when it is said Profectus est Romam, his apposition with Rome is conceived as the effect of his motion thither. From this idea of the

---

1 The Greek grammarians seem to have been aware of the nature of this case when they called it ἐνεργοῦς γενικός, or the general case; a name of which the Latin grammarians evidently mistook the meaning when they translated it causa genitiva, or the generative case, which is a name totally foreign to its nature. The chief objection to this conjecture, that the nominative and vocative were originally the same case, is taken from the Latin language, in which the nouns of the second declension ending in *us* terminate their vocative in *e*. But this is easily accounted for, as in such words was often dropped, as appears from the scanning of old Latin poetry; when this was done, the *u* being short, would naturally in pronunciation pass into *e*, a like short vowel; and thus, in the vocative case, *e* would in time be written instead of *u*. jects have therefore been classed into genera and species; and names given, not to each individual, but to each genus and species. By this contrivance of language, we are enabled to ascertain in some measure any individual that may occur, and of which we know not the proper name, only by referring it to the genus or species to which it belongs, and calling it by the general or specific name; but as there is frequent occasion to distinguish individuals of the same species from one another, it became necessary to fall upon some expedient to mark this distinction. In many languages general and specific terms are modified and restricted by three orders of words; the article, the adjective, and the oblique cases of nouns. The cases of nouns we have already considered; the adjective will employ our attention afterwards; at present our observations are confined to the article, a word so very necessary, that without it, or at least some equivalent invention, men could not employ nouns to any of the purposes of life, or indeed communicate their thoughts at all. As the business of articles is to enable us, upon occasion, to employ general terms to denote particular objects, they must be considered in combination with the general terms as merely substitutes for proper names. They have, however, been commonly called definitives; because they serve to define and ascertain any particular object so as to distinguish it from the other objects of the general class to which it belongs, and, of course, to denote its individuality. Of words framed for this purpose, whether they have by grammarians been termed articles or not, we know of no language that is wholly destitute. The nature of them may be explained as follows.

An object occurs with which, as an individual, we are totally unacquainted; it has a head and limbs, and appears to possess the powers of self-motion and sensation; we therefore refer it to its proper species, and call it a dog, a horse, a lion, or the like. If it belongs to none of the species with which we are acquainted, it cannot be called by any of their names; we then refer it to the genus, and call it an animal.

But this is not enough. The object at which we are looking, and which we want to distinguish, is not a species or a genus, but an individual. Of what kind? Known or unknown? Seen now for the first time, or seen before and now remembered? This is one of the instances in which we shall discover the use of the two articles A and THE; for, in the case supposed, the article A respects our primary perception, and denotes an individual as unknown; whereas THE respects our secondary perception, and denotes individuals as known. To explain this by an example: I see an object pass by which I never saw till now: What do I say? There goes a beggar with a long beard. The man departs and returns a week after: What do I then say? There goes the beggar with the long beard. Here the article only is changed; the rest remains unaltered. Yet mark the force of this apparently minute change. The individual once vague is now recognised as something known; and that merely by the efficacy of this latter article, which tacitly insinuates a kind of previous acquaintance, by referring a present perception to a like perception already past.

This is the explanation of the articles A and THE as given by the learned Mr Harris, and thus far what he says on the subject is certainly just; but it is not true that the article THE always insinuates a previous acquaintance, or refers a present perception to a like perception already past. I am in a room crowded with company, of which the greater part is to me totally unknown; I feel it difficult to breathe from the grossness of the enclosed atmosphere; and, looking towards the window, I see in it a person whom I never saw before. I instantly send my compliments to the gentleman in the window, and request that, if it be not inconvenient, he will have the goodness to let a little fresh air into the room. Of this gentleman I have no previous acquaintance; my present perception of him is my primary perception, and yet it would have been extremely improper to send my compliments to a gentleman in the window. Again, there would be no impropriety in saying, "A man whom I saw yesterday exhibiting a show to the rabble was this morning committed to jail charged with the crime of housebreaking." Notwithstanding the authority, therefore, of Mr Harris, and his master Apollonius, we may venture to affirm, that it is not essential to the article A to respect a primary perception, or to the article THE to indicate a pre-established acquaintance. Such may indeed be the manner in which these words are most frequently used; but we see that there are instances in which they may be used differently. What, then, it may be asked, is the import of each article, and in what respects do they differ?

We answer, that the articles A and THE are both of them definitives, as, by being prefixed to the names of genera and species, they so circumscribe the latitude of those names as to make them for the most part denote individuals. A noun or substantive, without any article to limit it, is taken in its widest sense. Thus the word man means all mankind;

The proper study of mankind is man; where mankind and man may change places without making any alteration in the sense. But let either of the articles of which we are treating be prefixed to the word man, and that word is immediately reduced from the name of a whole genus to denote only a single individual; and instead of the noble truth which this line asserts, the poet will be made to say, that the proper study of mankind is not the common nature which is diffused through the whole human race, but the manners and caprices of one individual. Thus far therefore the two articles agree; but they differ in this, that though they both limit the specific name to some individual, the article A leaves the individual itself unascertained; whereas the article THE ascertains the individual also, and can be prefixed to the specific name only when an individual is intended of which something may be predicated that distinguishes it from the other individuals of the species. Thus, if I say, a man is fit for treasons, my assertion may appear strange and vague; the sentence, however, is complete, and wants nothing to make it intelligible; but if I say, the man is fit for treasons, I speak nonsense; for as the article THE shows that I mean some particular man, it will be impossible to discover my meaning till I complete the sentence, and predicate something of the individual intended to distinguish him from other individuals.

The man that hath not music in himself... Is fit for treasons.

A man, therefore, means some one or other of the human race indefinitely; the man means, definitely, that particular man who is spoken of. The former is called the indefinite, the latter the definite, article.

The two articles differ likewise in this respect, that as the article A serves only to separate one individual object from the general class to which it belongs, it cannot be applied to plurals. It has indeed the same signification nearly with the numerical word one; and in French and Italian the same word that denotes unity is also the article of which we now treat. But the essence of the article THE being to define objects, by pointing them out as those of which something is affirmed or denied which is not affirmed or denied of the other objects of the same class, it is equally applicable to both numbers; for things may be predicated of one set of men, as well as of a single man, which cannot be predicated of other men. The use import of each article will appear from the following example: "Man was made for society, and ought to extend his good will to all men; but a man will naturally entertain a more particular regard for the men with whom he has the most frequent intercourse, and enter into a closer union with the man whose temper and disposition best suit with his own."

We have said that the article A cannot be applied to nouns, because it denotes unity. But to this rule there apparently a remarkable exception in the use of the adjectives few and many (the latter chiefly with the word gent before it), which, though joined with plural substantives, yet admit of the singular article A: as, a few men, several many men. The reason of this is manifest from its effect which the article has in these phrases. It means small or a great number collectively taken, to which it gives the idea of a whole, that is, of unity. Thus likewise a hundred, a thousand, is one whole number, an aggregate of many collectively taken, and therefore still retains the article A, though joined as an adjective to a plural substantive; as, a hundred years. The exception therefore is only apparent; and we may affirm that the article A universally denotes unity.

The indefinite article is much less useful than the other; and therefore the Greek and Hebrew languages have it not, though they have both a definite article. In languages of which the nouns, adjectives, and verbs, have no inflexion, no mistake can arise from the want of the indefinite article; because it can always be known by the terminations of the noun and the verb, and by the circumstances predicated of the noun, whether a whole species or one individual be intended. But this is not the case in English. In that language, the adjectives having no variation with respect to gender or number, and the tenses of the verbs being for the most part the same in both numbers, it might be often doubtful, had we not the indefinite article, whether the specific name was intended to express the whole species or only one individual. Thus, if we say in English, man was born sent from God, we must be understood to mean that the birth of every man is from God, because to the specific term the indefinite article is not prefixed. Yet the words, Εγὼ ἐξῆλθον, convey no such meaning to any person acquainted with the Greek language; as the word ἀπόστεις, without any article, is restricted to an individual in its concord with the verb and the participle; and the case of the passage is, A man was born, or existed, sent from God. But though the Greeks have no article corresponding to the article A, yet nothing can be more nearly related than their Ο to our THE; Ο βασιλεύς, the king, TO δώρον, the gift. In one respect, indeed, the Greek and English articles differ. The former is varied according to the gender and number of the noun with which it is associated, being ἀ, masculine, ἡ, feminine, τό, neuter, and οἱ, αἱ, τα, in the plural number; whereas the English article suffers no change, being invariably the same before nouns of every gender and in both numbers. There are, however, some modern languages which, in imitation of the Greek, admit of a variation of their article which relates to gender; but this cannot be considered as essential to this species of words, and it may be questioned whether it be an improvement to the language. In tongues of which the nouns have no inflexion, it can only serve to perplex and confuse, as it always presents a particular idea of sex where in many cases it is not necessary.

The articles already mentioned are allowed to be strict and properly such by every grammarian; but there are some words, such as this, that, any, some, all, other, &c., which are generally said to be sometimes articles and sometimes pronouns, according to the different modes of using them. That words should change their nature in this manner, so as to belong sometimes to one part of speech and sometimes to another, must appear very extraordinary; and if it were a fact, language would be a thing so equivocal, that all inquiries into its nature, upon principles of science and reason, would be vain. But we cannot perceive any such fluctuation in any word whatever; though we know it to be a general charge brought against words of almost every denomination, of which we have already seen one instance in the possessive case of nouns, and shall now see another in those words which are commonly called pronominal articles.

If it be true, as we acknowledge it to be, that the genuine pronoun always stands by itself, assuming the power of a noun, and supplying its place, then is it certain that the words this, that, any, some, &c. can never be pronouns. We are indeed told, that when we say this is virtue, give me that, the words this and that are pronouns; but that when we say, this habit is virtue, that man defrauded me, then are they articles or definitives. This, however, is evidently a mistake occasioned by overlooking those abbreviations in construction which are frequent in every language, and which, on account of that very frequency, have perhaps escaped the attention of grammarians whose sagacity has been successfully employed on matters less obvious. When we say this is virtue, it is evident that we communicate no intelligence till we add a substantive to the word this, and declare what is virtue. The word this can therefore in no instance assume the power of a noun, since the noun to which it relates, though for the sake of dispatch it may be omitted in writing or conversation, must always be supplied by the mind of the reader or hearer, to make the sentence intelligible, or this itself of any importance. When we have viewed speech analysed, we may then consider it as compounded. And here, in the first place, we may contemplate that synthesis which by combining simple terms produces a truth; then by combining two truths produces a third; and thus others and others in continued demonstration, till we are led, as by a road, to the regions of science. Now this is that superior and most excellent synthesis which alone applies itself to our intellect or reason, and which to conduct according to rule constitutes the art of logic. After this we may turn to those inferior compositions which are productive of the pathetic, &c. Here, if anywhere, the word this may be thought to stand by itself; and to assume the power of a noun; but let any man complete the construction of each sentence, and he will perceive that this is no more than a definite article. Thus, we may contemplate that synthesis which by combining simple terms produces a truth; then by combining two truths produces a third truth; and thus other truths and other truths in continued demonstration, till we are led, as by a road, into the regions of science. Now this combination of truths is that superior and most excellent synthesis which alone applies itself to our intellect or reason, and which to conduct according to rule constitutes the art of logic. After we have contemplated this art, we may turn, &c.

The word that is generally considered as still more equivocal than this; for it is said to be sometimes an article, sometimes a pronoun, and sometimes a conjunction. In the following extract it appears in all these capacities; and yet, upon resolving the passage into parts, and completing the construction, it will be found to be invariably a definite article. "It is necessary to that perfection, of which our present state is capable, that the mind and body should both be kept in action; that neither the faculties of the one nor of the other be suffered to grow lax or torpid for want of use; but neither should health be purchased by voluntary submission to ignorance, nor should knowledge be cultivated at the expense of health;" for that must enable it either to give pleasure to its possessor, or assistance to others." If this long sentence be resolved into its constituent parts, and the words be supplied which complete the construction, we shall see the import of the word that to be precisely the same in each clause. "The mind and body should both be kept in action; that action is necessary to that perfection of which our present state is capable; neither the faculties of the one nor of the other should be suffered to grow lax or torpid for want of use; the degree of action proper to prevent that laxness is necessary; but neither should health be purchased by voluntary submission to ignorance, nor should knowledge be cultivated at the expense of health; for that health must enable it either to give pleasure to its possessor, or assistance to others." Again:

He that's unskilful will not toss a ball.

A man unskilful (he is that) will not toss a ball. Here the word that, though substituted for what is called the relative pronoun, still preserves unchanged its definitive import; and in every instance, except where it may be used very improperly, it will be found to be neither more nor less than a definitive article.

It appears, then, that if the essence of an article be to define and ascertain, the words this and that, as well as any, some, all, &c., which are commonly called pronominal articles, are much more properly articles than any thing else, and as such should be considered in universal grammar. Thus, when we say, this picture I approve, but that I dislike, what do we perform by the help of the words this and that, but bring down the common appellative to denote two individuals, the one as the more near, the other as the more distant? So when we say, some men are virtuous, but all men are mortal; what is the natural effect of this all and some, but to define that universality and that particularity which would remain indefinite were we to take them away? The same thing is evident in such sentences as, "some substances have sensation, others want it; choose any way of acting, and some men will find fault," &c. For here some, other, and any serve all of them to define different parts of a given whole; some, to denote any indeterminate part; any, to denote any indefinite mode of action, no matter what; and other, to denote the remaining part, when a part has been assumed already.

We have said that the article is a part of speech so very necessary, that without it, or some equivalent invention, mankind could not communicate their thoughts; and that of words falling under this description we know of no language which is wholly destitute. We are aware that these positions may be controverted; and that the Latin may be instanced as a language which, without articles, is not only capable of communicating the ordinary thoughts of the speaker to the mind of the hearer, but which, in the hands of Cicero, Virgil, and Lucretius, was made to serve all the purposes of the most profound philosopher, the most impassioned orator, and the most sublime poet. That the Latin has been made to serve all these purposes cannot be denied, although Lucretius and Cicero both complain, that on the subject of philosophy, where the use of articles is most conspicuous, it is a deficient language. But should we grant what cannot be demanded, that these two great men were unacquainted with the powers of their native tongue, our positions would still remain unshaken; for we deny that the Latin is wholly without articles. It has indeed no word of precisely the same import with our the or the Greek ἡ; but the place of the indefinite article a might be always supplied, if necessary, with the numerical word unus. It may be so even in English; for we believe there is not a single instance where the words one man, one horse, one virtue, might not be substituted for the words a man, a horse, a virtue, without in the slightest degree altering the sense of the passage where such words occur. This substitution, however, can be but seldom or never necessary in the Latin tongue, of which the precision is much greater than that of the English would be without articles, because the oblique cases of the Latin nouns, and the inflexion of its verbs, will almost always enable the reader to determine whether an appellative represents a whole species or a single individual. The want of the definite article seems to be a greater defect; yet there are few instances in which its place might not be supplied by this or by that without obscuring the sense; and the Latin tongue is by no means deficient of articles corresponding to these two. Let us substitute the words one and that for a and the in some of the foregoing examples, and we shall find, though the sound may be uncouth, the sense will remain. Thus,

That man who hath not music in himself...

Is fit for treasons,

conveys to the mind of the reader the very same sentiment which the poet expresses by the words the man that hath not music, &c. Again, Man was made for society, and ought to extend his good will to all men; but one man will naturally entertain a more particular regard for those men with whom he has the most frequent intercourse, and enter into a closer union with that man whose temper and disposition suit best with his own. Now the words hic and ille being exactly of the same import with the words this and that, it follows, that wherever the place of the article the may in English be supplied by this or by that, it may in Latin be supplied by hic or by ille. This is the case with respect to Nathan's reproach of David, where the definite article is indeed most emphatical. The original words might have been translated into English, "thou art that man," as well as "thou art the man;" and in Latin they may with the utmost propriety be rendered, "Tu es ille homo." Indeed the words hic and ille, and we might instance many more, though they are commonly called pronouns, are in truth nothing but definite articles. Hic is evidently ἰδια; and ille is most probably derived from the Hebrew word al, in the plural ale, which may be translated indifferently, either the or that. But, what proves beyond dispute that these words are not pronouns, but articles, is, that in no single instance will they be found to stand by themselves and assume the power of nouns. For the sake of dispatch, or to avoid disagreeable repetitions, the noun may indeed be often omitted; but it is always supplied by the reader or hearer, when hic and ille appear in their proper place, and are seen to be invariably definitive articles. We shall give an example of the use of each word, and dismiss the subject.

In the first oration against Catiline, Cicero begins with addressing himself in a very impassioned style to the traitor, who was present in the senate-house. He then exclaims pathetically against the manners of the age, and proceeds in these words: Senatus haec intelligit, consilium videt, hic tamen vivit. Vivit? immo vero etiam in senatu venit; fit publici consilii participes. In this passage hic cannot be a pronoun; for from the beginning of the oration there occurs not a single noun of which it can possibly supply the place. When the orator uttered it, he was probably pointing with his finger at Catiline, and every

---

1 As in the Persian and other eastern languages, in which the place of our indefinite article is supplied by a termination to those nouns which are meant to be particularized. When Virgil says,

"Ile ego, qui quondam gracili modulatus avena Carmen."

It is obvious that he means, I am that man, or that poet, who sung, &c.; and though we may translate the words "I am he who tuned his song," yet when we construe the passage we are under the necessity of supplying either this or that, which shows that ille is nothing more than a definite article signifying that or the. - It appears, then, that the Latin tongue is not wholly destitute of articles, but few cases can occur where the Greek ει and our the may be supplied by the words hic and ille, which have in our opinion been very improperly termed pronouns. If there be any such cases, we can only confess that the Latin language is defective; whereas, had it no articles, it is not easy to conceive how it could answer, to a culti- vated people, the ordinary purposes of speech.

The articles this and that, unlike a and the, are varied according as the noun, with which they are associated, is in the singular or in the plural number. Thus we say, this and that man in the singular, and these and those men in the plural. The Latin articles hic and ille, for such we call them, are varied, like the Greek ει, not only with the number, but also with the gender of their nouns. In languages where the structure of a sentence may be so arranged from the order of nature, as it commonly is in Greek and Latin, and where the reader is guided, not by position, but by the terminations of the words, to those which are in concord and those which are not, these va- riations of the articles have their use; but in English they are of no importance. Were it not that the custom of the language, the forma loquendi as Horace calls it, has de- termined otherwise, there would be no more impropriety in saying this or that men, than in saying some men or the men.

As articles are by their nature definitives, it follows of course that they cannot be united with such words as are in their own nature as definite as they may be, nor with such words as, being undefinable, cannot properly be made overwise; but only with those words which, though in- definite, are yet capable through the article of becoming definite. Hence the reason why it is absurd to say, the I or the thou, because nothing, as will be seen afterwards, can make these pronouns more definite than they are of themselves; and the same may be said of proper names. Neither can we say, the both, because the word both is its own nature perfectly defined. Thus if it be said, "I have read both poets," this plainly indicates a definite pair, of whom some mention has been made already. On the contrary, if it be said, "I have read two poets," this may mean any pair out of all that ever existed. And since the numeral being in this sense indefinite, as in- deed are all others as well as itself, is forced to assume the article whenever it would become definite. Hence also its that as two, when taken alone, has reference to some primary and indefinite perception, whilst the article the has reference to some perception secondary and definite, it is bad language to say two the men, as this would be blending incompatibles, that is, it would be representing two men undefined and undefined at the same time. On the con- trary, to say, both the men, is good language; because the substantive cannot possibly be less apt, by being defined, to coalesce with a numeral adjective which is defined as well as itself. So likewise it is correct to say, the two men, these two men, or those two men; because here the article, being placed at the beginning, extends its power farther through the numeral adjective as the substantive, and tends equally to define them both.

As some of the above words admit of no article because they are by nature as definite as may be, so there are others which admit it not because they are not to be defined at all. Of this sort are all interrogatives. If we question about substances, we cannot say, the who is this, but who is this? And the same both as to qualities and quanti- ties; for we say, without an article, what sort of, how many, how great? The reason is, the article the respects beings of which we can predicate something; but interro- gatives respect beings about which we are ignorant, and of which we can therefore predicate nothing; for as to what we know, interrogation is superfluous. In a word, the natural associates of articles are all those common ap- pellatives which denote the several genera and species of beings; and it may be questioned whether in strictness of speech they are ever associated with any other words.

We have said that proper names admit not of the arti- cle, being in their own nature definite. This is true, whilst each name is confined to one individual; but as different persons often go by the same name, it is neces- sary to distinguish these from one another, to prevent the ambiguity which this identity of name would otherwise occasion. For this purpose we are obliged to have re- course to adjectives or epithets. For example, there were two Grecian chiefs who bore the name of Ajax; and it was not without reason that Mnestheus used epithets when his intention was to distinguish the one from the other. "If both Ajaxes cannot be spared," said he, "at least let mighty Telamonian Ajax come." But as epithets are diffused through various subjects, in as much as the same adjective may be referred to many substantives, it has been said to be necessary, in order to render both parts of speech equally definite, that the adjective itself assume an article before it, which may indicate a reference to some single person only. It is thus we say, Trypho the Grammarian, Apollodorus the Cyrenian, &c. This is the doctrine of Mr Harris, from which, with the highest re- spect for the learning of the author, we feel ourselves obliged to dissent. In the examples given, the article the is certainly not associated with the words Gramma- rian and Cyrenian, in the same manner in which it is as- sociated with the word man in the sentence, "The man that hath not music in himself." When we say Apollo- dorus the Cyrenian, we may, without folly or imperti- nence, be asked, the Cyrenian what? And the moment this question is answered, it will be seen that the article defines, not an adjective, but a substantive. If the answer be, the Cyrenian philosopher, the article the is associated with the word philosopher, and the phrase Apollodorus the Cyrenian is an abbreviation of Apollodorus the phi- losopher of Cyrene. In like manner, Trypho the gram- marian, is Trypho the grammarian writer, or Trypho the writer of grammar. Such abbreviations are very common. We say familiarly the Speaker, and are understood to mean a high officer in the British parliament; yet as speaker is a name common to many men, we may without impropriety be asked, what speaker we mean; and if so, we must reply, the Speaker of the House of Commons. But that which is eminent is supposed to be generally known; and there- fore, in common language, the speaker is deemed a suffi- cient designation of him who presides over the lower house of parliament. Hence, by an easy transition, the definite article, from denoting reference, comes to denote emi- nence also; that is to say, from implying an ordinary pre- acquaintance, to presume a kind of general and universal notoriety. Thus a king is any king, but the king is that person whom we acknowledge as our sovereign, the king of Great Britain. In Greek, too, as in English, the arti- cle is often a mark of eminence; for the Poet meant Homer, and the Stagyrite meant Aristotle; not but that there were many poets besides Homer, and many stagyrites besides Aristotle, but none equally illustrious. Pronouns. Before we dismiss the Article, we shall produce one example to show the utility of this species of words; which, although they may seem to be of small importance, yet, when properly applied, serve to make a few general terms sufficient for expressing with accuracy all the various objects about which mankind can have occasion to converse.

Let man be the general term, which I have occasion to employ for the purpose of denoting some particular. Let it be required to express this particular as unknown, I say a man; known, I say the man; definite, a certain man; indefinite, any man; present and near, this man; present and at some distance, that man; like to some other, such a man; different from some other, another man; an indefinite multitude, many men; a definite multitude, a thousand men; the ones of a multitude taken throughout, every man; the same ones taken with distinction, each man; taken in order, first man, second man, &c.; the whole multitude of particulars taken collectively, all men; the negation of that multitude, no man; a number of particulars present and near, these men; at some distance, or opposed to others, those men; a number of individuals separated from another number, other men; a small indefinite number, few men; a proportionally greater number, more men; a smaller number, fewer men. And so we might proceed on almost to infinitude. But not to dwell longer upon this subject, we shall only remark, that minute changes in principles lead to mighty changes in effects; so that principles are well entitled to regard, however trivial they may appear.

III.—Of Pronouns, or Substantives of the second order.

To men who are neither intoxicated with their own abilities, nor ambitious of the honour of building new systems, little pleasure can accrue from differing upon points of science from writers of great and deserved reputation. In such circumstances a man of modesty, although he will not upon the authority of a celebrated name adopt an opinion of which he perceives not the truth, must always advance his own notions with some degree of diffidence, as being conscious that the truth which he cannot perceive, may be visible to a keener and more perspicuous eye. In these circumstances we feel ourselves with regard to some of the most celebrated writers on grammar, from whom, concerning one or two points, comparatively indeed of but little importance, we have already been compelled reluctantly to differ. In treating of pronouns we are likely to deviate still farther from the beaten track; but that we may not be accused of acting the part of dogmatists in literature, and of claiming from others that implicit confidence which we refuse to give, we shall state with fairness the commonly received opinions, point out in what respects we think them erroneous, assign our reasons for calling them in question, and leave our readers to judge for themselves. The most celebrated writer in English who has treated of pronouns, and whom, since the publication of his Hermes, most other writers have implicitly followed, is Mr Harris, who, after a short introduction, proceeds thus:

"All conversation passes between individuals, who will often happen to be 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 this purpose. The first expedient upon this occasion seems to have been pointing, or indicating 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 pronouns. These also they distinguished into 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 subject of the conversation to be the speaker himself. Here, to supply the place of pointing by a word of equal power, the inventors of language 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."

The description of the different persons here given is taken, we are told, from Priscian, who took it from Apollonius. But whatever be the deference due to these ancient masters, their learned pupil, though guided by them, seems not to have hit upon the true and distinguishing characteristic of the personal pronouns. He supposes, that when the names of two persons conversing together are known to each other, they may, by the use of these names, express all that the personal pronouns express; but this is certainly not true. To us, at least, there appears to be a very material difference between saying "George did this," and "I did this;" nor do we think that the power of the pronoun would be completely supplied by the name, even with the additional aid of indication by the hand. So, when one man says to another with whom he is conversing, "James did so and so;" it is surely not equivalent to his saying "You did so and so." If such were the case, one might pertinently ask, when both persons are known to each other, Why do they use the personal pronouns? Mr Harris tells us, that "when the subject of conversation is the speaker himself, he uses I; and when it is the party addressed, he uses thou." But, in fact, the nature of the personal pronouns has no sort of connection with the subject of conversation, whether that conversation relate to the speaker, the party addressed, or a Greek book. In this sentence, "I say that the three angles of every triangle are equal to two right angles," the speaker is surely not the subject of the discourse; nor is the party addressed, but the truth of his assertion, the subject of discourse in the following sentence: "You say, that Horne Tooke's Diversions of Purley is the most masterly treatise on grammar, as far as it goes, that you have ever seen." Mr Harris uses the phrase "becoming the subject of conversation," in no other sense than that when the speaker has occasion to mention himself, he uses I; when the party addressed, thou; and when some other person or thing, he, she, or it. But we know that he may use other words by no means equivalent to the two first of these pronouns, which will sufficiently mark himself, and the party addressed; and that he may use indifferently, and without the smallest injury to the sense, either the third pronoun, or the word for which it is merely a substitute. A man who sustains various characters may design himself by any one of them. Thus Mr Pitt might have spoken of himself as the first lord of the treasury, chancellor of the exchequer, or member of the university of Cambridge; and in each case he would have been what Mr Harris calls the subject of conversation; yet every one feels that none of these designations would have been equivalent to I. What, then, is the use of the personal pronouns?

It appears to be simply this: The first denotes the speaker, as characterised by the present act of speaking, in contradistinction to every other character which he may bear. The second denotes the party addressed, as characterised by the present circumstance of being addressed, in contradistinction to every other character. And what is used the pronoun of the third person is merely a negation of the other two, as the neuter gender is a negation of both masculine and feminine. If this account of the personal pronouns be true, and we flatter ourselves that its truth will be obvious to everybody, there is but one way of expressing by other words the force of the pronouns of the first and second person. Thus, "The person who speaks to you did so and so," is equivalent to "I did so and so;" and "The person to whom I now address myself did so and so," is equivalent to "You did so and so."

Hence we see why it is improper to say the I or the thou; for each of these pronouns has of itself the force of a noun with the definite article prefixed, and denotes a person of whom something is predicted, which distinguishes him from all other persons. I is the person who now speaks, thou is the person who is now addressed by the speaker. Hence too we see the reason why the pronoun I is said to be of the first, and the pronoun thou of the second person. These pronouns can have place only in conversation, or when a man, in the character of a public speaker, addresses himself to an audience; but it is obvious that there must be a speaker before there can be a hearer; and therefore, that the pronouns may follow in order of nature, I, which denotes the person of the speaker, must take place of thou, which denotes the person of the hearer. Now the speaker and the hearer being the only persons engaged in conversation or declamation, I with great propriety called the pronoun of the first, and thou the pronoun of the second person. We have seen, that with respect to pronouns, the third person, as it is called, is merely a negative of the other two. This is evident from the slightest attention to the import of those words which are called pronouns of the third person. He, she, or it, denotes not the person either of the speaker or of the hearer; and, as we have just observed, no other person can have a share in conversation or declamation.

A absent person or an absent thing may be the subject of conversation, but cannot be the speaker or the person addressed. He, she, and it, however, as they stand by themselves, and assume the power of nouns, are very properly denominated pronouns; but they are not personal pronouns in any other sense than as the negation of sex or neuter gender.

We have already seen that nouns admit of number; pronouns, which are their substitutes, likewise admit of number. There may be many speakers at once of the same sentiment, as well as one, who, including himself, speaks the sentiment of many; speech may likewise be addressed to many at a time, as well as to one; and the subject of the discourse may likewise be many. The pronoun, therefore, of every one of the persons must admit of number to express this singularity or plurality. Hence Pronouns, the pronoun of the first person I, has the plural we; that of the second person thou, has the plural ye or you; and that of the third person he, she, or it, has the plural they, which is equally applied to all the three genders.

The Greeks and Romans, when addressing one person, used the pronoun in the singular number thou; whereas in the polite, and even in the familiar style, we, and many other modern nations, use the plural you. Although in this case we apply you to a single person, yet the verb must agree with it in the plural number; it must necessarily be, you have, not you hast. You was, the second person plural of the pronoun placed in agreement with the first or third person singular of the verb, is an enormous, though common solecism, which ought to be carefully avoided. In very solemn style, as when we address the Supreme Being, we use thou, perhaps to indicate that he is God alone, and that there is none like unto him; and we sometimes use the same form of the pronoun in contemptuous or very familiar language, to intimate that the person to whom we speak is the meanest of human beings, or the dearest and most familiar of our friends. A king, exerting his authority on a solemn occasion, adopts the plural of the first person, "we strictly command and charge;" meaning, that he acts by the advice of counsellors, or rather as the representative of a whole people. But in all cases in which the use of the pronoun deviates from the nature of things, the verb in concord deviates with it; for, as will be seen afterwards, these two words universally agree in number and person.

But though all these pronouns have number, neither in Greek, Latin, or any modern language, do those of the first and second person carry the distinctions of sex. The reason is obvious, namely, that sex, and all other properties and attributes whatever, except those mentioned above as descriptive of the nature of these pronouns, are foreign from the intention of the speaker, who, when he uses the pronoun I, means the person who now speaks, no matter whether man or woman; and when the pronoun thou, the person, no matter whether man or woman, to whom he now addresses himself; and nothing more. But the pronoun of the third person denoting neither the speaker nor the hearer, but the subject of the discourse, and being merely the substitute of a noun which may be either masculine, feminine, or neuter, must of necessity agree with the noun which it represents, and admit of a triple distinction significant of gender. In English, which allows its adjectives no genders, this pronoun is he in the masculine, she in the feminine, and it in the neuter; the utility of which distinction may be better found in supposing it away. Suppose, for example, that we should in history read these words, He caused him to destroy him, and were informed that the pronoun, 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 gender, how would it appear which was destroyed, which the destroyer, and which the cause that moved to the destruction? But there is no ambiguity, when we hear the genders distinguished: when we are told with the proper distinctions, that she caused him to destroy it, we know with certainty that the prompter was the woman; that her instrument was the hero; and that the subject of her cruelty was the unfortunate city. From this example we would be surprised how the Italians, French, and Spaniards could

The reason assigned by Mr Harris and his followers is, 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." This is perhaps the best reason which their description of the personal pronouns admits, but it is not satisfactory; for the speaker and hearer may meet in the dark, when different dresses cannot be distinguished. Pronouns express themselves with precision or elegance with no more than two variations of this pronoun.

Although in every language with which we are acquainted, there is but one pronoun for each of the first and second persons; and although it is obvious, from the nature and import of those words, that no more can be necessary; yet the mere English reader may perhaps be puzzled with finding three distinct words applied to each; I, mine, and me, for the first person; thou, thine, and thee, for the second. The learned reader will see at once that the words mine and me, thine and thee, are equivalent to the genitive and accusative cases of the Latin pronouns of the first and second persons. That mine is a pronoun in the possessive case is obvious; for if I were asked "whose book is that before me?" I should reply, "It is mine," meaning that it belongs to me. That the word me is the same pronoun in the case which the Latin grammarians call the accusative, is evident from the import of that word in the sentence he admires me, where the admiration is supposed to proceed from the person spoken of to the person who speaks. It appears, therefore, that though English nouns have only two cases, the nominative and possessive, the pronouns of that language have three, as, I, mine, me; thou, thine, thee; he, his, him, &c.; and that these are cases, can be questioned by no man who admits that mei, mili, me, are cases of the Latin pronoun ego. Both pronouns, the Latin and the English, are irregularly inflected; and perhaps those words which are called the oblique cases of each may have originally been derived from nominatives different from ego and I; but these nominatives are now lost, and mei and mine have, beyond all dispute, the effect of the genitives of the Latin and English pronouns of the first person. These variations, however, cannot be looked upon as an essential part of language, but only as a particular refinement, invented to prevent the disagreeable repetition of the pronoun, which must frequently have happened without such a contrivance. This seems to have been the only reason why pronouns have been endowed with a greater variety of cases than nouns. Nouns are in themselves greatly diversified. Every genus and every species of objects has a distinct name, and therefore the sameness of sound does not so often occur among them as it would among the pronouns, without cases, where the same I, thou, he, she, or it, answers for every object which occurs in nature; but by this diversity in the form of the words, the cacophonous, which would be otherwise disgusting, is in a great measure avoided. It is, probably, for the same reason that the plural of each of these pronouns is so very different from the singular. Thus from I, mine, me, in the singular, is formed, in the plural, we, ours, us; from thou, thine, and thee, ye or you, yours, you; and from he, she, it, his, hers, its, him, her, it, in the singular, they, theirs, them, in the plural. In all of which there is not the least resemblance between the singular and plural of any one word; and except in he, his, him; it, its; they, theirs, them; there is not any similarity between the different cases of the same word in the same number.

From the account here given of the personal pronouns, it appears that the first or second will, either of them, coalesce with the third, but not with each other. For example, it is good sense, as well as good grammar, to say in French any language, I am he, thou art he, we are they, you are they; but we cannot say, I am thou, nor thou art I, nor we are you, &c. The reason is, there is no absurdity for the speaker to be also the subject of the discourse, as when it is said, 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, is impossible; for which reason the coalescence of the pronouns of the first and second persons is likewise impossible.

I, thou, he, she, and it, are all that are usually called personal pronouns. There is another class of words, which are called sometimes pronominal adjectives, sometimes adjective pronouns, sometimes possessive pronouns; and by one writer of grammar they have been most absurdly termed pronominal articles. It is not worth while to dispute about a name; but the words in question are my, thy, her, our, your, their. These words are evidently in the form of adjectives; for, like other English adjectives, they have no variation to indicate either gender, number, or case; and yet they are put in concord with nouns of every gender and both numbers, as, my wife, my son, my book; her husband, her sons, her daughters, &c. But though in the form of adjectives, they have the power of personal pronouns in the possessive case: my book is the book of me, or the book of him who now speaks; our house is the house of us, or the house occupied by the persons who now speak; her husband, is the husband of a woman who can be known only from something preceding in the discourse; and their property is the property of them, of any persons, whether men or women, or both, who have been previously mentioned. Words which have the form of adjectives, with the power of pronouns, may, without impropriety, be called pronominal adjectives; and such is the name by which we shall henceforth distinguish them. To these pronominal adjectives, as well as to the personal pronouns, are subjoined the words own and self, and in the plural selves; in which case they are emphatical, and imply a silent contrariety or opposition. Thus, I live in my own house; that is, not in a hired house. This I did with my own hand; that is, not by proxy. This was done by myself; that is, not by another. The word self subjoined to a personal pronoun forms also the reciprocal pronoun; as, we hurt ourselves by vain rage; he blamed himself for his misfortune. Himself, itself, themselves, are supposed by Wallis to be put, by corruption, for his self, its self, their selves; so that self is always a substantive or noun, and not a pronoun. This seems to be a just observation; for we say, the man came himself, they went themselves; where the words himself and themselves cannot be accusatives, but nominatives, and were anciently written his self, their selves.

There are other words which are usually ranked under the class of pronouns; as who, which, what. These, when employed in asking questions, are called interrogative pronouns; though a name more characteristic might surely be found for them. Their import, however, will be more easily ascertained after we have considered another species of pronouns, which have been denominated relatives, and with which they are intimately connected.

The pronouns already mentioned may be called prepo-

---

1 If we mistake not, Dr Johnson has somewhere affected to ridicule Bishop Lowth for considering the word mine as the possessive case of the pronoun of the first person. According to the doctor, mine is the same word with the pronominal adjective my; and was anciently used before a vowel, as my was before a consonant. This is not said with the lexicographer's usual precision. That mine was anciently used before a vowel is certain; but it does not therefore follow that it is the same word with my. If it were, we might on every occasion substitute either of these for the other, without offending against grammar, however we might injure the sound; but we apprehend that this is not the case. "That book is mine," is good English; but "that book is my" would be a gross solecism. The reason is, that mine is a genuine pronoun, and stands by itself with the power of a noun; but my, being an adjective, cannot stand by itself. It is worthy of observation, however, that, repugnant as such expressions are to the present idiom of the English language, there is nothing in the nature of the thing that could render the use of them improper. All prepositions, as will be seen afterwards, are expressive of relations subsisting between those objects of which they connect the signs in discourse. Those objects may be denoted, either by single words, and then the preposition will govern a noun; or by assertions, and then it will govern a substantive as a verb. Thus, when it is said, "I came after his departure," the proposition after expresses the relation between two events, coming and his departure, and governs a substantive noun; but if it is said, "I came after he departed," the preposition in this case (for, as shall be shown afterwards, it is absurd to call it in the one instance a preposition, and in the other a conjunction) expresses the same relation as before, but governs a nominative and a verb.

This last expression is exactly similar to those employed above. When one says, for example, "the man of he speaks little is wise," however uncouth the expression may appear from its not being supported by the authority of custom, the preposition of is used precisely in the same manner, and serves the very same purpose, as when it is said, "the man of little speaking is wise." In both cases it denotes the relation between the two objects, man and little speaking; only in the one it is prefixed to a noun, in the other to an active clause of a sentence, the import of which is to be taken as a noun. Custom has indeed determined that prepositions shall more frequently govern a noun than a nominative and a verb; but they are, in their own nature, equally well adapted to answer both purposes.

But as the pronoun of the third person is merely the substitute of some noun, an objector may ask, What noun is here represented by he? The man of he speaks little is wise." Who is meant by the pronoun he? We answer, the man who is declared to be wise. The objection proceeds from inattention to the radical signification of the word of, which a late ingenious writer has shown to be the fragment of a Gothic or Anglo-Saxon word, signifying consequence or offspring. If this be admitted, and, after the proofs which he has given, we think it cannot be denied, the uncouth phrase, "The man of he speaks little is wise;" or, in other words, "The man, in consequence of his mind (he) speaks little, is wise;" or, in other words, "The man, in consequence of his speaking little, is wise." Mr Horne Tooke, has shown that of and for, though of different radical meanings, may often be substituted for one for the other without injury to the sense. Let this substitution be made in the present instance, and the propriety of the case will be apparent: "The man is wise for he speaks little." It must be remembered, however, that such a substitution cannot be made in every instance, because for signifies cause, and of signifies consequence. Pronouns. for that purpose they are well adapted. That pronoun seems to be of use only when there is a deficiency of adjectives or substantives to denote some complex attribute by which we want to limit a general term or expression. Where such adjectives or substantives exist in language, we may indeed use the relative or not at pleasure. Thus we may say, homo qui grandia loquitur, or homo grandilogus; because the adjective and the relative clause are precisely of the same meaning. But if the Latins were called upon to translate ἀνδρὸς καλότερος, we believe they must have made use of the relative pronoun, as we know not any corresponding adjective in their language.

The learned and ingenious Mr Harris, in his Treatise on Universal Grammar, has given an analysis of the relative pronoun very different from that which has been given by us. The result of his inquiry is, that the relative is equivalent to another pronoun, together with an expression of connection of that kind which is denoted by the particle and. This analysis he exemplifies, and endeavours to confirm, by the sentence, "Light is a body which moves with great celerity." Now, says he, instead of which substitute the word and it, and in their united powers you see the force and character of the pronoun here treated. But let any one attentively consider these two expressions, "Light is a body which moves with great celerity," and "Light is a body and it moves with great celerity;" and he will find that they are not precisely equivalent. For, to speak in the language of logic, there is in the first but one Proposition, of which the subject is light, and the predicate a complex term expressed by the words, body which moves with great celerity. In the second there are two Propositions, or two predications, concerning light; first, that it is a body; and, secondly, that it moves with great celerity. The relative clause, in the first case, expresses a property of the antecedent body, which, with that property, is predicated of the subject light; in the second case, this property is removed from the predicate of which it was an essential part, and is improperly converted into a new predication of the subject. The sentence may be resolved upon our principles, and its precise import preserved; as "Light is a body of it moves with great celerity;" the clause "it moves with great celerity," is conceived by the mind as having the force of an abstract substantive, and is connected with the antecedent body by the preposition of, answering to the termination of the genitive case. This abstract substantive thus connected expresses a quality of the body light. But by this example Mr Harris's doctrine is not exhibited in all its absurdity; it may be proper to try it by another.

Let us suppose that the following assertion is true: "Charles XII. was the only monarch who conquered kingdoms to bestow them on his friends." Here it is evident there is but one proposition, of which the predicate is expressed by the words, "only monarch who conquered kingdoms to bestow them on his friends;" so that the relative clause is a necessary part of the predicate, and has, like an abstract noun in the genitive case, the effect of modifying the general term monarch. Resolve this sentence on Mr Harris's principles, and you have two propositions of which the first is a notorious falsehood: "Charles XII. was the only monarch, and he conquered kingdoms to bestow them on his friends." But instead of and substitute of, thus, "Charles XII. was the only monarch of he conquered kingdoms to bestow them on his friends," and you preserve the true import of the expression.

Are there no cases, then, in which the relative may be resolved into the connective and with a prepositive pronoun? Undoubtedly there are, and we shall now endeavour to ascertain them.

Adjectives in language have two different effects upon the substantives to which they belong, according to the nature of the attribute which they express. If the attribute expressed by the adjective be competent to all the species of which the substantive is the specific name, it is plain that the adjective does not modify or limit the substantive, for this obvious reason, that nothing can modify which is not discriminative. Thus, when Horace says, Prata comis albicant pruinis, the adjective comis denotes a quality common to all boar-frost; and therefore cannot modify the substantive, because it adds nothing to the conception of which that substantive is the name. But when the attribute expressed by the adjective is competent to some individuals only of the species of which the substantive is the name, the adjective has then the effect of modifying or limiting the substantive. Thus when one says vir bonus, he makes use of an adjective which modifies the substantive, vir, because it expresses a quality or attribute which does not belong to all men.

The clause of a sentence, in which there is a relative, as it is in every other respect, so is it in this, equivalent to an adjective; it either modifies, or does not modify, the antecedent, according as the attribute which it expresses is or is not characteristic of the species to which the antecedent belongs. Thus, when it is said, "Man, who is born of woman, is of few days and full of trouble," the relative clause "who is born of woman," expresses an attribute common to all men, and therefore cannot modify. In like manner, when we say, "Socrates, who taught moral philosophy, was virtuous," the clause, "who taught moral philosophy," does not modify. In both these instances the relative clause might be omitted; and it might be said with equal truth, "Man is of few days and full of trouble," and "Socrates was virtuous."

But if it be said, vir sapit qui pauca loquitur, the relative clause, qui pauca loquitur, modifies the antecedent vir; for it is not affirmed of every man that he is wise, but only of such men as speak little. So "Charles XII. was the only monarch who conquered kingdoms to bestow them on his friends;" and, "the man that endureth to the end shall be saved;" with many more examples that will occur to every reader.

Now it will be found that it is only when the relative clause expresses such a property or circumstance of the antecedent as does not limit its signification, that the relative pronoun can be resolved into a prepositive pronoun with the conjunction and, and that in these cases the relative clause itself is of very little importance. Thus in the assertion, "Charles XII. was the only monarch who conquered kingdoms to bestow them on his friends," where the relative clause is restrictive, the who cannot be resolved into and he consistently with truth or common sense. But in the expression, "Man, who is born of woman, is of few days and full of trouble," the relative who may be so resolved, at least without violating truth, thus, "man is of few days and full of trouble, and he is born of a woman." The only difference between the sentence with the relative who, and the same sentence thus resolved, is, that, in the former case, it contains but one predication; in the latter two, and these but loosely connected.

Thus, then, it appears that the general analysis of the

---

1 Mr Harris was probably led into his opinion, from considering the Latin qui or quae as compounded of que and ei (See Horace, p. 81, 82, third edit.). But the notion of Perizonius is perhaps better founded, who, in his notes to the Menœceus of Sanctius, considers it as immediately taken from the Greek τις, which in the Doric is made εις, and in the Latin quis. For it seems highly probable, as some ingenious writers have endeavoured to show, that the Latin is a dialect of the Greek. Of this at least we may be certain, that many words in the former are immediately adopted from the latter. relative pronoun is into the particle of, and a prepositive pronoun; but that there are also occasions on which it may be resolved into a prepositive pronoun and the particle and, without materially altering the sense. Now, what is the reason of this distinction?

If the relative clause be equivalent to an adjective, or an abstract substantive in the genitive case, it is easy to see that the relative itself may, in every instance, be resolved into another pronoun and the particle of; but it is not perhaps so evident how it should in any instance be resolved by and. This last analysis has its foundation in the nature of the particles of and and; or, to speak more properly, in the nature of the attribute which the relative clause expresses. Both the particles of and and are used to link or join conceptions together; by this difference, that of has the effect of making all conceptions it connects figure in the mind as one object; whereas the conceptions connected by and are still conceived separately as before. To explain ourselves by example: suppose we take two words, man and virtue, which denote two distinct ideas or conceptions, and join them together by the particle of, saying man of virtue; it will no longer view them separately as significant of two conceptions, but of one. Take the same words, and join them together by the particle and, saying man and virtue; the conceptions denoted by man and virtue are still viewed separately as two, and notice is only given that they are collaterally connected.

This being the case, it follows, that when the relative modifies the antecedent, or, in other words, when the relative clause and the antecedent denote but one conception, the relative must then be resolved by of, in order to preserve this unity of conception. But when the relative does not modify the antecedent; that is, when its clause does not express any necessary part of a complex conception, then the conceptions or ideas denoted by the relative clause and the antecedent may be viewed separately as two; and therefore the relative may be resolved into the corresponding prepositive pronoun and the particle of.

To state this reasoning in a light somewhat different. Every relative clause, which expresses an attribute that is not applicable to a whole genus or species, must necessarily modify some general term, that is, restrict its signification; and as that general term must belong either to the subject or to the predicate of a proposition; it is evident, that every such relative clause is a necessary part of that subject or predicate in which its antecedent stands. If therefore a relative clause, which modifies, be taken away either from the subject or the predicate of a proposition; or if that connection, in consequence of which it modifies, be dissolved, which is always done when the relative is resolved by and; then the proposition itself will not hold true. The reason is, that the subject or the predicate becomes then too general; for, in the one case, something is predicated of a whole genus or species, which can be predicated only of some individuals of that genus or species; and in the other, a general predication is made where only a particular one can be applied. Thus, if it be said, "all men who transgress the laws are deserving of punishment;" the subject of the proposition is expressed by the words, "all men who transgress the laws." Take away the clause of the relative "who transgress the laws," and say, "all men are deserving of punishment;" and you have a proposition which is not true, because that is affirmed of the whole species which can be affirmed only of some individuals. Retaining now the clause of the relative, but resolving it by and, you have the same proposition as before; and together with it, in this instance, another which is equally false, namely, "all men, and thence transgress the laws, are deserving of punishment;" that is, "all men are deserving of punishment, and all men transgress the laws."

But when the attribute expressed by the clause of the relative is characteristic of the genus or species of the antecedent, and consequently applicable to every individual which that genus or species comprehends, the relative clause may be entirely omitted without affecting the truth of the proposition, which is already as general as it can be. As in this case the import of the relative clause is not restrictive of the signification of the antecedent, it is of little consequence whether the attribute be represented by the connective part of the relative, as of the antecedent, or be affirmed to belong to the antecedent in a separate assertion. Thus it matters not much, whether we say, "man, who is subject to death, ought not to be too much elated," that is, according to our analysis, "man of he is subject to death, ought not to be too much elated;" or whether, forming the relative clause into a separate assertion, and connecting the two by the particle and, we say "man, and he is subject to death, ought not to be too much elated." In the one sentence, indeed, the reason is implied why man should not be too much elated, his being subject to death; and in the other, no reason is assigned for this; we only affirm that man is subject to death, and likewise that he should not be too much elated; but as both affirmations are equally true and evident, it is of little consequence, in such a case as this, whether the reason upon which either is founded be implied or not.

From the whole of this investigation, the following conclusions may be deduced as sufficiently established: 1st, That the relative pronoun contains in itself the united powers of a connective and another pronoun. 2ndly, That of is the connective of which, together with another pronoun, it contains the powers, as in every possible instance it may be resolved into these constituent parts, and the import of the sentence in which it has place remain unaltered. 3rdly, That in the ancient languages the relative clause of a sentence has the import of an abstract substantive in the genitive case; in English, with the particle of prefixed. 4thly, That the relative pronoun is of necessary use only where there is a deficiency of adjectives or substantives to denote some complex attribute, by which we want to limit a general term or expression; but that where such adjectives or substantives exist in language, we may use the relative or not at pleasure. And, 5thly, That though, in cases where the relative clause does not limit a general term, the relative pronoun may, without violating truth, be analysed by and; yet such analysis is never proper, as it gives two predicates to the same subject, which, in the original proposition, had but one predicate.

If the clause of the relative be equivalent to an adjective, as in every instance it seems to be, it will naturally occur, that in the ancient languages, the relative should agree with its antecedent in gender, number, and case. They do agree for the most part in gender and number; in case they cannot often, because the very intention of introducing a relative into language is to represent the antecedent in a different case. Whenever we have occasion to use a substantive or noun in a clause of a sentence, and afterwards to express by another clause, in which there is a verb, an attribute of the object denoted by that substantive, we then employ the relative pronoun. Now it seldom happens that the two clauses admit of the same regimen; and hence the case of the relative is often necessarily different from that of the antecedent, as the case of each must be accommodated to the clause in which it is found. Thus we cannot say, "Deus qui colimus bonus est;" but, "Deus quem colimus bonus est;" because the regimen of the verb colo is always the accusative.

This shows the necessity of introducing a relative into Pronouns those languages which give inflexions to their nouns. Were all the nouns of a language indeclinable, there would be little occasion for a relative; and accordingly in English it is often omitted. Examples are frequent in our best authors. Suffice it to quote the following.

For I have business would employ an age.—Rowe. I had several men died in my ship of calamities.—Swift. They affect to guess at the object they cannot see.—Bolingbroke.

We are not ignorant that our most eminent grammarians consider such expressions as chargeable with impropriety; and we are far from recommending them in any dignified or solemn composition. But in the instances adduced there is not the smallest degree of obscurity; at least there is none occasioned by the omission of the relative. The reason seems to be, that the mind can easily, by an effort of its own, make the antecedent unite, first with the one clause, and then with the other. Thus, when it is said, "I have business would employ an age," the mind can, without any difficulty, as the word business has no inflexions, consider it first as the objective case after have, and then as the nominative to would employ; but this cannot be so easily done in the ancient languages, where the termination of the noun is changed by the variation of its cases.

Both in the learned and in the living languages the relative has different forms, corresponding to the different genders of nouns; and by these it gives notice whether it is applied to persons, or to things without life. Thus in the English language we say, The man or the woman who went to Rome; The tree which stands on yonder plain. It admits likewise, when applied to males or females, a variation of cases similar to that of the personal pronouns. Thus we say, The man whose book is now before me; The man or woman whom I saw yesterday; but the neuter admits of no such distinction,1 as we say, The tree which I saw, as well as The tree which stands on yonder plain. In modern languages the relative admits not of any distinction to denote number; for we say, The man or the men who came yesterday: The man or the men of whom I speak.

In English, the word that, which by some has been called a demonstrative pronoun, by others a pronominal article, and by us a definite article, is often used instead of the relative, as in the following examples: "He is the same man that I saw yesterday; he was the ablest prince that ever filled a throne." With regard to the principle upon which this acceptation of the word that depends, we beg to offer the following conjecture.

In English, from the phlegmatic arrangement of the language, occasioned by the want of inflexions and conjugations, the place of every part of a sentence is almost uniformly determined, and very little variety is allowed in the collocation of the words. The adjective is almost always placed in apposition with its substantive, and the nominative with its verb. In consequence of this uniformity in the collocation of the words, the mind acquires a habit of connecting in idea any kind of word with the place in which it is used to stand; and is naturally led to consider every word which stands in such a place as belonging to such a class. Hence it is, we imagine, that the definitive that passes into the nature of the relative pronoun; as in those instances in which it occupies the place of the relative, it was natural to consider it as having the same import. Yet the word that has undoubtedly in itself no more the force of the relative pronoun than the or this, or any other definitive whatever. In such expressions as the foregoing, it is not improbable that originally the clause of the definitive that, which we now call the relative clause, was thrown in as a kind of modifying circumstance, thus, "The book (I read that) is elegant;" where the speaker, finding the word book too general for his purpose, throws in a clause to qualify and restrict it, or to confine his affirmation to that particular book which he is then reading. We can easily suppose that through time the definitive that in such an expression might be transposed or removed from its own place to that of the relative; so that the expression would run thus, "The book that I read is elegant;" which would be considered as precisely equivalent to "The book which I read is elegant." This opinion is not a little confirmed by a similar use of the article in Greek, which, though undoubtedly a definitive like the English the, is often used instead of the relative pronoun. Numberless examples may be found in Homer and Herodotus, especially in the latter, who seldom uses what is properly called the relative. We shall produce one instance from each.

\[ \text{Εἰσίτω Ἀριστοῦ Ἀρχαιολόγῳ, τὸν περὶ τῶν Ἑλλήνων Στοιχεῖαν ἐπιμελῶς ἠκολούθησεν.—Homer.} \] \[ \text{Οὐαὶ γὰρ μὴ μάλιστα καταπονᾶται (scil. Ἀθηναῖοι) ὅτε σὺν αὐτῷ ἀποδεχόμενοι ἢ ἐπὶ Σάμου ἤ ἐπὶ Ἡρακλείου—Herodot.} \]

We have said that the interrogative pronouns, as they are called, who, which, what, are intimately connected with relatives; we now affirm, that the two first of these words are nothing but relatives, and that the last contains in itself the united powers of a relative and a definitive. With respect to cases, number, and gender, the words who and which, when employed as interrogatives, differ not from the same words when employed as relatives; and we hold it as a maxim, without which science could not be applied to the subject of language, that the same word has always the same radical import, in whatever different situations it may be placed. To understand this, it is necessary to observe, that all men have a natural propensity to communicate their thoughts in the fewest words possible; hence it follows, that words are often omitted which are necessary to complete the construction of the sentence; and this nowhere happens more frequently than in the use of who and which. In sentences where these words are confessedly relatives, we often find them without an antecedent; as,

Who steals my purse steals trash.—Shakespeare. Which who would learn, as soon may tell, &c.—Dryden. Qui Beatiem non edit, non tua carmina, Marci.—Virg.

That is, "He who steals my purse?" "Which he who would learn, as soon may tell?" and "Ille qui Beatiem non edit," &c. Such abbreviations occasion no obscurity, because from previous circumstances the hearer knows the mind of the speaker and the persons to whom he refers. But it is not with respect to the relative and antecedent only that such abbreviations have place; in sentences of a different form, whole clauses are sometimes omitted, whilst the meaning of the speaker is made sufficiently plain. Thus, when King Richard III. having lost his horse in battle, exclaims,

A horse! a horse! my kingdom for a horse! there is no complete thought expressed; but the circumstances in which the king then was, enabled those about him to understand that he wanted a horse. Accordingly Catesby answers him,

Withdraw, my lord; I'll help you to a horse.

In like manner, when a person asks a question, his expression is frequently incomplete; but the tone of his voice, or some other circumstance, enables us to ascertain his meaning, and to supply, if we please, the words that are omitted. Thus when it is said, An facisti? nothing more is expressed than If you did it; the Latin on being nothing else but the Greek ἂν, si; but some circumstance enables the person who hears it to know that the meaning is, "Say

---

1 "Whose" is by some authors made the possessive case of which, and applied to things as well as persons; I think, improperly—Lowth. Let us apply these observations to the words who and which. If these words be relatives, and if our analysis of the relative be just, it is obvious, that no complete meaning can be contained in the clause, "Who is your principal friend?" for that clause contains nothing more than the circumstance of being your principal friend applied to some unknown person; "of he is your principal friend." That this is indeed the case, every man may be convinced, by asking himself what he means by the interrogative who in such a sentence; for he will find it impossible to affix to it any meaning without supplying an antecedent clause, by which that which is called an interrogative will be immediately converted into the relative pronoun. The custom, however, of language, and the tone of voice with which the relative clause is uttered, intimates, without the help of the antecedent, the wish of the speaker to be informed by the person addressed of the name and designation of his principal friend; and we know that the sentence when completed is, "Tell me the name and designation of the person who is your principal friend." Again, when the prophet says, "Who is this that cometh from Edom, with dyed garments from Bozrah?" he utters part of a sentence, which when completed will run thus: "Describe the person who cometh from Edom (this is that person), with dyed garments from Bozrah." He sets a person coming from Edom, of whose name and designation he is ignorant; he calls upon some one for information concerning these particulars; and that there may be no mistake, he describes the unknown person as having "dyed garments from Bozrah;" but lest even that description should not be sufficiently accurate, he throws in the definitive clause, "this is that person," pointing at him, we may suppose, with his finger. Which, used as an interrogative, indicates a wish of knowing a particular person or thing out of more than one mentioned; as, "Which of the two did it?" that is, "Tell me the one of the two which did it?" for, in old English, which as a relative is often used, where in modern English we should say who; and that mode of speech is still retained when the antecedent is omitted, and the relative clause employed to indicate such a wish as that before us. What includes in itself the signification of a definitive and a relative pronoun; as, "from what has gone before, what follows may easily be guessed?" where the word what is equivalent to that which. When therefore we say, "What rude fellow is that?" our meaning is, "Describe that person who is that rude fellow." Upon the whole, then, it is evident, that the words called interrogatives are merely relative pronouns; and that interrogative sentences are relative clauses uttered in such circumstances as to enable the hearer to supply the antecedents necessary to complete the meaning.

In fine, we have seen that substantives are either primary or secondary; or, in other words, nouns or pronouns. Nouns denote substances, and those either natural, artificial, or abstract. They moreover denote things either general, or special, or particular; and a general or specific name is made to denote an individual by means of words called articles or definitives. Pronouns are the substitutes of nouns, and are either prepositive or subjunctive. The prepositive is distinguished into three orders, called first, the second, and the third person. The subjunctive, otherwise called the relative, includes the powers of all those three, having superadded as of its own the peculiarity of force of a connective.

IV. Of Verbs.

The words which we have hitherto considered are commonly called substantives primary or secondary, and definitives; because nouns are significant of substances; pronouns are the substitutes of nouns; and the article serves to ascertain the extent of the noun, and to determine whether on any occasion it be significant of a whole class of substances, or only of one individual. But substances are of importance to mankind only on account of their various qualities or attributes; for their internal texture is a thing of which we are profoundly ignorant, and with which we have no manner of concern. Thus, experience teaches us that certain vegetables are pleasant to the taste, and wholesome food; whilst others are unpleasant and poisonous. The former kinds are valuable only for their qualities or attributes; and they are the qualities or attributes of the latter that make them worthless or hurtful. A horse is strong, and swift, and docile, and may be trained to carry a man on a journey, or to drag a plough. It is for his strength, swiftness, and docility, that he is the most valuable of all quadrupeds. One man is brave, another learned, and another eloquent; and by possessing these different qualities or attributes, each is fitted for a different station in society. It is plain, therefore, that in contemplating substances, our attention must be principally bestowed upon their qualities, and that the words which serve to denote these qualities must be an essential part of language. Such words are in general called attributives, and are of three sorts, verbs, participles, and adjectives.

Of all the constituent parts of speech, none has given the grammarians greater trouble than the verb. The vast variety of circumstances which it blends together in one word throws very considerable difficulties in the way of him who attempts to analyse it and ascertain its nature; at the same time that, by its eminent use in language, it is entitled to all the attention which can be bestowed upon it. To the discussion of the verb, Mr Harris, whose notions of this as of the other parts of speech have been generally adopted by the subsequent writers on grammar, has dedicated a large proportion of his book, in which he has thrown out many excellent observations, mixed, as it appears to us, with several errors. We have already observed, that no man is ignorant when he uses what is called a verb and when a noun. Every schoolboy knows that the words is, lovethe, walketh, standeth, in English, and est, amat, amatur, ambulat, stat, in Latin, are verbs; he knows likewise that they are of different kinds, that some of them are said to be active, some passive, and some neuter. But it should seem that the first object of our investigation ought to be the characteristic of the verb, or that which all these words have in common, and which constitutes them verbs, distinguishing them from every other species of words. Now it is obvious to the slightest attention, that every verb, whether active, passive, or neuter, may be resolved into the substantive verb is, and another attributive; for lovethe is of the same import with is loving; walketh, with is walking; and amat, with amans est. But loving, walking, and amans, are not verbs; and hence it follows, that the characteristic of the verb, that which constitutes it what it is, and cannot be expressed by other words, must be that which is signified by the word is; and to us that appears to be neither more nor less than assertion.

Assertion, therefore, or predication, is certainly the very essence of the verb, as being that part of its office, and that part only, which cannot be discharged by other kinds of words. Every other circumstance which the verb includes, such as attribute, mode, time, it may be possible to express by adjectives, participles, and adverbs; but without a verb it is impossible to predicate, to affirm or deny, any one thing of any other thing. The office of the verb; then, when stript of all accidental circumstances, seems to be merely this: "To join together the subject and predicate of a proposition." Its powers are analogous to those of the sign \(+\) in algebra, which does not affect the separate value of the quantities between which it is placed, but only indicates their union or coalescence. To explain by an example: When we say, Cicero eloquens, Cicero wise: these are imperfect sentences, though they denote a substance and an attribute. The reason is, that they want an assertion, to show that such an attribute appertains to such a substance. But when we insert the word was, we join the substance and attribute together; we give notice that the wisdom and eloquence are applied to Cicero, and we do nothing more; we neither increase the wisdom nor diminish it, we neither make it real nor imaginary; for it was supposed in all its extent when the words Cicero and wise stood independent of each other. We may indeed use the verb in a form which implies not an assertion only, but likewise an attribute; as when we say, George writeth, or George walketh. But as whiteness or any other particular colour is not of the essence of a horse, an animal which is found of all colours; so, in the phrases quoted, the attribute, though implied, is not of the essence of the verb; for it may be equally well expressed by other words, as George is writing, and George is walking, which are phrases of the very same import with George writeth and George walketh.

In resolving every verb, whether active, passive, or neuter, into the substantive verb is and another attributive, we have the honour to agree with all the grammarians; but to the word is itself the learned author of Hermes has given a meaning which, as a verb, it does not admit. He observes, that before anything can be the subject of a proposition, it must exist; that all existence is either absolute or qualified, mutable or immutable; that the verb is can by itself express absolute existence, but never the qualified, without subjoining the particular form; and that it signifies both mutable and immutable existence, having in these cases different meanings, although the sentences which he gives as examples are evidently constructed in the same manner, and consist of the same parts of speech. His examples are: Of absolute existence, B is; of qualified, B is an animal; of mutable, This orange is ripe; of immutable, The diagonal of the square is incommensurable with its sides. But if predication be the essence of verb, all this is nothing to the purpose, and part of it is not true. It is not true that the verb is ever varies its signification; for it has as verb no connection with existence of any kind. All such circumstances are superadded to its verbal nature; or, to speak more accurately, we infer such circumstances from our previous knowledge of the objects concerning which the predication is made. When we say, "This orange is ripe," we do indeed mean, as Mr Harris observes, that it is so now at this present in opposition to past and future time; but it is not the verb is, but the definitive this, which fixes the time of maturity, as well as the place, of the orange. If we had said, Oranges are ripe, we might have been properly asked, When and where are they ripe? although the same verb is used in both sentences. Even in the sentence B is, absolute existence, the most simple of all, is inferred, and not expressed, by the verb; and the inference is made from this obvious principle, that when one utters a mark of predication, we naturally conclude that he means to predicate something of the subject. If he adds no specific predication, as B is noun, we apply to B the most general that we can; and what other species is so general as existence?

That the idea of existence, considered as mutable or immutable, is not contained in the verb is itself, but is derived from our knowledge of the objects concerning which the predication is made, appears manifestly from this, that if a person be supposed ignorant of the meaning of the words God and man, whilst he knows that of is, the uttering of the two propositions, God is happy, and This man is happy, will give him no notice of existence considered as mutable or immutable, temporary or eternal! His conclusion with respect to these modes of existence, if any such conclusion be drawn at all, must be derived entirely from his previous knowledge of the nature of God and the nature of man.

Some of our readers may possibly think this notion of verb too abstract and metaphysical; yet what other circumstance than mere predication is essential to that species of words? We say essential; for we are inquiring, not what is expressed by each individual verb, but what it is which is equally expressed by all verbs, and which distinguishes them from the other parts of speech. And if it be true, that every thing which the verb implies, predication alone excepted, may be expressed by other parts of speech, and that no other parts of speech can predicate, then we think ourselves warranted to affirm, that simple predication is the essential characteristic of verb; that every word which predicates is a verb; and that nothing is so which does not predicate.

It must not, however, be concealed, that a doctrine very different from this has been maintained by a writer of distinguished abilities. "We have energy expressed," says Dr Gregory, "and of course a verb constituted without affirmation, when we wish or command; without command, when we affirm or wish; without wish, when we command and affirm; yet in all these cases we have equally and indisputably a verb."

That in all these cases we have a verb, is indeed indisputable; but we hold it to be equally indisputable, that in all these cases we have affirmation. The ingenious author has given no direct example of a wish or command uttered without affirmation; and a feeling or sentiment which is not uttered has nothing to do with language; but he has given a sentence in which there are three verbs, that in his opinion denote no affirmation, but a very plain supposition. If a supposition can be expressed without affirmation, we shall very readily allow that a wish or command may be so expressed likewise. The doctor's supposition is thus expressed: "Had any punishment ever overtaken you for your broken vows; were but one of your teeth growing black, or even were but one of your nails growing less beautiful, I should believe you." It is almost superfluous to observe, that to every verb not in the infinitive mode there must be a nominative, and to every active verb an object, whatever be the arrangement of the sentence in which such verbs are found. These are truths known to every schoolboy; the reasons of them shall be afterwards given. It is likewise undeniable, that in the sentence before us the nominative to "had" is "any punishment;" to the first, "were one of your teeth;" and

---

1 The truth of this observation may be proved by experiment, as, for instance, by uttering to a man of good common sense these two propositions, taking care to express the words God and man in a language which he does not understand. Thus, Deus is happy, and hic homo is happy, uttered to a man totally unacquainted with the Latin tongue, will contain no notice of existence considered as mutable or immutable.

2 Theory of the Moods of Verbs, published in vol. ii. of the Transactions of the Royal Society of Edinburgh. the second, "one of your nails." But the sentence arranged in grammatical order, with the several nominatives before their respective verbs, is evidently elliptical; and the conjunction if must be supplied as well to complete its construction as to make sense of the passage. "If any punishment had ever overtaken you; if but one of your teeth were growing black, or even if but one of your nails were growing less beautiful, I should believe you." Now it has been proved, by such evidence as leaves no room for doubt, that if, though called a conjunction, is in fact a verb in the imperative mode, of the same import with give; so that we may substitute the one for the other without in the smallest degree altering the sense. The sentence will then run thus: "Give any punishment had ever overtaken you; give but one of your teeth were growing black, &c. I should believe you." It is therefore so far from being true, that had and were, when the sentence is completed, express no affirmation; that it is only upon granting the truth of the affirmation which they denote that the speaker says, "I should believe you." "Any punishment had ever overtaken you," is plainly an affirmation; if give that affirmation, admit its truth, "I should believe you." But it cannot be supposed that had and were change their significations by a mere change of place, or fit by being removed from the middle to the beginning of clause they lose their original import, and come to denote something entirely different. Were this the case, every attempt to ascertain and fix the general principles of grammar would be as ridiculous as an attempt to arrest the course of time. For what purpose, then, it may be asked, if the verb always denotes affirmation, is it removed from the middle to the beginning of the clause, when supposition is implied, as in the present instance? We answer, that supposition is neither more nor less than conditional affirmation; that when such affirmation is completely expressed, the verb is not removed to the beginning of the clause; and that such removal takes place only when the clause is elliptical, being merely an artificial contrivance in language, to show the reader or hearer the same word as if, demanding the truth of the affirmation, is omitted for the sake of dispatch. This is evident; for when the word requiring the affirmation to be granted is supplied, the verb must be restored to its place in the middle of the clause. Such abbreviations, or such contrivances to mark them, are frequent in all languages, as will be seen more clearly when we come to treat of modes and moods. Upon the whole, then, we are compelled reluctantly to differ from this writer, and still to think that simple predication is the essence of the verb.

Should we be required to exemplify our theory by language, and to produce instances of this simplified verb in practice, we might answer, that the not being able to produce such instances would be no good argument against the truth of our principles. It is the nature of language to express many circumstances by the same word, all of which, however, are not essential to distinguish the species to which that word belongs from the other species of words; and it is the nature of man to infer from discourse many things which are not actually expressed. Perhaps, however, something very nearly approaching to an exemplification of our idea of a simple verb will be found in the proposition, "The three angles of every plane triangle are equal to two right angles." What other office the verb are here performs than simply to join the subject and predicate, it is difficult to perceive. It does not give notice of time; or such notice, if given, is an imperfection; for the truth of the proposition is independent of time. Neither ought it to imply existence; for the proposition would be true were there neither a triangle nor a right angle in nature.

This idea of verb, when well considered, will be found just; but should any of our readers suspect it of novelty, and on that account be disposed to condemn it, we have only to request that he will restrain his censure till he has examined the writings of others, and nicely observed the several states of his own mind in discourse; for meditation may perhaps show him that our theory is not false, and inquiry will satisfy him that it is not novel.

But although it is certain that assertion, and assertion only, is essential to the verb, yet the greater part of that species of words which grammarians call verbs are used to denote an attribute as well as an assertion; or, in the language of logic, they express both the copula and the predicate of a proposition; thus, he liveth, he writeth, he walketh, are phrases equivalent in all respects to he is living, he is writing, he is walking. Now, of attributes, some have their essence in motion, as walking; some in the privation of motion, as resting; and others have nothing to do with either motion or its privation, as white and black. But all motion and all privation of motion imply time as their concomitant; and a substance may have an attribute to-day which it had not yesterday, and will not have to-morrow. This is self-evident; for a man may be at rest to-day who yesterday was walking, and to-morrow will be on horseback; and a sheet of paper may have been white yesterday which to-day is black, and at some future time will be of a different colour. As, therefore, all motions and their privations imply time; and as a proposition may be true at one time which is not true at another, all verbs, as well those which denote both an attribute and an assertion, as those which denote an assertion only, come to denote time also. Hence the origin and use of tenses, which are so many different forms assigned to each verb, to show, without altering its principal signification, the various times in which the assertion expressed by it may be true. Whether these various forms of the verb be essential to language, it is vain to dispute. They have place in every language with which we are acquainted; and as the use of the verb is to affirm one thing of another, it is absolutely necessary that the time when such or such an affirmation is true be marked by tenses, or some other contrivance. Concerning tenses, therefore, we shall throw together some observations, which, mutatis mutandis, are applicable to every language, nemissum only a general remark or two which seem necessary in order to proceed with precision.

Time, although its essence consists in succession continued and unbroken, may yet be considered by the mind as divided into an infinite number of parts. There is, however, one grand division which necessarily occurs, and to which the different tenses of verbs are in all languages adapted. Computing from some portion conceived to be

---

"Besides words, which are names of ideas in the mind, there are a great many others that are made use of to signify the connection that the mind gives to ideas or propositions one with another. The mind, in communicating its thoughts to others, does not need signs of the ideas it has thus before it, but others also to show or intimate some particular action of its own at that time relating to those ideas. This it does several ways; as is and is not are the general marks of the mind affirming or denying;" (Locke on Human Understanding.) "Verbum est pars orationis variabilis, aliquid de re aliqua dici seu affirmari significat. Vulgaris verbi definitio est, quod, sit pars orationis, quae agere, pati, aut esse significet. Sed nostra acculturatio, magisque ex ipsa verbi cujusvis natura petita videtur. Ceterum si affirmari laxiore hie sensu accipimus, pro eo quod predicari dialectici appellant, quo non nisi affirmaciones strictius sic dictae, sed negationes etiam interrogationesque includantur." (Ruddiman, Grammaticae Institutiones; see also Beattie's Theory of Language.)

OL. X. present, all time is either past or to come. Hence the tenses of verbs are threefold; some denoting time present, some time past, and others time future.

Again, from the very nature of time, it must be obvious that all its parts are relative; or, in other words, that no portion of it can be ascertained by any thing inherent in itself, but only by referring it to some other portion, with respect to which it is past, present, or to come. In this respect time is perfectly analogous to space; for as the space in which any object exists cannot be described but by stating its relation to some other space, so neither can the time of any attribute or action be determined, but by stating its relation to some other time. When, therefore, we would mark the time of any action or event, we must previously fix upon some point to which we may refer it. If this point be known, the time referred to it will also be known; but if the former be not known, neither will the latter.

Lastly, in contemplating an action, we may have occasion to consider it as going on, or as finished. This distinction is likewise denoted by the different tenses of verbs. In treating, therefore, of the tenses, there are two things to which attention ought principally to be turned—the relation which the several tenses have to one another in respect of time; and the notice which they give of an action's being completed or not completed.

Having premised these remarks, we proceed now to the tenses themselves; of which Mr Harris has enumerated no fewer than twelve. Of this enumeration we can by no means approve; for, without entering into a minute examination of it, nothing can be more obvious, than that his inceptive present, I am going to write, is a future tense; and his completive present, I have written, a past tense. But, as was before observed of the classification of words, we cannot help being of opinion, that to take the tenses as they are commonly received, and endeavour to ascertain their nature and their differences, is a much more useful exercise, as well as more proper for a work of this kind, than to raise, as might easily be done, new and hypothetical theories on the subject.

It has already been observed that all the tenses must necessarily mark relative time. In one sense this is extremely obvious. The present tense is used in contradistinction to both the past and future, and marks an attribute or action as existing in neither. The past and the future are in like manner used in contradistinction to the present, and mark an attribute or action which exists not now, but which in the one case has existed formerly, and in the other will exist at some time coming. But besides this relation of contradistinction subsisting among the tenses, there is another of co-existence, as we may call it, to which it is of great consequence to attend, especially in examining the nature of the present.

The present tense refers not only to something which is past or future, but also to something with which the attribute or action of the verb is contemporary. This reference is necessarily implied in its very name; for we cannot say of any thing that it is present, without implying at the same time that there is something else with which it is present. Hence it appears with how little reason Mr Harris and others have given us an aorist of the present, as marking present time indefinitely in contradistinction to other presents, which have been called inceptive, extended, and completive presents. For from what has been said, it follows, that the present tense is necessarily and from its very nature perfectly indefinite, and can of itself give notice of no precise or determinate portion or point of time whatever. A thing may have been present fifty years ago, may be present now, or at any future period. This tense implies the relation of co-existence between two or more things; but without some auxiliary circumstance, it cannot in any language mark the particular portion of time in which those things exist. The indefinite nature of this tense is indeed most clearly seen in that use of it in which Mr Harris has styled it the aorist of the present; that is, in cases where it is employed to denote the repetition of an action which the agent is accustomed frequently to perform, or to express propositions of which the truth is evinced by general experience; as in the following examples:

Hypocrisy...the only evil that walks Invisible, except to God alone.

Ad peritendum proposita qui eato judicat.

In these instances it is plain that there is no particular time pointed out; the propositions are true, or apprehended as true, at all times. Although the actions, therefore, of "walking" and "hastening" are expressed as present, it is impossible from the expressions to determine any precise point of time when they are present.

But if the present tense be thus indefinite, how, it may be asked, are we to ascertain the particular time which is intended? We answer, it is to be ascertained, either by stating the action of the verb as existing in some time already known, or by inference. If, for example, we say, "Millions of spiritual creatures walk the earth unseen," the proposition is general, and the time of walking undetermined. But if we add, "both when we wake and when we sleep," the time is by this addition ascertained and specified; for if the time when men wake and sleep be known, the time when these spirits walk the earth is also known. When no specifying clause is given by which to determine the time of the present tense, it is very commonly determined by inference. Thus, if one use such an expression as "He sleeps while I am speaking to him," the time of his sleeping is ascertained by the subsequent clause of the sentence; but if it be said simply, "he sleeps," without assigning any data from which it may be concluded when his sleeping is present, we very naturally infer that it is at the instant when we receive the information of his sleeping. Such inferences as this are common in language. The mind is desirous to obtain complete information on every subject, and therefore frequently supplies to itself what is not expressed in the speech of others.

Both these ways of ascertaining the precise time of the present tense are excellently illustrated by the use of the word present as applied to space. Take a familiar example: "His brother and he were present when I read the letter." It is at first sight evident that this expression is perfectly indefinite. But if it be said, "His brother and he were present at your house when I read the letter," the place of action is then determined by being referred to a portion of space which is known. If no such reference be made, the person who hears the speech uttered must either remain ignorant of the place intended, or he must ascertain it to himself by inference; and he will probably infer it to be that in which the speaker is at the time of his uttering the indefinite sentence. This leads us to observe that such inferences are not often made without sufficient foundation. Various circumstances may assist the reader or hearer in making them, and prevent all danger of mistake. He may have the evidence of sense, or of something preceding in the discourse, and a number of other particulars, to warrant his conclusion. Thus, if, when sitting by a large fire, any one pronounce the words "I am too warm," those to whom he addresses his speech are authorized to conclude that he is too warm at the time of speaking, unless he expressly prevent the drawing of that conclusion by adding some such clause as "when I wear a great coat."

It is strictly demonstrable, and has been in fact demonstrated by Mr Harris, that there is no such thing as present time. Yet do we not only conceive time as present and existing, but frequently as extended to a very great degree. We speak not only of the present instant, or the present day, but also of the present year, and even of the present century. This manner of conceiving time is indeed loose and unphilosophical, but it is sufficient for the ordinary purposes of language. To express time as it really is, we ought to say, the passing day, the passing year, and the passing century; but in common discourse we denote any portion of time present, in which the present now or instant is included, although it is obvious that part of that portion is past, and the remainder of it future. From the very nature of time thus conceived to be present, the tense now under consideration must represent the action of the verb as commenced, and not finished; for as time is in continued succession, and accompanies every action, when any action is not commenced, it exists in any time, although it may exist hereafter in time which is now future; and when it is finished, it exists no longer in time present, but in time past. Hence the absurdity of introducing into a theory of the tenses an inceptive present and a completive present; for these terms imply each a direct contradiction.

After having said so much respecting the present tense, we shall have but little to say of the preter-imperfect. It states an action in regard to time as past; and in respect of progress, as unfinished. Legebam, I was reading at some time, but my reading was then incomplete; I had not finished the book or the letter. We must here observe, however, as we did with respect to the present tense, that although the preter-imperfect represents the action as past, it does not inform us in what precise portion of past time the unfinished action was going on; this circumstance must either be given in separate words, or be inferred by the hearer. If one say simply, Legebam, the person to whom he addresses his speech will conclude that the time of his reading is past with respect to the present time of his speaking. But if he say Legebam antea venisti, he expressly states the action of reading as past with respect to the time in which his hearer came to the place where they both are at the time of speaking. The time of the preter-imperfect is always past with respect to the present instant when the imperfect is used, and of this the tense itself gives notice; but it may also boast with respect to some other time, and of this it conveys no information.

If we join two preter-imperfects together, the expression will state the co-existence of two progressive actions, both of which were going on at a time past in respect of some determinate time given or supposed. Cum tu scribis ego legebam; "when you were writing I was reading." Hence the preter-imperfect has by some grammarians been called the relative present; a name which, however, is by no means exclusively applicable to this tense. When the preter-imperfect is by the conjunction et joined in the same sentence with a plusquam-perfect, the two tenses express two actions, both prior to the time of speaking; but the one as having continued after the other was finished. Thus Æneas, speaking of the destruction of Troy, says, that after having escaped with his father and followers, he returned to the city in quest of his wife, and went directly to his own house; but there, contrary to his expectation, Danai, et teuctum omne tenebant, "the Greeks had rushed in," that action was over and completed before his arrival; but the act of "possessing the whole house," tenebant, was not over, but still continuing.

But it is necessary that the verb should denote actions which were complete or perfect in past time, as well as those which were incomplete or imperfect. For this purpose Greek and English verbs have an aorist, a preter-perfect, and a plusquam-perfect. Of these the Latin has only the two last. The preter-perfect in that language sustains a twofold character; it performs the office of the Greek and English aorist, as well as that of the preter-perfect properly so called; that is, it denotes a finished action at some indefinite past time, as well as at some time which is both past and definite.

In attempting to analyse the signification of complex terms, by which we here mean words that include in their signification a variety of particulars, it is of great advantage to have these particulars separately expressed by different words in another language. Now the English has resolved the tenses, which in the Greek and Latin languages are denominated the aorist and the preter-perfect, by means of what are commonly called auxiliary verbs, expressing the former by the verb did, and the latter by the verb have. In examining therefore the aorist and preter-perfect, it will be of use to inquire into the import of these verbs.

Did is evidently the aorist of the verb to do; a verb of the most general signification, as it denotes action of every kind. It expresses the finished performance of some action, the completion of which must of course have taken place in some portion of past time: "I did write, or I wrote (these expressions being equivalent) yesterday, a month, a year ago." But the import of did being so very general, it can convey no determinate meaning without being limited by the addition of some particular action; and this addition, however expressed, is to be considered in the same light as an accusative case, governed by the active verb did; for it produces exactly the same effect. ἔγραψα, scripsi, I did write; that is, "at some past time I performed the action of writing, and finished it."

The verb have, which is included in the preter-perfect, is plainly a verb of the present tense denoting possession, but a man may possess one thing as well as another; and therefore have requires limitation, for the very same reason that did requires it, namely, because its signification is perfectly general. Now this limitation, whatever it is, must be conceived as the thing possessed; and in instances where have is limited by a noun, this is obvious, and universally acknowledged. "I have a gold watch," means "I possess a gold watch." But to annex the same meaning to the word have, when used as an auxiliary verb, is an idea we believe not common, and which may perhaps be thought whimsical; yet what other meaning can be affixed to it? To suppose that words have not each a radical and determinate signification, is to suppose language a subject incapable of philosophical investigation; and to suppose, with Mr Harris, that there are words entirely devoid of signification, is at once to render all inquiries into the principles of grammar nugatory and ridiculous.

We conceive, then, that each of the phrases, ἔγραψα ἐπιστολήν, scripsi epistolam, I have written a letter, is equivalent to the phrase, "I possess at present the finished action of writing a letter." Such an expression may sound harsh to the ear, because it is not in use; but we often employ expressions, to the precise and proper meaning of which we do not attend; and if the above be attentively considered, however awkward it may at first appear, nothing will be found in it either improper or absurd.

The aorist, then, we conceive to state an action as performed and finished in some past portion of time; whilst the preter-perfect represents the past performance and completion of that action as now possessed. And here we may hazard a conjecture why have, when used as an auxiliary verb, is always joined with a past participle; whereas did is joined to a word expressing the simple action of the verb, or, as it is called, present infinitive. Of the expression, "I have written a letter," one part, namely, the verb have, denotes present time; the other part, viz. written, must denote past time, to give notice that the action is performed and finished. Did, on the other hand, implying past time, has no occasion for the past part of another verb to give notice of this circumstance; for "I did write a letter," is equivalent to, "at some past time I performed and finished the simple action of writing a letter."

The principal distinction in practice between the aorist and prater-perfect (for the difference seems little in their real import) consists in the time by which the performance of the action admits of being particularly specified. The prater-perfect is always joined with a portion of time which includes the present now or instant; for otherwise it could not signify, as it always does, the present possession of the finishing of an action. But the aorist, which signifies no such possession, is as constantly joined with a portion of past time which excludes the present now or instant. Thus we not only say, "I have written a letter this day, this week," but "I wrote a letter yesterday, last week;" and to interchange these expressions of time in Greek and English, where the aorist and prater-perfect have different forms, would be improper. In Latin, indeed, where they have but one form, the impropriety does not appear.

Besides the tenses already examined, which are expressive of past time, in most languages the verb has another tense called the plusquam-perfect, in which, however, no difficulty occurs to detain our attention. What the prater-imperfect is to the present tense, that the plusquam-perfect is to the prater-perfect. The verb had, by which it is resolved in English, being evidently the past time of have, sufficiently explains its meaning and relation to the other tenses. "I had written a letter," is equivalent to the phrase, "I possessed at some past time the finished action of writing a letter." It is justly observed by Dr Beattie, that the imperfect and plusquam-perfect are very useful, and may be the sources of much elegant expression; and that if one were not taught to distinguish, in respect of meaning as well as of form, these tenses from each other, and the praterite from both, one could not pretend to understand, far less to translate, any good classic author.

Having considered the tenses which imply present and past time, it now remains that we examine the import of those which are expressive of time future. In Latin and English there are two tenses for this purpose, of which the first represents an action in point of time as not yet existing, but as about to exist at some period to come; but it does not bring the completion of the action into view. The other asserts the futurity of an action together with its completion. Scribam, "I shall be writing," denotes future time and incomplete action; for it does not say whether I am to write for a long or for a short time, or whether I shall finish what I promise to begin. This part of the verb, therefore, to which the Greek ἔργον corresponds, is an imperfect future, or, in other words, indefinite. The futurity of any action, it should seem, may always be computed from the time of speaking; for every action must be future with respect to the time at which its futurity is declared; but the time of its futurity may be more precisely specified by fixing upon some other future time to which to refer it: "I shall be writing after he shall have departed." Shall or will refers to future time indefinitely; and write or writing refers to an action which is indeed to begin and so far to proceed, but of which nothing is said concerning the completion.

On the other hand, scripsero, "I shall have written," is a perfect future denoting complete action; for shall denotes future time; written, finished action; and have, present possession. So that the meaning of the whole assertion is, that "at some future period of time I shall possess the finished action of writing." The completion of the action, together with the possession of it, is always future with respect to the time of assertion; but, with respect to some other time expressed or understood, the completion of the action is to be past: Promittis te scribaturum si rogavero, "you promise to write if I shall have asked you." In this sentence the action of asking is future with relation to the time of promising, but it is past with relation to that of writing. This tense the Latin grammarians call the future of the subjunctive mode, but very improperly. The notice which it communicates respects not the power or liberty of acting, which, as will be seen by and by, is the characteristic of that mode; but only the action itself. It ought, therefore, to be ranked among the tenses of the indicative mode; for scripsero is, in every sense, as really indicative as scribam or scripseras ero.

These are all the tenses, essentially different from each other, which have place in the indicative mode (or mood) of any language with which we are acquainted; but as there are tenses in the mode called subjunctive, which bear the same names with those already examined, and which have yet a different import, it will be necessary to consider them before we dismiss the subject of tenses. Of modes in general something will be said hereafter; at present we shall only observe, that the mode with which we are now concerned is not very properly distinguished by the name assigned to it by the Latin grammarians. They call it the subjunctive, because it is often subjoined to another verb, and forms the secondary clause of a sentence; but the mode called indicative frequently appears in the same circumstances. The difference between these two modes appears to us to consist in this, that the indicative asserts something directly concerning the action, and the subjunctive, something concerning the power or liberty of the agent to perform it; for that the latter asserts as well as the former, admits not of dispute.

The present tense of the subjunctive mode, in the learned languages, answers to the English auxiliaries may and can. Let us consider these a little. May is evidently a verb of the present tense denoting liberty. When I assert that I may write, I give notice that I am under no compulsion to abstain from writing; that there is no impediment from without by which I am restrained from writing. Can is also a verb of the present tense, expressive of internal power or skill. "I can write" is equivalent to such an

---

1 "It will perhaps occur," says an elegant and ingenious writer, "that there are two Greek tenses of which I have given no account; namely, the second aorist and the second future. The truth is, that I consider them as unnecessary. Their place, for anything I know to the contrary, might at all times be supplied by the first aorist and the first future. Some grammarians are of opinion that the first aorist signifies time past in general, and the second indefinite time past; and that the first future denotes a nearer, and the second a remote, futurity. But this, I apprehend, is mere conjecture, unsupported by proof; and therefore I decline rather to adopt the sentiments of those who teach that the second future and the second aorist have no meaning different from that first future and the first aorist; and that they are the present and imperfect or some obsolete theme of the verb; and when the other themes came into use, happened to be retained for the sake of variety, perhaps, or by accident, with a pretence and future signification. Be this as it will, as these tenses are peculiar to the Greek, and have nothing corresponding to them in other tongues, we need not scruple to overlook them as superfluous." (Beattie's Elements of Language, part ii. chap. 2.) To these judicious observations we have nothing to add, except that they acquire no small degree of countenance from the circumstance that there are many Greek verbs which have no second future, and which are yet employed to denote every possible modification of future time. Of the pseudo-past-futures of the Greeks we have taken no notice, because it is found only in the passive voice; to which, if it were necessary, it is obvious that it would be necessary in all voices, as a man may be about to act as well as to suffer immediately. pression as this: "There is nothing in myself which incapacitates me for performing the operation of writing." This verb seems originally to have denoted knowledge or skill, and to have been afterwards extended to signify power or ability of any kind. There is little doubt of its being the same with the old English verb to con, which signifies to know. The difference between the import of these two verbs, may and can, will be best perceived in a familiar example. Suppose we say to any one, "You may write a treatise on grammar," to which he returns an answer, "I cannot;" our assertion evidently supposes him at liberty to write the treatise; his answer implies that he is unable or unskilled to do it. We may conclude, then, that the present tense of this mode contains a declaration of present liberty, ability, or skill; and its other tenses will be found to have reference to the same capacities.

The observation is here to be repeated which was enlarged upon when treating of the present of the indicative. The liberty or ability signified by this tense is always represented as present, but the time of this presence is indefinite. If no particular time be specified, we generally refer it to the time of speaking; but another point may be given from which we are to compute. "When he shall have finished, you may then proceed as you propose." Here the liberty of proceeding is stated as present, not at the time of speaking, but at the time of his finishing, which is future to the time of speaking. But though the liberty, ability, or skill, denoted by this tense, be represented as present, the action itself is stated as contingent; for it's not necessary that a man should perform an action because he has the capacity to perform it.

From this idea of the present of the subjunctive some of its most peculiar uses seem capable of being explained. And, in the first place, it appears to have a near affinity with the future of the indicative, insomuch that in many instances they may be used promiscuously. Without materially altering the effect of the expression, we may say, "Dico me facturum esse quo imperaret," or "quo imperaret." The reason of this, perhaps, may be, that with respect to us, futurity and contingency are in most cases nearly the same, both being involved in equal obscurity; and therefore it is often of little consequence which mode of expression we employ.

Secondly, the present of the subjunctive is used to denote the right of which a person is possessed. "I may, or can, sell this book." This application, which Dr Priestley considers as the primary signification of the tense, is easily deduced, or rather it follows immediately from the foregoing account of its import. For if one be under restraint, either external or internal, to prevent him from performing an action, he has surely a right to perform it.

Thirdly, the present of the subjunctive is often used to signify command or request; as when one says, "You may give my compliments to such a person." This use of the tense under consideration seems to have arisen from a desire to soften the harshness of a command, by avoiding the appearance of claiming superiority. When a man utters the above sentence, he certainly utters no command, but only asserts that the person to whom he speaks has liberty or power to do him a favour. This assertion, however, may contain no new information; and therefore the person addressed, reflecting upon the intention of the speaker in making it, infers that it indicates a wish or desire that his compliments should be made to such a person.

Of the subjunctive as well as of the indicative, the preterperfect is evidently the past time of the present. As the latter asserts liberty or ability to perform some action, as existing at present, the former asserts the same liberty or ability to have existed in time past; but the precise portion of time past in which these capacities existed must be specified by other words, or it will remain unknown. Thus, in the following sentence, "Dixi me facturum esse quo imperaret," the time of imperaret is referred to that of dixi; the person having the right to command is supposed to have had it at the time when the other said that he would obey. This tense, as well as the present, states the action as going on and incomplete; and also as future with respect to the liberty or ability to perform it. It is rendered into English by the verbs could or might, of which the first is the past time of can, and the second of may. From the near affinity which the present of the subjunctive has to the future of the indicative, the tense now under consideration appears in many instances as the past time of the latter as well as the former. Thus, Dixi me facturum quo imperaret, may be rendered "I said that I would do whatever he might, or whatever he should, command."

Of the preter-perfect it is sufficient to observe, that as the present states the agent as at liberty to be performing an unfinished action, so this tense states him as at liberty to perform the action considered as finished. "I may be writing a letter when you come;" that is, "I am at liberty to be writing a letter when you come." "I may have written a letter when you come;" that is, "I am at liberty to be in possession of the finished action of writing a letter when you come." It is a common mode of expression to say, "I may have done such or such a thing in my time," when he who speaks can have little doubt whether he has done the thing or not. In that case, the words may have done, cannot be considered as the preter-perfect of the subjunctive of the verb do; for it is nonsense to talk of liberty with respect to the performance of an action, which at the time of speaking is supposed to be past and completed. What, then, is the import of the phrase? We are persuaded that it is elliptical, and that the word say or affirm is understood: "I may (say that I) have done such or such a thing in my time;" for liberty or contingency can relate to actions only as they are conceived to be present or future.

Of all the tenses, the most complex is the plusquamperfect of this mode. It combines a past and a future time with a finished action. It may be considered as the past time both of the perfect future and the preter-perfect of the subjunctive; for it represents an action, future and contingent at some past time, as finished before another period specified; which period, therefore, though past at the time of speaking, was itself future with respect to the time when the futurity or contingency of the action existed. "Promisiisti te scripturum fuisse si regessis;" "You promised that you would write, if I should have asked you." Here the futurity of the action of asking, which is represented as complete and finished, is stated as co-existing with the past promise; but the action itself must be posterior to that promise. It is however supposed to be past with respect to the action of writing, which is also posterior to the promise.

Before we dismiss the subject of tenses, it may not be improper just to mention number and person; for these have place in every tense of the verb in the learned languages, and in many tenses even of the English verb. They cannot, however, be deemed essential to the verb; for affirmation is the same, whether it be made by you, by me, or by a third person, or whether it be made by one man or by a thousand. The most that can be said is, that verbs in the more elegant languages are provided with a variety of terminations which respect the number and person of every substantive, that we may know with more precision, in a complex sentence, each particular substance, with its attendant verbal attributes. The same may be said of sex with respect to adjectives. They have terminations which vary as they respect beings male or female; though it is past dispute that substances alone are susceptible of sex. We therefore pass over these matters, and all of like kind, as being rather amongst the elegancies of particular languages; and therefore to be learned from the particular grammar of each tongue, than amongst the essentials of language; which essentials alone are the subject of inquiry in a treatise on universal grammar.

Besides tenses, number, and person, in every tongue with which we are acquainted, verbs are subject to another variation, which grammarians have agreed to call modes or moods. Of modes, as of tenses, it has been warmly disputed whether or not they be essential to language. The truth seems to be, that the only part of the verb absolutely necessary for the purpose of communicating thought is the indicative mode; for all the others, as has been well observed by Dr Gregory, are resolvable, by means of additional verbs and a word denoting the action of the primary verb, into circuitous expressions which fully convey their meaning. But such expressions continually repeated would make language very prolix and wholly devoid of animation; and for this reason the import of each of the commonly received modes is a subject worthy of the philologist's investigation. About the number of modes, whether necessary or only expedient, as well as about the import of each, the writers on grammar have differed in opinion. Mr Harris, one of the most celebrated of those writers, has enumerated four modes of the verb, besides the infinitive: viz. the indicative or declarative, to assert what we think certain; the potential or subjunctive, for the purposes of whatever we think contingent; the interrogative, when we are doubtful, to procure us information; and the requisitive, to assist us in the gratification of our volitions. The requisitive, too, according to him, appears under two distinct species; either as it is imperative to inferiors, or precative to superiors. For establishing such a variety of modes as this, there appears to be no sort of foundation whatever. The same reasoning which induced the author to give us an interrogative and requisitive mode, might have made him give us a hortative, a dissuasive, a volitive, and innumerable other modes, with which no language is acquainted. But besides perplexing his reader with useless distinctions, we cannot help thinking that Mr Harris has fallen into some mistakes with regard to the import of those modes which are universally acknowledged. According to him, assertion is the characteristic of the indicative, and that which distinguishes it from the subjunctive or potential; but this is certainly not true, for, without an assertion, the verb cannot be used in any mode. Of this the learned author indeed seems to have been aware, when he observed respecting the subjunctive mode, that it is employed "when we do not strictly assert," and that "it implies but a dubious and conjectural assertion." The truth is, that the assertion implied in this mode, though it is not concerning the same thing, is equally positive and absolute with that conveyed by the indicative. An example quoted by himself should have set him right as to this matter:

Sed tacitus pasci si posset corvus, haberet Plus dapis.

Who does not feel that the assertion contained in haberet is as absolute and positive as any assertion whatever?

Perhaps we may be asked to define what we mean by a mode. We know not that we can define it to the satisfaction of all. Thus much, however, seems to be obvious, that those variations which are called modes do not imply different modifications of the action of the verb. Amo, Ameam, Ama, do not signify modes of loving; for modes of loving are, loving much, loving little, loving long, loving short, and the like. Shall we then get over the difficulty by saying, with Mr Harris, that "modes exhibit some way or other the soul and its affections"? This is certainly true, but it is nothing to the purpose; for it does not distinguish the meaning of mode from the object of language in general, all languages being intended to exhibit the soul and its affections.

Grammatical modes of verbs have been defined by Dr Gregory to be "concise modes of expressing some of those combinations of thoughts which occur most frequently, and are most important and striking." This is a just observation; but perhaps he would have given a more complete definition had he said, that grammatical modes of verbs are concise modes of expressing some of those combinations of thoughts which occur most frequently, and of which assertion is an essential part. This indeed seems to be the real account of the matter, especially if our notion of the nature of verb be well founded, that its essence consists in affirmation. And in this opinion we are the more confirmed, from a conviction that no man ever employs language on any occasion but for the purpose of affirming something. The speaker may affirm something directly of the action itself, something of the agent's power or capacity to perform it, or something of his own desire that it should be performed, but still he must affirm. If this be so, then are all the modes equally indicative. Some may be indicative of perceptions, and others of volitions; but still they all contain indications. On this idea the three foregoing modes of amo will be thus distinguished: When a man indicates his present feeling of the passion of love, he uses the first; when he indicates his present capacity of feeling it, he uses the second; and when he indicates his present desire that the

---

1. The imperative, for instance, may be resolved into a verb of commanding in the first person of the present of the indicative, and a word denoting the action of the primary verb, commonly called the infinitive mode of that verb. Thus, I name et versus teum meditare camovos, and Jubco te nunc ire et tecum meditari, &c. are sentences of the very same import. The subjunctive may be resolved in the same manner by means of a verb denoting power or capacity; for credam, and possum credere, may often be used indifferently. The indicative mode, however, is not thus convertible with another verb of affirming in the first person of the present of the indicative, and a word denoting the action of the primary verb; for Titius scribit, "Titius writes," is not of the same import with dico Titius scribere, quod Titius scribat, "I say that Titius writes." The first of these sentences, as has already been shown, contains but one assertion; the second obviously contains two. Titius writes, is equivalent to Titius is writing; I say that Titius writes, is equivalent to I am saying that Titius is writing. The reason why the imperative and subjunctive are resolvable into expressions into which the indicative cannot be resolved, will be seen when the import of each of those modes is ascertained.

2. Every verb, except the simple verb amo, art, is, &c. expresses without modes a combination of thoughts, namely, affirmation and an attribute. The affirmation, however, is alone essential to the verb, for the attribute may be expressed by other words. It is indeed extremely probable, that in the earliest ages of the world, the affirmation and attribute were always expressed by different words; and that afterwards, for the sake of conciseness, one word, compounded perhaps of these two, was made to express both the affirmation and the attribute; hence arose the various classes of verbs, active, passive, and neuter. Of a process of this kind there are evident signs in the Greek and some other tongues. But the improvers of language stopped not here. The same love of conciseness induced them to modify the compound verb itself, that it might express various combinations of thought still more complex; but in all these combinations assertion was of necessity included; for if the word had ceased to assert, it would have ceased to be a verb of any kind. person to whom he is speaking would entertain that passion, he uses the third.

As to what Mr Harris calls the interrogative mode, he observes that it has a near affinity to the indicative. It has in fact not only a near affinity to it, but, as far as language is concerned, there is not between the one and the other the slightest difference. For, in written language, take away the mark of interrogation, and, in spoken language, the peculiar tone of voice, and the interrogative and indicative modes appear precisely the same. That such should be the case is extremely natural. To illustrate this, let us for once speak in the singular number, and conceive one of our readers to be present. "I assert something, taking the truth of it for granted; but if you allow me to be wrong, I presume that you will set me right." In this case, assertion produces the same effect as interrogation. Instances of the same kind perpetually occur in common conversation. An acquaintance says to me, "You took a ride this morning." I answer yes or no, according to the case; and the same effect is produced as if he had said, "Did you take a ride this morning?" In this way simple assertions would at first be employed to procure information wanted. *Fecisti*, you did such a thing; *fecisti ne*, you did it not; either would produce the proper reply, and the information wanted would be gained. This being observed as language improved, men would accompany such a sentence with a peculiar tone of voice, or other marks, to signify more unequivocally that they wanted information, or that such information was the only object of their speech. Further progress in refinement would lead them to alter the position of the words of a sentence when they meant to ask a question, as we do in English, saying, when we assert, "You have read Euripides," and when we interrogate, "Have you read Euripides?"

In Greek and Latin, questions are asked commonly enough by the particles *ut* and *an*. These particles know to be exactly equivalent to the English particle *if*, at least to the sense in which that particle is commonly used. *An fecisti*, is "If you did it;" and the sentence may either be an abbreviation for *die an fecisti*, "tell me if you did it;" or *an* may perhaps be, as *if* certainly is, an imperative mode of some obsolete verb equivalent to *give*; and in that case, *an fecisti* will be a complete interrogative sentence, signifying, "you did it, give that." But the interrogative mode of Mr Harris we have said enough; perhaps our readers will think too much, since its useless distinction not found in any language. It will, however, be proper to say something of his precative mode, as far as it is the same with the optative mode of Greek grammarians.

And nothing, we think, can be clearer, than that the Greek optative constitutes no distinct mode of the verb, whatever meaning be annexed to the word mode. The different tenses of the optative are evidently nothing but past tenses of the corresponding tenses of the subjunctive. Thus, pres. subj. *τυρνα*, I may strike; pres. opt. *τυρνα*, I might strike, &c. This is proved to be indubitably the case by the uniform practice of the Greek writers. Examples might be found without number were we to read in search of them. The following sentence will illustrate our meaning: *Εξοχαιρον Ἀθηναίων ἐκεῖνοι Ἀγρίους*, "The Athenians came that they may assist the Argives." Here the leading verb *ἐκεῖνοι* being of the present tense, the dependent verb *ἐκεῖνοι* is the present subjunctive. But change the former to the past time, and the latter must also be changed. *Ἡρώτῳ Ἀθηναίων ἐκεῖνοι Ἀγρίους ταῦτα ἀποδείξαντες*, "The Athenians came that they might assist the Argives." Here it is plain that *ἐκεῖνοι*, the present of the optative, is the past time of *ἐκεῖνοι*, the present of the subjunctive; and the same holds in other instances.

It is almost unnecessary to add, that when this mode is employed to denote a wish, the wish is not expressed by the verb, but is understood. Such abbreviated expressions to denote a wish are common in all languages. Thus, in Greek,

*Τίπος πέτε δεῖσι, Ἀθηναίων ἐκεῖνοι Ἀγρίους*

signifies, "The gods might give you (or, as we say in English, changing the position of the verb, *might the gods give you*) to destroy," &c. So in Latin, *Ut te omnes dii adeoque perdant*, "That all the gods and goddesses may destroy you." Again, in English, "O that my head were waters, and my eyes a fountain of tears." In all these, and such like sentences, words equivalent to *I wish*, *I pray*, are understood. In Greek a wish is sometimes introduced by the particle *ι* or *ιν*, *ιφ*: as in Homer,

*Εἰ μήτερ ἐγὼ ἤγειρα, ἡμῶν ἀπολαύσαι*

"If it had been your fate not to be born, or to die unmarried." The supplement is, "It would have been fortunate for your country," or some such thing. In like manner, a poor person not uncommonly entreats a favour by saying, "Sir, if you would be so good." Here he stops; but the completion of his sentence is, "it would make me happy." In all these cases a wish is not formally expressed by the speaker, but inferred by the hearer. They are therefore instances of that tendency which mankind universally discover to abbreviate their language, especially in cases where the passions or feelings are interested.

The interrogative and optative modes being set aside as superfluous, it would appear from our investigation, that the real distinct modes of the verb, which are found in the most copious and varied language, are only three; the indicative, the subjunctive, and the imperative: And these are all that can be considered as necessary; the first to indicate the speaker's feeling or acting, the second to indicate his capacity of feeling or acting, and the third to indicate his desire that the person to whom he speaks should feel or act.

Here again we have the misfortune to differ in opinion from Dr Gregory, who seems to think that a greater number of modes, if not absolutely necessary, would, at any rate, be highly useful. His words are, "All languages, I believe, are defective in respect of that variety and accuracy of combination and of distinction, which we know with infallible certainty take place in thought. Nor do I know of any particular in which language is more deficient than in the expressing of those energies or modifications of thought; some of which always are, and all of which might be, expressed by the grammatical moods of verbs. Of this there cannot be a clearer proof than the well-known fact, that we are obliged to express by the same mood very different modifications or energies of thought. As, for instance, in the case of the grammatical mood

---

Of a question put in the form of an assertion we have a remarkable instance in the gospel of St Matthew. When Christ stood before Pilate, the governor asked him, saying, *Σὺ εἶ βασιλεὺς τῶν Ἰουδαίων*, "Thou art the king of the Jews." That this sentence was pronounced with a view to obtain some answer, is evident from the context: yet it is as plainly an affirmation, though uttered probably in a scoffing tone, as the serious confession of Nathaniel, *Σὺ εἶ Σωτὴρ ὁ Ἰσραήλ*, "Thou art the king of Israel." Had not the question been put in this form, which asserts Christ to be the king of the Jews, the reply could not have been *Σὺ λαβέσαι*, "Thou restest;" for without an assertion the governor would have said nothing. (See Dr Campbell's Translation of the Gospels, where the fun used in the original is with great propriety retained in the version.) called the imperative, by which we express occasionally prayer to God, command to a slave, request to a superior, advice to an equal or to any one, order as from an officer to his subaltern, supplication to one whom we cannot resist. If these be, as the author calls them, specific differences of thought, he will not surely object to their being all ranked under one genus, which may be called desire. That the internal feelings which prompt us to pray to God, to command a slave, to request a superior, to advise an equal, to give an order to an inferior, and to supplicate one whom we cannot resist, are all different in degree, cannot be denied. Each of them, however, is desire; and the predication by which the desire is made known to the person whom we address, is the same in all, when we utter a prayer as when we utter a command, when we request as when we supplicate. But predication alone is that which constitutes the verb; for desire by itself, however modified, can be expressed only by an abstract noun; and the mere energy of desire, when not applied to a particular agent, can be expressed only by a participle, or by what is commonly, though improperly, called the infinitive mode. Now it is certainly conceivable that a few shades of meaning, or a few degrees of one general energy, might be marked by corresponding variations of such verbs as combine energy with predication; and there could be no great impropriety in calling those variations modes, or rather modes of modes; but that such a multiplication of modes would be an improvement in language, is by no means evident. The verb, with the modes and tenses which it has in all languages, is already a very complex part of speech, which few are able, and still fewer inclined, to analyse; and it would surely be of no advantage to make it more complex by the introduction of new modes, especially when those degrees of energy which could be marked by them are with equal and perhaps greater precision marked, in the living speech, by the different tones of voice adapted to them by nature; and, in written language, by the reader's general knowledge of the subject, and of the persons who may be occasionally introduced. If there be any particular delicacy of sentiment or energy which cannot thus be made known, it is better to express it by a name appropriated to itself, together with the simple and original verb of affirmation, than to clog the compound verb with such a multiplicity of variations as would render the acquisition of every language as difficult as is said to be that of the Chinese written characters. The indicative, subjunctive, and imperative, are therefore all the modes of the verb which appear to be in any degree necessary or expedient; and they are in fact all the modes that are really found in any language with which we are acquainted.

For the infinitive, as has already been observed, seems on every account to be improperly styled a mode. To that name it has no title which we can perceive, except that its termination sometimes (for even this is not universally true) differs in the learned languages from the terminations of the other parts of the verb. Nay, if affirmation be, as it has been proved to be, the very essence of verb, it will follow that the infinitive is no part of the verb at all; for it expresses no affirmation. It forms no complete sentence by itself, nor even when joined to a noun, unless it be aided by some real part of a verb either expressed or understood. Scribo, scribēbam, scripsi, scripsīram, scribam, scripsīro, "I am writing, I was writing, I have written, I had written, I shall write, I shall have written," contain each of them an affirmation, and constitute a complete sentence; but scribere "to write," scripsisse "to have written," affirm nothing, and are not more applicable to any one person than to another. In a word, the infinitive is nothing more than an abstract noun, denoting the simple

---

1 Dr Gregory seems to think, that not merely a few, but a vast number, of these energies might be so marked. "Affirming, denying, testifying, foretelling, asking, answering, wishing, hoping, expecting, believing, knowing, doubting, supposing, stipulating, being able, commanding, praying, requesting, supplicating, loving, hating, fearing, despairing, being accustomed, wondering, admiring, wavering, swearing, advising, refusing, exhorting, demanding, encouraging, promising, threatening, &c." says he, "all admit very readily of being combined with the general import of a verb." And he adds, that "if every one of them had been expressed in all languages by variations as striking as those of ovō, currō, and servō, they must have been acknowledged as distinct moods of the verb."

If all these words denote different energies of thought, which, however, may be doubted, and if all those different energies, with many others for which, as the author justly observes, it is not easy to find names, could, like capacity and desire, be combined with the general action or energy of one verb; and if those combinations could be marked by corresponding variations of that verb, we should indeed acknowledge such variations to be distinct modes, or modes of modes, of the verb. But we doubt much if all this be possible. We are certain that it would be no improvement; for it seems to be evident, either that, in some of the modes, the radical letters of the original verb must be changed, and then it would cease to be the same verb; or that many of the modes must be expressed by words of very unmanageable length; not to mention, that the additional complication introduced by so many minute distinctions into a part of speech already exceedingly complex, would render the import of the verb absolutely unintelligible to nine-tenths even of those who are justly styled the learned.

2 In our idea of the infinitive, we have the good fortune to agree with the learned Ruddiman, whose words are, "Non incepte hic modus a veteribus quidamque verbis novae est appellatione. Est enim, si non vere ac sumper, quod nonnulli volunt, neomen substantivum, significatione certe et maxime affinitatis; quaeque vis sustinet per omnes casus. Et quidem manifeste videtur, cum adiectivum ei additur neutri generi; ut, Cic. Att. xii. 28, Cum eivere ipsius turpis sit multa. Pers. v. 53, Felle suum culque est. Cic. lib. i. 1, Totum hoc diaphanum philosophicum. Petron. lib. xvi. Meam intelligo, quaestio nulla possit venire. Item, absoque adjectivo, ut Ovid. Met. ii. 483, Poetae quoque critici, ut est, non solum loquendi, Plaut. Basi. ii. 2. 59, Id vereri perdiderit, id est, secundissimum. Cic. Tusc. v. 28, Longe de doctis hominibus et prudentibus, sed vivere et cogitare, id est, expedita est cognitio." See Grammatica Latina Institutiones, pars secunda, lib. i. cap. 2, where the reader will find examples of the infinitive used by the best Roman writers as a substantive noun in every case.

This opinion of Ruddiman and the ancient grammarians has been controverted with much ingenuity by Dr Gregory, who seems to think that in the infinitive alone we should look for the essence of the verb divested of every accidental circumstance, time only excepted. If this be indeed the case, almost everything which we have said of the verb, its tenses, and its modes, is erroneous; and he who takes his principles of grammar from the Encyclopedia Britannica will fill his head with a farrago of absurdities. But it may without much difficulty be shown, that the error does not attach to the doctrine respecting the infinitive, which has been stated and illustrated in this article, but to the strange inversion of all recognised principle involved in the notion which Dr Gregory has espoused, apparently by reason of its novelty alone.

The Doctor acknowledges (Transactions of the Royal Society of Edinburgh, vol. ii. liter. class, p. 195), that the infinitive is most improperly called a mode, and on that account he thinks we ought to turn our thoughts exclusively to it, "when we endeavour to investigate the general import of the verb, with a view to ascertain the accident which it denotes, and be led, step by step, to form a distinct notion of what is common in the accidents of all verbs, and what is peculiar in the accidents of the several classes of them, and thereby be enabled to give good definitions, specifying the essence of the verb," &c. It may be true, that to the infinitive exclusively we should turn our attention when we wish to ascertain the accident denoted by a particular verb or class of verbs; that is, the kind of action, passion, or state of being, of which, superadded to affirmation, that verb or class of verbs is expressive; but in accidents of this kind it may be doubted if there be any thing that with propriety can be said to be common to all verbs. There seems indeed to be nothing common to all verbs but that which is essential to them, and by which they are distinguished from every other part of speech; but every kind of action, passion, and state of being, may be completely expressed by participles and abstract nouns, and therefore such accidents we cannot find the essence of the verb, because such accidents distinguish it not from other parts of speech. Were man called upon to specify the essence of verse or metre, he would not say that it consists in the meaning of the words, or in the sing of these words according to the rules of syntax. In every kind of verse where words are used, they have indeed a meaning, and all good verses they are grammatically constructed; but this is likewise the case in prose, and therefore it cannot be the essence of verse. The essence of verse must consist in something which is not to be found in prose, namely, a certain harmonic succession of sounds and number of syllables; and the essence of the verb must likewise consist in something which is not to be found in any other art of speech, and that is nothing but affirmation. But if affirmation be the very essence of the verb, it would surely be improper, when we endeavour to ascertain the general import of that part of speech, to turn our thoughts exclusively to a word which implies an affirmation; for what does not affirm, cannot in strictness of truth be either a verb or the mode of a verb.

In the same way, it is said that "the infinitive denotes that kind of thought or combination of thoughts which is common to all the other modes." In that sense this is true we are unable to conceive; it denotes indeed the same accident, but certainly not the same thought or combination of thoughts. In the examples quoted, *Nos* *esse* *vivere*, *sed* *valere* *vidi*, the infinitives have evidently the force of abstract nouns, and not of verbs; for whereas there express the same states of being with *esse* and *valere*, they by no means express the same combination of thoughts. *Vivo* also affirms that *I am living*, and that *I am well*; and he who utters these words must think not of life and health in the abstract, but of life and health as belonging to himself. *Vivere* and *valere*, on the other hand, affirm nothing; and he who utters them thinks only of the states of living and of being in health, without applying them to any particular person.

The learned author of *The Origin and Progress of Language* having said that the infinitive is used either as a noun, or that it serves to connect the verb with another verb or noun, and so is useful in syntax, the doctor combats this opinion, and infers that the infinitive is truly a verb, because "the thought expressed by means of it may be expressed in synonymous and convertible phrases, in different languages, by means of other parts or moods of the verb." Of these synonymous and convertible phrases he gives several examples, the first of which is taken from Hamlet's soliloquy, "To be or not to be, that is the question," he thinks equivalent in meaning to, "The question is, whether we shall be or shall not be." But we are persuaded he is mistaken. Whether we shall be or shall not be involves a question, whether we shall exist at some future and indefinite time; but the subject of Hamlet's debate with himself was not, whether, if his conscious existence should be interrupted, it would be afterwards at some future and indefinite time restored, but whether it was to continue uninterrupted by his exit from this world. This, we think, must be self-evident to every reader of the soliloquy. It is likewise obvious that the word "question" in this sentence does not signify interrogatory, but object of debate, or matter to be examined; and that the word "that" serves no other purpose than to complete the verse, and give additional emphasis, perhaps, to an inquiry so important. "To be or not to be, that is the question," is therefore equivalent in all respects to "The continuance or non-continuance of my existence is the matter to be examined;" and the infinitive is here indisputably used as an abstract noun in the nominative case. Should it be said that the doctor may have taken the sentence by itself, unconnected with the subject of Hamlet's soliloquy, we reply, that the supposition is impossible; for, independently of the circumstances with which they are connected, the words "To be or not to be" have no perfect meaning. Were it not for the subject of the soliloquy, in which every reader supplies what is wanting to complete the sense, it might be asked, "To be or not to be"—what? A coward, a rinderer, a king, or a corpse? Questions all equally reasonable, and which in that case could not be answered.

With the same view, to prove that the infinitive is truly a verb, the doctor proceeds to remark upon the following phrases, *Dico*, *credo*, *puto*, *Titius existere*, *valere*, *jaceret*, *cederet*, *procureret*, *quaestio Maxima*, *projectione fuisse a Maxio*; which, he says, have the very same meaning with *dico*, *quod Titius existat*, *quod jaceat*, *quod cederit*, &c. He adds, that "the infinitives, as thus used, acquire no further meaning, in addition to the radical import of the verb with tense, like the proper moods; but the subjunctives after and lose their peculiar meaning as moods, and signify no more than bare infinitives." In the sense in which this observation is made by the author, the very reverse of it seems to be the truth. The infinitives, as thus used, acquire, at least in the mind of the reader, something like the power of affirmation, which they certainly have not when standing by themselves; whereas the subjunctives neither lose nor acquire any meaning by being placed after *quo*. *Dico*, *credo*, *puto*, *Titius existere*, *valere*, *jaceret*, &c., when translated literally, signify, I say, believe, think, Titius to exist, to be well, to lie along; a mode of speaking which, though now not elegant, is common with the best writers in the days of Shakespeare, and is frequently to be found in the writings of Warburton at a much later period. *Dico*, *credo*, *puto*, *Titius existere*, *quod jaceat*, &c., signifies literally, I say, believe, think, that Titius may exist, may lie along, &c. Remove the verbs, and the indicative sense of the former set of phrases, and it will be found that the infinitives had acquired meaning, when combined with them, which they have not when left by themselves, as *Titius existere*, *valere*, *Titius to exist*, to lie along, have no complete meaning, because they affirm nothing. On the other hand, when the indicative verbs are removed, together with the wonder-working *quo* from the latter set of phrases, the meaning of the subjunctives remains in all respects as it was before the removal; for *Titius existat*, *jacet*, &c., signify, Titius exist, may lie along, as well when they stand by themselves as when they form the final clauses of a compound sentence. Every one knows that *quo*, though often called a conjunction, is always fact the relative pronoun. *Dico*, *credo*, *puto*, *quod Titius existat*, must therefore be construed thus: *Titius existat* (*est id*) *quod dico*, &c., &c. "Titius may exist—is that thing, that proposition, which I say, believe, think." In the former set of phrases, the infinitives are used as abstract nouns in the accusative case, denoting, in conjunction with *Titius*, one complex conception, the existence, &c., of *Titius*: *Dico*, *credo*, *pulo*, "I say, believe, think;" and the object of my speech, belief, thought, is *Titius existere*, "the existence of Titius."

In confirmation of the same idea, that the infinitive is truly a verb, Dr Gregory quotes from Horace a passage, which, had we thought quotations necessary, we should have urged in support of our own opinion:

*Nec quicquam tibi prodet*

*Aurias tentasse domos, animoque retundum*

*Percurrire polum, moritura.*

Our apprehension, nothing can be clearer than that *tentasse* and *percurrire* are here used as nouns; for if they be not, where will we find a nominative to the verb *prodet*? It was certainly what was signified by *tentasse aerias domos, animoque retundum percurrire polum*, that is said to have been no advantage to Archytas at his death. This, indeed, if there could be any doubt about it, would be made evident by the two prose versions which the professor subjoins to these beautiful lines. The first of them runs as follows: *Nec quicquam tibi prodet quod aerias domos tentaveris, et animoque percurreris polum*; which must be thus constructed: *Aerias domos tentaveris, et percurreris animoque polum* (*est id* *quod nec quicquam tibi prodet*). This version, however, is not perfectly accurate; for it contains two propositions, whereas Horace's lines contain but one. The second, which, though it may be a cramped, and inelegant sentence, expresses the poet's sense with more precision, is in these words: *Nec quicquam tibi prodet moritura tua tentatio domum aeriarum, et cursum tuus circa polum*. Having observed, with truth, that this sentence has the very same meaning with the lines of Horace, Dr Gregory asks, "why are not *tentatio* and *cursum* reckoned verbs as well as *tentasse* and *percurrire*?" Let those answer this question who believe that any of these words are truly verbs; for they are surely, as he adds, all very near akin; fixed so near, that the mind, when contemplating the import of each, cannot perceive the difference. Meanwhile, we beg leave in founding commands with the execution of commands. But the learned writer proceeds to informs us, that "it is from the connection of futurity with commands that the future of the indicative is sometimes used for the imperative mode." The connection of which he speaks appears to us entirely imaginary; for futurity has nothing to do with commands, though it may with the execution of them. The present time is the time of commanding, the future of obeying. But supposing the connection real, it would not account for the future tenses being used imperatively. For although it were true, as it is evidently false, that commands are future, it would not follow that the relation is convertible, or that employing the future should imply a command. The principle upon which such expressions as, *Thou shalt not kill,* come to have the force of a command, seems to be this: When a person, especially one possessed of authority, asserts that an action depending on the will of a free agent, and therefore in its own nature contingent, shall or shall not actually take place; what are we to conclude from such an assertion? Why, surely it is natural to conclude that it is his will, his command, that his assertion be verified. The English word *shall,* if we be well informed, denoted originally obligation; a sense in which its past tense *should* is still commonly employed. In English, therefore, the foregoing process of inferring a command from an assertion of futurity seems to have been reversed; and the word *shall,* from denoting a command or obligation, has come to denote futurity simply.

Having considered the verb in its essence, its tenses, and its modes, we might seem to have exhausted the subject; but there is still something more to be done. Grammarians have distinguished verbs into several species; and it remains with us to inquire upon what principle in nature this distinction is made, and how far it proceeds. Now it must be obvious, that if predication be the essence of verbs, all verbs as such must be of the same species; for predication is the same in every proposition, under every possible circumstance, and by whomsoever it is made. But the greater part of verbs contain the predicate as well as the predication of a proposition; or, to speak in common language, they denote an attribute as well as an affirmation. Thus, *lego* is "I am reading;" *ambulo*, "I am walking;" *sto*, "I am standing;" *verbero*, "I am striking;" *verberor*, "I am stricken." But the attributes expressed by these verbs are evidently of different kinds; some consisting in action, some in suffering, and some in a state of being which is neither active nor passive. Hence the distinction of verbs, according to the attributes which they denote, into active, passive, and neuter. *Lego*, which is an assertion that I am employed in the act of reading, is an active verb; *verberor*, which is an assertion that I am suffering under the rod, is a passive verb, because it denotes a passion; and *sto*, which is an assertion that I am standing still, is said to be a neuter verb, because it denotes neither action nor passion. But it is self-evident that there cannot be action without an agent, nor passion without a passive being; neither can we make a predication of any kind, though it denote neither action nor passion, without predicating of something. All verbs, therefore, whether active, passive, or neuter, have a necessary reference to some noun expressive of the substance of which the attribute denoted by the verb is predicated. This noun, which in all languages must be in the nominative case, is said to be the nominative of the verb; and in those languages in which the verb has person and number, it must in these respects agree with its nominative.

Of action, and consequently of verbs denoting action, there are obviously two kinds. There is an action which passes from the agent to some subject upon which he is employed, and there is an action which respects no object beyond the agent himself. Thus *lego* and *ambulo* are verbs which equally denote action; but the action of *lego* refers to some external object as well as to the agent; for when a man is reading, he must be reading something, a book, a newspaper, or a letter, whereas the action of *ambulo* is confined wholly to the agent; for when a man is walking he is employed upon nothing beyond himself, his action produces no effect upon any thing external. These two species of verbs have been denominated transitive and intransitive; a designation extremely proper, as the distinction which gave rise to it is philosophically just. Verbs of both species are active; but the action of those only which are called transitive respects an external object; and therefore in those languages of which the nouns have cases, it is only after verbs which are transitive as well as active, that the noun denoting the subject of the action is put in the accusative or objective case. Verbs which are intransitive, though they be really active, are in the structure of sentences considered as neuter, and govern no case.

So much, then, for that most important of all words, the verb. We proceed now to the consideration of participles, adjectives, and adverbs, which, as they have a new relation to one another, we shall treat of under one head or division.

V.—Of Participles, Adjectives, and Adverbs.

1. Of Participles.

The nature of verbs being understood, that of participles is not of difficult comprehension. Every verb, except that which is called the substantive verb, is expressive of an attribute, of time, and of an assertion. Now if we take away the assertion, and thus destroy the verb, there will remain the attribute and the time; and these combined make the essence of that species of words called participles. Thus, take away the assertion from the verb γραφω, *write*, and there remains the participle γραφον, *writing*, which, without the assertion, denotes the same attribute and the same time. In the same manner, by withdrawing the assertion, we discover γραφας, written, in γραψας, wrote; γραφαν, about to write, γραψαν, shall be writing.* This is the doctrine of Mr Harris respecting participles; and, in our opinion, it is equally elegant, perspicuous, and just. But as it has been controverted by an author of some note in the republic of letters, we should perhaps be wanting in duty to our readers were we to pass his objections wholly unnoticed.

It is acknowledged by Dr Beattie, the writer to whom we allude, that the view which we have taken is the most convenient in which the participle can be considered in universal grammar; but he affirms that present participles do not always express present time, nor preterite participles past time; nay, that participles have often no connection with time at all. He thus exemplifies his assertion in Greek, in Latin, and in English:

"When Cebes says, Ἐπεριστρέφομαι εἰς τὸν Χρόνον ἤδη, 'We were walking in the temple of Saturn,' the participle of the present *walking* is, by means of the verb

our turn to ask, Why are not *tentatio* and *perenturiae* reckoned abstract nouns as well as *tentatio* and *cursum*? To this question it is not easy to conceive what answer can be returned upon the doctor's principles. In his theory there is nothing satisfactory; and what has not been done by himself, we need scarcely expect from any of his followers. On the other hand, our principles furnish a very obvious reason for excluding *tentatio* and *cursum* from the class of verbs; it is, because these words express no predication. *Tentatio* and *perenturiae* indeed denote predication no more than *tentatio* and *cursum*; and therefore upon the same principle we exclude them likewise from a class to which, if words are to be arranged according to their import, they certainly do not belong. shall, at some time posterior to that of speaking, have been loved; the expression shall have been loved denotes two times, both future with respect to the time of speaking; but when the time denoted by shall have comes to be present, that of the participle loved must be past, for it is declared that the action of it shall then be complete and finished.

We conclude, then, that it is essential to a participle to express both an attribute and time; and that such words as denote no time, though they may be in the form of participles, as docens, "learned," eloquens, "eloquent," &c., belong to another part of speech, which we now proceed to consider.

2. Of Adjectives.

The nature of verbs and participles being understood, that of adjectives becomes easy. A verb implies, as we have said, an attribute, time, and an assertion; a participle implies only an attribute and time; and an adjective implies only an attribute as belonging to some substance. In other words, an adjective has no assertion, and it denotes only such an attribute as has not its essence either in motion or its privation. Thus, in general, the attributes of quantity, quality, and relation, such as many, few, great, little, black, white, good, bad, double, treble, and the like, are all denoted by adjectives.

To understand the import and the use of this species of words, it must be observed that every adjective is resolvable into a substantive and an expression of connection equivalent to of. Thus, a good man is a man of goodness; where we see the attribute denoted by the adjective fully expressed by an abstract noun. But it is evident that the noun goodness does not express the whole meaning of the adjective good; for every adjective expresses not only an attribute, but also the connection between the attribute and its substance; whereas in the abstract noun, the attribute is considered as a substance unconnected with any other substance.

In the next place, it is to be observed, that the connection expressed by adjectives, like that expressed by of, is of a nature so general and indefinite, that the particular kind of connection must, in some languages, be inferred from our previous knowledge of the objects between which it subsists, or it will for ever remain unknown. This might be proved by a variety of examples, but will perhaps be sufficiently evident from the following: Color salubris signifies colour that indicates health; exercitatio salubris, exercise that preserves health; victus salubris, food that improves health; medicina salubris, medicine that restores health. In all these examples the connection expressed by the adjective form of salubris is different; and though it may be known from previous experience, there is nothing in any of the expressions themselves by which it can be ascertained. Thus, adjectives are each significant of an attribute and connection; but the particular kind of connection is ascertained by experience. The usual effect of adjectives in language, is to modify or particularize a general term, by adding some quality or circumstance which may distinguish the object meant by that term, from the other objects of the same species. I have occasion, for example, to speak of a particular man, of whose name I am ignorant. The word man is too general for my purpose, it being applicable to every individual of the human species. In what way then do I proceed, in order to particularize it, so as to make it denote that very man whom I mean to specify? I annex or conjoin to it such words as are significant of objects and qualities with which he is connected, and which are not equally applicable to others from whom I mean to distinguish him. Thus I can say, a man of prudence or a prudent man, a wise man, a good man, a Adjectives. brave man, and so on. By means of these additions the general term man is limited or modified, and can be applied only to certain men to whom belong the attributes expressed by the adjectives prudent, wise, good, and brave. If it be still too general for my purpose, I can add to it other qualities and circumstances, till I make it so particular as to be applicable to but one individual man in the universe.

This is the way in which adjectives are commonly used, but this is not the only way. Instead of being employed to modify a substantive, they sometimes appear as the principal words in the sentence, when the sole use of the substantive seems to be to modify the abstract noun, contained under the adjective to which that substantive is joined. In order to understand this, it will be necessary to attend to the following observations.

It may be laid down as a general proposition, that when any term or any phrase is employed to denote a complex conception, the mind has a power of considering, in what order it pleases, the simple ideas of which the complex conception is composed. To illustrate this observation by an example: The word eques in Latin denotes a complex conception, of which the constituent simple ideas are those of a man and a horse; with this connection subsisting between them, that the man is conceived as on the back of the horse. In the use of this word, it is well known that the idea first in order, as being the principal subject of the proposition, is commonly the man on the back of the horse; but it is not always so, for the mind may consider the horse as the principal object. Thus, when Virgil says,

Fraena Pelethrosii Lapitae gyronque dedere, Impetiti doros; atque equitem docuere sub armis Insultare solo, et gressus glomerare superbos—

the energies attributed to the object signified by equitem make it evident that the horse, and not the man, is meant; for it is not the property of a man, insultare solo, et gressus glomerare superbos.

The same observation holds true where the complex object is denoted by two or more words; an adjective, for instance, and a substantive. Thus in the phrase summus mons se inter nubila condit, the words summus mons represent a complex conception, of which the constituent ideas are those of height and mountain, connected together by the adjective form of summus. Either of these ideas may be the subject of the proposition; and the expression will accordingly admit of two different significations. If mons be made the subject of the proposition, the meaning will be, "the highest mountain hides itself amongst the clouds." If the substantive included in the radical part of summus be made the subject of the proposition, the expression will signify, "the summit, or highest part, of the mountain, hides itself amongst the clouds." The latter is the true import of the sentence.

From these observations and examples, we shall be enabled to understand the two uses of the adjective. It is either employed, as has been already observed, to restrict or modify a general term; or the abstract substantive contained in the adjective is modified by the noun, with which, in the concrete or adjective form, that abstract substantive is joined. The first may be called the direct, the second the inverse, acceptation of adjectives.

The inverse acceptation of adjectives and participles, for both are used in the same manner, has scarcely been noticed by any grammarian excepting Dr Hunter of St Andrews; yet the principle is of great extent in language. In order to explain it, we shall produce a few examples, which on any other principle it is impossible to understand. Livy, speaking of the abolition of the regal authority at Rome, says, Regnatum est Roma ab urbe condita ad liberatam annos ducentos quadriginta quatuor, "Monarchy subsisted at Rome (not from the city built, which would convey no meaning, but) from the building of the city to its deliverance." Both the participles condita and liberata are here used inversely; that is, the abstract substantives contained in condita and liberatam are modified or restricted by the substantives urbe and urbem, with which they unite. Again, Ovid, speaking of the contest between Ajax and Ulysses for the arms of Achilles, has these lines:

Qui, licet eloquio fidum quoque Nestora vincat, Haud tamen efficiet, desertum ut Nestora crimen Nullum esse rear.

Here also the adjective or participle desertum is taken inversely, and the general notion of desertion contained in it is modified or rendered particular by being joined with the substantive Nestora. The meaning of the passage is, "I will never be induced to believe that the desertion of Nestor was not a crime." Were desertum to be taken directly as an adjective modifying its substantive, the sentence must be translated, "I cannot believe that Nestor deserted was not a crime." But it is evident that this is nonsense; as Nestor, whether deserted or not deserted, could not be a crime.

It were easy to produce many more examples of adjectives taken inversely; but these may suffice to illustrate the general principle, and to show, that without attending to it, it is impossible to understand the ancient authors. We shall produce one instance of it from Shakespeare, to evince that it is not confined to the ancient languages, though in these it is certainly more frequent than in the modern:

Freeze, freeze, thou bitter sky, Thou canst not bite so nigh As benefits forget; Though thou the waters warp, Thy sting is not so sharp As friends remember'd not.

Here it is evident that the adjective forget is taken inversely; for it is not a benefit, but the forgetting of a benefit, which bites more than the bitter sky; and therefore, in this passage, the adjective serves not to modify the noun, but the noun benefits is employed to modify the abstract substantive contained in the adjective forget, which is the subject of the proposition, and the principal word in the sentence.

Had Mr Harris attended to this principle, and reflected upon what he could not but know, that all adjectives denote substances; not indeed subsisting by themselves, as those expressed by nouns, but concretely, as the attributes of other substances; he would not have classified adjectives with verbs, or have passed so severe a censure upon the grammarians for classing them with nouns. It matters very little how adjectives are classified, provided their nature and effect be understood; but they have at least as good a title to be ranked with nouns as with verbs, and in our opinion a better. To adopt Mr Harris's language, they are homogeneous with respect to nouns, as both denote substances; they are heterogeneous with respect to verbs, as they never do denote assertion.

Besides original adjectives, there is another class, which is formed from substantives. Thus, when we say, the party of Pompey, the style of Cicero, the philosophy of Socrates; in these cases the party, the style, and the philosophy spoken of, receive a stamp and character from the persons whom they respect; and those persons, therefore, perform the part of attributes. Hence they actually pass into attributives, and assume as such the form of adjectives. It is thus we say, the Pompeian party, the Ciceronian style, and the Socratic philosophy. In like manner, for a trumpet of brass, we say a brazen trumpet, and for a crown of gold, a golden crown, &c. Even pronominal substantives admit a similar mutation. Thus, instead of saying the book of me, and of thee, we say my book, and thy book; and instead of saying the country of us and of you, we say our country and your country. These words, my, thy, our, your, &c. have therefore been properly called nominonal adjectives.

It has been already observed, and must be obvious to all, that substances alone are susceptible of sex; and that therefore substantive nouns alone should have distinctions respecting gender. The same is true with respect to number and person. An attribute admits of no change in its nature, whether it belong to you or to me, to a man or a woman, to one man or to many; and therefore the words expressive of attributes ought on all occasions, and in every situation, to be fixed and invariable. For as the qualities good and bad, black and white, are the same, whether they be applied to a man or a woman, to many or to few; so the word which expresses any one of these attributes ought in strictness to admit of no alteration, with whatever substantive it may be joined. Such is the order of nature, and that order, on this as on other occasions, the English language most strictly observes; for we say equally, a good man or a good woman; good men or good women; a good house or good houses. In some languages, indeed, such as Greek and Latin, of which the nouns admit of cases, and the sentences of an inverted structure, it has been found necessary to endow adjectives with the threefold distinction of gender, number, and person; but as this is only an accidental variation, occasioned by particular circumstances, and not in the least essential to language, it belongs not to our subject, but to the particular grammars of these tongues.

There is, however, one variation of the adjective, which is place in all languages, is founded in the nature of things, and properly belongs to universal grammar. It is occasioned by comparing the attribute of one substance with a similar attribute of another, and falls naturally to be explained under the next section.

3. Of Adverbs, and the Comparison of Adjectives.

As adjectives denote the attributes of substances, so there is an inferior class of words which denote the modifications of these attributes. Thus, when we say, Cicero and Pliny were both of them eloquent; Statius and Virgil both of them wrote; the attributes expressed by the words eloquent and wrote are immediately referred to Cicero, Virgil, &c.; and as denoting the attributes of substances, these words, the one an adjective, and the other a verb, have been both called attributives of the first order. But when we say, "Pliny was moderately eloquent, but Cicero exceedingly eloquent; Statius wrote indifferently, but Virgil wrote admirably;" the words moderately, exceedingly, indifferently, and admirably, are not referrible to substantives, but to other attributes; that is, the words eloquent and wrote, the signification of which they modify. Such words, therefore, having the same effect upon adjectives that adjectives have upon substances, have been called attributives of the second order. By grammarians they have been called adverbs; and if we use the word verb in its most comprehensive signification, including not only verbs properly so called, but also every species of words which, whether essentially or accidentally, are significant of the attributes of substances, we shall find the name adverb to be a very just appellation, as denoting a part of speech, the natural appendage of such verbs. So great is this dependence in grammatical syntax, that an adverb can no more subsist without its verb, that is, without some word significant of an attribute, than a verb or adjective can subsist without its substantive. It is the same here as in certain natural subjects. Every colour, for its existence, as much requires a superficies, as the superficies for its existence requires a solid body.

Amongst the attributes of substance are reckoned quantity and quality; thus we say a white garment, a high mountain, and the like. Now some of these quantities and qualities are capable of intension or remission; or, in other words, one substance may have them in a greater or less degree than another. Thus we say, a garment exceedingly white, a mountain tolerably or moderately high. Hence, then, one copious source of secondary attributes or adverbs to denote these two, that is, intension and remission; such as greatly, tolerably, vastly, extremely, indifferently.

But where there are different intensions of the same attribute, they may be compared together. Thus, if the garment A be exceedingly white, and the garment B be moderately white, we may say, the garment A is more white than the garment B. This paper is white, and snow is white; but snow is more white than this paper. In these instances the adverb more not only denotes intension, but relative intension; nay, we stop not here, as we not only denote intension merely relative, but relative intension than which there is none greater. Thus we say, Sophocles was wise, Socrates was more wise than he, but Solomon was the most wise of men. Even verbs, properly so called, which denote an attribute as well as an assertion, must admit both of simple and of comparative intensions; but the simple verb to be admits of neither the one nor the other. Thus, in the following example, Fame he loveth more than riches, but virtue of all things he loveth most; the words more and most denote the different comparative intensions of the attribute included under the verb loveth; but the assertion itself, which is the essential part of the verb, admits neither of intension nor remission, and is the same in all possible propositions.

From this circumstance of quantities and qualities being capable of intension and remission, arise the comparison of adjectives, and its different degrees, which cannot well be more than the two species above mentioned; one to denote simple excess, and one to denote superlative. Were we indeed to introduce more degrees than these, we ought perhaps to introduce an infinite number, which is absurd. For why stop at a limited number, when in all subjects susceptible of intension, the intermediate excesses are in a manner infinite? Between the first simple white and the superlative whitest there are infinite degrees of more white; and the same may be said of more great, more strong, more minute, and so on. The doctrine of grammarians about three such degrees of comparison, which they call the positive, the comparative, and the superlative, must be absurd; both because in their positive there is no comparison at all, and because their superlative is a comparative as much as their comparative itself. Examples to evince this may be met with everywhere. Socrates was the most wise of all the Athenians; Homer was the most sublime of all poets. In this sentence Socrates is evidently compared with the Athenians, and Homer with all other poets. Again, if it be said that Socrates was more wise than any other Athenian, but that Solomon was the most wise of men, is not a comparison of Solomon with mankind in general, as plainly implied in the last clause of the sentence, as a comparison of Socrates with the other Athenians in the first?

---

1 Aristotle and his followers called every word a verb which denotes the predicate of a proposition. This classification was certainly absurd; for it confounds not only adjectives and participles, but even substantives, with verbs; but the authority of Aristotle is great; and hence the name of adverb, though that word attaches itself only to an adjective or participle, or a verb significant of an attribute, and does not attach itself to the pure verb. But if both imply comparison, it may be asked, in what consists the difference between the comparative and superlative? Does the superlative always express a greater excess than the comparative? It does not; for though Socrates was the most wise of the Athenians, yet is Solomon affirmed to have been more wise than he; so that here a higher superiority is denoted by the comparative more than by the superlative most. Is this then the difference between these two degrees, that the superlative implies a comparison of one with many, whilst the comparative implies only a comparison of one with one? No, this is not always the case neither. The Psalmist says, that "he is wiser (or more wise) than all his teachers;" where, though the comparative is used, there is a comparison of one with many.

The real difference between these two degrees of comparison may be explained thus:—When we use the superlative, it is in consequence of having compared individuals with the species to which they belong, or one or more species with the genus under which they are comprehended. Thus Socrates was the most wise of the Athenians, and the Athenians were the most enlightened of ancient nations. In the first clause of this sentence, Socrates, although compared with the Athenians, is at the same time considered as one of them; and in the last the Athenians, although compared with ancient nations, are yet considered as one of those nations. Hence it is that in English the superlative is followed by the preposition of, and in Greek and Latin by the genitive case of the plural number; to show that the object which has the pre-eminence is considered as belonging to that class of things with which it is compared. But when we use the comparative degree, the objects compared are set in direct opposition; and the one is considered not as a part of the other, or as comprehended under it, but as something altogether distinct and belonging to a different class. Thus, were any one to say, "Cicero was more eloquent than the Romans," he would speak absurdly; because every body knows, that of the class of men expressed by the word Romans, Cicero was one; and such a sentence would affirm that orator to have been more eloquent than himself. But when it is said that "Cicero was more eloquent than all the other Romans, or than any other Roman," the language is proper, and the affirmation true; for though the persons spoken of were all of the same class or city, yet Cicero is here set in contradistinction to the rest of his countrymen, and is not considered as one of the persons with whom he is compared. It is for this reason that in English the comparative degree is followed by a noun governed by the word of contradiction than, and in Latin by a noun in the ablative case governed by the preposition praet.1 either expressed or understood. We have already observed, that the ablative case denotes concomitancy; and therefore when an adjective in the comparative degree is prefixed to a noun, that noun is put in the ablative case, to denote that two things are compared together in company; but by means of the preposition, expressed or understood, that which is denoted by the comparative adjective is seen to be preferred before that which is denoted by the noun.

We have hitherto considered the comparative as expressed by the words more and most; but the authors or improvers of language have contrived a method of retrenching the use of these adverbs, by expressing their force by an inflexion of the adjective. Thus, instead of more fair, they say fairer; instead of most fair, fairest; and the same method of comparison takes place both in the Greek and Latin languages; with this difference, however, between the genius of these languages and ours, that we are at liberty to form the comparison either in the one method or in the other; whereas in those languages the comparison is seldom or never formed by the assistance of the adverb, but always by the inflexion of the adjective. Hence this inflexion is by the Greek and Latin grammarians considered as a necessary accident of the adjective; but it has reached no further than to adjectives, and participles sharing the nature of adjectives. The attributes expressed by verbs are as susceptible of comparison as those expressed by adjectives; but they are always compared by means of adverbs, the verb being too much diversified already to admit of more variations without perplexity.

It must be confessed, that comparatives, as well the simple as the superlative, seem sometimes to part with their relative nature, and to retain only their intensive. Thus in the degree denoting simple excess,

Tristior, et lacrymis oculos suffusa nitentest.—Virg.

Tristior means nothing more than that Venus was very sad. In the degree called the superlative this is more sad. Phrases extremely common are, Vir doctissimus, vir fortissimus, "a most learned man, a most brave man;" that is, not the bravest and most learned man that ever existed, but a man possessing those qualities in an eminent degree. In English, when we intimate that a certain quality is possessed in an eminent degree, without making any direct comparison between it and a similar quality, we do it by the intensive word very, more commonly than by most; as, Cicero was very eloquent, the mind of Johnson was very vigorous. This mode of expression has been called the superlative of eminence, to distinguish it from the other superlative, which is superlative upon comparison. Yet it may be said, that even in the superlative of eminence something of comparison must be remotely or indirectly intimated; as we cannot reasonably call a man very eloquent, without comparing his eloquence with the eloquence of other men. This is indeed true; but we cannot therefore affirm that comparison is more clearly intimated in this superlative than in the simple adjective eloquent; for when we say that a man is eloquent, we mark between his eloquence and that of other men a distinction of the same kind, though not in the same degree, as when we say that he is very eloquent.

In English we distinguish the two superlatives, by prefixing to the one the definite article the, to show that something is predicated of the object expressed by it, which cannot be predicated of any other object; and by subjoining the preposition of, to show that the objects with which it is compared are of the same class with itself; as, "Solomon was the wisest of men; Hector was the most valiant of the Trojans." To the other superlative2 we only prefix

---

1 See Ruddiman, Grammatica Institutiones, pars secunda, lib. i. cap. 2. Although it is certainly true, that when we use the superlative, we ought in propriety to consider the things compared as of the same class; and when we use the comparative, as of different classes; yet this distinction is not always attended to by the best writers in any language. In Latin and Greek the comparative is sometimes used, where in English we should use the superlative; as dextra est fortior manus, and in the Gospel it is said, that "a grain of mustard-seed is the smallest seed of all seeds, but when grown up it is the greatest (μεγαλύτερος) of herbs." Even in English, the custom of the language permits us not to say, "he is the tallest of the two," it must be the taller of the two; but we cannot say "he is the taller of the three;" it must be the tallest. For these and other deviations from the general rule no reason is to be found in the nature of things; they are errors made purely by use.

2 In English, the termination est is peculiar to the superlative of comparison, to which the definite article is prefixed. Thus we may say, "Homer was the sublimest of poets;" but we cannot say, "Homer was a sublimest poet." Again, we may say, "Homer was a very sublime poet;" but not, "Homer was the very sublime poet." As there are some qualities which admit of comparison, so there are others which admit of none; such, for example, as those which denote that quality of bodies arising from their figure, as when we say, a circular table, a quadrangular boxcourt, a conical piece of metal, and the like. The reason is, that a million of things participating the same figure, participate it equally, if they do it at all. To say, therefore, that whilst A and B are both quadrangular, A is more or less quadrangular than B, is absurd. The same holds true in all attributives denoting definite quantities of whatever nature; for as there can be no comparison without intension or remission, and as there can be no intension or remission in things always definite, therefore these attributes can admit of no comparison. By the same method of reasoning, we discover the cause why no substantive is susceptible of these degrees of comparison. A mountain cannot be said more to be or to exist than a mole-hill; but the more or less must be sought for in their quantities. In like manner, when we refer many individuals to one species, the lion A cannot be called more a lion than the lion B; but if any thing, he is more fierce, more swift, or exceeding in some such attribute. So, again, in referring many species to one genus, a crocodile is not more an animal than a lizard; nor a tiger more than a cat; but, if anything, the crocodile and tiger are more bulky, more strong, &c., than the animals with which they are compared; the excess, as before, being derived from their attributes.

The adverbs or secondary attributives already mentioned, those denoting intension and remission may be called adverbs of quantity continuous, as greatly, vastly, tolerably, &c.; once, twice, thrice, &c.; adverbs of quantity discrete, as ore and most, less and least, to which may be added equally, proportionally, &c., are adverbs of relation. There are others of quality; as when we say, honestly industrious, prudently brave; they fought bravely, he painted finely.

And here it may be worth while to observe, how the same thing, participating the same essence, assumes different grammatical forms from its different relations. For example, suppose it should be asked, How differ honest, honesty, and honesty? The answer is, they are in essence the same; but they differ in as much as honest is the attribute of a noun; honesty, of a verb or adjective; and honesty being divested of these its attributive relations, assumes the power of a noun or substantive, so as to stand by itself.

The adverbs hitherto mentioned are common to verbs of every species; but there are some which are confined to verbs properly so called, that is, to such verbs as denote actions or energies with their privations. All motion and rest imply time and place as a kind of necessary coincidences. Hence, when we would express the place or time of other, we have recourse to adverbs formed for this purpose of place, as when we say, he stood there, he went then, he came hither; of time, as when we say, he stood there he went afterwards, he travelled formerly. To these may be added the adverbs which denote the intensions and remissions peculiar to motion, such as speedily, hastily, surely, slowly, &c.; as also adverbs of place made out of prepositions, such as upwords and downwords, from up and down. It may, however, be doubted whether some of these words, as well as many others, which do not so properly modify attributes, as mark some remote circumstance attending an attribute or our way of conceiving it, are truly adverbs, though so called by the grammarians. The simple affirmative and negative yes and no are called adverbs, though they surely do not signify that which we hold to be the very essence of the adverb, a modification of attributes. "Is he learned? No." "Is he brave? Yes." Here the two adverbs, as they are called, signify not any modification of the attributes brave and learned, but a total negation of the attribute in the one case, and in the other a declaration that the attribute belongs to the person spoken of.

Adverbs are indeed applied to many purposes; and their general nature may be better understood by reading a list of them, and attending to their etymology, than by any general description or definition. Many of them seem to have been introduced into language in order to express by one word the meaning of two or three, and are mere abbreviations of nouns, verbs, and adjectives. Thus, the import of the phrase, in what place, is expressed by the single word where; to what place, by whither; from this place, by hence; in a direction ascending, by upwards; at the present time, by now; at what time, by when; at that time, by then; many times, by often; not many times, by seldom; and so of others.

Mr Horne Tooke has, with great industry and accuracy, traced many of the English adverbs from their origin in the ancient Saxon and other northern tongues, and shown them to be either corruptions of other words or abbreviations of phrases and sentences. He observes, "that all adverbs ending in ly, the most prolific branch of the family, are sufficiently understood; the termination being only the word like corrupted, and the corruption so much the more easily and certainly discovered, as the termination remains more pure and distinguishable in the other sister languages, in which it is written lick, lyh, lig, lignen." He might have added, that in Scotland the word like is, at this day, frequently used instead of the English termination ly; as for a goodly figure, the common people say a good-like figure. Upon this principle the greater part of adverbs are resolved into those parts of speech which we have already considered, as honestly into honest-like, vastly, into vast-like, &c.; so that when we say of a man he is honestly industrious, we affirm that he is honest-like industrious, or that his industry has the appearance of being honest. Adverbs of a different termination the same acute writer resolves thus; agnast into the past participle agazed;

The French exclaimed, the Devil was in arms. All the whole army stood agazed on him.—Shaksp.

Ago, into the past participle agone or gone. Asunder he derives from asundered, separated; the past participle of the Anglo-Saxon verb asundrien, a word which, in all its varieties, is to be found in all the northern tongues, and comes originally from sand, that is, sand. To wit, from witan, to know; as videlicet and scilicet, in Latin, are abbreviations of videere-licet and scire-licet. Needs, he resolves into need is, used parenthetically; as, "I must needs do such a thing." "I must (need is) do such a thing;" that is, "I must do it, there is need of it." Anon, which our old authors use for immediately, instantly, means, he says, in one, that is, in one instant, moment, minute. As,

And right anon withouten more abode.

Anon in all the haste I can.

Alone and only are resolved into all one, and one-like. In the Dutch, een is one; and all een alone; and all-teen-like, only, anciently alone. Alive is on live, or in life. Thus, in Chaucer,

When Pope says of a certain person, that he is "a tradesman, meek, and much a liar," the last phrase is the same with much unto lying, the word liar having the effect of an attributive.

These words were anciently written one's, two's, three's, and are merely the genitives of one, two, three, the substantive time or being omitted. Thus, How often did you write? Answer, Once, that is, one's time. (See Horne Tooke's Dissertions of Purity). Aught or ought; a whit or o whit; o being formerly written for the article a, or for the numeral one; and ukit or huit, in Saxon, signifying a small thing, as a point or jot. Awhile, which is usually classed with adverbs, is evidently a noun with the indefinite article prefixed; a while, that is, a time. Whilst, anciently and more properly whiles, is plainly the Saxon huelics, time that. Aloft was formerly written on-loft. As, in Chaucer,

And ye, my mother, my soverayne pleasance Over al thing, out take Christ on lofte.

Now, says Mr Horne Tooke, luft, in the Anglo-Saxon, is the air or the clouds, as, in lyfte cummede, coming in the clouds (St Luke). In the Danish, luft is air; and at spronge i luften, to blow up into the air, or aloft. So in the Dutch, de loef hebben, to sail before the wind; loeren, to ply to windward; loef, the weather gage, &c. From the same root are our other words, loft, lofty, to luft, lee, leeward, lift, &c. It would be needless, as the ingenious author observes, to notice such adverbs as, afoot, adays, ashore, astray, aslope, aright, abed, aback, abreast, afloat, aloud, aside, afield, aground, aland, &c. These are at first view seen for what they are. Nor shall we follow him through the analysis which he has given of many other adverbs, of which the origin is not so obvious as of these. Of the truth of his principles we are satisfied; and have not a doubt, but that upon those principles a man conversant with our earliest writers, and thoroughly skilled in the present languages, may trace every English adverb to its source, and show that it is no part of speech separate from those which we have already considered. The adverbs, however, of affirmation and negation, are of too much importance to be thus passed over; and as we have never seen an account of them at all satisfactory, except that which has been given by Mr Horne Tooke, we shall transcribe the substance of what he says concerning aye, yea, yes, and no. To us these words have always appeared improperly classed with adverbs, upon every definition which has been given of that part of speech. Accordingly, our author says, that aye or yea is the imperative of a verb of northern extraction; and means have, possess, enjoy. And yes is a contraction of ay-er, have, possess, enjoy, that. Thus, when it is asked whether a man be learned, if the answer be by the word yes, it is equivalent to have that, enjoy that belief or that proposition.

The northern verb of which yea is the imperative, is in Danish ejer, to possess, have, enjoy. Eja, aye or yea; ejer, possession; ejer, possessor. In Swedish it is ega, to possess, of which the imperative is ja, aye, yea; egare, possessor. In German, ja signifies aye or yea; eigenere, possessor, owner; eigen, own. In Dutch, eigenen is to possess; ja, yea. Greenwood derives not and its abbreviation no from the Latin; Minshon, from the Hebrew; and Junius, from the Greek. Our author very properly observes, that the inhabitants of the north could not wait for a word expressive of dissent till the establishment of those nations and languages; and adds, that we need not be inquisitive nor doubtful concerning the origin and signification of not and no; since we find that, in the Danish nodig, in the Swedish nodlig, and in the Dutch, noode, node, and no, mean averse, unwilling.

So that when it is asked whether a man be brave, if the answer be no, it is a declaration that he who makes it is averse from or unwilling to admit that proposition.

Most writers on grammar have mentioned a species of adverbs, which they call adverbs of interrogation, such as where, whence, whither, how, &c. But the truth is, that there is no part of speech which, of itself, denotes interrogation. A question is never asked otherwise than by abbreviation, by a single word, whether that word be a noun, a pronoun, a verb, or an adverb. The word where is equivalent to in what place; whence, to from what place; and how, to in what manner, &c. In these phrases, in what place, from what place, and in what manner, the only word that can be supposed to have the force of an interrogative, is what, which is resolvable into that which. But we have already explained, under the head of Pronouns, the principles upon which the relative is made to denote interrogation; and the same reasoning will account for the adverbs where, whence, whither, how, &c. being employed as interrogatives. When we say, "Where were you yesterday? whence have you come?" whether are you going? how do you perform your journey?" we merely use so many abbreviations for the following sentences: "Tell us, or describe to us, the place where (or in which) you were yesterday; the place whence (or from which) you have come; the place to which you are going; the manner in which you perform your journey?" And so much for adverbs. We now proceed to those parts of speech which are usually called prepositions and conjunctions, and of which the use is to connect the other words of a sentence, and to combine two or more simple sentences into one compound sentence.

VI.—Of Prepositions, Conjunctions, and Interjections.

It has been observed, that a man whilst awake is conscious of a continued train of perceptions and ideas passing in his mind, which depends little upon his own will; that he cannot to the train add a new idea; and that he can but very seldom break its connection. To the slightest reflection these truths must be apparent. Our first ideas are those which we derive from external objects making impressions on the senses; but all the external objects which fall under our observation are linked together in such a manner as indicates them to be parts of one great and regular system. When we take a view of the things by which we are surrounded, and which are the archetypes of our ideas, their inherent qualities are not more remarkable than the various relations by which they are connected. Cause and effect, contiguity in time or place, high and low, prior and posterior, resemblance and contrariety, with a thousand other relations, connect things together without end. There is not a single thing which appears solitary and altogether devoid of connection. The only difference is, that some are intimately and some slightly connected, some nearly and some at a distance. That the relations by which external objects are thus linked together must have great influence in directing the train of human thought, so that not one perception or idea can appear to the mind wholly unconnected with all other perceptions or ideas, will be admitted by every man who

---

1 The same resolution might probably be made of the Greek and Latin adverbs, were we as intimately acquainted with the sources of those tongues as Mr Horne Tooke is with the sources of the English language. "Many of the Latin adverbs," says Ruddiman, "are nothing else but adjective nouns or pronouns, having the preposition and substantive understood; as quo, eo, modo, for ad quo, eo, cadem (locis) or cui, ei, eidem (loco); for of old these datives ended in a. Thus, qua, hac, illae, ece, are plainly adjectives in the ablative singular feminine, the word via, a way, and the preposition in, being understood. Many of them are compounds; as, quemodo, i.e. quo modo; quemadmodum, i.e. ad quem modum; quamobrem, i.e. ob quem rem; quare, i.e. (pro) quo re; quaremus, i.e. versus quem (locum); scilicet, i.e. scire licet; videlicet, i.e. videre licet; illece, i.e. in loco; magnumper, i.e. magna opera; mirum, i.e. si (est) mirum; hodie, i.e. hoc die; postulare, i.e. postero die; pridie, i.e. praedie. Perfectio, certe, sunt, male, bene, plene, are obviously adjectives. Forte is the ablative of fortis; and if we had leisure to pursue the subject, and were masters of all the languages from which the Latin is derived, we doubt not but we should be able to resolve every adverb into a substantive or adjective. believes that his senses and intellect represent things as they are.

This being the case, it is necessary, if the purpose of language be to communicate thought, that the speaker be furnished with words, not only to express the ideas of substances and attributes which he may have in his mind, but also to indicate the order in which he views them, ad to point out the various relations by which they are connected. In many instances all this may be done by the parts of speech which we have already considered. The closest connection which we can conceive, is that which subsists between a substance and its qualities; and in every language with which we are acquainted, that connection is indicated by the immediate coalescence of the adjective with the substantive; as we say, *a good man*, *accarned man*; *vir bonus*, *vir ducis*. Again, there is a connection equally intimate, though not so permanent, between an agent and his action; for the action is really an attribute of the agent; and therefore we say, *the boy reads*, *the man writes*; the noun coalescing with the verb naturally, that no other word is requisite to unite them. Moreover, an action and that which is acted upon being contiguous in nature, and mutually affecting each other, the words which denote them should in language be mutually attractive, and capable of coalescing without external aid; as, *he reads a book*, *he builds a house*, *he breaks a bone*. Further, because an attribute and its modification are inseparably united, an adjective or a verb is naturally connected with the adverb which illustrates or modifies its signification; and therefore, when we say, *he walks slowly*, *he is prudently brave*, it is plain that no other word is necessary to promote the coalescence of the attributes "walking" and "bravery" with their modifications of "slowness" and "prudence." The agreement between the terms of any proposition which constitutes truth is absolutely perfect; but as either of the terms may agree with many other things besides its correlate, some word is requisite in every proposition to connect the particular predicate with the particular subject; and that is the office of the simple verb to be; as, the three angles of every triangle are equal to two right angles.

Thus we see that many of the relations subsisting between our ideas may be clearly expressed by means of nouns, adjectives, verbs, and adverbs; and in those languages of which the nouns have cases, there is perhaps no reason of much importance which might not be thus pointed out, without being under the necessity of employing the aid of any additional part of speech. In English, however, the case is otherwise; for were we to say, "He rode Edinburgh, went the parliament-house, walked his counsel the court met," we should speak unintelligibly; as in these expressions there is either a total want of connection, or such a connection as produces falsehood and nonsense. In order to give meaning to the passage, the several gaps must be filled up by words significant of the various relations by which the different ideas are connected in the mind; as, "He rode to Edinburgh, went to the parliament-house, and walked with his counsel till the court met." Of these connecting words, and with are called prepositions, and and till are usually called conjunctions. Although these prepositions and conjunctions are not so absolutely necessary in Greek or Latin as they are in English, yet as there is no language wholly without them, nor any language in which it is not of importance to understand their force, they will deserve a place in universal grammar.

The sole use of conjunctions and prepositions in language is to connect either sentences or other words; but the theory of these connectives themselves has certainly never been understood, unless Mr Horne Tooke has at last hit upon the truth. Mr Harris writes about them and

about them, quoting passages from Greek and Latin authors, and produces at last no information. His definitions of both, as parts of speech void of signification, are highly absurd; and even the principal distinction which he makes between them seems not to be well founded. Prepositions and conjunctions denote the relations subsisting between the ideas expressed by those words or sentences which they serve to connect; and as relations are contemplated by the mind as well as positive ideas themselves, the words which denote these relations cannot be insignificant. The essential difference between the conjunction and preposition, according to the same author, consists in this, that the former connects sentences, and the latter words; but the fact is often otherwise. An obvious example occurs where the conjunction and connects not sentences but words. "A man of wisdom and virtue is a perfect character." Here it is not meant to be asserted, "that the man of wisdom is a perfect character, and that the man of virtue is a perfect character;" for both these assertions would be false. This sentence therefore (and many such will occur) is not resolvable into two; whence it follows, that the conjunction and does not always connect sentences; and the same is frequently the case with other conjunctions.

Mr Horne Tooke's idea of prepositions and conjunctions is, that they do not form distinct classes of words, but are merely abbreviations of nouns and verbs; and, with respect to the English language, he has been remarkably successful in proving his position. But though such be undoubtedly the case in English, it would be rash to conclude a priori that it is so in all other tongues. To establish this general conclusion would require a long and tedious deduction in each particular language; and how much language, leisure, industry, and acuteness, such an undertaking would require, even in one tongue, it is not easy to determine. In the languages with which we are best acquainted, many conjunctions, and most prepositions, have the appearance at least of original words; and though this most acute grammarian, from his knowledge of the northern tongues, has been able to trace the most important of those in English to very plausible sources, the same thing would be difficult in other languages of which the sources are obscure, and absolutely impossible in those of which they are wholly unknown. It is, however, a strong presumption in favour of his opinion, that grammarians have never been able to assign any general characteristic of those species of words, which, did they constitute distinct parts of speech, one would think could not have so long remained undiscovered. It is a further presumption in his favour, that many words in Greek and Latin, as well as in English, which have been called conjunctions, are obviously resolvable upon his principles, and indeed discover their meaning and origin upon mere inspection. We shall therefore content ourselves with retailing the common doctrine respecting these parts of speech so far as it is intelligible, subjoining at the bottom of the page the analysis given by Mr Horne Tooke, of the most important English conjunctions and prepositions; and requesting our readers, who would understand the subject, to attend more to the relations between their various ideas, than to the frivolous distinctions which, in compliance with custom, we are compelled to lay before them. We shall treat first of the conjunction.

1. Of Conjunctions.

A conjunction is a part of speech of which, as its name indicates, the use is to connect either two or more words in a sentence, or to make of two simple sentences one compound sentence. It is usually said that conjunctions never connect words, but sentences only, and that this is the circumstance which distinguishes them from prepositions. We have already given one example which proves this distinction to be ill founded; we shall now give from Mr Horne Tooke one or two more, which will place its absurdity in a still clearer light: Two and two are four; John and Jane are a handsome couple; AB and BC and CA form a triangle. Are two four? Is John a couple and Jane a couple? Does one straight line form a triangle? It appears indeed that and may connect any two things which can be connected, as it signifies addition.

Conjunctions connecting sentences sometimes connect their meaning, and sometimes not. For example, let us take these two sentences, "Rome was enslaved, Caesar was ambitious," and connect them together by the conjunction because: "Rome was enslaved because Caesar was ambitious." Here the meanings, as well as the sentences, appear to be connected by that natural relation which subsists between an effect and its cause; for the enslaving of Rome was the effect of Caesar's ambition. That particular relation therefore is that which is denoted by the conjunction because, which would be improperly used to connect two sentences between which the relation of an effect to its cause exists not. But if it be said, "manners must be reformed, or liberty will be lost," here the conjunction or, though it joins the sentences, yet as to their meaning is a perfect disjunctive. Between the reformation of manners and the loss of liberty there is certainly a natural relation; but it is not the relation of contiguity or similitude, or of cause and effect, but of contrariety. The relation of contrariety therefore is the signification of the word or. And thus it appears, that though all conjunctions may combine sentences, yet, with respect to the sense, some are conjunctive and others disjunctive.

Those conjunctions which conjoin both sentences and their meanings are either copulatives or continuatives. The principal copulative in English is and, which we have already considered. The continuatives are much more numerous; if, an, because, therefore, wherefore, hence, &c. The difference between them is this: The copulative does no more than barely couple words or sentences, and is therefore applicable to all subjects of which the natures are not incompatible. The relation which it denotes is that of juxtaposition, or of one thing added to another. Continuatives, on the contrary, by a more intimate connection, consolidate sentences into one continuous whole; and are therefore applicable only to subjects which have an essential relation to each other, such as that of an effect to its cause, or of a cause to its effect. For example, it is no way improper to say, "Lysippus was a statuary, and Priscian a grammarian; the sun shineth, and the sky is clear;" because these are things that can co-exist, and yet imply no absurdity. But it would be absurd to say, "Lysippus was a statuary because Priscian was a grammarian;" though not to say, "the sun shineth because the sky is clear." With respect to the first, the reason is, that the word because denotes the relation which an effect bears to its cause; but the skill of Priscian in grammar could not possibly be the cause of Lysippus's skill in statuary; the coincidence between the skill of the one and that of the other, in arts so very different, was merely accidental. With respect to the shining of the sun and the clearness of the sky, the case is widely different; for the clearness of the sky is the cause of the sun's shining, at least so as to be seen by us.

As to the continuatives, they are either suppositives, such as if, an; or positives, such as because, therefore, &c. Take examples of each: "You will live happily if you live honestly; you live happily because you live honestly; you live honestly, therefore you live happily." The difference between these continuatives is this: The suppositives denote connection, but do not assert actual existence; the positives imply both the one and the other.

1 And is a Saxon word, being, according to Mr Horne Tooke, an abbreviation of ANAD, the imperative of the verb ANADAN, to add to or heap up. So that when we say two and two are four, we only declare that two added to two are four.

2 Because is compounded of the Saxon ME, by, and cause; and by some of our most ancient authors it was written BY CAUSE. "Rome was enslaved because Caesar was ambitious," is therefore equivalent to, "Rome was enslaved by the cause Caesar was ambitious;" taking the phrase, "Caesar was ambitious," as an abstract noun in concord with the other noun cause.

3 Or seems to be a mere contraction of the Saxon OMEN, which signifies other, that is, something different and often contrary. So that the conjunction or must always denote diversity, and very often contrariety.

4 As day and night, heat and cold; for we cannot say of the same portion of time, it is day and it is night; or of the same body, it is both hot and cold.

5 The reason of all this will be apparent from the analysis given by Mr Horne Tooke of those words which we have called suppositive conjunctions. If and an may be used mutually and indifferently to supply each other's place; for they are both verbs, and of the same import. If is merely the imperative of the Gothic and Anglo-Saxon verb GIPTAN, to give; and in those languages, as well as in the English formerly, this supposed conjunction was pronounced and written as the common imperative GIPT. Thus,

My supposse

Hath lotted her to be your butcher's mistress, Gif shee can be reclaimed; gif not, his prey. Sad Shepherd.

Gawin Douglass almost always uses gif for if, as the common people in some counties of Scotland do even at this day; and it is obvious, that our if always the signification of the English imperative give, and no other. So that the resolution of the construction in the sentence, "If you live honestly you will live happily," is simply this, "Give you live honestly (taking you live honestly as an abstract noun) you will live happily." Your living happily is declared to depend upon your living honestly as the condition; but give that, and your happiness is positively asserted. In like manner may such sentences be resolved as,

I wonder he can move! that he's not fixed!

If that his feelings be the same with mine.

Thus, "His feelings be the same with mine, give that, I wonder he can move," &c. And here we cannot forbear giving our assent to the truth of Mr Horne Tooke's observation, that when the datum upon which any conclusion depends is a sentence, the article that, if not expressed, may always be inserted. We do not, however, think the insertion at all times absolutely necessary to complete the syntax; for active verbs govern whole sentences and clauses of sentences as well as substantive nouns. Instances of this occur so frequently in the Latin classics, that they can have escaped no man's notice who has ever read Horace or Virgil with attention. We agree likewise with this most ingenious author, that where the datum is not a sentence, but some noun governed by the verb if or give, the article that can never be inserted. For example, if we be asked how the weather will dispose of us to-morrow, we cannot say: "If that fair, it will send us abroad; if that foul, it will keep us at home;" but "if fair, it will send us abroad," &c. The reason is obvious. The verb in this case directly governs the noun; and the resolved construction is, "give fair weather, it will send us abroad; give foul weather, it will keep us at home."

An, the other suppositive conjunction mentioned, is nothing else than the imperative of the Anglo-Saxon verb ANAN, which likewise means to give or to grant. As, "An you had an eye behind you, you might see more distraction at your heels than fortune before you;" that is, "Grant you had an eye behind you, you might see," &c. This account of the two conditional conjunctions in English is so rational and satisfactory, that we are strongly inclined to believe that all those words which are so called, are in all languages to be accounted for in the same manner. Not indeed that they must all mean precisely to give or grant, but some word equivalent; such as, be it, suppose, allow, permit; which meaning is to be sought for in the particular etymology of each respective language. The positives above mentioned are either causal, such as because, since, as, &c.; or collective, such as therefore, wherefore, &c. The difference between them is this: The causals subjoin causes to effects; as, "the sun is in eclipse, because the moon intervenes;" the collectives subjoin effects to causes; as, "the moon intervenes, therefore the sun is in eclipse." We therefore use causals in those instances where, the effect being conspicuous, we seek for its cause; and collectives, in demonstration and science, properly so called, where the cause being first known, by its help we discern effects.

As to causal conjunctions, we may further observe, that there is not one of the four species of causes which they are not capable of denoting. For example, the material case; "The trumpet sounds because it is made of metal." The formal; "The trumpet sounds because it is long and hollow." The efficient; "The trumpet sounds because an artist blows it." The final; "The trumpet sounds that it may raise our courage." It is worth observing, that the first three causes are expressed by the strongest affirmation; because if the effect actually be, these must be also. But this is not the case with respect to the last, which is only affirmed as a thing that may happen. The reason is obvious; for whatever may be the end which set the artist at work, that end it may still be beyond his power to obtain; as, like all other contingents, it may either happen or not. Hence also it is connected by a particular conjunction, that, absolutely confined to this cause.

We come now to the disjunctive conjunctions, a species of words which bear this contradictory name, because whilst they conjoin the sentences, they disjoin the sense; or, to speak a language more intelligible, they denote relations of diversity or opposition.

That there should be such words, whether called conjunctions or not, is extremely natural. For as there is a principle of union diffused throughout all things, by which this whole is kept together and preserved from dissipation; so is there in like manner a principle of diversity diffused throughout all, the source of distinction, of number, and of order. Now it is to express in some degree the modifications of this diversity, that those words called disjunctive conjunctions are employed.

Of these disjunctives some are simple and some adversative: Simple, as when we say, "either it is day or it is night;" adversative, as when we say, "it is not day, but it is night." The difference between these is, that the simple express nothing more than a relation of diversity; the adversative express a relation not barely of diversity, but also of opposition. Add to this, that the adversatives are definite, the simple indefinite. Thus, when we say, "the number three is not an even number, but an odd," we not only disjoin two opposite attributes, but we definitely affirm... Conjunctions firm the one to belong to the subject, and deny the other.

But when we say, "the number of the stars is either even or odd," though we assert one attribute to be, and the other not to be, yet the alternative is notwithstanding left indefinite.

As to adversative disjunctives, it has been already said, after Mr Harris, that they imply opposition; but the truth seems to be, that they only unite in the same sentence words or phrases of opposite meanings. Now it is obvious, that opposite attributes cannot belong to the same subject; as when we say, "Nereus was beautiful," we cannot superadd to this sentence that he was ugly; we cannot say, "he was beautiful but ugly." When there is opposition, it must be either of the same attribute in different subjects; as when we say, "Brutus was a patriot, but Cæsar was not;" or of different attributes in the same subject, as when we say, "Georgius was a sophist, but not a philosopher;" or of different attributes in different subjects, as when we say, "Plato was a philosopher, but Hippasus was a sophist."

The conjunctions used for all these purposes have been called absolute adversatives, we think improperly, as the opposition is not marked by the conjunctions, but by the words or sentences which they serve to connect. Mr Locke, speaking of the word but, says, that "it sometimes intimates a stop of the mind, in the course it was going, before it came to the end of it;" to which Mr Horne Tooke replies with truth, that but itself is the farthest of any word in the language from intimating a stop. On the contrary, it always intimates something to follow; insomuch that when any man in discourse finishes his words with but, instead of supposing him to have stopped, we always ask, but what?

Besides the adversatives already mentioned, there are two other species, of which the most important are unless and although. For example, "Troy will be taken, unless the palladium be preserved; Troy will be taken, although Hector defend it." The nature of these adversatives may thus be explained. As every event is naturally allied to its cause, so by parity of reason it is opposed to its preventive; and as every cause is either adequate or inadequate (inadequate when it endeavours without being effectual), so in like manner is every preventive. Now adequate preventives are expressed by such adversatives as unless; "Troy will be taken, unless the palladium be preserved;" that is, this alone is sufficient to prevent it. The inadequate are expressed by such adversatives as although; "Troy will be taken, although Hector defend it;" that is, Hector's defence will prove ineffectual. These may be called adversatives adequate and inadequate.

Such is the doctrine of Mr Harris, which, although we can discover in it no determinate meaning, we have ventured with others to retail, from respect to our readers, who may be more perspicacious than ourselves. The author was a man of great learning, and the subject, as he has treated it, appears to be intricate. But whatever sense or nonsense there may be in what he says of causes and preventives, adequate and inadequate, we have no hesitation in affirming that he has totally mistaken the import of the words unless and although. From these being called both preventives, the one adequate and the other inadequate, an unwary reader might be led to infer that they denote the same idea or the same relation; and that the whole difference between them is, that the expression of the one is more forcible than that of the other. Nothing, however, can be further than this from the truth. The meaning of unless is directly opposite to that of although. Unless and though are both verbs in the imperative mode; the former signifying take away or dismiss, the latter allow, permit, grant, yield, assent. This being the case, "Troy will be taken unless the palladium be preserved," is a sentence equivalent to "Remove the palladium be preserved (taking the palladium be preserved as an abstract noun, the preservation of the palladium). Troy will be taken." Again, "Troy will be taken, although Hector defend it," is the same as "Troy will be taken allow Hector (to) defend it." The idea, therefore, expressed by unless is that of the removal of one thing to make way for another; the idea expressed by although is that of allow-

that of not in the preceding line. The meaning of the couplet is, "Superadded (to something said or supposed to be said before) thy works shall endure in laude and glorie, ne out (id est, without) spot or fault!" In the following passage from Donne, the word out, although written in the same manner, is used in both its meanings: "You must answer that she was brought very near the fire, and as good as thrown into it; or else, that she was provoked to it by a divine inspiration. But that another divine inspiration moved the beholders to believe that she did therein a noble act, this act of her's might have been calumniated." That is, "You must answer that she was brought very near the fire." Acc. "Superadded (to that answer) ne out (or unless without), for, as will be seen by and by, all these words are of the same import) that another divine inspiration moved." &c. To these remarks and examples it may be worth while to add, that even now out is often used by the illiterate Scotch for without; as nothing is more common than to hear a clown say, "He came from home but his breakfast."

Having mentioned without as a word of the same import with but when distinguished from set, it may not be improper to consider that word here; for though in modern English it is entirely confined to the office of a proposition, it was formerly used indifferently either as a preposition or a conjunction. Without, then, is nothing but the imperative wyethan-utan, from the Anglo-Saxon and Gothic verb wyethan, witan, which in the Anglo-Saxon language is incorporated with the verb knos, knos. According to this derivation, which is Mr Horne Tooke's, the word without, whether called conjunction or preposition, is the same as be set; and such will be its import, should it after all be nothing more than a compound of with, which signifies to join, and sometimes to be, and set, out.

1 Either is nothing more than a distributive pronoun which every body understands, and or we have already explained.

2 So low down as in the reign of Queen Elizabeth, says Mr Horne Tooke, this conjunction was sometimes written oules or oules; but more anciently it was written oules and sometimes oulesse. Thus, in the trial of Sir John Oldcastle in 1413, "It was not impossible for them to make whole Christes cote without some, oulesse certeyn great men were brought out of the way." So, in "The image of governance," by Sir T. Elliot, 1451, "Men do fere to approache unto their sovereigne Lord oulesse they be called." So again, in "A necessary doctrine and erudition for any Christian man, set furthe by the king's majestie of England," 1543, "Oules ye believe, ye shall not understand." "No man shall be crowned, oulesse he lawfully fight." "The soul waxeth feeble, oulesse the same be cherished." "It cannot begynne, oulesse by the grace of God." Now oulesse is the imperative of the Anglo-Saxon verb oulesan, to dissolve or remove.

Les, the imperative of lesan, which has the same meaning as oulesan, is likewise used sometimes by old writers instead of oulesse. Instances might be given in abundance from Gawin Douglass and Ben Jonson; but perhaps it may be of more importance to remark, that it is this same imperative les which, placed at the end of nouns, and coalescing with them, has given to our language such adjectives as hopeless, restless, deathless, motionless, &c., that is, dimiss hope, rest, death, motion, &c.

Mr Horne Tooke observes, that all the languages which have a conjunction corresponding to les or oulesse, as well as the manner in which the place of these words is supplied in the languages which have not a conjunction correspondent to them, strongly justify his derivation which we have adopted. The Greek nas, the Latin nisl, the Italian se non, the Spanish sino, the French si non, all mean be it not. And in the same manner do we sometimes supply its place in English by but, without, be it not, lest if, &c. It may be proper just to add, that, according to the same author, the conjunction lest is a contraction of leced, the past participle of leken; and that lest with the article that, either expressed or understood, means no more than hoc diminuo or quo diminuo.

Although is compounded of al or all, and the', though, that, or, as the vulgar more purely pronounce it, thouf, thauf, and thouf. Now in one thing to co-exist with another, with which it is apparently incompatible.

Before we take leave of this subject, we might treat, as others have treated, of adverbial conjunctions, and conjunctions of various other denominations. But of multiplying subdivisions there is no end; and systems in which they abound convey for the most part no information. The nature of conjunctions can be thoroughly understood only by tracing each to its original in some parent or cognate tongue; and when that shall be done in other languages with as much success as it has been done by Mr Horne Tooke in English, then, and not till then, may we hope to see a rational, comprehensive, and consistent theory of this part of speech. Then too shall we get rid of all that farce of useless distinctions into conjunctive, adjective, indicative, subjunctive, copulative, continuative, substantive, positive, suppositive, causal, collective, preventive, adequate and inadequate, adversative, conditional, illative and the like, which explain nothing, and which serve only to veil ignorance and perplex sagacity.

That Mr Horne Tooke's principles will apply exactly to the conjunctions of every language both dead and living, is why our limited knowledge of these languages does not authorize us positively to affirm. It is, however, a strong presumption in favour of his opinion, that illiterate savages, the first cultivators of language, are little likely to have set out their faculties in quest of words to denote the abstract relations subsisting among their ideas, when we have such evidence as his book affords, that the names of the most common substances and qualities could answer that and every other purpose, which in the ordinary intercourse of life can be answered by the faculty of speech. It is a further presumption in his favour, that in the rudest languages there are few if any conjunctions; and that even in those which are the most highly polished, such as Greek and Latin, as well as English, many of those words which have been called conjunctions are obviously resolvable into another parts of speech. Thus ἀλλά, translated but, is evidently the neuter gender of either the nominative or accusative plural of ἄλλος, another; and when used as a conjunction, it intimates that you are going to add something to what you have already said. Ceterum has the same meaning, and is nothing but xxi irges. Mais, but, in French, is the Latin majus; ut, uti, ut, quod, is the relative pronoun. Of quoque, quia, praeterea, antequam, quemquam, quemvis, quantumvis, quamlibet, &c., the resolution is too obvious to require being mentioned. Where such resolutions as these can be made, or when the conjunctions of any particular tongue can be traced to their origin in any other, there needs be no dispute about their true import; but when the case is otherwise, and the conjunction either appears to be an original word, or is derived from a source to which it cannot be traced, we would advise such of our readers as wish to speak or write correctly, to dismiss from their minds all consideration of copulatives, continuatives, causals, and disjunctives, with the rest of that jargon which we have already mentioned; and to inquire diligently in what manner and for what purpose the conjunction in question is used by the best writers, both ancient and modern, of the particular language which they are studying. This will indeed be found a work of labour; but it appears to us to be the only means left of discovering the precise relations which such conjunctions were intended to express; and, by consequence, of knowing what words or sentences they are fitted to connect, so as to produce a style at once accurate and perspicuous.

2. Of Prepositions.

By Mr Harris and his followers, a preposition is defined to be a part of speech devoid itself of signification, but so formed as to unite two words that are significant, and that refuse to coalesce or unite of themselves. We have already expressed our opinion of that theory which holds certain words to be devoid of signification; but its absurdity, in the present instance, is more than ever glaring. Concerning the number of prepositions, it is well known that hitherto authors have never been agreed. The ancient Greek grammarians admitted only eighteen, the ancient Latin grammarians above fifty; though the moderns, Sanctius, Sciopius, Perizonius, Vossius, and Ruddiman, have endeavoured to lessen the number without fixing it. Bishop

---

**Table:**

| Gip | Gir | |-----|-----| | Anan | To give. | | Onlesan | To grant. | | Eakan | To dismiss. | | Getan | To add. | | Stellan | To get. | | Alesan | To put. | | Thafigan | To diminish. | | or | To allow. | | Thafian | To boot, or superadd. | | Botan | To be out. | | Beon-utan | To be out. | | Wyethan-utan | Dare congeriem. |

Lest is the participle lesed, of lesan, to dismiss.

Since is the participle of seon, to see.

That is the article or pronoun that.

As is es, a German article, meaning it, that, or which. And So is sa or so, a Gothic article of the same import with as. Wilkins thinks that thirty-six are sufficient; and Girard says that the French language has done the business effectually with thirty-two. But if prepositions be words devoid of signification, why should there be disputes respecting their numbers? or why in any language should there be more than one preposition, since a single unmeaning mark of connection would certainly answer the purpose as well as a thousand? The cipher, which has no value of itself, and only serves, if we may use the language of grammarians, to connote and consignify, and to change the value of figures, is not several and various, but uniformly one and the same. That "the preposition is so formed as to unite two words which refuse to coalesce or unite of themselves," is indeed true; and this union it effects, not by having no signification of its own, but by signifying the relation by which the things expressed by the united words are connected in nature. Prepositions are to be accounted for in much the same manner as the cases of nouns. The necessity of this species of words, or of some equivalent invention, follows from the impossibility of having in language a distinct complex term for each distinct collection of ideas which we may have occasion to put together in discourse. The addition or subtraction of any one idea to or from a collection of ideas, makes it a different collection; and if, after either of these operations, it were to be expressed by the same word as before, nothing could ensue but misrepresentation and falsehood. Now, to use in language a different and distinct complex term for each different and distinct collection of ideas, is equally impossible as to use a distinct particular term for each particular and individual idea. To supply, therefore, the place of the complex terms which are wanting in a language, are the cases of nouns and prepositions employed; by the aid of which, complex and general terms are prevented from being infinite or too numerous, and are used only for those collections of ideas which we have most frequent occasion to mention in discourse. By means of prepositions this end is obtained in the most simple manner. For, having occasion to mention a collection of ideas, for which there is no single complex term in the language, we either take that complex term which includes the greatest number, though not all, of the ideas we would communicate; or else we take that complex term which includes all and the fewest ideas more than those we would communicate; and then, by the help of the preposition, we either make up the deficiency in the one case, or retrench the superfluity in the other. For instance, having occasion to mention a house of a particular description, and knowing that the term house is too general for our purpose, and that the building we have in view has no appropriate name, we say, perhaps, "a house with a party-wall," or a "house without a roof." In the first instance, the complex term house is deficient, and the preposition directs to add what is wanting; in the second instance, the complex term is redundant, as it denotes a complete house. The preposition, therefore, directs to take away what is superfluous.

Now, considering prepositions in this the most simple light, as serving only to limit or modify general terms, it is absolutely necessary that they should have meanings of their own; for otherwise, how could we, in the instance before us, make known by them our intention, whether of adding to, or retrenching from, the same general term house. If, to a disciple of Mr Harris, we should say, a house join; he would reply join what? But he would not contend that join is an indeclinable word which has no meaning of its own, because he knows that it is the imperative of a verb, of which the other parts are still in use; and its own meaning is clear, though the sentence is not completed. If, instead of join, we should say to him, a house with, he would still ask the same question, what? But if we were to discourse with him concerning the word with, he would probably tell us that with is a preposition, an indeclinable word, which is itself devoid of signification, but so formed as to unite two words that are significant. And yet it would be evident by his question that he felt it had a meaning of its own; which is in reality the same as join. Indeed, so far has always been plainly perceived, that with and without are directly opposite and contradictory; and it would puzzle the most acute philosopher to discover opposition and contradiction in two words where neither of them had any signification. Wilkins, therefore, has well expressed their meaning, where he says, that with is a preposition "relating to the notion of social, or circumstance of society affirmed;" and the without is a preposition relating to the same notion of social, or circumstance of society denied."

But to denote the relations of adding and taking away, is not the only purpose for which prepositions are employed. They all indeed serve to modify some general term or general affirmation, but not precisely in the same way as with and without. It has already been observed, that words significant of those things which coincide in nature, coalesce with one another in syntax, without being beholden to any auxiliary tie. For instance, an adjective coalesces with its substantive, a verb with its nominative; a noun expressing an object acted upon, with a verb de-

---

1 This account of prepositions is taken from Mr Horne Tooke, who adds, that the only difference between the two words with and join, is, that the other parts of the Gothic and Anglo-Saxon verb withan, to join (of which with is the imperative), have ceased to be employed in the language. As with means join, so the corresponding French preposition avec means, and have that, or have that also. But though with, as the imperative of withan, means join, it has sometimes a very different signification. Mr Tyrwhit in his Glossary has truly observed, that with and by are often synonymous. They certainly are so; but then with seems to be an abbreviation of the imperative of wyrtan, to be; as without is of wyrtan-utan, to be out. This being the case, our two instances in the text will stand thus: a house join a party-wall; a house be-out a roof. Nor let any one be surprised that we make no difference between the conjunction without and the preposition without. The word is the same, whether it be employed to unite words or sentences. Prepositions were originally, and for a long time, clasped with conjunctions; and when first separated from them, they were only distinguished by the name of prepositive conjunctions. They are generally used to unite words, but not always; for we may say differently, "I came after his departure," or "I came after he departed." By the greater part of grammarians indeed, when employed as in the first sentence, is clasped with the prepositions; when employed as in the second, it is clasped with the conjunctions. The word, however, is the same in both sentences; its meaning is the same, and its effect precisely the same. The only circumstance of discrimination is, that in the first example it is prefixed to a noun, "his departure;" in the second, it is prefixed to a nominative and a verb, "he departed." But even the nominative and the verb, thus applied, express more than a specifying circumstance annexed to the other proposition, "I came;" and whenever they are rightly apprehended by the mind, they are stripped of their prepositional form, and considered abstractedly under a new phasis, "his departure." Thus then the two sentences are synonymous in every respect, excepting the apparent grammatical nature of the words "his departure," and "he departed;" and even these are reduced to one grammatical form in the mind, whenever the import of the propositions is rightly apprehended. Without, and many other prepositions, especially in the learned languages, are used exactly as over is used in the two instances which we have given. Mr Horne Tooke quotes Lord Mansfield for saying, "It cannot be read without the attorney-general consents to it." This, in modern English, is not the common phraseology; but it offends not against any principle of grammar. The nominative and the verb are here, as in the former instance, considered as an abstract noun. "It cannot be read WITHOUT the consent of the attorney-general." The splendid sun genially warmtheth through the fertile earth. But suppose we were desirous to modify the affirmation by the addition of other substantives, air, for instance, and beams; how would these coincide with the other words of the sentence, or under what character could they be introduced? Not as nominatives or accusatives to the verb, for both these places are already filled; the nominative by the substance sun, which is certainly the agent in this operation; the accusative by the substance earth, which is as certainly the object acted upon. Not as qualities of the sun and earth; for qualities inhering in these substances can only be expressed by adjectives, and the words air and beams are both substantives. Here then we must have recourse to prepositions; but we can employ only such prepositions as point out the relations which in air and the beams have to the sun warming the earth. In English we should say, "the splendid sun with his beams genially warmtheth through the air the fertile earth." The sentence, as before, remains entire and one; the substances required are both introduced; and not a word which was there before is detruded from its proper place. The import of with we have already discovered; it directs to unite the beams to the sun, as jointly with him performing the operation. But the air has no other connection with the operation, than as the medium or passage between the sun and the earth; and therefore the preposition through must denote that relation which subsists between an object in motion and the medium in which it moves; nor could a preposition of a different import have been employed, without altering the meaning of the whole sentence.

If Harris is of opinion that most, if not all, prepositions were originally formed to denote the relations of place. For this opinion we see not sufficient evidence. If indeed we could suppose the inventors or earliest improvers of language to have at all concerned themselves with relations as abstracted from the objects related, we must believe that those which first attracted their attention were the relations subsisting among themselves, and the various bodies with which they were surrounded. We must likewise agree with our author, that place is the grand relation which bodies or natural substances maintain at all times to one another; but we do not therefore think that it would attract the earliest notice of untought barbarians. On the contrary, we are of opinion that mankind must have made very considerable progress in science before they attempted to abstract place from body; an attempt which, according to some of the most profound philosophers, is not only difficult, but absolutely impracticable. But whatever be in this, the relations of cause and effect, of duration and motion, are in themselves as obvious, and as likely to arrest the attention and obtain names, as those of place. Amongst men totally illiterate they are evidently more so; for pain and pleasure would suggest some idea of cause and effect as matters of importance. There is, however, no probability that the inventors of any language had the least idea of abstract relations. They doubtless expressed complex conceptions by nouns and verbs, significant at once of the particular ideas and of the various relations by which they viewed those ideas as combined together in a complex conception. Afterwards, when men's minds became enlarged, and when, from the fluctuation inseparable from a living language, objects or ideas received new names, the old words, whether nouns or verbs, which were originally employed to express a particular complex conception, of which certain particular relations made a part, might be retained for the purpose of denoting those and all similar relations; and thus verbs and nouns would degenerate into particles bearing the names of prepositions and conjunctions. For instance, one Anglo-Saxon being desirous to communicate to another his own conception of a house with a party-wall, and having, we shall suppose, no such word in his tongue as a preposition, would naturally utter the word house, desiring his friend, at the same time, to add to that well-known sound another sound (uttering it) significant of the particular circumstance wanting to complete his complex conception, viz. a house with (that is join) a party-wall. The word with, as the imperative of a verb, denotes of course three ideas combined together, namely, a command or wish, an affirmation, and the idea of junction. But when the verb withan was dismissed from the English language, the imperative with was still retained; but losing its verbal and modal nature, it was thenceforth employed to denote only one of the three ideas for which it originally stood, viz. the idea of junction. And thus it is that verbs, and also nouns and adjectives, in passing from one language to another, may become prepositions and conjunctions. Thus too it is that some of those prepositions come to denote the contiguous, and some the detached, relation of body. The contiguous, as when we say, "Caius walked with a staff;" that is, "Caius join a staff, walked;" "the statue stood upon a pedestal;" that is, "the statue stood (the place of its standing) the higher part of a pedestal;" "the river ran..." over a sand," that is, "the river ran (the place of its running) the higher part of a sand." The detached relation, as when we say, "He is going to Italy," that is, "He is going, the end (of his journey) Italy;" "the sun is risen above the hills," that is, "the sun is risen (the place) the top of the hills;" "these figs came from Turkey," that is "these figs came beginning (their journey at) Turkey."

Besides the detached relation of body, Mr Harris is of opinion that the preposition from denotes two other relations not less different than those of motion and rest. Thus, if we say, "That lamp hangs from the ceiling," the proposition from assumes a character of quiescence. But if we say, "That lamp is falling from the ceiling," the proposition in such case assumes a character of motion. But this is evidently a mistake. The detached relation in the former instance of the figs, as well as the motion and rest in the present instances, are expressed not by the preposition, but by the verbs, came, falls, hangs. The word from has as clear, as precise, and at all times as uniform and unequivocal a meaning, as any word in the language. From means merely beginning, and nothing else. It is simply the Anglo-Saxon and Gothic noun frum, beginning, origin, source, fountain, author. Now if this meaning be applied to Mr Harris's instances, from will speak clearly for itself, without the assistance of the interpreting verbs, which are supposed by him to vary its character.

necessary to insert, between the verb and pedestal, a word significant of place, that pedestal may not be mistaken, by an ignorant person, for a portion of time, or any thing else connected with the standing of the statue.

That to is significant of detached relation, is the language of Mr Harris, which, though it may be allowed in a loose and vulgar sense, is certainly not philosophically just. The preposition to, in Dutch written toe and tot, is the Gothic substantive tau or taute, signifying act, effect, result, or consummation; which Gothic substantive is itself no other than the past participle tauid or taute of the verb tauljan, agere. And it is obvious that what is done is terminated, ended, finished. In the Teutonic this verb is written tuan or tuon; whence the modern German thun, and its preposition zu. In the Anglo-Saxon, the verb is teogan, and the preposition to. Do, the auxiliary verb, as it has been called, is derived from the same root, and is indeed the same word as to. The difference between a r and a d is so very small, that an etymologist knows by the practice of languages, and an antiquary by the reason of that practice, that in the derivation of words it is scarcely worth regarding. To support this etymology of to, Mr Horne Tooke gives a similar instance in the Latin tongue. The preposition ad, he says, is merely the past participle of agere, which past participle is likewise employed as a Latin substantive. He exhibits the derivation of ad thus:

\[ \begin{align*} \text{Agitum} & \rightarrow \text{agium} \\ \text{Agium} & \rightarrow \text{ad} \\ \text{or} & \rightarrow \text{of} \\ \text{Actum} & \rightarrow \text{act} \\ \text{Actum} & \rightarrow \text{at} \\ \end{align*} \]

The most superficial reader of Latin verse, he observes, knows how readily the Romans dropped their final s. And a little consideration of the organs and practice of speech will convince him how easily agit or act would become ad or at; as indeed this proposition was indifferently written either way by the ancients. By the later writers of Rome, the preposition was written ad with d only, in order to distinguish it from the other corrupt word called the conjunction at; which for the same reason was written with the c only, though that also had evidently been written, as the proposition, either ad or at. The proposition to and the conjunction ne in English, are both in syntax and in meaning used exactly as the proposition ad and the conjunction at in Latin. From the specimen prefixed to Johnson's Dictionary, as a history of our language, it appears that, as late as the reign of Elizabeth, the proposition and conjunction were both written with one o. And it has been shown in the first volume of the Transactions of the Royal Society of Edinburgh, that to and ne, as well as ad and at, are precisely of the same import. The only difference, in either language, between the proposition and the conjunction, is, that the former directs, as a modification of some previous proposition, the addition of some substantive or noun; the latter, sometimes a sentence or clause of a sentence considered abstractedly as a noun; and that, when the former is used, the preposition to which the modifying circumstance is to be added, is formally expressed, but omitted when the latter is employed. Thus Denham says,

Wisdom he has, and to his wisdom courage; Temper to that, and note all success.

In this example, every succeeding circumstance is by the preposition to marked as an addition to the preceding. "Wisdom he has, and courage additional to his wisdom." But Denham might with equal propriety have omitted the object which to governs, or to which it directs something to be added, though he must then, from the custom of the language, have employed the conjunction instead of the preposition. As,

Wisdom he has, and courage too.

That this mode of expression would have been more concise than, and as intelligible as, the other, "Wisdom he has, and courage to his wisdom," must be evident to every one.

Not only is the object governed by to omitted, when it is represented by a substantive in the context, but also when it is involved in a preposition; and then the conjunction, as it is called, is always used. Thus,

Let those eyes that view

The daring crime, behold the vengeance too.

So, "He made him prisoner, and killed his son." In the one example, the circumstance of beholding the vengeance is stated as an addition to the viewing of the crime; and in the other, the killing him is stated as an addition to the making him a prisoner. In both examples, the object governed by to is the amount of the preceding proposition taken abstractedly as a noun or substantive. Thus then it appears, that to and too, though clasped the one with the prepositions, and the other with the conjunctions, are really one and the same word. The same is true of ad and at. Thus, "An hoc, promissa barba et capilli effaverant sperantem oriri," signifies "Additional to this, his long beard and hair had given a wildness to his aspect." But when the object governed by ad is not formally stated, ad itself is clasped with the conjunctions, and written differently, at. Thus Terence, "Ph. Fui ut ut jusse, dedecorantur illi."

Pa. Faciam. Ph. At diligenter. Pa. Fiect. Ph. At mature. By the means of at, the circumstances of diligence and haste are superadded to the action commanded. Ph. It is not enough that you do it; you must do it carefully too. Pa. Well, it shall be carefully done. Ph. In good time too. At, taken in this sense, is most commonly employed, like the English but, to mark the unexpected union of incongruous objects; as, in Aulani tyrannus frequentavit, "at patre amabat;" literally, "He frequented the court of the tyrant; joined crew to that he loved his country;" "He was a coward, and a patriot too." But if ad and at in Latin, and to and too in English, be derived from verbs which signify to do or act, it may be asked how they come themselves to denote addition. The answer is obvious. If a man should utter a sentence, and to the end of it subjoin the very general word do, the person to whom he spoke, would naturally ask do what? and this question would, of course, produce an additional sentence or clause of a sentence. Besides, it is to be observed, that agere, from which the Latin proposition is derived, as well as the Gothic verb, which is the source of the English particles, means not only to do, but also to adduce or bring; so that when we say, "he is going to Italy," we do nothing more than affirm that "he is going," and desire the person to whom we speak, "to add Italy to the journey."

From this derivation of the preposition to, it will be seen at once upon what principle it is employed to mark the infinitive mode. In the learned languages that mode is generally known by its termination; but in English it would be impossible, without the aid of to, or of some other word significant of action, to distinguish the verb love from the noun or substantive.

This derivation is Mr Horne Tooke's; and he supports it by the following sentence: Ne reddid ge se se on framanne worldi, he wrohte wyrmpan and welfpan; which is the Anglo-Saxon of St Matthew, xix. 4, "Annon legisitis, quod qui eos in principio creavit, creavit eos marem et feminam?" These figs came from Turkey. That lamp falls from the ceiling. That lamp hangs from the ceiling.

"Come" is a complex term for one species of motion; "falls" is a complex term for another species of motion; and "hangs" is a complex term for a species of attachment. Have we occasion to communicate or mention the commencement or beginning of these motions, and of this attachment, and also the place where they commence or begin? To have complex terms for each occasion of this sort is absolutely impossible; and therefore nothing can be more natural or more simple than to add the signs of those ideas, viz., the word beginning, which will remain always the same, and the name of the place, which will perpetually vary. Thus,

These figs came—BEGINNING Turkey. That lamp falls—BEGINNING ceiling. That lamp hangs—BEGINNING ceiling.

That is, Turkey the place of BEGINNING to come. Ceiling the place of BEGINNING to fill. Ceiling the place of BEGINNING to hang.

It has been said by no less a man than Bishop Wilkins, that from refers primarily to place, and secondarily to time. By the truth is, that from relates to every thing to which beginning relates, and to nothing else.

From morn till night the eternal larum rang. That is, the "larum rang beginning morning," or morning being the time of its beginning, till night.

"To" always denotes beginning, so to and till always denote the end. There is, however, this difference between them, that to denotes the end of any thing; till the end only of time. We may say indifferently, "From morn to night," or "from morn till night, the eternal larum rang;" but we cannot say, "These figs came from Turkey till England." That till can, with propriety, be opposed to from only when we are talking of time, is evident; for it is a word compounded of to and while, that is, time.

And as the coalescence of these two words to-while took place in the language long before the present superfluous use of the article the, the phrase, "From morn till night" is either more nor less than "From morn to time night."

When we say, "from morn to night," the word time is omitted as unnecessary.

Besides from, Mr Harris mentions over as significant sometimes of motion and sometimes of rest; and quotes as instances the two following passages from Milton:

To support uneasy steps Over the burning marl.

He says he, over denotes motion. Again,

He with looks of cordial love Hung over her enamoured.

Here over denotes rest. But the truth is, that over denotes neither motion nor rest in either of these passages. In the first quotation, indeed, motion is implied; but it is implied in the word steps, and not in over, which denotes that the place of the steps was the top of the burning marl. In the second quotation, rest is implied, and too, the too a particular species of rest; but it is implied, or rather expressed, by the verb hung, and over denotes the place of that species of rest.

But though the original use of prepositions was to denote the relations of body, they could not be confined to this office only. They by degrees extended themselves to subjects incorporeal, and came to denote relations, as well intellectual as local. Thus, because in point of place he who is above has commonly the advantage of him who is below; hence we transfer over and under to dominion and obedience. Of a king we say, "he ruled over his people;" and of a soldier, "he served under his general." So too we say, with thought; without attention; thinking over a subject; under anxiety; from fear; through jealousy, and the like. All which instances, with many others of a similar kind, show that the first words of men, like their first ideas, had an immediate reference to sensible objects; and that in after days, when they began to discern with their intellect, they took those words which they found already made, and transferred them by metaphor to intellectual conceptions.

Amongst the relations which may be considered rather as intellectual than corporeal, are those of cause and consequence; and for the denoting of these we have two prepositions, which sometimes appear in direct opposition to one another, and at other times may exchange places without injury to the sense.

"Well! 'tis o'er so! I have the London disease they call love. I am sick of my husband, and for my gallant."

(Wycherley's Country Wife.)

Here of and for seem almost placed in opposition; at least their effects in the sentence appear to be very different; for, by the help of these two prepositions alone, and without the assistance of any other words, she expresses the two contrary affections of loathing and desire. The truth, however, is, that the author, if it had pleased him, might have used of where he has employed for, and for where he has put of. This is evident from the following quotation from the Sad Shepherd:

Marian. Come, Annie, you'll go with us. Annie. I am not well. Lucel. She's sick of the young shepherd that beheld her.

In the same manner we may, with equal propriety, say, "We are sick of hunger;" or, "We are sick for hunger." And in both cases we shall have expressed precisely the same thing, with only this difference, that, in the former sentence, we declare sickness to be a consequence; in the latter we declare hunger to be a cause. But to return to the Country Wife, that poor lady seems to have had a complication of distempers; she had at least two disorders; a sickness of loathing, and a sickness of love. She was sick for disgust, and sick for love. She was

Sick of disgust for her husband. Sick of love for her gallant. Sick for disgust of her husband. Sick for love of her gallant.

In the first sentence, as thus stated, sickness is declared to be the consequence of disgust, of which her husband is declared to be the cause. In the second, sickness is declared to be the consequence of love, of which her gallant is declared to be the cause. In the third sentence, disgust is declared to be the cause of her sickness, and the consequence or offspring of her husband. In the fourth, love is declared to be the cause of her sickness, and the consequence or offspring of her gallant.

These are complex terms because they are verbs. Each denotes an affirmation and time; and, combined with these, came and fell denote motion, and hangs denotes rest.

Under and beneath, though by the sound they seem to have little connection, are yet in fact almost the same word, and may very well supply each other's place. Under is nothing but on-nder, and beneath is compounded of the imperative be and the noun neth.

Neth, being compounded having slipped away from our language, would perhaps be unintelligible, had not the nouns nether and nethermost continued in common use. Neth, Anglo-Saxon, neððan, neððe; Dutch, neden; Danish, ned; German, niedre; and Swedish, ned; and neder, is undoubtedly as much a substantive, and has the same meaning, as the word neder. In common language it denotes the bottom. Thus, then, it appears, that though the first two of these sentences, taken entire, convey the very same meaning with the last two, yet the import of the preposition *for* is as different from that of *of*, as *cause* is from *consequence*. When two words or sentences are linked together by the former of these prepositions, the object expressed by the last word or sentence is declared to be the *cause* of that which is expressed by the preceding; when two words or sentences are linked together by the latter preposition, the object expressed by the first word or sentence is declared to be the *consequence* of, or to proceed *from*, the object expressed by the second. It is therefore a matter of perfect indifference to the sense whether we say "sickness of hunger," or "sickness for hunger;" "the man, of he speaks little, is wise," or "the man is wise, for he speaks little." By means of the preposition *of*, we declare sickness to be the *consequence* proceeding from hunger, and wisdom to be the *consequence* proceeding from the man's speaking little; by means of *for*, we declare hunger to be the *cause* of sickness, and the circumstance of speaking little to be the *cause* from which we infer the man's wisdom. In the one sentence, *of* is to be considered as a noun in apposition to *sickness*; in the other, as a noun in apposition to *the man is wise* (taken abstractedly as a noun). In the one sentence *for* (id est, *cause*) is to be considered as a noun in apposition to *hunger*; in the other, as the same noun in apposition to *he speaks little* (taken abstractedly as a noun).

In the foregoing use of prepositions, we have seen how they are applied by way of juxtaposition; that is to say, where they are prefixed to a word without becoming a part of it. But they are used also by way of composition; that is, they are prefixed to other words so as to become real parts of them. Thus in Greek we have ἵππος; in Latin, *intelligere*; and in English, *understand*. So also, to *ronstell*, or *overrock*, to *undervalue*, to *outgo*; and in Greek and Latin other instances innumerable. In this case the prepositions commonly transfuse something of their own peculiar meaning into the word with which they are compounded. For example, if we suppose some given space, *e* and *ex* signify *out of* that space; *per*, *through it*; *in*, *within it*; *sub*, *under it*. Hence, *e* and *per*, in composition, augment: *Enormis* is something not simply big, but big in excess; something got out of the rule, and beyond the measure. *Dico*, "to speak;" *edic*, "to speak out;" whence *edictum*, "an edict;" something so effectually spoken that all are supposed to hear and to obey. On the contrary, *in* and *sub* diminish and lessen. *Injustus*, *inxiquus*, "unjust, inequitable;" something that lies within justice and equity, that reaches not so far, that falls short of them. *Sumiger*, "blackish;" *surubicundus*, "reddish; tending to black, and tending to red;" but yet under the standard, and below perfection.

Before we dismiss this part of our subject, we shall make the same general remark upon prepositions which we formerly made on conjunctions, viz., that the precise import of each can with certainty be known only by tracing it to its source in some word of known and determinate meaning, either in the language where the preposition itself has place, or in some parent or cognate tongue. And I urge it may be laid down as an infallible rule, that where different languages use the same or a similar particle, that language ought to be considered as its legitimate parent, in which the true meaning of the word can be found, and where its use is as common and familiar as that of any other verbs and substantives. When prepositions can be traced to such sources as these, no room can be left for disputes concerning their meaning. In carrying on this etymological pursuit, we find advantages in the nature of prepositions which conjunctions do not afford us. *With* and *without*, *from* and *to*, with many other words belonging to this class, have meanings directly opposite and contradictory to each other. If, then, by the total or partial extinction of an original language, the root of any one preposition be lost, whilst that of its opposite remains, the philosopher ought to be satisfied with reasoning from contrariety, as nothing is more evident than that the meaning of a word is known when we know with precision the meaning of its opposite. When we meet, however, with a lackless preposition of which no root is left to be dug up, and which has itself no direct opposite in the language, nothing remains but that we inquire for what purpose it is used by the best writers both ancient and modern; and if we can fix upon one meaning which will apply, however awkwardly, to all the places where it occurs, or to the greater part of them, the probability is, that we have discovered the true and original meaning of the preposition; and by keeping that meaning constantly in view, we shall ourselves be enabled to use the word with perspicuity and precision.

3. Of Interjections.

Besides the above parts of speech, there is another acknowledged in all the languages of Europe, called the interjection; a word which cannot be comprehended under any of the foregoing classes. The genuine interjections are very few in number, and of very little importance, as they are thrown into a sentence without altering its form either in syntax or in signification. In the words of Mr Horne Tooke, the brutish inarticulate interjection has nothing to do with speech, and is only the miserable refuge of the speechless. The dominion of speech, according to the same author, is erected on the downfall of interjections. Without the artful contrivances of language, mankind would have nothing but interjections with which to communicate orally any of their feelings. "The neighing of a horse, the lowing of a cow, the barking of a dog, the purring of a cat, sneezing, coughing, groaning, shrieking, and every other involuntary convulsion with oral sound, have almost as good a title to be called parts of speech as interjections. In the intercourse of language, interjections are employed only when the suddenness or vehemence of some affection or passion returns men to their natural state, and makes them for a moment forget the use of speech; or when, from some circumstance, the shortness of speech will not permit them to exercise it." The genuine interjection, which is always expressive of some very strong sensation, such as *Ah!* when we feel pain, does not owe

---

1 Junius derives *for* from the Greek *πρό*; Skinner from the Latin *pro*; but I believe, says Mr Horne Tooke, that it is no other than the Gothic substantive *farihna*, *cause*. He imagines also that *of*, in the Gothic and Anglo-Saxon *af*, is a fragment of the Gothic and Saxon words *afara* and *afora*, *posterior*, *proles*. In a word, he considers *for* and *of* as nouns or substantives; the former always meaning *cause*, the latter always meaning *consequence*, *offspring*, *successor*, *follower*, &c. If this account of these words be just, and we have no doubt of it, the prepositions *for* and *of* are in syntax to be considered as nouns in apposition with other nouns, or with sentences taken abstractedly as nouns.

2 For instance, let us suppose that Mr Horne Tooke's derivation of *for*, from the Gothic substantive *farihna*, is fanciful and ill founded; yet there can be little doubt but *cause* is its true and original meaning, when it is found, that of sixteen examples brought by Greenwood and forty-six by Johnson, of different significations of the word *for*, there is not one where the noun *cause* may not be substituted instead of the preposition *for*; sometimes indeed awkwardly enough, but always without injury to the sense. Even where *for* seems to be *locu alterius*, which Lowth asserts to be its primary sense, it will be found to be *cause*, and nothing else. Thus, "He made considerable progress in the study of the law before be quitted that profession;" that profession *for* this of poetry;" that is, "before he quitted that profession, this of poetry" being the *cause* of his quitting it. characteristical expression to the arbitrary form of articulation, but derives its whole force from the tone of voice and modification of countenance and gesture. Of consequence, these tones and gestures express the same meaning, without any relation whatever to the articulation they may assume; and are therefore universally understood by all mankind. Voluntary interjections are used in books only for embellishment, and to mark forcibly a strong emotion. But where speech can be employed, they are totally useless, and are always insufficient for the purpose of communicating thought. Dr Beattie ranks strange, pious, amazing, wonderful, O dear, dear me, &c., when used alone, and without apparent grammatical syntax, amongst the interjections; but he might with as much propriety have considered hardly, truly, really, and even many Latin verbs, as interjections; for these too are often used alone to supply the place of whole sentences. The truth is, that all men, when suddenly and violently agitated, have a strong tendency to shorten their discourse by employing a single word to express a sentiment. In such cases, the word employed, whether noun, adjective, or verb, would be the principal word of the sentence, if that sentence were completed; and the agitation of the speaker is such, and the cause of it so obvious, that the hearer is no danger of mistaking the sense, and can himself supply the words which are wanting. Thus, if a person, after listening to a romantic narrative, were to exclaim, strange! could any man of common sense suppose, that the word strange, because uttered alone, had lost the power of an adjective and become an interjection? No, surely; every one sees that the exclamation is equivalent to, "that is strange," or "that is a strange story." Real interjections are never employed to convey truth of any kind. They are not to be found amongst laws, in books of civil institutions, in history, or in any treatise of useful arts or sciences; but in rhetoric and poetry, in novels, plays, and romances, where in English, so far from giving pathos to the style, they have generally an effect that is disgusting or ridiculous.

Having now analysed every part of speech which can be necessary for the communication of thought, or which acknowledged in any language with which we are acquainted, we shall dismiss the subject, after annexing an abstract, which may present at one view the several classes and subdivisions of words. Of the different modes of dividing the parts of speech, as well as of the little importance of systematic classifications, we have already declared our decided opinion; but for the sake of those who may think differently from us, we shall in the abstract subjoined adopt Mr Harris's classification as far as it is intelligible; after informing our readers that Mr Horne Cooke admits only three parts of speech, the article, the noun, and the verb, and considers all other words as corruptions or abbreviations of the two last of these.

GRAMMATICAL ABSTRACT,

Exhibiting a Systematic View of Words, as they are commonly arranged into distinct Classes, with the Subdivisions under each Class.

All language is composed of words, each of which may be defined, a sound significant of some idea or relation. These words may be arranged into four general divisions, namely:

I. SUBSTANTIVES; which are all those words that are expressive of things which exist, or are conceived to exist, of themselves, and not as the energies or qualities of anything else. These may be divided into two orders, viz.

1. Nouns, properly so called, being the names of all those things which exist, or are conceived to exist. These may be divided into three kinds, each of which admits of the subdivisions after mentioned, viz.

Natural, or those which are used as the names of natural substances; such as (genus) animal; (species) man, dog; (individual) Alexander, Cyrus, Cerberus, Argus.

Artificial, or the several names of artificial objects; such as, (genus) edifice; (species) house, church; (individual) Vatican, St Peter's, St Paul's.

Abstract, or those which are the names of qualities considered as abstracted from their substances; such as, (genus) motion; (species) flight, course; (individual) the falcon's flight, the greyhound's course.

Nouns of all kinds admit of the following accidents, viz.

Gender, which is a certain affection of nouns denoting the sex of those substances of which they are the names. For as in nature every object is either male or female, or neither the one nor the other, grammarians, following this idea, have divided the names of beings into three classes. Those which denote males are said to be of the masculine gender; those which denote females, of the feminine gender; and those which denote neither the one nor the other, of the neuter gender. The English is the only language of which the nouns are, with respect to sex, an exact copy of nature.

Number. As there is no object in nature single and alone, and as by far the greater part of nouns are the names of whole classes of objects, it is evident that every such noun ought to have some variation, to denote whether it is one individual of the class which is meant, or more than one. Accordingly we find, that in every language nouns have some method of expressing this. If one be mentioned, the noun is used in that form which is called the singular number; if more than one, it is used in a different form, which is called the plural number.

Cases. All nouns, excepting proper names, are general terms; but it is often necessary to use those general terms for the purpose of expressing particular ideas. This can be done only by connecting the general term with some word significant of a quality or circumstance peculiar to the individual intended. When that quality or circumstance is not expressed by an adjective, it is in English and most modern languages commonly connected with the noun by the intervention of a preposition; but in the Greek and Latin languages the noun has cases to answer the same, and even in English the noun has, besides the nominative, one case to denote possession.

2. Pronouns, which are a species of word invented to supply the place of nouns in certain circumstances. They are of two kinds, viz.

First, the prepositive; so called because they are capable of leading a sentence. These are divided into three orders, called the pronouns of the

First person. The pronoun of this person, in English I, denotes the speaker as characterized by the present act of speaking, in contradistinction to every other character which he may bear. It is said to be of the first person, because there must necessarily be a speaker before there can be a hearer; and the speaker and hearer are the only persons employed in discourse.

Second person. The pronoun of this person, in English thou, denotes the person addressed as characterized by the present circumstance of being addressed. It is said to be of the second person, because in discourse there cannot be a hearer till there be a speaker. The pronouns of the first and second persons have number and cases, for the same reason that nouns have these accidents; but in no language have they any variation denoting gender. The reason is, that sex, and all other properties and attributes whatever, excepting those just mentioned as descriptive of the nature of these pronouns, are foreign from the mind of the speaker when he utters I or thou in discourse.

Third person. The pronouns of this person, he, she, it, are employed to denote any object which may be the subject of discourse different from the speaker and the hearer. They are improperly said to be of any person; for there can be but two persons employed in discourse, the speaker and the party addressed. They are, however, pronouns; since they stand by themselves, and are the substitutes of nouns. He is the substitute of a noun denoting a male animal; she, of a noun denoting a female animal; and it, of a noun denoting an object which has no sex. All these, like the pronouns personal, admit of number and cases; but there is this peculiarity attending them, that though in every case of the singular number the distinction of gender is carefully preserved, in the plural it is totally lost; they, theirs, and them, being the nominative, possessive, and accusative cases of he, of she, and of it.

Secondly, the subjunctive; so called, because they cannot lead a sentence, but only serve to subjoin a clause to another previously enunciated. Of this kind are

Which and who. This subjunctive pronoun may be substituted in the place of any noun whatever, whether it be expressive of a genus, a species, or an individual; as, the animal which, the man who, Alexander who, &c. Nay, it may even become the substitute of the personal pronouns themselves; as when we say, I who now write, you who now read, thou who readest, he who wrote, she who spoke; where it is observable, that the subjunctive who adopts the person of that prepositive pronoun which it represents, and affects the verb accordingly. Who and which, therefore, are real pronouns, from substitution; and they have this peculiarity besides, that they have not only the power of a pronoun, but also of a connective of the same import with that which in English is expressed by the preposition of. The word that is now used indifferently for who or which, as a subjunctive pronoun; but it was originally used only as a definitive, and as such it ought still to be considered in philosophical grammar.

II. ATTRIBUTIVES; which are those words that are expressive of all such things as are conceived to exist, not of themselves, but as the attributes of other things. These are divided into,

1. VERBS, or those words which are expressive of an attribute and an assertion; as, I wrote.

The attributes expressed by verbs have their essence in motion or its privation; and as motion is always accompanied by time, therefore verbs are liable to certain variations called TENSES, viz.

The present, which represents the action of the verb as going on, and as contemporary with something else; as, I write, or I am writing, either just now, or when you are reading. The preter-imperfect, which represents the action of the verb as having been going on but not finished in some portion of past time; as, I was writing, no matter when, yesterday, last week, or last year. The aorist of the past, which represents the action of the verb as finished in some indefinite portion of past time; as, I wrote, or did write, yesterday, or last week. The preter-perfect, which represents the action of the verb as just now finished, or as finished in some portion of time within which the present instant is comprehended; as, I have written this day or this week. The plusquam-perfect, which represents the action of the verb as having been finished in some portion of time within which a determinate past instant was comprehended; as, I had written last week before I saw you. The first future, which represents the action of the verb as to be going on at some indefinite future time; as, I shall write or be writing to-morrow, or next week. The second future, which represents the action of the verb as to be completed at some definite future time; as, I shall have written when you come to-morrow, next week, next month, or next year.

Affirmation is the essence of every verb, insomuch that all verbs may be resolved into the substantive verb is, and another attributive. But a man may affirm something of the action of the verb directly, something of his liberty or capacity to perform that action, or something of his wish that another should perform it. To denote these several kinds of affirmation, all verbs have what grammarians call MODES, viz. the indicative, to denote the first kind of affirmation as, I write; the subjunctive, to denote the second; as, I may or can write; the imperative, to denote the third; as, write thou, or do thou write. Besides these, grammarians have given to every verb a mode, called the infinitive; as, to write. But this seems on every account to be improperly styled a mode. Nay, if affirmation be the essence of verb, the infinitive cannot be considered as any part of the verb at all, for it expresses no affirmation. It is indeed nothing more than an abstract noun denoting the simple energy of the verb in conjunction with time.

Verbs have likewise been distinguished into the following kinds, according to the nature of the attribute of which they are expressive, namely, 1st, Active-transitive, or those which denote an action that passes from the agent to some external object; as Caesar conquered Pompey; 2d, Active-intransitive, or those which express that kind of action which has no effect upon any thing beyond the agent himself; as, Caesar walked; 3d, Passive, or those which express not action but passion, whether pleasing or painful; as, Portia was loved, Pompey was conquered; 4th, Nuter, or those which express an attribute that consists neither in action nor in possession; as, Caesar stood.

2. PARTICIPLES, or those words which are expressive of an attribute combined with time. In English there are only two participles; the present, as writing, which expresses the action of the verb to write, as going on; and the past, as written, which expresses the action of the same verb as finished, and therefore past in time. In Greek and Latin there is a future participle, by which the attribute is represented as being in a state of exertion at some future time; as, γραφόμενος, scripturus, about to write.

3. ADJECTIVES, or those words which express as inhering in their substances the several qualities of things, of which the essence consists not in motion or its privation; as, good, bad, black, white, large, small, and the like. Attributes are the same whether they belong to males or females, to one object or to many, adjectives ought in strictness to admit of no variation respecting sex or number; and in English they actually admit of none. Some qualities, however, are of such a nature, that one substance may have them in a greater degree than another; and therefore the adjectives denoting these qualities admit in most languages of a variation which grammarians call the degrees of comparison. Thus Plato was wise, Socrates was wiser than he, but Solomon was the wisest of men. There is a species of adjectives derived from nouns, and even from pronouns; for we say, the Pompeian party, a brass trumpet, and my book, which are phrases equivalent to "the party of Pompey, a trumpet of brass, and the book of me."

4. ADVERBS, or those words which, as they denote the attributes of attributes, have been called attributives of the second order, to distinguish them from verbs, participles, and adjectives, which denote the attributes of substances, and are therefore called attributives of the first order. Adverbs are divided into two kinds, viz.

First, Those which are common to all attributives of the first order, that is, which coalesce equally with verbs, with participles, and with adjectives. These may be divided into adverbs of intensification and remission, or of quantity continuous; as moderately, vastly, exceedingly, &c. These, like adjectives of a similar nature, admit of the dif- different degrees of comparison. Of quantity discrete; as, once, twice, thrice, &c. These are not, in strictness of speech, adverbs, being in reality the possessive cases of me, two, three, &c. Of relation; as more, most, less, least, equally, proportionally, &c.

Second, Those which are confined to verbs properly so called, and which are of the following kinds: Of time; as, then, when, afterward, now, &c.; of place; as, here, there, where, hence, whence, &c.; and also adverbs derived from repositions; as, upward, downward, &c.; of intensions and remissions peculiar to motion; as, speedily, hastily, slowly, &c. We have given adverbs a place amongst the parts of speech necessary for the communication of thought; but it may be doubted whether they be entitled to this distinction. English adverbs at least seem to be nothing more than corruptions of nouns, adjectives, and verbs.

III. DEFINITIVES; which are all those words that serve to define and ascertain any particular object or objects as separated from others of the same class. These are commonly called ARTICLES, which are divided into two kinds, viz.

1. Indefinite, as a or an, which is prefixed to a noun or general term, to denote that but one individual is meant of that genus or species of which the noun is the common name. This article, however, leaves the individual itself quite indeterminate. Thus, man is the general name of the whole human race; a man is one individual, but that individual is unknown. Any is prefixed to a noun either in the singular or plural number, when it is indifferent as to the truth of the proposition what individual or individuals are supposed; thus, "any man will be virtuous when temptation is away." Some is prefixed to nouns in the plural number, to denote that only part of the species or genus is meant, leaving that part undetermined; thus, "some men are great cowards."

2. Definite, as the, which is prefixed to a noun, to denote one individual of the species of which something is predicated that distinguishes it from every other individual. Thus, "the man that hath not music in himself is fit for treasons." It is used before nouns in both numbers, and for the same purpose; for we may say, "the men who have not music in themselves are fit for treasons." This, prefixed to nouns in the singular number, denotes an individual as present and near at hand; as, "this man beside me." That, prefixed to a noun in the singular number, denotes an individual as present but at a little distance; as, "that man in the corner."

Besides these, there are many other articles both definite and indefinite.

IV. CONNECTIVES, or those words which are employed to connect other words, and of several distinct arts to make one complete whole. These may be divided into two kinds, viz.

1. Conjunctions, by which name are distinguished all those connectives which are commonly employed to connect sentences. These have been divided into two kinds, called conjunctives, or those words which conjoin sentences and their meanings also; and disjunctives, or those words which, at the same time that they conjoin sentences, distinguish their meanings. Each of these general divisions has again been subdivided; the former into copulatives and continuatives, the latter into simple disjunctives and adversative disjunctives. But the general division is absurd, and the subdivisions are useless. Conjunctions never diminish the meaning of sentences, nor have any other effect than to combine two or more simple sentences into one compound sentence. If those simple sentences be of opposite meanings before their combination, they will continue so after it, whatever conjunction be employed to unite them. In nature, different truths are connected, if they be connected at all, by different relations; and therefore, when the sentences expressive of those truths are connected in language, it must be by words significant of those natural relations. Thus, accidental addition is expressed by the conjunction and; as when we say, "Lyssipus was a statuary, and Priscian was a grammarian." The unexpected junction of contrary truths is expressed by but; as, "Brutus was a patriot, but Caesar was not." The relation of an effect to its cause is expressed by because; as, "Rome was enslaved because Caesar was ambitious." The relation of an effect to a cause of which the existence is doubtful, is expressed by if; as, "you will live happily if you live honestly;" the relation of a cause to its effect, by therefore; as, "Caesar was ambitious, therefore Rome was enslaved." The idea of simple diversity is expressed by either and or; "either it is day or it is night." Contrariety between two affirmations, which, though each may be true by itself, cannot both be true at once, is expressed by unless; as, "Troy will be taken unless the Palladium be preserved." Coincidence of two affirmations apparently contrary to each other is expressed by although; as, "Troy will be taken although Hector defend it."

2. Prepositions, or those connectives of which the common office is to conjoin words which refuse to coalesce; and this they can do only by signifying those relations by which the things expressed by the united words are connected in nature. The first words of men, like their first ideas, had an immediate reference to sensible objects; and therefore there can be no doubt but that the original use of prepositions was to denote the various relations of body. Afterwards, when men began to discern with their intellect, they took those words which they found already made, prepositions as well as others, and transferred them by metaphor to intellectual conceptions.

Prepositions viewed in this light are

Either proper, or those which literally denote the relations subsisting amongst the objects of sense, such as the accidental junction of two things between which there is no necessary connection; as, "a house with a party wall." The separation of two things which we should expect to find united; as, "a house without a roof, a man without hands." The relation subsisting between any thing and that which supports it; as, "the statue stands upon a pedestal." The relation of higher and lower; as, "the sun is above the hills," "to support uneasy steps over the burning marsh," "the sun is set below the horizon," "the shepherd reclines under the shade of a beech tree." The relation between any thing in motion and that in which it moves; as, "the rays of light pass through the air." The relation between any thing continued, whether motion or rest, and the point of its beginning; as, "the rays of light proceed from the sun," "these figs came from Turkey," "that lamp hangs from the ceiling." The relation between any thing continued and the point to which it tends; as, "he is going to Italy," "he slept till morning." The relation between an effect and its cause; as, "I am sick of my husband and for my gallant."

Or metaphorical. For as those who are above others in place have generally the advantage over them, the prepositions which denote the one kind of superiority or in- Relative Pronoun.

In the learned languages the antecedent is not frequently repeated along with the relative. In such expressions as, "Erant omnino itinera duo quibus itineribus donec exire possent," "Urbem quam statuo vestra est," "OTZ εν της τελευτας ΣΥΜΜΑΧΟΣ ἐκρινομένα, ἤγγιξεν ἰσχυρῶς, ἀπολαβὼν αὐτόν," the antecedent is repeated along with the relative, which agrees with it in the same manner as any adjective agrees with its substantive. The relative, therefore, must be an adjective, expressing an attribute of its antecedent, and representing it in some way or other. Again, the application of the relative is universal. There is no object whatsoever to which, in use, the relative may not refer, and which it may not represent. We may say, the animal which, the man who, the thing which, wisdom which; George who, I who, thou who, he that; and so in innumerable other instances. The attribute expressed by the relative is therefore extremely general, such indeed as may belong to any object whatsoever. The first matter for inquiry, then, is to endeavour to ascertain what the attribute in question really is, or, in other words, to determine the signification of the relative pronoun.

On this branch of the subject, however, it is not necessary to enlarge, after what has already been said respecting it. The analytical solution proposed by Mr Harris, on the authority of Apollonius De Syntaxi, namely, into the prepositional pronoun and the conjunction and, has already been noticed and refuted. The equation, as stated by him, and also by Dr Beattie, who adopts his view, does not consist of equivalents, either logical or grammatical. The solution is not, therefore, as it ought to be, of universal application. In some instances, indeed, it may appear to have the same effect in a sentence as the copulative conjunction and prepositional pronoun; but in others, and by far the greater number, it will be found that the effect produced by the substitution of the alleged equivalents is altogether different. Thus, if the sentence, "Light is a body which moves with great celerity," be resolved into, "Light is a body, and it moves with great celerity," as proposed by Mr Harris, the result is two distinct predication respecting light, viz. first, that "it is a body," and secondly, that "it moves with great celerity;" whereas, in the original sentence, "Light is a body which moves with great celerity," there is but one predication made concerning "light," namely, that "it is a body of great or astonishing velocity." It is also to be observed, that the prepositional pronoun in the one sentence, and the relative pronoun in the other, refer to different antecedents. Thus, in the sentence, "The magnet is a stone, and it attracts iron," magnet is the antecedent to it; but in the sentence, "The magnet is a stone which attracts iron," stone, and not magnet, is the antecedent to which. Nor is this all. There are many instances where the resolution proposed, by introducing a double instead of a single predication, would produce either nonsense or falsehood. We can say with perfect truth, that "Legendre was the first mathematician who remarked the coincidence of symmetrical magnitudes;" but if this be converted into "Legendre was the first mathematician, and he remarked the coincidence of symmetrical magnitudes," we obtain two propositions, the

---

1 Fateamur profecto opertet, simulque nos delectat fatari, quod nihil est hodie tentamen nisi prolatio, exigua quidem, praedictum docentissimi atque magistri Joannis Hunter, L.L.D. in Academia Andreapolitana Litterarum Humaniorum Professoria.

2 Cæsar, Comment. lib. i. c. 6.

3 Virgil. Æn. lib. i. l. 573.

4 Demosth. Olynth. B.

---

From his extensive acquaintance with classical literature, and the philosophical bent of his mind, Mr Harris was certainly a man eminently capable of making important philological discoveries, had he chosen, like Mr Horne Tooke and Dr Hunter, to think boldly for himself, instead of esteeming it a sort of heresy to differ from the opinions of the ancient grammarians. Of these, Apollonius was no doubt one of the most learned and acute, as we shall have occasion afterwards to show; and in a question of fact, his authority is not to be disputed. But the case is very different in regard to a theory, the truth or falsehood of which must be established by evidence, derived not from the dicta of grammarians, but from the principles and usages of language. It is an excessive and misplaced deference to authority, however, which has led Mr Harris to crowd his Hermes with a multiplicity of uncouth terms and unnecessary distinctions, which, though they give an air of learned precision to many of his propositions, have an obvious tendency to darken rather than elucidate the subject which he professes to explain. fewer of which contains an affirmation erroneous in point of fact. In like manner, if the expression, "Solomon was the son of David who rebelled against his father," be resolved into "Solomon was not the son of David, and he rebelled against his father," we have one negative proposition which is historically true, converted into two propositions, the one negative and the other affirmative, but both of them historically false. Mr Harris subjoins as a curious confirmation of his theory, "that the Latins in its structure of the relative seem to have well represented its compound nature, of part pronoun and part connective, in forming their qui and quis from qui and is, or, as Saliger would have it, from xai and i, xai and i." But, whatever there may be in this, the foregoing examples prove that the relative has often an effect very different indeed from that which is produced by combining a connective with the prepositive pronoun.

If, however, the connective (which is nothing but an arbitrary interpolation) be kept out of view, and the relation between the different pronouns be carefully considered, it will, on examination, be found that the relative is merely a particular explanation of the demonstrative pronoun. That the article is really a demonstrative pronoun, cannot be denied by any one who has attentively considered its true nature and import. The common application of ὁ μεν, ἢ δια, and το μεν, το δια, in Greek, renders this sufficiently obvious; and, besides, direct evidence may be produced to establish the fact. In the expression TOI ἐξωγενεῖς λέει, "illī quidem sunt daemones," ἐξωγενεῖς might have been used instead of τοι; and again, in the sentence, ἘΤΟ ξαι TO ἐξωγενεῖς ἀδελφοί τινες τοι ἦσαν ἀνθρώποι, "if man had done that and that (such and such things), it would not have been dead," ἐξωγενεῖς might have been used instead of τοι. Nor is this confined to Greek alone. In English also the article is used as a demonstrative pronoun. Thus, when Nathan said unto David, "Thou art the man," its expression is the same as if the warning minister of heaven had said, "Thou art that man." And the same practice likewise obtains in the Gothic and in the Anglo-Saxon, where, according to Manning, the article put demonstratively has the same signification as hic, hae, hoc; id est, id; illē, illa, illud.

But not only is the article employed as a demonstrative pronoun; in the more ancient Greek writers, particularly in Homer and Herodotus, it is frequently used as a relative pronoun. This has already been shown, and need not be further illustrated by examples, which might be produced in any number required. Ὀτι TOI Ὀλυμπίου ἰψοῦς is manifestly the same as Ὀτι Ὀλυμπίου ἰψοῦς. And, conversely, the relative is sometimes used as an article. "Imo, inquit Vigerus, absolute 'ΟΣ pro 'Ο nonnunquam usurpatur. Ὅριον (καὶ παράσημον) τῷ ἑαυτῷ ἀνεξάρτητος, ἀφ' ὅτι τὸ ἑαυτόν ἀνεξάρτητον 'quorum altero recte semper utens, altero indigentibus abunde subveniens; ubi ἡ ὉΤΙ pro ἀπὸ τοῦ ἑαυτοῦ. Anciently the article was used as the relative pronoun in English as well as in Greek. The Lord's Prayer, in its older form, commenced thus: "Our father the art in heaven," where the is obviously used as a relative pronoun. And the same thing obtains in the Gothic and Anglo-Saxon, in which last, according to Manning, the put after nouns of all the persons signifies qui, who; as Ic ἑαυτόν, Ego qui, I who. The same author adds, "Apud Gothos similiter articulus, per omnes causas, vicem gerit pronominis relative; ut sa in himinam, 'qui in celis.'" Again, the demonstrative pronoun that, in English, is used indiscriminately with the relative. Thus we say indifferently, "The man that endureth to the end," or, "The man who endureth to the end?" "The messenger that Alexander sent to Darius," or, "The messenger whom Alexander sent to Darius?" "The mountain that overhangs the lake," or, "The mountain which overhangs the lake;" and so universally. Upon the whole, therefore, it may be concluded that the relative is merely a particular application of the demonstrative pronoun; or, in other words, that the signification of the relative is radically the same with that of the demonstrative pronoun.

But what, it may be asked, is the precise signification of the demonstrative pronoun? When it is said, "Cæsar Dumorigem cepit atque interfecit eum," it is evident that the eum refers to and represents Dumorix. A question however arises, namely, under what character does it represent him? Cæsar took Dumorix; but eum does not represent him as a prisoner, and there is no other circumstance connected with Dumorix which the pronoun can suggest to the mind, except the previous mention of his name. From the reason of the thing, therefore, independently of any induction, it may be concluded that eum can only be equivalent to the said or before-mentioned; and, accordingly, it will upon trial be found that, wherever the demonstrative pronoun occurs, some such expression as said or before-mentioned may be substituted in its stead. And this view is confirmed by the etymology of the demonstrative pronoun in as far as it can be traced either in Greek or in any other language.

For example, the Greek pronoun αὐτός, corresponding to the Latin ipse or idem, and often to hic, appears to be the past participle of the old verb αὐτός, clamo or dico, formed upon the same analogy with the Latin past participle so Relative that *quoth*, the Greek demonstrative pronoun, is in fact nothing but *dictus* or *said*. Another familiar instance is the use of the word *ditto* in book-keeping; as when it is said "paid ditto in full." *Ditto*, as everybody knows, is the Italian *detto*, formed from the Latin *dictus*, in the same manner as *facto* from *factus*, *ritto* from *rectus*; and hence the expression, "paid ditto in full," is "paid said person in full," or, more simply, "paid him in full." Again, in our own language, Mr Horne Tooke has satisfactorily traced the pronoun *it*, anciently written *hit*, and still so pronounced in many parts of the country, to the past participle of the Anglo-Saxon verb *haitan*, *dicere*, *nominare*; whence it follows that *it* or *hit* signifies *said* or *named*. It may, therefore, with some confidence be concluded that *said* was the original import of the demonstrative pronoun; and, as the relative pronoun has been shown to be only a particular application of the demonstrative, the signification of the former must, *ex paritate rationis*, be identical or synonymous with that of the latter, and, consequently, equivalent to *said* or *named*.

In a learned essay, to which Dr Gray's prize was adjudged, on the import and application of the relative pronoun, the author, Mr H. Carmichael, shows that the apparent etymology of this pronoun, in the Gothic language, affords the strongest possible corroboration of the truth of the above conclusion; insomuch as the Gothic relative is merely, with a slight change of the aspirate into sigma, the past participle of the verb signifying *dicere*, to say. And the admitted connection traceable between the Gothic and the ancient dialects of the Greek language suggests another conjecture, which serves to strengthen the probability of the preceding etymology, and also to confirm the truth of the general proposition above enunciated. The observation of Mr Horne Tooke, that Anglo-Saxon verbs, having i as their characteristic letter, change that letter, in the formation of their past tense, into a broad, is equally applicable to Masso-Gothic verbs; thus *quiman* makes *quam*, *quitthan* *quath*, and so on. But as different persons pronounced differently, and also used different written characters as representatives of their sounds, this change of the characteristic letter was exhibited either by a broad, or by o, or by u; and hence it may be inferred from analogy that *quoth* would also be pronounced *quoth*, which accordingly is still met with in the colloquial dialect of several districts in Scotland. It would also appear that *quoth* was sometimes pronounced and written *quod*; thus, in Chaucer, "Sir Knight (quod he), my maister and my lord." *Quod* and *quoth* have therefore been in use. But, in Anglo-Saxon, *quid*, *dictum*, frequently occurs; and from the same analogy we may infer that *quith* was also in use at one period. It follows, then, that *quoth* and *quith* are in reality the past tense or past participle of the verb *quitthan*, *dicere*. But as the old form of the Greek relative pronoun was *σε*, different parts of which, as *σε*, *σε*, *σε*, occur frequently in Herodotus, it may not unreasonably be conjectured that the Gothic *quoth* formed the Greek *σε* (the *θ* becoming *γ* in the same manner as *δ* in the imperative of the second aorist, becomes first *δ*, and then *δ*), and that *quid* formed the Greek *σε*, afterwards written *τι*, and the Latin *quis*. Further, it has been observed, that in the Masso-Gothic, *sael*, *qui*, and the article *sa* corresponding sometimes to the Greek *δ*, and sometimes to *σε*, bear the strongest appearance of having formed part of a Gothic verb corresponding to the Anglo-Saxon *saxhan*, *dicere*; a circumstance which serves to confirm the truth of the position, that the real signification of the relative pronoun is *said* or *named*.

From this induction, which appears to be conclusive, we now proceed to consider the uses of this pronoun in language, and at the same time to inquire whether, in the different examples produced, the relative corresponds, in effect, to the signification which we have assigned to it. The principal uses of the relative are, first, when it renders the clause in which it stands, in some respect or other, descriptive of its antecedent, or, in other words, when the relative is equivalent to an adjective noun; and, secondly, when it intimates that the amount of the clause to which it is prefixed is to be taken as a substantive noun.

1. With regard to the first of these uses; namely, when the relative clause can be expressed by a single term, every one must perceive that it describes or qualifies the antecedent, and is therefore equivalent to an adjective noun. Thus, in the sentence "vir sapit qui paucia loquitur," the relative clause, *qui paucia loquitur*, may be expressed by the term *pauciloguous*, and "vir pauciloguous sapit" is therefore exactly the same as "vir sapit qui paucia loquitur"; that is, *qui paucia loquitur* is precisely equivalent to *pauciloguous*, or, in other words, the relative clause is equivalent to an adjective noun. But the relative is by no means restricted, in its application, to those cases where the clause which it introduces may be expressed by a single term; on the contrary, the latter may be as long and as complex as the writer or speaker may choose, although there be no single term in the language expressive of the same combination of ideas. When, for instance, we say, "He is the happy man, who, at any moment of his existence, can reflect on a life which has been spent in the conscientious discharge of his duty;" the clauses, "who can reflect on a life," and "which has been spent in the conscientious discharge of his duty," may not be expressible by single terms; but if they were, these terms would

---

**VERBS**

| Οράλλω | Susurrō | |--------|--------| | Καλῶ | Voco | | Κλέβω | Celebro | | Κρίνω | Orno | | Λέγω | Legō | | Αποκρίνω | Supplico | | Μύω | Misceo | | Νυμπτίας αμβίο | Nuptias ambio | | Ορίζω | Porrigō | | Πίνω | Mitto | | Πάγω | Pango | | Πίνω | Opto | | Ρεῖν | Dico | | Ρεῖν | Roboro |

**PAST PARTICIPLES**

| Οράλλω | παράλληλοςΤΟΣ | multum celebratus | |--------|----------------|------------------| | Καλῶ | παλλώςΤΟΣ | vocatus | | Κλέβω | πλεύςΤΟΣ | celebatus | | Κρίνω | πρόπτηςΤΟΣ | clausus | | Λέγω | πλεύςΤΟΣ | ornatus | | Αποκρίνω | πρόπτηςΤΟΣ | lectus | | Μύω | μύστηςΤΟΣ | ter-supplicatus | | Νυμπτίας αμβίο | μύστηςΤΟΣ | mistus | | Ορίζω | πρόπτηςΤΟΣ | desponsatus | | Πίνω | πρόπτηςΤΟΣ | correctus | | Πάγω | πρόπτηςΤΟΣ | missus | | Πίνω | πρόπτηςΤΟΣ | pactus | | Ρεῖν | πρόπτηςΤΟΣ | ter-optatus | | Ρεῖν | πρόπτηςΤΟΣ | dictus | | Ρεῖν | πρόπτηςΤΟΣ | robustus |

---

**VERBS**

| Στατεῖν | Tege | |---------|-----| | Στατεῖν | Pungo | | Τινεῖν | Ordino | | Τινεῖν | Tingo | | Τινεῖν | Perfiro | | Τινεῖν | Texo | | Θείνω | Emo | | Παίνω | Suffico | | Παίνω | Facie | | Παίνω | Aspergo | | Φέρω | Torreo | | Χρείω | Ungo | | Υποκρίνω | Tango | | ΑΤΤΩ | Dico |

**PAST PARTICIPLES**

| Στατεῖν | στατεῖνΤΟΣ | tectus | |---------|-------------|--------| | Στατεῖν | στατεῖνΤΟΣ | punctus | | Τινεῖν | τινεῖνΤΟΣ | ordinatus | | Τινεῖν | τινεῖνΤΟΣ | tinctus | | Τινεῖν | τινεῖνΤΟΣ | perforatus | | Τινεῖν | τινεῖνΤΟΣ | textus | | Θείνω | θείνωΤΟΣ | emptus | | Παίνω | παίνωΤΟΣ | suffocatus | | Παίνω | παίνωΤΟΣ | factus | | Παίνω | παίνωΤΟΣ | aspersus | | Φέρω | φέρωΤΟΣ | tostus | | Χρείω | χρείωΤΟΣ | uncetus | | Υποκρίνω | υποκρίνωΤΟΣ | tactus | | ΑΤΤΩ | ΑΤΤΩΤΟΣ | Dictus |

---

1 In legal phraseology *said* is itself used as a demonstrative pronoun; as "said city" and "said seisin" for "that city" and "that seisin." 2 This change, natural in itself, is of frequent occurrence; thus *hath* becomes *has*, *seeth* becomes *sees*, *doeth* becomes *does*, *speakest* becomes *speaks* and *sendeth* becomes *sends*. niversally be denominated adjective nouns. And this leads us to observe that there are two ways by which the relative qualifies the antecedent; the first, when the relative clause is merely descriptive of the antecedent; and the second, when it ascertains to what particular class or species the antecedent belongs. Thus, in the sentence, "The emperors, glorying in nothing but oppression and cruelty, issued many tyrannical edicts, which, as might be expected, fomented discontent and excited rebellion, instead of enforcing submission to their authority;" it must be obvious that the whole of the clause introduced by the relative is equivalent to a compound adjective descriptive of "tyrannical edicts;" and that if the word said were substituted instead of which (the noun "edicts" being understood), the sense and effect of the sentence would remain unaltered. Again, in the sentence, "homo sapiens caret, nil profuit hominibus," the relative cause quod sapientia caret, qualifies homo by ascertaining what particular class or species this "homo" belongs.

It though, in the former examples, it will readily be allowed that the word said may, with perfect propriety, supply the place of the relative pronoun, yet they who have not been accustomed to make the substitution may be prone to admit it in others, or to agree that the sentence "A man, who is destitute of wisdom, is a cipher in society," is exactly equivalent to "A man, said (man) is destitute of wisdom, is a cipher in society." Still the equivalence is undoubted; for the subject of the proposition, "a man," is, in both cases, qualified by words which have precisely the same effect in ascertaining to what class of men he belongs; and although, in the latter case, the relative clause assumes the form of a parenthetical sentence or interjected affirmation, yet if, in pronunciation, the word said be slurred over unemphatically, or articulated in the very same manner as the relative who, every one must be sensible that these two sentences are, in effect, identical. To prevent the appearance of violence in its substitution, it is necessary to repeat the antecedent along with said; but this arises solely from our not being accustomed to use said disjoined from its substantive, as we have long been accustomed to use the relative pronoun.

2. The second use of the relative pronoun is, when it intimates that the clause to which it is prefixed is to be regarded as the equation of a substantive noun; as in the sentence, "That you are disappointed, gives me pain," where the clause "that you are disappointed" is evidently the nominative to the verb "gives," and perfectly equivalent to "your disappointment." It is true that when the relative occupies this position, grammarians have been accustomed to call it a conjunction; but nothing, surely, can be more unphilosophical, or better calculated to produce endless perplexity and destroy the primitive simplicity of language, than to class the same word in different categories of speech according to its position and effect in a sentence. Such a mode of procedure, whilst it is but a clumsy method of screening ignorance, tends to retard the investigation into the real meaning of words, by deliberately substituting a technical name instead of a rational explanation; and it insidiously attempts to wrest from etymology the right of stamping upon terms their genuine signification. In opposition to the commonly received notion, however, it may safely be affirmed that, in known languages, the neuter of the relative pronoun is prefixed to a nominative and a verb, and that the sentence thus formed is, in amount, equivalent to a substantive noun. Thus, when Homer says, Ἀνερος τας Τευς εισηγησαι, "for you all perceive this at least, that my prize is going away to another," ὁ μετα τας εισηγησαι αλληλον is evidently the object or thing perceived by αλληλον; and hence, if this sentence could be expressed by a single term, it would be universally denominated a noun substantive, and would be in the accusative case governed by λαβομαι. The same observations are, mutatis mutandis, applicable to the following sentence:

where 'Ο είναι ἀνερος ἐπιστρέφων ἡμῖν ἀλληλον ἀποδέχεται is a complete substantive governed in the accusative by ἐπιστρέφων. Again, ἐπιστρέφων, the neuter of the relative pronoun ἐπιστρέφω (compounded of ἐπι- and στρέφω), is universally employed in the same manner. Thus, in the sentence, Ἀνερος τας Τευς εισηγησαι, ἐπιστρέφω τοῦτο ἡμῖν ἀλληλον, ὅτι ἐπιστρέφω τοῦτο ἡμῖν ἀλληλον, "even this alone, that you have taken your head safe and sound out of the mouth of a wolf," is sufficient reward for you," if ὅτι ἐπιστρέφω τοῦτο ἡμῖν ἀλληλον could be represented by a single word, every one would regard it as a substantive noun, and as the nominative to the verb ἐπιστρέφω.

In the Latin language, quod, the neuter of the relative, and ut, which is just the Greek ὅτι, are used in precisely the same manner. Thus, in the passage, "Hoc nunciantes orabant, ut opem ferret, hostemque haud dubium Romanis terra aut maritimis viribus arceret; qui ob nullam aliam causam, nisi quod imminerent Italiae, pateretur:" it is obvious that "ut opem ferret" and "ut hostem arceret" are compound substantives governed in the accusative by "orabant," and also that "quod imminerent" may be considered as a substantive under the regimen of ob, and expressed thus: "Qui (scilicet Oriconi) ob nullam aliam causam, nisi (ob) imminentiam Italiae, pateretur." It is evident, therefore, that quod, when thus prefixed to a nominative and a verb, is not a conjunction, but only the neuter of the relative pronoun; that when so used, the sentence is, in amount, equivalent to a substantive noun; that no real assertion is intended to be made by the relative clause; and that, on the contrary, the nominative and verb are merely an explanation of the relative which precedes them. In the expression, "I know that you are disappointed," that enunciates indefinitely the object known, and may apply to any object of knowledge whatsoever; but when the clause "you are disappointed" is subjoined, it limits the knowledge to a particular object, namely, your disappointment. It hence appears that, in making the words which explain that contain a real assertion, Mr Horne Tooke has admitted a fallacy into his ingenious resolution of the relative pronoun when thus employed; for, in such expressions as "I do not assert that you are disappointed," or, "I do not believe that you are disappointed," it would surely be absurd to maintain that I assert your disappointment, when I expressly say that I do not assert it, and do not even believe it. Upon the whole, then, it seems evident that, in this use of the relative, the pronoun itself is merely a general enunciation of the proposition which follows; and that this proposition being the equation of a substantive noun, or an explanation of the pronoun, is to be taken abstractedly, or, in other words, to be considered not as containing a formal assertion, but merely as something said or enunciated.

2. Of Comparison.

Comparison, comparatio, is derived from comparo, to put together, a word literally intimating the mode in which one thing is compared with another, namely, by collocation or juxtaposition; and hence, in all languages, this process is expressed in nearly the same manner. The degrees of comparison are commonly said to be three; the positive, comparative, and superlative. But the positive can scarcely be denominated a degree, as it simply states the epithet, and employs no comparison whatever; thus, albus, white, pulcher, fair. The comparative states a comparison between two objects only; the superlative between more than two. It is impossible, however, to express by different terms, or single words, the gradations or shades of difference which may exist between objects, as, for instance, in colours; and hence necessity obliged the framers of language to invent means by which these various gradations or qualities might be generally expressed. With respect to those words which admit of comparison, two things seem to be necessary; first, that they represent general ideas; and, secondly, that they represent objects which are capable of increase or decrease. In this view even nouns substantive, or generic terms, may be compared. Thus it may be said of a person possessing in an eminent degree those qualities which constitute a hero, that he is more a hero than another less liberally endowed in this respect; and, upon the same principle, we find in Greek ἀνδρεῖος, ἀνδρείτερος, ἀνδρειότερος. But the names of individuals, or, in other words, proper names, do not admit of comparison, because no individual can be another, and consequently no man can be more himself than another. The same observation extends to the personal pronouns, which, although in themselves general, do not admit of comparison, because they are particular in their application. Numbers, too, are the most general of all words; yet, as they do not admit of any addition or diminution, they are not susceptible of comparison; for, if we suppose two sets of any thing, each consisting of six, the one set cannot be more six than the other. Verbs, however, may admit of comparison, thus, amat magis, he loves more; and so also may participles as partaking of the nature of verbs. It is commonly objected, indeed, that the latter then become adjectives; but this is extremely incorrect; for, in such words as servans, servantior, amans, amantior; the meaning is not "loving," "more observant of love," but "loving at present," "more observant of love at present." Adverbs may also be compared; thus, convenienter, convenientissime, convenientissime; eleganter, elegantius, elegantissime. Prepositions, being a species of adjectives, and oblique cases of nouns, are likewise susceptible of comparison; thus we have in Greek ὑπὸ, ὑποστήριξις; now ὑποστήριξις; and, in English, before, further, foremost; hind, hindmost. Interjections, if they be signs significant of ideas or emotions of the mind, may, in like manner, be used in degrees of comparison. The repetition of the same word denotes a comparison, as alas, alas; and the expression ter quaterque beati, indicates a degree of comparison, perhaps the highest. Among the Hebrews the repetition of the same word, as good, good, is employed to express the superlative, and hence becomes synonymous with optimus, very good or best; and children make use of the same method to express the degrees of comparison. Of conjunctions nothing positive can be said in this respect, as their nature has not yet been satisfactorily ascertained.

But there is a much more refined and less obvious view of the subject of comparison, namely, that the comparative may be recomposed, and so also may the superlative; in other words, if language were fitted to express it, there may be an ascending or descending series of successive comparisons of comparatives and superlatives. Without attempting any thing of this sort, however, it may be shown that from one comparative another may be formed, and from one superlative another comparative and superlative, and so on to almost any extent whatever. Suppose three persons, John, James, and Peter: John is tall, James is taller, and Peter is tallest; but, instituting a second comparison, Peter is more taller than John than James is; that is, as Peter is the tallest of the three, and as James is taller than John, Peter, with reference to John, the common object, must be taller to a greater extent or degree than James, the intermediate subject of comparison, is. But this, which it is difficult to state intelligibly in words, may be better comprehended by the following simple diagram.

A ——— Here D is tall, and C is taller than D, B ——— but B is more taller than D than C is; C ——— and of the three, A, B, C, it is obvious D ——— that A is most the taller than D. In the first case, when we said that B is more taller than D than C is, we formed a new comparative of which the former comparative C is the positive; and, in the second case, we formed a new superlative, of which the former comparative was the positive. This may be visibly represented in the following manner:

Here we have three distinct groups or classes of objects, and A, B, and C are each the tallest of its respective class or group; yet these three in their relative excess may again be recomposed. Thus B may be more the tallest of its group than A is of its group, and of the three, A, B, and C, may be most the tallest of its group. This process, it is obvious, might be carried to any length, and we might form a triple, quadruple, quintuple, sextuple, &c. comparison; nor is there any logical incongruity in the expressions which we have employed, when these are properly applied and understood.

No limit is set to comparison in nature; the only boundary fixed is by our confined faculties. If we were to advance beyond the first, or at most the second step, we would become involved in relations so complicated that it would be impossible to conceive them; and convenience, therefore, obliges us to stop short. In English, we generally express such comparison by circumlocution. Thus, instead of saying that B is more greater than D than C is, we commonly say, that B exceeds D more than C does. But instances of the former are sometimes met with even in that language, and not unfrequently in Latin. "The Romans," says Mr Hume, in his History of England, "were more inferior to the Greeks in eloquence than in military glory;" that is, the Romans were inferior to the Greeks both in eloquence and in military glory; but they were more inferior in eloquence than they were inferior in military glory. The language of the historian, therefore, is both logically and grammatically correct, though its accuracy has been impeached by persons ignorant of the principle on which the comparison is made. Priscian the grammarian appears to have perfectly understood the rationale of such expressions; but Ruddiman, with all his ingenuity and erudition, seems to have been altogether in the dark respecting it, as is evident from the manner in which he speaks, in his Grammatica Major, of the instances of double comparisons quoted from Priscian. The resolution of all such examples, however, is, according to the principle here stated, equally certain and obvious. Thus, "Achilles Æneā magis fortior quam justior fuit;" that is, Achilles was more braver than Æneas, than he was juster than Æneas; in other words, Achilles exceeded Æneas both in bravery and in justice, but he exceeded him more in the one (bravery) than in the other (justice). The expression, therefore, is perfectly classical, and consistent with itself. Again, "Ajax est magis fortior Ulyssē quam Diomede;" that is, Achilles is braver than Ulysses, and he is also braver than Diomede, but he is more braver than Ulysses, than he is braver than Diomede, or, in other words, he exceeds Ulysses more in bravery than he does Diomede. There is a passage in the historian Florus, a writer of a vigorous imagination, and whose writings accordingly exemplify that characteristic, which, by reason of a double comparison, has been completely misunderstood. It is to be found in lib. iv. c. 46, of his history, and is as follows: "Sed nec minus admirabilior fuit illius exitus belli." In none of the editions of Florus is any notice taken of this passage; and upon consulting Clarke's translation it will be found that he has simply rendered minus admirabilior "less admirable," just as if it had been minus admirabilius. But it is not from the words themselves that we are enabled to make out the passage; it is from the context. Throughout the whole of it there is a train of closely-connected comparisons going on in the mind, and, in particular, three circumstances are mentioned, each having a tendency to create wonder in a greater degree, viz. 1st, "Nunquam ullo loco tantum virum populus Romanus vidit;" the forces were more numerous than at other times; 2d, "Nunquam imminentis ruinae manifestiora prodigia;" the presages of impending ruin were more striking than at other times; and, 3d, "Nunquam acerior neque alacrior exercitus Caesaris," the army was keener, and evinced greater alacrity, than on ordinary occasions. The circumstance, therefore, which led the author to use the expression "Sed nec minus admirabilior fuit illius exitus belli," was his having previously mentioned the three things first stated as more wonderful than at any other time. There is therefore a double comparison here, Sed nec minus admirabilior, &c. "But no less (or, in no less a degree) more wonderful was the issue of this war," viz. than the issue of other wars. The three particulars above stated were more wonderful than at any other time, and this was no less more wonderful. Without the double comparatives the real meaning of the author could not have been expressed. But for the same reason that it was "minus admirabilior," it might have been "magis admirabilior," more more wonderful. Magis, however, though used as a comparative, is in truth only a positive of which major is the comparative, and magissimus, i.e. maximus, the superlative; it is derived from the Greek μεγαλος, and literally signifies greatly; thus, "more elegant," was originally "greatly elegant." The comparative, therefore, produces a different effect in language from that which it did originally, when its import was not "more," but "greatly."

The intrepid apostle of the Gentiles, speaking of himself, in one of his epistles, says, ἐστιν ὁ διάβολος ἐπί τοῦ ἀρχέγονος, which is translated "less than the least," probably because "more the least" would have made but an awkward appearance in English. The superlative here is formed into a new comparative. But the Greek comparative ends in -τερος, which, when fully traced, seems to be τερος, the comparative of τος, one, joined to the word denoting the subject or thing compared, and signifies not more of one, but one of two; for we find the termination -τερος very frequently without any augment or increase of qualities, as τος, τοτερος, which of the two; τος, one, τοτερος, not more of one, or more of unity, but one of two; ἄλλος, another, ἄλλοτερος, not more of another, but another of two; ἐκ, who, ἐκτερος, which of the two, in Latin uter, abridged into uter; and, in the same manner, καλός, fair, καλότερος, the fair of the two. We may add, that whether, in English, means which of the two. Agreeably to this notion, then, the comparative being used as a partitive, major fratrium means the elder of the two brothers, or the one selected (on account of age) to the exclusion of the other.

Of the termination of the superlative no certain account can be given; it is a part of the subject on which all is conjectural; and although it could be traced, this would be attended with no practical advantage. Its Greek termination is in -τερος, and there is some coincidence between this termination and that of the ordinal numbers τρίτος, τετάρτος, πέμπτος, &c., which do not mean the aggregate numbers three, four, five, but express each an ordinal unity; thus τρίτος signifies the third, and often one of the three. Verbs. The same coincidence between the superlative and the numerals is observable in Latin; thus, optimus, &c., primus, quintus, &c., the first, the fifth, always denoting unity, that is, unity in a numerical series or order. Now, since in these two languages there is such a coincidence between the superlative and the numerals, let us for a moment suppose the superlative to be made up of unity, or to contain in it unity; then doctissimus Romanorum, resolved on this assumed principle, will be doctus unus Romanorum, "the one learned or learned one of the Romans," making thus a selection of one to the exclusion of all others. Unus is taken two ways, either definitely or indefinitely; either as one only to the exclusion of all others, or as one of any number. Horace has an example which seems to favour this view, and in which for unus potes, one that is powerful, we might substitute potissimus, the most powerful. And Quintilian says, "Demosthenes eminens unus in omni genere loquendi est," in which expression eminens unus is the equation of eminentissimus. The superlative, like unus, is used both definitely and indefinitely; definitely when optimus, for instance, means the best; and indefinitely when it is rendered very good. Horace, when inviting Meceenas to a feast, desires him to write back word, "Tu, quotus esse velis, scribe," where quoit, combined with unity, signifies "of how many you wish to be one;" and so perhaps the superlative is composed of unity and the word expressive of that which is susceptible of comparison, namely, the positive.

Lastly, the comparative is often expressed by quam, as "Virtus est praestantior quam aurum." But tam and quam correspond to each other as antecedent and relative, and are really such themselves; tam being the accusative of the Greek article το, τος, and quam the accusative of the relative, and their effect being tam, to the degree, quam, to which degree. Hence the expression may be resolved, "Virtue is generally excellent to the degree to which gold is," and more. According to the idiom of the French language, it would have been "quam aurum non est," to which degree gold is not. In certain cases of apparent comparison, it is said by many that potius and magis are understood, as "Res autem admonet cavere quam consultare," where it is commonly said that potius ought to be supplied; but surely this would be a very violent supplement, and the more natural one, undoubtedly, would be the antecedent of quam. Thus, "Res autem admonet tam cavere quam consultare," "Circumstances warn us to be on our guard to the degree to which they warn us to consult" or consider what is best to be done. The import of than in English, however, is not so easily ascertained as that of quam in Latin. Where potius and magis are used they seem to be of the same import with the Greek prepositions εἰς and ἐπί, and significant of apposition or juxtaposition. To in English denotes apposition and likewise comparison. "These blunders are nothing to what we find in Grenada," that is, compared with "what we find in Grenada."

Such is an abridged, and, we fear, very imperfect, view of the ingenious and original speculations of Dr Hunter on the subject of comparison; speculations with which all his pupils must be familiar, but which, notwithstanding their extreme refinement, have not hitherto, so far as we know, been submitted to the public, even in an intelligible abstract.

3. Analytical Resolution of the Moods and Tenses in the Greek and Latin Verbs.

The subject of the verb, and of verbal forms, has already been treated of at considerable length, but chiefly, as the reader will have perceived, with reference to the Eng- lish language. It may not be improper therefore to extend the application of the same principles to the resolution of the moods and tenses of the Greek and Latin verbs, and to establish, both analytically and inductively, their universality as laws of human speech, or rather of the human mind adapting language to the expression of its thoughts.

It occurred to Dr Hunter, early in life, that by separating the time from the other circumstances involved in those forms of the Latin verb called the tenses of the indicative and subjunctive moods, the explanations of the tenses, and also of these moods themselves, would be much simplified. The view which then occurred to him he continued to give annually, in his grammatical prelections, during a period of about half a century; and though, without calling in the aid of other principles, it is insufficient for the explanation of certain uses of the moods and tenses, yet, as far as it goes, it has contributed to the elucidation of a subject respecting which grammarians had, with a few exceptions, being long accustomed to wander in the dark. Mr Dawes, in his Miscellanea Critica, first published in 1743, reached the very verge of Dr Hunter's opinion; and Matthis, in his copious grammar of the Greek language, published in 1809, explained some passages of the Greek authors conformably to a general principle found exemplified in the structure of every language. But as he does not seem to have been at all aware of the principle itself, and of the universality of its application, Dr Hunter was induced, several years ago, to publish an introductory view of the moods and tenses of the Greek and Latin verbs, in which both the principle and its application are most clearly and beautifully unfolded, and which, with his consent and approbation, we are enabled to subjoin. Before entering on the discussion of the particular verbal forms, however, this venerable patriarch of philology premises some general observations connected with the subject, and necessary to a just view of it.

And, first, in the polished and refined state in which the Latin language has come down to us, it is perhaps impossible to decompose the various moods and tenses of its verbs, so as to exhibit with certainty the elements, or the generic verbs, by the coalescence or combination of which those moods and tenses have been originally formed. The surest, perhaps the only, guide, to which we can have recourse, is furnished by the decomposed state in which these moods and tenses are represented in our own or other modern languages, by means of those generic verbs termed auxiliary. Thus, scripsero, when decomposed into its simple elements, we represent, by recombining these elements in this manner, I shall—have—written.

Secondly, whosoever, therefore, shall attempt a complete and satisfactory explanation of the moods and tenses of the learned languages, which is still a desideratum in our grammars, must attend particularly to the literal import and meaning of those auxiliary verbs. But even this will not always be sufficient. He will often find that more is meant than meets the ear; and that it is further necessary to call to his aid our modes and habits of thinking, and the inferences which mankind naturally and habitually draw from the verbal statements given. In the expression I will write, all that is stated is my present will, inclination, or intention to write. There is nothing in it expressive of a future action; yet every person understands it as a declaration, or assertion, of future writing. Now, as this is not the amount of the statement literally taken, the effect must be the consequence of an inference drawn from the present volition to the future action. From the present volition the future action is inferred.

Thirdly, The verbs which our grammarians have termed auxiliary are chiefly these: am, do, have, shall, will, may, can. One or other of these we employ, or may employ, in rendering literally into English the moods and tenses of the Greek and Latin verb. These verbs, however, when combined with a noun or subject, constitute each of them, the predicate of a proposition. They all of them express an assertion, affirmative or negative, and consequently they are all verbs in the indicative mood; as I am, I am not; I do, I do not; I have, I have not; I shall, I shall not; I will, I will not; I may, I may not; I can, I cannot. From this view of the subject, some important consequences seem deducible.

1. These auxiliaries, being thus all verbs in the indicative mood, and all of them, when combined with a noun or subject, expressing an assertion, it follows that the tenses of the Latin subjunctive, or potential, or optative, as in certain instances it has been called, as well as the subjunctive and optative of the Greek verb, which involve those auxiliaries, and are literally rendered into English by means of them, are also indicative. To consider certain of those forms or tenses as exclusively indicative or assertive, and others as not so, seems to be founded completely in misapprehension, and tends to perplex and mislead. "I may write," "I might write," are as much assertions, as "I do write," and "I did write." The thing indicated or asserted is different; but, in as far as assertion is concerned, they are completely similar, and stand upon precisely the same footing.

2. Such expressions as, "I may read," "I may write," "I might read," "I might write," "I may have read," "I may have written," "I might have read," "I might have written," are incorrectly and improperly considered as moods, or modes, of reading and of writing. It is the verb may that is generic, or the term to be modified, and the other verbs annexed to it contain the modification. The verb may predicates liberty in general of its nominative or subject; may read, may write, predicate liberty to read and to write. They express not liberty in general, but liberty modified or particularized. They ought therefore to be considered as modes or moods of the verb may, not of the verbs read and write. The same observations are applicable to all our other auxiliaries. The verb can predicates power or ability; "can read," "can write," predicate particular power, the power or ability to perform those particular actions; and so of the rest. What has now been said of the English auxiliaries is applicable to the moods and tenses of the learned languages. If the generic verb be prefixed, as it is in our own language, it remains a separate word; and has been termed an auxiliary verb. If it be subjoined, or annexed at the end, being unemphatical, it becomes enclitic; and, adhering to the verbs to which it is thus annexed, comes at last to be considered as a flexion, or termination. The French future J'ai à or de porter, is an instance of the one process, and Je porter-ai is an instance of the other. And, in the progress of elision or abbreviation, the auxiliary may be so disguised as to be no longer distinguishable. It may even be doubted if Je porter-ois will now be recognised as a compound, or be considered to be the abbreviated form of Je porter-arvais, "I had to carry." Similar to these last examples appear to be the moods and tenses of the Greek and Latin verb. The auxiliaries, being annexed at the end, have come to adhere to the verbs to which they were annexed, and to be considered as merely changes of termination; and they are with difficulty, if at all, distinguishable as having been separate verbs. Yet the foregoing considerations and examples render it extremely probable that they were originally separate verbs, not prefixed, like the English auxiliaries, but subjoined.

---

1 See Miscellanea Critica, p. 65, Dr Burgess' edition. 2 Matthis's Greek Grammar, sect. 522. and at last attached to and incorporated with the words to which they were subjoined.

But, whatever opinion may be adopted as to the origin and history of these flexions, no doubt whatever can be entertained that they are similar in their nature to our auxiliaries; that the terminations contain the generic part of the expression, and are modified by the radical part of the verbs combined with them. The forms *scrib-am*, *leg-am*, *teg-am*, are therefore improperly considered as moods, or modes, of the verbs *scribo*, *lego*, *tego*. It is the termination -am, the common part, that represents I may, or I can, and which receives modification from *scrib-* leg-, teg-. The general notion of liberty or power is modified by the notion of writing, reading, and covering, with which it is combined; the termination -am, by the radical part of the verbs, not the radical part by the termination.

3. As the English auxiliary verbs determine the nature of the moods of the learned languages, and, as has already been seen, lead us to conclude, that, by whatever name they may have been distinguished, subjunctive, optative, or potential, they are all equally indicative or assertive; so they also, if we listen to their suggestions, will be found equally decisive as to the tenses or times of those moods. These auxiliaries are all verbs. They are all verbs of the indicative mood; and they are all verbs of the present time or tense, each of them having a corresponding form appropriated to past time. Thus,

**Pres. Am**, do, have, shall, will, may, can.

**Past Was**, did, had, should, would, might, could.

The terminations or flexions of the verbs of the learned languages, which contain or involve those auxiliaries, should therefore also be either of the present or of the past time or tense. Whatever they predicate of their subject or nominative, whether existence, action, possession, duty or obligation, volition, liberty or possibility, ability or power, they will be found, like the auxiliaries which they involve, to have two corresponding forms, one predicating those attributes of the present time, the other of the past. If then we separate the time from the other circumstances involved in those forms of the Greek and Latin verb called the moods and tenses, the time thus separated, or separately attended to, will, in every one of them, be found to be either the present or the past. One primary object of these forms seems to have been to give the present and the past as two fixed points of time, with a reference to which all the other ideas involved in them are to be estimated and determined. They are all past, present, future, contingent, &c. with reference to one or other of these two fixed points.

These preliminary observations seemed necessary to prepare the way for the proposed elementary view of the tenses of the Latin verb; and in the sequel that view will be further illustrated and confirmed by the application of it to the moods and tenses of the Greek verb. The tenses of the Latin verb, arranged as present and past, are as follow:

| Present | Corresponding Past | |---------|-------------------| | Indic. Pres. Scribo, | Scriberem, Indic. Imperf. | | Perf. Script, | Scripseram, Pluperf. | | Fut. Scribam, | Seriiberem, Subj. Imperf. |

Let us adopt the words now and then, as short and convenient expressions to indicate the present and the past. Scribo, the present, then, represents my writing as now doing, as now in progress; scriberam, the corresponding past, as then doing, then in progress. Scripsi, the perfect, represents it as now done, or past; scripsieram as then done or past. Scribam, whether considered as the future indicative, or as the present subjunctive, represents my writing in the one case as now future, as a thing that is to be; in the other as a thing that may be. Although there may appear to be a marked difference between these two when they are uttered separately, and not combined with another verb, yet, when they are found in combination, as generally happens, the effect of them is in most cases precisely the same. They equally present the verbal state as future, and equally present it as contingent or possible. Thus, Si jubebis or jubeas faciam, "If you order me, I will do it." Both forms represent my doing it as depending on your ordering me; and both, at the same time, mark the relation of that order to the present time; the one as an event now future, the other as an event now contingent or possible. Jubebis, however, in this combination, leaves it uncertain whether you shall order me or not; and jubeas leads to the inference, that, if you order me at all, that order is future. What the one states as future, is at the same time stated by si as uncertain; what the other states as uncertain, is at the same time perceived to be future. The two modes of conceiving the event are so blended together, that it is often a matter of indifference whether we employ the one tense or the other; and in numerous instances the present subjunctive may be rendered by shall. It is from this affinity or similarity between these two modes of conceiving the action or verbal state, that the terms expressing them often coincide in form. Thus scribam, legam, audiam, are, in all our elementary books, set down both as futures of the indicative, and rendered shall or will, and as presents of the subjunctive, and rendered may or can; and, whilst amemus is rendered may or can love, scripsieram of the same termination is rendered shall or will write. It is from the same affinity of signification that both these tenses always unite, in one past form, the imperfect of the subjunctive; as scriberem, which, as the past time of the future indicative, represents my writing as then future; and, as the past time of the present subjunctive, represents it as then contingent or possible, as a thing that might (may-ed) be.

There is, in like manner, so great a similarity of meaning in the forms scripsieram and scripsero, that, excepting the first person singular, they coincide completely, and always unite in one past form, scripsisse. Scripsieram represents, not my writing, but my having written (the act of writing being conceived, not as doing, but as done), as now contingent or possible, as a thing that may be, or, which amounts to the same thing, it states my writing as a thing that may have been. Scripsero states my having written, the completion of my writing, as now future, or as a thing that shall be, or it states my writing as a thing that shall have been. The former is accordingly rendered by

---

1 This tense Dr Clarke has, with great correctness of thought, called praesens perfectum, the time being really the present. There is, however, no incongruity in combining with this form an expression of the past time at which the action was performed; as, for example, in stating an action as now past, and at the same time as having taken place last year, or even last century; or in stating an event of last year as now past. The usage of the English language does not admit of combining with the perfect, expressed by have, any expression of time that excludes the present; at least we do not commonly say, "I have written the letter last week," but "I wrote the letter last week."

2 This form ought to be called praeteritum perfectum, in conformity with the view on which the tenses have received their denominations, the time being really past (praeteritum), and the action or verbal state represented as then finished (perfectum), and not in progress. Helvetii jam per angustias et fines Sequanorum suas copias transduerant, et in Eduovium fines pervenerant, comunque agere populumentum. (Cæsar in Bell. Gall. lib. i. c. 11.) The time of these three verbs is a past time, and in this instance they refer to the same point of past time; but the first two express what was then done or past, the third what was then doing. Mr Ruddiman, I may have written; the latter, I shall have written. Scripsissem expresses precisely the same conception with the two last, differing from them in no circumstance whatever but the time, and expressing the same thing with reference to past time which they express with reference to the present. It is the past time of scripsissem and scripsero. It states my having written as then contingent or possible; I might (may-ed) have written; or as then future, I should (scould-ed, Saxon) have written.

Let us apply this view to the various forms in combination with another verb.

And, first, of those which have respect to present time.

| Quae imperat, | se facturos pollicentur. | | --- | --- | | imperavit, | | imperabit, | | imperet, | | imperaverit (perf.) | | imperaverit (fut.) |

This set of examples are all presents, each of them containing in itself an expression of present time with reference to which all the other ideas involved in it are to be estimated. "They promise now, that they will do

\[ \begin{align*} \text{he now commands;} \\ \text{he now hath commanded;} \\ \text{he shall command;} \\ \text{he may command;} \\ \text{he may have commanded;} \\ \text{he shall have commanded.} \end{align*} \]

In this combination, imperat states the act of commanding as now present, at the time when they promise; imperavit states it as now past, imperabit as now future; imperet as now contingent or possible, as an event that may take place; imperaverit (perf.) states the having commanded as now contingent or possible; imperaverit (fut.) states the having commanded as now future.

It is necessary further to observe, that all these forms contain only the reference or relation of the command to the present time, and that all of them, excepting the first, leave the time of commanding, the time when the command is actually given, completely indefinite. Imperaret, for example, merely states the act of commanding as now past, but leaves the time of giving the command unascertained. The command may have been given last moment, last week, last month, last year, or last century. In the same manner, imperabit states the act of commanding as now future. It expresses barely the relation of the command to the present time, but leaves the time of commanding completely indefinite. It may be next moment, next day, next week, next year, next age. If the time of the command be necessary to be ascertained, an expression of future time, or the statement of a future event, with which it shall be contemporaneous, is required for the purpose; as to-morrow morning, next week, at sunset, or at sunrise, &c.; or it may be obvious from other circumstances in the context or narrative. In like manner, imperaverit states the having commanded as an event now possible, or now future; but of itself it indicates neither the time of giving the command, nor any time or event to which the command shall be prior, or before which it may or shall have taken place. From the desire, however, which we have to obtain precise rather than vague information, we naturally avail ourselves of the circumstances in the context or in the combination in which it stands; to ascertain what the form itself leaves undefined. In the example given above, imperaverit may be considered as standing in a twofold connection, and as having a double reference; one to the present time, the time of the promise (pollicentur), with respect to which the act of commanding is stated as contingent or future; the other to the future time of performing the promise (facturos), with respect to which the act of commanding shall be past. The whole statement may be fully brought into view by a paraphrastic or periphrastic development, thus: "They promise that they will execute his commands, when the act of commanding, now contingent, or now future, shall once be past, or when the command shall once have been given."

Secondly, Of those tenses which have reference to past time.

| Quae imperabat, | se facturos pollicerantur. | | --- | --- | | imperaverat, | | imperaret, | | imperavisset, |

This second set of examples are all past tenses, each of them containing in itself an expression of past time, with reference to which all the other ideas contained in it are to be estimated. "They promised then, that they would do

\[ \begin{align*} \text{he then commanded;} \\ \text{he had then commanded;} \\ \text{he should command;} \\ \text{he might command;} \\ \text{he might have commanded;} \\ \text{he should have commanded.} \end{align*} \]

In this combination imperabat exhibits the act of commanding, as then doing, having the same reference to past time that imperat (commands) has to the present; and therefore may justly be considered as the past time of imperat. Imperaverat represents the act of commanding as then past, and is therefore the past time of imperat, which represents it as now past. Imperaret represents the act of commanding either as then future, or as then contingent or possible. In the former case it may be considered as the past time of imperabit: imperabit stating the act of commanding as now future, or future at the present time; imperaret as then future, future at a past time, which in this instance is the time of the promise. They are therefore related to each other as present and past. On the other hand, if imperaret be considered as representing the act of commanding as then contingent or possible, a thing that might be, it will be the past time of imperet; the latter exhibiting the conception in reference to the present time, and the former exhibiting the same conception in relation to past time. The affinity of signification between the future indicative and the present subjunctive, as formerly observed, is so great, that their terminations often coincide; and hence they always unite in one past form, the imperfect of the subjunctive.

What increases the difficulty of this part of the subject

---

1 These two, like the present subjunctive and the future indicative, approximate so nearly in signification, that it is often a matter of indifference which of the two be employed.

Si senexero hodie quicquam in his te nuptiis Fallacie conari, quo flant minus; Te, Dave, in pistum sedem usque ad necem. En lege atque omine, ut, si te inde exemerim, ego pro te malam.

Ter. And. I. ii. 25.

In this passage the future and perfect subjunctive may change places without in the least varying the ultimate effect. My perception of your attempting deception is stated as now future, but to be past or to have taken place when I commit you to the grinding-house; and my releasing you from thence is stated as now contingent or possible, and consequently now future, but to be past when I grind in your stead. My perception is conceived to stand here in a double relation: one to the present time, with respect to which it is future, and the other to the commitment, with respect to which it is to be past. Both events are represented as future with respect to the present time, but the one as past with respect to the other. The same observation is applicable to exemerim and mislem. says that he will write," is incontrovertibly classical Greek, the expression having a reference to the present time. Let be changed into ἤρησεν the past, ἤρησεν, the indicative, must be changed into the corresponding optative ἤρησε. Ἐφαίνεται οὖν ἂγγέλων, "He said that he would write." The conception in both examples is exactly the same, except as to time; the former expressing it with reference to the present time, the latter with reference to the past. Should any doubt be entertained as to what has now been stated, the following passage from the second book of Xenophon's Hellenica will be found to contain as many examples of the optative of the future as may perhaps establish the point by induction. It is part of the defence made by Theran menes, one of the Thirty Tyrants, when he was impeached by Critias, another of them, because he would not concur in all their measures of cruelty and rapine.

"...'Εκεῖ τὸ εἶδος πήρανεν ὅμοιον ἴδομεν καὶ ἠγώνετο ἐπελάσσειν, ἐν τῶν γὰρ ἔργων, ἡμέτερον τοῦτον τοῦτον ἡρώεσθε. Ἄφεσιν γὰρ ἦν, ἀσυμμέτρητον μᾶς Ἀνίκους τοῦ Σαρπηνοῦ, ἤδη τοῦτο χειρὸν ἵππον ἄγειν, ἡμεῖς νύν μὲν ἐν τῷ ἐντοίνῳ ὑποχωρεῖσθε· ὁμοίως δ᾽ ἄλλων τῆς ἑταίρης, ἠγώνετο ἐπελάσσειν τοῦτο τοῦτον τοῦτον, ἕτερος ἢ Ἱππίππους. Ἡτοῖμα βέβαιον τῷ τοῦτο, ἀπεκεκλήτομεν ἰέναι ἴσως ἐπιστάν· ἀλλ᾽ ἡρώεσθε ἡμῶν τοῦτον τοῦτον."

In this passage every one of the indicatives is of the past time, and all the verbs subjunctive are optatives of the future, "ἦν ἥξειν etc.," I had seen, and consequently I knew, that when Leon of Salamis was put to death, the persons similar to him (ὁμοίων) would be afraid, and being afraid, οὔλιον would be hostile to this establishment of ours. I judged too (τὸ πράγμα) that when Niceratus, the son of Nicias, was apprehended, the persons in a similar situation with him would become (ὑποκαταλαβομενοι) ill-affected to you. And likewise, when Antiphon was put to death by you, who during the war furnished two swift-sailing triremes, I knew (ἐπέβολον) that those who had been prompt and eager to serve the state would be suspicious of you (ὑποστάτων). And when they said that each of us must seize one of the persons (ὁμοίων), I opposed the measure. For it was manifest (ἐξαίρετο) that, if these were put to death, all the persons (ὁμοίων) would be hostile (πολεμεῖν) to the establishment."

Let us now imagine that, instead of this narrative recital of his conduct and sentiments, Theronemenee deems it proper to urge the same considerations to his colleagues in the di- rect or dramatic form. Every one of his statements must now be made by verbs, not of the past, but of the present time; and the indicatives of the past being changed into indicatives of the present, the verbs subjunctive must also be changed from optatives of the future to indicatives of the future, that is, from the past to the present, so as no longer to express what was then future, but what is now future. Change ἠρώμενον the past into ἠρώμενον the present (I have seen, consequently)" I know that if Leon of Salamis be put to death, the persons similar to him (ὁμοίων) will be afraid, and being afraid, οὔλιον will be hostile," &c. In like man- ner, if ἐπιρρέοντο, ἑπικεχείρην, ἡρώεσθε, Ἃρόν, be changed into ὑπολαβόντο, ὑπολαβόντο, ὑπολαβόντο, ἡρώεσθε, Ἀρόν, οὔλιον, must also be changed into the corresponding indicatives, Ἢρόν κατέσχε, ἥρόν, and ἥρόν, which state these events as now future. The forms stating the events as future at a past time, will necessarily pass into those stat- ing them as future now or at the present time.

The first of the two positions may be illustrated from the same part of Xenophon's narrative, in which he relates in detail the proceedings of the Thirty, and the measures which they adopted to gratify their hatred and establish their power.

"Τούτων δὲ γεγράφησαν, ὡς ἐκεῖ ἦν τῶν παιώνων ἀνδρῶν, ἐν Βουλῇ ΟΙΝΤΟ, συλλήθησαν μετὰ ἑρμηνείας ἀπὸ ΑΙΤΕΡΕΝΤΙΟΝ, συνέλεξαν τοὺς Ἀχαιῶν ἐκ τῆς Ἑλλάδος ἀνδρῶν, ὅτι ἐν Βουλῇ ΟΙΝΤΟ καὶ τοῖς Ἀχαιοῖς ἐν τῷ Ἰόνιῳ ἤδη ἦσαν ἀνδρῶν ἀποστολῆς, τὰ ἐν Ἀχαιοῖς αὐτῶν ἀποστολῆς ἀποστολῆς ἐν τῷ Ἰόνιῳ ἤδη ἦσαν ἀνδρῶν ἀποστολῆς."

Here the whole texture of the narrative shows that the author is relating past events, the verbs being all of the past time, ἀποστολῆς, ἀποστολῆς, ἀποστολῆς, &c. The ends proposed to be accomplished should, therefore, agreeably to the first position, be subjoined in the optative. "There being nothing to prevent their doing whatever they might choose (ἐκλέγοντες), they put many to death out of hatred, and many on account of their money. And, that they might have (ἔχοντες) money to give to the guards, they thought it proper (ἰκανοῦντες) that each of them should seize one of the μαρτυροῦντες, and put them to death; and they ordered (ἐκλέγοντες) Theramenes likewise to seize whichever of them he might choose (ἐκλέγοντες)."

If the narrative be given in the present tense, the leading verbs will then be ἀποστολῆς, ἀποστολῆς, ἀποστολῆς, and all the optatives subjoined in dependence on them must be changed into the corresponding subjunctives; thus: "There being now nothing to prevent their doing whatever they may choose (ἐκλέγοντες), they put to death (ἀποστολῆς) many out of hatred, and many for the sake of their money. And, that they may have (ἔχοντες) money likewise to give to the guards, they think it proper (ἰκανοῦντες) that each of them shall seize one of the μαρτυροῦντες and put him to death. And they order (ἐκλέγοντες) Theramenes likewise to seize whichever of them he may choose (ἐκλέγοντες), &c."

In giving the reply of Theramenes to this requisition of his colleagues, the author makes a transition from the indirect or narrative form, to the direct or dramatic. He personates the speaker (Theramenes), and gives the ἰσχυρὰ verba supposed to have been spoken by him on the occasion. "It does not appear to me," said he, "to be honourable for those who profess themselves the best men of the state, to act more unjustly than the sycophants. For they (the sycophants) generally suffered those to live (ἰκανοῦντες) from whom they might receive money (ἔχοντες ἐκ τῆς Ἑλλάδος ἀνδρῶν); but we are going to put to death (ἰκανοῦντες ἐκ τῆς Ἑλλάδος ἀνδρῶν) persons guilty of no injustice, that we may receive their money (ἐκ τῆς Ἑλλάδος ἀνδρῶν)."

This reply of Theramenes deserves to be particularly attended to, on account of the transition from the past to the present which it contains, and the corresponding transition from the optative to the subjunctive. Had the author intended to exemplify the first of the above positions, he could not have expressed himself in terms more appropriate, or more completely adapted to the purpose.

From what has been stated, it is easy to perceive why those statements, which, in the direct or dramatic form, are given in the subjunctive or in the future indicative, are in the indirect form, or narrative recital of them, expressed by the optative, precisely in the same manner as in our own language we employ, in direct statements, the present forms may and shall, and in the narrative recital of those statements the past forms might and should. Thus, the direct or dramatic statements, "I may perhaps go to London next week," and "I shall perhaps go to London next week," become in the narrative recital of them, I said that I might perhaps go to London next week," and "I said that I should perhaps go to London next week." So, when Agamemnon had retired wounded from the field of battle, Hector, in addressing the Trojans and Lycians, employs the subjunctive, a present form:

"Προσέρχομαι ἐκ τῆς Ἑλλάδος ἀνδρῶν, ἐν Βουλῇ ΟΙΝΤΟ, συλλήθησαν μετὰ ἑρμηνείας ἀπὸ ΑΙΤΕΡΕΝΤΙΟΝ, συνέλεξαν τοὺς Ἀχαιῶν ἐκ τῆς Ἑλλάδος ἀνδρῶν, ὅτι ἐν Βουλῇ ΟΙΝΤΟ καὶ τοῖς Ἀχαιοῖς ἐν τῷ Ἰόνιῳ ἤδη ἦσαν ἀνδρῶν ἀποστολῆς, τὰ ἐν Ἀχαιοῖς αὐτῶν ἀποστολῆς ἀποστολῆς ἐν τῷ Ἰόνιῳ ἤδη ἦσαν ἀνδρῶν ἀποστολῆς."

Drive (now) your horses straight against the valiant Greeks, that you may bear away still greater glory." But, when the statement is given in the narrative form, as a past event, the subjunctive ἀποστολῆς is changed into the optative ἀποστολῆς, the corresponding form of the past time:

"Εἶπεν ἐκ Τούβης Διομέδῃς Παλλάδῃ Ἀθήνῃ ΔΟΚΕ πινεῖν μελ ἑρμηνείας, ἐν Βουλῇ ΟΙΝΤΟ, ἐν Ἀχαιοῖς ἐν τῷ Ἰόνιῳ ἤδη ἦσαν ἀνδρῶν ἀποστολῆς."

Then Minerva gave strength and courage to Diomede, that he might (may-ed) become eminently conspicuous, and might bear away distinguished glory."

It is to be noticed, however, that Homer does not always observe this, the usual and strictly grammatical, sequence of these moods. In the speech of Minerva to Diomede, we find a past tense followed by the subjunctive, "I removed the mist from your eyes that you may distinguish (ἴδοντες ἐκ τῆς Ἑλλάδος ἀνδρῶν) a deity from a man in the field of battle." But, in the indirect form, or narrative recital, of this speech of Minerva, given by Plato, it is ἔξωσα γνώσαις. His words are, "Homer says that Minerva removed the mist from the eyes of Diomede, that he might distinguish." But even in Homer the expression might have been, ἔξωσα γνώσαις, ἐκ τῆς Ἑλλάδος ἀνδρῶν, ἐν Βουλῇ ΟΙΝΤΟ, ἐν Ἀχαιοῖς ἐν τῷ Ἰόνιῳ ἤδη ἦσαν ἀνδρῶν ἀποστολῆς. I removed that you might distinguish. Wherein, it will be said, consists the difference? There is none in the ultimate effect. ἔξωσα would be used in reference to the time of removal, expressed by ἐκ τῆς Ἑλλάδος, a verb of the past time, to which it would correctly correspond. Homer has chosen, however, to change the reference, and to substitute the subjunctive, a form of the present time, thereby marking a reference to the time of Minerva's speaking, not to that of the removal. I (just now) removed that you may (now, or hereafter) distinguish. ἔξωσα states the removing as past, and may be used with equal propriety, whether the removal took place a moment or a century before. This appears to be a more correct view of the passage than that given by Dawes, who, to preserve a correspondence in the times, is obliged to make ἐκ τῆς Ἑλλάδος a preterperfect. Such changes of reference are, we believe, to be found in every language, and they are strikingly frequent in the Greek. A person speaking of himself might say in English, "I laboured hard in my youth, that I may enjoy ease and repose in the evening of life." These changes of reference too are often alternate, as, in historical narrative, a transition is often made from the past to the present time, and from the present to the past.

In the course of this discussion, it has been stated that all the tenses of the subjunctive mood are indicative or assertive. If that observation be admitted to be just, it will supersede the controversy, whether the futures in -ro belong to the indicative or the subjunctive mood. It has also been suggested, that what have been considered as the modes or moods of the Latin verb are incorrectly so conceived; that these flexions are either generic verbs themselves, or equivalent to generic verbs, indicating liberty, power, &c.; and that, instead of modifying the radical part

---

1 It is foreign to the object of these observations to enter on any discussion of the meaning of the word μαρτυροῦντες. It seems to mean dissolute strangers.

2 In Aleibid. II. propæ finem. These passages are quoted both by Dawes and by Matthise. of the verbs to which they are annexed, they are really modified by it. It has been shown, that all the tenses, both of the indicative and of the subjunctive mood, whatever they may further involve, are every one of them either of the present or of the past time; that they express of themselves one or other of these times, and may be classed in two sets, the one containing those forms which are of the present time, and the other those of the past time; that the present and past are two fixed points, with reference to which all the other ideas contained in those complex forms, called moods or tenses, are to be estimated; and that these other ideas are to be estimated as past, present, future, contingent, &c., with reference to these two fixed points.

This is the elementary view which it was proposed to give of the tenses of the Latin verb. It contains, as far as we are competent to judge, the only ground on which a complete and satisfactory account of these tenses, in all their various occurrences and applications, can be raised as a superstructure; and, if admitted, it will enable us to decide several grammatical controversies; so that, for example, between the acute and bold Sanctius and his learned editor Perizonius, concerning the tenses of the infinitive and the participles, in which the former maintained, that strictly speaking, time or tense does not belong to these, and that they are found of all times. It is obvious that the infinitive and participles do not, like the definitive forms, contain in themselves the two fixed points of estimation above mentioned; but, on the contrary, that their statements, when they are estimated at all, are to be estimated either from the verb with which they stand connected, or from some circumstance given in the context. They are always estimated from something extraneous, and not, like those of the definitive forms, from any fixed point of time which they themselves express. Thus scribere represents the act of writing as doing or in progress, but contains in itself no expression of the time when doing. The time may be any whatsoever, past, present, or future. It is generally, but not always, determined by that of the verb with which the infinitive is combined. Thus, in dico te scribere, dixi te scribere, dicam te scribere, the time of the writing is determined, not by the infinitive itself, but by that of dico, dixi, and dicam, the verbs with which it is here combined, and is different in each of these examples. The same observation is equally applicable to the participles, as is manifest from the following examples: video te scribentem, vidi te scribentem, cras video te scribentem. It has been said above, that the time of the infinitive is not always determined by that of the verb with which the infinitive stands combined. Dico te scribere may equally signify, I say that you were then writing, as that you are now writing; and dico te scripsisse may equally signify, I say that you had then written, as I say that you have now written, if any circumstance in the context require it to be so understood.

In further illustration of the subject, it has been stated, that through the whole of the Greek verb except the future, the subjunctive and corresponding optative differ only as present and past; consequently, if the subjunctive be rendered by may, or may have, the optative will be rendered by might, or might have; and when the subjunctive is rendered by shall, or shall have, as frequently happens, the optative will be rendered by should, or should have. It also appears that the future of the Greek verb may be considered as itself subjunctive, and of the present time, and the optative of the future as the corresponding past time.

These views the very learned and ingenious author has sought to establish by facts; and if the inductive conclusion appear to be founded on too limited a number of examples, the number of these may be increased by any person ad libitum. He proceeds:

The celebrated Matthaeus has attempted to explain the common use of the subjunctive mood, when combined with certain particles of time, by the hypothesis, that it then expresses an action frequently repeated at the present time. But meeting in Homer with these words, addressed by Andromache to Hector,

εἰς τὸν ἀνάγκην, ἵνα εὖ τοῦτο ἐπιστρέψῃ,

he is obliged to admit, that the subjunctive with these particles does not always express an action frequently repeated; and for this good reason, that a person can die but once. Yet that learned author does not seem to perceive that this single instance in Homer is fatal to his hypothesis.

Mr Seyer of the university of Oxford, in his work on the syntax of the Latin verb, has assumed, that, in certain cases, the subjunctive is strictly synonymous with the indicative, and has proposed to distinguish it in such cases by the appropriate name of the False Subjunctive. Nothing could have been devised more unphilosophical than this proposal, which, had it been successful, and generally adopted, would have effectually dashed all spirit of inquiry, and prevented all further investigation of the subject. Can it for a moment be supposed that an enlightened people, provided with two forms of the verb for the expression of two different and distinct conceptions, should reject the form appropriate to one of them, and employ in preference the other form, which did not correctly and adequately express the conception? If the learned author had seen the following example, of authority that will not be questioned, "I regret that it should have been found necessary to add to the burdens of the people," he might perhaps have been induced to pause a little, and to ask himself whether delicacy of feeling, arising from some peculiarity in the circumstances of the speaker, or in those of the party addressed, or in the occasion, may not have led him to employ the subjunctive form, should have been found, in preference to the indicative forms, has been found, and was found.

---

1 The Prince Regent's speech in proroguing the Parliament in July 1819. 2 We subjoin here Dr Hunter's conclusive reply to the objections stated by Professor Dunbar in his work on the Greek and Latin Languages. It is reprinted from the postscript to Dr H.'s Disquisition on the Latin and Greek Verbs. 3 Mr Dunbar, the learned professor of Greek in the university of Edinburgh, in a work entitled An Inquiry into the Structure and Affinity of the Greek and Latin Languages (p. 250), has made some animadversions on one part of the foregoing essay. I subjoin the passage in his own words:—I must here state that Dr Hunter's ideas respecting the subjunctive mood appear to me not well founded. He has derived his notions of it from the English auxiliary verbs may, can, &c., which he says are all indicative or assertive. This is no doubt true of these verbs when taken by themselves; but when joined with others, they evidently convey the notion of some condition or contingency, arising either from ability, inclination, duty, or obligation, in the same manner as those parts of Latin verbs that are subjoined to the tenses of the indicative, and give them a conditional meaning. It is surely not the same thing to say, I am writing, and I may or can write. The one implies the present act unconditionally; the other, a future act, and conditional. I may write if I please. I can write to him, if you wish it, &c. 4 I must observe, that there is nothing in language more subtle than the use of our auxiliary verbs, and of the verbal forms of the Greek and Latin which contain or involve them. They require more depth of thought, and more patient investigation, than yet appears to have been bestowed upon them. I cannot, therefore, at present enter on the subject further than to observe, that, as far as I am competent to judge, the strictures of my learned critic are founded in misapprehension, and completely erroneous. The sen- Of all the philological speculations of Dr Hunter, one of the most beautiful, perhaps, is that concerning the analogy observed in the formation of some of the tenses of the Greek verb, of which, under the modest title of "Conjectures," he published an outline in the ninth volume of the Transactions of the Royal Society of Edinburgh. The importance of this speculation, which, notwithstanding the heading given to it, is very far indeed from being merely "conjectural," consists in the advance it makes towards establishing juster rules for the formation of the tenses in some instances, and particularly in accounting for certain forms of the verb in Homer and Hesiod, which, being apparent violations of the usual analogy, have perplexed grammarians, and reduced them to the necessity of imagining new presents, without authority, and for no other reason but to account for such anomalous forms. The following is the clear and concise statement of his views on this matter, given by Dr Hunter himself.

1. The perfect middle has not a middle signification. (See Kuster De Voce Media, sect. i.) Its form, too, is active; and therefore it probably belongs to the active voice, which, on this supposition, becomes more regular and uniform, having a double preterite, as well as a double future and aorist. Thus,

\[ \text{τετάξω} \quad \text{τετάξα} \quad \text{τετάξω} \quad \text{τετάξα} \quad \text{τετάξω} \quad \text{τετάξα} \]

2. It is no objection to this account, that the second preterite may sometimes be found in a middle signification; for active verbs in all languages are sometimes used as middle, the reciprocal pronoun being conceived, though not expressed. Thus μαθησόμενος in Greek, accingo in Latin, prepare in English: 'omnes accingunt operi,' they all prepare (themselves) for the work.

3. In the second preterite, or preterite middle, uniformly arises from \( \tau \) in the present, and from nothing else:

\[ \text{λέγω}, \quad \text{λέγομαι} - \text{λέγω}, \quad \text{βλέπω}. \] \[ \text{λέγω}, \quad \text{λέγομαι} - \text{λέγω}, \quad \text{συνεχώ}. \]

That σύνεχος, μακρός, ἀγαθός, and other such verbs having a liquid before \( \alpha \), make ἑρέχω, μακρός, ἀγαθός, and not μακρός, ἀγαθός.

4. On the Analogy observed in the Formation of some Tenses of the Greek Verb.

\[ \text{ἐδίδασκα}, \quad \text{ἐδιδάσκα}, \quad \text{ἐδίδασκα}, \quad \text{ἐδιδάσκα}, \quad \text{ἐδίδασκα}, \quad \text{ἐδιδάσκα}, \]

is no exception from the rule. For all such verbs probably had of old not \( \epsilon \), but \( \alpha \), only, in the penult. Thus φέρω seems to have been of old φέρει, whence the old or Ἑολίκ future φέρει. So παρά was παρά, whence μαζεύω, παρά; and the old or Ἑολίκ form διαγένετο is still to be found. In like manner, βαλλόμενος seems, as Dr Moor has observed, to be formed, not from βαλλόμενος, but from the old verb βάλλω, whence βαλλόμενος, ἰακοῦμενος.

5. Although the first preterite, or perfect active, generally follows the analogy of the first future, yet it sometimes, too, observes that of the second. Thus,

\[ \text{Στελλάω} \quad \text{not ἱστορήσω, from στελλάω, 1. Fut.} \] \[ \text{but ἱστορήσω, from στελλάω, 2. Fut.} \]

In the same manner, τίθηναι, κτίζω, ἐχωρίζω, στεγάζω, &c.

6. On the other hand, although the second preterite, or perfect middle, generally follows the analogy of the second future, yet it sometimes, too, observes that of the first. Thus,

\[ \text{Στελλάω} \quad \text{not ἱστορήσω, from στελλάω, 2. Fut.} \] \[ \text{but ἱστορήσω, from στελλάω, 1. Fut.} \]

In the same manner, τίθηναι, κτίζω, ἐχωρίζω, στεγάζω, &c.

Nay, from \( \sigma \) in the penult, there often arises \( \sigma \) in the first preterite, or perfect active. Thus, στεγάζω, ἐλέγχω, ἐλέγχω, ἐλέγχω, from στεγάζω, λέγω, ἐλέγχω, ἐλέγχω, &c. And this seems of old to have been still more frequent, and to have extended also to the passive voice. Thus, μακρός, σύνεχος, in the old or Ἑολίκ dialect, from μακρός (i.e., μακρός) and σύνεχος.

7. In the same manner, although the first aorist generally follows the analogy of the first future, yet it also seems, particularly in the more early periods of the language, to have sometimes followed that of the second future. Thus, ξεῖν, ἤξει; so ἵστα, ἵππα, ὁράω, ἄλλαμα, ἴππα, &c.

8. Since, then, there are in the Greek verb two futures, and two preterites, each of which preterites follows the analogy of either future, and since there are two aorists, and that the first aorist also follows the analogy of either future; would it not, from these facts, be rather probable than otherwise, that the second aorist should also follow the analogy of either future; that, as from τῶν there is formed ἴππον, so from τῶν there should be formed ἴππον? And

---

tence, I may write, states my present liberty to write. It states that there is at present no external hindrance to prevent my writing. The words contain a proposition true and complete in itself, without further addition; and my learned critic is obliged to annex to it the conditional clause, if I please, to make it suit his purpose; not aware, it should seem, that it is the actual writing that is dependent on my pleasure, not the present liberty to write. In like manner, I can write, states my present power or ability to write; and this capacity exists independently of your wishes. My actual writing may depend on your wish, but not my ability to write. I can write whether you wish it or not. The freedom from external preventive and the inherent power or capacity frequently exist together; but either may exist without the other. Ἐξερευνάμενος ἂν ἦν ἢ ἔχοντας ἢ τὸ γεγραμμένον. There is from without no hindrance to my writing, but I am unable to do it because of the gout in my hand. Ἐξερευνάμενος ἂν ἦν ἢ ἔχοντας ἢ τὸ γεγραμμένον, but not the ἄγχος.

My learned critic, under a misapprehension for which it is difficult to account, has found fault with my calling the future in the future of the subjunctive, imagining, from my employing that designation of it, that I intended to exclude that future from the indicative mood. I must observe, however, that although, to prevent being misunderstood, I was obliged to call the future in the name by which it was generally known, and by which it is distinguished in our elementary books; yet the general tenor of my essay, and even the following paragraph of it alone, ought to have prevented all possibility of misapprehension on the subject: 'In the course of this discussion it has been stated that all the tenses of the subjunctive mood are indicative or assertive. If that observation be admitted to be just, it will supersede the controversy, whether the futures in -ro belong to the indicative or to the subjunctive mood.' P. 14.

I have stated (p. 4) that 'one primary object of these forms called the moods and tenses seems to have been to give the present and the past as two fixed points of time, with reference to which all the other ideas involved in them are to be estimated and determined. They are all past, present, future, contingent, &c., with reference to one or other of these two fixed points.' And I have there given a scheme of the tenses of the Latin verb arranged, agreeably to that principle, as present and past. This unfortunate future in -ro, however, Professor Dunbar has, in exhibiting that scheme, studiously transferred from the class of presents to that of pasts, and has thereby marred instead of mending the consistency of the scheme of arrangement. For it is quite evident that, as scribamus states my writing as now future, and therefore, as he admits, is properly arranged as a present; so, for the same reason, scripsimus, which states my having written, or the completion of my writing, as now future, must also be arranged in the class of presents; the completion of my writing being future relatively not to the past, but to the present time.'

We have only to add, that when Professor Dunbar says, 'It is surely not the same thing to say, I am writing, and I may or can write,' he not only misunderstands Dr Hunter, but confounds two things not very difficult to distinguish; namely, assertion generally, and the thing asserted or indicated in particular verbal forms. Dr Hunter never said or imagined that, 'I am writing,' and 'I may or can write,' are identical propositions. What he really stated, and has always maintained, is, that 'I may write,' 'I might write,' are equally assertions as 'I do write,' 'I did write.' The thing asserted or indicated is different; but, in as far as assertion is concerned, these expressions are completely similar, and stand upon precisely the same footing. if this were admitted, would it not account for such words as καταβάλλω, ἰδούσαι, λέγειν, ἐρῶ, ὄντος, ἔστω, ἀξίω, ἀξίων, ἀξίων, &c., which have so much perplexed the celebrated Dr Clarke (see Notes on Iliad, B. 35. E. 109; I. 618), in a manner more satisfactory than is done by the arbitrary method of imagining new presents in -εῖν, from which these forms may be deduced? For it is only to account for such anomalous forms that these new presents have been imagined; and the grammarians would no doubt have assumed imaginary new presents also, to account for such first aorists as ἤξει, ἤξει, if, by so doing, these aorists could have been reduced to the usual analogy.

"Now this supposition, namely, that the second aorist, as well as the first, follows the analogy of either future, which is thus rendered in some degree probable, has actually taken place, at least in some instances. Thus the compounds of τίμω,

(from the first future ἀποτιμᾶν) has in

the second norist part. ἀποτιμᾶν, as well

(as from the second fut. ἀποτιμᾶν, ἀποτιμᾶν).

So ἤξει, they came, II. 23. v. 38; καταβάλλει, II. 24. v. 191, both of which follow the analogy of the first futures, viz. ἤξει (from ἤξει, ἤξει) and καταβάλλει. And every person acquainted with the elements of Greek is aware, that, as ἤξει is the primary form, whence, by the common reduplication, arises, first, ἤξει, and then, dropping the short i, by what the grammarians call syncope, ἤξει, and as ἤξει is the primary form, whence, in like manner, is derived γίγνομαι (Latin gigno) and γίγνομαι; so ἤξει is the primary form, which, by like reduplication and syncope, gave birth to ἤξει, cado, from which primary form ἤξει, making in the first future ἤξει, there is formed the second aorist ἤξει, in common use; which, in so far, both illustrates and confirms the foregoing hypothesis."

These observations were written about thirty years anterior to the date of their publication in the Edinburgh Transactions (1823), and, of course, before either Villoison's or Heyné's editions of the Iliad were published. Subsequently to their appearance, Dr Hunter judged it proper to subjoin some facts contained in these two celebrated editions, together with a few observations suggested by these facts, and tending to confirm the opinion which he had been led to form of the anomalous flexions above mentioned of the Greek verb. The additional illustrations, derived from these sources, together with the critical remarks by which they are accompanied, we shall now place before our readers. (See Edinburgh Transactions, ubi supra.)

"From the ancient scholia, quoted by Heyné in his observations on the various readings of the Iliad, and from the large volume of scholia published by Villoison (as well as from those contained in Barnes's edition of the poems of Homer, published in 1711), it seems to be clearly established, that, in the days of the Ptolemys, this form of the second aorist, following the analogy of the first future, was more frequent in the writings of Homer than that of the first aorist, formed upon the analogy of the same future; and that it was the reading received and approved by the most eminent of the Alexandrian grammarians, by Zenodotus, the keeper of the Alexandrian library, and by Aristarchus himself.

"Whilest, however, the Greek scholiasts have recorded these as the ancient and genuine forms used by Homer, they have, I suspect, greatly erred in their manner of accounting for them. After the lapse of so many centuries, these Homeric forms had, in a great measure, become obsolete, and the other form, that of the first aorist, had come into general use; and those grammarians could devise no other method of accounting for these old and now obsolete forms, but by imagining new presents in -εῖν, of which they supposed these forms to be the regular imperfects. This manner of solving the difficulty has been implicitly followed by Eustathius, in his Greek Commentary on the Poems of Homer, and (I am sorry to be obliged to add) by my late invaluable friend Lord Monboddo; of whom, and of the dies Attici, now long gone by, in which I enjoyed his learned and elegant society, I never can think but with an indescribable feeling of satisfaction and regret.

"If, however, we acquiesce in this solution of the difficulty, we must be content to abate somewhat of our admiration of the Greek language, which has been heretofore held to express, with the utmost clearness and precision, all the various and diversified conceptions of the human mind, even to their minutest shades of difference. Indeed, that the Greeks, or any enlightened people, should employ those forms of their verbs as presents, which were familiar to every ear as appropriate expressions of the future, or, as Eustathius often expresses it, should draw back the future to express the present, is, when duly considered, a supposition in itself so unreasonable, not to say absurd, that it would require better and more conclusive evidence to support it, than has yet been produced. Besides, where was the necessity of such an innovation? what advantage was to be gained by it? The original presents remained in common use. Why then employ the futures, ἤξει and ἤξει, for example, as presents, when ἤξει and ἤξει, and ἤξει and ἤξει, present forms in common use, would answer the same purpose?

"There is still another consideration, to which, in a question like this, due weight ought to be given. There is no evidence whatever that those present-futures, or future-presents, in -εῖν and -εῖν, ever existed. They are not to be found in the poems of Homer or Hesiod. They are not to be found in the Greek authors posterior to Homer and Hesiod; and of Greek authors prior to the age of Homer and Hesiod we know absolutely nothing.

"From these considerations, it seems to be evident that the existence of these future-presents is merely a gratuitous assumption of grammarians, to account for the Homeric forms ἤξει, ἤξει, &c., which they could in no other way reduce to any known analogy of formation.

"This, however, the learned Dr Clarke had too much acuteness and good sense to admit; and, accordingly, we find him regarding these presents as altogether fictitious. His words on the passage 'ἐν ἀγαθῷ ἀναβαίνει, ἀναβαίνει,' are these: 'Editi plurimi habent αναβαίνει, a verbo fiecto ἀναβαίνει.' And, to obviate the difficulty, he changes αναβαίνει here into αναβαίνει; and, when αναβαίνει, or any of its compounds, occurs, his general practice is to substitute the form of the first aorist, ἀναβαίνει. But, even if it would remove the difficulty, this change is unauthorized. The other form in -εῖν is attested as genuine by the Alexandrian grammarians, is supported by the Greek scholiasts, by the ancient manuscripts of the highest authority, and by the early editions. In Villoison's edition of the Iliad, which bears to be an exact transcript of a manuscript of the tenth century, in the library of St Mark at Venice, the form is uniformly -εῖν, except in a single instance, and in that one instance I suspect -εῖν to be an error of the press.

"Dr Clarke, however, soon finds that this change of -εῖν into -εῖν, violent and unauthorized as it is, will not remove the difficulty. In the address of Diomedes to Sthenelus, he meets with the imperative, καταβάλλει, which he attempts to account for in this manner: 'Vide- tur mihi imperativus ex futuro deductus, licet id non agnos- cant grammatici; errareque eos, qui verbum hic in pre- senti fingunt xaraSvou. Quanquam, analogia haud dis- simili perspexi usurpare visum est Homerus verbum bavou, boou, boou, &c. Nothing can show more strongly the difficulty which Dr Clarke must have felt in accounting for these forms, than that he is here obliged to have recourse to another gratuitous assumption of an imperative of the future, a thing unheard of before; and almost to admit bavou as a present, from which to deduce boou as an im- perfect. The fact is, that he found the authorities in fa- vour of boou and boou to be so numerous, and of such weight, that he did not venture, in these examples, to change -iv- into -ov.

"Afterwards, however, he found himself obliged to aban- don this hypothesis of an imperative of the future, as well as his original position, that these new presents in -ov and -ovou are imaginary; and he appears at last (he or his son, it is uncertain which) to adopt, in its full extent, the explanation given by Eustathius. That commentator, whose remarks would have been invaluable had his taste and judgment been equal to his means of information, if I rightly understand him, has this observation on v. 154 of the 20th book of the Odyssey: 'Oouou is sanctioned by a poetic rule of formation (xovouvou xovouvou) as the im- perfect of the verb boou, the future having retrograded into a present. For, of boou, the imperfect is boou, the third person boou, the imperative the same in sound, also boou, of which the plural is boou.' Dr Clarke or his son adds in continuation: 'Atque analogia quidem haud dissimili sepe usurpare videtur. poetar verba boouou, boouou, boouou, &c. Vide autem ad Il. v. 109' (where, however, he reproaches the hypothesis of these new presents in -ovou). 'Occurrit eadem vox (viz. boou) Il. v. 103, 6. 718, et apud Theocrimum, Idyll. xxiv. 48. Porro fuerunt, teste Barne- sio, qui boou ubique legendum statuerunt, sed nullo pro- fectu. Nam, uti ipse annotat, occurrit apud Homerum boou in imperativo, numero singulari, ubi isti emendationi locus nullus est, &c. et apud Callimachum, sive erat boou, boou. Hymn. in Cer. v. 137.'

"All the attempts, however, to account for these forms in particular verbs are partial and unsatisfactory. As the facts to be accounted for are all similar, the true solution must be general, and such as will embrace them all. Whe- ther the method which I have ventured to suggest be of this description, does not belong to me to determine. It possesses at least this indispensable requisite, that it ac- counts for all the phenomena in question.

"It would be tiresome to enter upon a particular examina- tion of every individual verb of the kind which we are con- sidering. It may not be improper, however, to select one example for the purpose of illustration.

"There are not many verbs of more frequent occurrence in Homer than boou, boou, in one or other of its parts. It is found in the present:

Enov (e Cypro) boou houos ovv 'Iko, xovouvou xovouvou. Odys. v. 444.

and the future, boouou, occurs once in Hesiod, and ten or twelve times in Homer; and, in not one of these instances, can it by any sophistry be tortured into a present, or forced into an expression of present time, in the sense in which present time is generally understood. I shall note the places of occurrence, that any person may, if he chooses, examine them, and judge for himself. They are these: Hesiod. Op. et Dies, v. 477. Il. v. 240; 7. 363, 502; v. 182; o. 505; v. 47; o. 728. Od. v. 515; v. 198; v. 276.

"boouou, the future of boou, then, must be decidedly ex- cluded from the catalogue of new presents, devised by the Greek scholiasts to account for the supposed imperfects, so frequent in Homer. And yet a new present, boouou, is equally necessary in this, as in any other verb of the kind, to account for boouou, they came, the past tense, formed upon the analogy of the first future; which, though not noticed either by Dr Clarke or by Heyne, is no less frequent in Homer than the future itself. Thus,

'Αλλ' ἐν τῷ Τροικῷ ἀναπαύεται ἡ πόλις. Ἡλ. i. 773. At quando jam Trojani VENERUNT, fluviorum labentes.

See also Il. x. 470; v. 433; v. 1; o. 692. Od. v. 5, 31, 495; v. 1; v. 194. Hymn. in Apoll. 411, 438, 230, 278.

"Nor is there a single unequivocal example of the first aorist of this verb to be found in the poems either of Ho- mer or of Hesiod, except one, viz. boouou, which occurs in the Hymn to Apollo, v. 223. In that same hymn, how- ever, there are no fewer than four examples of the form in -ov (boouou), and another still in the Hymn to Mercury, v. 398.

"This instance of boouou, so frequent in Homer, may there- fore be fairly admitted to be, in the language of the cele- brated Bacon, an instantia crucis, clearly pointing out the true method of accounting for all the other Homeric forms of the same description. If so, there will no longer be any necessity for the gratuitous assumption of new presents in -boou and -ovou, to account for these forms; nor for suppos- ing, with Dr Clarke, that xaraSvou is an imperative of the future; nor for changing, as he has generally done, the second aorists boouou, xaraSvouou, &c. into regular first aorists in -ovou. We ought rather to restore the Homeric second aorists in -ovou, recognised as the genuine read- ings by the grammarians of Alexandria, and the Greek scholiasts, and confirmed by the ancient manuscripts of highest authority, and the early editions.

"If the view which I have ventured to give of these Homeric forms in -ovou be well founded, it will hardly be necessary to notice the canon of criticism attempted to be established by the learned Heyne, in his luminous ed- ition of the Iliad, as to -ovou and -ovou, and the reasons for preferring the one to the other in particular passages. Of the futility of that attempt every classical scholar who will read with sufficient attention the last seventeen verses of the third book of the Odyssey, will be fully convinced. The learned German has adopted, without scruple or hesi- tation, the explanation of these Homeric forms given by the Greek scholiasts; and, in allusion, I presume, to the difficulty felt by Dr Clarke, in admitting the existence of new presents in -ovou and -ovou, he states his opinion, and the grounds of it, with decisive confidence, in the follow- ing words: 'Ese autem talem formam praesentis boouou dubitari nequit; nam est v. c. O. 105. Αλλ' ἐν τῷ Τροικῷ ἀναπαύεται ἡ πόλις,' p. 506; v. 4. And on B. v. 35, he makes the following remark, not quite consistent with itself: 'Quid hoc loco interstit, non video; nisi quod aeterno antiquo- ris vocis speciem habet: nova forma subnatu, ut boouou, boouou, et alia.' How the nova forma should have anti- quioris vocis speciem, I do not clearly comprehend. But a gleam of truth will sometimes burst through the darkness of inveterate prejudice."

These views, which to most persons will probably ap- pear to carry their own evidence along with them, have, however, had the misfortune to incur the disapprobation and censure of Professor Dunbar, who, in his very meri- torious work on the Greek and Latin languages (p. 188, et seqq. Edinburgh, 1827, in 8vo), has treated "one of the most acute scholars and grammarians of the present day" (these are his own words) with far less consideration than might have been expected from such an admission, and who contradicts speculations which it is impossible to refute, in a peremptory tone, neither called for by the subject itself, nor much in keeping with the courtesy hat usually obtains amongst men of letters. It may not improvement unrestricted, therefore, to show, by a few examples, that the critical strictures of Professor Dunbar are as deficient in point of accuracy, as his mode of expressing himself is, we are sorry to say, wanting in that respect which is due to the venerable age and to the great attainments of "the most acute scholar and grammarians" of his time.

1. After citing the greater part of the fifth paragraph of the above quotation, Mr Dunbar observes, "Dr Hunter then gives as an example the compound verb διαγραφώ from the first future of which, says he, comes the second aorist participle διαγράθη, as well as from the second future διαγράσω, ἵστιμεν." Where Dr Hunter found the second aorist participle διαγράθη, I do not know, as I never observed it; the other form, διαγράσω, occurs often in the Greek authors." Now, it will not be very difficult to show where Dr Hunter found διαγράσω, or at least sufficient authority for employing it, and where Mr Dunbar might have made the discovery. In the Cyclopedia of Xenophon (lib. v. c. 2, l. 9), there is the following passage:

«Ἀλλὰ τοὐ μὲν διαχώρα, ὅτι ἐκεῖνος ἑδήποτε πρὸς εἶον, πωλεῖεν γείτην, ἐκεῖνος ἄλλω» ERETEMENIT, ὅće παν πρὸς ὁράνη n., r.» For a reason which will be obvious to every scholar, we abstain from translating this passage, in which Gorbyra describes the treatment to which a certain ancient lover of surgery was in the habit of subjecting some of his boon companions. But as we have here τὸ, the particle of which is ἀντίστην, it is difficult to conceive why Mr Dunbar should proscribe διαγράσω, or at least διαγράσωλα, which only differs from the word employed by Xenophon in having the preposition αὐτό instead of the preposition ἓι or ἐι prefixed to it. Ὀσυσίον means to cut away or off; ὠσυσίον signifies to cut out, or, in other words, indicates a somewhat more radical and complete form of the operation. If it be good Greek to use such a form of the verb as ῠσυσίον, it can scarcely, we should imagine, be bad Greek to employ διαγράσωλα, the same verb, in the same sense, varied only in its sense by the preposition prefixed.

2. "I am, however, decidedly of opinion," continues Mr Dunbar, "that no nor aorist was ever formed from a second future; and in the examples which the doctor has given, 'he has assumed what never existed'." Here again the learned professor cuts a little before the point; for the assumption on which Dr Hunter has grounded his ingenious solution of these apparent anomalies, so far from being gratuitous and unauthorized, as Mr Dunbar asserts, is distinctly countenanced by Apollonius of Alexandria, whom Priscian calls summus auctor artis grammaticae, and Mr Harris, in his Hermes, describes as "the most acute author that ever wrote on grammar." In his treatise De Syntaxi (lib. i. c. 36, p. 76), this learned and acute Grammarian, speaking of imperatives, says,

"Quod est hisc verba non ridiculi existimentur, qui negant esse verba imperativa futuri temporis, cum omnia cognatur ad futuri sensum? nam ea imperamus, quae aut non sunt, aut non facta sunt. Quae autem non sunt, aut non facta sunt, et tamen ut singli apti nata sint, ea futuri temporis sunt: siquidem et ea que impera non sunt, sive si quis imperat; si vero, cum negant participia... Ergo ex futuri sensu imperamus ut fiat aliquid, aut non fiat, additione negato participi, imperfecte quidem, ut exacteri sim deloque; perfecte vero, ut exacteri sim deloque. Quid autem decept eo quis点缀riveret futuri imperativi? Γεγκλιπ γεγεγκγεγκπ-too; ἐν τῷ λοιπῷ᾽ ἦτο διαγράσωλα: qua quanquam absolutus a nihilis in libro De Imperativis diluntur, tamen quia locus facit hic idem exigimus, paucis quadam adjungamus. Nimirum illud γεγκλιπ γεγεγκγεγκπ τῶν poetaic licentia per exitium usurpata pra voces γεγκλιπ γεγέγραπται: non diverso tempero sed directo verbis, quod, utra voce velis, recte preferitur: quemadmodum in indicativo modo quoestio ejusmodi ante orta est, in γεγκλιπ γεγεγκγεγκπ nonque τῷ γεγεγκγεγκπ-τι, inquam, sed γεγεγκγεγκπ, quandoquidem in istis nonnullis disciplinis differim nonse, sed tantum in scriptura; quae assumpti potest non sine ratione probabililies sic demonstratus in capitio De Verbis."

Grammar.

Verba, bonus in subilo De Imperativis dulsiamur, tamen quia locum facit hic idem exigimus, paucis quartam adjungimus. Nimirum illud γεγκλιπ γεγεγκγεγκπ τῶν poetaic licentia per exitium usuclipta prae voces γεγκλιπ γεγέγραπται: non diverso tempero sed directo verbis, quod, utra voce velis, recte preferitur: quemadmodum in indicato modo participatedae ante orta est, in γεγκλιπ γεγεγκγεγκπ nonque τῷ γεγεγκγεγκπ-τι, inquam, sed γεγεγκγεγκπ, quandoquidem in istis nonnullis disciplinis differim nonse, sed tantum in scriptura; quae assumpti potest non sine ratione probabilileges sic demonstratus in capitio De Verbis." such verbs were ever employed as presents? I hope to be able to prove, both from examples and analogy, that the old established opinion, viz. that the futures of certain verbs, whose original presents had become obsolete, were sometimes used as presents, and that those forms which the doctor supposes to be second aorists, were imperfects, used, however, as aorists, i.e. with indefinite time." We shall have occasion immediately to examine some of the "examples" and "analogies" by means of which the learned professor labours to establish his position. It will be sufficient here to advert to the confusion of ideas which reigns in the passage we have just cited. How, says Dr Hunter, can you account for such forms as ἀναβάσηνος, ἰδὼν, ἰὼν, ἰὼν, &c., and the like, if they are not to be considered as second aorists formed on the analogy of the first futures of the verbs to which they belong? Perfectly, responds Mr Dunbar; they are neither first nor second aorists (Apollonius, as we have just seen, thought otherwise), but only imperfects, "used, however, as aorists," and performing all the functions of these particular verbal forms; that is, they are not aorists, but imperfects, which imperfects, however, are in reality aorists. Nor is even this all. The learned professor, misled apparently by the literal meaning of the term, is betrayed into an entire misconception in regard to the time indicated by the aorist, which he considers as "indefinite." He could not easily have committed a greater mistake, as has already been shown in treating of the verb. The time of the aorist, strictly considered, is absolute past, and if it ever appear to be "indefinite," as it unquestionably in some instances does, that arises, not from any vagueness or uncertainty in the tense itself, but from an inference drawn by the mind from those particular predications which it is employed to express. Thus, ἐγὼ ἤρξα τὸν πόλεμον εὐθὺς διαλύσω, "A short time dissolved (or did dissolve) the associations of bad men." This is all that the words actually express. But, as we know from experience that the laws of nature, moral as well as physical, are uniform, we infer that what occurred formerly, happens now, and will take place hereafter; and hence, by an easy and gradual coalition of the secondary inference with the primary and invariable signification of the verb, we come at length to say that "a short time always dissolves the associations of the wicked." The expression in Scripture, ὁ ὑμῶν ἐστιν ὁ ἀγαπητός μου ἐν ἡλικίᾳ, rightly translated in our version, "This is my beloved Son, in whom I am always well pleased," and all others of a similar description, are to be resolved and explained upon precisely the same principle.

4. In rejecting Dr Hunter's solution, which proceeds upon ascertained and determinate data, Professor Dunbar adopts one which he admits to be purely imaginary. "I believe," says he, "the present of ἵνα is nowhere to be found in the language, because it must have been the same in appearance and sound as εἰς, πρὸς: ἵνα is commonly used instead of it. But Pindar, in his fourth Pythian Ode (v. 181), employs εἰς as a present.

"I say that I obtain instruction from Chiron," &c. In this passage Professor Dunbar appears to have committed several mistakes. Without adverting to his contradictory expressions in regard to εἰς, we may observe, in the first place, that if Pindar had really employed εἰς here as a present, the language which he puts into the mouth of Jason would, upon that supposition, have conveyed a statement contrary to the fact. Jason, it is true, was educated by Chiron; but long before the period referred to in this passage, his education had been concluded; and there is nothing said to warrant the assumption that, on his return from the Argonautic expedition, which is celebrated in this ode, he had resumed his studies under the Centaur, and was, at the time of speaking, actually obtaining or receiving "instruction from Chiron." He had no doubt, according to his own statement, been to visit that famous personage; but his object, as we shall presently show, was not for the purpose of being taught, instructed, or educated by Chiron (that had been accomplished long before), but to obtain information from the Centaur to enable him to make good his claims, which, it seems, had been disputed. Adjured to speak the truth, Jason accordingly says, "I will relate what I have learned (ἐξερευνήσας) from Chiron; for I have just come from the cave near Charilo and Philyre:" and, in conformity with this declaration, he proceeds to state the substance of the information he had obtained from the Centaur respecting his rights. In what follows there is not a word said respecting "instruction" or education. The narrative of Jason relates to a subject of more immediate interest to him, one indeed in regard to which he had engaged that he would produce (εἰς) what he had learned (ἐξερευνήσας) from Chiron during the visit to his cave, and tell nothing except on the authority of his old master, or in accordance with the information which he had received from him. In our view, therefore, the common interpretation of this passage, Institutus sum a Chirone, is erroneous, notwithstanding it has been adopted by so distinguished a scholar as Heyné, who conceives ἵνα ἐστὶ ἐν ἡλικίᾳ Χίρων, εἰς equivalent to ἐξερευνήσας τοῦ Ἠρακλέους. But, however this may be, it is evident from the words here quoted, that the celebrated German scholar does not consider εἰς as a present tense; and that, whether our interpretation of the passage be right or wrong, Heyné, in adopting the ordinary one, has distinctly corroborated the view which we have taken of εἰς.

But, not to dwell longer on this, other examples of the nature of instantaneity crucis may be produced to disprove Mr Dunbar's assertion that εἰς is used "as a present." In the ninth Pythian Ode of Pindar (l. 104) we meet with the following passage:

Τοῦτο τελεῖ τετράς, ἐν ἡλικίᾳ, Ἐρμῆς Ἐκδιδοῦσας ὁμοίως καὶ Γάρης, ἀπὸ τῆς ἰδίας ἰδίας ἰδίας, Οἴδητε.

In this sentence, if εἰς were a present, as Mr Dunbar contends, how is its time to be reconciled with that of εἰς; ἐστί; If the learned professor be right, Pindar has, in this case at least, been guilty of a gross anachronism. Again, in the seventh book of the Iliad (l. 81), Hector says,

Εἰδοὺς εἶ ἐν τῷ τετάρτῳ, ἔχω ἂν ἀπολύσω Ἀχαιοὺς, Τυράννον ἀπολύσω, ὅτιδ' ἂν ἀπολύσω Δαρεῖον, Καὶ ἐπὶ τοῦτο τοῦτο Ἀχαιοὶ ἀπολύσω.

"Sin ego illum interfecerò, dederitque mihi gloriam Apollo, arma detracta feram ad Ilium sacrum, et suspendam ad templum Apollinis longe jaculantis:" "But if I shall be fortunate enough to slay him, and Apollo give me the glory of such a victory, then, having stripped him of his arms, I will carry them to sacred Ilium, and will hang them up at the temple of the far-darting Apollo." By no sophistry can the verb εἰς in this passage be tortured into a present; and the same observation equally applies to εἰς in the foregoing extract from Pindar. But if εἰς be a first future, as it undoubtedly is, can it for a moment be doubted that εἰς is an aorist formed upon the analogy of that future; and if so, is it not clear that, in some instances at least, the second aorist follows the analogy of the first future?

(A.)