Whatever opinion may be formed on this very trifling point, D'Alembert is entitled to equal praise for the ingenious application he has made of the story to illustrate the extensive and powerful influence of language upon thought. It seems to have struck Locke's fancy very strongly in this point of view, for he has alluded to it repeatedly in different parts of his Essay.

SECTION IV.—MISCELLANEOUS OBSERVATIONS ON LANGUAGE.

The latest inquirers into that branch of literature which relates to language, have directed their attention chiefly to the etymological study of different tongues, considered as a guide to our conclusions concerning the origin and migrations of the various tribes of our species. As this view of the subject has, at first sight, but a remote connexion with the Philosophy of the Human Mind, I shall dismiss it with a few miscellaneous remarks. Some of these will be found, I flatter myself, not altogether foreign to the design of this work.

The first author of eminence who led the way in these etymological speculations was Leibnitz, whose various contributions to this branch of knowledge occupy no inconsiderable space in the general collection of his works. In his earliest publication on the subject, which forms the first article in the first volume of the Memoirs of the Berlin Academy, he has very clearly and concisely stated the important purposes to which he conceived such researches to be subservient, as well as the leading principle which he proposed to keep in view in carrying them on. I shall only quote the introductory paragraph; but the whole paper well deserves the attention of those whose taste leads them to similar pursuits.

"Cum remotæ gentium origines historiam transcendant,

serve to prove, that the anecdotes of the noble author are not always to be implicitly relied on, even with respect to his own contemporaries. His incorrectness, in the present instance, is the

more wonderful, as he was himself, (as I happen to know from good authority,) a pupil of Sanderson's, and lived with him in habits of intimacy, while a student at Cambridge.

linguæ nobis præstant veterum monumentorum vicem. Et vetustissima linguarum vestigia supersunt in nominibus fluviorum atque sylvarum, quæ mutatis accolis plerumque persistunt; proximæ sunt locorum ab hominibus constitutorum appellationes; quanquam enim multæ villæ, multa oppida à conditoribus nominentur, quod in Germania valde est frequens, quæ serius exculta est; alia tamen loca à situ, à proventu, à cæteris qualitatibus appellantur, et vetustiorum difficilis est etymologia. Nomina etiam vetera hominum, quorum nulla Germaniæ gens plura Frisiis retinuit, ducunt nos in sacraria, ut sic dicam, veteris linguæ. Illud enim pro axiomate habeo, omnia nomina quæ vocamus propria, aliquando appellativa fuisse; alioqui ratione nulla constarent.1 Itaque quoties vocabulum fluminis, montis, sylvæ, gentis, pagi, oppidi, villæ, non intelligimus, intelligere debemus, ab antiqua nos lingua discessisse."2

In various other parts of his works, Leibnitz has enlarged on the ignorance in which we are left by our historical records, with respect to the earlier migrations of the human race; and has fully established this principle, in which all the soundest philosophers now acquiesce, that if any new lights shall ever be thrown on this part of the history of mankind, it is from these etymological researches, conducted by extensive learning under the guidance of sober judgment and good sense, that they are chiefly to be expected.

In confirmation of this idea, I shall quote a passage from Mr. Horne Tooke, an author whose opinion is justly entitled to considerable deference in all etymological discussions, and whose good sense has preserved him, on most occasions, from giving up his judgment to those fanciful resemblances and analogies, which had so powerful an influence on the theories.

of most of his predecessors in this department of philology. The passage which follows, I introduce merely as a proof of the confidence with which, in the opinion of Tooke, we are entitled to reason concerning the migrations of our race in past ages, from the affinities between the languages which have prevailed in different parts of the globe.

"It is a great mistake, into which both the Italian and Latin etymologists have fallen, to suppose that all the Italian must be found in the Latin, and all the Latin in the Greek, for the fact is otherwise. The bulk and foundation of the Latin language is Greek; but great part of the Latin is the language of our northern ancestors, grafted upon the Greek. And to our northern language the etymologist must go for that part of the Latin which the Greek will not furnish; and there, without any twisting or turning, or ridiculous forcing and torturing of words, he will easily and clearly find it. We want, therefore, the testimony of no historians, to conclude that the founders of the Roman state, and of the Latin tongue, came not from Asia, but from the north of Europe, for the language cannot lie, and from the language of every nation we may with certainty collect its origin. In the same manner, even though no history of the fact had remained, and though another Virgil and another Dionysius had again, in verse and prose, brought another Eneas from another Troy to settle modern Italy, after the destruction of the Roman government, yet, in spite of such false history, or silence of history, we should be able, from the modern language of the country, (which cannot possibly lie,) to conclude with certainty, that our northern ancestors had again made another successful irruption into Italy, and again grafted their own language upon the Latin, as before upon the Greek, for all the Italian which cannot be easily shown to be Latin, can be easily shown to be our northern language."1

1 Diversions of Purley, vol. ii. p. 140, quarto edit. A similar observation is made by Leibnitz in the memoir above quoted. "Ex Celtis, id est Germanis Gallisque, Alpes Pyrenæosque transgres

sis, Italiam et Hispaniam habitatores accepisse credibile est, longe ante illos qui memorantur Livio Gallorum posteriorum adventus. Aborigines Italiæ intelligo, id est incolas Græcis, Lydis,

In this department of literature important future discoveries may, I think, be confidently anticipated, in proportion as a more extensive communication among the different tribes of mankind shall be opened, and as the affinities between their various languages shall be more accurately investigated. Those which have been already traced among some of the most remote and apparently unconnected tongues, are sufficient to demonstrate how imperfectly we are acquainted with the past migrations of our race; and, if they are not yet sufficient to afford a direct proof of the common descent of the different inhabitants of our globe, unquestionably tend, the farther the subject is prosecuted, to lend additional presumptions in favour of that conclusion.

In order, however, to render the study of the affinity of languages a solid foundation for our conclusions, it is necessary that those who devote themselves to it, (keeping in mind the limited grasp of the human powers,) should guard against the danger of rendering their labours fruitless, by aiming at what is wholly beyond the comprehension of our faculties. A few languages, grammatically and critically possessed, would enable them to add more usefully to the mass of philological knowledge, than the almost miraculous gift of tongues displayed in the labours of Adelung and some of his successors.1 When I say this, I would not be understood to detract from the great and acknowledged merits of these literary prodigies. What

Phrygibus, Phœnicibus aliisque mari advectis antiquiores. Nam antiquissimæ migrationes omnes terra factæ sunt, serius et ægrè magna multitudo navigavit: quanquam Tacito contrarium exciderit. Græcorum deinde multæ in Italia maritimæ coloniæ conditæ sunt: inde lingua Latina ex Celticâ Græcâque mistis nata. Postea pars Italiæ citerior Galliæ Cisalpinæ, ulterior magnæ Græciæ nomen tulit; in medio Latini et Tusci cum vicinis plurimum utrinque trahebant. Hetruscam antiquam non intelligimus, ac quæ ipsius in lapidibus nonnullis supersunt, ne legimus quiVOL. IV.

dem."-P. 10. [Opera, Dutensii, vol. pars ii. p. 193.]

iv.

On this subject see also Dictionnaire Etymologique de la Langue Latine, with the Discours Préliminaire, in the sixth volume of Court de Gebelin's Monde Primitif.

I call it an almost miraculous gift, because in looking over such tables as that exhibited in the Supplement to the Encyclopædia Britannica, (see Article Languages,) I can only wonder and admire at faculties to which I am unconscious of possessing in myself anything at all analogous,

E

they have done, was necessary to prepare the way for the minuter industry of their successors. The great outlines being now marked out, and the limits of the chief departments settled, it will be comparatively easy for inferior artists to complete the details of the survey. Much may be expected, for much has been already done, by those missionaries to foreign parts, who, in the midst of their more important occupations, have not thought it an unworthy employment of their time to make an accurate study of the languages spoken by the nations they wish to convert. In India it is well known what they have accomplished, by translating the Scriptures into different tongues, even into the sacred language of the Bramins; with which I have little doubt some of them now possess a more intimate and familiar acquaintance than had previously been acquired by any other class of scholars who have visited the East.

On the other hand, it is impossible not to feel some scepticism concerning etymological researches, when they turn on languages which are accessible only to a small number of scholars; the deceptions which are commonly practised in support of such theories being facilitated in proportion to the When I say obscurity in which the subject is involved. deceptions, I would not, by any means, insinuate, that they are always intentional and culpable. It is a natural and pardonable weakness to overvalue even a smattering of knowledge which is possessed by few; and one of the adepts may be readily excused, if he addresses the public on his favourite subject, in a more confident tone than he would have assumed, had the points in question lain more open to general examination. Add to this, that when once the mind is intoxicated with a theory, it eagerly grasps at every shadow of evidence which seems to favour it, and is generally the first dupe to the system it has created.1 I well remember the impression uni

1 In the hope of guarding my younger readers against lending too easy a faith to the seducing theories of etymologists, I shall subjoin the sober opinion of a writer, who, of all our contemporaries,

was best entitled, from his own unparal leled acquisitions, to form a competent judgment on this subject;-I need scarcely say that I allude to Sir William Jones: