In carrying on such etymological enquiries, atten tion should first be paid to the languages in which the real primitives are to be found. Thus in analysing the modern English, the Anglo-Saxon, the Greek, the Latin, must all be referred to, as each of these have contributed something to the present stock, and without a knowledge of each, the investigation must be very incomplete. In the actual deduction of the words as they occur, discri minating attention is often required. Sometimes, indeed, the origin is too manifest to admit of dubiety. Terms of art, ecclesiastical denominations, and forensic phrases, if transplanted from one language to another, generally bear such manifest marks of their peculiar origin, that they can at once be referred to their primitives; but nouns and verbs in common use, particles, prepositions, and conjunctions, are often so changed in passing from one people to another, that it requires an accurate scru tiny before we can ascertain their parent stern. In such cases, various circumstances call for attention. The successive revolutions in the language which we are at tempting to analyse should be studied ; changes, both in orthography and in pronunciation, gradually taking place, and these sometimes so considerable, that without remounting to the earlier stages, the connecting links would quite disappear. The varieties of enunciation even of the same words, particularly in the vowels, which accident may have introduced, and habit continued ; the interchange of consonants of the same order, as certain organs of speech in diffct cm countries arc more or less exercised, and the alteration of letters for the sake of more harmonious sound, arc all of them fruitful sources of apparent discrepancy, in words of similar origin and import ;—while, at the same time, metaphorical applica tions, restricted senses, and analogical modifications, are no less apt to produce great variations in meaning. All these it is the business of a skilful etymologist to analyse, till the primitive word can be traced through all its chan ges. Examples of all kinds may be found in every mo dern dictionary. While the etymologist, by attention to these circumstances, can proceed successfully in his la bours, great caution is necessary, on the other hand, to avoid the fanciful and absurd deductions in which lexi cographers and antiquarians have too rashly indulged. Etymologies far-fetched, orbuilt upon resemblances pure ly accidental, ought in general to be disregarded ; and unless historical circumstances can be traced, sufficient at least to render the supposed connection or descent pro bable, if not certain, a few similarities of speech would be too weak a foundation for a system of filiation. Occa sional coincidences may arise in languages quite uncon nected, and which have no radical affinity ; it is only when, by following up the analysis, we can trace the re semblance in the radical stems, that we are warranted to infer the actual descent.
Upon the principles now laid down, etymological enquiries into the structure of different languages might be conducted with success, proceeding by a strict inductive process from the derivative languages ac tually existing, to the remoter primitive tongues, which furnished the roots whence these have gradu ally emanated. When arrived at these primitive lan guages, it might still be an object of curious research to continue the investigation, and endeavour to analyse these primitive languages themselves, till som.:. view could be obtained of the first elements, and subsequent progress of human speech. In every primitive lan guage, it seems by no means improbable, that if the roots could be fully investigated upon just and philoso phical principles of etymology, and the application of the elemental sounds in the formation of each distinctly marked out, a peculiar and specific idea would be found attached to the use of every individual cons°• nant, prevailing in some form or other, through all the subsequent ramifications. Not that the general idea could be in view when the words were formed, but that the generalisation arose from the repeated applica tion of the same sound, to express individual objects or individual feelings, concurring in the same common quality. By such an analysis, a theory of language might be Formed, according with the progress of na ture, and an excellent tract would be opened, for tra cing the procedure of the human intellect in the ar rangement of ideas and formation of speech. A com parison of these primitive roots, too, would best enable us to determine what degree of affinity existed among the languages to which they belong, and, perhaps, throw some light on the much-agitated question, whether all the languages of the world were really derived from one. Probable as the opinion of their common origin must be allowed to be, and strengthen ed by many striking proofs of actual coincidence, still, to establish it completely, such evidence seems to be wanting, as a radical analysis, upon the principles now mentioned, alone could furnish.
Etymology has chiolly been cultivated by lexicogra phers and grammarians, a class of writers too nume rous to be particularized. Philosophical etymology has principally been studied by the moderns; a few ex amples of it occur in the writings of Plato and Aristo tle, and in some of the works of the Platonists of the .Alexandrian school, but these arc only incidentally, and not systematically touched upon. In modern times, some of the Dutch philologists, particularly llemster huis and his disciples, have distinguished themselves in this line. Horne Tooke has thrown great light on Eng lish etymology, and Dr Jamieson has furnished much valuable information on that or the Northern languages. The expected work of Dr Murray on the languages of Europe, there is reason to hope, will prove a most valuable acquisition to the lovers of etymological re seal ell. (<0