4. THE second class above mentioned, is of those words which, in one language, do, but imperfectly, correspond to any of the words of another language compared with it. Of this kind will be found, if properly attended to, most of the terms relating to morals, to the passions and matters of sentiment, or to the objects of the reflex and internal senses, in regard to which, it is often impossible to find words in one language, that are exactly equivalent to those of another. This holds in all languages, less or more, according as there is more or less, uniformity, in the constitution, religion, and laws, of the nations whose languages are compared; on which constitution, religion, and laws, as was observed, the sentiments, manners, and customs of the people, in a great measure, depend. Herein consists one principal difficulty which translators, if persons of penetration, have to encounter. Finding it sometimes impossible to render fully the sense of their author, they are constrained (if I may borrow a term from the mathematicians) to do the best they can by approximation.

To come to examples: To the Greek words apɛτης σωφροσύνη, εγκρατεια φρόνησις, ελεος, the Latin words, virtus, temperantia, continentia, prudentia, misericordia, are not entirely equivalent; still less the English words virtue, temperance, continence, prudence, mercy: for, though these last are manifestly formed from the Latin words, one would think that, by being adopted into another country, they had all, more or less, changed their nature with the climate.

Those persons whose knowledge, in such matters, is but superficial, will not enter readily into these sentiments. They are accustomed to consider certain words, in the different languages, as respectively correspondent. The grammars, lexicons, and common translations, lead them to conclude so, and they inquire no further. But those who are conversant with authors of reputation, in these different tongues, will need no arguments to convince them of the truth of what has been advanced.

Who knows not that the Latin word virtus would, in many instances, be but weakly, not to say improperly, rendered by the English word virtue; as that word, in Roman authors, comes often nearer the import of what we call valour or fortitude, sometimes even brute force? We should not readily ascribe virtue to wild beasts; yet Tacitus so applies the term virtus: "Fera animalia, si clausa teneas, virtutis obliviscuntur." And if some of our words have too great latitude of signification to answer always to their Latin etymons; some have, on the contrary, too little. For example, the English word temperance is too confined in meaning to answer to the Latin temperantia, which implies moderation in every desire, and is defined by Cicero, in one place, " moderatio cupiditatum rationi obediens ";" and in another, "temperantia est quæ in rebus aut expetendis aut "fugiendis, rationem ut sequamur, monet 3." Now all that is implied in the English word is almost only

37 De Fin. l. ii.

37

38 De Fin. 1. i.

that species which he denominates "temperantia in victu." And, though the differences may not be so considerable in all the other related words above. mentioned, it were easy to shew that they cannot, in every instance, be made to tally.

It requires, indeed, but a very small skill in languages to enable us to discover that etymology is often a very unsafe guide to the proper acceptation of a term. It will not be doubted that the Latin word sobrius is the root of the English word sober, and their term honestum of our term honesty ; but every body knows that the related words, in the two languages, will not always answer to each other. Nay, to shew, in the strongest manner, how much more difficult it is, than is commonly imagined, to apprehend the precise import, and proper application, of words of this order in dead languages, I shall transcribe a short passage from the fourth book of the Tusculan Questions, where the author explains the generic word ægritudo, with the various names of species comprehended under it. Amongst other observations are the following: "Agritudo est opinio recens mali presentis, in quo demitti contrahique animo rectum esse videatur. Egritudini subjiciuntur angor, mæror, dolor, luctus, ærumna, afflictatio: angor est ægritudo premens, mæror ægritudo flebilis, ærumna ægritudo laboriosa, dolor agritudo crucians, afflictatio ægri. "tudo cum vexatione corporis, luctus ægritudo ex ejus, qui carus fuerat, interitu acerbo." "Let any "one," says D'Alembert ", "examine this passage

ვი

39 Sur l'Harmonie des Langues, et sur la Latinité des Modernes.

"with attention, and say honestly, whether, if he "had not known of it, he would have had any idea "of these nice shades of signification here marked; "and whether he would not have been much embar"rassed, had he been writing a dictionary, to distinguish with accuracy the words ægritudo, mæror, "dolor, angor, luctus, ærumna, afflictatio. If Cicero, "the greatest philosopher as well as orator that ever "Rome produced, had composed a book of Latin

synonymas, such as that which Abbe Girard did "of French; and if this work had but now for the "first time been produced in a circle of modern La

tinists, I imagine it would have greatly confound"ed them, in showing them how defective their knowledge is of a subject of which they thought "themselves masters."

I have brought this quotation, not to support D'Alembert's opinion, who maintains that it is impossible for any modern to write Latin with purity; ⚫ but only to shew how much nicer a matter it is than is commonly supposed, to enter critically into the peculiarities of a dead language. It might be easily shown, were it necessary, that distinctions like those now illustrated in the nouns, obtain also in the verbs of different languages. Under this class those words also may be comprehended which are not barely the names of certain things, or signs of particular ideas, but which express also the affection or disposition of the speaker, towards the thing signified. In every language, we shall find instances wherein the same thing has different names, which are not perfectly sy

nonymous; for though there be an identity of subject, there is a difference of manner, wherein the speaker appears affected towards it. One term will convey the idea with contempt, another with abhorrence, a third with some relish, a fourth with affection, and a fifth with indifference. Of this kind are the diminutives and amplificatives which abound so much in the Greek, and Italian, languages.

It is this principally which justifies Girard's observation, that there are much fewer words in any language which are, in all respects, synonymous than is commonly imagined. And it is this which makes the selection of apposite words so much, and so justly, the study of an orator: for when he would operate on the passions of his hearers, it is of the last consequence, that the terms he employs not only convey the idea of the thing signified, which may be called the primary use; but that, along with it, they insinuate into the minds of the hearers, the passion of the speaker, whatever it be, love, or hatred, admiration or contempt, aversion or desire. This, though the secondary use of the word, is not the less essential to his design. It is chiefly from the associated affection that these different qualities of synonymous words taken notice of by Quintilian must be considered as originating: "Sed cum idem frequentissime plura significent, quod ovvovvua vocatur, jam sunt "alia aliis honestiora, sublimiora, nitidiora, jucundiora, vocaliora." The last is the only epithet which regards merely the sound. The following will serve for an example of such English synonymas, public