Page:Language and the Study of Language.djvu/276

254 necessity, a sound induction from observed facts, which brings us to the conclusion that all linguistic elements possessing distinct meaning and office, variously combined and employed for the uses of expression, are originally independent entities, having a separate existence before they entered into mutual combination.

In the light of these considerations let us examine a single word in our language, the word irrevocability. It comes to us from the Latin, where it had the form irrevocabilitas (genitive -tatis). It is clearly made by the addition of ty (tas, tatis) to a previously existing irrevocable (irrevocabili-s), just as we now form a new abstract noun from any given adjective by adding ness: for example, doughfacedness. Again, revocable (revocabilis) preceded irrevocable, as dutiful preceded undutiful. Further, if there had been no verb to revoke (revocare), there would have been no adjective revocable, any more than lovable without the verb to love. Yet once more: although we in English have the syllable voke only in composition with prefixes, as revoke, evoke, invoke, provoke, yet in Latin, as the verb vocare, 'to call,' it is, of course, older than any of these its derivatives, as stand is older than understand and withstand. Thus far our way is perfectly clear. But while, in our language, voke appears as a simple syllable, uncombined with suffixes, this is only by the comparatively recent effect of the wearing-out processes, formerly illustrated (in the third lecture); in the more original Latin, it is invariably associated with formative elements, which compose with its forms like vocare, 'to call,' vocat, 'he calls,' vocabar, 'I was called;' or, in substantive uses, vocs (vox), 'a calling, a voice,' vocum, 'of voices;' and so on. There is nothing, so far as concerns the formative elements themselves, to distinguish this last class of cases from the others, before analyzed; each suffix has its distinct meaning and office, and is applied in a whole class of analogous words; and some of them, at least, are traceable back to the independent words out of which they grew. The only difference is that here, if we cut off the formative elements, we have left, not a word, actually employed as such in any ancient language of our family, but a significant