seen itself forming a suffix, in double, triple, and so forth, and which conveys the idea of 'bending' or 'folding.' By successive extensions and modifications of meaning, by transferral from one category to another through means of their appropriate signs, we have developed this simple idea into a form which can only be represented by the long paraphrase 'numerous conditions of being not able to bend (or fit) to something.'
With but few exceptions—which, moreover, are only apparent ones—all the words of our language admit of such analysis as this, which discovers in them at least two elements, whereof the one conveys the central or fundamental idea, and the other indicates a restriction, application, or relation of that idea. Even those brief vocables which appear to us of simplest character can be proved either to exhibit still, like am for as-mi, the relic of a mutilated formative element, or, like is for as-ti, to have lost one which they formerly possessed. This, then, in our language (as in the whole family of languages to which ours is related), is the normal constitution of a word: it invariably contains a radical and a formal portion; it is made up of a root combined with a suffix, or with a suffix and prefix, or with more than one of each. In more technical phrase, no word is unformed; no one has been a mere significant entity, without designation of its relation, without a sign putting it in some class or category.
It is plain, therefore, that a chief portion of linguistic analysis must consist, not in the mere dismembering of such words as we usually style compounded, but in the distinction from one another of radical and formal elements; in the isolation of the central nucleus, or root, from the affixes which have become attached to it, and the separate recognition of each affix, in its individual form and office. But our illustrations have, as I think, made it not less plain that there is no essential and ultimate difference in the two cases: in the one, as in the other, our process of analysis is the retracing of a previous synthesis, whereby two independent elements were combined and integrated. That this is so to a certain extent is a truth so palpable as to admit of neither
