The vagueness of the concept of mimesis reached its apex in those writers who gave it as a general title to any sort of work that had nature for its object, employing the Aristotelian phrase to affirm that "_omnis ars naturae imitatio est,_"[38] or saying, like the painter Eupompus when he blamed his servile imitators, that "_natura est imitanda, non artifex._"[39] And those who wished to escape this vagueness did not know how to do so, save by conceiving the activity of imitation as the practical producer of duplicates of natural objects, a prejudice bora in the bosom of the pictorial and plastic arts, against which Philostratus perhaps intended to argue, in common with the other advocates of imagination. [Sidenote: _Speculations on language._] The speculations upon language had a close connexion with those upon the nature of art begun by the sophists, for whom it became a matter for wonder that sounds could signify colours or things inaudible; that is to say, _speech_ presented itself as a _problem._[40] It was then discussed whether language was by nature (φύσει or by convention νόμω). By nature was sometimes understood mental necessity, and by convention what we should call a merely natural fact, psychological mechanism or sensationalism. In that sense of the terms, language would have been better called φύσει than νόμω. But at other times the distinction led to the question whether language answers to objective or logical truth and to the real relations between things (όρθότης των ονομάτων); and in this case, those would seem to be nearer the truth who proclaimed it to be conventional or arbitrary in respect to logical truth: νόμω or θέσει, and not φύσει Two different questions were consequently being treated together, and both were confusedly and equivocally discussed. They find their monument in the obscure _Cratylus_ of Plato, which seems to fluctuate between different solutions. Nor did the later affirmation that the word is a sign (σημείον) of the thought solve anything, for it still remained to be shown in what way the sign was to be understood, whether φύσει or νόμω. Aristotle, who looked upon words as imitations (μιμηματα), in the same way as poetry,[41] made an observation of first-rate importance: in addition to the _enunciative_ propositions, which express the (logically) true or false, there are others which do not express either the (logically) true or false, as for example the expressions of aspirations and of desires (εύχή), which therefore belong, not to logical exposition, but to poetical and rhetorical exposition.[42] And in another place we find him affirming in opposition to Bryson (who had said that a base thing remained such with whatever word it were designated) that base things can be expressed both with words that place them beneath the eye in all their crudity, and with other words which surround them with a veil.[43] All this might have led to the separation of the