In this much-vexed passage a form or variant of ναγνώρισς is described and exemplified. It is said to be συνθετ, which word all the translations I have seen render ‘compound’ or ‘composite,’ but their authors either omit or (me judice) clearly fail to explain in what sense this form is said to be ‘composite.’ I believe that this translation is wrong, and that the word here means something else, συνθετος λόγος is good Greek for ‘a made-up tale’ (cf. Aesch. Prom. Vinct. 686, where συνθετο λόγοι are also called μθοι ψευδεῖς). The point is not that what is said or told is necessarily a lie, but that it is untrue, baseless, not founded on fact. All this is supported by the use of the corresponding verb in e.g. Thuc. I. cc. XXI. and XCVII., where it is opposed to the truthful recording of facts (συγγρϕειν). Further, it justifies Dacier's reading in 1459a 21, where the relative clause refers to ἱστοραις not to τς συνθσεις. The characteristic of a poetic σνθεσις is that it is a connected narrative which has a causal nexus between its parts, and so is of μα πρξις, whereas an ἱστορα is a record of disconnected happenings to a single person or group of persons during a certain period. Such a σὐνθις may be merely made-up or fictional. It is specially so where it has no basis in fact (or prior accepted legend or ‘epic’), but is the sheer invention of the poet. Clearly such a fiction must be rendered acceptable or plausible, and this can only be done through someone being misled, and this misleading implies that that someone draws a wrong conclusion from evidence present or presented to him— i.e. it presupposes a παραλογισμός.