^{page 76 note 1} Only a very rash man ventures new suggestions on the Poetics without consulting a consilium amicorum; I am much indebted to A. M. Dale, Hermann Fränkel, John Moore, Brooks Otis, and T. B. L. Webster for helpful discussion and criticism.

^{page 77 note 1} For the technique of Lysias in accomplishing this purpose cf. Usher, S., ‘Individual Characterisation in Lysias’, Eranos lxiii (1966), 99–119.Google Scholar

^{page 78 note 1} Following the lead of Spengel, Vahlen, Ueberweg, and Gudeman, many German scholars seem to have accepted the reading as firmly established; it is also adopted by Hatzfeld-Dufour and Butcher. Bywater, Hardy (Budé ed.), Rostagni, Kassel, and Else prefer the compound. But they give no reasons for their preference. In Rhet. ad Alex. 1438^b26 the compound is certainly preferable, as it describes an orator's choice of material: .

^{page 79 note 1} Bonitz, Index Aristotelicus, under with accusative, gives as the invariable meaning; a check of the passages which he quotes shows that he is right, that it is not used to indicate purpose (for which Aristotle regularly uses ).

^{page 79 note 2} Pearson, L., The Lost Histories of Alexander the Great (American Philological Assoc. Monographs xx, 1960), esp. Chapter 8.Google Scholar

^{page 79 note 3} Cf. Walbank, F. W., ‘Tragic History: a Reconsideration’, BICS ii (1955), 4–14,Google Scholar and ‘History and Tragedy’, Historia ix (1960), 216–34.Google Scholar

^{page 79 note 4} Bywater translates: ‘We maintain that Tragedy is primarily an imitation of action and that it is mainly for the sake of the action that it imitates the personal agents.’ But he cannot be right; Aristotle has made it clear that representation of action comes first in time as well as in importance, and characterization second. Gudeman renders ‘der Handlung zu Liebe’ in the earlier passage, but ‘vermittelst der Handlung’ here. His commentary, however, shows that he is not happy about the argument at all.

^{page 80 note 1} Else translates ‘for that reason rather than any other’. I should prefer ‘mainly by this means’.

^{page 80 note 2} So also in the discussion of characterization in Poetics 1454^a we see that character in tragedy is created as speech or action reveals motivation. A single may be insufficient to indicate moral quality, but if the reason for the action is made clear, we shall have characterization: .

^{page 80 note 3} Ueberweg evidently thought that Aristotle could not have argued like that, and therefore emended in the final phrase to . This was apparently what the Arabic translator read here. Cf. Solmsen, F., ‘The Origin and Method of Aristotle's Poetics’, CQ xxix (1935), 193.Google Scholar Solmsen is ‘hardly convinced that Aristotle really wanted the solution of the to depend upon the of the characters’

^{page 81 note 1} The simple compound can mean ‘acquire’ or ‘attain’ as well as ‘include’ or ‘contain’. Cf. Hdt. 5. 23, 8. 106, Isaeus 8. 37, Isoc. Paneg. 45, Nic. 22, where it means ‘get control of’; and for a less material sense Antid. 184 (). The verb's range of meaning is particularly well illustrated by Isocrates. A meaning more like ‘define’ seems to be indicated in Antid. 217 and Ad Nic. 9, whereas in Antid. 8 and Plat. 63 it suggests ‘accomplishing purposes all at the same time’, ‘encompassing’, or in more colloquial language ‘managing’, as we might say to a man with numerous tasks to do or heavy burdens to carry, ‘Can you manage all that?’ Cf. also the use of in [Plutarch] Mor. 14 c, where the object of the verb, though the text is in doubt, must be something like ‘tasks’ or ‘precepts’.

^{page 81 note 2} Latin writers use involvere in a similar way. Cf., e.g., Virgil, Aen. 6. 336, 12. 689.

^{page 82 note 1} Cf. Gregory of Nyssa, Ep. 3. 12: . Note also his use of in Ep. 17. 3.

^{page 82 note 2} The modern use of ‘involve’ and ‘imply’ is not strictly justified by any examples in classical Latin. Krebs-Schmalz, Antibarbarus der lateinischen Sprache, s.v. Involvo and Implico, give stern warning that it is neulatfinisch. The earliest use in the modern sense is hard to trace; it is not recognized in the Thesaurus linguae latinos. Aquinas uses implicare in the sense of ‘imply’, and the New English Dictionary notes how already in the seventeenth century English, unlike French or Italian, uses ‘imply’ and ‘employ’, both derived from implicare, in entirely different meanings.

One can see the germ of the development in passages like Tac. Ann. 3. 63, si qua iniquitas involveretur, where the modern meaning ‘involve’ would be possible, though the passage should more properly be compared with Virgil, Aen. 6. 100, obscuris vera involvens.