^{page 42 note 1} Reference to this work will be by chapter and Bekker page and line number only.

^{page 42 note 2} Hamblin, C. L., Fallacies (London, 1970)Google Scholar, especially chapters 1 to 4.

^{page 42 note 3} Patzig, G., Die aristotelische Syllogistik² (Göttingen, 1963)Google Scholar.

^{page 42 note 4} Perhaps the soundest available comment is in the Introduction to Brunschwig, J., Aristote: Topiques, Tome 1 (Paris, ‘Les Belles Lettres’, 1967)Google Scholar. But I must immodestly ask those who wish for a more accurate assessment of this question to await my forthcoming work on Aristotle's concept of dialectic.

^{page 42 note 5} See the reviews by Kneale, Martha, Philosophical Quarterly XXI (1971), 184Google Scholar, and Englebretsen, George, Dialogue XII (1973), 153Google Scholar.

^{page 42 note 6} Probably the most influential among the many presentations of this view has been that of Solmsen, F., Die Entwicklung der aristotelischen Logik und Rhetorik (Berlin, 1929), especially pp. 70–6Google Scholar.

^{page 42 note 7} For examples, see Hamblin, , Fallacies, p. 13Google Scholar.

^{page 43 note 1} 1. 164 b 25–165 a 4; 17. 175 a 31–6.

^{page 43 note 2} 3. 165 b 18–22. At 14. 173 b 23–5 (cf. 32. 182 a 20–4, 182 b 3–5) Aristotle speaks of ‘syllogising solecisms’, thereby indicating that arguments which are concerned with solecisms are posterior in definition to syllogisms and refutations simpliciter. On solecisms see further p. 51 below.

^{page 43 note 3} 2. 165 b 7–8; 34. 183 a 27–8.

^{page 43 note 4} 11. 172 a 11–13; Met. K₃. 1061 a 28–b 11.

^{page 44 note 1} 8. 170 a 12–19.

^{page 44 note 2} Top. A 10–11; ϴ 5; see my ‘Aristotle on Relativism’, Philosophical Quarterly XXIV (1974), esp. p. 202Google Scholar.

^{page 44 note 3} See Patzig, Syllogistik, chapter 3.

^{page 44 note 4} 9. 170 b 5–8.

^{page 44 note 5} Rhet. A2. 1356 b 28–1357 a 7; E.N. A₃. 1095 a 2–11.

^{page 44 note 6} 10. 170 b 32–5.

^{page 44 note 7} 10. 171 a 23–7.

^{page 45 note 1} 10. 170 b 38–40.

^{page 45 note 2} 8. 169 b 37–170 a 11.

^{page 45 note 3} 8. 169 b 18–20.

^{page 45 note 4} See Kirwan, Christopher, Aristotle's Metaphysics Books Г, Δ, Ε (Oxford, 1971), pp. 84–5Google Scholar, who is, however, over-cautious about this. To the passages which link dialectic and peirastic add Top. ϴ 11. 161 a 25.

^{page 46 note 1} 9. 170 a 20–34.

^{page 46 note 2} 9. 170 a 35–9; 11. 172 a 11– b 1.

^{page 46 note 3} 11. 171 b 34–172 a 2, and especially 172 a 9–12.

^{page 46 note 4} 11. 172 b 1–8.

^{page 47 note 1} 22. 178 a 16–28.

^{page 47 note 2} 20. 177 a 35.

^{page 47 note 3} 20. 177 a 38–b. 9.

^{page 47 note 4} 4. 166 a 14–21.

^{page 47 note 5} This could provide an argument to back up the claim at 4. 165 b 27–30.

^{page 48 note 1} See Stalnaker, Robert C., ‘Pragmatics’, in Semantics of natural languages, edd. Davidson, D. and Harman, G. (Dordrecht, 1972), pp. 380–97CrossRefGoogle Scholar.

^{page 48 note 2} 6. 168 b 27–31; 7. 169 b 6–7; 8. 170 a 4–5.

^{page 48 note 3} 6. 169 a 5.

^{page 48 note 4} See Quine, W. V., From a logical point of view² (Cambridge, Mass., 1961), pp. 139–59Google Scholar.

^{page 48 note 5} My purpose in using this unaristotelian language is to show how a modern exponent of the distinction between contexts of reference would comment on the syllogistic, which he would probably interpret as talking about properties. Aristotle's different way of describing these matters follows from his distinction between essence and accident.

^{page 48 note 6} See Heintz, John, ‘Property existence and identity’, Journal of Philosophy LXX (1973), 734–43, esp. 738–9Google Scholar.

^{page 49 note 1} Ryle, Gilbert, Plato's progress (Cambridge, 1966), p. 143Google Scholar; Hamblin, , Fallacies, p. 86Google Scholar.

^{page 49 note 2} Met. E₂; An. Pr. A₁₃. 32 b 19; An. Post. B₃₀. All syllogistic requires at least one universal premiss.

^{page 49 note 3} 24. 179 a 35–9.

^{page 49 note 4} 24 179 b 7–11.

^{page 49 note 5} 24. 179 b 15–21.

^{page 49 note 6} 24. 179 b 27–33.

^{page 49 note 7} 25. 180 a 32–b 6.

^{page 50 note 1} The example of a B₃ argument at 26. 181 a 8–11 strongly recalls the disputed argument at 24. 179 b 7–11.

^{page 50 note 2} 20. 177 b 27–34.

^{page 50 note 3} 24. 179 b 37–180 a 22.

^{page 50 note 4} 30. 181 b 19–24. On the other hand the comparison of A₁ and B₇ at 17. 175 b 39–176 a 18 has no tendency to undermine the A/B distinction.

^{page 50 note 5} 17. 175 b 18–27. This obscure passage examines an escape route from the argument that the reference of an ambiguous name is determined by whatever makes true a statement in which it occurs. The suggestion contests this argument by maintaining that while the object of reference is fixed, the object-aspect (‘this Coriscus’) is not. Aristotle thinks that this suggestion concedes too much, and appeals to the case of visible objects – viz. those where the definiteness of reference of the name is uncontestable.

^{page 50 note 6} The prime context is the discussion of the object of wish in E.N. Г 4. See ‘Aristotle on Relativism’, pp. 198–9, and an extended treatment in my forthcoming work on dialectic referred to above.

^{page 51 note 1} For a parallel conception of method in zoological classification, see Balme, D. M., Aristotle's De Partibus Animalium I and De Generatione Animalium I (Oxford, 1972), p. 105Google Scholar.

^{page 51 note 2} 14. 174 a 5–9.

^{page 51 note 3} For this reason I treat ‘language’ as a count-noun, always speaking of ‘a language’ when I gloss Aristotle's views. The use of ‘language’ as a mass-noun tends to beg the question in favour of deep structure.

^{page 51 note 4} For a sound treatment of Aristotle's ideas on the relation between the investigation of words and things, see Bambrough, Renford, ‘Aristotle on justice: a paradigm of philosophy’, in New essays on Plato and Aristotle, ed. Bambrough, Renford (London, 1965), pp. 159–74Google Scholar.

^{page 51 note 5} 5. 167 a 35.

^{page 52 note 1} 6. 168 b 11–16.

^{page 52 note 2} 1. 165 a 6–13.

^{page 52 note 3} 33. 182 b 6–12.