^{page 221 note 1} I am indebted to Miss M. E. Hubbard and to Professors J. L. Ackrill and H. Lloyd-Jones for valuable criticism and advice.

^{page 221 note 2} This view is not of course Victorian at all; it is found in Vettori's edition of 1560, and often later (see Bremer, J. M., Hamartia, 1968, 69 ff.).Google Scholar

^{page 221 note 3} Dacier, A. (ed.), 1692, 190Google Scholar (cf. below, p. 226 n. 4); Butcher, S. H., Aristotle's Theory of Poetry and Fine Art (1897),Google Scholar 4th ed. 1907, 317–22 (it is Butcher's misfortune that although he makes it perfectly clear that he understands in the widest possible sense and explains why, he is constantly criticized for restricting it to the moral sense); Glanville, I. M., ‘Tragic error’, C.Q. xliii (1949), 47–57CrossRefGoogle Scholar (cf. below, p. 234 n. 3); Grube, G. M. A., Aristotle on Poetry and Style, 1958, xxivGoogle Scholar f., The Greek and Roman Critics, 5965, 79–80; Kaufmann, W., Tragedy and Philosophy, 1969, 61–2.Google Scholar

^{page 222 note 1} Bremer, op. cit. (p. 221 n. 2), 31–60. This valuable study sets out all the available evidence, with full references to earlier work.

^{page 222 note 2} Hey, O., 'AMAPTIA, Philologus (1928), 1–17, 137–63.CrossRefGoogle Scholar

^{page 222 note 3} Pol. 1336^a2; Ath. Pol. 8. 4, 16. 2; E.N. 1107^a15, 1110^b29, 1119^a34. Hey, op. cit., 140, 149, 153–4, 157. Professor Ackrill points out that in fact the use in 1110^b29 need not be a ‘moral’ use, because of the restriction .

^{page 222 note 4} Met. 1396^a21 (emended by Hey after the Latin version, 145); 1412^b28 (regarded by Hey as a colourless use, 151). Bremer, op. cit., 55 f.

^{page 223 note 1} Hey, 146 ff.; Bremer, 53 f.

^{page 223 note 2} Bremer, 54.

^{page 223 note 3} It would, however, be better supported by E.N. 2. 2 and 5–6 than by 1125^a18, since this, as Professor Ackrill points out to me, does not apply to all virtue-vice triads, but wily to the special case in which the extreme are not called because they are not particularly harmful in their effects, cf. 1123^a31–3.

^{page 223 note 4} This is not to say that Aristotle could not distinguish between acts that are wrong because they break the rules and acts that are wrong because of their motives, which is the distinction Bremer (54 n. 92) seems to be making (cf. E.N. 2. 4); or that he could not distinguish between a mistaken moral judgement and a morally wrong act due, e.g., to . The point is that he sometimes uses the same language to cover both types of situation (cf. p. 226).

^{page 223 note 5} I take to mean ‘come to be reckoned as’, ‘count as’, rather than ‘become’ (causal); but the causal interpretation is possible, and the argument is not affected. (Cf. 1108^a5 for a similar ambiguity.)

^{page 223 note 6} For Aristotle the ignorance is not purely intellectual, since rightness or wrongness about the end is inseparably connected with : cf. E.N. 6. 12–13, esp. 1144^a8, 20, 35; 1145^a5. ( is often in fact not morally neutral: it means ‘culpable ignorance’, ‘stupidity’, as opposed to plain ignorance, . Cf. E. H.F. 347, and Wilamowitz's note.)

^{page 224 note 1} Or Bywater's wake; e.g. Rostagni, ed. 1927 (2nd ed. 1945); Gudeman, ed. 1934; House, H., Aristotle's Poetics, 1956Google Scholar; Else, Gerald F., Aristotle's Poetics: the Argument, 1957Google Scholar; Jones, John, On Aristotle and Greek Tragedy, 1962;Google Scholar D. W. Lucas, ed. 1968; Dawe, R. D., ‘Some reflections on Ate and Hamartia’, H.S.G.P. lxxii (1968), 89–123Google Scholar; Bremer, op. cit.; Lloyd-Jones, H., The Justice of Zeus, 1972Google Scholar; Vickers, Brian, Towards Greek Tragedy, 1973.Google Scholar

^{page 224 note 2} Strictly Aristotle does not call these acts , but the point is well taken, since he describes their agents as .

^{page 225 note 1} Op. cit. (p. 224 R. 1), 93.

^{page 225 note 2} The point in ch. 9 is rather different, that incidents arouse more pity and fear they are manifestly (or apparently) ‘caused’, i.e. seem to be the result of some action, not (coincidences, etc., which are in principle unpredictable, Phys. B. 5–6). In E.N. 5. 8, are distinguished from in that the is within the agent, not outside, as it is if the consequence of his act is unpredictable (see p. 226, and p. 232 with n. t). So here: the fall from prosperity must manifestly result from some action.

^{page 226 note 1} The general point is fully treated by Hare, R. M.in The Language of Morals (1952). Cf. also p. 223 n. 4 above.Google Scholar

^{page 226 note 2} I am indebted to Miss Hubbard for pointing out to me this possibility.

^{page 226 note 3} See p. 228 and p. 232 n. s below.

^{page 226 note 4} Dacier, op. cit. (p. 221 n. 3), having rightly assigned Oedipus Tyrannus to the category of acts done through ignorance, spoils it by finding Oedipus' moral failings, such as they are, more important than his incest and parricide. This was intended neither by Aristotle nor by Sophocles (see p. 240 n. 2 below).

^{page 227 note 1} e.g. in his discussion of Anaxagoras he gives as examples of the elements fire and water, which are for him but not for Anaxagoras (Met. 984^a14, where see Ross).

^{page 227 note 2} Else, op. cit. (p. 224 n. 1), 391 n. 86; Webster, T. B. L., Hermes lxxxii (1954), 305, compares Antiphanes fr. 191K, where it is implied that the madness version was well known.Google Scholar

^{page 228 note 1} Story Patterns in Greek Tragedy, 1964 10, 19 (so also Vickers, op. cit. [p. 224 n. s], 63–4).Google Scholar

^{page 228 note 2} ‘Disjunction’ is a convenient term here it is of course a misnomer: see pp. 225 f.3

^{page 228 note 3} I shall render gicolScaos by ‘intentional’, by ‘intentional’, by ‘not intended’, as being more in accordance with English usage than the traditional ‘voluntary’, ‘involuntary’, ‘non-voluntary’. This translation will not always work, and I am forced to abandon it in the discussion of E.E. 2. 8 (pp. 239–40); nor does it work throughout in E.N. 3. I, where Aristotle debates whether a wrong act done by one acting is : such an act could not be ‘intentional’. In fact no pair of meaningful terms can be found to correspond exactly to and , because Aristotle's various criteria for distinguishing between them cannot always be consistently applied. (I owe this point to Professor Ackrill.)

^{page 229 note 1} Thus in real life (and this is what Aristotle is talking about in E.N.) we might pity someone simply for the mental suffering caused by an act for which he was completely exonerated and suffered no material disadvantage at all, e.g. if he ran over a child who had suddenly darted in front of his car; though Aristotle, who nowhere mentions mental suffering in connection with pity (see Cope on Rhet. 2. 8. 2), might not have allowed this extreme case.

^{page 229 note 2} 111^a1, Cf. 1109^b32.

^{page 230 note 1} Bremer, 40 f., 49.

^{page 230 note 2} Radermacher, L., Artium Scriptores, SB Wien 227, 3 (1951), 69; Bremer, 40 n. 47.Google Scholar

^{page 230 note 3} Bremer, 48.

^{page 231 note 1} Hey, op. cit. (p. 222 n. 2), 138–40.

^{page 232 note 1} This is generally taken, perhaps rightly to mean ‘when the agent is responsible fol his own ignorance’, whether or not Jackson's is read for ; the whole em phasis being on (cf. Rhet 1374^b7). M. Schofield has, however, arguer (P.C.P.S. N.S. xix [1973], 66–70) that Aristotle is not here primarily concernec with negligence, but with unintended act: for which the agent cannot plead in excuse mischances outside his control; he adopt: Burnet's rendering, ‘when the charge (i.e the charge against him) originates in him’. (For this view, cf. Rhet. ad Alex. 1427^a34–6.) The interpretation of Joachim-Rees, ‘the origin of the cause of the act qua injurious is within him’ (presumably with a similar emphasis) seems to me very difficult. The point does not affect my argument.

^{page 232 note 2} 7. 6. 1149^b4 … cf. 1145^a10 ff. and 7. 5 on the (see Glanville, op. cit. [p. 221 n. 3], 49 nn. 5 and 8).

^{page 233 note 1} Bremer, 45–54.

^{page 233 note 2} Op. cit. (p. 221 n. 3), 48 ff.

^{page 234 note 1} Glanville, 49, cl. E.N. 1110^a23.

^{page 234 note 2} See p. 252 below.

^{page 234 note 3} Op. cit. 53: ‘Too narrow a view of the meaning of here is precluded by Aristotle's list of suitable men’ (cf. pp. 226–8 above). She includes the whole range of apapria in E.N.: a mistake in the practical syllogism; wrong acts due to some kinds of ; and any act ‘not ’, cl. 1115^b15, 1119^a34. Dawe describes Glanville's article as ‘an interesting and valuable complement to Hey's discussion’ (90 n. 1). Since Glanville's views for the most part run counter to Hey's, perhaps ‘corrective’ would be a better word than ‘complement’.

^{page 235 note 1} Dawe, op. cit., passim; Bremer, dm iv-vi (N.B. p. 99 n. I).

^{page 236 note 1} Ostwald, M., ‘Aristotle on Hamartia and Sophocles’ Oedipus Tyrannus', Festschrift Kapp, 1958, 93–108.Google Scholar

^{page 236 note 2} It might be argued that made against this: that although a misdeed, as well as a mistake, could be called meaning ‘momentous’ (in its consequences), the same phrase applied to a flaw of character or disposition must mean a serious defect in the character per se, not only serious for its consequences (so M. E. Hubbard, Ancient Literary Criticism [ed. D. A. Russell and M. Winterbottom], 1972, 107 n.). But this would be wrong. A machine can have a defect which is at the same time serious (in its consequences), in that it may be dangerous, and slight, in that it is easily remedied or does not seriously impair the machine's functioning. So a relatively slight defect of character in a man, such as hastiness or irritability, may have grave consequences (through hasty decisions, etc.), and this would be properly termed .

^{page 237 note 1} Cf. Lucas on 1448^a2, cit. Vahlen⁴, 278–8 (cf. also Beiträge zu Aristoteles Poetik, 1865, 78).

^{page 237 note 2} ‘The sense of the word is sufficiently fixed by its opposite,, as well as by the equivalent expression, ’, Twining; so most modern editors.

^{page 238 note 1} For a cogent criticism of Adkins's view see Creed, J. L., ‘Moral Values in the Age of Thucydides’, C.Q. N.s xxiii (1973), 213–31.CrossRefGoogle Scholar

^{page 238 note 2} is a painful feeling caused by the spectacle of undeserved suffering (other factors may contribute, but this one is necessary). , its counterpart, is a painful feeling caused by undeserved prosperity (which in Po. 13 is called ). Both situations are unjust, and this is what disturbs us (Rhet. 2. 8–9). Excessively unjust suffering so disturbs us that we feel not pity but moral outrage. A situation is if it satisfies our moral sense: deserved suffering or deserved prosperity. (That is connected with ‘moral sentiment’ was rightly seen by Twining [ed.2, 1812, ii. 104–6] cf. Schadewaldt, , Hermes, lxxxiii (1955), 137Google Scholar. This is rather different from the fifth-century use, in which means more generally sympathy with, compassion for one's fellow men. Pohlenz's attempt to mediate between these senses [ibid. lxxxiv (1956), 59] is unsuccessful.) In this note I am much indebted to Miss Hubbard.

^{page 238 note 3} See Fritz, K. von, Antike and modern Tragödie, 1962,Google Scholar‘Tragische Schuld’, Studium Generale 8 [1955]).Google Scholar

^{page 239 note 1} Ancient views on the matter were rather different: a man's character and position in his city, and the fact that he had helped it, might be invoked as a ground for an-quitting him (Lys. 21. 15); cf. Dover, K. J.Greek Popular Morality, 1974, 292 ff. Considerations other than moral were relevant to the notion of in the fifth century (cf. Schadewaldt, op. cit., 141; Adkins, op. cit., 91–5), and possibly also in Aristotle, though here the moral element is necessary and predominant.Google Scholar

^{page 239 note 2} Identification is not always a helpful concept, nor is ‘mea res agitur’ always appropriate. It is better to speak more generally of sympathy or involvement (cf. Vickers, op. cit. [p. 224 n. 1],57–9).

^{page 239 note 3} This alienation of sympathy has of course nothing to do with Brechtian alienation; in fact the emotional involvement it implies belongs to a kind of drama Brecht was concerned to repudiate (though it is a commonplace of criticism that Brecht was a great dramatist despite his theoretical aims and not because of them: it is precisely because we respond to Galileo, for all the faults he is given, that the play has the impact it does).

^{page 239 note 4} There is some analogy to the limitation of by in its limitation by : although we tend to pity those who are like us (Rhet. 1386^a24), we do not pity the suffering of those who are closely related to us (), but find it terrible and are shocked by it (ibid. I386^a 17–23). But this clearly has nothing to do with alienation of sympathy.

^{page 240 note 1} It is not of course true of the Theatre of the Absurd; a difference which might seem to make against Jan Kott's claim (if I understand him rightly) that most Greek tragedies can be placed in this category (unpublished lecture delivered in the University of Toronto, 1972, cf. The Eating of the Gods, 1974 (1974 9, 43, tot, 145, etc.).

^{page 240 note 2} ‘On Misunderstanding the Oedipus Rex’, G. & R. xiii (1966), 37–49. The same pointis briefly made by von Fritz, op. cit. (p. 238n. 3), 8, 467–8.Google Scholar

^{page 240 note 3} Cf. the remarks of von Fritz, op. cit.,5–6.

^{page 240 note 4} Lord Hailsham, as reported in The Times, 4 Dec. 1973; cf. Paul Oestricher's letter in reply, ibid.., 8 Dec.

^{page 241 note 1} I Op. cit. (p. 224 n. I), 36–40.

^{page 242 note 1} Kleines Organon fiir das Theater (1948), nr. 1–12,Google Scholar 33–5 = Gesammelte Werke (Werkausgabe, 1967), 16. 2, 663–7, 676–8. Cf. Fritz, K. von, Antike u. moderne Tragödie, 1962, x-xiv.CrossRefGoogle Scholar

^{page 242 note 2} The distinction has often been blurred, from the Frogs onwards. The plays, and the myths they embody, show an overriding concern for moral or religious sanctions and their violation, which doubtless reflects a central preoccupation of early Greek society (cf., e.g., Kirk, Myth, , its Meaning and Function, 1970, 190Google Scholar ff.; The Nature of Greek Myth, 1974, passim; Vickers, op. cit. [p. 224 n. 1], 163–337). Given this preoccupation, which most societies to some extent share, it is not surprising that tragic conflict should often be projected in moral terms. This does not mean that these terms can be simply defined, or that the poet is preaching.Google Scholar

^{page 243 note 1} 681 ff., with 438 ff., 1007 ff. (Adkins op. cit. 94 n. 4). But the Philoctetes, he says is a special case, being ‘advanced’ in other respects (ibid. 85).

^{page 243 note 2} See Kannicht on Hel. 417–18.

^{page 243 note 3} Io is more pitiable because she is in. nocent (P. V. 577: that is the point of the appeal); Prometheus asks for sympathy (ibid. 274) because his punishment is out of proportion to his deserts (268). (Prometheus also considers that he deserves pity because he pitied mankind (239–41)—a co-operative virtue (), though no doubt traditional was needed to give it effect (235).) At Ag. 1526 (if Page is right to keep the paradosis) Iphigeneia merits pity because she did not deserve what Agamemnon did to her; he merits none because he deserved what he got. Antigone is in her death (S. Ant. 694) not because of her status, but because she does not deserve to die—rather she deserves a reward (699): ;

^{page 244 note 1} Dawe, 95 ff.; cf. Stallmach, J., Ate: Zur Frage des Selbst- and Weltuerstandnisses des frügriechischen Menschen, 1968, (Diss. Göttingen 1950), who leaves the question open.Google Scholar

^{page 245 note 1} Op. cit. 110f.; argued in full in Eranos, lxiv (1966), 1–21, esp. 6–13.Google Scholar

^{page 245 note 2} Cf. Lesky's, interpretation in J.H.S. lxxxvi (1966), 80–3Google Scholar (this still seems to me the most helpful account, despite the emphasis on theology criticized by Dover, , J.H.S. xciii [1973] 58–69). I do not see that the words show that the (i.e. ) in 223 refers to the decision, not the act (Dawe, 110)—they imply merely that the act was wicked, deliberate, and harmful in its consequences. Of course in Aristotelian terms the deliberation is part of the action, but Aeschylus' account need not be so precisely analytical.Google Scholar

^{page 245 note 3} See Peradotto, J., ‘The Omen of the Eagles and the ethos of Agamemnon’, Phoenix xxiii (1969), 237CrossRefGoogle Scholar–61

^{page 246 note 1} Kitto, H. D. F., Poiesis, 1966, ch. iv, esp. 577Google Scholar (cogently criticized by Easterling, P. E, B.I.C.S. xv [1968], 58–68).Google Scholar

^{page 246 note 2} The Justice of Zeus, 1971, 521–3. Pro fessor Lloyd-Jones tells me that I misunderstand his position: he holds that the curse has no special dramatic importance in the play and therefore receives no special emphasis, but none the less is essential to the scheme of the plot. I accept the correction, but would still maintain that anything essential to the plot of a Greek tragedy is always emphasized in the play.Google Scholar

^{page 246 note 3} As Lloyd-Jones maintains, op. cit.(above, n. 2), 117–19.

^{page 246 note 4} He considers only those plays in which ‘a noble person goes to his doom’.

^{page 247 note 1} See Winnington-Ingram, R. P.in Entretiens Hardt vi (1958), 175–6.Google Scholar

^{page 247 note 2} So the scene is in effect interpreted by Dodds, (C.R. xxxix [1925], 102Google Scholar–4), and—with some difference of emphasis—by Barrett (on 333–5), E. R. Schwinge (Die Verwendung der Stichomythie in den Dramen des Euripides, 1969, 182Google Scholar–4), and Segal, C. P. (Hermes xcviii [1970], 284Google Scholar ff.). But John Gould has argued, perhaps rightly (J.H.S, . xciii [1973], 86–7; cf. Vickers, op. cit. [p. 224n. 1], 359 n. 6), that by the rules of supplication Phaedra must yield at this point. Her plea of cannot then be dismissed as an excuse, and her confession cannot technically be faulted, though it is still a wrong choice in a conflict of duties. So even if the wrong choice is due to weakness, it will not be a case of , properly so called.Google Scholar

^{page 247 note 3} This motive, though certainly made clear in the text, is not stressed and should not be over-estimated. We can ask: would Phaedra have acted otherwise without this additional provocation, or does the logic of the play demand that she accuse him anyway?—and answer, perhaps, yes. But this question does not I think arise for Euripides: that her motivation is over-determined at the human level causes no more logical difficulty than its over-determination by the duality of divine and human intentions.

^{page 247 note 4} Dodds would add (3) that Hippolytus is culpably over-scrupulous in keeping his oath; but there is no warrant in the play for supposing that Euripides meant his audience to take this point.

^{page 247 note 5} How far his arrogance and vehemence would in real life be causally connected with his sexual abnormality is not in question: again, if Euripides had intended us to take this point he would have made it clear.

^{page 248 note 1} Knox, B. M. W., Y.C.S. xiii (1952), 3–4, 19 ff.Google Scholar

^{page 248 note 2} I think it likely that Teiresias was punished with all the rest, this being indicated in the gap in our text before 1330. Admittedly this cannot be proved, but lack of mention in Apsines' brief summary does not make against it. Cadmus and Teiresias are not, however, strictly tragic figures.

^{page 249 note 1} Those who think that Euripides is talking about orgiastic religions like that of Sabazius, or about mass hysteria like St, Vitus's dance, mistake the model for the concept, and miss the whole point of the play's universality. The Bacchae is about liberation and its containment, one or other of which is likely enough to destroy our own society in my lifetime.

^{page 249 note 2} Whether Pentheus' aggressive hostility to Dionysus would in real life be connected with his voyeurism is another question which does not arise, since Euripides does not invite us to consider it; any more than he invites us to connect Hippolytus' vehemence with his abnormal sexuality, see p. 247 n. 5 above.

^{page 250 note 1} That is, in E.N.; at E.E. 1225^a28 observes: ‘we say that those divinely pos sessed () are not in control or themselves: it was not within their own power to say or do what they said or did’ (cf. 233) Madness, epilepsy, etc., are regarded Aristotle in E.N. as a special kind of ( or ) to which the normal terms of moral action do no directly apply (cf. 7. 5, esp. 1149^a9–20). a court of law the plea ‘I was sent mad a god’ would doubtless not have carrie much weight, since its truth could alway be denied, or the divine visitation ascribe to some sin (see Dover, K. J., Greek Popula Morality, 1974, 151–2).Google Scholar But in the Heracle both objections are ruled out (cf. ibid. 154) (This dramatic premiss does not of toursi mean that Euripides is tilting against the author of .)

^{page 250 note 2} The strain of the labours has caused a nervous crisis (Dodds, , C.R. xliii [1929], 99)Google Scholar or an obsession (Kamerbeek, , Mnemos. 19 [1966], 15)Google Scholar or toppled a megalomaniac over the edge (Wilamowitz, Herakles ², 128: Heracles' superhuman heroism is grounded in violence, and ‘leads not to heaven but to madness’); the madness is due to an excess of black bile in a natural melancholic (Pohlenz, , Die griechische Tragodie ²;, 1954, 298–300)Google Scholar; Heracles has usurped divine functions and honours, and is struck down to protect the divine prerogative (Burnett, A. P., Catastrophe Survived, 1971, 177–8). See also n. 3 below.Google Scholar

^{page 250 note 3} Wilamowitz, op. cit., 129; Pohlenz, loc. cit.; Murray, O.C.T., app. crit. to v. 575; Blaiklock, E. M., The Male Characters of Euripides, 1952, 128–9.Google Scholar

^{page 251 note 1} Lloyd-Jones reminds me that Poseidor in the Ȯdyssey punishes Odysseus foi blinding the Cyclops, and that Athena anc Aphrodite do not brook the insult to their offered by Ajax or Hippolytus: ‘this is not the same as the justice of Zeus.’ Bur Odysseus survives, and the blinding of tin Cyclops is no subject for serious drama while Ajax and Hippolytus have done someo thing to merit the gods' anger. All Heracle: has done is to be begotten by Zeus as tin enemy of Hera (1263 ff.).

^{page 251 note 2} 1307 '; cf. Hipp. 120, Batch. 1348, Bellerophon, fr. 292.

^{page 251 note 3} ‘ and in Euripides’ Herakles', J.H.S. lxxxii (1962), 7–18.Google Scholar

^{page 251 note 4} C. Q. N. S, . xvi (1966), 209 ff.Google Scholar

^{page 251 note 5} This example, and the substance of p. 252 n. I I owe to Dr. Matthew Dickie. Cf. J. L. Creed, op. cit. (p. 238 n. 1), esp. 218 ff.; Andrewes-Dover on Thuc. 5. 505. 4.

^{page 251 note 6} Wilamowitz's (for ) is certainly right.

^{page 251 note 7} Op. cit. (n. 4 above), 212.

^{page 252 note 1} Hope is often a bad thing in archaic and classical Greek thought, as a source of delusion (cf. Cornford, F. M., Thucydides Mythistoricus, 1907, 167Google Scholar f.; Fränkel, H., Dichtung and Philosophie, 1951,Google Scholar see index b. S.V. ); but it is not always so (e.g. S. Trach. 124–6, 723–6; fr. 196; E. frr. 408, 409, 761; D. tS, 97), and in particular it is the mark of an to show courage in a tight place, cf. P. I. 8. 16 , ibid. 7. 35 . Amphitryon takes the view of the Melians at Thuc. 5. 102, which the Athenians cynically rebut; there is nothing untraditional about his attitude, as Adkins seems to think (op. cit. [P. 251 n. 4], 213). See p. 251 n. 4 above.

^{page 252 note 2} Hel. 560; see Kannicht's note.

^{page 253 note 1} Op. cit., 54–6, the conclusion of her whole argument.

^{page 253 note 2} See Glanville, 55 nn. 6, 7, cl. Rhet. 1385^b13; 1386^a34; 1382^a21.

^{page 253 note 3} The argument about fear I owe to Miss Hubbard.

^{page 253 note 4} See Coles, R. A., A New Oxyrhynchus Papyrus: the Hypothesis of Euripides' Alexandros, B.I.C.S. Suppl. no. 32 (1974).Google Scholar