1. Cp. Republic, 419a, ff., 421b, 465c, ff., and 519e; see also chapter 6, especially sections II and IV.

2. I am thinking not only of the medieval attempts to arrest society, attempts that were based on the Platonic theory that the rulers are responsible for the souls, the spiritual welfare of the ruled (and on many practical devices developed by Plato in the Republic and in the Laws), but I am thinking also of many later developments.

3. I have tried, in other words, to apply as far as possible the method which I have described in my The Logic of Scientific Discovery.

5. In my story there should be ‘no villains … Crime is not interesting … It is what men do at their best, with good intentions … that really concerns us’. I have tried as far as possible to apply this methodological principle to my interpretation of Plato. (The formulation of the principle quoted in this note I have taken from G. B. Shaw’s Preface to Saint Joan; see the first sentences in the section ‘Tragedy, not Melodrama’.)

8. I suppose that what I call the ‘strain of civilization’ is similar to the phenomenon which Freud had in mind when writing Civilization and its Discontents. Toynbee speaks of a Sense of Drift (A Study of History, V, 412 ff.), but he confines it to ‘ages of disintegration’, while I find my strain very clearly expressed in Heraclitus (in fact, traces can be found in Hesiod)—long before the time when, according to Toynbee, his ‘Hellenic society’ begins to ‘disintegrate’. Meyer speaks of the disappearance of ‘The status of birth, which had determined every man’s place in life, his civil and social rights and duties, together with the security of earning his living’ (Geschichte des Altertums, III, 542). This gives an apt description of the strain in Greek society of the fifth century B.C.

9. Another profession of this kind which led to comparative intellectual independence, was that of a wandering bard. I am thinking here mainly of Xenophanes, the progressivist; cp. the paragraph on ‘Protagoreanism’ in note 7 to chapter 5. (Homer also may be a case in point.) It is clear that this profession was accessible to very few men.
I happen to have no personal interest in matters of commerce, or in commercially minded people. But the influence of commercial initiative seems to me rather important. It is hardly an accident that the oldest known civilization, that of Sumer, was, as far as we know, a commercial civilization with strong democratic features; and that the arts of writing and arithmetic, and the beginnings of science, were closely connected with its commercial life. (Cp. also text to note 24 to this chapter.)

10. Thucydides, I, 93 (I mostly follow Jowett’s translation). For the problem of Thucydides’ bias, cp. note 15 (1) to this chapter.

11. This and the next quotation: op. cit., I, 107. Thucydides’ story of the treacherous oligarchs can hardly be recognized in Meyer’s apologetic version (Gesch. d. Altertums, III, 594), in spite of the fact that he has no better sources; it is simply distorted beyond recognition. (For Meyer’s partiality, see note 15 (2) to the present chapter.)—For a similar treachery (in 479 B.C. on the eve of Plataea) cp. Plutarch’s Aristides, 13.

12. Thucydides, III, 82–84. The following conclusion of the passage is characteristic of the element of individualism and humanitarianism present in Thucydides, a member of the Great Generation (see below, and note 27 to this chapter) and, as mentioned above, a moderate: ‘When men take revenge, they are reckless; they do not consider the future, and do not hesitate to annul those common laws of humanity on which every individual must rely for his own deliverance should he ever be overtaken by calamity; they forget that in their own hour of need they will look for them in vain.’ For a further discussion of Thucydides’ bias see note 15 (1) to this chapter.

13. Aristotle, Politics, VIII, (V), 9, 10/11; 1310a. Aristotle does not agree with such open hostility; he thinks it wiser that ‘true Oligarchs should affect to be advocates of the people’s cause’; and he is anxious to give them good advice: ‘They should take, or they should at least pretend to take, the opposite line, by including in their oath the pledge: I shall do no harm to the people.’

18. It must however be kept in mind that, as the reactionaries complained. slavery was in Athens on the verge of dissolution. Cp. the evidence mentioned in notes 17, 18 and 29 to chapter 4; furthermore, notes 13 to chapter 5, 48 to chapter 8, and 27–37 to the present chapter.

20. Thucydides, VII, 28; cp. Meyer, op. cit., IV, 535. The important remark that ‘this would yield more’ enables us, of course, to fix an approximate upper limit for the ratio between the taxes previously imposed and the volume of trade.

21. This is an allusion to a grim little pun which I owe to P. Milford: ‘A Plutocracy is preferable to a Lootocracy.’

22. Plato, Republic, 423b. For the problem of keeping the size of the population constant, cp. note 7, above.

25. This is admitted by Meyer (op. cit., IV, 433 f.), who in a very interesting passage says of the two parties: ‘each of them claims that it defends “the paternal state” …, and that the opponent is infected with the modern spirit of selfishness and revolutionary violence. In reality, both are infected … The traditional customs and religion are more deeply rooted in the democratic party; its aristocratic enemies who fight under the flag of the restoration of the ancient times, are … entirely modernized.’ Cp. also op. cit., V, 4 f., 14, and the next note.

26. From Aristotle’s Athenian Constitution, ch. 34, §3, we learn that the Thirty Tyrants professed at first what appeared to Aristotle a ‘moderate’ programme, viz., that of the ‘paternal state’.—For the nihilism and the modernity of Critias, cp. his theory of religion discussed in chapter 8 (see especially note 17 to that chapter) and note 48 to the present chapter.

27. It is most interesting to contrast Sophocles’ attitude towards the new faith with that of Euripides. Sophocles complains (cp. Meyer, op. cit., IV, III): ‘It is wrong that … the lowly born should flourish, while the brave and nobly born are unfortunate.’ Euripides replies (with Antiphon; cp. note 13 to chapter 5) that the distinction between the nobly and the low born (especially slaves) is merely verbal: ‘The name alone brings shame upon the slave.’—For the humanitarian element in Thucydides, cp. the quotation in note 12 to this chapter. For the question how far the Great Generation was connected with cosmopolitan tendencies, see the evidence marshalled in note 48 to chapter 8—especially the hostile witnesses, i.e. the Old Oligarch, Plato, and Aristotle.

28. ‘Misologists’, or haters of rational argument, are compared by Socrates to ‘misanthropists’, or haters of men; cp. the Phaedo, 89c. In contrast, cp. Plato’s misanthropic remark in the Republic, 496c–d (cp. notes 57 and 58 to chapter 8).

29. The quotations in this paragraph are from Democritus’ fragments, Diels, Vorsokratiker⁵, fragments number 41; 179; 34; 261; 62; 55; 251; 247 (genuineness questioned by Diels and by Tarn, cp. note 48 to chapter 8); 118.

33. Herodotus’ work with its pro-democratic tendency (cp. for example, III, 80) appeared about a year or two after Pericles’ oration (cp. Meyer, Gesch. d. Altertums, IV, 369).

34. This has been pointed out for instance by T. Gomperz, Greek Thinkers, V, 13, 2 (Germ. edn, II, 406 f.); the passages in the Republic to which he draws attention are: 557d and 561c, ff. The similarity is undoubtedly intentional. Cp. also Adam’s edition of the Republic, vol. II, 235, note to 557d26. See also the Laws, 699d/e, ff., and 704d–707d. For a similar observation regarding Herodotus III, 80, see note 17 to chapter 6.

35. Some hold the Menexenus to be spurious, but I believe that this shows only their tendency to idealize Plato. The Menexenus is vouched for by Aristotle, who quotes a remark from it as due to the ‘Socrates of the Funeral Dialogue’ (Rhetoric, I, 9, 30 = 1367b8; and III, 14, 11 = 1415b30). See especially also end of note 19 to chapter 6; also note 48 to chapter 8 and notes 15 (1) and 61 to the present chapter.

36. The Old Oligarch’s (or the Pseudo-Xenophon’s) Constitution of Athens was published in 424 B.C. (according to Kirchhoff, quoted by Gomperz, Greek Thinkers, Germ. edn, I, 477). For its attribution to Critias, cp. J. E. Sandys, Aristotle’s Constitution of Athens, Introduction IX, especially note 3. See also notes 18 and 48 to this chapter. Its influence upon Thucydides is, I think, noticeable in the passages quoted in notes 10 and 11 to this chapter. For its influence upon Plato, see especially note 59 to chapter 8, and Laws, 704a–707d. (Cp. Aristotle, Politics, 1326b–1327a; Cicero, De Republica, II, 3 and 4.)

37. I am alluding to M. M. Rader’s book, No Compromise—The Conflict between Two Worlds (1939), an excellent criticism of the ideology of fascism.
With the allusion, later in this paragraph, to Socrates’ warning against misanthropy and misology, cp. note 28, above.

38. *(1) For the theory that what may be called ‘the invention of critical thought’ consists in the foundation of a new tradition—the tradition of critically discussing the traditional myths and theories—see my ‘Towards a Rational Theory of Tradition,’ The Rationalist Annual, 1949; now in Conjectures and Refutations. (Only such a new tradition can explain the fact that, in the Ionian School, the three first generations produced three different philosophies.)*
(2) Schools (especially Universities) have retained certain aspects of tribalism ever since. But we must think not only of their emblems, or of the Old School Tie with all its social implications of caste, etc., but also of the patriarchal and authoritarian character of so many schools. It was not just an accident that Plato, when he had failed to re-establish tribalism, founded a school instead; nor is it an accident that schools are so often bastions of reaction, and school teachers dictators in pocket edition.
As an illustration of the tribalistic character of these early schools, I give here a list of some of the taboos of the early Pythagoreans. (The list is from Burnet’s Early Greek Philosophy², 106, who takes it from Diels; cp. Vorsokratiker⁵, vol. I, pp. 97 ff.; but see also Aristoxenus’ evidence in op. cit., p. 101.) Burnet speaks of ‘genuine taboos of a thoroughly primitive type’.—To abstain from beans.—Not to pick up what has fallen.—Not to touch a white cock.—Not to break bread.—Not to step over a crossbar.—Not to stir the fire with iron.—Not to eat from a whole loaf.—Not to pluck a garland.—Not to sit on a quart measure.—Not to eat the heart.—Not to walk on highways.—Not to let the swallows share one’s roof.—When the pot is taken off the fire, not to leave the mark of it in the ashes, but to stir them together.— Not to look in a mirror beside a light.—After rising from the bedclothes, to roll them together and to smooth out the impress of the body.

39. An interesting parallelism to this development is the destruction of tribalism through the Persian conquests. This social revolution led, as Meyer points out (op. cit., vol. III, 167 ff.), to the emergence of a number of prophetic, i.e. in our terminology, of historicist, religions of destiny, degeneration, and salvation, among them that of the ‘chosen people’, the Jews (cp. chapter 1).
Some of these religions were also characterized by the doctrine that the creation of the world is not yet concluded, but still going on. This must be compared with the early Greek conception of the world as an edifice and with the Heraclitean destruction of this conception, described in chapter 2 (see note 1 to that chapter). It may be mentioned here that even Anaximander felt uneasy about the edifice. His stress upon the boundless or indeterminate or indefinite character of the building-material may have been the expression of a feeling that the building may possess no definite framework, that it may be in flux (cp. next note).
The development of the Dionysian and the Orphic mysteries in Greece is probably dependent upon the religious development of the east (cp. Herodotus, II, 81). Pythagoreanism, as is well known, had much in common with Orphic teaching, especially regarding the theory of the soul (see also note 44 below). But Pythagoreanism had a definitely ‘aristocratic’ flavour, as opposed to the Orphic teaching which represented a kind of ‘proletarian’ version of this movement. Meyer (op. cit., III, p. 428, § 246) is probably right when he describes the beginnings of philosophy as a rational counter-current against the movement of the mysteries; cp. Heraclitus’ attitude in these matters (fragm. 5, 14, 15; and 40, 129, Diels⁵; 124–129; and 16–17, Bywater). He hated the mysteries and Pythagoras; the Pythagorean Plato despised the mysteries. (Rep., 364e, f.; cp. however Adam’s Appendix IV to Book IX of the Republic, vol. II, 378 ff., of his edition.)

40. For Anaximander (cp. the preceding note), see Diels⁵, fragm. 9: ‘The origin of things … is some indeterminate (or boundless) nature; … out of those things from which existing things are generated, into these they dissolve again, by necessity. For they do penance to one another for their offence (or injustice), according to the order of time.’ That individual existence appeared to Anaximander as injustice was the interpretation of Gomperz (Greek Thinkers, Germ. edn, vol. I, p. 46; note the similarity to Plato’s theory of justice); but this interpretation has been severely criticized.

41. Parmenides was the first to seek his salvation from this strain by interpreting his dream of the arrested world as a revelation of true reality, and the world of flux in which he lived as a dream. ‘The real being is indivisible. It is always an integrated whole, which never breaks away from its order; it never disperses, and thus need not re-unite.’ (D5, fragm. 4.) For Parmenides, cp. also note 22 to chapter 3, and text.

45. In the Gorgias, which is, I believe, Socratic in parts (although the Pythagorean elements which Gomperz has noted show, I think, that it is largely Platonic; cp. note 56 to this chapter), Plato puts into the mouth of Socrates an attack on ‘the ports and ship-yards and walls’ of Athens, and on the tributes or taxes imposed upon her allies. These attacks, as they stand, are certainly Plato’s, which may explain why they sound very much like those of the oligarchs. But I think it quite possible that Socrates may have made similar remarks, in his anxiety to stress the things which, in his opinion, mattered most. But he would, I believe, have loathed the idea that his moral criticism could be turned into treacherous oligarchic propaganda against the open society, and especially, against its representative, Athens. (For the question of Socrates’ loyalty, cp. esp. note 53 to this chapter, and text.)

46. The typical figures, in Plato’s works, are Callicles and Thrasymachus. Historically, the nearest realizations are perhaps Theramenes and Critias; Alcibiades also, whose character and deeds, however, are very hard to judge.

48. Cp. Meyer, Geschichte des Altertums, V, 38 (and Xenophon’s Hellenica, II, 4, 22). In the same volume, on pp. 19–23 and 36–44 (see especially p. 36) can be found all the evidence needed for justifying the interpretation given in the text. The Cambridge Ancient History (1927, vol. V; cp. especially pp. 369 ff.) gives a very similar interpretation of the events.
It may be added that the number of full citizens killed by the Thirty during the eight months of terror approached probably 1,500, which is, as far as we know, not much less than one-tenth (probably about 8 per cent.) of the total number of full citizens left after the war, or 1 per cent. per month—an achievement hardly surpassed even in our own day.
Taylor writes of the Thirty (Socrates, Short Biographies, 1937, p. 100, note 1): ‘It is only fair to remember that these men probably “lost their heads” under the temptation presented by their situation. Critias had previously been known as a man of wide culture whose political leanings were decidedly democratic.’ I believe that this attempt to minimize the responsibility of the puppet government, and especially of Plato’s beloved uncle, must fail. We know well enough what to think of the shortlived democratic sentiments professed in those days at suitable occasions by the young aristocrats. Besides, Critias’ father (cp. Meyer, vol. IV, p. 579, and Lys., 12, 43, and 12, 66), and probably Critias himself, had belonged to the oligarchy of the Four Hundred; and Critias’ extant writings show his treacherous pro-Spartan leanings as well as his oligarchic outlook (cp., for instance, Diels⁵, 45) and his blunt nihilism (cp. note 17 to chapter 8) and his ambition (cp. Diels⁵, 15; cp. also Xenophon’s Memorabilia, I, 2, 24; and his Hellenica, II, 3, 36 and 47). But the decisive point is that he simply tried to give consistent effect to the programme of the ‘Old Oligarch’, the author of the Pseudo-Xenophontic Constitution of Athens (cp. note 36 to the present chapter): to eradicate democracy; and to make a determined attempt to do so with Spartan help, should Athens be defeated. The degree of violence used is the logical result of the situation. It does not indicate that Critias lost his head; rather, that he was very well aware of the difficulties, i.e. of the democrats’ still formidable power of resistance.
Meyer, whose great sympathy for Dionysius I proves that he is at least not prejudiced against tyrants, says about Critias (op. cit., V, p. 17), after a sketch of his amazingly opportunistic political career, that ‘he was just as unscrupulous as Lysander’, the Spartan conqueror, and therefore the appropriate head of Lysander’s puppet government.
It seems to me that there is a striking similarity between the characters of Critias, the soldier, æsthete, poet, and sceptical companion of Socrates, and of Frederick II of Prussia, called ‘the Great’, who also was a soldier, an æsthete, a poet, and a sceptical disciple of Voltaire, as well as one of the worst tyrants and most ruthless oppressors in modern history. (On Frederick, cp. W. Hegemann, Entlarvte Geschichte, 1934; see especially p. 90 on his attitude towards religion, reminiscent of that of Critias.)

49. This point is very well explained by Taylor, Socrates, Short Biographies, 1937, p. 103, who follows here Burnet’s note to Plato’s Euthyphro, 4c, 4.—The only point in which I feel inclined to deviate, but only very slightly, from Taylor’s excellent treatment (op. cit., 103, 120) of Socrates’ trial is in the interpretation of the tendencies of the charge, especially of the charge concerning the introduction of ‘novel religious practices’ (op. cit., 109 and 111 f.).

50. Evidence to show this can be found in Taylor’s Socrates, 113–115; cp. especially 115, note 1, where Aeschines, I, 173, is quoted: ‘You put Socrates the Sophist to death because he was shown to have educated Critias.’

51. It was the policy of the Thirty to implicate as many people in their acts of terrorism as they could; cp. the excellent remarks by Taylor in Socrates, 101 f. (especially note 3 to p. 101). For Chaerephon, see note 56, (5) e ₆ to the present chapter.

52. As Crossman and other do; cp. Crossman, Plato To-Day, 91/92. I agree in this point with Taylor, Socrates, 116; see also his notes 1 and 2 to that page.
That the plan of the prosecution was not to make a martyr of Socrates; that the trial could have been avoided, or managed differently, had Socrates been prepared to compromise, i.e. to leave Athens, or even to promise to keep quiet, all this seems fairly clear in view of Plato’s (or Socrates’) allusions in the Apology as well as in the Crito. (Cp. Crito, 45e and especially 52b/c, where Socrates says that he would have been permitted to emigrate had he offered to do so at the trial.)

57. There is a famous and rather puzzling passage in the Second Letter (314c): ‘There is no writing of Plato nor will there ever be. What goes by his name really belongs to Socrates turned young and handsome.’ The most likely solution of this puzzle is that the passage, if not the whole letter, is spurious. (Cp. Field, Plato and His Contemporaries, 200 f., where he gives an admirable summary of the reasons for suspecting the letter, and especially the passages ‘312d–313c and possibly down to 314c’; concerning 314c, an additional reason is, perhaps, that the forger might have intended to allude to, or to give his interpretation of, a somewhat similar remark in the Seventh Letter, 341b/c, quoted in note 32 to chapter 8.) But if for a moment we assume with Burnet (Greek Philosophy, I, 212) that the passage is genuine, then the remark ‘turned young and handsome’ certainly raises a problem, especially as it cannot be taken literally since Socrates is presented in all the Platonic dialogues as old and ugly (the only exception is the Parmenides, where he is hardly handsome, although still young). If genuine, the puzzling remark would mean that Plato quite intentionally gave an idealized and not an historical account of Socrates; and it would fit our interpretation quite well to see that Plato was indeed conscious of re-interpreting Socrates as a young and handsome aristocrat who is, of course, Plato himself. (Cp. also note 11 (2) to chapter 4, note 20 (1) to chapter 6, and note 50 (3) to chapter 8.)

58. I am quoting from the first paragraph of Davies’ and Vaughan’s introduction to their translation of the Republic. Cp. Crossman, Plato To-Day, 96.

61. Menexenus, 235b. Cp. note 35 to this chapter, and the end of note 19 to chapter 6.

62. The claim that if you want security you must give up liberty has become a mainstay of the revolt against freedom. But nothing is less true. There is, of course, no absolute security in life. But what security can be attained depends on our own watchfulness, enforced by institutions to help us watch—i.e. by democratic institutions which are devised (using Platonic language) to enable the herd to watch, and to judge, their watch-dogs.

63. With the ‘variations’ and ‘irregularities’, cp. Republic, 547a, quoted in the text to notes 39 and 40 to chapter 5. Plato’s obsession with the problems of propagation and birth control may perhaps be explained in part by the fact that he understood the implications of population growth. Indeed (cp. text to note 7 to this chapter) the ‘Fall’, the loss of the tribal paradise, is caused by a ‘natural’ or ‘original’ fault of man, as it were: by a maladjustment in his natural rate of breeding. Cp. also notes 39 (3) to ch. 5, and 34 to ch. 4. With the next quotation further below in this paragraph, cp. Republic, 566e, and text to note 20 to chapter 4.—Crossman, whose treatment of the period of tyranny in Greek history is excellent (cp. Plato To-Day, 27–30), writes: ‘Thus it was the tyrants who really created the Greek State. They broke down the old tribal organization of primitive aristocracy …’ (op. cit., 29). This explains why Plato hated tyranny, perhaps even more than freedom: cp. Republic, 577c.—(See, however, note 69 to this chapter.) His passages on tyranny, especially 565–568, are a brilliant sociological analysis of a consistent power-politics. I should like to call it the first attempt towards a logic of power. (I chose this term in analogy to F. A. von Hayek’s use of the term logic of choice for the pure economic theory.)—The logic of power is fairly simple, and has often been applied in a masterly way. The opposite kind of politics is much more difficult; partly because the logic of anti-power politics, i.e. the logic of freedom, is hardly understood yet.

64. It is well known that most of Plato’s political proposals, including the proposed communism of women and children, were ‘in the air’ in the Periclean period. Cp. the excellent summary in Adam’s edition of the Republic, vol. I, p. 354 f., *and A. D. Winspear, The Genesis of Plato’s Thought, 1940.*

65. Cp. V. Pareto, Treatise on General Sociology, §1843 (English translation: The Mind and Society, 1935, vol. III, pp. 1281); cp. note 1 to chapter 13, where the passage is quoted more fully.

66. Cp. the effect which Glaucon’s presentation of Lycophron’s theory had on Carneades (cp. note 54 to chapter 6), and later, on Hobbes. The professed ‘a-morality’ of so many Marxists is also a case in point. Leftists frequently believe in their own immorality. (This, although not much to the point, is sometimes more modest and more pleasant than the dogmatic self-righteousness of many reactionary moralists.)

68. The group-spirit of tribalism is, of course, not entirely lost. It manifests itself, for instance, in the most valuable experiences of friendship and comradeship; also, in youthful tribalistic movements like the boy-scouts (or the German Youth Movement), and in certain clubs and adult societies, as described, for instance, by Sinclair Lewis in Babbitt. The importance of this perhaps most universal of all emotional and æsthetic experiences must not be underrated. Nearly all social movements, totalitarian as well as humanitarian, are influenced by it. It plays an important rôle in war, and is one of the most powerful weapons of the revolt against freedom; admittedly also in peace, and in revolts against tyranny, but in these cases its humanitarianism is often endangered by its romantic tendencies.—A conscious and not unsuccessful attempt to revive it for the purpose of arresting society and of perpetuating a class rule seems to have been the English Public School System. (‘No one can grow up to be a good man unless his earliest years were given to noble games’ is its motto, taken from Republic, 558b.)
Another product and symptom of the loss of the tribalistic group-spirit is, of course, Plato’s emphasis upon the analogy between politics and medicine (cp. chapter 8, especially note 4), an emphasis which expresses the feeling that the body of society is sick, i.e. the feeling of strain, of drift. ‘From the time of Plato on, the minds of political philosophers seem to have recurred to this comparison between medicine and politics,’ says G. E. G. Catlin (A Study of the Principles of Politics, 1930, note to 458, where Thomas Aquinas, G. Santayana, and Dean Inge are quoted to support his statement; cp. also the quotations in op. cit., note to 37, from Mill’s Logic). Catlin also speaks most characteristically (op. cit., 459) of ‘harmony’ and of the ‘desire for protection, whether assured by the mother or by society’. (Cp. also note 18 to chapter 5.)

70. The tribal paradise is, of course, a myth (although some primitive people, most of all the Eskimos, seem to be happy enough). There may have been no sense of drift in the closed society, but there is ample evidence of other forms of fear—fear of demoniac powers behind nature. The attempt to revive this fear, and to use it against the intellectuals, the scientists, etc., characterizes many late manifestations of the revolt against freedom. It is to the credit of Plato, the disciple of Socrates, that it never occurred to him to present his enemies as the offspring of the sinister demons of darkness. In this point, he remained enlightened. He had little inclination to idealize the evil which was to him simply debased, or degenerate, or impoverished goodness. (Only in one passage in the Laws, 896e and 898c, there is what may be a suggestion of an abstract idealization of the evil.)

71. A final note may be added here in connection with my remark on the return to the beasts. Since the intrusion of Darwinism into the field of human problems (an intrusion for which Darwin should not be blamed) there have been many ‘social zoologists’ who have proved that the human race is bound to degenerate physically, because insufficient physical competition, and the possibility of protecting the body by the efforts of the mind, prevent natural selection from acting upon our bodies. The first to formulate this idea (not that he believed in it) was Samuel Butler, who wrote: ‘The one serious danger which this writer’ (an Erewhonian writer) ‘apprehended was that the machines’ (and, we may add, civilization in general) ‘would so … lessen the severity of competition, that many persons of inferior physique would escape detection and transmit their inferiority to their descendants.’ (Erewhon, 1872; cp. Everyman’s edition, p. 161.) The first as far as I know to write a bulky volume on this theme was W. Schallmayer (cp. note 65 to chapter 12), one of the founders of modern racialism. In fact, Butler’s theory has been continually rediscovered (especially by ‘biological naturalists’ in the sense of chapter 5, above). According to some modern writers (see, for example, G. H. Estabrooks, Man: The Mechanical Misfit, 1941), man made the decisive mistake when he became civilized, and especially when he began to help the weak; before this, he was an almost perfect man-beast; but civilization, with its artificial methods of protecting the weak, leads to degeneration, and therefore must ultimately destroy itself. In reply to such arguments, we should, I think, first admit that man is likely to disappear one day from this world; but we should add that this is also true of even the most perfect beasts, to say nothing of those which are only ‘almost perfect’. The theory that the human race might live a little longer if it had not made the fatal mistake of helping the weak is most questionable; but even if it were true—is mere length of survival of the race really all we want? Or is the almost perfect man-beast so eminently valuable that we should prefer a prolongation of his existence (he did exist for quite a long time, anyway) to our experiment of helping the weak?
Mankind, I believe, has not done so badly. In spite of the treason of some of its intellectual leaders, in spite of the stupefying effects of Platonic methods in education and the devastating results of propaganda, there have been some surprising successes. Many weak men have been helped, and for nearly a hundred years slavery has been practically abolished. Some say it will soon be re-introduced. I feel more optimistic; and, after all, it will depend on ourselves. But even if all this should be lost again, and even if we had to return to the almost perfect man-beast, this would not alter the fact that once upon a time (even if the time was short), slavery did disappear from the face of the earth. This achievement and its memory may, I believe, compensate some of us for all our misfits, mechanical or otherwise; and it may even compensate some of us for the fatal mistake made by our forefathers when they missed the golden opportunity of arresting all change—of returning to the cage of the closed society and establishing, for ever and ever, a perfect zoo of almost perfect monkeys.

Notes to Volume II

Notes to Chapter Eleven

1. That Aristotle’s criticism of Plato is very frequently, and in important places, unmerited, has been admitted by many students of the history of philosophy. It is one of the few points in which even the admirers of Aristotle find it difficult to defend him, since usually they are admirers of Plato as well. Zeller, to quote just one example, comments (cp. Aristotle and the Earlier Peripatetics, English translation by Costelloe and Muirhead, 1897, II, 261, n. 2), upon the distribution of land in Aristotle’s Best State: ‘There is a similar plan in Plato’s Laws, 745c seqq.; Aristotle, however, in Politics 1265b24 considers Plato’s arrangement, merely on account of a trifling difference, highly objectionable.’ A similar remark is made by G. Grote, Aristotle (Ch. XIV, end of second paragraph). In view of many criticisms of Plato which strongly suggest that envy of Plato’s originality is part of his motive, Aristotle’s much-admired solemn assurance (Nicomachean Ethics, I, 6, 1) that the sacred duty of giving preference to truth forces him to sacrifice even what is most dear to him, namely, his love for Plato, sounds to me somewhat hypocritical.

2. Cp. Th. Gomperz, Greek Thinkers (I am quoting from the German edition, III, 298, i.e. Book 7, Ch. 31, § 6). See especially Aristotle’s Politics, 1313a.
G. C. Field (in Plato and His Contemporaries, 114 f.) defends Plato and Aristotle against the ‘reproach … that, with the possibility, and, in the case of the latter, the actuality of this’ (i.e. the Macedonian conquest) ‘before their eyes, they … say nothing of these new developments’. But Field’s defence (perhaps directed against Gomperz) is unsuccessful, in spite of his strong comments upon those who make such a reproach. (Field says: ‘this criticism betrays … a singular lack of understanding.’) Of course, it is correct to claim, as Field does, ‘that a hegemony like that exercised by Macedon … was no new thing’; but Macedon was in Plato’s eyes at least half-barbarian and therefore a natural enemy. Field is also right in saying that ‘the destruction of independence by Macedon’ was not a complete one; but did Plato or Aristotle foresee that it was not to become complete? I believe that a defence like Field’s cannot possibly succeed, simply because it would have to prove too much; namely, that the significance of Macedon’s threat could not have been clear, at the time, to any observer; but this is disproved, of course, by the example of Demosthenes. The question is: why did Plato, who like Isocrates had taken some interest in pan-Hellenic nationalism (cp. notes 48–50 to chapter 8, Rep., 470, and the Eighth Letter, 353e, which Field claims to be ‘certainly genuine’) and who was apprehensive of a ‘Phœnician and Oscan’ threat to Syracuse, why did he ignore Macedon’s threat to Athens? A likely reply to the corresponding question concerning Aristotle is: because he belonged to the pro-Macedonian party. A reply in Plato’s case is suggested by Zeller (op. cit., II, 41) in his defence of Aristotle’s right to support Macedon: ‘So satisfied was Plato of the intolerable character of the existing political position that he advocated sweeping changes.’ (‘Plato’s follower’, Zeller continues, referring to Aristotle, ‘could the less evade the same convictions, since he had a keener insight into men and things …’) In other words, the answer might be that Plato’s hatred of Athenian democracy exceeded so much even his pan-Hellenic nationalism that he was, like Isocrates, looking forward to the Macedonian conquest.

3. This and the following three quotations are from Aristotle’s Politics, 1254b–1255a; 1254a; 1255a; 1260a.—See also: 1252a, f. (I, 2, 2–5); 1253b, ff. (I, 4, 386, and especially I, 5); 1313b (V, 11, 11). Furthermore: Metaphysics, 1075a, where freemen and slaves are also opposed ‘by nature’. But we find also the passage: ‘Some slaves have the souls of freemen, and others their bodies’ (Politics, 1254b). Cp. with Plato’s Timaeus, 51e, quoted in note 50 (2), to chapter 8.—For a trifling mitigation, and a typically ‘balanced judgement’ of Plato’s Laws, see Politics, 1260b: ‘Those’ (this is a somewhat typical Aristotelian way of referring to Plato) ‘are wrong who forbid us even to converse with slaves and say that we should only use the language of command; for slaves must be admonished’ (Plato had said, in Laws, 777e, that they should not be admonished) ‘even more than children.’ Zeller, in his long list of the personal virtues of Aristotle (op. cit., I, 44), mentions his ‘nobility of principles’ and his ‘benevolence to slaves’. I cannot help remembering the perhaps less noble but certainly more benevolent principle put forward much earlier by Alcidamas and Lycophron, namely, that there should be no slaves at all. W. D. Ross (Aristotle, 2nd ed., 1930, pp. 241 ff.) defends Aristotle’s attitude towards slavery by saying: ‘Where to us he seems reactionary, he may have seemed revolutionary to them’, viz., to his contemporaries. In support of this view, Ross mentions Aristotle’s doctrine that Greek should not enslave Greek. But this doctrine was hardly very revolutionary since Plato had taught it, probably half a century before Aristotle. And that Aristotle’s views were indeed reactionary can be best seen from the fact that he repeatedly finds it necessary to defend them against the doctrine that no man is a slave by nature, and further from his own testimony to the anti-slavery tendencies of the Athenian democracy.
An excellent statement on Aristotle’s Politics can be found in the beginning of Chapter XIV of G. Grote’s Aristotle, from which I quote a few sentences: ‘The scheme … of government proposed by Aristotle, in the two last books of his Politics, as representing his own ideas of something like perfection, is evidently founded upon the Republic of Plato: from whom he differs in the important circumstance of not admitting either community of property or community of wives and children. Each of these philosophers recognizes one separate class of inhabitants, relieved from all private toil and all money-getting employments, and constituting exclusively the citizens of the commonwealth. This small class is in effect the city—the commonwealth: the remaining inhabitants are not a part of the commonwealth, they are only appendages to it—indispensable indeed, but still appendages, in the same manner as slaves or cattle.’ Grote recognizes that Aristotle’s Best State, where it deviates from the Republic, largely copies Plato’s Laws. Aristotle’s dependence upon Plato is prominent even where he expresses his acquiescence in the victory of democracy; cp. especially Politics, III, 15; 11–13; 1286b (a parallel passage is IV, 13; 10; 1297b). The passage ends by saying of democracy: ‘No other form of government appears to be possible any longer’; but this result is reached by an argument that follows very closely Plato’s story of the decline and fall of the state in Books VIII–IX of the Republic; and this in spite of the fact that Aristotle criticizes Plato’s story severely (for instance in V, 12; 1316a, f.).

4. Aristotle’s use of the word ‘banausic’ in the sense of ‘professional’ or ‘money earning’ is clearly shown in Politics, VIII, 6, 3 ff. (1340b) and especially 15 f. (1341b). Every professional, for example a flute player, and of course every artisan or labourer, is ‘banausic’, that is to say, not a free man, not a citizen, even though he is not a real slave; the status of a ‘banausic’ man is one of ‘partial or limited slavery’ (Politics, I, 14; 13; 1260a/b). The word ‘banausos’ derives, I gather, from a pre-Hellenic word for ‘fire-worker’. Used as an attribute it means that a man’s origin and caste ‘disqualify him from prowess in the field’. (Cp. Greenidge, quoted by Adam in his edition of the Republic, note to 495e30.) It may be translated by ‘low-caste’, ‘cringing’, ‘degrading’, or in some contexts by ‘upstart’. Plato used the word in the same sense as Aristotle. In the Laws (741e and 743d), the term ‘banausia’ is used to describe the depraved state of a man who makes money by means other than the hereditary possession of land. See also the Republic, 495e and 590c. But if we remember the tradition that Socrates was a mason; and Xenophon’s story (Mem. II, 7); and Antisthenes’ praise of hard work; and the attitude of the Cynics; then it seems unlikely that Socrates agreed with the aristocratic prejudice that money earning must be degrading. (The Oxford English Dictionary proposes to render ‘banausic’ as ‘merely mechanical, proper to a mechanic’, and quotes Grote, Eth. Fragm., vi, 227 = Aristotle, 2nd edn, 1880, p. 545; but this rendering is much too narrow, and Grote’s passage does not justify this interpretation, which may originally rest upon a misunderstanding of Plutarch. It is interesting that in Shakespeare’s Midsummer Night’s Dream the term ‘mere mechanicals’ is used precisely in the sense of ‘banausic’ men; and this use might well be connected with a passage on Archimedes in North’s translation of the Life of Marcellus.)
In Mind, vol. 47, there is an interesting discussion between A. E. Taylor and F. M. Cornford, in which the former (pp. 197 ff.) defends his view that Plato, when speaking of ‘the god’ in a certain passage of the Timaeus, may have had in mind a ‘peasant cultivator’ who ‘serves’ by bodily labour; a view which is, I think most convincingly, criticized by Cornford (pp. 329 ff.). Plato’s attitude towards all ‘banausic’ work, and especially manual labour, bears on this problem; and when (p. 198, note) Taylor uses the argument that Plato compares his gods ‘with shepherds or sheep-dogs in charge of a flock of sheep’ (Laws, 901e, 907a), then we could point out that the activities of nomads and hunters are quite consistently considered by Plato as noble or even divine; but the sedentary ‘peasant cultivator’ is banausic and depraved. Cp. note 32 to chapter 4, and text.

5. The two passages that follow are from Politics (1337b, 4 and 5).

6. The 1939 edition of the Pocket Oxford Dictionary still says: ‘liberal … (of education) fit for a gentleman, of a general literary rather than technical kind’. This shows most clearly the everlasting power of Aristotle’s influence.
I admit that there is a serious problem of a professional education, that of narrow-mindedness. But I do not believe that a ‘literary’ education is the remedy; for it may create its own peculiar kind of narrow-mindedness, its peculiar snobbery. And in our day no man should be considered educated if he does not take an interest in science. The usual defence that an interest in electricity or stratigraphy need not be more enlightening than an interest in human affairs only betrays a complete lack of understanding of human affairs. For science is not merely a collection of facts about electricity, etc.; it is one of the most important spiritual movements of our day. Anybody who does not attempt to acquire an understanding of this movement cuts himself off from the most remarkable development in the history of human affairs. Our so-called Arts Faculties, based upon the theory that by means of a literary and historical education they introduce the student into the spiritual life of man, have therefore become obsolete in their present form. There can be no history of man which excludes a history of his intellectual struggles and achievements; and there can be no history of ideas which excludes the history of scientific ideas. But literary education has an even more serious aspect. Not only does it fail to educate the student, who is often to become a teacher, to an understanding of the greatest spiritual movement of his own day, but it also often fails to educate him to intellectual honesty. Only if the student experiences how easy it is to err, and how hard to make even a small advance in the field of knowledge, only then can he obtain a feeling for the standards of intellectual honesty, a respect for truth, and a disregard of authority and bumptiousness. But nothing is more necessary to-day than the spread of these modest intellectual virtues. ‘The mental power’, T. H. Huxley wrote in A Liberal Education, ‘which will be of most importance in your … life will be the power of seeing things as they are without regard to authority … But at school and at college, you shall know of no source of truth but authority.’ I admit that, unfortunately, this is true also of many courses in science, which by some teachers is still treated as if it was a ‘body of knowledge’, as the ancient phrase goes. But this idea will one day, I hope, disappear; for science can be taught as a fascinating part of human history—as a quickly developing growth of bold hypotheses, controlled by experiment, and by criticism. Taught in this way, as a part of the history of ‘natural philosophy’, and of the history of problems and of ideas, it could become the basis of a new liberal University education; of one whose aim, where it cannot produce experts, will be to produce at least men who can distinguish between a charlatan and an expert. This modest and liberal aim will be far beyond anything that our Arts Faculties nowadays achieve.

7. Politics, VIII, 3, 2 (1337b): ‘I must repeat over and again, that the first principle of all action is leisure.’ Previously, in VII, 15, 1 f. (1334a), we read: ‘Since the end of individuals and of states is the same … they should both contain the virtues of leisure … For the proverb says truly, “There is no leisure for slaves”.’ Cp. also the reference in note 9 to this section, and Metaphysics, 1072b23.
Concerning Aristotle’s ‘admiration and deference for the leisured classes’, cp. for example the following passage from the Politics, IV (VII), 8, 4–5 (1293b/1294a): ‘Birth and education as a rule go together with wealth … The rich are already in possession of those advantages the want of which is a temptation to crime, and hence they are called noblemen and gentlemen. Now it appears to be impossible that a state should be badly governed if the best citizens rule …’ Aristotle, however, not only admires the rich, but is also, like Plato, a racialist (cp. op. cit., III, 13, 2–3, 1283a): ‘The nobly born are citizens in a truer sense of the word than the low born … Those who come from better ancestors are likely to be better men, for nobility is excellence of race.’

8. Cp. Th. Gomperz, Greek Thinkers. (I am quoting from the German edition, vol. III, 263, i.e. book 6, ch. 27, § 7.)

9. Cp. Nicomachean Ethics, X, 7, 6. The Aristotelian phrase, ‘the good life’, seems to have caught the imagination of many modern admirers who associate with this phrase something like a ‘good life’ in the Christian sense—a life devoted to help, service, and the quest for the ‘higher values’. But this interpretation is the result of a mistaken idealization of Aristotle’s intentions; Aristotle was exclusively concerned with the ‘good life’ of feudal gentlemen, and this ‘good life’ he did not envisage as a life of good deeds, but as a life of refined leisure, spent in the pleasant company of friends who are equally well situated.

10. I do not think that even the term ‘vulgarization’ would be too strong, considering that to Aristotle himself ‘professional’ means ‘vulgar’, and considering that he certainly made a profession of Platonic philosophy. Besides, he made it dull, as even Zeller admits in the midst of his eulogy (op. cit., I, 46): ‘He cannot inspire us … at all in the same way as Plato does. His work is drier, more professional … than Plato’s has been.’

11. Plato presented in the Timaeus (42a f., 90e f., and especially 91d f.; see note 6 (7) to chapter 3) a general theory of the origin of species by way of degeneration, down from the Gods and the first man. Man first degenerates into a woman, then further to the higher and lower animals and to the plants. It is, as Gomperz says (Greek Thinkers, book 5, ch. 19, § 3; I am quoting from the German edition, vol. 11, 482), ‘a theory of descent in the literal sense or a theory of devolution, as opposed to the modern theory of evolution which, since it assumes an ascending sequence, might be called a theory of ascent.’ Plato’s mythical and possibly semi-ironical presentation of this theory of descent by degeneration makes use of the Orphic and Pythagorean theory of the transmigration of the soul. All this (and the important fact that evolutionary theories which made the lower forms precede the higher were in vogue at least as early as Empedocles) must be remembered when we hear from Aristotle that Speusippus, together with certain Pythagoreans, believed in an evolutionary theory according to which the best and most divine, which are first in rank, come last in the chronological order of development. Aristotle speaks (Met., 1072b30) of ‘those who suppose, with the Pythagoreans and Speusippus, that supreme beauty and goodness are not present in the beginning’. From this passage we may conclude, perhaps, that some Pythagoreans had used the myth of transmigration (possibly under the influence of Xenophanes) as the vehicle of a ‘theory of ascent’. This surmise is supported by Aristotle, who says (Met., 1091a34): ‘The mythologists seem to agree with some thinkers of the present day’ (an allusion, I suppose, to Speusippus) ‘… who say that the good as well as the beautiful make their appearance in nature only after nature has made some progress.’ It also seems as if Speusippus had taught that the world will in the course of its development become a Parmenidian One—an organized and fully harmonious whole. (Cp. Met., 1092a14, where a thinker who maintains that the more perfect always comes from the imperfect, is quoted as saying that ‘the One itself does not yet exist’; cp. also Met., 1091a11.) Aristotle himself consistently expresses, at the places quoted, his opposition to these ‘theories of ascent’. His argument is that it is a complete man that produces man, and that the incomplete seed is not prior to man. In view of this attitude, Zeller can hardly be right in attributing to Aristotle what is practically the Speusippian theory. (Cp. Zeller, Aristotle, etc., vol. II, 28 f. A similar interpretation is propounded by H. F. Osborn, From the Greeks to Darwin, 1908, pp. 48–56.) We may have to accept Gomperz’s interpretation, according to which Aristotle taught the eternity and invariability of the human species and at least of the higher animals. Thus his morphological orders must be interpreted as neither chronological nor genealogical. (Cp. Greek Thinkers, book 6, ch. 11, § 10, and especially ch. 13, §§ 6 f., and the notes to these passages.) But there remains, of course, the possibility that Aristotle was inconsistent in this point, as he was in many others, and that his arguments against Speusippus are due to his wish to assert his independence. See also note 6 (7) to chapter 3, and notes 2 and 4 to chapter 4.

12. Aristotle’s First Mover, that is, God, is prior in time (though he is eternal) and has the predicate of goodness. For the evidence concerning the identification of formal and final cause mentioned in this paragraph, see note 15 to this chapter.

13. For Plato’s biological teleology see Timaeus, 73a–76e. Gomperz comments rightly (Greek Thinkers, book 5, ch. 19, § 7; German edn, vol. II, 495 f.) that Plato’s teleology is only understandable if we remember that ‘animals are degenerate men, and that their organization may therefore exhibit purposes which were originally only the ends of man’.

14. For Plato’s version of the theory of the natural places, see Timaeus, 60b–63a, and especially 63b f. Aristotle adopts the theory with only minor changes and explains like Plato the ‘lightness’ and ‘heaviness’ of bodies by the ‘upward’ and ‘downward’ direction of their natural movements towards their natural places; cp. for instance Physics, 192b13; also Metaphysics, 1065b10.

15. Aristotle is not always quite definite and consistent in his statements on this problem. Thus he writes in the Metaphysics (1044a35): ‘What is the formal cause (of man)? His Essence. The final cause? His end. But perhaps these two are the same.’ In other parts of the same work he seems to be more assured of the identity between the Form and the end of a change or movement. Thus we read (1069b/1070a): ‘Everything that changes … is changed by something into something. That by which it is changed is the immediate mover; … that into which it is changed, the Form.’ And later (1070a, 9/10): ‘There are three kinds of substance: first, matter …; secondly, the nature towards which it moves; and thirdly, the particular substance which is composed of these two.’ Now since what is here called ‘nature’ is as a rule called ‘Form’ by Aristotle, and since it is here described as an end of movement, we have: Form = end.

16. For the doctrine that movement is the realization or actualization of potentialities, see for instance Metaphysics, Book IX; or 1065b17, where the term ‘buildable’ is used to describe a definite potentiality of a prospective house: ‘When the “buildable” … actually exists, then it is being built; and this is the process of building.’ Cp. also Aristotle’s Physics, 201b4 f.; furthermore, see Gomperz, op. cit., book 6, ch. 11, § 5.

17. Cp. Metaphysics, 1049b5. See further Book V, ch. IV, and especially 1015a12 f., Book VII, ch. IV, especially 1029b15.

18. For the definition of the soul as the First Entelechy, see the reference given by Zeller, op. cit., vol. II, p. 3, n. 1. For the meaning of Entelechy as formal cause, see op. cit., vol. I, 379, note 2. Aristotle’s use of this term is anything but precise. (See also Met., 1035b15.) Cp. also note 19 to chapter 5, and text.

19. For this and the next quotation see Zeller, op. cit., I, 46.

20. Cp. Politics, II, 8, 21 (1269a), with its references to Plato’s various Myths of the Earthborn (Rep., 414c; Pol., 271a; Tim., 22c; Laws, 677a).

21. Cp. Hegel, Lectures on the Philosophy of History, transl. by J. Sibree, London 1914, Introduction, 23; see also Loewenberg’s Hegel—Selections (The Modern Student’s Library), 366.—The whole Introduction, especially this and the following pages, shows clearly Hegel’s dependence upon Aristotle. That Hegel was aware of it is shown by the way in which he alluded to Aristotle on p. 59 (Loewenberg’s edition, 412).

22. Hegel, op. cit., 23 (Loewenberg’s edition, 365).

23. Cp. Caird, Hegel (Blackwood 1911), 26 f.

24. The next quotations are from the place referred to in notes 21 and 22.

25. For the following remarks, see Hegel’s Philosophical Propaedeutics, 2nd Year, Phenomenology of the Spirit, transl. by W. T. Harris (Loewenberg’s edition, 68 ff.). I deviate slightly from this translation. My remarks allude to the following interesting passages: § 23: ‘The impulse of self-consciousness’ (‘self-consciousness’ in German means also self-assertion; cp. the end of chapter 16) ‘consists in this: to realize its … “true nature” … It is therefore … active … in asserting itself externally …’ § 24: ‘Self-consciousness has in its culture, or movement, three stages: … (2) in so far as it is related to another self …: the relation of master and slave (domination and servitude).’ Hegel does not mention any other ‘relation to another self’.—We read further: ‘(3) The Relation of Master and Slave … § 32: In order to assert itself as free being and to obtain recognition as such, self-consciousness must exhibit itself to another self … § 33: … With the reciprocal demand for recognition there enters … the relation of master and slave between them … § 34: Since … each must strive to assert and prove himself … the one who prefers life to freedom enters into a condition of slavery, thereby showing that he has not the capacity’ (‘nature’ would have been Aristotle’s or Plato’s expression) ‘… for his independence … § 35: … The one who serves is devoid of selfhood and has another self in place of his own … The master, on the contrary, looks upon the servant as reduced, and upon his own individual will as preserved and elevated … § 36: The individual will of the servant … is cancelled in his fear of the master …’ etc. It is difficult to overlook an element of hysteria in this theory of human relations and their reduction to mastership and servitude. I hardly doubt that Hegel’s method of burying his thoughts under heaps of words, which one must remove in order to get to his meaning (as a comparison between my various quotations and the original may show) is one of the symptoms of his hysteria; it is a kind of escape, a way of shunning the daylight. I do not doubt that this method of his would make as excellent an object for psycho-analysis as his wild dreams of domination and submission. (It must be mentioned that Hegel’s dialectics—see the next chapter—carries him, at the end of § 36 here quoted, beyond the master–slave relation ‘to the universal will, the transition to positive freedom’. As will be seen from chapter 12 (especially sections II and IV), these terms are just euphemisms for the totalitarian state. Thus, mastership and servitude are very appropriately ‘reduced to components’ of totalitarianism.) With Hegel’s remark quoted here (cp. § 35) that the slave is the man who prefers life to freedom, compare Plato’s remark (Republic, 387a) that free men are those who fear slavery more than death. In a sense, this is true enough; those who are not prepared to fight for their freedom will lose it. But the theory which is implied by both Plato and Hegel, and which is very popular with later authors also, is that men who give in to superior force, or who do not die rather than give in to an armed gangster, are, by nature, ‘born slaves’ who do not deserve to fare better. This theory, I assert, can be held only by the most violent enemies of civilization.

26. For a criticism of Wittgenstein’s view that, while science investigates matters of fact, the business of philosophy is the clarification of meaning, see notes 46 and especially 51 and 52 to this chapter. (Cp. further, H. Gomperz, ‘The Meanings of Meaning’, in Philosophy of Science, vol. 8, 1941, especially p. 183.) For the whole problem to which this digression (down to note 54 to this chapter) is devoted, viz. the problem of methodological essentialism versus methodological nominalism, cp. notes 27–30 to chapter 3, and text; see further especially note 38 to the present chapter.

27. For Plato’s, or rather Parmenides’, distinction between knowledge and opinion (a distinction which continued to be popular with more modern writers, for example with Locke and Hobbes), see notes 22 and 26 to chapter 3, and text; further, notes 19 to chapter 5, and 25–27 to chapter 8. For Aristotle’s corresponding distinction, cp. for example Metaphysics 1039b31 and Anal. Post., I, 33 (88b30 ff.); II, 19 (100b5).
For Aristotle’s distinction between demonstrative and intuitive knowledge, see the last chapter of the Anal. Post. (II, 19, especially 100b 5–17; see also 72b 18–24, 75b31, 84a31, 90a6–91a11.) For the connection between demonstrative knowledge and the ‘causes’ of a thing which are ‘distinct from its essential nature’ and thereby require a middle term, see op. cit., II, 8 (especially 93a5, 93b26). For the analogous connection between intellectual intuition and the ‘indivisible form’ which it grasps—the indivisible essence and individual nature which is identical with its cause—see op. cit., 72b24, 77a4, 85a1, 88b35. See also op. cit., 90a31: ‘To know the nature of a thing is to know the reason why it is’ (i.e. its cause); and 93b21: ‘There are essential natures which are immediate, i.e. basic premises.’ For Aristotle’s recognition that we must stop somewhere in the regression of proofs or demonstrations, and accept certain principles without proof, see for example Metaphysics, 1006a7: ‘It is impossible to prove everything, for then there would arise an infinite regression …’ See also Anal. Post., II, 3 (90b, 18–27).
I may mention that my analysis of Aristotle’s theory of definition agrees largely with that of Grote, but partly disagrees with that of Ross. The very great difference between the interpretations of these two writers may be just indicated by two quotations, both taken from chapters devoted to the analysis of Aristotle’s Anal. Post., Book II. ‘In the second book, Aristotle turns to consider demonstration as the instrument whereby definition is reached.’ (Ross, Aristotle, 2nd edn, p. 49.) This may be contrasted with: ‘The Definition can never be demonstrated, for it declares only the essence of the subject …; whereas Demonstration assumes the essence to be known …’ (Grote, Aristotle, 2nd edn, 241; see also 240/241. Cp. also end of note 29 below.)

28. Cp. Aristotle’s Metaphysics, 1031b7 and 1031b20. See also 996b20: ‘We have knowledge of a thing if we know its essence.’

29. ‘A definition is a statement that describes the essence of a thing’ (Aristotle, Topics, I, 5, 101b36; VII, 3, 153a, 153a15, etc. See also Met., 1042a17)—‘The definition … reveals the essential nature’ (Anal. Post., II, 3, 91a1).—‘Definition is … a statement of the nature of the thing’ (93b28).—‘Only those things have essences whose formulæ are definitions.’ (Met., 1030a5 f.)—‘The essence, whose formula is a definition, is also called the substance of a thing.’ (Met., 1017b21)—‘Clearly, then, the definition is the formula of the essence …’ (Met., 1031a13).
Regarding the principles, i.e. the starting points or basic premises of proofs, we must distinguish between two kinds. (1) The logical principles (cp. Met., 996b25 ff.) and (2) the premises from which proofs must proceed and which cannot be proven in turn if an infinite regression is to be avoided (cp. note 27 to this chapter). The latter are definitions: ‘The basic premises of proofs are definitions’ (Anal. Post., II, 3, 90b23; cp. 89a17, 90a35, 90b23). See also Ross, Aristotle, p. 45/46, commenting upon Anal. Post., I, 4, 20–74a4: ‘The premises of science’, Ross writes (p. 46), ‘will, we are told, be per se in either sense (a) or sense (b).’ On the previous page we learn that a premise is necessary per se (or essentially necessary) in the senses (a) and (b) if it rests upon a definition.

30. ‘If it has a name, then there will be a formula of its meaning’, says Aristotle (Met., 1030a14; see also 1030b24); and he explains that not every formula of the meaning of a name is a definition; but if the name is one of a species of a genus, then the formula will be a definition.
It is important to note that in my use (I follow here the modern use of the word) ‘definition’ always refers to the whole definition sentence, while Aristotle (and others who follow him in this, e.g. Hobbes) sometimes uses the word also as a synonym for ‘definiens’.
Definitions are not of particulars, but only of universals (cp. Met., 1036a28) and only of essences, i.e. of something which is the species of a genus (i.e. a last differentia; cp. Met., 1038a19) and an indivisible form, see also Anal. Post. II, 13., 97b6 f.

31. That Aristotle’s treatment is not very lucid may be seen from the end of note 27 to this chapter, and from a further comparison of these two interpretations. The greatest obscurity is in Aristotle’s treatment of the way in which, by a process of induction, we rise to definitions that are principles; cp. especially Anal. Post., II, 19, pp. 100a, f.

32. For Plato’s doctrine, see notes 25–27 to chapter 8, and text.
Grote writes (Aristotle, 2nd ed., 260): ‘Aristotle had inherited from Plato his doctrine of an infallible Nous or Intellect, enjoying complete immunity from error.’ Grote continues to emphasize that, as opposed to Plato, Aristotle does not despise observational experience, but rather assigns to his Nous (i.e. intellectual intuition) ‘a position as terminus and correlate to the process of Induction’ (loc. cit., see also op. cit., p. 577). This is so; but observational experience has apparently only the function of priming and developing our intellectual intuition for its task, the intuition of the universal essence; and, indeed, nobody has ever explained how definitions, which are beyond error, can be reached by induction.

33. Aristotle’s view amounts to the same as Plato’s in so far as there is for both, in the last instance, no possible appeal to argument. All that can be done is to assert dogmatically of a certain definition that it is a true description of its essence; and if asked why this and no other description is true, all that remains is an appeal to the ‘intuition of the essence’.
Aristotle speaks of induction in at least two senses—in a more heuristic sense of a method leading us to ‘intuit the general principle’ (cp. An. Pri., 67a22f., 27b25–33, An. Post., 71a7, 81a38–b5, 100b4 f.) and in a more empirical sense (cp. An. Pri., 68b15–37, 69a16, An. Post., 78a35, 81b5 ff., Topics, 105a13, 156a4, 157a34).
A case of an apparent contradiction, which, however, might be cleared up, is 77a4, where we read that a definition is neither universal nor particular. I suggest that the solution is not that a definition is ‘not strictly a judgement at all’ (as G. R. G. Mure suggests in the Oxford translation), but that it is not simply universal but ‘commensurate’, i.e. universal and necessary. (Cp. 73b26, 96b4, 97b25.)
For the ‘argument’ of Anal. Post. mentioned in the text, see 100b6 ff. For the mystical union of the knowing and the known in De Anima, see especially 425b30 f., 430a20, 431a1; the decisive passage for our purpose is 430b27 f.: ‘The intuitive grasp of the definition … of the essence is never in error … just as … the seeing of the special object of sight can never be in error.’ For the theological passages of the Metaphysics, see especially 1072b20 (‘contact’) and 1075a2. See also notes 59 (2) to chapter 10, 36 to chapter 12, and notes 3, 4, 6, 29–32, and 58 to chapter 24.
For ‘the whole body of fact’ mentioned in the next paragraph, see the end of Anal. Post. (100b15 f.).
It is remarkable how similar the views of Hobbes (a nominalist but not a methodological nominalist) are to Aristotle’s methodological essentialism. Hobbes too believes that definitions are the basic premises of all knowledge (as opposed to opinion).

34. I have developed this view of scientific method in my Logic of Scientific Discovery; see, e.g. pp. 278 ff. and pp. 315 ff., for a fuller translation from Erkenntnis, vol. 5 (1934) where I say: ‘We shall have to get accustomed to interpreting sciences as systems of hypotheses (instead of “bodies of knowledge”), i.e. of anticipations that cannot be established, but which we use as long as they are corroborated, and of which we are not entitled to say that they are “true” or “more or less certain” or even “probable”.’

35. The quotation is from my note in Erkenntnis, vol. 3 (1933), now retranslated in The Logic of Scientific Discovery, pp. 312 ff.; it is a variation and generalization of a statement on geometry made by Einstein in his Geometry and Experience.

36. It is, of course, not possible to estimate whether theories, argument, and reasoning, or else observation and experiment, are of greater significance for science; for science is always theory tested by observation and experiment. But it is certain that all those ‘positivists’ who try to show that science is the ‘sum total of our observations’, or that it is observational rather than theoretical, are quite mistaken. The rôle of theory and argument in science can hardly be overrated.—Concerning the relation between proof and logical argument in general, see note 47 to this chapter.

37. Cp. e.g. Met., 1030a, 6 and 14 (see note 30 to this chapter).

38. I wish to emphasize that I speak here about nominalism versus essentialism in a purely methodological way. I do not take up any position towards the metaphysical problem of universals, i.e. towards the metaphysical problem of nominalism versus essentialism (a term which I suggest should be used instead of the traditional term ‘realism’); and I certainly do not advocate a metaphysical nominalism, although I advocate a methodological nominalism. (See also notes 27 and 30 to chapter 3.)
The opposition between nominalist and essentialist definitions made in the text is an attempt to reconstruct the traditional distinction between ‘verbal’ and ‘real’ definitions. My main emphasis, however, is on the question whether the definition is read from the right to the left or from the left to the right; or, in other words, whether it replaces a long story by a short one, or a short story by a long one.

39. My contention that in science only nominalist definitions occur (I speak here of explicit definitions only and neither of implicit nor of recursive definitions) needs some defence. It certainly does not imply that terms are not used more or less ‘intuitively’ in science; this is clear if only we consider that all chains of definitions must start with undefined terms, whose meaning can be exemplified but not defined. Further, it seems clear that in science, especially in mathematics, we often first use a term, for instance ‘dimension’ or ‘truth’, intuitively, but proceed later to define it. But this is a rather rough description of the situation. A more precise description would be this. Some of the undefined terms used intuitively can be sometimes replaced by defined terms of which it can be shown that they fulfil the intentions with which the undefined terms have been used; that is to say, to every sentence in which the undefined terms occurred (e.g. which was interpreted as analytic) there is a corresponding sentence in which the newly defined term occurs (which follows from the definition).
One certainly can say that K. Menger has recursively defined ‘Dimension’ or that A. Tarski has defined ‘Truth’; but this way of expressing matters may lead to misunderstandings. What has happened is that Menger gave a purely nominal definition of classes of sets of points which he labelled ‘n-dimensional’, because it was possible to replace the intuitive mathematical concept ‘n-dimensional’ by the new concept in all important contexts; and the same can be said of Tarski’s concept ‘Truth’. Tarski gave a nominal definition (or rather a method of drafting nominal definitions) which he labelled ‘Truth’, since a system of sentences could be derived from the definition corresponding to those sentences (like the law of the excluded middle) which had been used by many logicians and philosophers in connection with what they called ‘Truth’.

40. If anything, our language would gain precision if we were to avoid definitions and take the immense trouble of always using the defining terms instead of the defined terms. For there is a source of imprecision in the current methods of definition: Carnap has developed (in 1934) what appears to be the first method of avoiding inconsistencies in a language using definitions. Cp. Logical Syntax of Language, 1937, §22, p. 67. (See also Hilbert-Bernays, Grundlagen d. Math., 1939, II, p. 295, note 1.) Carnap has shown that in most cases a language admitting definitions will be inconsistent even if the definitions satisfy the general rules for forming definitions. The comparative practical unimportance of this inconsistency merely rests upon the fact that we can always eliminate the defined terms, replacing them by the defining terms.

41. Several examples of this method of introducing the new term only after the need has arisen may be found in the present book. Dealing, as it does, with philosophical positions, it can hardly avoid introducing, for the sake of brevity, names for these positions. This is the reason why I have to make use of so many ‘isms’. But in many cases these names are introduced only after the positions in question have been described.

42. In a more systematic criticism of the essentialist method, three problems might be distinguished which essentialism can neither escape nor solve. (1) The problem of distinguishing clearly between a mere verbal convention and an essentialist definition which ‘truly’ describes an essence. (2) The problem of distinguishing ‘true’ essential definitions from ‘false’ ones. (3) The problem of avoiding an infinite regression of definitions.—I shall briefly deal with the second and third of these problems only. The third of these problems will be dealt with in the text; for the second, see notes 44 (1) and 54 to this chapter.

43. The fact that a statement is true may sometimes help to explain why it appears to us as self-evident. This is the case with ‘2 + 2 = 4’, or with the sentence ‘the sun radiates light as well as heat’. But the opposite is clearly not the case. The fact that a sentence appears to some or even to all of us to be ‘self-evident’, that is to say, the fact that some or even all of us believe firmly in its truth and cannot conceive of its falsity, is no reason why it should be true. (The fact that we are unable to conceive of the falsity of a statement is in many cases only a reason for suspecting that our power of imagination is deficient or undeveloped.) It is one of the gravest mistakes if a philosophy ever offers self-evidence as an argument in favour of the truth of a sentence; yet this is done by practically all idealist philosophies. It shows that idealist philosophies are often systems of apologetics for some dogmatic beliefs.
The excuse that we are often in such a position that we must accept certain sentences for no better reason than that they are self-evident, is not valid. The principles of logic and of scientific method (especially the ‘principle of induction’ or the ‘law of uniformity of nature’) are usually mentioned as statements which we must accept, and which we cannot justify by anything but self-evidence. Even if this were so, it would be franker to say that we cannot justify them, and leave it at that. But, in fact, there is no need for a ‘principle of induction’. (Cp. my The Logic of Scientific Discovery.) And as far as the ‘principles of logic’ are concerned, much has been done in recent years which shows that the self-evidence theory is obsolete. (Cp. especially Carnap’s Logical Syntax of Language and his Introduction to Semantics.) See also note 44 (2).

44. (1) If we apply these considerations to the intellectual intuition of essences, then we can see that essentialism is unable to solve the problem: How can we find out whether or not a proposed definition which is formally correct is true also; and especially, how can we decide between two competing definitions? It is clear that for the methodological nominalist the answer to a question of this kind is trivial. For let us assume that somebody maintains (with the Oxford Dictionary) that ‘A puppy is a vain, empty-headed, impertinent young man’, and that he insists upon upholding this definition against somebody who clings to our previous definition. In this case, the nominalist, if he is patient enough to do so, will point out that a quarrel about labels does not interest him, since their choice is arbitrary; and he may suggest, if there is any danger of ambiguity, that one can easily introduce two different labels, for example ‘puppy₁’ and ‘puppy₂’. And if a third party should support that ‘A puppy is a brown dog’, then the nominalist will patiently suggest the introduction of the label ‘puppy₃’. But should the contesting parties continue to quarrel, either because somebody insists that only his puppy is the legitimate one, or because he insists that his puppy must, at least, be labelled ‘puppy₁’, then even a very patient nominalist would only shrug his shoulders. (In order to avoid misunderstandings, it should be said that methodological nominalism does not discuss the question of the existence of universals; Hobbes, accordingly, is not a methodological nominalist, but what I should call an ontological nominalist.)
The same trivial problem, however, raises insurmountable difficulties for the essentialist method. We have already supposed that the essentialist insists that, for instance, ‘A puppy is a brown dog’ is not a correct definition of the essence of ‘puppiness’. How can he defend this view? Only by an appeal to his intellectual intuition of essences. But this fact has the practical consequence that the essentialist is reduced to complete helplessness, if his definition is challenged. For there are only two ways in which he can react. The one is to reiterate stubbornly that his intellectual intuition is the only true one, to which, of course, his opponent may reply in the same way, so that we reach a deadlock instead of the absolutely final and indubitable knowledge which we were promised by Aristotle. The other is to admit that his opponent’s intuition may be as true as his own, but that it is of a different essence, which he unfortunately denotes by the same name. This would lead to the suggestion that two different names should be used for the two different essences, for example ‘puppy₁’ and ‘puppy₁’. But this step means giving up the essentialist position altogether. For it means that we start with the defining formula and attach to it some label, i.e. that we proceed ‘from the right to the left’; and it means that we shall have to attach these labels arbitrarily. This can be seen by considering that the attempt to insist that a puppy₁ is, essentially, a young dog, while the brown dog can only be a puppy₂, would clearly lead to the same difficulty which has driven the essentialist into his present dilemma. Accordingly, every definition must be considered as equally admissible (provided it is formally correct); which means, in Aristotelian terminology, that one basic premise is just as true as another (which is contrary to it) and that ‘it is impossible to make a false statement’. (This seems to have been pointed out by Antisthenes; see note 54 to this chapter.) Thus the Aristotelian claim that intellectual intuition is a source of knowledge as opposed to opinion, unerringly and indubitably true, and that it furnishes us with definitions which are the safe and necessary basic premises of all scientific deduction, is baseless in every single one of its points. And a definition turns out to be nothing but a sentence which tells us that the defined term means the same as the defining formula, and that each can be replaced by the other. Its nominalist use permits us to cut a long story short and is therefore of some practical advantage. But its essen-tialist use can only help us to replace a short story by a story which means the same but is much longer. This use can only encourage verbalism.
(2) For a criticism of Husserl’s intuition of essences, cp. J. Kraft, From Husserl to Heidegger (in German, 1932). See also note 8 to chapter 24. Of all authors who hold related views, M. Weber had probably the greatest influence upon the treatment of sociological problems. He advocated for the social sciences a ‘method of intuitive understanding’; and his ‘ideal types’ largely correspond to the essences of Aristotle and Husserl. It is worth mentioning that Weber saw, in spite of these tendencies, the inadmissibility of appeals to self-evidence. ‘The fact that an interpretation possesses a high degree of self-evidence proves in itself nothing about its empirical validity’ (Ges. Aufsaetze, 1922, p. 404); and he says quite rightly that intuitive understanding ‘must always be controlled by ordinary methods’. (Loc. cit., italics mine.) But if that is so, then it is not a characteristic method of a science of ‘human behaviour’ as he thinks; it also belongs to mathematics, physics, etc. And it turns out that those who believe that intuitive understanding is a method peculiar to sciences of ‘human behaviour’ hold such views mainly because they cannot imagine that a mathematician or a physicist could become so well acquainted with his object that he could ‘get the feel of it’, in the way in which a sociologist ‘gets the feel’ of human behaviour.

45. ‘Science assumes the definitions of all its terms …’ (Ross, Aristotle, 44; cp. Anal. Post., I, 2); see also note 30 to this chapter.

46. The following quotation is from R. H. S. Crossman, Plato To-Day 1937, pp. 71 f.
A very similar doctrine is expressed by M. R. Cohen and E. Nagel in their book, An Introduction to Logic and Scientific Method (1936), p. 232: ‘Many of the disputes about the true nature of property, of religion, of law, … would assuredly disappear if the precisely defined equivalents were substituted for these words.’ (See also notes 48 and 49 to this chapter.)
The views concerning this problem expressed by Wittgenstein in his Tractatus Logico-Philosophicus (1921/22) and by several of his followers are not as definite as those of Crossman, Cohen, and Nagel. Wittgenstein is an anti-metaphysician. ‘The book’, he writes in the preface, ‘deals with the problems of philosophy and shows, I believe, that the method of formulating these problems rests on the misunderstanding of the logic of our language.’ He tries to show that metaphysics is ‘simply nonsense’ and tries to draw a limit, in our language, between sense and nonsense: ‘The limit can … be drawn in languages and what lies on the other side of the limit will be simply nonsense.’ According to Wittgenstein’s book, propositions have sense. They are true or false. Philosophical propositions do not exist; they only look like propositions, but are, in fact, nonsensical. The limit between sense and nonsense coincides with that between natural science and philosophy: ‘The totality of true propositions is the total natural science (or the totality of the natural sciences).—Philosophy is not one of the natural sciences.’ The true task of philosophy, therefore, is not to formulate propositions; it is, rather, to clarify propositions: ‘The result of philosophy is not a number of “philosophical propositions”, but to make propositions clear.’ Those who do not see that, and propound philosophical propositions, talk metaphysical nonsense.
(It should be remembered, in this connection, that a sharp distinction between meaningful statements which have sense, and meaningless linguistic expressions which may look like statements but which are without sense, was first made by Russell in his attempt to solve the problems raised by the paradoxes which he had discovered. Russell’s division of expressions which look like statements is threefold, since statements which may be true or false, and meaningless or nonsensical pseudo-statements, may be distinguished. It is important to note that this use of the terms ‘meaningless’ or ‘senseless’ partly agrees with ordinary use, but is much sharper, since ordinarily one often calls real statements ‘meaningless’, for example, if they are ‘absurd’, i.e. self-contradictory, or obviously false. Thus a statement asserting of a certain physical body that it is at the same time in two different places is not meaningless but a false statement, or one which contradicts the use of the term ‘body’ in classical physics; and similarly, a statement asserting of a certain electron that it has a precise place and momentum is not meaningless—as some physicists have asserted, and as some philosophers have repeated—but it simply contradicts modern physics.)
What has been said so far can be summed up as follows. Wittgenstein looks for a line of demarcation between sense and nonsense, and finds that this demarcation coincides with that between science and metaphysics, i.e. between scientific sentences and philosophical pseudo-propositions. (That he wrongly identifies the sphere of the natural sciences with that of true sentences shall not concern us here; see, however, note 51 to this chapter.) This interpretation of his aim is corroborated when we read: ‘Philosophy limits the … sphere of natural science.’ (All sentences so far quoted are from pp. 75 and 77.)
How is the line of demarcation ultimately drawn? How can ‘science’ be distinguished from ‘metaphysics’, and thereby ‘sense’ from ‘nonsense’? It is the reply given to this question which establishes the similarity between Wittgenstein’s theory and that of Crossman and the rest. Wittgenstein implies that the terms or ‘signs’ used by scientists have meaning, while the metaphysician ‘has given no meaning to certain signs in his propositions’; this is what he writes (pp. 187 and 189): ‘The right method of philosophy would be this. To say nothing except what can be said, i.e. the propositions of natural science, i.e. something that has nothing to do with philosophy: and then always when someone else wished to say something metaphysical, to demonstrate to him that he had given no meaning to certain signs in his propositions.’ In practice, this implies that we should proceed by asking the metaphysician: ‘What do you mean by this word? What do you mean by that word?’ In other words, we demand a definition from him; and if it is not forthcoming, we assume that the word is meaningless.
This theory, as will be shown in the text, overlooks the facts (a) that a witty and unscrupulous metaphysician every time he is asked, ‘What do you mean by this word?’, will quickly proffer a definition, so that the whole game develops into a trial of patience; (b) that the natural scientist is in no better logical position than the metaphysician; and even, if compared with a metaphysician who is unscrupulous, in a worse position.
It may be remarked that Schlick, in Erkenntnis, 1, p. 8, where he deals with Wittgenstein’s doctrine, mentions the difficulty of an infinite regress; but the solution he suggests (which seems to lie in the direction of inductive definitions or ‘constitutions’, or perhaps of operationalism; cp. note 50 to this chapter) is neither clear nor able to solve the problem of demarcation. I think that certain of the intentions of Wittgenstein and Schlick in demanding a philosophy of meaning are fulfilled by that logical theory which Tarski has called ‘Semantics’. But I also believe that the correspondence between these intentions and Semantics does not go far; for Semantics propounds propositions; it does not only ‘clarify’ them.—These comments upon Wittgenstein are continued in notes 51–52 to the present chapter. (See also notes 8 (2) and 32 to chapter 24; and 10 and 25 to chapter 25.)

47. It is important to distinguish between a logical deduction in general, and a proof or demonstration in particular. A proof or demonstration is a deductive argument by which the truth of the conclusion is finally established; this is how Aristotle uses the term, demanding (for example, in Anal. Post., I, 4, pp. 73a, ff.) that the ‘necessary’ truth of the conclusion should be established; and this is how Carnap uses the term (see especially Logical Syntax, § 10, p. 29, § 47, p. 171), showing that conclusions which are ‘demonstrable’ in this sense are ‘analytically’ true. (Into the problems concerning the terms ‘analytic’ and ‘synthetic’, I shall not enter here.)
Since Aristotle, it has been clear that not all logical deductions are proofs (i.e. demonstrations); there are also logical deductions which are not proofs; for example, we can deduce conclusions from admittedly false premises, and such deductions are not called proofs. Non-demonstrative deductions are called by Carnap ‘derivations’ (loc. cit.). It is interesting that a name for these non-demonstrative deductions has not been introduced earlier; it shows the preoccupation with proofs, a preoccupation which arose from the Aristotelian prejudice that ‘science’ or ‘scientific knowledge’ must establish all its statements, i.e. accept them either as self-evident premises, or prove them. But the position is this. Outside of pure logic and pure mathematics nothing can be proved. Arguments in other sciences (and even some within mathematics, as I. Lakatos has shown) are not proofs but merely derivations.
It may be remarked that there is a far-reaching parallelism between the problems of derivation on the one side and definition on the other, and between the problems of the truth of sentences and that of the meaning of terms.
A derivation starts with premises and leads to a conclusion; a definition starts (if we read it from the right to the left) with the defining terms and leads to a defined term. A derivation informs us about the truth of the conclusion, provided we are informed about the truth of the premises; a definition informs us about the meaning of the defined term, provided we are informed about the meaning of the defining terms. Thus a derivation shifts the problem of truth back to the premises, without ever being able to solve it; and a definition shifts the problem of meaning back to the defining terms, without ever being able to solve it.

48. The reason why the defining terms are likely to be rather less clear and precise than the defined terms is that they are as a rule more abstract and general. This is not necessarily true if certain modern methods of definition are employed (‘definition by abstraction’, a method of symbolic logic); but it is certainly true of all those definitions which Crossman can have in mind, and especially of all Aristotelian definitions (by genus and differentia).
It has been held by some positivists, especially under the influence of Locke and Hume, that it is possible to define abstract terms like those of science or of politics (see text to next note) in terms of particular, concrete observations or even of sensations. Such an ‘inductive’ method of definition has been called by Carnap ‘constitution’. But we can say that it is impossible to ‘constitute’ universals in terms of particulars. (With this, cp. my The Logic of Scientific Discovery, especially sections 14, pp. 64 ff., and 25, p. 93; and Carnap’s ‘Testability and Meaning’, in Philosophy of Science, vol. 3, 1936, pp. 419 ff., and vol. 4, pp. I ff.)

49. The examples are the same as those which Cohen and Nagel, op. cit., 232 f., recommend for definition. (Cp. note 46 to this chapter.)
Some general remarks on the uselessness of essentialist definitions may be added here. (Cp. also end of note 44 (1) to this chapter.)
(1) The attempt to solve a factual problem by reference to definitions usually means the substitution of a merely verbal problem for the factual one. (There is an excellent example of this method in Aristotle’s Physics, II, 6, towards the end.) This may be shown for the following examples. (a) There is a factual problem: Can we return to the cage of tribalism? And by what means? (b) There is a moral problem: Should we return to the cage?
The philosopher of meaning, if faced by (a) or (b), will say: It all depends on what you mean by your vague terms; tell me how you define ‘return’, ‘cage’, ‘tribalism’, and with the help of these definitions I may be able to decide your problem. Against this, I maintain that if the decision can be made with the help of the definitions, if it follows from the definitions, then the problem so decided was merely a verbal problem; for it has been solved independently of facts or of moral decisions.
(2) An essentialist philosopher of meaning may do even worse, especially in connection with problem (b); he may suggest, for example, that it depends upon ‘the essence’ or ‘the essential character’ or perhaps upon ‘the destiny’ of our civilization whether or not we should try to return. (See also note 61 (2) to this chapter.)
(3) Essentialism and the theory of definition have led to an amazing development in Ethics. The development is one of increasing abstraction and loss of touch with the basis of all ethics—the practical moral problems, to be decided by us here and now. It leads first to the general question, ‘What is good?’ or ‘What is the Good?’; next to ‘What does “Good” mean?’ and next to ‘Can the problem “What does ‘Good’ mean?” be answered?’ or ‘Can “good” be defined?’ G. E. Moore, who raised this last problem in his Principia Ethica, was certainly right in insisting that ‘good’ in the moral sense cannot be defined in ‘naturalistic’ terms. For, indeed, if we could, it would mean something like ‘bitter’ or ‘sweet’ or ‘green’ or ‘red’; and it would be utterly irrelevant from the point of view of morality. Just as we need not attain the bitter, or the sweet, etc., there would be no reason to take any moral interest in a naturalistic ‘good’. But although Moore was right in what is perhaps justly considered his main point, it may be held that an analysis of good or of any other concept or essence can in no way contribute to an ethical theory which bears upon the only relevant basis of all ethics, the immediate moral problem that must be solved here and now. Such an analysis can lead only to the substitution of a verbal problem for a moral one. (Cp. also note 18 (1) to chapter 5, especially upon the irrelevance of moral judgements.)

50. I have in mind the methods of ‘constitution’ (see note 48 to this chapter), ‘implicit definition’, ‘definition by correlation’, and ‘operational definition’. The arguments of the ‘operationalists’ seem to be in the main true enough; but they cannot get over the fact that in their operational definitions, or descriptions, they need universal terms which have to be taken as undefined; and to them, the problem applies again.
A few hints or allusions may be added here concerning the way we ‘use our terms’. For the sake of brevity, these hints will refer without explanation to certain technicalities; they may therefore, in the present form, not be generally understandable.
Of the so-called implicit definitions, especially in mathematics, Carnap has shown (Symposion I, 1927, 355 ff.; cp. also his Abriss) that they do not ‘define’ in the ordinary sense of this word; a system of implicit definitions cannot be considered as defining a ‘model’, but it defines a whole class of ‘models’. Accordingly, the system of symbols defined by a system of implicit definitions cannot be considered as a system of constants, but they must be considered as variables (with a definite range, and bound by the system in a certain way to one another). I believe that there is a limited analogy between this situation and the way we ‘use our terms’ in science. The analogy can be described in this way. In a branch of mathematics in which we operate with signs defined by implicit definition, the fact that these signs have no ‘definite meaning’ does not affect our operating with them, or the precision of our theories. Why is that so? Because we do not overburden the signs. We do not attach a ‘meaning’ to them, beyond that shadow of a meaning that is warranted by our implicit definitions. (And if we attach to them an intuitive meaning, then we are careful to treat this as a private auxiliary device, which must not interfere with the theory.) In this way, we try to keep, as it were, within the ‘penumbra of vagueness’ or of ambiguity, and to avoid touching the problem of the precise limits of this penumbra or range; and it turns out that we can achieve a great deal without discussing the meaning of these signs; for nothing depends on their meaning. In a similar way, I believe, we can operate with these terms whose meaning we have learned ‘operationally’. We use them, as it were, so that nothing depends upon their meaning, or as little as possible. Our ‘operational definitions’ have the advantage of helping us to shift the problem into a field in which nothing or little depends on words. Clear speaking is speaking in such a way that words do not matter.

51. Wittgenstein teaches in the Tractatus (cp. note 46 to this chapter where further cross-references are given) that philosophy cannot propound propositions, and that all philosophical propositions are in fact senseless pseudo-propositions. Closely connected with this is his doctrine that the true task of philosophy is not to propound sentences but to clarify them: ‘The object of philosophy is the logical clarification of thoughts.—Philosophy is not a theory but an activity. A philosophical work consists essentially of elucidations.’ (Op. cit., p. 77.)
The question arises whether this view is in keeping with Wittgenstein’s fundamental aim, the destruction of metaphysics by unveiling it as meaningless nonsense. In my The Logic of Scientific Discovery (see especially pp. 311 ff.), I have tried to show that Wittgenstein’s method leads to a merely verbal solution and that it must give rise, in spite of its apparent radicalism, not to the destruction or to the exclusion or even to the clear demarcation of metaphysics, but to their intrusion into the field of science, and to their confusion with science. The reasons for this are simple enough.
(1) Let us consider one of Wittgenstein’s sentences, for example, ‘philosophy is not a theory but an activity’. Surely, this is not a sentence belonging to ‘total natural science (or the totality of the natural sciences)’. Therefore, according to Wittgenstein (see note 46 to this chapter), it cannot belong to ‘the totality of true propositions’. On the other hand, it is not a false proposition either (since if it were, its negation would have to be true, and to belong to natural science). Thus we arrive at the result that it must be ‘meaningless’ or ‘senseless’ or ‘nonsensical’; and the same holds for most of Wittgenstein’s propositions. This consequence of his doctrine is recognized by Wittgenstein himself, for he writes (p. 189): ‘My propositions are elucidatory in this way: he who understands me finally recognizes them as senseless …’ The result is important. Wittgenstein’s own philosophy is senseless, and it is admitted to be so. ‘On the other hand’, as Wittgenstein says in his Preface, ‘the truth of the thoughts communicated here seems to me unassailable and definite. I am, therefore, of the opinion that the problems have in essentials been finally solved.’ This shows that we can communicate unassailably and definitely true thoughts by way of propositions which are admittedly nonsensical, and that we can solve problems ‘finally’ by propounding nonsense. (Cp. also note 8 (2, b) to chapter 24.)
Consider what this means. It means that all the metaphysical nonsense against which Bacon, Hume, Kant, and Russell have fought for centuries may now comfortably settle down, and even frankly admit that it is nonsense. (Heidegger does so; cp. note 87 to chapter 12.) For now we have a new kind of nonsense at our disposal, nonsense that communicates thoughts whose truth is unassailable and definitive; in other words, deeply significant nonsense.
I do not deny that Wittgenstein’s thoughts are unassailable and definitive. For how could one assail them? Obviously, whatever one says against them must be philosophical and therefore nonsense. And it can be dismissed as such. We are thus faced with that kind of position which I have described elsewhere, in connection with Hegel (cp. note 33 to chapter 12) as a reinforced dogmatism. ‘All you need’, I wrote in my Logik der Forschung (now translated as The Logic of Scientific Discovery: see p. 51), p. 21, ‘is to determine the conception of “sense” or of “meaning” in a suitably narrow way, and you can say of all uncomfortable questions that you cannot find any “sense” or “meaning” in them. By recognizing the problems of natural science alone as “meaningful”, every debate about the concept of meaning must become nonsensical. Once enthroned, the dogma of meaning is for ever raised above the possibility of attack. It is “unassailable and definitive”.’
(2) But not only does Wittgenstein’s theory invite every kind of metaphysical nonsense to pose as deeply significant; it also blurs what I have called (op. cit., p. 7) the problem of demarcation. This he does because of his naïve idea that there is something ‘essentially’ or ‘by nature’ scientific and something ‘essentially’ or ‘by nature’ metaphysical and that it is our task to discover the ‘natural’ demarcation between these two. ‘Positivism’, I may quote myself again (op. cit., p. 8), ‘interprets the problem of demarcation in a naturalistic way; instead of interpreting this question as one to be decided according to practical usefulness, it asks for a difference that exists “by nature”, as it were, between natural science and metaphysics.’ But it is clear that the philosophical or methodological task can only be to suggest and to devise a useful demarcation between these two. This can hardly be done by characterizing metaphysics as ‘senseless’ or ‘meaningless’. First, because these terms are better fitted for giving vent to one’s personal indignation about metaphysicians and metaphysical systems than for a technical characterization of a line of demarcation. Secondly, because the problem is only shifted, for we must now ask: ‘What do “meaningful” and “meaningless” mean?’ If ‘meaningful’ is only an equivalent for ‘scientific’, and ‘meaningless’ for ‘non-scientific’, then we have clearly made no progress. For reasons such as these I suggested (op. cit., 8 ff., 21 f., 227) that we eliminate the emotive terms ‘meaning’, ‘meaningful’, ‘meaningless’, etc., from the methodological discussion altogether. (Recommending that we solve the problem of demarcation by using falsifiability or testability, or degrees of testability, as criterion of the empirical character of a scientific system, I suggested that it was of no advantage to introduce ‘meaningful’ as an emotive equivalent of ‘testable’.) *In spite of my explicit refusal to regard falsifiability or testability (or anything else) as a ‘criterion of meaning’, I find that philosophers frequently attribute to me the proposal to adopt this as a criterion of meaning or of ‘meaningfulness’. (See, for example, Philosophic Thought in France and in the United States, edited by M. Farber, 1950, p. 570.)*
But even if we eliminate all reference to ‘meaning’ or ‘sense’ from Wittgenstein’s theories, his solution of the problem of demarcating science from metaphysics remains most unfortunate. For since he identifies ‘the totality of true propositions’ with the totality of natural science, he excludes all those hypotheses from ‘the sphere of natural science’ which are not true. And since we can never know of a hypothesis whether or not it is true, we can never know whether or not it belongs to the sphere of natural science. The same unfortunate result, namely, a demarcation that excludes all hypotheses from the sphere of natural science, and therefore includes them in the field of metaphysics, is attained by Wittgenstein’s famous ‘principle of verification’, as I pointed out in Erkenntnis, 3 (1933), p. 427. (For a hypothesis is, strictly speaking, not verifiable, and if we speak loosely, then we can say that even a metaphysical system like that of the early atomists has been verified.) Again, this conclusion has been drawn in later years by Wittgenstein himself, who, according to Schlick (cp. my The Logic of Scientific Discovery, note 7 to section 4), asserted in 1931 that scientific theories are ‘not really propositions’, i.e. not meaningful. Theories, hypotheses, that is to say, the most important of all scientific utterances, are thus thrown out of the temple of natural science, and therefore put on a level with metaphysics.
Wittgenstein’s original view in the Tractatus can only be explained by the assumption that he overlooked the difficulties connected with the status of a scientific hypothesis which always goes far beyond a simple enunciation of fact; he overlooked the problem of universality or generality. In this, he followed in the footsteps of earlier positivists, notably of Comte, who wrote (cp. his Early Essays on Social Philosophy, edited by H. D. Hutton, 1911, p. 223; see F. A. von Hayek, Economica, VIII, 1941, p. 300): ‘Observation of facts is the only solid basis of human knowledge … a proposition which does not admit of being reduced to a simple enunciation of fact, special or general, can have no real and intelligible sense.’ Comte, although he remained unaware of the gravity of the problem hidden behind the simple phrases ‘general fact’, at least mentions this problem, by inserting the words ‘special or general’. If we omit these words, then the passage becomes a very clear and concise formulation of Wittgenstein’s fundamental criterion of sense or meaning, as formulated by him in the Tractatus (all propositions are truth-functions of, and therefore reducible to, atomic propositions, i.e. pictures of atomic facts), and as expounded by Schlick in 1931.—Comte’s criterion of meaning was adopted by J. S. Mill.
To sum up. The anti-metaphysical theory of meaning in Wittgenstein’s Tractatus, far from helping to combat metaphysical dogmatism and oracular philosophy, represents a reinforced dogmatism that opens wide the door to the enemy, deeply significant metaphysical nonsense, and throws out, by the same door, the best friend, that is to say, scientific hypothesis.

52. It appears that irrationalism in the sense of a doctrine or creed that does not propound connected and debatable arguments but rather propounds aphorisms and dogmatic statements which must be ‘understood’ or else left alone, will generally tend to become the property of an esoteric circle of the initiated. And, indeed, this prognosis seems to be partly corroborated by some of the publications that come from Wittgenstein’s school. (I do not wish to generalize; for example, everything I have seen of F. Waismann’s writing is presented as a chain of rational and exceedingly clear arguments, and entirely free from the attitude of ‘take it or leave it’.)
Some of these esoteric publications seem to be without a serious problem; to me, they appear to be subtle for subtlety’s sake. It is significant that they come from a school which started by denouncing philosophy for the barren subtlety of its attempts to deal with pseudo-problems.
I may end this criticism by stating briefly that I do not think that there is much justification for fighting metaphysics in general, or that anything worth while will result from such a fight. It is necessary to solve the problem of the demarcation of science from metaphysics. But we should recognize that many metaphysical systems have led to important scientific results. I mention only the system of Democritus; and that of Schopenhauer which is very similar to that of Freud. And some, for instance those of Plato or Malebranche or Schopenhauer, are beautiful structures of thought. But I believe, at the same time, that we should fight those metaphysical systems which tend to bewitch and to confuse us. But clearly, we should do the same even with un-metaphysical and anti-metaphysical systems, if they exhibit this dangerous tendency. And I think that we cannot do this at one stroke. We have rather to take the trouble to analyse the systems in some detail; we must show that we understand what the author means, but that what he means is not worth the effort to understand it. (It is characteristic of all these dogmatic systems and especially of the esoteric systems that their admirers assert of all critics that ‘they do not understand’; but these admirers forget that understanding must lead to agreement only in the case of sentences with a trivial content. In all other cases, one can understand and disagree.)

53. Cp. Schopenhauer, Grundprobleme (4th edn, 1890, p. 147). He comments upon ‘intellectually intuiting reason that makes its pronouncements from the tripod of the oracle’ (hence my term ‘oracular philosophy’); and he continues: ‘This is the origin of that philosophic method which entered the stage immediately after Kant, of this method of mystifying and imposing upon people, of deceiving them and throwing dust in their eyes—the method of windbaggery. One day this era will be recognized by the history of philosophy as the age of dishonesty.’ (Then follows the passage quoted in the text.) Concerning the irrationalist attitude of ‘take it or leave it’, cp. also text to notes 39–40 to chapter 24.

54. Plato’s theory of definition (cp. note 27 to chapter 3 and note 23 to chapter 5), which Aristotle later developed and systematized, met its main opposition (1) from Antisthenes, (2) from the school of Isocrates, especially Theopompus.
(1) Simplicius, one of the best of our sources on these very doubtful matters, presents Antisthenes (ad Arist. Categ., pp. 66b, 67b) as an opponent of Plato’s theory of Forms or Ideas, and in fact, of the doctrine of essentialism and intellectual intuition altogether. ‘I can see a horse, Plato’, Antisthenes is reported to have said, ‘but I cannot see its horseness.’ (A very similar argument is attributed by a lesser source, D.L., VI, 53, to Diogenes the Cynic, and there is no reason why the latter should not have used it too.) I think that we may rely upon Simplicius (who appears to have had access to Theophrastus), considering that Aristotle’s own testimony in the Metaphysics (especially in Met., 1043b24) squares well with this anti-essentialism of Antisthenes.
The two passages in the Metaphysics in which Aristotle mentions Antisthenes’ objection to the essentialist theory of definitions are both very interesting. In the first (Met., 1024b32) we hear that Antisthenes raised the point discussed in note 44 (1) to this chapter; that is to say, that there is no way of distinguishing between a ‘true’ and a ‘false’ definition (of ‘puppy’, for example) so that two apparently contradictory definitions would only refer to two different essences, ‘puppy₁’ and ‘puppy₂’; thus there would be no contradiction, and it would hardly be possible to speak of false sentences. ‘Antisthenes’, Aristotle writes about this criticism, ‘showed his crudity by claiming that nothing could be described except by its proper formula, one formula for one thing; from which it followed that there could be no contradiction; and almost that it was impossible to make a false statement.’ (The passage has usually been interpreted as containing Antisthenes’ positive theory, instead of his criticism of the doctrine of definition. But this interpretation neglects Aristotle’s context. The whole passage deals with the possibility of false definitions, i.e. with precisely that problem which gives rise, in view of the inadequacy of the theory of intellectual intuition, to the difficulties described in note 44 (1). And it is clear from Aristotle’s text that he is troubled by these difficulties as well as by Antisthenes’ attitude towards them.) The second passage (Met., 1043b24) also agrees with the criticism of essentialist definitions developed in the present chapter. It shows that Antisthenes attacked essentialist definitions as useless, as merely substituting a long story for a short one; and it shows further that Antisthenes very wisely admitted that, although it is useless to define, it is possible to describe or to explain a thing by referring to the similarity it bears to a thing already known, or, if it is composite, by explaining what its parts are. ‘Indeed there is’, Aristotle writes, ‘something in that difficulty which has been raised by the Antisthenians and other such-like uneducated people. They said that what a thing is’ (or the ‘what is it’ of a thing) ‘cannot be defined; for the so-called definition, they say, is nothing but a long formula. But they admit that it is possible to explain, for example of silver, what sort of a thing it is; for we may say that it is similar to tin.’ From this doctrine it would follow, Aristotle adds, ‘that it is possible to give a definition and a formula of the composite kind of things or substances, whether they are sensible things, or objects of intellectual intuition; but not of their primary parts …‘(In the sequel, Aristotle wanders off, trying to link this argument with his doctrine that a defining formula is composed of two parts, genus and differentia, which are related, and united, like matter and form.)
I have dealt here with this matter since it appears that the enemies of Antisthenes, for example Aristotle (cp. Topics, I, 104b21), cited what he said in a manner which has led to the impression that it is not Antisthenes’ criticism of essentialism but rather his positive doctrine. This impression was made possible by mixing it up with another doctrine probably held by Antisthenes; I have in mind the simple doctrine that we must speak plainly, just using each term in one meaning, and that in this way we can avoid all those difficulties whose solution is unsuccessfully attempted by the theory of definitions.
All these matters are, as mentioned before, very uncertain, owing to the scantiness of our evidence. But I think that Grote is likely to be right when he characterizes ‘this debate between Antisthenes and Plato’ as the ‘first protest of Nominalism against the doctrine of an extreme Realism’ (or in our terminology, of an extreme essentialism). Grote’s position may be thus defended against Field’s attack (Plato and His Contemporaries, 167) that it is ‘quite wrong’ to describe Antisthenes as a nominalist.
In support of my interpretation of Antisthenes, I may mention that against the scholastic theory of definitions, very similar arguments were used by Descartes (cp. The Philosophical Works, translated by Haldane and Ross, 1911, vol. I, p. 317) and, less clearly, by Locke (Essay, Book III, ch. III, § 11, to ch. IV, § 6; also ch. X, §§ 4 to 11; see especially ch. IV, § 5). Both Descartes and Locke, however, remained essentialists. Essentialism itself was attacked by Hobbes (cp. note 33 above) and by Berkeley who might be described as one of the first to hold a methodological nominalism, quite apart from his ontological nominalism; see also note 7 (2) to chapter 25.
(2) Of other critics of the Platonic-Aristotelian theory of definition, I mention only Theopompus (quoted by Epictetus, II, 17, 4–10; see Grote, Plato, I, 324). I think it likely that, as opposed to the generally accepted view, Socrates himself would not have favoured the theory of definitions; what he seems to have combated was the merely verbal solution of ethical problems; and his so-called attempted definitions of ethical terms, considering their negative results, may well be attempts to destroy verbalist prejudices.
(3) I wish to add here that in spite of all my criticism I am very ready to admit Aristotle’s merits. He is the founder of logic, and down to Principia Mathematica, all logic can be said to be an elaboration and generalization of the Aristotelian beginnings. (A new epoch in logic has indeed begun, in my opinion, though not with the so-called ‘non-Aristotelian’ or ‘multi-valued’ systems, but rather with the clear distinction between ‘object-language’ and ‘meta-language’.) Furthermore, Aristotle has the great merit of having tried to tame idealism by his common-sense approach which insists that only individual things are ‘real’ (and that their ‘forms’ and ‘matter’ are only aspects or abstractions). *Yet this very approach is responsible for the fact that Aristotle does not even attempt to solve Plato’s problem of universals (see notes 19 and 20 to chapter 3, and text), i.e., the problem of explaining why certain things resemble one another and others do not. For why should there not be as many different Aristotelian essences in things as there are things?*

55. The influence of Platonism especially upon the Gospel of St. John is clear; and this influence is less noticeable in the earlier Gospels, though I do not assert that it is absent. Nevertheless the Gospels exhibit a clearly anti-intellectualist and anti-philosophizing tendency. They avoid an appeal to philosophical speculation, and they are definitely against scholarship and dialectics, for instance, that of the ‘scribes’; but scholarship means, in this period, interpreting the scriptures in a dialectical and philosophical sense, and especially in the sense of the Neo-Platonists.

56. The problem of nationalism and the superseding of Jewish parochial tribalism by internationalism plays a most important part in the early history of Christianity; the echoes of these struggles can be found in the Acts (especially 10, 15 ff.; 11, 1–18; see also St. Matthew 3, 9, and the polemics against tribal feeding taboos in Acts 10, 10–15). It is interesting that this problem turns up together with the social problem of wealth and poverty, and with that of slavery; see Galatians 3, 28; and especially Acts 5, 1–11, where the retention of private property is described as mortal sin.
The survival in the Ghettos of eastern Europe, down to 1914 and even longer, of arrested and petrified forms of Jewish tribalism is very interesting. (Cp. the way in which the Scottish tribes attempted to cling to their tribal life.)

57. The quotation is from Toynbee, A Study of History, vol. VI, p. 202; the passage deals with the motive for the persecution of Christianity by the Roman rulers, who were usually very tolerant in matters of religion. ‘The element in Christianity’, Toynbee writes, ‘that was intolerable to the Imperial Government was the Christians’ refusal to accept the Government’s claim that it was entitled to compel its subjects to act against their conscience … So far from checking the propagation of Christianity, the martyrdoms proved the most effective agencies of conversion …’

58. For Julian’s Neo-Platonic Anti-Church with its Platonizing hierarchy, and his fight against the ‘atheists’, i.e. Christianity, cp. for example Toynbee, op. cit., V, pp. 565 and 584; I may quote a passage from J. Geffken (quoted by Toynbee, loc. cit.): ‘In Jamblichus’ (a pagan philosopher and number-mystic and founder of the Syrian school of Neo-Platonists, living about A.D. 300) ‘the individual religious experience … is eliminated. Its place is taken by a mystical church with sacraments, by a scrupulous exactness in carrying out the forms of worship, by a ritual that is closely akin to magic, and by a clergy … Julian’s ideas about the elevation of the priesthood reproduce … exactly the standpoint of Jamblichus, whose zeal for the priests, for the details of the forms of worship, and for a systematic orthodox doctrine has prepared the ground for the construction of a pagan church.’ We can recognize in these principles of the Syrian Platonist and of Julian the development of the genuine Platonic (and perhaps also late Jewish; cp. note 56 to this chapter) tendency to resist the revolutionary religion of individual conscience and humaneness by arresting all change and by introducing a rigid doctrine kept pure by a philosophic priest caste and by rigid taboos. (Cp. text to notes 14 and 18–23 to chapter 7; and chapter 8, especially text to note 34.) With Justinian’s prosecution of non-Christians and heretics and his suppression of philosophy in 529, the tables are turned; it is now Christianity which adopts totalitarian methods and the control of conscience by violence. The dark ages begin.

59. For Toynbee’s warning against an interpretation of the rise of Christianity in the sense of Pareto’s advice (for which cp. notes 65 to chapter 10 and 1 to chapter 13) see, for example, A Study of History, V, 709.

60. For Critias’ and Plato’s and Aristotle’s cynical doctrine that religion is opium for the people, cp. notes 5 to 18 (especially 15 and 18) to chapter 8. (See also Aristotle’s Topics, I, 2, 101a30 ff.) For later examples (Polybius and Strabo) see, for example, Toynbee, op. cit., V, 646 f., 561. Toynbee quotes from Polybius (Historiae, VI, 56): ‘The point in which the Roman constitution excels others most conspicuously is to be found, in my opinion, in its handling of Religion … The Romans have managed to forge the main bond of their social order … out of superstition.’ etc. And he quotes from Strabo: ‘A rabble … cannot be induced to answer to the call of Philosophic Reason … In dealing with people of that sort, you cannot do without superstition.’ etc. In view of this long series of Platonizing philosophers who teach that religion is ‘opium for the people’ I fail to see how the imputation of similar motives to Constantine can be described as anachronistic.
It may be mentioned that it is a formidable opponent of whom Toynbee says, by implication, that he lacks historical sense: Lord Acton. For he writes (cp. his History of Freedom, 1909, p. 30 f., italics mine) of Constantine’s relation to the Christians: ‘Constantine, in adopting their faith, intended neither to abandon his predecessor’s scheme of policy nor to renounce the fascinations of arbitrary authority, but to strengthen his throne with the support of a religion which had astonished the world by its power of resistance …’

61. I admire the mediæval cathedrals as much as anybody, and I am perfectly prepared to recognize the greatness and uniqueness of mediæval craftsmanship. But I believe that æstheticism must never be used as an argument against humanitarianism.
The eulogy of the Middle Ages seems to begin with the Romantic movement in Germany, and it has become fashionable with the renaissance of this Romantic movement which unfortunately we are witnessing at the present time. It is, of course, an anti-rationalist movement; it will be discussed from another point of view in chapter 24.
The two attitudes towards the Middle Ages, rationalism and anti-rationalism, correspond to two interpretations of ‘history’ (cp. chapter 25).
(1) The rationalist interpretation of history views with hope those periods in which man attempted to look upon human affairs rationally. It sees in the Great Generation and especially in Socrates, in early Christianity (down to Constantine), in the Renaissance and the period of the Enlightenment, and in modern science, parts of an often interrupted movement, the efforts of men to free themselves, to break out of the cage of the closed society, and to form an open society. It is aware that this movement does not represent a ‘law of progress’ or anything of that sort, but that it depends solely upon ourselves, and must disappear if we do not defend it against its antagonists as well as against laziness and indolence. This interpretation sees in the intervening periods dark ages with their Platonizing authorities, their hierarchies of priest and tribalist orders of knights.
A classical formulation of this interpretation has been made by Lord Acton (op. cit., p. 1; italics mine). ‘Liberty,’ he writes, ‘next to religion, has been the motive of good deeds and the common pretext of crime, from the sowing of the seed at Athens, two thousand five hundred and sixty years ago … In every age its progress has been beset by its natural enemies, by ignorance and superstition, by lust of conquest and by love of ease, by the strong man’s craving for power, and the poor man’s craving for food. During long intervals it has been utterly arrested … No obstacle has been so constant, or so difficult to overcome, as uncertainty and confusion touching the nature of true liberty. If hostile interests have wrought much injury, false ideas have wrought still more.’
It is strange how strong a feeling of darkness prevails in the dark ages. Their science and their philosophy are both obsessed by the feeling that the truth has once been known, and has been lost. This expresses itself in the belief in the lost secret of the ancient philosopher’s stone and in the ancient wisdom of astrology no less than in the belief that an idea cannot be of any value if it is new, and that every idea needs the backing of ancient authority (Aristotle and the Bible). But the men who felt that the secret key to wisdom was lost in the past were right. For this key is faith in reason, and liberty. It is the free competition of thought, which cannot exist without freedom of thought.
(2) The other interpretation agrees with Toynbee in seeing, in Greek as well as in modern rationalism (since the Renaissance), an aberration from the path of faith. ‘To the present writer’s eye’, Toynbee says (A Study of History, vol. V, pp. 6 f., note; italics mine), ‘the common element of rationalism which may be discernible in the Hellenic and Western Civilization is not so distinctive as to mark this pair of societies off from all other representatives of the species … If we regard the Christian element of our Western Civilization as being the essence of it, then our reversion to Hellenism might be taken to be, not a fulfilment of the potentialities of Western Christendom, but an aberration from the proper path of Western growth—in fact, a false step which it may or may not be possible now to retrieve.’
In contrast to Toynbee, I do not doubt for a minute that it is possible to retrieve this step and to return to the cage, to the oppressions, superstition, and pestilences, of the Middle Ages. But I believe that we had much better not do so. And I contend that what we ought to do will have to be decided by ourselves, through free decisions, and not by historicist essentialism; nor, as Toynbee holds (see also note 49 (2) to this chapter), by ‘the question of what the essential Character of the Western Civilization may be’.
(The passages here quoted from Toynbee are parts of his reply to a letter from Dr. E. Bevan; and Bevan’s letter, i.e. the first of his two letters quoted by Toynbee, seems to me to present very clearly indeed what I call the rationalist interpretation.)

62. See H. Zinsser, Rats, Lice, and History (1937), pp. 80 and 83; italics mine.
Concerning my remark in the text, at the end of this chapter, that Democritus’ science and morals still live with us, I may mention that a direct historical connection leads from Democritus and Epicurus via Lucretius not only to Gassendi but undoubtedly to Locke also. ‘Atoms and the void’ is the characteristic phrase whose presence always reveals the influence of this tradition; and as a rule, the natural philosophy of ‘atoms and the void’ goes together with the moral philosophy of an altruistic hedonism or utilitarianism. In regard to hedonism and utilitarianism, I believe that it is indeed necessary to replace their principle: maximize pleasure! by one which is probably more in keeping with the original views of Democritus and Epicurus, more modest, and much more urgent. I mean the rule: minimize pain! I believe (cp. chapters 9, 24, and 25) that it is not only impossible but very dangerous to attempt to maximize the pleasure or the happiness of the people, since such an attempt must lead to totalitarianism. But there is little doubt that most of the followers of Democritus (down to Bertrand Russell, who is still interested in atoms, geometry, and hedonism) would have little quarrel with the suggested re-formulation of their pleasure principle provided it is taken for what it is meant, and not for an ethical criterion.

Notes to Chapter Ten

Notes to Volume II

Notes to Chapter Eleven