to memory, so sensation depends on thinking, on mental construction (to become effective as knowledge).

It is, indeed, far more likely that Buddhist teaching made little of and passed lightly over this question of a physical basis of thought or mind. It was too closely involved with the animistic point of view-how closely we may see, for instance, in the Brihadāraṇyaka Upanishad. When King Milinda puts a similar question respecting the subject of sensations, he does so from so obviously animistic a standpoint that the sage, instead of discussing mano, or heart, with him, argues against any one central subjective factor whatever, and resolves the process of cognition into a number of connate' activities. The method itself of ranking mental activity as though it were a sixth kind of sense seems to point in the same direction, and reminds us of Hume's contention, that when he tried to catch himself' he always tumbled on some particular perception. Indeed it was, in words attributed to Gotama himself, the lesser blunder in the average man to call 'this four-elementish body' his soul than to identify the self with what is called cittam, that is, mano, that is, viññāņam.' For whereas the body was a collocation that might hold together for many years, mind, by day and by night is ever arising as one thing, ceasing as another!'3

Impermanence of conscious phenomena was one of the two grounds of the Buddhist attack. So far it was on all fours with Hume. The other ground was the presence of law, or necessary sequence in mental procedure. The Soul was conceived as an entity, not only above change, an absolute constant, but also as an entirely free agent. Both

Cf. the interesting inquiry into the various modes of association in remembering, given in Mil., pp. 78, 79, and 77, 78.

* Mil. 54. He calls it vedagu (knower), and, when cross-examined, abbhantare jivo (the living principle within).

S. ii., pp. 94-96.

grounds, be it noted, are laid down on psychological evidence on the testimony of consciousness. And both grounds were put forward by Gotama in his very first sermon. The standard formula for the latter only is reproduced in our Manual. And it is interesting to see the same argument clothed in fresh dress in the dialogue with Milinda referred to above. The point made is this: that if any one of the skandhas could be identified with a self or soul, it would, as not subject to the conditions of phenomena, act through any other faculty it chose. It would be a principle, not only of the nature of what we should call will, but also of genuine free will.3 Soul and Free Will, for the Buddhist, stand or fall together. But, he said, what we actually find is no such free agency. We only find certain organs (doors), with definite functions, natural sequence, the line of least resistance and association. Hence we conclude there is no transcendent 'knower' about us.

Here I must leave the Buddhist philosophy of mind and theory of intellection. We are only at the threshold of its problems, and it is hence not strange if we find them as baffling as, let us say, our own confused usage of many psychological terms-feeling, will, mind-about which we ourselves greatly differ, would prove to an inquiring' Buddhist. If I have not attempted to go into the crux of the sankhara-skandha, it is because neither the Manual nor its Commentary brings us any nearer to a satisfactory hypothesis. For future discussion, however, the frequent enumerations of that skandha's content, varying with every changing mood, should prove pertinent. In every direction there is very much to be done. And each addition to the texts edited brings new light. Nor can philosophic interest fail in the long-run to accumulate about a system

1 Vin. i. 14;= M. i. 138, 300; S. iii. 66; cf. iv. 34.

2 P. 257 et seq.

3 Cf. the writer's article on the Vedalla Suttas, J. R. A. S., April, 1894. • Mil., loc. cit. f

of thought which at that early time of day took up a task requiring such vigour and audacity-the task, namely, of opposing the prevailing metaphysic, not because problems of mind did not appeal to the founders of that system, but because further analysis of mind seemed to reveal a realm of law-governed phenomenal sequence for which the ready hypothesis of an unconditioned permanent Self super grammaticam was too cheap a solution.

VIII.

On the Buddhist Notions of Good, Bad, and Indeterminate.' By way of dhammā, rūpam and cittam, by way of Buddhist phenomenology and psychology, we come at last to the ethical purport of the questions in the Manual. Given a human being known to us by way of these phenomenal states, what is implied when we say that some of them are good, some bad, others neither?

The Dhamma-Sangani does not, to our loss be it said, define any one of these concepts. All it does is to show us the content of a number of thoughts' known as one or the other of these three species of dhamma. In a subsequent passage (pp. 315-318) it uses the substantival form of 'good' (ku salata; another form is ko sallam) in the sense of skill or proficiency as applied to various kinds of insight, theoretical or practical.

Now if we turn to the later expression of old tradition in the Commentaries, we find, on the one hand, an analysis of the meaning of 'good'; on the other, the rejection of precisely that sense of skill, and of that alone out of four possible meanings, with respect to 'good' as used in Book I. Kusalam, we read,' may mean (a) wholesome, (b) virtuous, (c) skilful, (d) felicific, or productive of happy result. The illustrations make these clear statements clearer. E.g. of (a), from the Dasaratha Jataka: 'Is it good for you, sir, is it wholesome?" Of (b) What, sir, is good

1 Asl. 38.

The two adjectives are kusalam, anamayam.

behaviour in act? Sire, it is conduct that is blameless (ana vajjo).' Of (c) 'You are good at knowing all about the make of a chariot." Again: The four girl-pupils are good at singing and dancing.' Of (d) 'Good states, brethren, are acquired through good karma having been wrought and stored up.'

Of these four, (c) is alone ruled out as not applicable to the eight types of good thoughts constituting dhammā kusalā. In so far, then, as we suffer the Buddhist culture of the fifth century to interpret the canon for us, good,' in the earlier ethics, meant that which insures soundness, physical and moral, as well as that which is felicific.

The further question immediately suggests itself, whether Buddhism held that these two attributes were at bottom identical. Are certain states' intrinsically good, i.e., virtuous and right, independently of their results? Or is ' good,' in the long-run at least, felicitous result, and only on that account so called? Are Buddhists, in a word, Intuitionists, or are they Utilitarians? Or is not a decidedly eclectic standpoint revealed in the comprehensive interpretation given of ku salam?

These are, however, somewhat modern-I am tempted to say, somewhat British-distinctions to seek in an ancient theory of morals. They do not appear to have troubled Buddhism, early or late. The Buddhist might possibly have replied that he could not conceive of any thought, word, or deed as being intrinsically good and yet bad in its results, and that the distinction drawn by the Commentator was simply one of aspects.

If pressed, however, we can almost imagine the Buddhist well content with the relative or dependent good of Utilitarianism, so closely is his ethics bound up with cause and effect. Good, for him, is good with respect to karma —that is, to pleasurable effect or cudæmonia.

With respect to the supremely good effect, to arahatship

1 Cf. M. ii. 94.

or Nirvana, he might, it is true, have admitted a difference, namely, that this state was absolutely good, and not good because of its results. It was the supreme Result or Fruit, and there was no beyond.' But then he did not rank Nirvana exactly in the category of good, and precisely for this reason, that in it moral causation culminated and ceased. He spoke of it as Indeterminate, as without result-as a Freedom, rather than as a Good.

He would not then have fallen in with Aristotle's definition of Good in terms of aim, viz., as that at which everything aims.' Good was rather the means by and with which we aim. But that at which we aim is, in all lower quests, Sukham, in the one high quest, Vimutti (emancipation), or Nirvana.

Nor must the substitution of these two last terms for that well-being, that well-ness, Tò cù r,' which is the etymological equivalent of sukham, be taken as indicating the limit of the consistent Hedonism or Eudæmonism of the Buddhist. For he did not scruple to speak of these two also (Emancipation and Nirvana) in terms of pleasurable feeling. Gotama attaining his supreme enlightenment beneath the Bo-tree is said to have experienced Emancipation-bliss' (vimutti-sukhapatisam vedi). And to King Milinda the Sage emphatically declares Nirvana to be 'absolute (or entire) happiness' (e kanta-sukham). And we know, too, that Buddhism defined all right conduct and the sufficient motive for it in terms of escape from ill (dukkham, the antithesis of sukham) or suffering. Here then again their psychological proclivity is manifested. They analyzed feeling, or subjective experience, into three modes: sukham, dukkham, adukkham-asukham. And in Good and Bad they saw, not ends or positions of attainment, but the vehicles or agencies, or, to speak less in abstractions, the characteristic mark of those kinds of

1 Cf. p. 12, n. 3. 3 Mil. 313.

2 Vin. i. 2, 3, quoted Jāt. i. 77.