What defect then does Mr Russell find in this proposed ground? The answer is quite unambiguous. The 'law of sufficient reason' upon which the ground is itself grounded is untrue. For what is the 'reason' for a proposition? 'The reason for a proposition is always expected to be one or more simpler propositions. Thus the law of sufficient reason should mean that every proposition can be deduced from simpler propositions. This seems obviously false, but in any case it cannot be relevant in considering Idealism, which holds propositions to be less and less true the simpler they are, so that it would be absurd to insist on starting from simple propositions.' 1

1 Philosophical Essays, p. 164.

Now if the 'law of sufficient reason' does in fact mean what Mr Russell in this last citation says that it means, I am ready to grant that it does not hold good. But it is perfectly certain that this is not what the law means for any Idealist who introduces the conception to explain his adherence to the internality of relations. What the Idealist means when he holds that 'nothing can be just a brute fact, but must have some reason for being thus and not otherwise,' is, quite simply, that the intellect cannot accept an 'ungrounded' union of differents, a 'bare conjunction.' But the 'ground' which the intellect posits is not, of course, supposed to be 'one or more simpler propositions.' Indeed Mr Russell shows, in the passage quoted above, that he himself is quite well aware of this, and why he should imagine that his argument still touches Idealism is far from clear. The 'ground' which must be posited on Idealist principles is, as we have sufficiently seen earlier in this chapter, a system within which the differents connected are conceived as mutually implicatory elements. If Mr Russell's criticism is to be relevant to the Idealist defence of 'internality,' what he ought to assail is the line of argument which contends for the intellectual necessity of positing such a ground if differences are to be united in a way acceptable to thought. For if it be true that the union of differences does require a ground of this nature, it is impossible to hold that the possession of a relation, or any other character, by a term 'makes no difference' to the nature of the term.

Let us pass on to consider Mr Moore's treatment of the matter in his well-known essay, 'External and Internal Relations.'1 We shall see, I think, that Mr Moore is quite as far removed as Mr Russell from an appreciation of the real foundations of the Idealist doctrine.

1 Philosophical Essays, p. 165.

With Mr Moore's statement of what is meant by 'internality' the Idealist will have no quarrel. After a somewhat lengthy process of eliminating abstractly possible meanings, Mr Moore reaches the position that 'one thing which is always implied by the dogma that, "All relations are internal," is that, in the case of every relational property2 it can always be truly asserted of any term A which has that property that any term which had not had it would necessarily have been different from A.'3 This is, I think, a perfectly fair statement. And we can agree with Mr Moore further when he asserts that the 'difference' alleged in the 'dogma' is not merely 'numerical difference' but also 'qualitative difference'.

What then is Mr Moore's own position with regard to the 'dogma' so stated? He is prepared to agree that some relational properties are certainly internal to their terms in the sense that the absence of these relational properties necessarily involves numerical difference in the terms. He is prepared to agree that it is probable, although not certain, that some relational properties are internal to their terms in the fuller sense that the absence of these relational properties necessarily involves qualitative difference in the terms. But he insists that there remain over many quite obvious cases in which relational properties are not internal to their terms, either in the one sense or in the other. And this, of course, is more than sufficient for the denial of the doctrine of internal relations.

The Idealist reader of Mr Moore's essay arrives at the present stage in a mood of eager expectancy. There is every reason to believe that Mr Moore is about to grapple seriously with a matter which has always been for Idealism of bed-rock importance. But nothing could in fact be more disappointing than Mr Moore's actual treatment. What he has to show, in order to overthrow the opposing doctrine, is that some relational properties at any rate are not internal. He proposes to do so by demonstrating that some relational properties are not internal even in the 'minor' sense (i.e. entailing numerical difference in the terms), and therefore a fortiori not internal in the 'major' sense. But the actual argument is at bottom little more than an appeal to the supposed 'plain facts' of common sense. Indeed the impression with which one is left is that, in Mr Moore's opinion, the main work of refuting the doctrine of internal relations is completed when that doctrine has once been clearly stated (as by himself). When clearly stated it refutes itself.

1 Philosophical Studies, chap. ix.

2 Mr Moore's distinction of relational property from relation is, I think, a contribution to clarity, but does not specially affect our present argument.

3 Philosophical Studies, p. 284.

Take the case of a coloured patch, says Mr Moore, half red and half yellow. Now the whole patch has the relational property of possessing the red patch as a spatial part. This relational property is, Mr Moore agrees, clearly internal in that 'any whole, which had not contained that red patch, could not have been identical with the whole in question.'1 But to find a relational property that is not internal we have only to turn to the relational property owned by the red patch, that of being a spatial part of this whole. 'It seems quite clear,' says Mr Moore, 'that... the red patch might perfectly well have existed without being part of that particular whole.' 2 Yet the 'dogma' of internal relations implies 'that any term which does in fact have a particular relational property could not have existed without having that property.'3 'In saying this,' Mr Moore goes on, 'it obviously flies in the face of common sense. It seems quite obvious that in the case of many relational properties which things have, the fact that they have them is a mere matter of fact: that the things in question might have existed without them.'1

1 Philosophical Studies, p. 288. 2 Ibid. 3 Ibid.