John Foster argues that our common-sense view of the world as something whose existence is logically independent of the human mind and also metaphysically fundamental is quite mistaken. Instead, he claims, the existence of the world is ultimately constituted by facts about human sensory experience. Our sensory experience is organised in such a way as to create the physical world by disposing things to appear ‘systematically worldwise’. The organisation of our sensory experience has this nature because God has so ordained it, and it is this which gives our experience the objectivity which makes it experience of a real world. This, in brief, is Foster's thesis, one which, as he acknowledges, might strike us as bizarre.

Before looking at the case for this position, two fairly obvious questions press for answers. The first is whether one can accept that a benevolent God has arranged things in such a way that we all have a view of reality which is in fact a delusion. Descartes, by contrast, argued that the certainty of the existence of a benevolent deity gives us certainty of the existence of a mind-independent world. Second, the problem of evil is raised in an even more pressing form: why should such a deity decide to give us experiences as of the Black Death, earthquakes, and all the other evils of the natural world?

Foster begins his case by arguing that neither a direct realist nor a representationalist view of perception can give us an understanding of what it might be to be in contact with a mind-independent world. (Foster in fact uses different terminology, but the more familiar terms would have been better.) Against direct realism, Foster advances again the case of illusions such as that of the stick which looks bent in water, colour blindness, astigmatism, etc., considerations which direct realists have found unpersuasive. Secondary qualities are discussed, with emphasis on the claim that colour does not feature in the scientific account of the physical world. I don't think that the direct realist need object to the claim that taste and smell exist merely as powers in objects to evoke certain sensations in sentient beings, but the case of colour has always been more troubling, though I doubt if any direct realist will consider that the issue of the status of colour in itself poses an insuperable problem for that position.

The representationalist position is rejected because on this view we fairly clearly do not have direct access to the world. The claim that the best explanation of our sensory experience is that it relates to and is caused by a mind-independent reality beyond it is not discussed at this point, though it surely should have been. There is also a curious argument that on the representationalist view we are forced to say that concepts such as ‘apples’, ‘trees’, ‘rivers’, and so on have no application (38). It is difficult to see what support there could be for this claim. To say that we perceive such objects by means of a mental representation is not on the face of it to imply that the relevant concepts have no application.

Once the secondary qualities are excluded from the physical world, the discussion moves on to the status of primary qualities, and the focus is on Locke's claim that the one property which makes matter, in all circumstances, the sort of thing that it is, and which furnishes each material object with its intrinsic content, is solidity, or impenetrability. Foster briefly examines the idea that impenetrability amounts to no more than a principle of individuation, but rightly rejects it (57-8). Foster argues that even if we take impenetrability to be a causal power, a disposition to prevent penetration, and can take solidity to be the intrinsic property on which, in the framework of the laws of nature, this power is grounded, ‘none of this, of course, tells us what solidity as such is’(59). I find it difficult to understand why, given our first-hand experience of solidity and impenetrability, together with the scientific explanation and description of what constitutes this feature of the physical world, Foster should feel that we still have no grasp of what solidity as such is, and that our knowledge of it can only be ‘opaque’. Foster's explanation of his use of the terms ‘transparent’ and ‘opaque’ seems to be of no help. Conceiving of triangularity as triangularity is transparent, Foster says, but conceiving of it as the geometrical figure discussed in the fourth chapter is opaque (54). If that is what opacity is, it is difficult to see how one's being struck by a solid object gives one only an opaque conception of solidity in any way parallel to this. Moreover, in the example about triangularity it is clear what a transparent conception is, but we have no idea, apparently, of what it could be to have a transparent conception of solidity, of what solidity as such might be. We are simply told that neither our experience, nor that experience in conjunction with a scientific account of solidity, can give us such a transparent conception, and that therefore a realist conception of the physical world leads to the view that ‘intrinsic content’ is ‘inscrutable’.

The refutation of realism is spelt out in chapter 4 and proceeds by means of an elaborate, and quite difficult to follow, speculation. The gist of the argument is that what underlies the appearance of a single unified physical world may in fact be a series of discrete spaces, or collections of particles, which are so organised as to present the appearance of a single unified space. This is a possibility that the realist must accept, and Foster claims that the fact that the realist has to allow such possible cases amounts to a refutation of realism. It would, Foster thinks, ‘be odd if realism were the right philosophical account of the world as it actually is… but did not generalise to how things would stand if one of the less probable outcomes obtained. We surely expect realism as a philosophical thesis to be a response to our fundamental conception of the world, rather than to what we take to be the most plausible account of the specific circumstances of its existence’(149).

It seems to me that what Foster is claiming here is that we must reject realism because it is unable to counter certain sceptical speculations with a complete refutation of their possibility. As opposed to this, I would claim that we surely ought to have learnt that there is no such thing as ‘the refutation of scepticism’. What should govern our thinking here are two principles: first, Occam's razor, which demands in this case that it is illegitimate to posit a number of entities (spaces, etc.) when the requisite explanatory work is effected by positing only one, and second, the inference to the best explanation, which will yield the result that the best explanation of the appearance of a single, unified physical world is the actual existence of such a world. But none of this amounts to a demonstration that the sort of speculation that Foster entertains is logically ruled out of court. It is therefore very difficult to accept that Foster has offered anything which could be taken to be a convincing refutation of realism.

A final puzzle arises with regard to Foster's treatment of the obvious problem for an idealist of what to say about the apparent truth that the world existed long before consciousness arrived on the scene. Foster leads in to his answer to this by first considering the apparent causal interaction between mind and body. How can there be any such thing if the physical world is ontologically dependent on the human mind? The answer is that, while ‘in the context of the idealist's account of the ultimate metaphysical status of the physical world… the realist's thesis of ontological independence must be denied… with respect to what gets idealistically created, human subjects and the physical world form interactive partners in an integrated psychophysical whole’(184). But even if we accept that somehow causal interaction between mind and body is itself part of what is idealistically created, it is very much more difficult to accept Foster's extension of this approach to make sense of the truth that the world existed for billions of years before human beings appeared on the scene. Foster is of course aware that the idealist seems committed to saying that ‘what is created as a world for human subjects – created in part by the way their sensory experience is organised – is created as something whose existence precedes them’ (185). His answer is that ‘with respect to what is created, it is correct … to say that the universe comes into existence at t′ [i.e., billions of years ago], and with respect to its mode of creation, it is correct to say that it came into existence at t′ [i.e., when human subjects came into existence] (185). I find it impossible to understand how a universe billions of years old can be what is idealistically created billions of years later, or, though Foster does not mention this, how a universe the vast majority of which is beyond what is embraced in our sensory organisation can also be part of what is thus created. Altogether, the case Foster develops for what he calls phenomenalistic idealism is a very puzzling one.