Appearance Pluralism, Perception, and Causation Guy Longworth1 University of Warwick March 08 1. My aims here are fairly limited. They’re also a little incestuous, since my main target is the first main chapter of my colleague A. D. Smith’s book The Problem of Perception. Smith’s book is mainly about a version of the Argument from Illusion and how he thinks we should respond to that argument. My aims today are the following. First, I want to sketch the Argument, as Smith appears to understand it, pausing every so often to worry at some of Smith’s formulations. Second, I want to focus more closely on a particular sort of response to the Argument, a response that Smith associates with the early 20th Century movement called New Realism, but which I think can be detached from that setting. Here, I want to consider what Smith thinks is wrong with that response, adding some considerations from the work of Sydney Shoemaker. I won’t be able to respond fully to the complaints pressed by Smith and Shoemaker, mainly because I’m not yet in a position to mount such a response. Rather, my aim is limited to pointing to some hoops that someone who seeks to press their complaints would need to jump through, or push us through, before it would be clear that there was really something wrong with the broadly New Realist response to the argument that Smith presents. Let me say from the outset what I’m not trying to do. First, I’m not trying to argue that the New Realist type response provides a blanket answer to the Argument from Illusion. For one thing, even in Smith’s hands, the Argument from Illusion appears to shade into an Argument from Hallucination. And the New Realist type response to the Argument from Illusion is implausible as a response to the Argument from Hallucination. Second, I’m not suggesting that the New Realist type response provides the optimal, or even a viable, response with respect to any case of illusion. All I’m attempting here is a sketch of some of the ground over which future dispute over the standing of the New Realist response should take place. Third, then, I’m not suggesting that there aren’t other, better objections to the New Realist response than 1 A version of this note was presented at the London Logic and Metaphysics Forum, March 2008. Thanks to Mark Kaldron, M.G.F. Martin and other members of the audience on that occasion. Thanks especially to Matthew Soteriou who has done his utmost to make my engagements with this topic less Naïve. 1 those presented by Smith and Shoemaker. I’d be pleased to hear about such objections in discussion. But my present concern is limited to a particular class of objections. Finally, I’ll be focusing on visual perception and take no stand about how the issues discussed here might play out with respect to other forms of perception. 2. Let me begin by sketching Smith’s version of the Argument from Illusion. The argument is designed to undermine what Smith calls “Direct Realism”, a view according to which what we see, or what we see directly, includes mind-independent objects and perhaps mind-independent features or properties of mind-independent objects. In the first instance, I think “Direct Realism” is supposed to be construed as a form of Naïve Realism, so to include at least the following two components: (1) The phenomenal character of our perceptual experience is at least partly constituted (as I’ll sometimes put it shaped) by objects and properties of which we are aware in having the perceptual experience. (2) The objects and properties of which we are aware in having perceptual experience and that go towards shaping the phenomenal character of our perceptual experience are mind-independent objects and properties, at least to the extent that they exist independently of being experienced and are possible objects of experience for perceivers other than us. I think that Smith takes Naïve Realism also to include a third component, according to which the objects and properties that we see play a particular role in determining the phenomenal character of our visual experience. (3) The form of awareness involved in the perceptual experience we suffer through vision, and which shapes the phenomenal character of our experience, is the form seeing. Thus, the objects and properties that partly determine the phenomenal character of the perceptual experience involved in our seeing are just the objects and properties that are seen by us. Obviously, there is much more to be said in spelling out this type of position. And there is no reason to expect that all who take themselves to endorse Naïve Realism 2 should endorse all components on all spellings out. But the sketch here will do for present purposes. Smith’s own response to the Argument from Illusion involves dividing up the objects and properties that play a role in shaping our experience into, on one hand, those that go towards determining the phenomenal character of our experience and, on the other hand, those that we see. He thinks that by drawing that distinction, he is able to explain phenomenal character by appeal to mind-dependent properties without thereby threatening the idea that what we see are mind-independent objects and properties. His idea is that the Argument from Illusion shows that we need minddependent properties to explain the phenomenal character of our perceptual experience, since there is insufficient variety in the world to explain the various ways the world appears to us in experience. And he thinks that if those mind-dependent properties were also the things we see in undergoing perceptual experience, they would occlude the mind-independent objects and properties that we naively take ourselves to see. The problem of perception, for Smith, is to explain how there can be space for his favoured combination of views. As we’ll see, the New Realist type response takes a slightly different tack. The idea driving the response is that the objects and properties awareness of which shapes perceptual experience are mind-independent objects and properties. In order to sustain that idea, the New Realist appeals to the further idea that the mind-independent world contains a sufficient plurality of properties to explain the vicissitudes of experience to which the Argument from Illusion makes appeal. Hence, according to the New Realist response, there is no need to appeal to a plurality of mind-dependent properties in order to sustain the first component of Naïve Realism. But I think that the New Realist can afford to remain neutral about whether the mind-independent properties to which they appeal in order to explain phenomenal character are properties that we ordinarily see. The more minimal version of their view is that they are properties of which we are aware, whether or not our awareness takes the form of seeing them. So, Smith endorses Naïve Realism (1) and amends Naïve Realism (2) in line with his distinction between two roles that objects and properties of awareness can play in shaping experience. And he rejects Naïve Realism (3). The New Realist type 3 response, as understood here, endorses Naïve Realism (1) and Naïve Realism (2) and is neutral with respect to Naïve Realism (3). 3. The crux of the dispute between Smith and his more naïvely realist opponents is the status of Naïve Realism (1) and Naïve Realism (2). Smith’s version of the Argument from Illusion is designed to show that at least one of these components must be revised or rejected and that the favoured revision involves giving up the claim that the phenomenal character of perceptual experience is shaped by mind-independent objects and properties. How is the Argument from Illusion supposed to sustain that trick? 3.1. Let’s begin with the first stage of Smith’s presentation of the Argument, including his account of what an illusion is (or, rather, what “illusion” will extend over for purposes of his argument). Our argument begins…with the premise that perceptual illusion can occur. The term “illusion” is to be understood here as applying to any perceptual situation in which a physical object is actually perceived, but in which that object perceptually appears other than it really is, for whatever reason. [Footnote omitted] It is therefore irrelevant whether the subject of an illusion is fooled by appearances or not. (p.23) This understanding of “illusion” is narrower than some natural understandings, since we might allow that illusion can strike when no “physical object” is involved. Smith himself appeals to the possibility of auditory illusions, and one might think that these involve sounds. Perhaps Smith includes those amongst the physical objects. Or perhaps he thinks that when one experiences a sound, one typically hears the producer of the sound, often a physical object. But there are other cases. For instance there might be illusions involving rainbows, and there might be illusions involving events. Smith’s understanding of illusion might also be thought broader than (some) ordinary understandings, as Smith notes: 4 For example, the world appears differently to those who are colour-blind and to those who are not. This involves illusion, in the possibly unnatural sense here employed. For if I, not being colour-blind, cannot tell red and green things apart, but you can, at least one of these colours must look different to the two of us. So, for at least one of us, that colour cannot look the way it really is. (p.23) 3.2. Why is it supposed to follow from a physical object’s looking different to the two subjects that, for at least one of the subjects, the colour cannot look a way it really is? Smith’s explains his reason in the following way: That there “really” is a way the object is “objectively” is a presupposition of the Direct Realism that is under investigation here, and so will not come up for discussion. (p.23) The most important feature of this passage is the claim that “Direct Realism” is committed to there being a way a perceived object is, however many ways it can appear to different subjects. Of course, there being a way is consistent with there also being other ways. But the way this claim appears to function in Smith’s argument strongly suggests that he intends the quantifier to rule out the possibility of there being other ways. Otherwise, it would be consistent with the view that the colourblind subject and the non-colour-blind subject are both aware of ways the target object is, even though those ways are different. The obvious understanding of the claim then is this: “Direct Realism” presupposes that there is exactly one way a perceived object really is. That appears to be a remarkably strong presupposition. Some natural questions at this point are the following. Is the presupposition tenable, independently of consideration of illusion (on Smith’s understanding of the latter)? Aren’t there often many ways an object is, even according to natural scientific descriptions of the object, including ways described in physics, chemistry, and perhaps biology? Who is the “Direct Realist” who is supposed to make this presupposition? 5 Perhaps Smith’s view can be clarified in the following way. Suppose that there is exactly one way each physical object is. Then we might suppose that that way will often include a variety of more specific ways. For instance, the fact that a physical object is one way—for instance, spherical—appears not to rule out its also being a different way—for instance, red. In that case, when we say that there is exactly one way the physical object is, we might mean something like the following. There are a variety of specific ways the object is, and if we take each of those ways and conjoin them, we get a maximally specific way the object is. And there is only one such way the object is. Smith’s argument would then appear to require something like the following. Each specific way an object “really” is involves a single dimension of variation. With respect to each such dimension, there is only one specific location the object can occupy, so only one specific way the physical object can be. Hence, if it appears to different subjects to occupy different positions along a single dimension, then at least one of those appearances must fail to correspond with the way the object is, for the object can occupy at most one location along any such dimension. With respect to the present example, the idea would be that there is a single dimension of colour variation, and that an object can really occupy at most one point along that dimension. Hence, the way the object really is cannot constitute the way it appears to both our subjects, since some coloured objects present different colour appearances to the two subjects. 4. That’s the First Stage of the Argument, summarized by Smith in the following way: (P1.) [T]here is no type of physical feature that may not appear differently from the way it really is to any sense that could possibly perceive it. From there, we move on to the Second Stage, P2: (P2) [From P1] …[W]henever something perceptually appears to have a feature when it actually does not, we are aware of something that does actually possess that 6 feature. So, if you are looking at a white wall, which because of the illumination looks yellow to you, you are aware of something yellow. From the perspective of Naïve Realism, as understood here, the following component of this premise seems more or less incontrovertible. Since phenomenal character is constituted by the objects and features that we’re aware of, something of which we are aware must have the features required to constitute the phenomenal character of the experiences that we enjoy. So, if the phenomenal character of our perceptual experience is of a type that could only be explained by our awareness of something with some property, F, then in order to enjoy experience with that character, we must be aware of something with that property, F. There is a get out clause, since Naïve Realism is only committed to partial constitution of phenomenal character by objects and properties of awareness. But the get out clause is in place to allow that general conditions on experience or ways of experiencing might play a role in shaping perceptual character, and it would be against the spirit of the account to allow specific features of specific experiences to have their phenomenal character shaped by minddependent objects or properties. So for present purposes we can allow that the Naïve Realist will accept P2. However, it does not yet follow that that property must be one that, in cases of illusion, is not instanced by the mind-independent objects of which we are aware. 4.2. The Third Stage of the argument is P3: P3. …[S]ince the appearing physical object does not possess that feature which, according to the previous step, we are immediately aware of in the illusory situation, it is not the physical object of which we are aware in the illusory situation, it is not the physical object of which we are aware in such a situation; or, at least, we are not aware of it in the direct, unmediated way in which we are aware of whatever it is that possesses the appearing feature—that direct way in which we formerly took ourselves to be generally aware of normal physical objects. (p.25) 7 If we endorse the three premises, as Smith understands them, we appear to be forced (give or take the get out clause mentioned earlier) to give up Naïve Realism (1) or Naïve Realism (2), or both. We might worry that all the work is being done by the first premise, which is in effect the premise that some cases of things looking some way to us cannot be explained by appeal to mind-independent objects or their properties. And Smith’s argument to this point does not rule out the following type of response. The phenomenal character of our perceptual experience is partly shaped by the features of things we are aware of. And, where the experience takes the form of visual perception, those features are mind-independent features. However, they are not always the features we say that the things we see really possess; sometimes they are mere appearances. However, that is perfectly consistent with the ways things appear being mind-independent features, features things would possess whether or not we became aware of them. What, if anything, does Smith have to say in response to that type of response to his argument? Put another way, what argument does Smith have to the effect that some cases of illusion cannot be explained in line with Naïve Realism (1) and Naïve Realism (2)? What constraints are there on the range of mind-independent objects and features that are able to sustain that result? 5. Smith considers a response to his argument at least close to the one just sketched, attributing it to the New Realist movement that flourished in the United States at the beginning of the last century. He writes: A number of philosophers…have attempted to block [the argument] at Stage Three. The New Realists accept the sense-datum inference: when the tomato looks black to me, I am aware of something black. They deny, however, that I am therefore not aware of a normal physical object… For why should not the blackness that I see be a genuine feature of the tomato? (29) 8 Although Smith’s characterisation of New Realism brings that position close to the sort of response sketched above, there are two points at which it departs slightly from that sketch. First, the initial way in which Smith describes New Realism has it committed to objects’ instancing a variety of colour properties, so that the tomato instances both redness and also blackness. But the response sketched here is less radical. According to that response, the tomato instances both the property of appearing red and the property of appearing black, where that is perfectly consistent with it instancing neither redness or blackness. Second, Smith characterises the New Realists as attempting to block his argument at Stage Three, our P3. But an equally natural view would be that they seek to challenge the argument at Stage One, our P1, by seeking to reject the idea that illusion is possible, at least where illusion is understood in the particular way that Smith recommends. On Smith’s understanding, there are possible illusions that involve physical objects’ lacking any property corresponding with the way they appear. In denying the possibility of that sort of illusion, the present response has it that appearance properties are ways things are, so that the only sort of illusion that is possible is one involving things not really being a way they appear, despite really having appearance properties able to shape the phenomenal character of experiences of that thing. Smith seems to correct both points in the following passage, while introducing another optional element to the New Realist type account: [The New Realists]…were actually engaged in questioning our everyday conception of physical reality, and replacing it with a picture of physical objects as extremely complex systems of variegated views and appearances. Reality is now regarded as being sufficiently rich to allow all the features present in our varying perceptions to be genuine constituents of physical reality. (30) The optional element that Smith introduces here is the idea that the ‘variegated views and appearances’ that are to play a role in constituting the variegated characters of our 9 perceptual experiences are to replace, rather than to supplement, those aspects of our everyday conception of physical reality that involve physical objects and some of their properties being comparatively stable elements in our experience, by contrast with the complex systems of appearances elements of which play a more ephemeral role in shaping our experience. What, then, does Smith think is wrong with the New Realist style of response? There are two main strands to his objection, both in the following passage. Our straightforward understanding of physical objects is replaced in such an account by a bewildering system of properties from which the fairly stable everyday characteristics of things can be abstracted only with great ingenuity… The supposedly physical difference [presented by a single physical object that presents two appearances to a subject, or subjects] is a difference that makes no physical difference at all. A merely apparent difference is not, to be sure, wholly without possible causal consequences. Such consequences are, however, restricted to causal contexts that involve the psychological realm… The point is that there is no possible non-psychological, merely physical, situation in which the supposedly physical difference…counts for anything. This invalidates the claim that the difference in question is indeed physical. (31) Let’s leave aside the optional component of the New Realist type response according to which our straightforward understanding of physical objects is replaced, rather than supplemented, by a bewildering system of appearance properties. We can then focus on two broad lines of objection contained in this passage. The first line of objection is that, if we allow physical objects to bear the required variety of appearance properties, this is bound to ‘bewilder’ our attempts to see through their array in order also to perceive the ‘fairly stable…characteristics of things’. There are at least two ways of developing this sort of concern, neither of which I want to spend time on here. The first way takes the form of a challenge to the New Realist to explain how we are able to perceive objects and the properties they really have in 10 the face of the bewildering plurality of appearances. While I think that challenge is a good one, it is one that would appear to confront most theorists in this area, since most theorists allow that there is a plurality of appearances and differ only over whether those appearances are taken to be independent of our experience of them. In particular, the challenge confronts Smith to the same extent that it confronts the New Realist. The second way of developing this type of concern also takes the form of a challenge, in this case to explain why our awareness of appearances does not serve to screen us off from objects and their fairly stable characteristics. Since it is possible for something to appear some way F without being F, we might think that awareness of the appearance of F is compatible with a failure to be aware of F. In that case, we might wonder how, in cases where something not only appears F, but also is F, we can see through the appearance in order to see the instancing of F. Again, I think this is a good question, but it seems an equally good question for someone who believes that appearance properties are mind-dependent, so is not of immediate concern in the present context. The second line of objection is the one I want to focus on for the remainder of the talk. According to this second line, the problem is supposed to attend the idea that appearance properties are “physical”, or perhaps more minimally that such properties are causally efficacious with respect to what is “physical”. Again, there are a variety of ways of developing this sort of concern, and I shall focus on only a selection. A first way of developing the concern is by appeal to the idea that “physical” properties must have a variety of possible effects. In particular, according to this idea, “physical” properties must do more than simply affecting the phenomenal character of perceptual experience. One way of responding to this development is to point out that, if that is what the requirements are on “physical” properties, as Smith understands them, then there is no immediate reason for the Naïve Realist to care about whether the properties that shape phenomenal character are “physical”. What the Naïve Realist is committed to is only 11 that the shaping properties are mind-independent; they have no further commitments regarding the specific nature of those mind-independent properties. In fact I don’t think that Smith intends his classification of properties as “physical” to do that sort of work. Rather, he intends “physical” and “objective” to do the same sort of work, akin to the work of “mind-independent” as I’m using it. His thought, then, is that in order for a property to count as mind-independent it must have effects on things other than our minds. One response to this development would be to question the requirement. We might challenge the claim that mind-independence entails the power to affect things other than the mind. However, we might then find it somewhat mysterious that mindindependent properties could be so limited in their range of effects given that they are constituted independently of those effects. If we do find that mysterious, then a better response would be to point out that it appears to be only a contingent feature of appearance properties that their effects are limited to perceptual experiences. Insofar as we think of appearance properties as mind-independent we might also think of them as the sorts of properties that might be detected by a physical machine, perhaps a machine similar in construction to the physical systems that enable human experience. Unless there is reason to think that no such machine is possible, its possibility would seem to underwrite sufficient width of cosmological role for appearance properties that we can happily view those properties as mind-independent. Although further discussion is warranted, I predict that the upshot will be unfavourable to Smith’s case. Either Smith will be forced to appeal to more stringent conditions on what is to count as mind-independent, or “objective”, or “physical”, at which point his Naïve Realist opponent will be apt to view those conditions as optional. Or he will appeal to conditions that the Naïve Realist is independently disposed to accept, at which point his Naïve Realist opponent will be able to show that appearance properties can meet those conditions. It seems to me then that Smith’s argument runs aground at this point. There is no immediate reason, based upon the premise that they only have impact on experience, to reject the view that appearance properties are mind-independent. However, there is 12 a quite different way of developing this general type of concern by appeal to a converse premise. Taking this line, Sydney Shoemaker has attempted to argue that mind-independent appearance properties would, in at least some possible cases, be precluded from having an impact on experience. It is to his argument that I now turn. Shoemaker’s argument can be sketched in the following way. We begin by supposing, in line with Naïve Realism, that there might be two subjects, Jo and Kim, who are so constructed that, when confronted by a particular type of physical property, P, the phenomenal character of Jo’s experience is shaped by appearance F and not by appearance G, while the phenomenal character of Kim’s experience is shaped by appearance G and not by appearance F. The question at issue is whether the phenomenal characters of both experiences are shaped by mind-independent appearance properties, F and G, respectively. Shoemaker argues that whether that is possible depends upon the specific way in which the difference between Jo and Kim is explained. In particular, Shoemaker argues that it appears to be possible that the “proximal effect” on Jo and Kim of the physical property, P, together with whatever appearance properties it supports, might be the same, consistently with their experiences having different phenomenal characters. This might be so, for instance, if the immediate impact on their perceptual systems, perhaps at the retinal surface, is the same, and the differences in their respective experiences arise as a result of further processing in the system’s innards. Shoemaker’s idea, in short, is that what I’ll call the “surface” of the two systems screens off their operative elements—those involving the system’s innards—from causal interaction with the two appearances. Hence, the different mind-independent properties are inefficacious in producing the different experiential effects enjoyed by Jo and Kim. That may be consistent with the appearance properties being mind-independent. But on Shoemaker’s view, it is not consistent with their shaping Jo and Kim’s experiences in the way the New Realist supposes. Hence, if Shoemaker is right, it is not consistent with the scouted response to Smith’s argument. I don’t have space to engage in detail with Shoemaker’s argument. Instead, I want to briefly sketch some hoops that the purveyor of the argument will need to jump through (or push us through) before the New Realist, or appearance pluralist, need be troubled by it. 13 1. Although we might well believe that there are causal constraints on what can be perceived, it is not trivial that these constraints apply at the level of experience and experienced property. The Naïve Realist takes the relevant form of experience, awareness of a property, to be a type of relation holding between a subject and the objects or properties of which they are aware. And it is apt to seem mysterious that a property could cause the holding of a relation between a subject and that property, in part because causation is often held to involve an independence condition, apparently failed in the required type of case. However, Mark Kalderon has a lovely example which at least suggests that there is nothing immediately problematic about such cases of causation. He has us consider the power of the wind to cause a weather vane to point in its direction. We could attempt to question whether Kalderon’s case is really of the required sort. And we could try to question whether whether the principles governing the case are guaranteed to hold in the case of awareness. For in the case of awareness, the relations are supposed to play a constitutive role with respect to experience, while weather vanes and their directions can be characterised without appeal to the wind. But a better response would be to shift tack slightly. Instead of using Naïve Realism in an attempt to argue that causation couldn’t play the constraining role, we might use it instead as the basis for an argument to the effect that it needn’t play that role. In the first place, the thought would be that required distinctions between cases of perception and other cases can be captured by appeal to whether the required experiential relation holds, without needing additional input from causal conditions. In the second place, it might be suggested that, insofar as causal conditions are relevant to whether an experiential relation is in place, those conditions apply at a lower level, requiring, for instance, that the physical properties that support the experienced properties interact causally with parts of the physical system that supports the experience of those properties. At that point, the apparent difficulty for the Naïve Realist would lapse. 2. Suppose, however, that reason can be provided for a causal condition on awareness that applies at an appropriate level. The next hoop involves the Naïve Realist’s 14 opponent providing reason to think that Jo, Kim and their respective appearance properties fail to meet that causal condition. If, for instance, the condition is framed in terms of chained counterfactuals, then there appears to be no immediate reason for thinking that the fact that Jo and Kim share an experiential surface undermines the obtaining of those chained counterfactuals. What appears to be driving Shoemaker’s argument at this point is a more substantive conception of causal interaction than is supplied by appeal solely to counterfactuals. Perhaps, for instance, he is appealing to some sort of physical model of causation, according to which causal interaction involves transfer, or preservation, of some conserved quantity like energy. Such a model might ground the idea that the systems’ surfaces screen off the appearance properties from experiential effects. There are two questions to press at this point. The first is, Should we buy the proposed account of causation? Jonathan Schaffer argues that our ordinary conception of causation appears happily to allow cases of negative causation, where cessation of a process sustaining some effect is taken to cause cessation of the effect. One example of this is where a bullet to the heart is allowed to cause death by inducing cessation of the process whereby the brain is oxygenated. If Schaffer is right, then we have some reason for rejecting the type of physical model of causation that appears to be required by Shoemaker’s argument. Alternatively, suppose that Schaffer is wrong. The second question is whether we should retain our initial commitment to there being causal conditions on awareness, so understood. That is, discovering that causation proper involves energy transfer or preservation might make us rethink our initial view that causation proper is required for awareness. Perhaps, for example, for the work the causal condition is required to do in governing awareness, a merely counterfactual condition is adequate and causation proper is redundant. 3. Suppose however that those concerns have been addressed. The third hoop for the opponent of Naïve Realism involves providing a principled reason for restricting the proximal effect of appearance properties in a way that might show them to be inefficacious in the production of experience. 15 One question here is, Why should we view the effects of experience as being anything less than the whole system? Shoemaker appears to be motivated by the idea that we can think of the surface of the experiential system as a point of initial impact, with effects within the surface transmitted by events involving the surface rather than events involving the distal precursors of those surface events. But we might question that conception of causal interaction. In homage to Thompson Clarke, we might ask whether our conception of causal relations supports answers to questions like “How much of the operation of the system was caused by the impact of light on its surface?” Consider an analogy. I might throw a stone at a cup, causing the cup to break. How much of the cup’s breaking was caused by the initial impact of the stone on its surface? How much was due instead to the internal structure of the cup? Insofar as we find questions like this difficult to answer, or even to understand, that is a sign that we shouldn’t rush to accept the sort of answer Shoemaker thinks that we should give in the case of perceptual systems. Suppose, however, that we allow that the immediate effects of the physical object presented to Jo and Kim are surface effects. Shoemaker’s argument then requires that those surface effects act as a sort of screen, or bottleneck, with respect to the plurality of properties we are assuming to be carried by the physical object. His idea at this point appears to be that, since the surface of each of the systems is of the same type— perhaps of the same physical type—then it must transmit the same effects into the system’s innards. However, it should be obvious that the appearance pluralist will be unwilling to allow that argument to stand. For the appearance pluralist is already committed to the possibility of a single physical object, with a single range of physical properties, carrying a plurality of appearance properties. So a natural response to the latest round of argument is that it is also possible for the inner surface of the system to carry a plurality of properties and for it to be the causal impacts of one rather than another of this plurality that causally explains the different experiences suffered by Jo and Kim. 6. As I said at the start, I haven’t attempted to show that the New Realist, or appearance pluralist, type of response to the Argument from Illusion is adequate. I 16 haven’t even attempted to show that no causal argument of the sort presented by Smith and by Shoemaker can be used to undermine that type of response. What I hope to have done is to indicate some points at which more work will be required before we’re in a position properly to assess appearance pluralist response to the Argument from Illusion. 17