Gregg Rosenberg on Consciousness and Causation -- Underrated Philosophy?

While waiting for reader submissions, I suppose I will start things off. My selection today for "underappreciated philosophy" is Gregg Rosenberg's 2005 Oxford University Press book, A Place for Consciousness: Probing the Deep Structure of the Natural World.

I'm not sure that Rosenberg's book counts as underappreciated simpliciter. It seems to be appreciated by a certain type of philosopher (it has been cited 74 times according to Google Scholar, mostly by people sympathetic with panpsychism and Russellian/Neutral Monism). My sense is that it is underappreciated by philosophers at large, however, and by people interested in the metaphysics of consciousness and causation.

I guess I will begin with a little background on Rosenberg. He was, apparently, a student under (CORRECTION: with) David Chalmers back when Chalmers was at Indiana. According to the back flap of the book, Rosenberg left the discipline after grad school, patented some form of technology, started a venture-capital funded company, and sold it (for a lot of money presumably) in 2001. He then went back to his dissertation, wrote it up as a book, and got it published through OUP. Pretty remarkable.

Anyway, here's why I think the book is underappreciated.  Here's a first pass: it's a book that more or less singlehandedly caused me, a one-time staunch physicalist, to become a mind-body dualist. I know this is just a personal anecdote, but still, I think it is worth dwelling on for a moment. I started out my undergraduate career doing philosophy of mind at Tufts with Dan Dennett, one of the hardest-core physicalists out there. I was completely on board with him. Dualism had always seemed silly to me, and completely at odds with any scientifically respectable account of reality. And reading David Chalmers' book The Conscious Mind didn't sway me at all. The Zombie Argument -- the argument that Chalmers' entire book was based on -- immediately struck me then (just as it does now) as utterly question-begging. It seemed to me that will share Chalmers' intuition that zombies are conceivable, and so metaphysically possible, if one antecedently finds dualism attractive. Since I didn't find dualism attractive in the slightest, the Zombie Argument seemed silly to me.

Anyway, I more or less remained a physicalist...until I read Rosenberg's book. What was it about the book that did it for me? What was it that converted me into dualist? The first preface of Rosenberg's wonderful premise hits the nail on the head. Rosenberg writes on p. ix:

My intention in writing this book was to create something whose importance lies beyond the details of its arguments. I myself consider this primarily a book of ideas. Of all my hopes, my dearest is this: that A Place for Consciousness should provide inspiration to those like me who were raised with the physicalist orthodoxy, accepting it but not fully comfortably, whose disquiet always has been silenced at the end by the baffling question: How could it be otherwise? I believe this book points to a place in the space of philosophical ideas where something truly new and interesting exists. I am, above all, trying to lead readers to that place so that they can return without me to explore it on their own...

...

Both Russell and Whitehead argued that physical science reveals only a structural aspect to nature. If physics is all structure, it is natural to suppose that intrinsic properties related to the intrinsic properties we experience in consciousness are the intrinsic content of the physical. The suggestoin raises several questions: (1) Why should the intrinsic properties of a physical system be experiential? (2) Why do they exist above the level of the microphysical...? (3) Why should they form a unity of the kind we are acquainted within consciousness? and (4) Why should phenomenal content, as the intrinsic content of the physical, correspond so closely to the information structure within the brain? By constitutively linking experience and causation, I answer these questions from first principles. (my bolding/italics)  

The sections I have bolded here are key. Professional philosophers today seem to emphasize rigor above all else. This has long puzzled me (see here, here, here, and here), given that most of the Greats (the Platos, the Aristotles, the Kants, etc.) almost "to a T" went beyond what they could rigorously show, opening up important new ways of thinking in the process. Kant never successfully established categorical normativity -- but so what? He was the first person to even get us thinking about it explicitly. And that was a great advance. 

Rosenberg's book accomplished something similar for me. In contrast to other work defending some form of dualism (such as Chalmers' work), Rosenberg's book painted the first picture of reality to me that made dualism seem more plausible than physicalism. This was in large part because -- and here's the real thing that gripped me about his book -- Rosenberg shows (pretty convincingly in my view) that it is not just consciousness that science has trouble accounting for; science cannot account for causation either. And it can't account for it for the same simple reason: science deals with structure, whereas consciousness and causation -- properly understood -- both seem utterly simple and intrinsic. (note: it was really the second half of Rosenberg's book -- the part on causation, not the first part (on consciousness) -- that "got" me)

Indeed, Rosenberg got me to fundamentally "see the world in a different way" -- a way in which the idea of there being intrinsic qualities (such as primitive consciousness and causation) just have to be a part of any remotely plausible metaphysical picture of reality. For here, essentially, is what the book got me to see. First, every substance or property (note: I think there is no coherent distinction between substances and properties) we would ordinarily call physical -- mass, charge, the weak and strong nuclear forces, books, motorcyles, whatever -- are all fundamentally relational, or structural, in nature. An electon just is defined in terms of a particular function. So is mass. So is charge. Etc. But -- and this is the second thing that Rosenberg's book got me to see -- you can't have relations without relata. There have to be intrinsic properties "behind" the relations we observe (as physical things) for there to be those relations in the first place. This is, by the way, something that I think Kant knew too (with the phenomena/noumena distinction), and which Daniel Stoljar has focused on recently as well (though I think he is wrong in his analysis of its implications). 

So, anyway, it wasn't the detailed arguments of Rosenberg's book that got me. I suspect some (or even much) of his own theory of causation, and how consciousness emerges from large-scale causal nexus in the brain, is false. Be that as it may, Rosenberg got me to thinking, for the very first time -- on broadly Kantian grounds, but also simply focusing on the phenomena of consciousness, causation, and structure -- that any world at all (let alone ours) just has to be dualistic in nature: at least in the sense of there being (A) "physical"/structural properties that can be measured by science, and (B) intrinsic/non-structural properties which cannot. (Readers might also note that these are ideas I explore and put to use in the new metaphysical model of the world and quantum mechanics that I developed in my recent Phil Forum paper "A New Theory of Free Will". Among other things, I argue that *any* world whatsoever must be constructed on a hardware/software duality that maps onto precisely the structural/intrinsic distinction that motivates mind-body dualism).

That, in a nutshell, is why I am a dualist today. I think Rosenberg's point is fundamentally right, that it persuasively shows what's wrong with physicalism, and that it persuasively shows what is fundamentally wrong with all purported defenses of physicalism, including the phenomenal concept strategy (which I agree with Chalmers to be a blind alley: either phenomenal concepts can be made sense of in relational/physical terms, in which case they fail to account for the intrinsic/non-relational elements of experience, or else they are understood in a non-physical way, in which case the strategy is self-defeating). And that is why I think Rosenberg's book is underappreciated.

Comments

[Gregg Rosenberg]

Hi Marcus,

I was pointed to your blog post by Axel Barcelo, a logician friend of mine. I just wanted to thank you for taking the time to write it. As far as I know, you are the first commenter to focus in on the book as a book of ideas, which is how I hoped it would be received. As I say in the preface, I supply many detailed and rigorous arguments in the book, but they are not the reason for writing it or where the enduring value is (though predictably I'm more hopeful at least some of them are sound than you are!)

BTW, a small correction: David Chalmers and I were graduate students together at Indiana, though he graduated first and was very kind to agree to serve on my dissertation committee after he graduated. It's probably more accurate to say I studied with him than under him, though he had significant influence on some of my thinking.

best wishes, and thank-you,
--Gregg

[Marcus Arvan]

Hi Gregg: mea culpa! For some reason I thought you were Dave's student. Anyway, I'm more sympathetic with your arguments than my post might indicate -- though I do have worries about the distinction you want to draw between consciousness and protoconscioisness that some of your critics have had. If you don't mind a minor plug of my own work, I think you might find the paper I recently published on free will (linked to in the post) of interest. As I indicate above, the model of reality and quantum mechanics it defends is very much inspired by thoughts that germinated in my head a while after reading your book. I think (or at least hope!) you might find the model attractive, particularly in light of the many philosophical problems I think it sheds new light upon.

Cheers,
Marcus

[Eric Thomson]

Good to see a discussion of Rosenberg's work. A few questions I have always had about his work, that have contributed a lot to my underappreciation :)

1. I have never been on board with the claim that the natural sciences cannot tell us about intrinsic properties in the world. For instance, aren't many anatomical or structural features intrinsic features? E.g., E.g., this molecule is shaped like a double helix. Your heart's mass is 0.75 kg. Etc..

2. Let's say that my experience of red indeed has such-and-such intrinsic properties. Neuroscientists have hypothesized that this property of brains, and the results have been tremendous, with ongoing fecundity. What explanatory power is purchased by the view that there are actually two sets of properties, that happen to map onto each other perfectly?

3. How does Rosenberg avoid the standard problem with panpsychism: epiphenomenalism? I know he focuses a great deal on causality and experience, but can't we just imagine the intrinsic causal powers obtaining without experiences? E.g., like in a vat of helium gas? What does it give you to add experiences to the helium gas, over and above what his story about causality, gives you? Given that our best evidence is that experiences are associated with complex brain states, I need a good reason to allow experiences to bleed into helium gas...

Anyway, those have always been my concerns (I have not read his book, but only saw him speak at Indiana on his theory).

[Marcus Arvan]

Hi Eric: thanks for your comment. I can't speak for Rosenberg, but here are my thoughts in reply.

1. Although you are using the standard philosophical conception of what makes a property "intrinsic", I think -- and I think Rosenberg thinks -- that that conception fundamentally misses the boat. For let's think about the kinds of properties you are calling "intrinsic": the shape of a double-helix, or my heart's mass. These properties are *fundamentally* structural. A helical shape is defined in terms of relationships between a things parts, etc. The same is true of mass, including my heart's mass. A thing's mass in *defined* in terms of how the thing relates to *other* things (e.g. how much force it takes it move it, etc. - and force *in turn* is relationally defined!). This is Rosenberg's point, and I think it is fundamentally right: the "physical" world -- mass, charge, helical structures, etc. -- is *entirely* relational all the way down. (He's not the only one to think this, by the way. Russell, Whitehead, Blackburn and many others -- not to mention Kant! -- have made a similar point. Indeed, in one sense I think it's so obvious that it almost goes without saying. All interactions with physical objects are just that: *interactions*, or relationships. Science cannot, as such, get at the things that stand behind the relationships: the truly intrinsic *relata* (or hardware) that comprises the relations. And I think -- and I think Rosenberg thinks (and Kant thinks, etc.) -- that there can't even *be* the physical relations at all (i.e. the physical world) without such fundamentally intrinsic relata. Thus, natural science *inevitably* leaves out something that any world must contain (I am going beyond Rosenberg here; these are sort of my own thoughts).

2. I think your point about red and neuroscience misses the boat on precisely the same count. We can explain structural features of brains, including how they work (or function). But relational structures can always be *described* (that's just what it is for something to be relational; one can, in principle, describe any relation you like). What red *looks* like, on the other hand, is utterly indescribable. (Color neuroscientists still admit this, by the way; I've read several articles over the past several years by color neuroscientists admitting we can't ever know if "my green is your red" precisely because the colors themselves can't be described).

3. I'll break my response to this one into two parts: (A) what I think Rosenberg would say, and (B) what I would say.

(A) I think Rosenberg would say that without the intrinsic stuff (panpsychism), you don't even have a *causal* order. All you have is ordered series' of information. So, it's not that consciousness (or protoconsciousness) "does nothing" (as epiphenomanlism implies). No, on Rosenberg's picture, it does something very important: it *links* ordered series' of information into a true causal order.

(B) In my recent paper in Phil Forum, "A New Theory of Free Will" (which, as I note above, has Rosenbergian roots), I provide a new model of quantum mechanics in which consciousness is *not* epiphenomenal, but rather figures into a new explanation of quantum mechanics (including quantum superposition, indeterminacy, and quantum collapse) that in turn illuminates a number of longstanding philosophical problems in the philosophy of mind, time, and personal identity. But I won't get onto my soapbox here. If you want to see how I think consciousness can be truly causally effective in an apparently causally-closed physical order, you'll have to read the paper. :) (The short story, though, is this: there are reasons to think that our world may be structurally identical to a peer-to-peer networked virtual reality simulation. Actual P2P simulations replicate the phenomena of (a) quantum superposition, (b) quantum indeterminacy, (c) quantum collapse, (d) the mind-body problem, (e) the problem of personal identity, while finally, (f) providing a clear model and mechanism according to which genuine free will/causal efficacy of mind in a higher reference-frame can result in an illusion of complete causal closure in a lower-level reference frame (namely, our reference frame within the purely relational/physical order we perceive through our senses).

[Eric Thomson]

Marcus, just focusing on point #1. My point was that structural properties can be intrinsic. It is an intrinsic structural property of this object that it is cube-shaped, for instance. There may be extrinsic structural properties too (e.g., this cube is above that cube).

More to the point, biology can give description of a thing's intrinsic properties without having to give a description of every intrinsic property associated with the thing. For instance, we can know that someone's brain is undergoing large-scale voltage fluctuations, even if the fundamental nature of voltage is not specified in that description. That doesn't mean that such voltage fluctuations aren't intrinsic features of brains, with measurable consequences in those brains. It just means that there are other features that this description leaves out.

Consciousness seems to be such a higher-level feature, not a fundamental feature like charge. Indeed, the main conclusion that the panpsychist needs is that consciousness is more like charge, less like all the other higher-level biological phenomena we study all the time like action potential generation, muscle contraction, and such. E.g., panpsychism is often the result when people try to approach the brain as a physicist rather than a biologist.

[Eric Thomson]

By relational properties I just mean those properties partly constituted by relations to properties of other objects, and am making no claims about whether the relata are ordinary objects (whatever that means).

Also, even assuming the contents of experience seem simple isn't enough to establish they are fundamental. The basis of an experience could be quite complicated, just as our ability to recognize a face seems simple or basic but we know the underlying mechanisms are extremely complex.

For that matter, don't most panpsychist thinks our particular species of experiences result from some kind of superposition of simpler protoexperiences? Because of that, I am not convinced that even under panpsychism our experiences should be taken as 'utterly simple.'

[Marcus Arvan]

Hi Eric: A few minutes after posting my last comment, I deleted it because I wasn't happy with what I wrote. I figured I'd gotten to it before anyone would have a chance to read it, but judging by your newest comment I was wrong.

Anyway, I was thinking it would be best to set the intrinsic/extrinsic distinction aside, given that we seemed to be using different notions, and instead focus on the relational/non-relational distinction. Because that's really the point. Science deals with relations -- with structures -- whereas things like redness seem non-relational.

I actually think your example of facial recognition, far from making the point you want it to, serves my side of the argument. Faces are defined structually (they are comprised by noses, eyes, mouths, etc. *related* in specific ways). That is why there is no Hard Problem of Facial Recognition. Although coding a facial recognition program isn't easy, one can see -- structurally, and functionally -- what a program would have to do to recognize faces. Yet redness is simple. It is just RED. Unlike a face, it cannot be described in structural terms. It cannot be described in any terms at all. It's just red.

As to your point about what panpsychists in general think, maybe you're right. I was just trying to state my own view.

[Eric Thomson]

Marcus my point in bringing up facial recognition was that subjective simplicity doesn't imply something is metaphysically fundamental or simple. That was the extent of that analogy. Ontology is hard, if not impossible, to read from phenomenology.

While I understand that 'redness' may seem nonrelational, that doesn't imply its underlying basis is nonrelational or simple.

[Marcus Arvan]

Eric: Yes, but my point was that neither (A) the analogy, nor (B) the ontological lesson you want to draw from it (viz. you can't read ontology off of phenomenology) is correct.

For my point was this: that however "simple" face recognition may seem phenomenologically, it quite easy to see how its apparent simpleness can be given a structural/functional explanation. I don't think this is the case for redness. Redness is a property. It is ontologically *there*, right before one's eyes. And so any adequate ontology must make sense of it. But -- I (and people like Rosenberg) say -- you just can't do it. Its nature -- right before one's eyes -- is simple in a way that facial recognition is not (faces have structure, redness doesn't).

In short: I (like Rosenberg, like Chalmers, etc.) think one *can* read ontology off of certain phenomenological experiences. The nature of redness is "right there" when you look at it.

[Marcus Arvan]

Here's another way of putting the point. When one recognizes a face, there is indeed a phenomenological sense in which the experience of recognition is simple. One has an "aha, that's a face" experience. But this is perfectly describable, and communicable in language. Phenomenal redness, on the other hand, is not *merely* simple. It has a certain inexplicable *quality* to it -- namely, looking red -- that is real but cannot be described. Given that it is a real quality, it is something -- a property, ontologically -- that exists and cannot be captured in the quantitative, relational terms of science.

[Gregg Rosenberg]

Eric, my treatments of these ideas are related to conceptual systems of things, within which categories of objects and properties exist. If it helps, here are the definitions I use in Chapter 12 of the book, where I deal with these issues in depth.

Definition 12.1 - Carriers - Objects or properties whose natures outrun the categories of a given schema but which can enter into the appropriate relations with one another [[to instantiate properties of the schema]] when put into the proper combinations.

Definition 12.2 - P is an extrinsic property within a system S if, and only if, P is instantiated within an instance of S and P has a nature that is not exhausted by its relations to other elements as they are defined within S.

Definition 12.3 - An internal contrast exists between A and B if, and only if, there is a comparative relation R such that necessarily, if A exists and B exists, then R(A,B).

A property is intrinsic "to a system" if its identity conditions can be given entirely by relations to other entities within a system of things to which it belongs.

A property is extrinsic "within a system" if a property is present within a system and it has a nature that is not exhausted by its relations to other elements as they are defined within that system.

A property is intrinsic tout court if it is not a property intrinsic to any system.

Your examples of intrinsic properties are ones that are intrinsic to systems, often with elements that are extrinsic to those systems. Marcus is referring to properties that are intrinsic tout court (i.e., have an element of their nature which is extrinsic within any system in which they exist). For example, a color is intrinsic tout court, because the nature of a color is not fully specified once you've described all its relations to other colors.

[Eric Thomson]

Thanks a lot Gregg, when I get time off work I will have to read more closely. As I said above, I was basing all my concerns on your talk you gave to prospective grad students at Indiana U (I think it was 1996), so I appreciate your patience here, will definitely take a look.

[Eric Thomson]

a color is intrinsic tout court, because the nature of a color is not fully specified once you've described all its relations to other colors

But I wouldn't claim that a color is specified (at an ontological level) by leaving it as a basic term, and specifying its relations to other such terms. That might be enough for a behavioral examination of color, but not its basis.

My claim is that that this particular color is identical to such-and-such neural process. E.g., let's say a large-scale thalamocortical oscillation of a certain type, just for argument's sake. I am fine with taking color as basic in some epistemic or phenomenological sense. But the move from there, to ontology, is where I do not buy into any of this. The evidence (e.g., from the physiology of color vision) simply doesn't support such a claim.

[Eric Thomson]

Murat Aydede and Güven Güzeldere's underappreciated article '"Cognitive Architecture, Concepts, and Introspection: An Information-Theoretic Solution to the Problem of Phenomenal Consciousness' has some relevant, and beautiful ideas that bear on this. E.g., they convincingly argue for the contingency of what we take as 'basic' experiences (e.g., colors), with a very nice discussion (Section 3 of the paper). Indeed I was just considering submitting this paper for this blog, and would if I had an extra 6 hours to write it up.

[Marcus Arvan]

Eric: once you admit that colors are intrinsic tout court, you have already given the game away -- for *nothing* in the physical sciences is intrinsic tout court: physical properties are *relational* all the way down. And you cannot save your position by saying that colors are epistemically intrinsic tout court but not ontologically (viz. in terms of identity to certain neural functions). The reason for this is simple: the neural firings are *not* intrinsic tout court; they are functional/relational structures/processes. And you cannot -- in principle -- explain a property that is intrinsic tout court by reference to properties that are not. An analogy: why do we have no Hard Problem of Water -- that is, a problem of explaining how water (the liquid stuff around us) is identical to H2O? The answer is simple: we can *explain* all of the properties of water (it's liquidity, etc.) in *terms* of the molecular chemistry. But this is precisely what cannot be done with colors. You can say that colors are identical to certain neural firings all you want, but as long as a physical explanation of those neural firings leaves out what it is like to actually experience the color -- then, by Leibniz's Law, the experience is not, after all, identical to the neural firings. If it were, then by the law of identity a complete physical explanation simply would be a complete qualitative explanation of the experience. Which it isn't.

[Eric Thomson]

Marcus I didn't admit that colors are intrinsic tout court: I was quoting Gregg in my first paragraph but my italics tags didn't convert.

[Eric Thomson]

Quoting Marcus:
"You can say that colors are identical to certain neural firings all you want, but as long as a physical explanation of those neural firings leaves out what it is like to actually experience the color -- then, by Leibniz's Law, the experience is not, after all, identical to the neural firings."

If, by hypothesis, to have the experience is to be in that brain state, then I would no more expect to have that experience by knowing about that brain state, than I would expect to photosynthesize because I know about the physiology of photosynthesis. Indeed, I would predict that I do not experience red just by learning about someone else experiencing it.

But this is just steering away from Rosenberg's argument, into standard Hard Problem/Bat/Gap motifs that have a very well-worn path on the internet and don't need a voice here.

[Marcus Arvan]

Eric: but on my understanding that is Rosenberg's argument. For my part, I don't think the first part of Rosenberg's book -- the part on consciousness -- advances past the traditional rendering of the Hard Problem. His point there is simply the point I'm making: that science is structure, and things like color-experiences outstrip mere structure. It's the second part of R's book -- the part on causation -- that I find fascinating. His point is that it isn't just consciousness that has non-structural elements (as though there is something unique about *only* consciousness that defies physical/structural explanation). If it were merely consciousness that had non-structural feasutes, then perhaps the cognitive story you want to tell would be cogent (viz. you can know everything about photosynthesis without photosynthesizing) -- though, again, it don't think it is cogent. The point is, once we see that it is not *merely* consciousness that presents this problem -- not something merely in our brains -- but causation too that your kind of response (and common physicalist responses to the Hard Problem) has no purchase. For you can't explain away the fact that -- if Rosenberg is right -- that *causation* must have intrinsic properties tout court in terms of cognition. So now we have two elements of reality -- consciousness and causation -- that *both* present the same, physically insoluble problem. And that is what I take to be of such value in R's book. The lesson is that for reality to have physical structure at all, it *must* have intrinsic features tout court that are, by definition, beyond the scope of functional/scientific inquiry. And R's point is that things like color experience -- consciousness itself -- are the only things we are acquainted with that qualify (they cannot be described structurally). This is how his argument goes:

1. We know (from the case of causation) that the world must have intrinsic properties tout court that are simple, indescribable and beyond the scope of science.

2. Conscious qualities (such as colors) present themselves as simple and indescribable.

Thus 3. By inference to the best explanation, conscious qualities are the intrinsic tout court properties beyond the scope of science.

So, I'm not just rehashing old versions of the Hard Problem. I'm trying to press R's entire argument, which hinges crucially on the intrinsic-tout-court aspect of causation.

[Marcus Arvan]

A simpler way to put my last point: if R is right that causation has to have intrinsic elements beyond the scope of science, attempts to explain away the Hard Problem of consciousness in physicalist terms (the cognitive explanations you seem to want to appeal to) are under-motivated. If R is right, we already have independent reasons to think that one thing -- causation -- is beyond the scope of science. Once we know that, the next question is: which sort of thing *could* be the intrinsic features of causation? R says: consciousness fits the bill perfectly. Thus, the best explanation of the Hard Problem isn't any kind of cognitive limitation; the best explanation is an ontological one -- namely, dualism (or what R likes to call "liberal naturalism", but which by my lights is nothing less than a form of dualism).

[Gregg Rosenberg]

From the end of chapter 13 in the book, in support of Marcus' assertion about my argument,

"When one is first exposed to the idea of [[causal receptivity]],one's intellectual interest in the idea is sometimes accompanied by real misgivings...It can seem very hard to visualize appropriately. On the evidential side, people worry how we could ever get access to the supposed facts about receptivity. Because the physical facts underdetermine the receptive facts, the story about [[receptive causation]] may seem like rationalist speculation, unfettered. ...Even if receptive connections seem metaphysically required to account for the causal nature of the world, they seem conceptually obscure and epistemically opaque, and we may react with an urge to explain them away."
...
"Repeating, [[receptive causation]] seems to be metaphysically required as part of the causal structure of the world, yet it seems conceptually obscure and epistemically opaque. I now point out that consciousness presents itself as epistemically transparent and conceptually immediate: We have observational knowledge of it. Yet is seems metaphysically baroque, so we do not have full confidence in our observations. Not only are the phenomenal qualities that exist within consciousness brute features of nature, but also many of the apparent features of that experiential context seem extravagant, and accounting for them is awkward. The entire package is unmotivated by any deeper naturalistic considerations. Why should the world contain such a thing?"

"As an intellectual poser, consciousness is the mirror image of [[receptive causation]]. The problem is not observing that it exists or that it has many of those strange features. Evidentially, it and they are presented to us. But it is very difficult to believe what is presented to us because, metaphysically, it seems too queer and unmotivated a kind of thing: it has no natural place in the world. It is just a strange "nomological dangler" on an otherwise internally complete and self-consistent machine, a physical machine belonging wholly to the physical world."

"By adopting the Consciousness Hypothesis, the Liberal Naturalist can use the non-mysterious features of each entity (consciousness and causation) to address mysteries arising with regard to the other. A Theory of Natural Individuals incorporating receptivity predicts many of the most troubling aspects of phenomenology on independent grounds. The attractiveness of this strategy is obvious because it is clear that, if we adopt it, each strange phenomenon, consciousness and [[receptive causation]], undercuts the motivation for skepticism about the other. The receptive connection is epistemically opaque, but consciousness is not, and so it can be a model for a real, live, caught-in-trap receptive field. The phenomenal field of consciousness seems too strange to be what it seems, to arbitrary and brute, but the characteristics of the causal nexus created by receptivity are not arbitrary. Consciousness is strange in just the way [[our theory would predict]] a carrier of nomic content has to be strange."