Chapter 2: The Argument against Physicalism

[hypnagogue]

I should note that in this post, the claim that the arguments in this chapter should essentially polarize thinkers between eliminative materialism and various forms of antiphysicalism is mine. Rosenberg doesn't state this anywhere in the chapter, although by my sights it's where his argument winds up taking us.

 

[hypnagogue]

I think the point being made here by a couple of us is that he has not shown Q -> ~P. He assumes it.

Perhaps you aren't swayed by the argument, but you certainly can't say he assumes it. Pages 21-25 are dedicated to defending the premise that bare difference does not entail qualitative content. Rosenberg's basic strategy here is to show how bare difference characterizations of qualitative content consistently conflict with what we observe via first person introspection.

There is a sense in which the argument hits a certain rock bottom, in that it ultimately appeals to empirical, first person observation rather than continuing an abstract train of logic. For instance, on page 22:

[phenomenal] colors are contents instantiating a structure of difference relations, not structures instantiated merely by difference relations ... our acquaintance with the phenomenal qualities yields information about them as contents occupying slots within these difference structures. Reification of the difference structure as basic ignores the grounding of those differences in each specific case and so ignores the content instantiating those structures.

The argument cannot go much further than this, because the premises here are empirical claims that cannot be supported by further discourse, but must be verified via first person observation of qualia. On a glance, this might seem to have the flavor of an assumption, because we cannot dig any deeper into the argument on a purely abstract level. But Rosenberg clearly does not assume the truth of the premise; he says the truth of the premise is born out empirically by one's own subjective experience.

So now the question becomes, even if the above is not an assumption, is it true? This can only be answered on an individual basis, but if everyone's subjective experience is more or less similar to mine, I cannot see how one could dispute the above claim. The different aspects to my subjectively experienced color space are not mere differences. I see this as different from this not because they just 'are' somehow fundamentally different and that's that; I see them as different because the qualitative content of the former is different from that of the latter.

There may be a cause of some confusion here that should be clarified. When Rosenberg talks about structures of difference relations for colors and the like, he is restricting his commentary purely to the facts about subjective experience. He is not referring to structures of difference relations that might exist in the brain; rather, he is referring to structures of difference relations that can be abstracted from subjective experience. So the relevant difference relations about colors here would not of the type "neural assembly X in V1 activates under such-and-such conditions," etc. Rather, they would be of the type, "phenomenal red is different from phenomenal violet"; "phenomenal orange appears 'closer' to phenomenal red than it does to phenomenal violet in the phenomenal color space"; and so on. It seems obvious upon introspection that we do not e.g. judge red's difference from violet to be a fundamental, ungrounded fact, but rather that we judge red and violet to be different because their respective qualitative contents are different.

 

[honestrosewater]

I don't think there is any substantial difference. It's just more convenient to define something as an observable, as opposed to having to say something like "an observable phenomenon" over and over.

The difference I saw as that an observable is an object while being observable is a property. But if an object is an observable, it has the property of being observable and vice versa. Eh, at least I saw the trees.

I have a different interpretation of what Rosenberg means by observables. It seems to me that observables could exist in a pure Life world. See my last response to Fliption's post.

Yeah, I was interpreting "awareness" and "we" in the narrower, more usual way.

I don't know if this has been a point of confusion for you, but if it helps, all the argument forms you listed are part of one overarching argument. The main argument is what you have listed as As; the Bs are a separate argument in defense of the premise A1; the Cs and Ds are used to argue for premise B2. They are not independent arguments, but rather, they are nested inside of each other.

edit: And just to be perfectly clear... the Cs are not an argument, but a definition.

Yeah, that's why I was hoping to condense them.

So now the question becomes, even if the above is not an assumption, is it true? This can only be answered on an individual basis, but if everyone's subjective experience is more or less similar to mine, I cannot see how one could dispute the above claim. The different aspects to my subjectively experienced color space are not mere differences. I see this as different from this not because they just 'are' somehow fundamentally different and that's that; I see them as different because the qualitative content of the former is different from that of the latter.

There may be a cause of some confusion here that should be clarified. When Rosenberg talks about structures of difference relations for colors and the like, he is restricting his commentary purely to the facts about subjective experience. He is not referring to structures of difference relations that might exist in the brain; rather, he is referring to structures of difference relations that can be abstracted from subjective experience. So the relevant difference relations about colors here would not of the type "neural assembly X in V1 activates under such-and-such conditions," etc. Rather, they would be of the type, "phenomenal red is different from phenomenal violet"; "phenomenal orange appears 'closer' to phenomenal red than it does to phenomenal violet in the phenomenal color space"; and so on. It seems obvious upon introspection that we do not e.g. judge red's difference from violet to be a fundamental, ungrounded fact, but rather that we judge red and violet to be different because their respective qualitative contents are different.

That isn't what I was thinking (about what's happening in the brain), but I think I get it now. We can find or abstract bare difference structures both above and below the level of qualitative content. He was talking about the above level, and I was talking about the below level. The above and below levels correspond to what's usually called mental and nonmental, respecitvely.?

 

[honestrosewater]

B1. The fundamental properties of a pure Life world consist of bare differences.
B2. Facts about phenomenal consciousness include facts about qualitative content.
B3. Facts about bare differences cannot entail facts about qualitative content.
B4. Therefore, some facts about phenomenal consciousness are not entailed by pure Life facts.

D1. Some thoughts and memories are observables.
D2. If thoughts and memories are observables, then the evidence for them is observable.
D3. Phenomenal contents (i.e. qualia) provide evidence for observable kinds of thoughts and memories.
D4. Therefore, qualia are observable.

E1. If x has the status of being an observable, then the evidence for x must also have the status of being observable.

By D1 (and the Cs), some facts about bare differences are observables, right? By D3, the evidence for some facts about bare differences is qualitative content, right? So, by E1 (and D2), in some way, facts about bare differences entail facts about qualitiative content (the fact or status of being an observable- Edit: and thus existing). How does this not apply to B3? B3 applies to only certain kinds of facts?

Edit: Also, if qualia are observables, what observable evidence do we have for them? Do qualia provide evidence for themselves or not need evidence? If so, why could thoughts and memories not provide evidence for themselves or not need evidence? I'm not clear on the relationship between an observable and its evidence.

 

[honestrosewater]

I think the point being made here by a couple of us is that he has not shown Q -> ~P. He assumes it. He therefore leaves it as possible that both P and Q may be true statements.

Specifically, as I think someone has already said, he has failed to prove (or convince us) that a pure Life world does not have qualitative content.
C: If the world does not have qualitative content (~Q), then the world's ontology is not composed of bare differences (~P).
D: If a pure Life world does not have qualitative content (R), then a pure Life world's ontology is not composed of bare differences (S).
S is false by definition, but he hasn't proven (or convinced us) that R is true.

 

[Canute]

Perhaps you aren't swayed by the argument, but you certainly can't say he assumes it. Pages 21-25 are dedicated to defending the premise that bare difference does not entail qualitative content.

Yes, he does argue this. I'm suggesting that he has not succeeded. It is not hard to show that our experience of q-content requires the existence of something more than bare differences. But this 'something more' does not itself consist of qualitative content, for q-content would not exist without this other thing, the thing that discriminates between qualities. He has not shown that q-content in itself does not consist of bare differences. This is a subtle distinction but I feel it's an important one.

But Rosenberg clearly does not assume the truth of the premise; he says the truth of the premise is born out empirically by one's own subjective experience.

Our own subjective experience does not show that the difference between red and blue is anything more than a bare difference. However it does show that we are aware of the difference, whether it is bare or not.

So now the question becomes, even if the above is not an assumption, is it true? This can only be answered on an individual basis, but if everyone's subjective experience is more or less similar to mine, I cannot see how one could dispute the above claim. The different aspects to my subjectively experienced color space are not mere differences. I see this as different from this not because they just 'are' somehow fundamentally different and that's that; I see them as different because the qualitative content of the former is different from that of the latter.

This illustrates why I don't like reifying differences in q-content. You say that the difference between red and blue is not a bare or mere difference. But I would argue that it is. It is the fact that we experience at all that clinches his argument, not the fact that red and blue are experienced as being different to each other and are therefore not bare differences.

Anyway, for myself I'm happy to leave this point. I accept his overall conclusion.

 

[Canute]

Specifically, as I think someone has already said, he has failed to prove (or convince us) that a pure Life world does not have qualitative content.

Yeah, that's it really. I'm convinced, but not because of the argument given here.

 

[Steve Esser]

Excellent discussion. The relationship between phenomenal content and the experience of it definitely gets addressed later in the book and it will be interesting to see what you guys think of it (although there is significant heavy lifting on causality in-between).

 

[Fliption]

It is not hard to show that our experience of q-content requires the existence of something more than bare differences.

He has not shown that q-content in itself does not consist of bare differences.

I'm not sure how these two statements can be consistent. I'm not sure I understand this point.

 

[hypnagogue]

Yes, he does argue this. I'm suggesting that he has not succeeded. It is not hard to show that our experience of q-content requires the existence of something more than bare differences. But this 'something more' does not itself consist of qualitative content, for q-content would not exist without this other thing, the thing that discriminates between qualities. He has not shown that q-content in itself does not consist of bare differences. This is a subtle distinction but I feel it's an important one.

But if q-content would not exist without experience, and experience is not entailed by bare difference, then it follows that q-content is not entailed by bare difference. (Bare difference can't entail experience, and q-content can't exist without experience, therefore bare difference can't entail q-content.)

Another perspective: If q-content really can't exist without experience, doesn't this indicate that there is something to q-content that is more than bare difference as well? If not, why couldn't q-content exist in a bare difference world that is devoid of experience? What is the nature of experience's contribution to the existence of q-content, and why could this contribution not be made by a structure of bare differences?

I think you're running into problems here because you're making a complete distinction between q-content and experience. I mentioned before that later on Rosenberg explicitly considers them to be 'two sides of the same coin,' so to speak, but even from just consideration of this chapter, I don't think we can completely separate the two. All the discussion has been about q-content as it is experienced in the first person, i.e., the q-content of phenomenal consciousness. I'm not sure it makes any sense to speak of such q-content in the absence of experience. If phenomenal red exists without being experienced, in what sense can we really say it is phenomenal red? It seems to me that it cannot be phenomenal red in any sense at all. The notion of q-content of p-consciousness seems to already suppose the experiential aspect you're considering here; we cannot talk about the q-content with already assuming the experience of it.

Our own subjective experience does not show that the difference between red and blue is anything more than a bare difference.

Rosenberg says of bare differences on pages 18-19:

bare differences are defined circularly in terms of their difference from each other. ... I say the difference is bare because it does not rest on any further categorical facts about the properties ... It is a difference that is ungrounded by any further facts about internal structural differences between those entities or internal relations of difference or contrast between unspecified intrinsic contents.

What could a bare difference between red and blue be like? If the difference between red and blue were bare, then we could say nothing more about them than that
1. They are defined circularly in terms of their difference ('red' is not 'blue'; 'blue' is not 'red').
2. They are associated with differing dynamics (e.g. they are associated with propensities to say things like "the sky is blue"; "the fire hydrant is red"; etc.)

That just sounds like a zombie to me. What makes us distinct from zombies is precisely that our acquaintance with phenomenal properties is not bare; there is more to what we experience about red and blue than what is included in 1 and 2.

Our notion of phenomenal red and phenomenal blue is not just that they are circularly defined as different from one another. We begin with independent, self-contained notions of the two: It's like something to see red, and it's like something to see blue. From these, we can observe that the difference between red and blue is not bare, but is grounded in the further fact that what it is like to see them is different. So in place of propsition 1 above, we would have something like

1'. It is like something to see phenomenal red (this).
2'. It is like something to see phenomenal blue (this).
3'. What is like to see phenomenal red is not what it is like to see phenomenal blue, therefore phenomenal red and phenomenal blue are different.

 

[Canute]

Originally Posted by Canute

It is not hard to show that our experience of q-content requires the existence of something more than bare differences.

He has not shown that q-content in itself does not consist of bare differences.

I'm not sure how these two statements can be consistent. I'm not sure I understand this point.

I can see that it looks like that. I'll try to answer, and this should answer Hypnagogue's point as well.

I agree with H that the existence of q-content shows that something more exists than just bare differences, and in this sense agree with Rosenberg's argument. What I was objecting to was that rather than take experience as fundamental, with q-content as the evidence for it, he takes q-content itself as fundamental. This gives the impression that he is arguing that q-content itself is fundamental and does not consist of bare differences.

I find this not so much wrong as confusing. Q-content in itself may be no more than bare differences, and many people argue that it is. If you are right about mind and matter being aspects of one fundamental kind of 'stuff', as you suggest elsewhere, then all differences are bare differences, whether they are qualitative of quantatitive. What cannot be just bare differences is the experience of those differences (whether they are bare differences or not).

It sounds like a pedantic point, but I was trying to say that just because we experience red and green, say, as qualitatively different does not mean that the difference between them is, in the final analysis, any more than a bare one. We experience/apprehend the gliders and other entities within Lifeworld as q-content also, so if the difference between red and green is not a bare difference then neither is the difference between gliders and other entities within Lifeworld (once they are in our consciousness). And what of the difference between red and a glider?

It seems to me that it is our ability to experience bare differences as different that shows that there is something more to the world than bare differences, not the existence of non-bare differences. It is true, as Hypnagogue says, that in a way q-content is the same thing as experience, since without experience there would be no q-content. But Rosenberg does not argue this, he stops at q-content expressed in terms of differences (red/green etc).

So while the existence of q-content shows that experience exists and that experience is something more than a bare difference, he does not show that, for instance, the q-difference between red and green is more than a b-difference.

I can see that this sounds like an odd objection, but can't see how to explain clearly why I feel it's an important one at the moment. I'll try to come back with something more to the point. Either way it's not an issue that need hold up the disussion. Steve Esser informs us that the difference between phenomenal content and experience gets dealt with later, so it can wait until then, and maybe it'll turn out to be a non-issue.

Perhaps another way of saying it that one cannot argue for qualities as being fundamental, but can argue for something that can experience those qualities as being fundamental, and Gregg's argument is a bit ambiguous in this respect. I think a committed physicalist would use this ambiguity to wriggle off the hook.

 

[honestrosewater]

Just a quick question- If all processes in a pure Life world are digital, does it matter if processes (especially processes in the brain) in a pure physical world are analog or a hybrid of digital and analog? Perhaps the only relevant point is that a pure physical world is bare difference structure.

 

[hypnagogue]

Just a quick question- If all processes in a pure Life world are digital, does it matter if processes (especially processes in the brain) in a pure physical world are analog or a hybrid of digital and analog? Perhaps the only relevant point is that a pure physical world is bare difference structure.

This concern is addressed on pages 25-26. Rosenberg acknowledges that the kinds of bare differences in physics are more subtle, complex, etc. than those of a pure Life world. But ultimately, what matters to the argument is not the subtlety or complexity of an ontology's bare differences, but that the ontology posits a pure bare difference world at all.

 

[honestrosewater]

If no one understood post #39, try this. If observables exist in a pure Life world, then qualia exist in a pure Life world, since, acc. to D4, qualia are observables.
I realize Rosenberg never claimed observables exist in a pure Life world. Still, he admits that information and instruments exist in pure (Life and physical) worlds. I don't see how he can deny that information and instruments also exist in impure worlds. It seems Rosenberg would agree that qualitative contents are information (they are meaningful to us in some context) and that we are instruments. If qualitative contents don't exist in pure worlds, either qualitative contents are a special type of information or we are a special type of instrument or both. By my understanding, as instruments, we are equivalent to zombies. And, by my understanding, zombies can exist in a pure world. If zombies can exist in a pure world, qualitative contents must be a special type of information. What is so special about qualitative contents per se (in themselves)?
IOW, using "detection" (and "detectibles", "detectors", etc.) as an analogue to "observation" (and "observables", 'observers", etc.), the difference between detection and observation is that detection can produce only facts about bare differences while observation can produce facts about qualitative content. There don't appear to be any differences between observers and detectors, per se (i.e. they function the same way). The difference between observation and detection must lie in observables and detectibles. If there's no difference between observers and detectors, the difference between observables and detectibles can't be the way they are processed by observers and detectibles. So what's the difference?

Perhaps I've made a mess of things, but it's because I can't see what is so special about qualitative contents. If it's that qualitative contents are a type of information that can't be derived from other types of information (and, furthermore, a particular qualitative content can't even be derived from other information of the same type), that leads me back to the processing of information which is presumably identical in both pure and impure worlds, especially in both humans and zombies.

 

[Canute]

I think this is the problem I've been talking about. There is nothing special about observables. However there is something special about observers. Or perhaps it would be better to say that observables are special, but it is the observer who makes them special, not some property intrinsic to the observable.

 

[hypnagogue]

If observables exist in a pure Life world, then qualia exist in a pure Life world, since, acc. to D4, qualia are observables.

Claiming that qualia are observables does not imply that all observables are qualia.

I don't see how he can deny that information and instruments also exist in impure worlds.

He doesn't deny that.

If qualitative contents don't exist in pure worlds, either qualitative contents are a special type of information or we are a special type of instrument or both.

You are right to point out that Rosenberg will need to explain what it is that differentiates the causal aspects of subjective experience from the causal aspects that are understood to belong to 'pure' physical systems. We won't get there for a while, though, as it will take most of the rest of the book to build up to the point where we do have such a causal theory of subjective experience (or at least the outlines of such a theory).

Perhaps I've made a mess of things, but it's because I can't see what is so special about qualitative contents. If it's that qualitative contents are a type of information that can't be derived from other types of information (and, furthermore, a particular qualitative content can't even be derived from other information of the same type), that leads me back to the processing of information which is presumably identical in both pure and impure worlds, especially in both humans and zombies.

Is it that you fail to see any relevance or force to the antiphysicalist argument in itself, or is it that you find the implications of accepting the argument to just lead to even more problems?

If it's the latter, then you're not alone. Thinking about consciousness seems to inevitably put us on that carousel. If we accept the antiphysicalist argument and accept the causal closure of physics, we seem to have no room left for p-consciousness to play any causal role, and we wind up with an unsatisfying epiphenomenalism. What makes this book original is that it proposes a new way to escape this quandary, but it will take some work to get there. You can view the rest of the book partially as an effort to answer to the concerns you voice here.

 

[hypnagogue]

I think this is the problem I've been talking about. There is nothing special about observables. However there is something special about observers. Or perhaps it would be better to say that observables are special, but it is the observer who makes them special, not some property intrinsic to the observable.

There is never any claim that there is any particular problem about observables. There is a claim that qualia in particular present problems. The only reason observables are brought into the discussion is to affirm that we do indeed have epistemic access to qualia, and therefore that we can even say something meaningful about them in the first place. Some deflationary arguments try to undermine qualia by establishing that we don't even have epistemic access to them, so Rosenberg needs to argue that we do.

And again, the special quality of phenomenal observers that you reference is so intimately bound up with the special qualities of qualia that we can't substantively differentiate the two. The problem of qualitative content already is the problem of a subjective experiencer. Affirming the 'special qualities' of qualitative content already is an affirmation of the special qualities of the subject of experience. We cannot coherently recognize or deny one without likewise recognizing or denying the other.

 

[Canute]

There is never any claim that there is any particular problem about observables. There is a claim that qualia in particular present problems. The only reason observables are brought into the discussion is to affirm that we do indeed have epistemic access to qualia, and therefore that we can even say something meaningful about them in the first place. Some deflationary arguments try to undermine qualia by establishing that we don't even have epistemic access to them, so Rosenberg needs to argue that we do.

And again, the special quality of phenomenal observers that you reference is so intimately bound up with the special qualities of qualia that we can't substantively differentiate the two. The problem of qualitative content already is the problem of a subjective experiencer. Affirming the 'special qualities' of qualitative content already is an affirmation of the special qualities of the subject of experience. We cannot coherently recognize or deny one without likewise recognizing or denying the other.

That's fine. I don't disagree (and haven't been disagreeing) with any of that.

 

[honestrosewater]

Claiming that qualia are observables does not imply that all observables are qualia.

Eh, guess I wasn't really thinking about that- just trying to introduce the subject of my post- which was about two different types of observables.

He doesn't deny that.

I meant I was saying something he would almost certainly agree with. Perhaps I should have said "could" instead of "can".

Is it that you fail to see any relevance or force to the antiphysicalist argument in itself, or is it that you find the implications of accepting the argument to just lead to even more problems?

The former. I finally deleted my last post after editing it a dozen times trying to clarify my problem. I'm starting to think I was just expecting too much, I don't know, detail or such. Focusing on the formal aspects of bare difference structures, I find your comments,

Our notion of phenomenal red and phenomenal blue is not just that they are circularly defined as different from one another. We begin with independent, self-contained notions of the two

quite interesting. If they lead me anywhere, I'll share, but feel free to move on.

 

[loseyourname]

Fair enough, but it seems to me that any construal of the qualitative content of p-consciousness that describes it as not intrinsic is just a variation of eliminative materialism. It seems that any acknowledgment of 'qualitative content' on this view is a rather hollow one that is closer to denying qualitative experience (writing it off as a misguided illusion of some sort) than it is to embracing it as an actually existent phenomenon.

You might think, but that certainly isn't what I'm proposing. I'll elaborate on this below.

If a complete physical theory doesn't give a complete description of the objective aspects of social behavior, it is because of difficulties in practice such as intractability, not because of difficulties in principle (at least, according to physicalism).

That isn't necessarily the thinking of scientists outside of physics. For instance, a complete accounting of all of the chemical reactions taking place in the head of a member of the Green party won't tell us that he opposes the WTO. The best we can do is study his brain and discover a correspondence between certain reactions and thoughts and feelings that he tells us he is having. If we had no one to tell us what he was thinking, however, we could never do this, but we could certainly still have the thoughts and feelings. Imagine that you are a neuroscientist dropped into a world full of beings that you have absolutely no capacity to communicate with in any way. You could study the reactions taking place in their brains all you want, but you will never learn the content of their thoughts or of what they communicate to each other. To go a little further, let's say their world is a pure physical world and that furthermore, they are all zombies. They can still communicate using qualitative concepts such as a hatred of their world's version of the WTO, and such communication will have qualitative content that your physical understanding of their world will never tell you anything about.

The antiphysicalist argument is that physics could not completely account for the facts about p-consciousness, even in principle. That's a much stronger claim than, and a claim of an entirely different nature from, the claim that physics can't completely describe social behavior, as you have cast it above. The latter is not a threat to physicalism, but the former is.

I'm saying that it might very well be that the former claim is not much different than the latter. It seems to me that physics by itself can never tell us anything about qualitative content, but that qualitative content can still exist in a pure physical world or even in a zombie world. As you said, zombie's can still write novels. Novels have qualitative content such as themes and character traits. This qualitative content clearly can exist in a zombie world, a world that you will agree is a pure physical world. It should be clear from this example alone that it is wrong to say that our world is not a pure physical world simply because it contains qualitative content.

Physics can tell us everything about the brains of humans, in which concepts such as grammar and narrative structure reside.

Let's say that all conscious beings in the universe died, and a piece of paper was left behind with the sentence "Bob is tall." Would not that sentence still contain a subject and predicate? By the same token, are you prepared to say that scrolls written in dead languages that no one in this world can understand have no grammatical structure or narrative content? That might be the case, but I hardly think that position or its contrary is uncontentious.

Of course zombies could write novels; that's the whole point. If zombies could not write novels, then there would indeed be something about novels that is fundamentally elusive to the physical account.

So are you saying that novels in zombie worlds have no themes or character traits or linguistic meaning? These are all qualitative attributes of novels. It seems to me that Gregg is really trying to get at something a little more fundamental than qualitative content. I think he has to, because it seems clear to me that physics can easily entail qualitative content. It just can't give any meaningful description of it.

 

[Steve Esser]

I think the anti-physcialist argument is using the term 'qualitative content' in a very specific way which doesn't exactly map to the use of the term in other contexts.

Late in Ch. 2, GR mentions a paper by Daniel Stoljar (you can find it http://philrsss.anu.edu.au/~dstoljar/onlinepapers/2COP.html ) which I got around to reading last weekend. In it, Stoljar proposes a distinction between 2 types of physicalism, t-physicalism (t stands for theory) and o-physicalism (o stands for object).

He says the distinction rests on 2 theses (which each have good support): first, physical theory of the sort we have (t-physicalism) only tells us about dispositional properties of physical objects (I think this is equivalent to saying relational or extrinsic properties). Second, for there to be dispositional properties there must also be categorical properties (intrinsic properties) such that the instantiation of the latter is metaphysically sufficient for the instantiation of the former. O-physicalism acknowledges the existence of categorical or intrinsic properties.

This argument, tracing its roots to B. Russell, is closely related to GR's bare difference structure vs. qualitative content. I thought the paper was helpful in drawing out this distinction further, although GR's work ultimately goes much further.

 

[loseyourname]

I think the anti-physcialist argument is using the term 'qualitative content' in a very specific way which doesn't exactly map to the use of the term in other contexts.

Late in Ch. 2, GR mentions a paper by Daniel Stoljar (you can find it http://philrsss.anu.edu.au/~dstoljar/onlinepapers/2COP.html ) which I got around to reading last weekend. In it, Stoljar proposes a distinction between 2 types of physicalism, t-physicalism (t stands for theory) and o-physicalism (o stands for object).

I believe I might have read that already. If not, I'm familiar by now with the distinction between the different kinds of physicalism. I'm perfectly sympathetic to this argument against physicalism. I do think that the physics we have describes only extrinsic relations without any attempt at substance ontology. I can see how intrinsic properties and the objects that bear them are never touched on. If "qualitative" in this case is meant by Gregg (and by hypnagogue) to be synonymous with "intrinsic," then I really have no problem with any of this chapter or the supporting arguments posed in this thread. The problem with using this argument as an argument against the physical entailment of consciousness remains, however. The qualitative content of conscious experience might not be intrinsic at all, despite the claims made in places that we know it is from experience (I certainly don't). I'll agree that his general argument against physicalism (t-physicalism anyway) is sound and I'll agree with his conclusion that not all facts are either facts of physics or entailed by physics. I just can't make the jump from there to concluding that facts about consciousness are in the set of facts not entailed by physics.

 

[honestrosewater]

Our notion of phenomenal red and phenomenal blue is not just that they are circularly defined as different from one another. We begin with independent, self-contained notions of the two: It's like something to see red, and it's like something to see blue. From these, we can observe that the difference between red and blue is not bare, but is grounded in the further fact that what it is like to see them is different. So in place of propsition 1 above, we would have something like

1'. It is like something to see phenomenal red (this).
2'. It is like something to see phenomenal blue (this).
3'. What is like to see phenomenal red is not what it is like to see phenomenal blue, therefore phenomenal red and phenomenal blue are different.

Am I completely out of it? In a pure Life world, it must be like something to detect "on" and/or it must be like something to detect "off". It has to be known that this state is "on" (or this state is "off"). It isn't enough to know "on" is not "off" and "off" is not "on". If I presented you with a Life grid of cells all in the same state and asked you if the cells would change states in the next time step, you couldn't answer unless you knew if the cells are currently "on" or "off". No? So how is detecting "on" different from seeing "red"? Why does Rosenberg not have to explain now the difference between observables that are qualia and observables that are not qualia?

Edit: Can I say that Rosenberg deals with three things: (1) formal systems, (2) interpreted formal systems, and (3) instantiated interpreted formal systems? Using the game of Life as an example, (1) would be the rules, (2) could be a grid drawn on a piece of paper representing the "cells", a cell having a penny placed inside of it representing the "on" state, an empty cell representing an "off" state, etc., and (3) would be me playing the game (moving the pennies according to the rules).
I don't see him examining a pure Life world, a pure physical world, and an impure world each as (1), (2), and (3). An examination of a pure Life and impure world would be enough. He examines a pure Life world as (1) and an impure world as (3) (when appealing to our subjective experience of qualia), but where does he examine a pure Life world as (3) or an impure world as (1)? What I have been asking for is an examination of both worlds as (2)- which he starts but doesn't complete, IMO. Until he examines the two worlds at the same level, all he's saying is that math isn't physics isn't the physical world- which isn't very impressive (Edit: with all due respect). Am I wrong? I want to see a pure Life world and an impure world examined at the same level- as formal systems, interpreted formal systems, or instantiated interpreted formal systems. Am I asking for too much?
Rosengogue: Do you have qualia (or epistemic access to qualitative contents through observation, whatever)?
Me: Yes.
Rosengogue: Well, here's a formal system, and it doesn't have qualia.
Me: So? It doesn't have me either.
Rosengogue: No, wait, come back. This system has something we can interpret as being you, as playing the role of you.
Me: And this system still doesn't have anything that plays the role of qualia?
Rosengogue: No.
Me: If I have qualia, but this system doesn't have anything that plays the role of qualia, how can it have something that plays the role of me?
Rosengogue: ??

 

[Steve Esser]

The problem with using this argument as an argument against the physical entailment of consciousness remains, however. The qualitative content of conscious experience might not be intrinsic at all, despite the claims made in places that we know it is from experience (I certainly don't). I'll agree that his general argument against physicalism (t-physicalism anyway) is sound and I'll agree with his conclusion that not all facts are either facts of physics or entailed by physics. I just can't make the jump from there to concluding that facts about consciousness are in the set of facts not entailed by physics.

OK, that's clear.

The other arguments needed are that we
1.have epistemic access to qualia thru experience.
2.qualia are not constituted by bare differences (if they are not extrinsic, they must be intrinsic (I think).

But then I definitely see how this second step argument is clouded at this stage by the still fuzzy (inter)relationship between qualia and experience which you, honestrosewater and Canute have identified as an issue. I think the problem comes in particular because this is a kind of static analysis: if qualitative content is a kind of stuff out there, why couldn't it be composed of bare differences? It is really the action of experiencing qualia which is at issue here.

I'm more easily swayed by the argument even at this stage because I guess I've always been inclined to see first-person experience (of qualia) as being essentially an access to the intrinsic nature of things, while the third-person (really inter-subjective) construction of physical theory as being necessarily about extrinsic properties.

 

[honestrosewater]

The other arguments needed are that we
1.have epistemic access to qualia thru experience.
2.qualia are not constituted by bare differences (if they are not extrinsic, they must be intrinsic (I think).

Qualia do have some extrinsic property(ies) by being observables, right? If so, they can be included in a formal system- an impure world or interpreted impure world- and that system can possibly be shown to be inconsistent or incomplete internally or relative to the real world- an instantiated interpreted impure world. If qualia can't be included in any way in a formal system, it seems to me that his argument is missing something. (Oy, I'm starting to annoy myself now )

 

[hypnagogue]

In a pure Life world, it must be like something to detect "on" and/or it must be like something to detect "off". It has to be known that this state is "on" (or this state is "off"). It isn't enough to know "on" is not "off" and "off" is not "on".

A system in a pure Life world could detect whether a certain cell was on or off at a given time step by observing the causal dynamics of the surrounding cells over time. The point is that there is nothing more to the ontology of on and off than just their differing dispositional properties over time.

Detection of 'on' in a pure Life world does not imply that it is like something (in the sense of Nagel) for the detecting system to detect 'on.' You seem to be using a much broader and more colloquial sense of the phrase that is not faithful to Nagel's usage.

Edit: Can I say that Rosenberg deals with three things: (1) formal systems, (2) interpreted formal systems, and (3) instantiated interpreted formal systems?

It would be more accurate to say that he deals with 'pure' formal systems and our universe, which he argues is not a pure formal system. (2) and (3) do not really add anything to (1); they do not tell us anything that doesn't already logically follow from (1). If something about (2) and (3) did not follow from (1), then we would have some sort of strongly emergent phenomena on our hands, and we'd have to add some extra rules to the rules already contained in (1). But that would break our stipulation that (1) is already the complete set of rules. So by definition, everything about (2) and (3) logically follows from (1) (given some set of initial conditions).

He examines a pure Life world as (1) and an impure world as (3) (when appealing to our subjective experience of qualia), but where does he examine a pure Life world as (3) or an impure world as (1)?

Examining a pure Life world as (1) is equivalent to examining it as (3), since everything about (3) logically follows from (1), by definition of what it means to be a pure Life world. On the other hand, it is up for grabs whether our world is a pure physical world; what we have to do is observe it (I suppose this meets your criterion for (3)) and see whether what we observe can follow from (1), the formal rules given by physics.

edit: Sorry, let me clarify something here. The initial conditions of any given Life world will not follow from the rules governing its evolution; to get a Life world off the ground, we need to suppose some set of initial conditions in addition to the dynamical rules. If this is what you meant by (2), then (2) does not logically follow from (1). But (3) certainly follows from a combination of the rules and a set of initial conditions. In any case, the basic point stands that nothing is really to be gained from an analysis of the evolution of a particular Life world. The argument is an argument in principle, and will apply to any instantiation of any given pure Life world.

Rosengogue: No, wait, come back. This system has something we can interpret as being you, as playing the role of you.
Me: And this system still doesn't have anything that plays the role of qualia?
Rosengogue: No.

You seem to be missing the essential point of the argument, which is that there is more to qualia than just whatever functional or dispositional roles they may play in a cognitive context. The physical causal dynamics of your zombie twin are identical to your own, so (barring interactionist dualism) if qualia play some effective causal role in you, there will be a corresponding effective causal role in the zombie. What is different is that the playing out of causal roles in you is associated with what it is like to be you (i.e. your p-consciousness), whereas for your zombie twin there is no p-consciousness associated with the causal roles; it's just the effective causal roles and nothing else.

 

[honestrosewater]

Detection of 'on' in a pure Life world does not imply that it is like something (in the sense of Nagel) for the detecting system to detect 'on.' You seem to be using a much broader and more colloquial sense of the phrase that is not faithful to Nagel's usage.

Yes, for comparison.

A system in a pure Life world could detect whether a certain cell was on or off at a given time step by observing the causal dynamics of the surrounding cells over time.

The game doesn't start until the initial conditions are specified. I wasn't sure why "The initial state of this cell is on" is not an "independent, self-contained notion" like "It is like something to see phenomenal red."

The rest of your post clears things up. Thank you very much.

 

[hypnagogue]

The qualitative content of conscious experience might not be intrinsic at all, despite the claims made in places that we know it is from experience (I certainly don't).

Rather than continue our discussion on peripheral issues about qualitative content and physics, for now I'd like to focus on this claim. I agree with Steve that that when Rosenberg talks about qualitative content, he essentially means a kind of intrinsic property. We can infer this from the form of his argument; he contends that qualia cannot be analyzed into structures of bare differences, and 'bare difference' is defined as a kind of dispositional property with no grounding intrinsic content.

Rosenberg argues that qualia have qualitative content (essentially, that they have intrinsic properties) in section 2.5, and I've presented partial summaries of the argument here and here. If you don't find this reasoning persuasive, can you indicate what premises or inferences you find contentious?

 

[hypnagogue]

The game doesn't start until the initial conditions are specified. I wasn't sure why "The initial state of this cell is on" is not an "independent, self-contained notion" like "It is like something to see phenomenal red."

I probably worded that poorly. What I meant to say was that in a pure Life world, the difference between 'on' and 'off' is a primitive, fundamental fact. They are not different in virtue of some further fact; they just are different, and there's nothing more to it than that. In the case of phenomenal red and phenomenal blue in our world, the experience of phenomenal red and the experience of phenomenal blue are the fundamental facts, and the difference between them is derivative on the fact that the qualitative experience of these colors is distinct.

 

[loseyourname]

Rather than continue our discussion on peripheral issues about qualitative content and physics, for now I'd like to focus on this claim. I agree with Steve that that when Rosenberg talks about qualitative content, he essentially means a kind of intrinsic property. We can infer this from the form of his argument; he contends that qualia cannot be analyzed into structures of bare differences, and 'bare difference' is defined as a kind of dispositional property with no grounding intrinsic content.

Rosenberg argues that qualia have qualitative content (essentially, that they have intrinsic properties) in section 2.5, and I've presented partial summaries of the argument here and here. If you don't find this reasoning persuasive, can you indicate what premises or inferences you find contentious?

Well, in the first case, the text you cited makes no argument. It's an appeal to empirical observation. He can contend that experiential content is intrinsic, but I'm just not sure why. I think it's a little difficult to make any claim about the fundamental nature of either observation itself or of the objects we observe. I just can't get my head around the problem of trying to figure out why it is that he thinks he can empirically "see" that content he experiences is intrinsic in nature.

What could a bare difference between red and blue be like? If the difference between red and blue were bare, then we could say nothing more about them than that
1. They are defined circularly in terms of their difference ('red' is not 'blue'; 'blue' is not 'red').
2. They are associated with differing dynamics (e.g. they are associated with propensities to say things like "the sky is blue"; "the fire hydrant is red"; etc.)

I can see the analogy Gregg tries to draw between a pure Life world and the world of physics, but they aren't the same. Physics does not define its properties as bare differences. Mass isn't "that which is not momentum" and so forth. Objects are defined in physics by the measurable quantities that they possesses at any given moment. It's hard to say that the difference from one physical object to another is ever "bare" in the same sense that the difference between an "on" cell and an "off" cell is in a Life world. I'm not certain that this matters, but certainly the book (at least in chapter two) doesn't make much of an effort to demonstrate that it doesn't.

If bare differences are really only defined circularly, as your above argument contends, then they really don't exist in the world of physics. The example you made with colors can be extended to quantitative properties as well. For instance:

What could a bare difference between 5 meters and 3 meters be like? If the difference between 5 meters and 3 meters were bare, then we could say nothing more about them than that
1. They are defined circularly in terms of their difference ('5 meters' is not '3 meters'; '3 meters' is not '5 meters').

If we're going to draw the same conclusion you did about color, then quantities such as 5 meters and 3 meters cannot exist in a pure Life world. Clearly they can, however, which makes the claim that content not defined in terms of bare differences cannot exist in a pure Life world dubious. Distance clearly can, as can many others things.

To be honest, it seems that what I am saying here is rather trivial. Even Gregg states that economies might conceivably exist in a Life world. I doubt that myself, but if he feels that way, I have no reason to argue as of yet. Economies clearly cannot be defined in terms of bare differences, which means that Gregg admits that properties not defined as bare differences can be entailed by a physics built up from bare difference structures. With this in mind, it seems that colors cannot be excluded from a Life world only on the basis of their not being bare difference structures.

 

[loseyourname]

OK, that's clear.

The other arguments needed are that we
1.have epistemic access to qualia thru experience.
2.qualia are not constituted by bare differences (if they are not extrinsic, they must be intrinsic (I think).

I would say we have experiential access to qualia through experience. That may sound trite, but accepting 1 depends on what you mean by "epistemic access." I will agree that we have access to the knowledge that something exists and that something else exists that experiences that something that exists. What I don't see us as having access to is the precise nature of this something that exists, qualia in this case.

Regarding 2, I'd like to comment on the equation you are making between bare differences and extrinsic properties. If bare differences are meant in the sense hypnagogue (and Rosenberg, IIRC) has been using them - x is x because it is not y, etc. - then that is certainly not what is meant by extrinsic. Extrinsic simply refers to a property of an object as it relates to other objects. The nature of these properties is not defined in terms of differences, but rather in terms of how the relationship can be expressed. It can be expressed qualitatively, but generally the physical sciences will express the relationship mathematically.

I'm more easily swayed by the argument even at this stage because I guess I've always been inclined to see first-person experience (of qualia) as being essentially an access to the intrinsic nature of things, while the third-person (really inter-subjective) construction of physical theory as being necessarily about extrinsic properties.

Well, that's just it. If you've always felt this way, then you're likely to be easily swayed by any argument that reaches the same conclusion you've already reached, especially when it appeals to your direct experience. Personally, I don't know whether or not I've ever experienced a quality that was intrinsic to the object that caused my experience. I'll agree that physical theories are descriptions of extrinsic properties, but that is simply because extrinsic properties are the only properties that the physical theorist can investigate empirically. It always seemed to me that there is no way to access intrinsic properties, because they are necessarily cut off from the causal chain that leads to experiential access. If they were properties that could be the cause of other properties, then they would be properties of objects as they relate to other objects, which seemingly makes them extrinsic. I know Gregg has worked a new theory of causation that supposedly removes this little difficulty, but I will go into those chapters with a sceptical mind.

 

[Fliption]

Well, in the first case, the text you cited makes no argument. It's an appeal to empirical observation. He can contend that experiential content is intrinsic, but I'm just not sure why. I think it's a little difficult to make any claim about the fundamental nature of either observation itself or of the objects we observe. I just can't get my head around the problem of trying to figure out why it is that he thinks he can empirically "see" that content he experiences is intrinsic in nature.

This was my exact point when I posted ealier in this thread about the argument for qualia being consider qualitative content i.e. intrinsic i.e. not entailed by bare differences...

Now it appears to me that this argument that qualia has qualitative content isn't a whole lot different from other arguments we've heard before. Unfortunately, these past arguments have always had rebuttals and I see that this one doesn't seem to be convincing to some of the participants here as well. It still seems as if we're stuck with "it is qualitative because it seems qualitative". Has any progress been made in this regard with this book or are we missing something? I personally agree with the idea. I'm just not sure the arguments in this book are any different or stronger than previous arguments. (I'm referring specifically to the point of qualia being qualitative and therefore not subject to explanation by physics)

 

[Steve Esser]

Regarding 2, I'd like to comment on the equation you are making between bare differences and extrinsic properties. If bare differences are meant in the sense hypnagogue (and Rosenberg, IIRC) has been using them - x is x because it is not y, etc. - then that is certainly not what is meant by extrinsic. Extrinsic simply refers to a property of an object as it relates to other objects. The nature of these properties is not defined in terms of differences, but rather in terms of how the relationship can be expressed. It can be expressed qualitatively, but generally the physical sciences will express the relationship mathematically.

Hi. On this point, the distinction you see between bare differences and extrinsic (or relational) properties doesn't seem like a distinction which matters. I guess I'm agreeing again with the author here (sorry - later on I'm sure I'll be more critical):
(p.26-top):
"The theoretical character of the basic properties is just the same in both cases: One stipulates at first that they are distinct and fleshes out their natures by designating laws that describe how they behave. The only real differences between Life and physics lies in such attributes as the complexity of the laws, the number and kind of dimensions the cells exist in, and perhaps non-local causation."(etc)

 

[loseyourname]

"The theoretical character of the basic properties is just the same in both cases: One stipulates at first that they are distinct and fleshes out their natures by designating laws that describe how they behave. The only real differences between Life and physics lies in such attributes as the complexity of the laws, the number and kind of dimensions the cells exist in, and perhaps non-local causation."(etc)

Well, I don't think he's taking the matter seriously enough, to be honest. I'll agree with him that a pure physical world or a pure Life world, lacking some form of grounding ontology, probably cannot exist. Just going on the assumption that it cannot, I think we can make a couple of basic inferences about the nature of this grounding ontology in a physical world and in a Life world. In a Life world, we already know exactly what it is. There is the circular definition of states as simply two states that are distinct from each other. In this case, in fact, the nature of the grounding ontological objects doesn't matter aside from the fact that there are two kinds of objects and that each object is both 1) extrinsically the same as other kinds of objects in its kind and 2) extrinsically not the same as objects not in its kind.

In a physical world, though, I don't think we can have that same kind of grounding ontology. Particle A and particle B relate to each other by means of quantities that are various in number and can themselves vary in degrees (though the degrees are quantized, they are certainly more than two). This isn't just a matter of having more complex relational laws - it is also a matter of having a more complex and rich grounding ontology.

I know that, in a sense, what I'm saying here is not important to the central point that Gregg is making, but I still think it should be pointed out simply as a matter of literary criticism, if nothing else. His basic argument really just seems to be something along the lines of:

1. All physical data is extrinsic in nature. (uncontentious from my vantage point)
2. Experiential data is intrinsic in nature. (supposedly provided by direct experience, but I will argue this one)
3. No intrinsic data can be entailed by extrinsic data. (by the definition of the terms being used, I have no qualm with this)
Therefore, experiential data cannot be entailed by physical data.

Why not just leave it at that? He may as well admit that he is making a judgement call to some degree, by saying that experiential data seems to him to be intrinsic in nature (though I still can't conceive of what the difference would be between experiencing extrinsic v. intrinisic data). It seems that, by making analogy to a Life world - a simplified world in which, again, it just seems obvious that experience could not exist - he has proven that experience cannot exist in a physical world. I've mulled this point with hypnagogue by saying it is clear there are things that can exist in a physical world but cannot exist in a Life world, so this is not a convincing argument. He's countered by saying that isn't the argument, that there is something else special about experience that makes it impossible in both a Life and a physical world. But if this is the case, why make the argument that physical facts cannot entail experience because Life facts cannot entail experience (the argument explicitly made at the outset of the chapter). Doing so only serves to confuse and lead us off on tangents such as these. Just stick with the argument I've presented above and admit that still nothing has been proven. The rest of the book can continue and that's fine. I'm guessing he might make a very compelling case for his view of causation and experience later in the book. So be it. He doesn't need to pretend that he has disqualified opposing theories before he can present his own. If his theory has additional explanatory power, then it is worth investigating on its own merit.

 

[hypnagogue]

Well, in the first case, the text you cited makes no argument. It's an appeal to empirical observation. He can contend that experiential content is intrinsic, but I'm just not sure why. I think it's a little difficult to make any claim about the fundamental nature of either observation itself or of the objects we observe. I just can't get my head around the problem of trying to figure out why it is that he thinks he can empirically "see" that content he experiences is intrinsic in nature.

The text I cited appeals to empirical observation, but this appeal is made in the service of a deductive argument. Here's the basic form:

1. If a difference is bare, then it is a difference that does not rest on any further facts. (pg. 18)
2. The difference between phenomenal red and phenomenal blue rests on the further fact that what it is like to see this is different from what it is like to see this.
3. Therefore, the difference between phenomenal red and phenomenal blue is not a bare difference. They are not merely different, rather they are different in virtue of the fact that what it like to see one is not what it is like to see the other.

Rosenberg argues that a bare difference analysis of subjective experience (whether it characterizes phenomenal colors as bare differences, or as being entailed by bare differences) will leave out the phenomenal qualities. For instance, suppose we have a complete bare difference analysis of phenomenal colors (call it B) that consists of a set of statements like "p-red is different from p-blue"; "p-orange appears 'closer' to p-red than p-violet"; etc. B does not exhaustively characterize the phenomenal color space; it leaves out the phenomenal content that grounds the differences and relationships among the colors. B does not entail the facts about phenomenal content either, because (for instance) if we inverted the phenomenal color spectrum (so that p-red refers to this, p-violet refers to this, etc.), all the facts in B would remain true but would now be consistent with an entirely different phenomenal color space.

If we're going to draw the same conclusion you did about color, then quantities such as 5 meters and 3 meters cannot exist in a pure Life world. Clearly they can, however, which makes the claim that content not defined in terms of bare differences cannot exist in a pure Life world dubious. Distance clearly can, as can many others things.

To be honest, it seems that what I am saying here is rather trivial. Even Gregg states that economies might conceivably exist in a Life world. I doubt that myself, but if he feels that way, I have no reason to argue as of yet. Economies clearly cannot be defined in terms of bare differences, which means that Gregg admits that properties not defined as bare differences can be entailed by a physics built up from bare difference structures. With this in mind, it seems that colors cannot be excluded from a Life world only on the basis of their not being bare difference structures.

Bare differences are "ungrounded by any further facts about internal structural differences ... or internal relations of difference or contrast between unspecified intrinsic contents" (pg. 19). Economies in our world and prospective economies in a pure Life world cannot be characterized as a set of bare differences, but they do have further facts about internal structural differences that can be ultimately accounted for by bare differences. If physical spacetime can be shown to be a phenomon that arises from a more abstract set of functional relationships, then the same is true of physical distance. In a pure Life world, the difference between a string of 5 cells and a string of 3 cells could be reduced to bare difference facts about individual cells.