I was recently reading through a new paper of Uriah Kriegel’s called The Same-Order Monitoring Theory of Consciousness where he says this
If consciousness were indeed a relational property, M’s being conscious would fail to contribute anything to M’s fund of causal powers. And this would make the property of being conscious epiphenomenal (see Dretske 1995: 117 for an argument along these lines).
This is, by all appearances, a serious problem for HOMT [higher-order monitoring theory a.k.a. Higher-order thought theory]. Why have philosophers failed to press this problem more consistently? My guess is that we are tempted to slide into a causal reading of HOMT, according to which M* produces the consciousness of M, by impressing upon M a certain modification. Such a reading does make sense of the causal efficacy of consciousness: after M* modifies M, this intrinsic modification alters M’s causal powers. But of course, this is a misreading of HOMT. It is important to keep in mind that HOMT is a metaphysical, not causal, thesis. Its claim is not that the presence of an appropriate higher-order representation yields, or gives rise to, or produces, M’s being conscious. Rather, the claim is that the presence of an appropriate higher-order representation constitutes M’s being conscious. It is not that by representing M, M* modifies M in such a way as to make M conscious. Rather, M’s being conscious simply consists in its being represented by M*.
So far this is all right (notice how Uriah has Rosenthal’s account correctly formulated in such a way as to be immune from certain unicorn arguments). I have also pointed out how this implicit assumption about what the higher-order thought theory is keeps people from thinking the theory is anti-Cartesian in certain important respects.
But Uriah goes on to say that
When proponents of HOMT have taken this problem into account, they have responded by downplaying the causal efficacy of consciousness. But if the intention is to bite the bullet, downplaying the causal efficacy is insufficient – what is needed is nullifying the efficacy. The charge at hand is not that HOMT may turn out to assign consciousness too small a fund of causal powers, but that it may deny it any causal powers. To bite the bullet, proponents of HOMT must embrace epiphenomenalism. Such epiphenomenalism can be rejected, however, both on commonsense grounds and on the grounds that it violates what has come to be called Alexander’s dictum: to be is to be causally effective. Surely HOMT would be better off if it could legitimately assign some causal powers to consciousness. But its construal of consciousness as a relational property makes it unclear how it might do so.
Now Rosenthal will be speaking about this issue at the ASSC, and Uriah is right that Rosenthal does not think that there is much, if any, function to consciousness qua consciousness, so I don’t want to get into that stuff. What I want to question is whether or not anyone who agrees with Rosenthal is committed, in the way that Uriah seems to think that they are, to saying that consciousness is epiphenomenal.
The brunt of the challenge seems to come from the claim that our being conscious of a mental state, and hence that mental state being conscious, does not change, or modify the first-order state in any way and so its causal powers are unaffected by being conscious. I think that this is right; in fact I use this as a premise in my argument that higher-order theories are committed to there being something that it is like for a creature to have a conscious thought. But does this claim entail that consciousness is epiphenomenal? I am not sure that it does.
I think that someone who like the higher-order theory could say that, while the first-order state does not come to have any new causal properties when it become conscious, the creature in which the state occurs does. So, at the very least, even by Rosenthal’s lights, we get the ability to report (as opposed to express) our mental states when they are conscious, and we get the ability to introspect our mental states and thereby come to know what it is like for us to have them.
Now whether they can say there is more to the function of consciousness than this is another question, but at the very least, one does not have to dine on the bullet that Uriah has prepared.
Richard Brown 
“notice how Uriah has Rosenthal’s account correctly formulated in such a way as to be immune from certain unicorn arguments”
No, I didn’t notice. What was the immunity granting-portion of the Uriah quote? The main thing I noticed was his “M’s being conscious simply consists in its being represented by M*” which would put it right in the path of the Unicorn insofar as the Unicorn involves arguing that there’s no such thing as being represented.
On to the topic of your post…
You write: “I think that someone who like the higher-order theory could say that, while the first-order state does not come to have any new causal properties when it become conscious, the creature in which the state occurs does.”
Two questions:
1. But isn’t whatever causal properties the creatrue gains due to the tokening of the HOT and not due to the state of which the HOT is a representation?
2. If so, isn’t the case that it’s not state consciousness that is adding causal oomph to the creature?
P3 of the Unicorn says “Representing something does not suffice to confer a property to that thing,” (C1 in the book version)…Uriah correctly points out that Rosenthal agrees with this and that his theory is not committed to it. So he shouldn’t be concerned with the Unicorn.
Answer to question 1: Yes, that’s the idea
Answer to question 2: No; state consciousness consists in there being a suitable HOT around and so the new causal powers are directly due to state consciousness, just not to the first-order state
So it can be true both that having conscious mental states serves a (selectively advantageous) function and that this is not due to the first-order state comming to have some new (causally efficacious) property…
Re: unicorn stuff.
The following quote needs to be read in such a way that it isn’t simply contradictory.
“It is not that by representing M, M* modifies M in such a way as to make M conscious. Rather, M’s being conscious simply consists in its being represented by M*.”
If a “modifies M” is read as entailing any change in its properties regardless of whether the properties are intrinsic or relational, then it is contradictory to say that M becomes conscious while undergoing no modification. However, it is pretty clear that what Uriah has in mind is a reading of “modifies M” wherein it is read as entailing only change in intrinsic properties, since earlier in the passage you quote Uriah talks about the causal reading he and Rosenthal rejects in terms of “intrinsic modification”.
All of this would immunize against the Unicorn if the Unicorn was about whether being represented is an intrinsic property. However, the Unicorn involves denying that being represented is any kind of property whatsoever, intrinsic or relational.
—-
Re: the causal stuff. I don’t get it.
Help me understand why you aren’t comitting yourself to all three parts of the following inconsistent triad:
X consists in Y.
Y has causal powers.
X does not have causal powers.
If you want, let X = M’s being conscious
Y = M*’s being a representation of M
Or…maybe you think the triad isn’t inconsistent? Maybe there’s some wiggle-room in “consists in” that I’m missing?
I don’t think that it is intrinsic v, relational that is at play here. It is that the higher-order state does not make any non-cambridge change to the first-order state, but the real issue is just that P3/C1 is not somethng that Rosenthal is committed to and since that is the main target of the Unicorn, Rosenthal doesn’t have to worry about it…that is unless you can reformulate it so that P3/C1 does not play such a pivotal role in the argument…
As for the inconsistent triad…I don’t see how I am committed to all three of those; in facrt I have alread denied the third…X does have (new) causal powers, just not in virtue of the first-order state.
Re: unicorn stuff.
Switching to an interpretation of the Uriah quote in terms of cambridge changes vs. non-cambridge changes doesn’t acheive immunization either. Even if the change to being represented is a cambridge change it is still vulnerable to the Unicorn insofar as the Unicorn argues that there is no such thing as being represented.
Re: causal stuff. I am totally lost. I assume that M’s causal power’s supervene on its intrinsic properties. M’s change to being represented changes none of its causal powers. So M’s being represented is a causally inefficacious cambridge relational property of M. Thus “X does not have causal powers” where “X = M’s being conscious”. How can you deny this part of the triad? Is it that you reject the assumption that causal powers supervene on intrinsic properties?
Re US:
There really isn’t any ‘switching’ as this is the way that Uriah characterizes the dispute earlier, but that is irrelevant…the point is that Rosenthal agrees with your P3/C1, which you take to show that his theory is false, so something is going on here…
The Unicorn does give an argument for there being no such thing as being represented, but that argument is likely unsound…in particular, I think Rosenthal would accept the first premise, and deny the second (because he has a suitably loose sense of what a property is)
Re CS:
I do not deny the supervienence claim.
M’s being conscious has nothing to do with M (on Rosenthal’s view)…in fact he says that M’s being conscious is not really a relational property that it (M) has, but that this is a loose way of talking that suits his purposes. So, on his view, M’s being conscious consists solely in there being a suitable HOT, and that can happen (again, on his view) even if the first-order state is not present. So the causal powers that the HOT has=the causal powers that M’s being conscious has.
After re-reading section 4 of Ch. 3, let me forstall an objection that I can hear you about to make.
Saying this does not mean that the HOT is the conscious state. For the HOt to be conscious, we would need another state that targets it. M’s being conscious consists in there being a suitable HOT, that does not make the HOT conscious, that make M conscious.
[…] an earlier post (The Function of Consciousness in Higher-Order Theories) I argued that higher-order theories were committed to saying that consciousness had very little […]
[…] Disregarding the ‘functionally/physically identical’ bit, a zombie on the higher-order theory of consciousness is a creature that has all of my first-order states but none of my higher-order states. There will be nothing that it is like for this creature to have any of its mental states, even though he and I will be pretty much behaviorally indistinguishable (since conscious mental states have very little function on the higher-order theory (but not ‘no function’, as I argued in The Function of Consciousness in Higher-Order Theories)). […]
[…] objection from the possibility of the HOT occuring in the absence of the first-order state, or as I argued, from Uriah’s charge that higher-order theories, like Rosenthals’s, that claim that the […]
[…] is still there but is inhibited…this might pose a problem for Rosenthal’s argument that conscious states do not have any fucntion…but I will leave that for another […]
[…] The Function of Consciousness in Higher-Order Theories […]