Unicorns and Monitoring Theories of Consciousness

I just finished a draft of a paper and put it up on my website. Comments welcome.

Link: Beware of the Unicorn: Consciousness as Being Represented and Other Things that Don’t Exist

According to monitoring theories of consciousness (including both higher-order representational theories and same-order representational theories), a mental state’s being conscious consists in a mental state’s being represented. I argue that while there is such a property as a state’s being conscious there is no such property as a state’s being represented. I can mentally represent unicorns without their existing. Failing existence, unicorns instantiate no properties, including the so-called property of being represented. Whatever representation amounts to, representing something does not entail the instantiation of the property of being represented. If there is no such property as being represented, then the property of being conscious cannot consist in it. While my primary targets are monitoring theories, I briefly consider whether first-order representational theories might be vulnerable to similar objections. First-order representational theories arguably may be read as holding that the phenomenality of a property consists in its being represented. If there is such a property as being phenomenal but no such property as being represented, then a property’s being phenomenal cannot consist in its being represented.


10 Responses to “Unicorns and Monitoring Theories of Consciousness”

  1. Brian says:

    P2 of the Unicorn seems extremely strong! In its support, you appear to give an argument of the following form:

    (A) We think about Unicorns, even though they don’t exist.
    (B) Unicorns don’t instantiate the property of being thought about.
    (C) The property of being thought about doesn’t exist.

    But see the following argument with the same form:

    (a) Little girls love Unicorns, even though they don’t exist.
    (b) Unicorns don’t instantiate the property of being loved.
    (c) The property of being loved doesn’t exist.

    Surely something is wrong here! It would be very strange indeed if we accepted the conclusion that being loved is not a property, loves is not a relation etc. on the basis of this kind of argument. Maybe there is support for P2 coming from elsewhere, and I am just not seeing it?

    Still, I think that if one were to tinker with P2 a bit, then a version of the Unicorn argument would be effective against second-order versions of representationalism (e.g. Rosenthal’s theory). However, I don’t think that the Unicorn can be made to work against same-order or ‘reflexive’ versions of representationalism, since on these theories the existence of the relevant state is guaranteed (so you don’t get the unicorn-style cases; this is a big advantage of the SOR theories IMHO).

    It’s interesting to think about whether a version of the Unicorn might work against first-order representationalist theories, but this is a comlicated issue… I look forward to reading the end of your paper and seeing what you have to say about this.

  2. Pete Mandik says:

    Brian, Ithanks for talking a look at my paper and giving me some feedback. I find your remoarks helpful. Here are some further thoughts:

    I’d agree with you that the argument about love is just as strong as the argument about thinking. My reaction, though is “too bad for love”. I think a convincing demonstration that the argument is fallacious would have to involve something that had a much more obviously false conclusion. (Note that none of my remarks are necessarily unromantic: I affirm the existence of both love and lovers!)

    Regarding whether even an improved version would be ineffecacious against same-order theories, I adress that in the paper. SOR theories seem to just tack on the existence of the intentional objects of monitoring states. In brief, my line against this is something like “existence isn’t a property, so I don’t see how adding existence to being represented (which isn’t a property) gives you a property”.

    Re: applying the Unicorn against first-order theories, I say a bit more about that at the following post:

    Phenomenal Consciousness of Inexistent Colors

  3. Richard Brown says:

    Hey Pete,

    You know after rereading this unicorn paper I wonder if you know that the ‘revision’ that you propose is actually Rosenthal’s unrevised original position? It is possible on his account for a HOT to occur unacompanied by a first order state. In that case one is none the less conscious of oneself as being in that state even though they are not. You claim that if he says this then he does not implement the transitivity principle, but you don’t take into account his anti-kripkean line on aboutness. On his view your thought about a unicorn is really about unicorns even though they don’t exist and so to you HOT is really about the first order state that it represents you as being in. So he does think that this implements TP. NOw don’t get me wrong I am not agreeing with this, my sympathies are with Kripke, but I just thought you should know.

  4. Pete Mandik says:

    Richard, you’re right to point that out. Don’t you agree though, that its really puzzling for Rosenthal to say that TP would be implemented in such cases (regardless of what you think of aboutness)? Keep in mind that in such cases the so-called conscious state is a state that doesn’t exist. In Rosenthal’s 2002 “How many kinds of consciousness?”, he says of such cases that “there is no problem about how a nonexistent state can have the monadic property of being conscious” and my big problem with this just is the Unicorn problem: nonexistent states have no properties whatsoever.

    If this is what consciousness amounts to, why not just say that there is no such property as being conscious?

  5. Richard Brown says:

    Yeah I do think that there SEEMS to be a tension between two things that are independently quite plausible. On the one hand we have TP which seems very near to our common sense conceptions. A conscious state is simply one that we are aware of being in. This would make it seem very odd to have a conscious state that was not there! On the other hand we have the theoretical claim that the way in which we are aware of being in those states is by our representing ourselves as being in them. One of the things about representation is that it comes with the possibilityof misrepresentation, so it seems perfectly natural to talk about representing ourselves as being in a state that we are not in fact in.

    Now is there REALLY a tension between these? Not when one sees that TP does not commit us to the view that the first order state literally gets transformed into a conscious state. Rather a conscious state is one that we are conscious of ourselves as being in and we can be consious of ourselves as being in a state that we are not in. This is still an implemention of TP because I am conscious OF myself as being in a certain state. You get your unicorn-type intuitions because you want TP to mean that I am conscious of THAT state, but for Rosenthal you are not conscious of THAT state (pointing at a state that isn’t there) you are conscious of yourself AS BEING in that state (which isn’t really there).

    Of course one may at this point try to mount an argument against TP from another direction; namely that the contrast between the Dretske-style view and the Rosenthal-style view seems to collapse…maybe we can talk about this later tonight (if we don’t have anything better to do, which I hope we do!)

  6. Pete Mandik says:

    Hey Richard,

    I plan on being too busy tonight laying down a totally heavy bass line to say this then, so I’ll say this now:

    1. Thanks for all the feedback on the unicorn stuff. I find this very helpful in clarifying my thoughts on the matter. (also: fun).

    2. I should clear up a confusion that has creeped up in these comments. In my paper I do not accuse this particular interpretation you are discussing as failing to implement TP. What I say is along the lines of: if you revise HOT theory so that the HOTs themselves are the conscious states, then you have failed to implement TP.

    3. Maybe you are right about where I get my unicorn-type intuitions, but here’s how I’d put the point. I get my intuitions from wanting there to be such a property as being conscious. If the consciousness conferring powers of HOTs reach out into the realm of unicorns and leprechauns, how is this different from fictionalism about consciousness? As you yourself say, on such an interpretation “TP does not commit us to the view that the first order state literally gets transformed into a conscious state.” Maybe this is a true theory of consciousness, but now what the response to my argument becomes is to deny my premise that there is such a property as a state’s being conscious.

    4. Let the record show, the view you are describing is committed to the following: two people can be in subjectively identical global mental states, global states that do not differ in what it is like to be in them, yet differ in whether they have conscious states because one has a true HOT and the other has a HOT that refers to an non-existent state. My objection to this is the “too weird” objection and it goes like this: That is too weird!

  7. [...] Are there belief-dependent properties? More specifically, does having a belief that a is F suffice to confer any properties whatsoever to a? If you were a certain kind of idealist, you would think that such a belief sufficed to confer the property of being F to a’s. If you weren’t an idealist, but nonetheless bought into certain kinds of relational accounts about belief, then you would think that the belief in question at least conferred to a’s the property of being believed. As I’ve argued in “Unicorns and Monitoring Theories of Consciousness” there are no such properties and this spells bad news for various theories of consciousness. Here’s Dennett on belief dependent properties. And snakes: (1) Many people (wrongly) believe that snakes are slimy. [...]

  8. Richard Brown says:

    For the record, those two people do not differ in conscious states. They each are aware of themselves as being in the same state and what it is to be in a conscious state is to be aware of yourself as being in it. One of them is right the other wrong. It really is no big deal, just as if I was hallucinating a knife and you were veridically seeing a knife. Seems the same to both of us, who cares? Also we could use third person techniques to determine who actually did have the first order state and who didn’t because that state will opperate in its normal way independently of my being aware of myself as in it.

  9. Pete Mandik says:

    Richard, do you agree, though, that only one of them has a conscious state that actually exists?

  10. [...] One thing I find problematic with such a view is that it involves the identification of what something is (in this case, with what its location is) with how it is represented. Or, more briefly, it identifies something’s being F with its being represented as F. I think this notion is deeply problematic and one of the main problems I have with it is that I think nothing can consist in its being represented. A very brief argument for this view goes as follows. We can represent things that don’t exist. But failing existence, those “things” instantiate no properties whatsoever. Thus, representing is not property confering. Nothing instantiates a property in virtue of being represented. (I say more about this line of thinking in my “Unicorns and Monitoring Theories of Consciousness”) [...]