Tuesday, February 1, 2011

what it's like: what's HOT got to do with it?

hot hot hot hot hot hot hot
Here's a little something for you aficionados of the debate over the higher-order thought theory of consciousness. Adherents of the theory, HOT-heads, need a response to problems surrounding the puzzling case of empty HOTs: higher-order thoughts that, so the story goes, would make a state conscious if only there existed a first-order mental state for the HOT to be about, but are unaccompanied by any such first-order state. What to say about such a case? Well, one thing they can and do say (sometimes) about such a case is to embrace it as an unproblematic possibility: sometimes there's a HOT, but no first order target, and thus, no state that is conscious. Call this scenario (H &  ~C).

Is embracing (H &  ~C) bad for the HOT-heads? It would be bad if they also embraced an entailment from there being a HOT to there being a conscious state, call this (H -> C). Clearly (H & ~C) and (H -> C) can't both be true. So, what to do? If you're a HOT-head who's also embracing (H & ~C), then you better find some grounds for denying (H -> C). Can any grounds be found? Let's see.

One way an objector, an anti-HOT-head, might try to defend (H -> C) is by linking H to C by way of the notion of what-it's-like. So, the existence of a HOT entails the existence of a state in virtue of which there's "something it's like", a what-it's-like-ness, and the existence of a what it's like state entails the existence of a conscious state. Abbreviating: (H -> W) & (W -> C).

At this point the HOT-head can attempt a case against this linking move by suggesting the separable instantiability of state consciousness and what-it's-like consciousness. Just because there's a state in virtue of which there's something it's like, they might say, it doesn't follow that that very state is one that is conscious. Going just a bit further, the HOT-head might, following a recent suggestion by Richard Brown, say that in the empty HOT case, even though there's no state that has state consciousness, there's a state that has phenomenal consciousness, and further, the phenomenally conscious state is the HOT itself.

At this point, an unsympathetic reader may find the resultant view insufficiently motivated. Here's what strikes me as a problem: If state consciousness and phenomenal consciousness are separably instantiable like this, then what motivates saying, for instance, that phenomenality or what-it's-like-ness attaches only to HOTs? Given the scouted severing of what it's like and state consciousness, why couldn't a plain-old first-order thought give rise to "something it's like"?

Consider: why is it plausible that HOTs give rise to what-it's-like-ness? Well, it seems (pun!) to do with the fact that HOTs give rise to (or are) appearances: if I have a HOT to the effect that I have a first-order green perceiving, then that's how things will seem to me regardless of whether the HOT is true, false, or empty. Phenomena, phenomenality, phenomenology, phenomenal consciousness  -  all those fancy "ph" ways of talking about what's like - are all in the service of tracking appearances, the ways things seem. But appearances go along with first-order states too: If I have a first order thought that there's a dog on the blanket, then that's how things seem to me regardless of whether the thought is true, false, or empty. It will seem like there's a dog on the blanket. What is it like to be me? It's like thinking that there's a dog on the blanket, dude!

So what's the big deal about HOTs vis a vis phenomenal consciousness? Why not phenomenal consciousness without HOTs, say, first-order cognitive phenomenology? One sort of answer I've come across on a few occasions goes like this: what's crucial concerning what it's like is what it's like for me, and in order for some mental representation to give rise to the relevant appearance, it has to represent me in some way, a way that makes it higher order. The plain old first order thought that there's a doggy on the blanky doesn't represent me and thus it doesn't give rise to an appearance of how things are for me.

I'm having a hard time seeing this line of thought here as at all convincing, so maybe I'm misremembering it. But anyway, I'm not seeing why the representation's being in me doesn't suffice to make the subsequent appearances for me. It seems to me that once you get comfortable with the idea that HOTs give rise to there being something it's like, AND you're happy severing phenomenal consciousness from state consciousness, then there's no real basis for denying phenomenality to mere first order thoughts.

Anyway, I'm undoubtedly missing something here. Maybe some friendly HOT-head or HH sympathizer will drive by and lay some knowledge on me?


  1. Nicely put, that.

    I'd add that the problem gets still harder for lovers of transitivity. Many HOT folks explain state consciousness in virtue of transitive consciousness, a move that does motivate positing HOTs as the right tools for the right job. Sever phenomenal consciousness from state consciousness -- and a transitivity role -- altogether, and it's not at all clear why HOTs get to be the bearers of phenomenal consciousness.

    And then everything Pete said again.

  2. Thanks, Roblin. You raise a nice point about the role of transitivity here. I wonder, though, for the non-severing HOT theorist, how best to connect transitivity and WIL. Something like this?: A state *of which* we are conscious is a state...
    (a) in virtue of which there's something it's like;
    (b) there is something it's like to think about;
    (c) blah-bitty blah...somthin'-somethin'...what it's like!

  3. Good point. I recollect (or reconstruct) the HOT argument for WIL as a function of HOT content (viz. transitivity) as being something like:

    (1) Changes in WIL could be a function of target states with additional properties (sensory qualities) or HOTs with different content.

    (2) We can provide cases where (arguably) the target states remain the same (two occasions of drinking the same wine, e.g.) yet the WIL differs.

    (3) Hence, HOTs determine WIL.

    (4) BONUS TRANSITIVY-BASED THEORIZING: Changes to the content of HOTs allow one to be conscious of sensory qualities already present in the target.

    RESULT IF YOU ASK ME: "blah-bitty blah...somethin'-*about*-somethin'...bitty...what it's like!"

  4. Ahh, thanks Roblin. That helped knock a few cobwebs loose. That sounds right.

  5. Notes from a HOT-head:

    1. Us HOT folk reject H&~C. We hold that H->C, with the clarification that C is a state the subject is aware of himself as being in. There's no need to pull apart C and WIL if you go this way, though obviously there are other issues!

    2. When you're aware of yourself as being in a state, there's something it's like to be you, for you.

    3. FORs don't make you aware of yourself as being in states (that's not their functional role). Since you need to be aware of yourself as being in a state for their to be something it's like for you, FORs can't account for WIL for you. (Support for 3: nonconscious FORs in masked priming, inattentional blindness, etc.)

    4. Why think any of this? Well, here's one way to put it:

    Start with a folk psychological platitude purportedly explaining the difference between conscious and nonconscious mental states: conscious states are ones that we're aware of being in. This is just a rough starting point for the theory, but ya gotta begin somewhere. This claim isn't irrevisable or anything, but it allows us to start with a (relatively) neutral characterization. There are of course other starting points, but this one isn't bad, and it's got a decent pedigree (Aristotle, Locke, etc.).

    Now form a theory of the underlying mechanisms explaining the folk-psych principle. Rosenthal argues that HOT is the way to go here. Accepting that, we're faced with the empty HOT possibility, for Humean reasons spelled out by Armstrong. What to do? Answer: get as close as you can to folk usage, but in the end, revise. So it's counterintuitive that conscious states are merely states we represent ourselves as being in. That happens sometimes in science.

    And who cares what we call things here, anyway? So long as I have a mechanism explaining what goes on when there's something it's like for subjects (they are representing themselves with HOTs as being in states), then I don't care if we say that this state is "state conscious" and that state is "phenomenally conscious" and this or that state has the (awful!) property of what-it's-like-ness. This is just metaphysical bookkeeping or ordinary language therapy.


  6. And howdy, Roblin! Good to (virtually) see ya!

  7. Hey Josh. Thanks for the helpful points.

    I'm having trouble with your #1, though.

    If, as you say,

    A: "conscious states are ones that we're aware of being in"

    and, as I would have thought you'd agree,

    B: We are sometimes aware of ourselves as being in states that we aren't actually in (false and empty HOT scenarios)


    shouldn't you be *affirming* H&~C ???

    (If I emphasize here that the negation involves an actually existing, not-merely-notional conscious state, does that change things?)

    I'm not getting it. So, sock it to me.