About Me

My photo
I am a former middle and high school science teacher pursuing a doctorate in Science Ed. at George Mason University, with a concentration in cognitive science and the evolution of cognition and learning. Postings on this blog represent my own views, not those of my employer or school. All writing displayed on this page is original work unless otherwise noted, and thus copyrighted.

08 December 2009

I was starting to feel a bit stupid...

Donald Hoffman provies what he regards as a strong criticism to Dehaene on the Edge.org site here. The initial attempts at refutation revolve around the idea of quantum computation, wherein Hoffman states that quantum computation (citing a study in Nature in 2007) is responsible for photosynthesis in plants and green sulfur bacteria, and therefore could be coopted by natural selection for other purposes, such as consciousness. Well, that's a nice thought, but doesn't that bold an assertion require some sort of proof? Perhaps beyond the contributions of evolutionary game theory to thought experiments, which is what is proposed by Hoffman? If Hoffman wants to claim that the brain is a quantum computer, regardless of objections made by Dehaene about quantum computers not functioning at the temperatures needed for organisms to maintain chemical reactions (i.e. life), then Hoffman, or someone who does more than theoretical masturbation, might want to examine the human (or some other supposedly conscious creature's) brain for the wave functions and transmissions thereof that are the hallmark of quantum computing. That is, after all, what those researchers who published to Nature did.
No, the portion that made me feel stupid for a moment was Hoffman's referral back to his 2006 paper on "The Scrambling Theorem" (Science Direct Link). This beast reads like a post-modern masterpiece, and with good reason despite the claims of empiricism: He has nothing on which to actually base his argument. I don't mean he doesn't have much evidence, I mean that he literally hasn't a shred. Hoffman merely states the following: to disprove reductive functionalism in consciousness (the idea that one can reduce a cognitive process to the mathematical function, however complex, that explains it. In effect, that cognitive processes are mathematical functions carried out by neurons), one simply needs to be able to imagine that consciousness can work in another manner, thus making it logically possible (anyone see the flaw in that one?). He then goes on to say how, if he can logically and mathematically demonstrate that people can have different colour experiences from the same stimuli, yet perform the same on colour tasks, then he has effectively disproved reductive functionalism. Beautiful... and here's where I thought I was lost. Hoffman then jumps directly into a model that supposedly backs his claim, but fails to demonstrate how any of the pieces of his model are tied to reality, and that's purely because, as he admits later, they aren't empirically tied to anything at all. Translation from academia to English: Hoffman made this up, out of his own imagination, which brings us right back to the argument he claims he wasn't using. Sorry Donald, but just because I can dream up another way that might coffee might brew or my herb garden might grow doesn't mean that the manner in which we understand those processes is incorrect because of it. I can fully imagine Hoffman's paper being written by a 16 year old with a thesaurus and a horrendous grasp of elementary logic, but that doesn't make it so, althought it does, to some degree, resemble the product.
The invocation of probability at the end doesn't make this any better; If we want to show that Jack sees orange when Jill sees blue, we need to measure the signals coming from both of their optic nerves upon viewing identical wavelengths of light, and then test how each individual responds to the stimuli. ("what colour do you see?" "orange"). Hoffman's general point is that we can't tell, even from that sort of experiment, that Jack and Jill are in fact experiencing the same stimuli, and I suppose that's correct. It is, however, certainly not correct that it then precludes the conscious experience of either of them from being the same as the function used to represent that experience, it simply means that we've somehow substituted, via the internal system of the mind, a different value in the function for each of them. Hoffman, then, has demonstrated that we can't use the same function for each person, and that's about as useful as Hume demonstrating that we, logically, can't know anything. It isn't, at all.

* Whenever someone chooses to use an analogy on the internet, it is, shortly afterward, demonstrated just how poorly constructed that analogy actually is. If that doesn't occur in this case, I'll put it down to a lack of readership.*


  1. I would reply, regarding Hume, that he's arguing more selectively: we can't inductively know anything at all, because it's not possible to support induction logically without resorting to induction: "Why do you think the sun will rise tomorrow?" "Because it's always happened that way." "So why do you think predicting the future based upon the past is useful?" "Because in the past, that has been a successful way of predicting the future."

    And yet, of course, the empiricists acted upon conviction as did anyone else: their objection, rally, was that at some point, we have to base epistemic probability on something other than logic.

  2. Thanks for clearing that up... since my examination of Hume has either been secondhand or cursory, I was wondering if I'd missed something... he seemed more reasonable than that from what I could find.