• apokrisis
    6.8k
    Again I am gob-smacked that you simply repeat my own arguments back to me.

    The only difference is that I emphasise the complementary logic involved. The brain has an obvious interest in doing as much at an automatic learnt level as possible, because by doing that, attentional level deliberation is by default reserved to deal with whatever else turns out to be unexpected, novel, or otherwise most significant about some passing moment.

    And I've long been championing functional models - like Grossberg's ART neural nets, or Friston's Bayesian brain - which best make that point.

    To the extent that the brain can make its environment predictable, it doesn't really have to pay attention to it. It already knows what is going on before it happens. The other side of that coin is then that when events start not meeting expectations, the brain knows to flip to the complementary form of analysis - the one we call attentional and deliberative. Rather than the smooth and skilled stereotype response, the higher brain can enter a creative and exploratory mode of thinking, remembering and learning. With the frontal planning areas and working memory engaged, the world can be kept in mind long enough to try stuff until some new understanding appears to have a good fit.

    So it is an approach that accounts for the phenomenology pretty easily. It explains stuff like how we can hit tennis balls or drive cars when attentional processing is much too slow and much too tentative to account for such real world skill.

    I don't know why you would start banging on about pre-conscious habit being "primitive". It should be clear enough that habit equates with accumulated wisdom. Attention and habit are two ends of the one dynamical process of coming to understand the world in useful fashion. So there is no evolutionary sequence here. Both had to arise together because each is formally the other's opposite as a style of processing.

    But again, that is a point that is difficult to understand unless you are a Peircean or systems thinker.

    Reductionists think that complexity builds hierarchically from the ground up. You have primitive unthinking creatures that are a bundle of reflexes. Then evolution keeps adding more intricate processing and suddenly out pops a self aware mind. It is the same computational paradigm that leads people to expect awareness to pop out as part of an information processing sequence that culminates in some final data display.

    But natural philosophy understands hierarchical organisation to be triadic. Everything starts with a symmetry breaking that then progresses in two complementary directions. The two orthogonal poles of organisation that result - the local and the global - can then interact. You have a holistic system which self-organises.

    So when it comes to brains, or simpler nervous systems, you can't talk about which came first - attention or habit. They have to arise together as a way of mutually breaking some vaguer state of uncertainty or indeterminacy.

    A jumping spider has a brain the size of a poppy seed and yet it still has this same contrast between attentional and habitual processing. It can creep around the back of dangerous prey after it has paused long enough to assess the situation.

    Now of course a jumping spider is not conscious like a cat let alone a child. But - if we define consciousness vs unconscious in terms of a functional contrast of processing strategies - then we can say it is also a conscious creature as we can stick electrodes in its head and observe the same fundamental attention vs habit distinction.

    Our intuition that the consciousness of the jumping spider is hardly as good as ours is then also accounted for by the fact it has to pause and consider. It has to pounce rather than smoothly pursue. It is more robotic in that its levels of neural performance are not so integrated in "real time", nor are they integrated in a general fashion over scales of minutes, hours and days.

    So definitely we can see a clear difference in scale - without having to claim an essential difference in kind.

    But if we get down to a worm or a jellyfish, neither attentional-processing nor habit-forming exist except in the most neurally reduced form. You can demonstrate the habituation of reflexes. So there is a bit of "in the moment" learning to go with a bit of genetically-inherited instinct. There is some kind of contrast in adaptive behavioural response - the precursors to attention and habit. Yet also it is getting about as attenuated as we can imagine.

    So I am speaking to a different model of a system - the model of an organism rather than a machine. A system that processes signs rather than information. And that is just a different paradigm with its own developmental and self organising logic.
  • Forgottenticket
    212


    I'm curious, what are your thoughts on the global workspace theory stuff?
    (global neuronal workspace et al.)
  • apokrisis
    6.8k
    I'm curious, what are your thoughts on the global workspace theory stuff?JupiterJess

    I don't think it is wrong so much as just clunky. It is still stuck in essentially a representational/computational paradigm with its flaws.

    But on the other hand - in the mid-1990s - I thought it also clearly the leader in terms of that approach. It got the neural basics right, like the two stage habit~attention distinction, and the contextual, or constraints-based, approach to processing.

    I knew Baars, so we discussed this quite a bit. At the time, the debate for me was about how to reconceive brain function as self-organising dynamics instead of data-processing computation. Both paradigms appeared to have a lot of correctness, yet how could they be married? That was when I got into the emerging bio-semiotic approach in theoretical biology. Semiotics does marry the dynamical and computational views in the one idea of the sign relation.

    So in summary, 20 years ago, the global workspace was a competent summary of the neurocognitive evidence. But the philosophy of mind issue of "what paradigm" was equally clearly not solved by that. It still awaits its semotic revolution. :)
  • charleton
    1.2k
    There is no "calculation'. Calculation implies units and measurement. A dog catching a ball is a pure analogue experience where the motion of two things comes together. There is no measurement and no values. Call it what you want, but it is not calculation.
123456Next
bold
italic
underline
strike
code
quote
ulist
image
url
mention
reveal
youtube
tweet
Add a Comment