Sunday 8 July 2012

Malcolm MacIver: Sensory and Motor Spaces and the Emergence of Multiple Futures


      Abstract: Back in our watery days as fish, we lived in a medium that was inherently unfriendly to seeing things very far away. The technical way this is measured is the "attenuation length" of light through the medium. After light travels the attenuation length through a medium, about 63% of the light is blocked. The attenuation length of light in water is on the order of tens of meters. For a beast of a meter or two in length, which moves at a rate of about a body length or two per second, that's a pretty short horizon of time and space. In just a few seconds, you'll reach the edge of where you were able to see. If you're down in the depths at all, or in less clear water, you may reach the edge of your perceptual horizon in about a second. In this talk, I'll explore the quantification of sensory and motor spaces, developed through our work on the weakly electric fish, a popular model system of sensory neurobiology. I discuss the relationship between behavioral control and the relative size of these spaces. Finally, I'll discuss whether emergence on to land, where the attenuation length of light is essentially infinite, may have been a key step in producing favorable conditions for the evolution of the ability to plan over multiple possible futures.

      Neuroethology From Morphological Computation to Planning
      Omnidirectional Sensory and Motor Volumes in Electric Fish

Comments invited

19 comments:

  1. THERE IS NO "ACCESS" CONSCIOUSNESS

    There is only one consciousness: feeling.

    "Access" refers to whether or not information (data) is accessed.

    If the accessed information is felt, it's conscious.

    If the accessed information unfelt, it's not conscious (hence no kind of consciousness)

    If the information is not accessed, it's not accessed.

    ReplyDelete
    Replies
    1. Thanks for this clarification. For me it was not clear what access consciousness should really mean.

      Delete
    2. So in this scheme, would there be feeling, and then different feelings such as being in the world (like a fish in the water), and the feeling of future choices enabled by long range high resolution imaging (this would be two varieties or contents of consciousness)? That does seem cleaner, and clearer.

      Delete
    3. There are 2 different questions: (1) How/why do organisms feel anything at all and (2) how/why do they feel this rather than that. The hard problem is (1). If/when (1) is answered, (2) becomes another of the "easy" problems: correlations between brain states and felt states (Dan Dennett's "heterophenomenology"). But without an answer to (1), (2) is just weather-forecasting and hermeneutics!

      Delete
  2. I like the idea that jump on land had these consquences in terms of the sensorimotor capabilities. I am wondering if that fish "just" uses that electric sense to detect/sense the prey or put it in a different way: how much do lower animals actually perform multisensory integration / processing. This is related also to The first comment from Matt i think: could that jump to land living also have promoted the integration of vision and audition, so to perform a more multisensory representation of space? Do fish or other lower animals have multisensory regions?

    ReplyDelete
    Replies
    1. No, multisensory integration is a big deal in these brains. Colleagues have shown neurons that multisensorily fuse different varieties of electrosense, electrosense and vision, and electrosense and lateral line. Likewise in other fish brains.

      However, it is still the case that changing the properties of any of the modalities will have its neuronal repercussions, and big changes happen with big changes in lifestyle - such as from water to land, or from land to air, or water to air. This will result in different requirements for multisensory integration.

      Delete
  3. Electric fish are a great example of the embodiment problem that Edelman mentioned in one discussion. It is difficult to imagine consciousness is like through a sensory system with which we have no experience.

    ReplyDelete
    Replies
    1. This one of the reasons people were initially attracted to the system. The thought was that with no native sense that maps on to what electrosense is like, we'd be less prone to make mistakes of the naive introspection variety when it comes to reasoning about sensory processing.

      Delete
  4. Thanks to Juliette for her comment at the end of Malcom's talk. That was quite a relevant and actually a very fundamental one : we behave a certain way because it is adaptive, but surely we do not behave one way or another because we know it is adaptive.
    Understanding why we behave one way or another relies on the ultimate level of explanation and this level cannot be neglected if we want a complete understanding of it. That's why, as least as much as the other levels of explanation (physiological, ontogenetic and phylogenetic), the ultimate bases of behavior is more than worth studying.

    ReplyDelete
    Replies
    1. Is Juliette the one who thanked me for raising the limits of consciousness point? If so, then I'm not completely sure I'm getting your idea. My take is that the SV:MV arrangement we have is very well suited to our ancestral condition and needs. We've greatly augmented our SV through symbolic exchange like TV etc. But the problem is now the problems are needing vastly larger SVs, in some sense. Symbol exchange helps, but where it really starts to have traction on our core behavioral reactivity is when it "puts us there" such as with feeback systems. This is why people find twitter sometimes irksome (I'm at the party you were not invited to right now) while they don't care so much when they've heard about it at some later time -- e.g. http://nyti.ms/Ngtnrc.

      Delete
  5. Thanks for an amazing talk - I have tried (unsuccessfully) to mentally marry the ideas of Neuroscience and the Environment for several years, and you've given us a beautiful and eloquent framework from which to think about these links, on so many levels...
    On another note - I was wondering if Dr. MacIver has any insight into what cortical mapping of the external sensory world might look like in these electric fish - what do the receptive fields of sensory neurons look like? Fascinating stuff.

    ReplyDelete
    Replies
    1. Thanks. Receptive field mapping in these animals in the first nucleus they go to show conventional center surround structure, both excitatory and inhibitory styles. Further in you find multiple topographic maps, with enhanced area for higher sensor densities. No surprises. What is interesting to think about, to put in terms easily parsed by our visually dominated awareness, is how it would feel to see in all directions around the body at once.

      Delete
  6. Great talk! I just have one quibble. One implicit suggestion in Dr. MacIver’s presentation is that all the species that lived in water before the Tiktaalik were always directly reacting to sensory inputs. Their motor neurons were, in some sense, directly wired to their sensory inputs. Once this connection was severed, land animals developed highly integrated cognitive systems that allowed them to anticipate and plan their actions. I wonder, however, if walking and seeing on land really were necessary conditions in order to develop such systems. Think of Dr. Mather’s evidence that octopi create some kind of mental maps of their surrounding in order to go back to their resting place using different paths. They certainly aren’t simply reacting to their sensory inputs. They might even have some kind of an anticipation system, as she plausibly suggested. So, even though seeing on land might have helped us to develop our cognitive systems, it is not out of question that some kind of cognitive evolution had begun before the first land animals.

    ReplyDelete
  7. I find thrilling the prospect that one could examine the evolution of perceptual spaces. The transition from “collision mode” to “reactive” and “deliberate” modes of perceptual understanding seems, IMHO, to be an excellent way to address the question of how we developed such intricate decision-making processes. The idea that access consciousness is the operation of plan-executing mechanism, inherently related to an organism’s perceptual and motor spaces, is particularly appealing to me. Thanks to Doctor MacIver for the interesting talk.

    ReplyDelete
  8. It seems to me that animals that would be beneath the "deliberate" mode wouldn't be able to have a ToM. Tentatively, I think we could say that once the deliberative mode was possible, since you can mentally rehearse and weigh different possibilities, you have the capacity to do so with other animals and/or moving objects, hence giving them an intentional capacity.
    And I want to thank you Dr MacIver for your talk (in the top 5 of the school IMHO) and relevant interventions throughout the summer school.

    ReplyDelete
  9. I also was very impressed by Dr. MacIver's talk and the very elegant hypotheses he put forth. One thought that his ideas and findings generated: as any proponent of monism would agree, for 'feeling' to happen - like anything else - some finite amount of time is required. Could it be that organisms in 'collision mode' or 'reactive' mode literally have no time to 'feel'? For these organisms, fight or flight responses have to be packed into a relatively small window of time. In contrast, survival instinct for organisms in 'deliberate' modes of perceptual understanding can be fit into a much larger time span, which allows for other 'things' to happen (very crudely put). This needs to be hashed out, but implications would carry to p-consciousness too (to use terms emplyed by Dr. McIver...).

    ReplyDelete
  10. I still am not able to point out exactly why but the point about a "deliberative" vs "reactive" or "collision" being created by the information the senses receive blew my mind.

    As far as making a list for conditions that require consciousness like Eva Jablonka suggested happens, this seems like it should definitely be on there.

    Also: if Dolphins have as much of a range with echolocation in the water as we do with vision and that they're also rather social... what does that mean for the potential of higher order cognition etc. ?

    ReplyDelete
  11. conscience is reaction to stimuli in a symbiotic environment no?

    ReplyDelete
    Replies
    1. No, reaction is reaction, as in robots. And "symbiotic" is (presumably) interaction. All of that is possible without consciousness. And that's to problem (in explaining what causal role consciousness plays).

      Delete