Wednesday 4 July 2012

Ezequiel Morsella: The Primary Function of Consciousness in the Brain


Abstract: Despite the challenges in unraveling how the nervous system gives rise to consciousness, a consensus has been growing that (a) consciousness is associated with only a subset of all nervous regions and processes, and (b) the primary function of consciousness is to integrate processes/information that would otherwise be independent (the integration consensus).  Recent research illuminates the subset of areas and processes that are most closely related to conscious processing.  These investigations reveal that consciousness serves to integrate only certain kinds of information/processes.  Many forms of integration can occur unconsciously.  The peculiar form of integration associated with consciousness involves a form of information broadcasting that is intimately related to what is casually referred to as 'voluntary' action and to the skeletal muscle output system.  All these developments are synthesized in Supramodular Interaction Theory (SIT).  During this lecture, I will review evidence for the integration consensus, SIT, and other notable contemporary reductionistic approaches.

    Morsella, E. (2005). The function of phenomenal states: Supramodular interaction theory. Psychological Review, 112, 1000-1021. LINK:http://bss.sfsu.edu/emorsella/images/MorsellaPsychRev.pdf

Comments invited

44 comments:

  1. I find Morsella's theory very interesting concerning actions/senses etc. But how can this conflict and integration theory be applied to our consciousness of "internal states" like beliefs, understanding, ideas, etc?

    ReplyDelete
    Replies
    1. The example E. Morsella just proposed, someone's urge taking its origin from the unconscious sight of a burden cigarette moments prior to the urge, is responding to your question. I imagine that two thalamocortical loops controlling for two semi-independent certain unconscious processes would just become aware to the individual once these get into conflict, just like it would be in a Stroop task. It seems to be a bit synthetic, but with a little of effort I can imagine it.

      Delete
    2. My first reaction would be to point out that these things are all related to language. Well, I'm not clear on whether understanding (depending on one understands it... ) is a matter of language, but things like ideas, beliefs and so on are, I think strongly related to the human faculty of language. What Morsella describes is a very basic kind of consciousness, one that certainly does not require linguistic capabilities. So now, there's two options: either you try to see how language can be superimposed on top of that basic consciousness and how it modifies or extends it, or you argue that conscious states must include these "internal" states.

      Delete
    3. Read in Baars "The biological cost of consciousness" (2012): What about the fact that conscious waking happened hundreds of millions of years ago, way before the appearance of language (only 100,000 years ago)? Would that mean that prior to the development of language, beings weren't capable of understanding? I don't think so.

      Delete
    4. I thought you made a very good point Marjorie. My first thought was: If we are conscious when two processes are in conflict, maybe it works as well for 'internal states'. (If you look at your mind as a pool of thoughts/ideas/... all in conflict to reach consciousness)
      Laurie-anne Dion

      Delete
    5. Obviously not. However, there's a difference between leanring how to walk (understanding what I need to do to walk) and understanding what Kant meant to say in the Critique of Pure Reason (which I might not actually understand, but that's besides the point). I'd have to read Baars' paper to be clear on what he means by "conscious walking", but in the case of learning motor skills, you obviously don't need language. If by understanding, one means the grasping of complex reasoning, then yes, I think you need language.

      Delete
    6. I wrote "conscious waking", not "conscious walking".
      But yes, of course I wouldn't even be thinking about attributing to tetrapods the capacity of understanding Kant's reasoning. ;)

      Delete
    7. Oh my, I'm terribly sorry. Should have read more carefully. Still, my point stands: I don't mean to say that understanding necessarily requires language, but that internal, reflective understanding does require it, along with beliefs (as propositional attitudes) and ideas.

      Delete
    8. I don't think language is required. To me it sounds a bit like the visuocentrism thing. We think with language, because it has become such an important modality for humans, but don't we sometimes think in terms of images... In regard to beliefs, history is filled of example of people following blindly symbols... A second point, aren't beliefs (value system) responsible for a large part of internal conflicts?

      Delete
    9. Laurie-Ann suggests that two thoughts could be in conflict, but in the examples that Morsella gave, the conflicts were arising from systems that had different types of control, such as a "reflexive" behavior (like breathing) conflicting with contextual information (like presence of poisonous gas). Any suggestions on how ideas/understanding could arise from different systems like this?

      I imagine we could have "reflexive" internal states (maybe anger?), that we have to integrate with knowledge to suppress an action in an inappropriate context.

      Delete
    10. These are all excellent points. In accordance with the work of Jeffrey Gray and Panksepp (and many others, including psycholinguists), one may propose that there is good evidence that consciousness does not require language, though, as you mention, language does at times influence the contents of consciousness. This is an excellent point that I will ponder further, especially when working on EEG experiments that involve 'subvocalization' (i.e., "talking in one's head"). Thank you so much.

      Delete
  2. INTEGRATION AND INTEGRATION

    There's no question that there are some things of which we are conscious, others not; some things we do (or can do) consciously, others not.

    Perhaps the things that are conscious involve certain special kinds of integration that the unconscious things do not.

    But why should those special integrations be conscious (felt) -- rather than just integrations? To say that they are conscious integrations rather than just integrations because conscious integrations are faster or stronger is not yet to explain how and why the fact that they are conscious makes them faster or stronger.

    After all, on the face of it, integration is just integration!

    ReplyDelete
    Replies
    1. I don't think that Morsella's theory only pointed out faster/stronger integrations but more importantly widespread networks involving different brain regions and loops. He suggests that one of the these region might be the frontal cortex (one hypothesis among many others because we don't know the exact cerebral region of consciousness). Maybe this is why and how we can understand non-automatic action/behavior/reaction, etc since the frontal lobe is known to be associated with more higher functions, for example the capacity to understand consequences of an action, introspection, social behaviors, emotion-related memories, etc. For my point of view, consciousness is a mental function just as executive functions or memory are, and that only some species have developed it through evolution based on adaptative needs (human, non human primates, maybe not aplysia !) .. but this come from an experimental neuropsychology student !
      Claudia Polevoy - UQAM

      Delete
    2. Dear Prof. Harnad, this is an excellent point. The approach simply identifies integrations that reliably occur consciously and integrations that can occur unconsciously, with no claims about conscious integration being stronger or faster than its unconscious counterpart, as commented above by Ms. Claudia Polevoy.
      The approach is "descriptive," simply pointing out, among other things, the integrations that can occur unconsciously. Conscious integrations seem to be associated with, not "afference binding" (e.g., intra- and inter-sensory binding) or efference binding (i.e., stimulus-response binding), but "integrated behavior" through "efference-efference binding," which is peculiar to the skeletal muscle output system. Yours is an excellent point: Why subjectivity is part of this process remains a mystery from this standpoint, unfortunately.

      Delete
    3. Regarding neural processing, one of the main points of the talk was that there is no consensus regarding the neuroanatomical correlates of consciousness, unfortunately. There is strong evidence for each of the many contemporary 'neuroanatomical hypotheses.' Your comment about frontal cortex is an excellent point to keep in mind.

      Delete
    4. And back to the question WHY (which is constantly and pertinently pointed out by Pr. Harnad), even a consensus regarding the neuroanatomical correlates of conscioussness wouldn't solve it. Even if substrats are not causes, "we will never have access at the origins, so we can only focus on charaterizing conditions that allowed language[conscioussness/feelings], without absolute causality or determinism" (this sentence is from Rastier, 2006, "De l’origine du langage à l’émergence du milieu sémiotique" where he was talking about the relevance of working on the origins of language).

      Delete
    5. IMHO, what is baffling about the problem of how consciousness emerges from information integration is less the problem of subjectivity—theories like Baars’ and especially Merker’s could account for the subjective aspect of experience—but rather the problem of determining why any of the integrated information generates a phenomenological experience. Indeed, models like Merker’s give us mathematical explanations for why experience should be constructed as centered on a subjective, first-person vantage point. This has to do with information integration, and is arguably an “easy” problem: how should one generate, from the data available to our sense organs, a 1st-person vantage point?

      The challenge Doctor Harnad is proposing is to determine at what point, in the integration of information flows, something like felt experience begins to emerge. If consciousness is a feature of certain types of information processing, then indeed such a point will be determinable, qua feature of information flow and integration. However, if consciousness is a more basic characteristic of biological systems, such an approach might be misleading…

      Delete
    6. Disagreements between scientists regarding “what is the location of consciousness” in the brain sounds to me like a like a man asking: “where is the Internet on earth?” Many different people can go in many different countries and report that “they have found Internet there, therefore, country X is the source of Internet”. This disagreement in localization originates from a misunderstanding of what consciousness, or Internet, is. Dr Morsella’s analogy to a global broadcasting system certainly helps resolving this issue.

      Delete
  3. What about cognitive conflicts? Sloman's dual process theory explain that there is two system: and A and B, called heuristic and analytic systems by Evans. These two systems differ by the way they work. For those who are familiar with this theory, could we make a comparison of the heuristic/analytic system and the smooth/skeletal muscle systems? The problem I can find about this is that to "experience" conflicts in the muscular system it has to be two skeletal muscle system that are in interaction and it can't be a smooth and skeletal muscle. Or there would be no conflit (maybe I misunderstood this part). But in the dual process theory, the heuristic system (wich would be the equivalent of smooth muscle system) and the analytic system are sometimes in conflict and the analytic have to override the heuristic system.

    ReplyDelete
  4. I wouldn't know for sure, but if you take the Stroop task where people are most likely to read the word written than to tell in what colour it's printed, the conflict might be resolved according to how previous experiences (your life experience) consolidated common sense: you're most likely to need to read a word than to tell the colour. The reading module is probably stronger than the telling-the-colour one. Would that be high-level decision? I don't know; the decision to read the word and not the colour is unconscious and out of our voluntary action. If we were really pushing our higher-level capacities unto the task, we wouldn't be sometimes mistaking ourselves in reading the words.

    ReplyDelete
  5. CONSCIOUSNESS AS A MEDIATOR OF CONFLICT: There is something in conflict that may be key to understanding the function of consciousness. There is conflict happening constantly in behavioural plans. Each subsystem of the brain is pushing to do something, but only one behavior comes out. Every time 2 or more planned behaviors are in conflict, there needs to be a resolution. If each of these systems wants to push out its own behavior and there is no universal rule to resolve conflicts, then each conflict becomes a separate battle of subsystems. There is where consciousness may have come into play as a mediator. It receives influences from all those subsystems at a time, and integrates these influences, giving each a particular weight, to come up with a unique an integrated behavioural output.

    ReplyDelete
  6. Je me demandais comment la théorie de Morsella s'applique aux animaux, notamment les animaux avec des systèmes nerveux très simples. En effet, si la conscience sert principalement à la résolution de conflits, il semblerait que tous les organismes qui ne font pas face à de tels conflits ne pourront être considérés comme conscients.

    ReplyDelete
    Replies
    1. Si j'ai bien compris, son but est de trouver la forme la plus fondamentale de "conscience". Sous cette interprétation, les organismes qui ne font pas face à de tels conflits ne sont pas conscients, parce qu'ils n'ont pas besoin de la conscience. Ce qui serait intéressant serait de regarder qu'est-ce qui fait qu'un organisme fait face à de tels conflits (par exemple, y a-t-il un type de problème adaptatif particulier à régler). Ça permettrait peut-être de s'approche d'une explication écologique de l'émergence de la conscience. Enfin, peut-être.

      Delete
    2. Mais je ne crois pas que E. Morsella ait dit que la conscience serve principalement à la résolution de conflits. Il a dit que dans le cas d'un "synchrony blindness", on aurait seulement conscience que deux processus divergents aient eu cours au moment où ceux-ci rentrent en conflit l'un avec l'autre. Je ne crois pas qu'il ait nécessairement dit qu'un processus non conflictuel soit moins "conscient" qu'un qui implique un conflit.
      Ce qui est intéressant, je trouve, avec le "synchrony blindness" est que même s'il y a différents processus survenant en même temps, donc couvrant une plus large région d'activation cérébrale, cela n'amène pas nécessairement une conscience des processus en cours.

      Delete
    3. Je pense qu'il a dit assez clairement dans la période de questions que la présence de conflit relié aux 'skeletal muscles' n'est pas à une condition nécessaire pour être conscient à un moment donné (cependant, elle est peut-être suffisante). On peut être conscient même lorsqu'on écoute une conférence par exemple, sans ressentir aucun conflit. Donc, de ce point de vue, si la présence de conflit n'est pas nécessaire, les animaux pourraient (en théorie) être conscients sans jamais avoir à résoudre de conflits.

      Delete
  7. Right. Couldn't (unconscious) attention modulate Stroop-like conflicts?

    Less important but maybe interesting for Morsella's theory: Amir Raz showed that the hypnotic suggestion to experience the words of the Stroop Task as meaningless squiggles overrides the Stroop effect, i.e., eliminates the conflict. So here the experience of the words modulates whether or not conflict is perceived at all, which seems to decouple conflict resolution and consciousness.

    ReplyDelete
  8. An issue i thought about this: at what point does the consciousness experience start.. In the moment the conflict is detected, or when it is subsequently handled? I think he answered before that the post-conflict can be handled unconsciously again. However, then is it more than drawing attention to the conflict? Is consciousness than equal to attention?

    ReplyDelete
    Replies
    1. I was also wondering this! Also, how exactly do we differentiate/relate/define the terms, aware/awareness, access and attention in relation to consciousness?

      Does being aware of something make us feel it? accessing certain information make us feel? ‘paying attention’ to something bring it into consciousness? Can’t we feel what is like to be a human, or at least what it is like to be ourselves, without any outside input to pay attention to? Is ‘internal input’ or I guess thoughts just as salient/relevant in terms of consciousness as external input?

      Izabo Deschênes

      Delete
    2. Izabo: I cannot think of a situation where there is no "outside input" to pay attention to, as long as we are in a body.

      Delete
    3. Martha: what about dreams?

      Izabo: I agree, the field of consciousness would greatly benefit from agreement on how to define the awareness/attention/consciousness/feeling. Although I disagree with Pr. Harnad when it comes to doing away with 'weasel words', I do think feeling=consciousness is a good place to start at. Awareness I would characterize as a passive state elicited and modulated by stimulus/input intensity. I would say that Attention is different from Awareness in that it is an active state that acts on what is available through awareness. Thoughts?

      Delete
  9. The fact that hypnotic suggestions can resolve the conflict through diminished awareness to the words is very interesting!
    Thank you!

    ReplyDelete
  10. From what I understand from Professor Morsella's talk, the function of consciousness would be to resolve conflicts while integrating information from parallel processes involving skeletal muscles.

    First, I must say I find this view particularly interesting.

    Then consider this: why is it that the only stage of sleep during which we have the impression of being conscious is REM sleep (where dreams occur), which happens to also be the stage of sleep during which the skeletal muscles have the least (if at all...) activity?

    ReplyDelete
    Replies
    1. Yea, interesting observation... along these lines, in REM one often looses the ability to ground your dream-self and dream-experiences into your physical body - we often see ourselves flying, see our minds hovering above our physical body - or have an inability to translate our dream-thoughts into bodily action. We need a constant stream of proprioceptive feedback to create accurate body-maps, so perhaps the lack of motor efferent/afferent activity during REM party explains why REM feels so qualitatively different from just a day-dream, say.

      On the topic of atypical states (not that there's anything really atypical about REM...), what about phantom-limb sensations? What about individuals paralyzed from the neck downwards?

      Delete
    2. This is an excellent point. I should first mention that the theory pertains only to normal waking behavior, which is already complicated enough at this stage of scientific understanding :) Yet, your astute observation leads one to consider that both Sechenov and William James believed that conscious thoughts are intimately related to action. (Sechenov proposed that conscious thoughts are inhibited actions, which is an intriguing notion.) With these ideas in mind, one has to remember that, during rapid eye movement (REM) sleep, while one is having conscious dream content all the muscles are in a state of "sleep paralysis," except, that is, for the eye muscles. (I should mention that I am not a sleep expert but that this is what I have learned during the years.) It has been
      documented that, when such paralysis fails, animals (e.g., cats) act out their dreams, which is of course potentially dangerous for them. Although the theoretical approach presented during my talk pertains only to normal
      waking behavior, these observations associated with sleep paralysis (and the lack thereof) reveal the intimate link between conscious thought and skeletomotor action,
      consistent with some of the ideas of James, Sechenov, and the approach I presented. Again, you raise an excellent point.

      Delete
  11. Drawing from literature on unconscious goal pursuit, I found that the fact that the conflict between an unconscious and a conscious action brings the unconscious action to awareness, contrasts with the efficiency observed in unconscious goal pursuit when the unconscious goal-oriented actions are conflicted with temptations or important goals. Less self-regulatory energy is exerted and less conflict is felt if the gosl has become so automatic that it is unconscious, when confronted to other goals or temptations, to the point of the decision to keep on pursuing the unconscious goal, and not the others, is automatic, and un-noticed. Actually, i am not sure if one is unaware of the decision process or is on deliberated consciously but without much effort. I will have to review Morsella and Bargh's articles or chapters to find my answers...


    Pier-Eric Chamberland, UQTR

    ReplyDelete
  12. And... What happens when you realize you cannot do something that you are used to do unconsciously on that specific situation? The conflict to do vs refrain from doing the action, bringing the conflict to attention in an over-actived fashion? A sort of Zeigarnik effect of actions or goals? Let's remind ourselves that we crave to drop or put on the ground the porcelain full of hot soup, until we are able to do so on the table. Then we stop thinking about it...

    ReplyDelete
  13. Ezequiel, how is this different from global workspace theory? - B

    ReplyDelete
    Replies
    1. The framework presented during my talk builds on the pioneering Global Worskpace Theory (GWT) only in a few ways, including specifying which integrations in the brain seem to require conscious states and which integrations appear to not require such states. The approach, based
      on introspective data and synthesized in Supramodular Interaction Theory, begins to illuminate why conflicts involving perceptual processing (e.g., McGurk and ventriloquist effects) and smooth muscle conflicts are seldom conscious, but why certain conflicts (e.g., holding one's breath or Stroop conflict) reliably involve conscious processing. The approach also explains why skeletal muscle has been referred to as "voluntary" muscle. In short, as mentioned in my reply above, in which I mention James and Sechenov, consciousness constrains what has historically been referred to as "voluntary action." This form of
      action is a property of the skeletal muscle system. Thus, the approach builds on the classic and groundbreaking GWT.
      ______________________________

      Delete
  14. As it has been pointed out a couple of times since the beginning of this summer school, despite some scepticism on a real adaptive function of consciousness, consciousness has to have an adaptive function, otherwise it wouldn't have remained through evolution.

    Ezequiel Morsella brought for the first time in this summer school the fundamental idea that any organism with conscious processes also have unconscious processes. And maybe the key component to explain why consciousness appeared and remained through evolution relies on the fact that we have contents that will be always processed by uncouscious processes and will never reach the level of consciousnes whereas others do. So what would be the adaptive function of separating data into conscious versus unconcious contents?

    ReplyDelete
  15. in my point of view it's all manner of learning. Here's a brief explanation of what i mean by this statement:
    every day our brain is computing and guiding unconsciously a lot of different tasks in the same time but we are not aware of them. the brain can perform an amazing multitasking and mutimodal works unconsciously.but counsiously we can barely perform three task in the same time. What we are doing every day is constant learning or attempt to learn. what i mean by learning is recalling or transforming conscious processies to unconscious process.
    the best example is a baby who attend to control his body to grasb an object. this baby is extremly feeling his body movement and tring different combination of body controle to reach the target.
    so then every thing we have ben faced at since birth and we have learned from it, should be first felt (conscious) before to be send to unconscious processies. that why we no longer need to be conscious of the grasping target processies for example. because it became unconscious.

    ReplyDelete
    Replies
    1. Learning and integrating are probably the keypart of conscious behavior. However, you have to be careful with bodily analogies such as a baby struggling to grab hold of his body.

      A common mistake is to think that babies actually build their motor skills out of nowhere. It looks a lot more like inhibiting neuronal motor commands than building. As illustrated by Parkinson's decease or even drug-related ticks and unvoluntary facial expressions, bodily control isn't as much a building as it is a constraint over the unrelanting neurological surges that go through the brain.

      Think of the thoughts that go through your mind as you drift asleep: they go from anything to anywhere, surfing on unreliable causal and affective links, and they do not make much sense. If you come back to yourself without falling asleep, you'll probably remember the last thing you thought about without being able remember neither the thought before that or which path lead you there.

      All in all, consciousness has got to be part of a restrictive process, similar to focus or attention.

      Delete
  16. FROM EZEQUIEL MORSELLA:

    Thank you all for the wonderful comments presented here, presented after my talk, and presented throughout the days of this singular symposium, which has been an incredible learning experience for me. Each of these comments provides substantial food for thought.

    If there is one point to remember from my talk it would be that not all integrations in the brain require consciousness and that integrations that require consciousness seem to have certain
    properties. I explain this more thoroughly in my article in Psychological Review, available at the following link. http://bss.sfsu.edu/emorsella/publications.html

    Thank you again for all the feedback.
    Best regards,
    Ezequiel Morsella

    ReplyDelete
  17. I wonder how much of the integration is done post-hoc to the doing and how this relates to retrospective judgements of agency...

    ReplyDelete
  18. Thousands of years old mystery of consciousness solved http://www.consciousnessexplained.org

    ReplyDelete