Consciousness and the Social Brain: A review

What is consciousness?  What is the inner experience that we all feel we have?  Is it an illusion or an ontological reality?

Michael Graziano thinks it is a reality, and he thinks he has found a scientific theory for it, which he describes in his book: ‘Consciousness and the Social Brain‘.

The quest for a workable theory

Graziano, an expert in neuroscience and psychology at Princeton University, starts off by pointing out that most theories of consciousness have a couple of major flaws.

The first flaw is that they involve a magic step, a mysterious mechanism that can’t be explained.  For example, the common theory that consciousness emerges from complexity can’t explain the transition from non-conscious complexity to conscious complexity.  Consciousness simply arises in some mysterious fashion.

The second flaw is that the theories usually focus on how the brain produces consciousness, but ignore how consciousness produces behavior.  This ignores the only objective thing about consciousness, which is that we can state that we have it.

Graziano’s goal is to provide a theory that explains the mechanism of consciousness without either flaw.  He compares such a theory to natural selection, the core mechanism behind evolution.  Generated from fossil evidence, the idea of evolution existed for several decades before Darwin’s theory of natural selection provided a scientific non-magical mechanism to explain it.  Graziano hopes to identify a similar mechanism for consciousness.

The theory

Graziano starts by describing the phenomenon of attention.  Attention is the process of the brain figuring out which inputs, which signals to focus its resources on.  It is a process of various coalitions of signals competing for those resources.  We attend to something when one of those coalitions “wins”.  At least until the next ascendant coalition takes over.

Attention is an emergent process.  It can be driven by the signals themselves in a bottom up fashion, or it can be biased from top down preferences to focus on certain signals.  One can override the other based on their strength.  For example, maybe you choose, in a top down fashion, to focus your attention on this blog entry, but suddenly a spider crawls on your arm, causing the bottom up spider related signals to override the top down directive.

Some theories stop here and say that this is consciousness, and that perception of anything else is an illusion.  This view equates awareness with attention.  But these theories have a problem.  Often attention and awareness do seem to be the same, but not always.  Your attention can be focused on something without your awareness.

If you doubt this, check out the Apollo Robbins TED talk I linked to the other day.  Also consider things like subliminal messages, and ailments like blindsight or hemispatial neglect.  (Graziano actually discusses hemispatial neglect at length in the book.)

But if awareness is not attention, then what is it?  Here we get to the core of Graziano’s theory.

He sees awareness as being an internally constructed model of attention.  It is information held in a part of the brain about the emergent and messy process of attention.  Like all models, it is a simplification of the real process, and there is much going on in the real phenomenon (attention) that the model (awareness) isn’t privy to, so it is incomplete.

Graziano uses the example of a general plotting strategy with toy soldiers on a table map.  The toy soldiers and map are a model of the real situation involving real troops and battlefields.  The model is a simplification, and an incomplete one, but it is useful enough for the general to make decisions about where to focus his resources.

In the same fashion, awareness is a predictive model of attention that the brain uses to influence attention in the top down fashion mentioned above.  But like the general’s strategy map, awareness, the model, the schema of attention, is a simplified and incomplete version of what is going on.

Compare this to what we know of consciousness.  It is aware of a lot of our perceptions, and a lot of the things that we focus attention on, but not all of them.  It is unaware of a lot of things that are going on in our subconscious.  It is incomplete.  But it is still useful.

Graziano calls his theory of consciousness the attention schema theory.  Awareness is the attention schema.  (For the remainder of this post, the terms will be used interchangeably.)

An important concept to understand about the attention schema is that we don’t just have one for ourselves (self awareness), we also build one for other people or entities that we perceive to be conscious (social awareness).  The schema we have for others is of course not as detailed as the one for ourselves, but both are information.

A quick note.  The attention schema is not the homunculus, the little person in the head observing and controlling actions.  It’s a feedback mechanism that allows the brain to review its own attentional state.  But it may provide a powerful explanation for the intuition that leads many to believe in the homunculus.

Reviewing fMRI and brain damage studies, Graziano also theorizes about the location in the brain where awareness may be located.  He focuses on the superior temporal sulcus and the temporal parietal junction.  He is careful to say that there may be other locations involved, but these two seem to be involved in most awareness processing.

Relation to other theories

Graziano notably discusses his theory’s relationship to two classes of theories: social theories and information integration theories.

Social theories, which begin with the observation that the cognitive machinery involved in understanding and predicting other minds is also the same machinery in understanding ourselves.  In other words, consciousness is our theory of mind turned toward ourselves.  Graziano accepts these theories as valid, but incomplete.  They explain self awareness and social awareness, but not awareness itself, which the attention schema theory is focused on.

Information integration theories, the idea that consciousness arises from the integration of information, are also incomplete.  Graziano notes that while information integration is certainly a requirement for consciousness, it isn’t sufficient to explain it, and that it holds one of the flaws mentioned above, requiring a magic mysterious step.

He points out that advocates of information integration as the sole explanation for consciousness would have to explain why the internet or the US income tax code are not conscious, since both are complicated and highly integrated.  (That is, aside from asserting that they are conscious but have no way to communicate to us that they are, which is an untestable proposition.)

Philosophical ground

Graziano weighs in on free will at the neuroscience level.  Deliberately stepping around issues like determinism, quantum uncertainty, and compatibilism, he restricts his scope to whether or not awareness, consciousness, has any control of our actions.

Many theories assert that consciousness is only an observer, an after the fact rationalizer, not an active controller.  Graziano disagrees with this.  He agrees that major role of consciousness is as an observer, rationalizer, etc, but that it also controls many of our decisions and actions, pointing out that patients with their awareness damaged are also impaired in behavior and decision making.

Graziano also speculates about whether animals have consciousness.  Most animals obviously have attention.  The question is whether or not they have awareness, the internal model of attention.  He thinks most mammals and birds do, although perhaps not at the same level as humans.

He also speculates that aliens could, conceivably, be fully intelligent without having the attention schema model, without being what we would call conscious.

Graziano discusses what a computer would need in order to be conscious: an attention mechanism, a sub-system to model that mechanism, and having the model data available to the components of the overall system.  He points out that many modern supercomputers currently have as much processing power as the human brain but, lacking the structure he lays out, they show no signs of being conscious.  Using the theory, he predicts that a conscious computer could be developed within a few years.

In perhaps the part of the book most vulnerable to misinterpretation, he points out that if awareness, the attention schema, actually is just information, then the inanimate objects, animals, natural forces, or computers that we attribute consciousness to, actually have a certain kind of consciousness, even if it is only located in our own brain in a low resolution fashion.

We attribute awareness to these things in a similar way that we attribute awareness to ourselves or to other people.  We hold an attention schema for them in the same manner that we hold an attention schema for ourselves.  Both are information.  One is not necessarily any more real than the other.

This means that spirits, ghosts, and related concepts exist, as attention schemas.  A version of the consciousness of deceased people do live on in the memories of their living friends and family.

Graziano is careful to clarify that he’s not saying these things exist outside of the mind, only that our spiritual intuitions may have foundations in our mental architecture.

My take

This book has changed my views on consciousness in a number of ways, which I consider the best compliment I can give it.  Going in, I have to admit I leaned heavily toward the social theories he discussed, that consciousness is just our theory of mind turned inward.  I also viewed internal experience as largely an illusion, a cognitive confusion, a subjective bias not confirmed by the facts.

But if Graziano’s theory is correct, and it seems to have a lot going for it, then consciousness exists as an actual attribute that autonomous agents may or may not hold, inner experience is real, and we may have a scientific theory for it.

Graziano is realistic.  His theory is based on solid empirical foundations, but he acknowledges that it will require years of additional scientific work to see if it is substantially the whole story.

I’m not sure yet if I agree with Graziano’s assertion that consciousness has “control” of our actions.  It may be more accurate to say it has information that may influence our actions.  But this could be semantic fussiness on my part in an area rife with disagreements that only amount to semantic fussiness.

On computer consciousness, I tend to suspect that while the human architecture is one path to awareness, to an inner experience, there may be other more streamlined options available for an artificial mind.  All that may be required is a subsystem modeling the whole system, and making that information available for use to the overall system, an internal feedback mechanism.

Whether or not this theory is the natural selection of consciousness, or only a step toward that comprehensive theory, I can’t say.  But it feels like is at least a major step.

Can we say that the attention schema theory solves the hard problem?  I think it depends on your definition of the hard problem.  It provides a possible explanation for experienceness, but its counter-intuitive assertion that awareness is information may invalidate it as a solution for many who are interested in the hard problem.

You may read the details in the book and come to different conclusions.  If you are interested in consciousness, I highly recommend it.

31 thoughts on “Consciousness and the Social Brain: A review

  1. David Deutsch maintains that we could build a conscious AI using today’s processing power if we only knew how. It sounds like Graziano has laid out a high-level blueprint for how this could be done.


    1. Interesting. I’ve seen Deutsch’s TED talks but didn’t know he had weighed in on this. I think they’re right. Consciousness is a matter of programming, not processing power.

      Of course, as a programmer, I can cheerfully say that the devil is in the details. It’s rarely as easy as it looks from a distance.


    2. I know I’m conscious. I don’t know that David Deutsch is conscious and I would not, could not know that a machine claiming to be conscious was conscious. Or rather, I have no grounds upon which to judge such claims. Other people may be conscious, of course, but I must doubt it (I am obliged to doubt it, and in any case, have never seen the slightest evidence for it in what is now a long life – and, reader, neither have you). So far from my awareness of others’ consciousness somehow constituting a reflexively self-constituting awareness of my own, my undoubtable consciousness is the sole grounds for doubting that of others (oddly, unexpectedly, there is a discussion of this regression in Sartre’s ‘Being and Nothingness’, of all places). Graziano has no more succeeded in closing the perpetually diminishing gap between a reductionist description and a viable explanation than John Bickle did, but does not match Bickle’s understanding of the nature of the problem. The author and many of his reviewers might usefully consult A. R. Louch’s ‘Explanation and Human Action’ (1966) as a primer in this special problem in explanation. There are no conceivable scientific, but plenty of conceivable philosophical accounts. By and large, the more strenuous the scientific pursuit, the less rigorous the philosophical technique and the less critical the meta-theoretical basis for the broad ontological assumptions involved – note the proliferating hypostatisations in Graziano’s essay.

      Liked by 1 person

      1. Emm,
        I wonder if you meant that you can’t be certain of the existence of us readers, though not quite that you doubt our existence? If you like I’ll take a Turing test for you. Regardless I’m a strong Cartesian myself (naturalistically of course), so I do hope that you stick around. Apparently Mike just put up a modern Graziano post. I’m not sure if I’ve read this old one though.

        Liked by 1 person

  2. Thanks ‘SAP’, my interest is piqued, and though I think I’ll have to read his ‘God Soul Mind Brain’ first, I can see myself reading:

    ‘Human consciousness and its relationship to social neuroscience: A novel hypothesis’, Graziano/Kastner 2011

    Click to access Cog_Neurosci2011_98.pdf

    … before the week is out.

    On the ‘Memories …’ thread you commented that as you read you were beginning to see consciousness as a tool of the unconscious. Did that thought hold through to the end?


    1. I can definitely see you reading ‘God Soul Mind Brain’ first. It sounds much more in line with your interests.

      The paper you link to looks like a good summary of what he covers in the book. I see many of the same illustrations.

      On consciousness as a tool, I’m not sure. As I said above, it might be a matter of semantics. Awareness is a model of attention which arises through several brain processes. Awareness is information used to make decisions, although it’s not all of the information used in decision making. But knock awareness out, and decision making ability seems to disappear. So, does that amount to “control”? I’m not sure.

      When we decide to “consciously do something”, which part of the brain is deciding to do it “consciously”. The parts involved in generating awareness (TPJ, STS ,etc) or other parts? Do we only mean awareness when we say consciousness or do we include the executive portions? If so, what happens when the executive does things we’re not aware of? (What does “we” even mean in this context.)

      I do feel comfortable saying ‘awareness’ is a tool, a subsystem, used by the brain.


      1. I’ll have to read that paper and come back and reread your post and reply – hopefully that’ll give me a better handle on the terminology.

        Regarding “When we decide to “consciously do something”, …”:

        ‘The mind’s best trick: how we experience conscious will’, Wegner 2003

        Click to access wegner-trick.pdf

        ‘Précis of The illusion of conscious will’, Wegner 2004

        I don’t know, maybe we don’t(I’m just reading the 2nd one now)?


        1. Wow, the 2nd one is vast and diverse. I like the word these papers use: “cause”. Based on Graziano’s information, I feel comfortable saying that awareness causes a lot of our actions, albeit indirectly, by supplying information to the executive functions. Indeed, it seems to be a necessary ingredient for many of those actions. I’m much more comfortable with that word than “controls”.


          1. On ‘Précis of …’, yes, it is – not sure I’ll need to read the book if I can get through it. I’ve been at it for 1 1/2(on&off) and only as far as ‘2. Apparent mental causation (Ch. 3)’ – think I’ll have to continue in the morning 🙂

            I just scanned ahead and found out it actually ends on page 11 and is followed by 20 pages of open peer commentary and the Wegner’s response. This will be interesting!


          2. I admire your fortitude. My brain (or maybe just my awareness) wimpers at the prospect of reading the whole thing. I didn’t mention it in the review, but while Graziano is an excellent and engaging writer, reading that material can tie your mind in knots.

            I’m currently reading ‘Farewell to Reality’ by Jim Baggott, which looks like it will also be mentally taxing. I’m almost certainly going to have to take a break after this and read some fiction for a while.


  3. I am probably twisting this a bit based on my interpretation and not having read the book, but it seems that attention interacts with the subsystem and not with reality directly. The subsystem is a model of reality based upon input from senses and thought, but not reality itself. The things that we perceive to exist, exist in the form we perceive them because we are there to perceive them. In that way, I think that everything has consciousness. I think this is what you are saying above, especially in regard to spirits and the like. Consciousness is localized from an individual perspective, but expands outward into everything that is perceived and imagined.


      1. You’re welcome on the link. It is Zemanta but I think it helps both our blogs. Anyway, I appreciate you stopping by!

        On the sequence, here’s a way to think of it:
        1. Signals come in from our senses.
        2. The signals vie for cognitive resources in a never ending competition. This competition isn’t coordinated. It’s a random messy emergent process.
        3. Some combination of signals win and receive the resources, for a time. This is attention.
        4. A part of the brain models the results of attention (or at least some of them), both for purposes of prediction and influencing it.
        5. The result of 4 is awareness, or consciousness. It is information that is made available to other parts of the brain for decisions and other purposes.

        Hope that helps.


        1. This helps, but with what you are saying in 4, doesn’t this mean that the model also feeds into what signals are given cognitive resources? A signal may not fit based on the experience of the model. Some things could be let in for a short time or flat out denied access to resources. This is done consciously and sub-consciously. To me this says that attention operates inside of the model focusing on the model’s representation of reality.

          Whether the model is another input (or signal) to attention or attention operates only within the model may be moot or one in the same. My point is that attention is directed to what is determined to be important based upon a model of reality and not a result of solely the signals. Thus, we are operating in augmented reality and not direct reality. Or at best only a portion of reality that is let in by this whole process.


          1. I think you’re right. Awareness affects attention. Attention can be biased in a top down manner, and that biasing can only happen from the information that comes from awareness. The key word is ‘biased’ rather than ‘controlled’, since bottom up signals can override that biasing, such as the spider-on-the-arm example I gave in the post.

            I agree that reality as we understand it is a construction in our brain. The best example Graziano gives is the color white. White is actually a dirty conglomeration of every color, but we perceive it as pure luminance. The perception of clean white is a construction of the brain. Of course, the perceptions of color itself is also such a construction. In nature, there are no colors, only varying wavelengths of electromagnetic radiation.


  4. Well, we’ll see how the commentaries go – some may get skimmed or even skipped, but I will finish the article tomorrow likely and the FAQ’s at the end(Wegner’s response).

    ‘Farewell to Reality’? “… a pointed critique of modern theoretical physics.” – the publisher

    Yikes! You’re talking *real* fortitude there 🙂

    I finally finished a repeat skim through ‘Thinking, F&S’ to make some notes and want to do the same with ‘Who’s In Charge?’. After that I’ve got 4 waiting but ‘Religion, Brain & Behavior’ extended their open access through February so the books can wait(I have peeked at the intro to Haidt’s a couple of times).

    I’m going to look at a bit more of ‘Philosophy in the Flesh’ too, at least until it gets too philosophical for me 🙂


  5. Awareness (any description) is the subsystem that interfaces mind to sensory data… stop
    It functions as a subsystem of the simulation of the world around us by guiding input to the simulation. When the simulation has no input, there is no consciousness/decision making.
    Intelligence without changing input is a paperweight. Even if you hold all the sensory data to relative stasis, the intelligence turns itself inside out… like a wild animal caged in a small zoo.


Your thoughts?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.