Deflating Consciousness

I wrote this a while back in response to Alex Rosenberg's the Atheist Guide to Reality. I have let most of it stand. The main idea is one that has been grappled with for a long time about mechanizing the world, especially that of human activity. We can see these worries from earlier thinkers, but such thoughts became more complicated in the 20^th century as we began creating machines that could do complex intellectual activities better than conscious human beings could do these things. The argument that brain/mind processing is mechanistic all-the-way-down thus became more plausible. In addition to creating semi-intelligent machines, in general our ideas about what it is happening within brain/minds, including with what we call conscious processing, was improving.

The point of this argument is really to deflate what consciousness is, and to deflate other ideas that are usually tied to the power of human consciousness, such as reasoning, creativity, free will, and moral decisions. In this argument I assume a given conscious property of humans, but I take a more non-existent stance on consciousness, or claim that it is better spoken of in representational and information processing terms. That is, “consciousness” is a representational state, or some mix of world/self modeling, perceptual information and emotion/feeling/bodily structures. For a fuller expression of these ideas see previous posts on subjectivity and Michael Graziano.

One further note, and you should be able to intuit something like this, I recently read Ray Kurzweil's book How to Make a Mind and he makes some of these similar arguments. I thought the book was lacking in depth and coherence on some of these core issues, though he lays out the general problems well. It was also a fairly enjoyable, fast paced read.

_________

An argument against free will, consciousness, and intentionality.

A) Watson and Deep Blue (machines) process information (reason, intentionalize*, choose courses of action) in a competitively useful way that matches their human opponents’ processing of information as regards the games of Jeopardy! and chess.

B) Assuming that the human processing of information has conscious elements and the computers don't, whatever structures and functions consciousness grants human beings during this type of information processing (reasoning, intentionality, decisions) is not special**. That is, whatever properties or functions adhere to human consciousness in this decision making can be duplicated or outdone by non-conscious structures, as it has been outdone in these games.

B-2) Other processing of information in different games or language use is of similar structure as that in these games.

C) There is no reason to believe that the processing of information during moral and social decisions is of a different structure (or is benefited by a different structure) than the processing of information during Chess games or Jeopardy.

D) Whatever properties and functions make up consciousness they are not special, that is, they do not grant us capacities different than that of non-free-willed***, non-conscious, behaviorally determined entities that process information and make decisions based solely on whatever their internal state is at that time, the environmental inputs, and whatever algorithmic procedure incurs.

Conclusion:

Consciousness and free will play no useful, functional or structural role during Chess or Jeopardy decisions that could not be equally structured non-consciously. Consciousness and ‘free will’ add nothing to moral and social decision making as well.

*By intentionality here I mean the relational status in the processing between, say, thinking about "Paris" and the actual Paris. I follow Dennett and Rosenberg (among others) in saying that original intentionality never coheres and that the intentional state is a functional representational state that provides a brain or computer with appropriate structural formation and behavioral responses. The "appropriate structural formation" being that there is a correlation in the structure of, say, the real geographic relationship between Paris to Lyon or between two sides of a triangle and the brain/mind representation and perceptual models of the relationship between Paris and Lyon or between two sides of a triangle. This structural formation of brain/mind, whether human or computer, thus grants “appropriate” or shared behavioral repertoires. As was shown, consciousness does not grant us greater (more useful) intentional structures as we play Chess or Jeopardy. In other instances of human intentionality, say about a moral decision, the intentional structures that adheres to human information processing (including consciousness’s role) is of a similar intentional structure as to what Watson and humans do when they are processing information about “Paris,” and we have to assume that such grants us no behavioral or processing capabilities that could not be granted non-consciously.

**Conscious creatures may be "special" in the sense that there is "nothing else it is like to be that thing," but such specialness is probably blown out of proportion by our yearning to be special (not to mention that any complex representational system is unique in the relational qualities it is representing). That is, the things that truly make human beings special, our complex societies, projects, imagining of a great many worlds and how we can respond and build different ones, is not granted to us because of consciousness: those abilities could have been and can be produced through other non-conscious means. Though, evolutionarily speaking, that beings like us would be conscious may have been the most likely outcome.

***On free will, we can, of course, take the compatibilist route and say that both Deep Blue and Deep Blue's opponent (a human) had "free will" and were making "free choices," and that moral and social decisions are "freely chosen" in the same way. I am taking free will in the incompatibilist sense, but also hope to push the idea that the compatibilist notion makes us nothing better than complex computers, which hopefully drives a wedge into the multitude of connotations that compatibilists sneak in when they claim, baldly, that "free will" is real. Such claims of free will often sneaks in the idea that we do something substantially different in decision making and choice making than Deep Blue or Watson or some other machine when they make choices. Again, though consciousness may make a description of human choice making more complex, there will not be some strong dynamic that makes our decision making radically, functionally superior than a non-conscious computer's choice making. At least that has to be our tentative conclusion.

On subjectivity

A little more on subjectivity

I have always felt intuitively that the idea of the first-person private sphere is overly stated. There may be a significant impracticality to understanding what is happening in the brain-that-is-you. That is, there may be an impracticality to understanding your experience. But assuming that we as individuals are machines programmed by our histories, with our own models and representations, and endless personal knowledge and therefore associational structures that are only affiliated to us as individuals, then the idea that you have some impossible-to-see first-person experience makes sense.

Except that I strongly deny that it makes sense, or at least, that it gives us any epistemological purchase about the nature of consciousness. The reason why I am walled off from your experience is because in order to have “your experience” I would have to be a machine built by genes and environment to be the exact machine that you are at a given moment within that exact environmental space. Through story telling and background sharing and explaining the connections of the immediate environment that we as individuals are fixated on, one individual can share certain relational aspects of their experience with any other. In other words, there is a reason why us Americans like the same television shows. They elicit some similar experiences that we all share. Enough of our modeling, association structures, and perceptual structuring is shared, and therefore great similarities in our current representation of the present moment, say a TV show, is shared.

Through nature and culture similar machines (us people) are built across a society or a family. Therefore our experiences are close enough aligned to elicit similar reactions. That seems like a banal statement except so many people hang so much on what such subjectivity amounts to.

Now, can I experience the show in the exact same way as the person sitting next to me experiences it? No. But again, given the kind of machines that we are that are delicately programmed over many years to form the exact repertoire of associations and emotions that we do when we are present in a given environment, to enter into your exact experience would require me to have your (near) exact programming. What walls me off from understanding in perfect detail “your experience” is not some divide of the specialness of consciousness. It is instead the practical impossibility of programming enough of your associations and models to elicit the exact way that your perceptual representations and brain/body processes will play out.

Some Examples

Given the internal model or representation of what it is like when two humans experience a needle in the arm, including shared structures of pain response as well as shared cultural inculcation of pain response behavior (etc.), our internal representation, our experience, will have certain shared qualities. It will have other personal, subjective qualities that other people do not have because their associational, representational, and bodily responses are slightly different. For example, two ten year old's who are scared of shots will share more inner representational similarities than those ten year old's will share with the greatest stoic out there while she is getting a shot. Though, still, there are surely at least some inner representational similarities, for instance, some aspects of what-it-is-like when someone touches you on the arm will be shared by all three. We can assume this is partially true given that we are all generally wired in similar ways, our bodily representation models share similarities, say.

The simple story is that experience, that what it is like, is some kind of internal monitoring of our conglomerate representations. These ten years old's have significant representations that are centrally focused on the representations of,  “touching arm; this is supposed to really hurt.” And the stoic has significant representations that internally get focused on, “touching arm, pain means nothing in the world.” I put the “representation” in linguistic form there, but obviously there is immense complexities there which only at times are influenced by social, linguistic mediation.

The individuality and uniqueness of our experience, of our internal representations of our self as we interact with the world is the only way that possibly makes sense. If my computer was internally representing everything within its self and what processes it was carrying out, it would quickly be having unique representational structures. It is the only computer with its repertoire of documents and programs.

Likewise, if a rock was given a modeling and representational system, it mapped the information of different parts of its self, it would be the only one with an internal representation of atoms in that exact arrangement. If we then started talking about the rocks awareness of its own self, it would be the only one that was presently mapping that exact arrangement, as well as internally representing that it is mapping that mapping of that exact arrangement. Hence, it is subjective. It also makes perfect sense that its internal representational system (its information system) may have significantly different qualities than the next rock over. Say one rock is using a scanning electron microscope to array its atoms location and it draws such maps in crayons which it then internally is shown a partial, singularly focused map of (its center of awareness). The other rock uses an internal fMRI and draws maps or representations in computer models which it then accesses and is centrally focused on (or represents its attentional system as singularly focusing on). The individuality of those internal representations, and the specificity of the given awareness beam, is going to follow from those individual structural systems.

From these examples, it follows that the internal representation of the bat is likely more walled off from visual creatures than the internal representation of visual creatures is to other visual creatures. We may be able to glean some similar inner representational similarities, say if bats have spatial mapping that is something similar to our spatial mapping, there may be the slightest overlap in what our representational systems are doing. In those ways perhaps a bat and a human will have some similar experiences, like the kids and stoic getting a shot above. That there may also be unimaginable differences in internal representation because of significant structural features, in this case perceptual differences, is no more a useful insight than that you can never fully internally represent the exact experience of the person next to you. 

Lastly, language may be needed for some of the most acute internal representations and modelings. A bat and dog may have awareness and have robust experiences, but only humans (that we imagine for now) have an awareness, a representation, that we have awareness. There is good reason to believe that language provides much of the scaffolding for that kind of self-awareness. Language allows for an explosion of representing our selves as we are representing, and for delegating and dictating in finer detail all of the distinctions in the world, including as we attend to those different distinctions. It allows for thoughts on thoughts. For us linguistic beings, we may have a far tougher time imagining the experience of any nonlinguistic being than we do trying to understand specifically the bats perceptual differences.

As to feelings . . .

Whatever the painfulness of pain amounts to, there is good reason to believe it was in animals 50 million years ago. It is not feelings that have allowed humans to do the incredible things they do. Assuming it is feelings which many think gives consciousness and humans special capacities, there is good reason to think they are wrong.

Some thoughts on consciousness

Update: For a more coherent explanation of Graziano's program check out selfawarepatterns. 

Here are a couple of articles by Michael Graziano, one older from Aeon and the other recently from the NYtimes. I agree with his general program towards how we are going to deal with consciousness. The Aeon article is older than his short blurb in the NYTimes, and I think the former suffers from the fact that he was not upfront about eliminating and reducing consciousness, which it looks like he is now more willing to immediately own.

Just for the record, I am an eliminativist and reductionist. I believe certain parts of the concept of consciousness are mistaken ascriptions from our metacognitive stances, in which mistakes have arisen because our first hand experience is a very poor tool for analyzing what consciousness is. Much of our other ascriptions of the concept of consciousness can be reformulated into a representational or information type structure. Your brain is forming models and representing (loosely speaking) different perceptual experiences and how those present experiences fit into other models and other previous representations that we have. Emotions and feelings are bodily processes, some of which, particularly emotions, can also be set into representational structures as we represent the state of our being. Feelings are more tricky, but we can probably ascribe some bodily activity to them or put them into some kind of representation schema (pain is a representation of the pin pricking your arm, say). Some day I will deflate (in the least) the rest of those feelings. I will show that, on the bare, those feelings and the power of such do not amount to much and were already well established in lower animals. The rest of what we ascribe to consciousness, as above, is window dressing and can be seen as complex information crunching given the kind of powerful, parallel machines that we are. 

Some influences here include Dennett, Patricia Churchland, Thomas Metzinger, Stanislas Dehaene, Jesse Prinz, and Antonio Damasio. 

But, for now, my thoughts on Graziano.

The argument is going to be that what we call “consiousness” is really an attentional, representational structure. It is a model of a self (body) that is in certain relation to all of these other models of the world and other bodies (humans, animals). Conscious awareness, under Graziano, is a robust, continuous representation of our self's attention towards various objects. It is that internal representation. Or “consciousness” is what we ascribe to that constant process of relational representations at the center of that focal process. Awareness of our continuous attention processes is what we have modeled and classified as consciousness. Awareness here is really just another representational configuration. 

We also have this to different degrees. Early mammals and even reptiles have a great deal of representation of the world, say through visual processes. They also have complex attention processes; they focus their activities and perceptions onto different objects. Over time animals developed more ability to direct their attention. A monkey, I assume, has greater attentional control as well as a larger perceptual repertoire than a lizard. In the end, animals have a representation of a certain part of the world (their visual focus) and they also have the beginnings of rudimentary modeling and representing of self/world relations, which allows them to engage with more complex attention and discriminating processes. 

That kind of attentional visual awareness is still alive in homo lineages and more modern human consciousness that arrive later. As language comes about, we now have even more complex models and representations of self/word, including in the end literally creating the concept of consciousness. As Graziano talks about, animals are already modeling other animals behaviors and anticipating their behaviors, which is a gradual buildup to a rudimentary other-mind-analysis. In humans we started from those more limited world/selves modeling capacities, which are most robust in mammals, especially in certain pack mammals, and maybe in some birds. This kind of strong visual being with consistent, careful streaming of the behavior of others as well as one's self, is where human consciousness starts. But as we add language we arrive at a full blown theory of mind, that is, we start actually pondering about the inner workings of other people, and asking about their desires and intentions (etc.), as well as coming to better grips with our own immediately perceived psychological architecture. And eventually deciding that we are conscious.

Trying to place other animals' consciousness as compared to humans is a frequent pastime. When it comes to something like taking in a specific visual scene, there are two bases of consciousness that seem central. One is simple visual processing. In that sense, the chimp's immediate visual processing of red seems like it has to have many similarities to our visual processing of red, at least at some basic level. Now, the main difference stems that an adult human's self/world modeling is far more complicated, and any single experience is not experientially bared to only visual processing. That is, our consciousness of red is not some isolated percept. Some kind of Buddhist like understanding of what we do when “we experience” is far removed from what the concept of experiencing usually means. Any single experience of red is tied into other representations and to our modeling altogether. This is where some of the higher-order-representation and self-representational theories come into the picture. We are usually not only immediately processing shape and color, but our experiencing is going to be saturated by other representations, such as the background idea that "I am Lyndon and I am experiencing red.”

It makes sense that at least part of our present experiencing were there in chimps and in early-non-linguistic humans. The reason why we think a chimp “experiences red” differently than a gorilla “experiences red” and much differently than how a human “experiences red” is because of the entire nesting of the visual processes into their representational schemata, as well as obviously differences in color discrimination.

This also explains why any individual's experience of red is different at different times. You may be seeing and processing the same shade of red at the art gallery as you do when you are watching a firetruck at a parade, but we cannot rend an “experience” from a broader context, at least I do not think it makes sense. If we can find some common ground to both experiences, then it will be by stripping out other contexts of that representation and not fully relaying “your experience” as you “saw a red firetruck at the parade.” How we describe your experience of red in that particular situation will encompass more than the simple experience (representation) of merely seeing red. Those representations, under my formula, also includes whatever feelings/emotions that are created in such a scene. Say, if we find some underlying singular program of "seeing red," where we try to isolate a particular representational state of seeing red, including the arising of particular emotions/feelings that the brain produces along with that representation, it will not describe the actual experiencing of any particular life-event. Some shallower categorical move can be made, such that there are similarities across our processing of a particular wavelength of red that will be the same, but that simple categorization will never exhaust any singular experience that we have.

As far as uniqueness and subjectivity. Your particular model and present representation (I am John in Seattle getting coffee and my dog is at home waiting to be fed, I need to call my doctor . . .) is of course unique. Given that this is an “experience,” you are the only one with anything close to that particular representation and model structure. Furthermore, the exact set of emotions and feelings that surround your personal representation is also unique. Given the above thought, the way that such an immediate thought has a certain construction of bodily processes, feelings, and emotions tied to it is rather uninteresting. Likewise, certain socially-shared representations, given nature's structures, will have some shared emotional responses with many other people who have similar experiences. Being stood-up for a date, say, is a common experience that has similar representations as well as similar visceral reactions across people. The fact that your representation/modeling is slightly different than anyone else's (no other John was stood-up by Julia Roberts on May 24^th), is entirely uninteresting. It is also uninteresting that a complexity of that experience is tied to a particular set of emotions, which have been tempered by all the other experiences you have had in the past as well as homo sapiens's givens. It makes sense that we can find similarities across other peoples' similar experiences and across other of your own experiences. The uniqueness of the present situation, the uniqueness and subjective factors of the exact representation of you amid your world/self models with the exact emotions that arise at this particular time, makes sense given the kind of representational and bodily creatures that we are.

Furthermore, we can also shrug off as uninteresting the idea that your particular representational substrate, that you represent certain wavelengths of light within the representational schema that you do, is a mystery fact. Models and representations are made for functional use. It does not matter whether your map is drawn in crayon or pencil, what matters is that such is functionally usable. Now, some internal representation of what that map looks like will be slightly different based on our's or nature's choosing of the exact substrate of that map (pencil or crayon; emotion-based-thought-processing or cold-calculating-thought-processing), but again, the fact that such an internal model has a particular, unique representational experience, since one was drawn in pencil and one in crayon, is something we can shrug at. It seems obvious that a chimp's location mapping system may have a different "feel," or different internal representational characteristics, than a bat's location mapping system, given that one was first structured mainly by visual processes, say, while the other was structured by sonar processes. Or something to that effect.

The bat's world modeling and perceptual representations being difficult for us to represent, that is they are hard to image in the way the bat does, makes perfect sense. Given the bat's perceptual structures and brain structures, its particular experience of world representing and modeling is not something that should throw us for a loop.

Overcoming Underachieving, Ruth Peters (2000)

This book is pop-psychology or self help, but I somehow stumbled upon it a while ago and gave it a quick read. I am not going to review it, but I do agree with the main message which she delivered well. Her main message, or the one I took, is that all children (generally speaking) are capable of being successful in school and all it requires is enough work. She argues there are ways of building familial relationships and attitudes towards school that can get children to put in the necessary work. If they do such, they will be successful in school and probably in life more generally. They will be happier. Very well. A good message that I wholeheartedly support. Success in school is about the amount of work you put in, for the most part.

But I thought I would bring this book up for a slightly different reason. First, the good, and rereading this quote it really does embrace much of what I believe, what I want this blog to be about.

As we’ve seen, academic achievement is more a product of appropriate placement of priorities and responsible behavior than it is of intelligence. There are lots bright kids who flunk classes, brilliant high schoolers who drop out and quick-witted adults who can’t seem to hold a job or effectively face challenges at home or in the workplace. Almost any child can do well at school whether they have low, average, or superior ability. What seems to count the most is one’s work ethic and willingness to tolerate frustration—to continue to tackle tough math problems, even though your first impulse is to give up, and to bring home and study the science text, even though tonight is the season finale of a favorite television sitcom. This type of prioritizing is often not innate—it’s learned by watching others (generally parents), receiving guidance and supervision (by Mom and Dad), as well as by making the connection between behavior and consequences (yep, performed by Mom and Dad again!). . . . The single best predictor of academic success is parental involvement. That’s not only a stunning statement but a heavy obligation. . . . This book is about academic success. Of all the bases of self-concept described above, this area is the one most within your child’s control and your influence. Appropriate study skills guarantee good grades and excellent performance. Study skills are merely actions, and self-esteem is based on esteemable acts. It follows, therefore, that the greater the quantity and quality of study actions your child performs, the greater number of esteemable acts will be attributed to her. The larger the number of esteemable acts, the more solid and stable will be her self-concept. The connection between parents motivating kids to employ good study skills and the resulting self-esteem level is strong. . . . Academic achievers are also seen as school leaders. In grade school they get picked more often to clean the blackboard, to mentor younger kids, and are trusted by teachers to take notes to the office. . . . Remember, the difference between an A student and a C student is often not based in intelligence, it’s a matter of sweat equity—how much one is willing to put into studying and completing tasks.

That's all nicely said. I will preface this next part by saying that perhaps we should not look to cheap writers for more endurable thoughts. The author has torn down some basic problems of socialization/education, but she is far from giving us the best understanding of some of the more entrenched relationships of self and world. The more difficult analysis of the relation between identity, behavior, social norms, and institutions are necessary to engage in. This passage is the extension from above (p. 201). So, I give you a book-throwing-quote:

And the cycle of self-concept continues as we mature . . . Sure, looking like Julia Roberts or Brad Pitt doesn't hurt and social extroverts tend to have an easier go at it than do the painfully shy, but what really comes to the fore is the work ethic developed in the earlier academic settings. Employers look for grade-point averages when they hire, and responsibility behaviors and productivity when they promote. Prospective spouses, often initially attracted by physical appearance, remain impressed by earning capacity, responsibility, commitment, and stability (often outcomes of learning good frustration tolerance as a kid). Marriages continue not because Dad can shoot hoops well, but because he provides for the family by finding and keeping a good job and acting in a loving and responsible manner. Mom remains attractive to Dad not only because she works out or looks great in a bathing suit, but because she’s up on the daily news, has interests and therefore is interesting, and is fulfilled in her career at the workplace or does a good job keeping the house and kids on track.

Peters' general program is not that bad, and I am sure she helped at least some families approach school and family life in a better way, both in her practice and through writing. But the above quote is insane. She is concentrating on how to be “successful” in the world that is provided to us, but the reproduction of certain familial structures and personal relationships are capable of being structured in better (more useful) ways for the individual. This is especially true for gender and family relationships, and part of becoming “educated” and knowledgeable must also include the capacity to reflect on one’s own self and institutions where one can. Obviously, certain social institutions and behavioral norms are outside one's control. That does not mean they are beyond rethinking, and that does not mean we cannot form communities, families, and discourses that open up those institutions and norms to reflection. And I would hope, in time, that we would make significant changes to our social world and our selves. But, pointedly, when well educated people turn to talking about improving our “self-concepts” they cannot ignore such institutional and discursive structures. There are many points of our social world where it needs to be continually stated that such states of affairs do not have to be reproduced.

I often give a caricature of Mary Wollstonecraft's argument. Such as: 

“It is not the genes or willpower of “Woman” that means that she is a doltish, emotive, shallow individual who cannot carry on a conversation or be a decent object of pleasure to “Man,” it is only from an absurd environment and cultural expectations that we set up around her. Give “Her” the education and socialization of “Him” and you will see females become more worthy life partners to males.

And we thus embarked on feminism. I read Wollstonecraft as arguing against the essence of gender characteristics, or against the genetic structuring of gender expressions, at least in the manner they had inculcated her society. When Peters encourages the cultural reproductions above without providing for ways in which we do not have to reproduce that world, she does a great disservice to helping people understand their self-concept and to create better social worlds. The quote from Peters above grates in my mind against the most basic belief about dichotomizing gender differences, and doing it in a spurious way. Many people who are reading Peters or talking to her in her practice as a counselor may not blink an eye at such ideas. But the line that to become a more attractive wife today a woman should manage to occasionally read the news (because she is self-disciplined), and therefore can have a decent conversation with her husband, is straight out of the 18^th century.

Just for charity I will give you the rest of the passage from above, which shows that the message is for the most part a good message:

So if you want to help your kid with her self-concept, there are many things you can do. Save up for the braces and figure out how she can have some nice outfits to wear. Shoot hoops and throw the football with your son. Talk to your children about social skills, such as good eye contact, smiling at others, telling jokes, and becoming a good listener. But most of all—set the state for them to feel smart. Teach good study skills and and habits, expect classwork and homework to be completed on a daily basis. Talk about going to college as a strong possibility . . .

One final note. This quote from Peters is bad. I think this kind of empty thinking leads to worse worlds for all of us, especially for females as we continue to reproduce a great deal of the society we found. Attacking Peters on a personal level does absolutely nothing towards uncovering the complexities of our social organization. To write long winded critiques with harsh and piercing language for the ideas beholden is necessary (such critiques can be seen as essentially personal attacks at times). The quote from Peters, and all the social structures that make it a sustainable thought, leads to an unequal world, one that I think most people are trying to move away from, including Peters. That does not mean we need to accuse her or her work of sexism or mysogyny. Among people willing to come to the table of reflecting on beliefs and our cultural world, as most people who readily read books and most blogs do, there should be more charity and careful analysis of ideas. That can include reflecting on social acts by prominent people, but still, even here, what is to be attacked are ideas and practices, not people. This is how you get under what are very complicated issues, and how you make careful, willful changes. I am confident that this is how useful, sustained social revolutions will happen. Human beings are complex creatures whose social embeddedness requires a delicate knife to understand. Coming to grips with these questions is not helped by tribalism, but by slow, steady analysis.  

Different Senses of Self

By dissecting the self, I very much mean analyzing institutions, really the environment in general, and how that environment creates the individuals that we are as it interacts with our genetic structures. What we are looking at there are structures of identity and behavior. 

The self is also analyzed in the phenomenological sense, in the entity that we postulated at the center of recurring thoughts, at the center of our head, as that which streams the flow of perceptions. I follow people like Thomas Metzinger (The Ego Tunnel) and Bruce Hood (The Self Illusion) among many other cognitive scientists and philosophers (etc.) who focus on the idea that the concept of self that we naively postulated in the past does not match with our present best understanding. Some of what we engulfed in the sense of self was posited there because of transparent structures of brain to mind. We could say something like the little voice that runs through your head, your immediate experiencing of world and body and thoughts, encouraged us to postulate an entity, the self, that is no longer holding up under the closer inspection of neuroscience. Before the arising of cognitive science, broadly speaking, there were thoughtful criticizers from philosophy, psychology, and Buddhism.

There is often wrangling around the term illusion, which should often just be ignored in cases where conversations break down into such. There are a few good things to keep in mind. When people say the self does not exist or that it is an illusion, they often are arguing against its conceptualization as humans have created it from various experiences. There is usually reference to some continuous, seamless, unified and unitary entity that we postulate as who we are. As we dig deeper into the brain we are finding something that is different than many of those things. It makes some sense to me that we would call our original postulations illusions, as they were arrived at out of naive reflections on our experiences. There are other senses of self, of what is usually conceptualized as what it means to be a self, that are not illusions. We are brains and bodies, and if by self we just mean that generally persistent entity encompassing such, including much of the persistent brain/mind structure, then that meaning of self is not an illusion.

One of the main themes around this subject is that our mental thoughts were not structured to be faithful analyzers of what those mental thoughts are, and thus eventually what human beings postulated as the concept of self naturally consumes misleading information. On a similar structure, we wan look at the phenomenology of making a choice, of having contemplations and desires running through our head as we actively select among various options, without having any access to the working of the brain and much of the other constraining and thought-organizing factors. That structure all means that we experience an unconstrained, unlimited control of our choices from the first person perspective. I think it is one of the reasons why compatibilism will always be empty. It makes far more sense that our concept of free will is wrapped up in this constant experience of choice making, and that this experience has a very misleading or even self-unanalyzable phenomenological aspect. As to the general problem, R. Scott Bakker at the Three Pound Brain gives good arguments to these themes of misleading phenomenology, and he even goes further to claim that such a gap is unbridgeable. We are doomed to bad theories in some ways. I am not that pessimistic. I think we will eventually get a grasp on consciousness and these other mental phenomena. 

I advocate, or playfully enjoy, cleaning up language and concepts. There are always equally adequate ways of getting across the same concept, and making those changes allows us to control connotation as we try to get to what is important. So, with that, I am happy to discard the self and free will to the dust bin. What's important is trying to figure out what kind of creatures that we are, what our experiences consist in, and in what kind of selves and society we wish to create.