by Tom Pepper
Over the last year or so, I have many times dismissed Thomas Metzinger’s “Self-Model Theory of Subjectivity” as flawed, and claimed that it functions to produce a capitalist ideology of the subject under the guise of science. This summary rejection of Metzinger’s project has angered and frustrated a few participants here, and I have only ever given brief and cursory explanations of the problems I find in this particular form of cognitive reductivism. Here, I propose to give a somewhat fuller account of what is wrong with Metzinger’s theory, with the hope of avoiding having this same argument every time his name comes up.
I am going to focus on his essay “Précis: Being No One,” in which he sketches the argument of his very long book. This is not meant as a formal academic refutation of all of his work, by any means. The “Précis” is sufficient because it focuses on exactly those parts of his book that I find to be most problematic; it is also useful because, being fairly short, any readers not familiar with Metzinger can get a good idea of his project by reading it themselves (I provide a link to the pdf version below). It is my hope that after reading this, any reader can extend the argument against Metzinger’s theory to the rest of his project on her own. There has been little criticism of Metzinger from the discipline of philosophy in the English-speaking world; no doubt this is largely because the philosophical errors in his work are so startlingly obvious that most philosophers would be embarrassed to take the time to point them out. It would be a bit like asking Freud to engage in debate with Dr. Phil. However, I am no Freud, and don’t have quite as far to stoop. There have, however, been some responses to Metzinger from philosophers, and I give links to some of them below; although the emphasis is different from my own, they basically find the same fundamental faults in his work. Perhaps if Metzinger did not claim to have finally solved once and for all the perennial mind/body, and then offered only a recycled solution that any undergraduate philosophy student should be embarrassed not to recognize as a familiar error, no philosopher at all would have wasted any time on him. If one makes grand claims, one ought to at least make grand errors.
What I propose here is simply to point out the major problems with the Self-Model Theory of Subjectivity, explaining why it fails to solve the problem it claims to have solved. In fact, it fails to even recognize what the problem actually is. Then I will proceed to explain how this theory functions as capitalist ideology, working to ensure the reproduction of the existing relations of production.
Problems With the Self-Model Theory
I will begin with the most damaging error—which is also the one most often pointed out by those who criticize Metzinger. In philosophy, this error has many names, such as the “homunculus problem” or the problem of “infinite regress”; I will here refer to it as the “subtle atman” problem. This is when we depend on a transcendent and non-determined “consciousness” to explain how our consciousness is completely determined and not at all transcendent. When someone asserts that there is no ultimate essence or core self, but his account of how the apparent self (conventional self, phenomenal self, illusory self, etc.) arises requires the existence of an ultimate consciousness, then he is simply sneaking an atman in the back door. This is the problem Hume faced when he attempted in his Treatise on Human Nature to account for the mind completely from within a naturalist and reductive model. Here is how Galen Strawson explains this problem in his book The Evident Connection:
Within a year, Hume sees that he can’t maintain the view that this is the true idea of the mind, although his empiricist principles commit him to the view that it is. Or rather, he sees that it’s not the idea of the mind that he’s worked with in his philosophy, although his empiricist principles commit him to working with no other. This is his problem: the empiricistically ‘true’ idea of the mind isn’t consistent with his philosophical commitments considered as a whole. . . He realizes that he has throughout his philosophy made us of—presupposed—a notion of the mind or self that is open to the objection that it is not legitimate by his own lights. (33-34, emphasis added.)
Now, Hume realized the error he was making, and had the courage not to hide behind the lame excuse that this is just a matter of the conventions of language. He understood that his empiricist model of the mind required the existence of a Cartesian separate consciousness—that it is not merely our conventional way of speaking that makes this error, but a failure of this explanation; the problem was that the working of the empiricist model always required the existence of some separate, undetermined, guiding consciousness, and Hume rejected the possibility of anything resembling a soul or any kind of transcendent essence. Ever since, this has been a problem that no reductionist model of the mind has been able to solve, and a problem that every undergraduate philosophy student is required to understand.
Metzinger, however, clearly remains oblivious to this centuries-old error. His Self-Model Theory consistently makes exactly the same error, and refuses to accept or recognize that it is a real problem. He asserts that his theory claims that “no such things as selves exist in the world,” that there is no “unchangeable essence or …thing” but only a conceptual error in which a model of a self is mistaken for such a core essence or thing (3). And here is the problem. Who is making this mistake? What consciousness is making the error of falsely believing that a “representation” of a self is an actual self? If a mistake is being made, who is making it? This rhetorical sleight-of hand runs throughout Metzinger’s book and articles, but I will offer just a handful of examples from the “Precis” alone:
“autoepistemic closure”…refers to an a “inbuilt blind spot”, a structurally anchored deficit in the capacity to gain knowledge about oneself.(2)
Who is “blind” to this particular spot? What transcendent consciousness is unable to gain this particular self-knowledge?
The phenomenal property of selfhood as such is a representational construct…It is a truly phenomenal property in terms of being an appearance only. (3)
To whom is this representation made? Who is mistaking this “appearance” for a real thing?
The transparency of the self-model is a special form of inner darkness. It consists in the fact that the representational character of the contents of self-consciousness is not accessible to subjective experience. (4)
How can we possibly understand this “subjective experience” which is denied access to the workings of the brain, except as some kind of transcendent consciousness?
The contents of conscious experience are characterized by my ability to react directly to them with a multitude of my mental and bodily capacities: I can direct my attention toward a perceived color or bodily sensation. (6)
Who is doing the “directing” here? Who is the (undetermined, freely choosing) owner of “capacities” and how does she choose where to direct those capacities?
I could extend these examples for pages, even without leaving this one essay. But it should be clear enough that, like Hume in the treatise, Metzinger makes use of a philosophical concept of a free and undetermined consciousness that is not legitimate according to his own theory.
As with Hume, this goes beyond merely accidentally lapsing into the errors of ordinary language. Metzinger’s model requires the existence of a transcendent and undetermined consciousness, a subtle atman. As with cognitive science generally, which takes the computer as the model of the mind, there must be a programmer or operator who remains separate from the system and directs its actions and evaluates outcomes. Metzinger proposes the existence of a “model” of the self, and suggests that there is not self because this model is only a model, not an actual self. We act in the world, on Metzinger’s model, by observing this model and evaluating how it relates to “objects” in the world, and then choosing to “direct” our mental capacities to act in a certain way; the guiding consciousness which does this directing does so completely undetermined by the system, except for the limitations on its knowledge which arise from mistaking the self-model for an actual self. This error, for Metzinger, is functional—it is biologically and evolutionarily useful—but it might at some point become limiting of the capacity of this transcendent and undetermined consciousness—this atman—to acquire correct knowledge.
There is some passing attempt to defend against this error, but it is a rather weak defense. Either Metzinger doesn’t really see the extent of the problem, or doesn’t want to call attention to the complete failure of his system. He gives the tired claim that this is a matter of “linguistic tools from public space” (19), but he doesn’t seem to be completely aware that he has made use of those “tools” throughout his own argument, and that it fundamentally depends upon them. Clearly we could avoid the use of conventional language in academic discourse—Metzinger’s many invented or redefined terms should be evidence of this fact. This is not a matter of a necessity of common language and clarity, because he is not hesitant to redefine terms throughout his project and insist that we ignore their common meaning. And while we may frequently fall into the use of linguistic convention in, for instance, explaining how our computer works (the computer thinks this is an error, etc.), we could, and often do, avoid this kind of language when we need to. Metzinger also tries to avoid this problem with the mind-as-computer analogy: “Any machine can do self-directed modeling”(22). Of course, this doesn’t solve the problem at all, because there is always a separate consciousness working to direct the machine to do this purportedly “self-directed” work.
This one problem alone should be enough to discredit Metzinger’s repackaged empiricism. I will, however, point out three other problems that are made particularly clear in this essay. This doesn’t even begin to cover all of the errors in the book of which this is the précis. It will, however, give us some indication of where to look for the ideological purpose of Metzinger’s pseudo-science.
One enormous problem is the insistence that this is a neurological account of the mind, an account that explains how the mind arises from the brain. This is asserted, but there is in fact no evidence whatsoever that this theory corresponds to anything in the brain. It is simply assumed that if we limit all research to trying to find “neural correlates” for the “constraints” of the system, we will find them eventually. For each “constraint,” however, we are informed that there is no evidence whatsoever to support the theory: “Currently no detailed theories concerning possible neural correlates…for the appearance of a coherent, conscious model of the world do exist. However, there are a number of interesting speculative hypotheses”(7); “Neural correlates of the window of presence Very little is known in terms of implementational details” (10); “Neural correlates of convolved holism Once again, we have to admit that not enough empirical data are currently available in order to be able to make and precise statements”(16). In the absence of any neurological evidence for the theory, and given the obvious lack of conceptual coherence, why should we accept this theory at all? Why would anyone even consider it? I will return to the ideological motivations for accepting a theory which lacks both empirical support and theoretical coherence.
Another philosophical error is the assumption, common in empiricism, that objects have an inherent givenness which we passively perceive. The world, on this understanding, organizes itself naturally, and we are simply machines designed to detect and adapt to it. Since Heidegger, the problem with this conception should be too obvious to need explanation, but I will do so, briefly, anyway. Metzinger assumes that our perception of objects in the world is related to an erroneous self-model, but that the objects as such are simply given. That is, we may be mistaken about how an object can be of use to us, but we are always seeing the thing as it naturally occurs, always “carving nature at the joints,” in our perceptions of it. We may have varying degrees of detail in our perception, but the discrete objects we perceive are objects “from their own side,” as it were. All perception is the accumulation of “information”: “complex information pertaining to dynamical subject-object relations can be extracted from reality and used for selective and flexible further processing”(4, emphasis added). The “first-person perspective” is always “trying to give fine-grained and clear descriptions of target phenomena”(5), and the limitations on this are always at the “neurobiological level”(5). In a revealingly Kantian example, Metzinger explains that when we look at “a beautiful landscape” we can achieve a “briefly emerging integrated whole”(15). Kant assumes that the aesthetic appreciation of a landscape is evidence that our sense perceptions are in correct alignment with our concepts, that we are perceiving the world correctly and so are thinking correctly, and are correctly adjusted to the world. If we do not perceive as aesthetically pleasing what is generally perceived as beautiful, this is an indication of an error in our thought or senses. Metzinger’s theory depends very heavily on this Kantian aesthetics of the beautiful, meant to encourage political stability and acceptance of the existing social system, and to pathologize any discontent—I will return to this point later.
The problem with this assumption of a “reality” which we detect in perceptual “information” is that it just doesn’t correspond in any way to how people actually live and interact in the world. As Andrew Collier puts it, empiricism “can take experience itself to be an authority above criticism, unaware of the way experience can confirm our prejudices, since we may see what we have been taught to see”(72, emphasis added). The beauty of a landscape is a learned perception. In fact, many more of our perceptions of the world are learned in language and social practices—even including visual perceptions and sensations of taste and smell. Metzinger’s model very powerfully insists on an atomistic subject correctly perceiving the world, with possible errors or limits in our evolutionary biology. He cannot consider the role of the social formations in producing the subject at all.
This leads to the final problem I will discuss here: the ridiculously naïve theory of language. For Metzinger, we all develop a “private language” along with our model of our self, and then, once we have developed this (illusory) self and (neurologically determined) language, we are complete atomistic subjects who can enter into social relationships and “the transition from biological to cultural evolution”(31) begins. As Metzinger explains the process, it begins with “processing by subsymbolic mechanisms like attention or implicit memory,” gradually working up to “concept formation, metacognition and verbal report”(6). These perceptions are “entirely determined by the properties of your brain”(24), and even the sense of “self” we develop is an empirical and biologically determined concept, never socially constructed. Metzinger claims that his theory is “truly radical” because it asserts that “no such things as selves exist in the world”(24). The self is only an error that the transcendent consciousness (which does not exist) makes when it mistakes the biologically determined “self-model” for its own mirror image.
Again, this concept of language acquisition is probably too obviously absurd today for there to be any necessity to point out the error, but I will do so anyway. Clearly we do not produce a language from neurologically predetermined perceptual dispositions. It should be obvious to anyone who speaks a language that we enter into a pre-existing language because of our social experiences, not neurological predispositions. This theory of language is required by the Self-Model Theory of Subjectivity, however. If language were socially produced, it could be a result of social formations and a site of social struggle, and this would require us to look for the mind and the subject in social formations instead of in the brain the computer metaphors of cognitive science.
I have offered only a cursory indication of the most glaring errors of Metzinger’s theory. Still, it should be enough to make anyone wonder why anyone would be duped into accepting it. In fact, I have to admit that while writing this I feel a bit like someone who is offering scientific arguments against the existence of fairies, since I don’t personally know of any academic in the U.S. who does take Metzinger seriously. Nevertheless, it seems that several participants on this blog think he has something important to say, and so I offer a short account of why I think giving Metzinger serious attention is a mistake. I will proceed, now, to explain the ideological implications of Metzinger’s project, and try to account for why it might be appealing despite the total absence of empirical support or theoretical coherence.
Ideological Function of the Theory
Once again, I am using the term ideology in the Althusserian sense. An ideology is not an error or illusion or deception or trick or “false consciousness”!!! An ideology is a belief imbedded in a practice that functions to reproduce the existing relations of production. Ideology is our relations to our relations of production, the way we experience our existence in the world, and so is necessarily very real, with real causal powers. It is not “real” in any mind-independent way, however. While the existence of the planets and the stars that make up our solar system are completely mind-independent, in that they would continue to exist regardless of whether there were any conscious beings to know about them, our ideologies would cease to exist if we did. This does not make them any less real. The United States is a very real entity, with real causal powers, but it is completely ideological, in this sense, because it is mind-dependent, would cease to exist if there were no people left to believe in it and follow its laws, and it functions to reproduce the global capitalist economic system.
Of course, ideology can be based on error, and can require deception. It doesn’t need to, though. Perhaps the worst kind of error is when an ideological concept is mistaken for a scientific truth. The idea that “hard work” is good is an ideological concept—it cannot be “proven” false, because it is true to the extent that we believe in it, and that we get some benefit and satisfaction from doing whatever counts as “hard work” in our culture. However, if we claim that the virtue of hard work is a “natural” need of the species, that it is a concept that is true separate from our particular cultural construction of “work” and subjective experience of satisfaction, we are attempting to reify an ideology and delude people, including ourselves. Whenever an ideology requires deception or error, whenever it can only be perpetuated by limiting knowledge or requiring belief in a falsehood, it is a safe bet that it is an ideology that is oppressive and that we is not in our best interest to perpetuate.
How might Metzinger’s theory function as an ideology? It clearly requires both ignorance and error, so it is likely an ideology we would not willingly accept and participate in if we were made aware of it. Let’s consider some of the possible ideological functions this theory might serve.
Political quietism: As the very Kantian aesthetic example mentioned above would suggest, one function of the theory would seem to be to produce subjects who would seek to adjust themselves to the existing social system, and never consider that their difficulties result from that system. Any problems would result in some internal error in “representation” or information processing, and the solution would be to make adjustment to the atomistic machine that is the human individual until she can adequately function in whatever role she is put in. The best indication of proper functioning, of course, is proper aesthetic sensibility: the appreciation of a beautiful landscape, and the total absence of any annoying concepts concerning the social construction of taste, or landscapes, would be an indication that the system is working well. Once this occurs, the soul trapped in this machine would feel at home, and the machine itself would seem “transparent”; and, of course, that soul/consciousness could be safely ignored.
Naturalizing emotions: Reproducing Romantic ideology, Metzigner separates out emotion and thought, assigning the former “natural” and causal status and suggesting that the latter has no real causal power and can be changed at will (apparently at the will of the transcendent consciousness that does not exist). Discussing “system-related information,” he asserts that there is a “degree of flexibility and autonomy in dealing with the contents of self-consciousness”(19). Emotions are similar to bodily sensations of pain and hunger, naturally occurring and highly inflexible, while “thoughts are something that may not even be determined…before being spoken loud or writing down”(19). The suggestion is that our emotions are not culturally influenced, but hard-wired, while our thoughts are more ephemeral and ultimately not really important at all, since they are the outcome of our actions, mere epiphenomenon which never motivate action themselves. I won’t bother to argue for the position that emotions are very much cultural constructs here; the point is that this assumption that the particular emotions we happen to have are natural and motivating reifies our existing ideology. In addition, Metzinger simply assumes without even feeling the need to argue the point that thought cannot motivate action, that it is powerless and a mere epiphenomenon.
To offer just one contrasting position, Spinoza argues that emotions in fact are thoughts, they are simply thoughts which are “inadequate” or imprecise, and so lead to difficulty in acting in the world. Clear and correct thought, on the other hand, can both motivate our actions (we do things for reasons we have in mind, not only because of feelings in our “gut”), and can lead to more successful and correct action, provided we participate in a collective which produces correct ideas together (on this argument, see Balibar, Spinoza and Politics, particularly chapter 5). Metzinger’s naturalization of emotions and reduction of thought to epiphenomena functions, once again, to reify our existing ideology; more troubling yet, it suggests that only imprecise and inadequate ideas are motivating and “true” ideas, and therefore that we can only be motivated by an ideology which distorts reality. This particular point could be developed more rigorously, using Metzinger’s Being No-one, but I’m not sure it is necessary, seeing that Metzinger isn’t really taken that seriously by mainstream thought; I offer it here only to indicate the problem.
Atomist and private property: As I have already pointed out, Metzinger has a completely atomistic concept of human minds, in which each individual consciousness is produced by an individual body’s interaction with the “information” in the world, and only once the brain has produced a consciousness and a private language does the individual interact with other individuals. Along with this absolute atomism, he assumes the completely natural and biological occurrence of the most important “phenomenal property” in the whole functionalist account of the subject: “The property of mineness (also often called the ‘sense of ownership’)” (19). The mistaken belief that we have a private consciousness which is not at all socially constructed is a central error of capitalist social formations, leading to frustration and suffering in our everyday lives. “The private consciousness and the mystified consciousness go hand in hand” Henri Lefebvre explains in Critique of Everyday Life, Vol. I, “reinforcing each other and becoming increasingly entrenched as a result of instabilities which have their origins in real life” (153). Metzinger places the concept of the atomistic consciousness in the category of scientific fact (it really is completely individually produced according to neurological necessity), and the ideological concept of “private property” becomes simply a necessary error, one we must have because it is necessary for the working of the entire model: it “bundles [the] differing forms of phenomenal content”(19). Lefebvre, again, explains the ideological importance of this pairing of our existence and ownership: “Under capitalist regimes, ‘to exist’ and ‘to have’ are identical…And this situation is not a theoretical one…it is an ‘absolutely desperate’ reality; the man who has nothing finds himself ‘separated from existence in general’” (155). Metzinger simply takes this capitalist social formation for an eternal scientific truth.
Structural Mystification: Perhaps most importantly, should Metzinger’s theory be widely accepted, it would produce a discipline which functions to produce a certain kind of “knowledge” while obscuring and preventing the production of other kinds of knowledge. Gary Potter uses the term “structural mystification” to name a particular pair of contradictory concepts concerning knowledge production: “1) the production and dissemination of knowledge is an essential characteristic of all educational systems; and 2) the obfuscation of the production of knowledge and the restriction (and sometimes outright prevention) of the dissemination of knowledge are essential characteristics of the educational systems” (134). The point is that mystification works not by preventing the production of knowledge, but by insisting that one kind of knowledge is the “right kind” to be produced, and thereby framing the problem in certain disciplinary ways that will function exactly to prevent the problem from ever being solved. If the study of the structure of the subject were pursued in the way Metzinger would have us pursue it, we would seek the “neural correlates” which he is sure, despite his own admission of an absolute absence of evidence, must someday be found. We would try to determine what kinds of perceptions and emotions a subject should be exposed to in order to produce the right kind of behavior in the world. But there would be no discipline in which we could consider the social formations which actually give rise to our “mind,” and so no way to examine how we might change those social formations to reduce suffering or, to put it more positively, to increase our capacity to interact with and enjoy the world and one another.
The ultimate goal of Metzinger’s project, I would suggest, is to frame the problem in such a way that it become completely impossible to see it clearly, much less to solve it. In this way, it works to ensure the reproduction of the existing relations of production, and to avoid any production of knowledge of those relations and their oppressive effects. And then we must, as Metzinger does, fall back on the assertion that “certain aspects of consciousness are ineffable” (The Ego Tunnel, 9).
What I have offered in this brief, informal essay is not a thorough critique of Metzinger’s theory. Such a critique would require that I trace these errors through his entire system, point out all the other many errors the system contains, and demonstrate that they are unquestionably fatal errors, that they cannot be overcome without rejecting all of the foundational premises. Personally, I don’t think Metzinger’s work is worth the time or effort it would take to do this kind of critique; it simply isn’t sufficiently popular or sufficiently “convincing” to spend the time on it.
What I have offered instead is a mere indication of the most obvious and damaging errors, and the ideological function such a project might serve. I hope that anyone who might be mislead or confused by Metzinger’s “Self-Model Theory of Subjectivity” will be able to see these errors in the rest of his writings, and move on to more productive kinds of intellectual work.
Can we now put the Metzinger debate to rest, folks?
Metzinger’s essay “Précis: Being No One”:
Some short critiques of Metzinger:
Sonja Rinofner-Kreidl: “The Limits of Representationalism: A Phenomenological Critique of Thomas Metzinger’s Self-Model Theory.” Synthesis Philosophica, 40 (2/2005) pp. (355–371).
Hamish Thompson: “Review of Being No-one: The Self-Model Theory of Subjectivity.” Essays in Philosophy, 6 (1/2005) http://commons.pacificu.edu/cgi/viewcontent.cgi?article=1202&context=eip
Lynne Rudder Baker: “Naturalism and the First-Person Perspective” http://people.umass.edu/lrb/files/bak07natM.pdf
Lynne Rudder Baker: “Does Naturalism Rest on a Mistake.” American Philosophical Quarterly 48.2 (2011): 161-173. http://people.umass.edu/lrb/files/bak07natM.pdf
Baliber, Étienne. Spinoza and Politics. London: Verso, 2008.
Collier, Andrew. Critical Realism: An Introduction to Roy Bhaskar’s Philosophy. London: Verso, 1994.
Lefebvre, Henri. The Critique of Everyday Life: Introduction (Vol. I). London: Verso, 2008.
Potter, Garry. “Power and Knowledge: A Dialectical Contradiction.” Journal of Critical Realism. 9.2 (2010): 133-154.
Strawson, Galen. The Evident Connection: Hume on Personal Identity. Oxford: Oxford University Press, 2011.