What is phenomenal selfhood? What, precisely, is the nonconceptual sense of ownership going along with the phenomenal experience of selfhood, or of “being someone”?
In the discussion, he makes a striking comment related to the reality of the self. If the phenomenal self-model (PSM) is “of a nonhallucinatory kind”:
…the system then represents certain aspects of reality as being parts of itself, and it does so correctly. What it achieves is not only self-experience but self-knowledge. (Metzinger 2004, p 607)
In reading this passage, I wondered how Metzinger can reconcile it with his claim that ‘no such things as selves exist in the world.’ Here he says that the system represents itself to itself by means of its PSM, and that it does so “correctly.” Metzinger certainly admits that systems exist. Are we not, then, such systems?
If the self-model models the system—whether accurately or not—then something exists of which it is a model. Metzinger’s self-model is a model of his system. Instead of denying that selves exist, why does Metzinger not conclude that selves are the systems so modelled? Why does he not make a case for conceptual reform?
“No Such Things as Selves Exist in the World”
Let’s examine Metzinger’s argument for his ‘main thesis,’ as given in the final section of Being No One.
Metzinger’s primary focus is on the phenomenology of the self.
What exactly does it mean to have the conscious experience of being someone? In this limited sense, the folk-phenomenological notion of “being someone” denotes a phenomenal property like many others, a property like the scent of mixed amber and sandalwood or the gustatory experience of cinnamon, a property like the emotional experience of elation, or the sense of surprise going along with a sudden cognitive insight. It is just a way of experiencing reality: currently, you are someone. What makes consciously experienced selfhood special, and different from all the other forms of experiential content, is the fact that—in non-pathological standard situations and in beings like ourselves—it is highly invariant. It is always there. (Metzinger 2004, p 626)
If the experience of being someone is ‘highly invariant,’ that may explain why it is so difficult to notice and characterize. If it is always part of my experience, and I cannot even imagine what my experience would be like if it were absent, then how can I attentionally isolate it from the rest of my experience? If I cannot isolate it, can I be sure I am aware of it? If I cannot be sure I am aware of it, how do I know it is there at all?
In the previous post I argued that the phenomenological markers of the self—the qualities of “me-ness”—are not much like the taste of cinnamon. They resemble emotional qualities more than sensory qualities. I recognize myself by my emotional dispositions—by the way I feel, affectively, towards myself. Those markers can be manipulated. The rubber-hand experiment demonstrates that I can be induced to feel that an inanimate object is actually part of my body.
The rubber-hand illusion can be used to drive a wedge between the experience of self—the quality of “me-ness”—and experience of the rest of reality. It allows me to compare the unusual experience of my self-model invading a stuffed dish-glove, to my ordinary experience in which the dish-glove appears to me like any other ordinary lifeless object. By paying attention to the difference, I can begin to isolate the quality of “me-ness.”
Metzinger suggests that the invariance of the experience of ‘being someone’ gives rise to intuitions about the self that have traditionally misled not only philosophers, but almost everyone else:
This phenomenally transparent representation of invariance and continuity constitutes the intuitions that underlie many traditional philosophical fallacies concerning the existence of selves as process-independent individual entities, as ontological substances that could in principle exist all by themselves, and as mysteriously unchanging essences that generate a sharp transtemporal identity for persons. But at the end of this investigation we can clearly see how individuality (in terms of simplicity and indivisibility), substantiality (in terms of ontological autonomy), and essentiality (in terms of transtemporal sameness) are not properties of selves at all. At best, they are folk-phenomenological constructs, inadequately described conscious simulations of individuality, substantiality, and essentiality. And in this sense we truly are no one. We now arrive at a maximally simple metaphysical position with regard to selves: No such things as selves exist in the world. At least their existence does not have to be presupposed in any rational and truly explanatory theory. Metaphysically speaking, what we called “the self” in the past is neither an individual nor a substance, but the content of a transparent PSM. There is no unchanging substance, but a complex self-representational process that can be interestingly described on many different levels of analysis at the same time. For ontological purposes, “self” can therefore be substituted by “PSM.” (Metzinger 2004, p 626)
Because there is constancy in our experience, we infer an unchanging cause: the ‘essential’ self, a substance which persists throughout our biological lives. A substance which is the real subject of questions about personal identity. The enduring substance which gives me a reason to look out for my own interests—a reason which applies to me alone—a prudential reason. There is no such enduring substance, says Metzinger. Our intuitions about the self are informed by a misleading experience. The self, as portrayed by a traditional folk-psychological theory, does not exist.
This conclusion is a bit of a let-down. It is as though, 600-odd pages into his book, Metzinger announced that there are no Cartesian egos. Metzinger seems to admit as much:
However, this first reading of the concept of “being no one” is only an answer to the crude traditional metaphysics of selfhood, and I think as such it is a rather trivial one. (Metzinger 2004, p 626)
What is interesting is not the conclusion that there is no such thing as selves (or persons) as traditionally conceived, but the new theory which Metzinger erects to replace the old one, and the explanation it offers as to why we have had, and continue to have, such a stubborn tendency to go wrong.
Metzinger proceeds to ask, “on a somewhat deeper level,” whether the structure of our experience “makes us constitutionally unable to see certain truths.” His question is aimed at the mind sciences, including cognitive neuroscience, AI, and philosophy of mind.
[C]ould it be that the conscious experience of being someone itself hinders growth of knowledge in these disciplines, by making certain theoretical positions or solutions of problems look utterly implausible, dangerously provocative, absurdly humiliating, or simply inconceivable to beings like ourselves? (Metzinger 2004, p 627)
Perhaps implausible, perhaps dangerously provocative. But inconceivable? Metzinger takes a strong position:
…you could never really believe that the SMT, the self-model theory of subjectivity, actually is true. You cannot believe in it. … “Being convinced,” like smelling mixed amber and sandalwood or being someone, is here interpreted as a phenomenal property. But for the current theory you cannot in principle have that property, because phenomenally simulating the truth of the SMT would involve a cognitively lucid, nonpathological way of dissolving your sense of self. It would involved being convinced and phenomenally being no one at the same time. (Metzinger 2004, p 627)
Isn’t there something wrong with Metzinger’s interpretation of “being convinced” as a phenomenal property? Belief is paradigmatically cognitive, not experiential. Is Metzinger not convinced of the truth of his own theory? On the evidence of the considerable efforts he has made to argue for it, I would conclude that he is.
What is Metzinger trying to say here? Clues lie in his thumbnail summary of the SMT, which fills the place of the ellipsis in the above quote:
Take what may be the central idea, the idea that metaphysically speaking no such things as selves exist in the world; that the conscious experience of selfhood is brought about by the phenomenal transparency of the system-model; and that what philosophers call the epistemic irreducibility of conscious experience—the fact that it is tied to a first-person perspective—can be exhaustively analyzed as a representational phenomenon, which in the future will likely be fully explained on functional and neurological levels of description.
Recall that, for Metzinger, the first-person perspective is primarily the subject-object relation (not, as might be supposed, the ‘centeredness’ or perspective properties of the experience of embodied organisms). To believe, or be convinced of, something, is to be a subject. But to believe something is not necessarily to believe that one believes it. Nor is it necessarily to have any sort of phenomenal experience of believing something. Metzinger himself provides the counter-example. He clearly believes that the SMT is true (his philosophical writings are ample evidence that he believes his stated theory); yet he claims to be constitutionally unable to have the corresponding phenomenal property of belief (a claim which I have no reason to doubt).
The above synopsis of the SMT states “that the conscious experience of selfhood is brought about by the phenomenal transparency of the system-model”. In my last post, I questioned Metzinger’s views that the self-model is necessarily transparent to itself, and that transparency of the model is what makes the subject of the model appear real to the system doing the modeling. I argued that transparency does not explain why our self-models feel as real and compelling as they do. A likelier explanation involves the emotional attachment which we have to the parts of reality which we call ourselves, an attachment strengthened by millions of years of natural selection.
Metzinger thinks the self-model is not only transparent, but necessarily so; and that is what makes it “inconceivable to beings like ourselves” that there is no real entity which the model represents. (For now, I’ll ignore Metzinger’s above-mentioned statement, which seems at odds with his main thesis, that the self-model is an accurate representation of the system to itself.) Is it true that we are constitutionally incapable of regarding the self-model as illusory—possibly as a kind of hallucination?
It’s worth pointing out that people are capable of learning that certain kinds of experience are illusory, misleading, or hallucinatory. Ramachandran describes a patient who was diagnosed with Charles Bonnet Syndrome after a head injury from an automobile accident. Larry’s world was filled with visual and auditory hallucinations. One day, while reporting symptoms of his disease to Ramachandran, Larry mentioned ‘seeing’ a monkey in the doctor’s lap. Ramachandran’s account of their conversation is telling:
“As I look at you, there is a monkey sitting on your lap,” Larry announced.
“Yes, right there on your lap.”
I thought he was joking. “Tell me how you know you’re hallucinating.”
“I don’t know. But it’s unlikely there would be a professor here with a monkey sitting in his lap so I think there probably isn’t one.” He smiled cheerfully. “But it looks extremely vivid and real.” I must have looked shocked, for Larry continued, “For one thing, they fade after a few seconds or minutes, so I know they’re not real. And even though the image sometimes blends quite well into the rest of the scene around it, like the monkey on your lap,” he continued, “I realize that it is highly improbable and usually don’t tell people about it.” Speechless, I glanced down at my lap while Larry just smiled. “Also, there is something odd about the images—they often look too good to be true. The colors are vibrant, extraordinarily vivid, and the images actually look more real than real objects, if you see what I mean.” (Ramachandran, 1998, p 107)
Before it starts to fade, the monkey appears real to Larry—even ultra-real. Larry’s hallucinations were sufficiently veridical that he often had trouble distinguishing them from real objects. When dressing in the morning, he had trouble finding his shoes in the closet because its floor was littered with hallucinated shoes. With respect to Metzinger, two points should be made about Larry’s experience of the monkey. The first is that “earlier processing stages” were not attentionally available. The monkey just appeared to him the way a real monkey sitting in Ramanchandran’s lap would have appeared, and the way the hallucinated shoes appeared on his closet floor. Therefore, attentional availability of earlier processing stages is not necessary for someone to be fully aware (as Larry was) that an aspect of his experience is hallucinatory. The second point is that Larry reached his awareness by means of a rational, cognitive process. He knew it was extremely improbable that a real monkey was sitting in Ramachandran’s lap; he knew well (five years after his accident) that he was subject to hallucinations; therefore he concluded that the monkey was a hallucination, and he treated it as such. Because Larry was wise to what was going on, the monkey hallucination was fully opaque. Functionally, Larry was able to isolate the monkey as a non-representational component of his experience.
The same thing can happen with respect to our experience of ourselves. When I experience the rubber hand illusion, I know the rubber hand is not my hand, although I feel that it is. Because I know it is not, I allow the experimenter to smash the rubber hand with a mallet. Even though I have a visceral, emotional involvement with the rubber hand, as a result of my body image having invaded it, my knowledge of the truth dissociates the concern I feel from the value I place on my real, biological hand. My real hand is useful to me, and I do not want it to be injured by a heavy blow from a mallet. Nor do I want the excruciating pain which would accompany such a blow. My cognitive belief that the rubber hand is not my real hand largely overcomes the odd emotional sense, induced by the experimental procedure, that the rubber hand is part of my body. The experimentally-induced emotion is enough to give me a twinge of panic when the mallet swings down on ‘my’ hand; it is enough to cause the muscles in my real arm to tense; but it is not enough to overcome my determination to leave my real arm, which I know is safe, where it is. The illusion is functionally, for me, just an illusion.
If I were cognitively persuaded that my entire self-model is based on the same kind of underlying mechanism as the rubber-hand illusion, could I begin functionally to treat the model as illusory? This is too complex a question for a simple answer. If I am right that my self-model, which includes the hand portion which invaded the rubber glove, is phenomenally grounded in my emotional attachment to what I call myself, and that this attachment is just something I produce, a product of my psychological and neurological makeup, something for which the best explanation is an evolutionary one, something that only contingently and partially coincides with my physical body, then perhaps I can learn to manipulate it with more facility than I now can. To the extent I could do so, my self-model would become opaque to me, in Metzinger’s sense; I would be treating it as a model rather than as the real entity which the model purports to represent. This is a more general case of what I do when I allow my wife to smash ‘my’ rubber hand with a mallet; and it has much in common with Larry’s tolerant and unexcited response to the vision of a live monkey in Ramachandran’s lap.
If I were persuaded that my self-model was illusory in that way, how might it change my life? Instead of unthinkingly trusting and obeying the impulses that arise from my emotional investment in my self-model, I would discount them—much as I discount the alarm I feel when I see the mallet swinging down on my rubber hand. I would realize that these emotional reponses are a product of natural selection, and that they may or may not coincide with my most deeply held, chosen values. Before acting on my personal fears and desires, I would try to evaluate them, again with reference to my independently held values. Those that I judged to be in conflict with my independently held values, I would discount further, suppress or ignore. I would do what I could to avoid acting on them—just as Edwin, in an earlier example, overcame his craving for poutine in the interests of healthy eating.
What I have just described is not unusual or exotic—it is what we all do when we engage in moral deliberation. All but the most self-indulgent among us sometimes curb the impulses of self-concern that urge us to do things that are contrary to our independently-held values. But there is a difference in scope. Moral deliberations usually involve a choice between two or more values: often between selfish and unselfish values. Both selfish and unselfish values are given weight by the deliberator. What is radical about the idea that the self is illusory—that there is no reason for self-concern—is that it has potential to discount the weight of selfish values to zero. It challenges us, when faced with moral dilemmas, to find reasons for acting in our personal interests that are independent of the values to which self-concern gives rise.
Why We Should Doubt our Intuitions about the Self
One of Metzinger’s most valuable insights is that biology—in particular, the structure of the self-model, which is a product of our biological heritage—largely determines our intuitions about the self. This remains true and important regardless of the fate of some of the details of Metzinger’s theory, which may not survive close scrutiny. And it highlights a difference between our intuitions about the self and intuitions about other less personal subjects. Metzinger contrasts his theory of mind to contemporary physics, which is also counter-intuitive.
Yet most of us believe that these theories are among the best mankind has so far created. Basically, we trust those physicists. In the mind sciences things are different, and in an interesting way. (Metzinger 2004, p 627)
We respect the achievements of quantum physics; but we live our daily lives as though we lived in a Newtonian world. At the scale of most human activity, the Newtonian model (which most people, these days, find intuitively plausible) is good and useful. Quantum physics does not threaten the engineering principles which we use to design bridges.
Is all of this a problem? Yes and no. It is a problem if—as opposed to other, for example, physical theories about the nature of reality—we impose the additional constraint of intuitive plausibility on theories of consciousness, the phenomenal self, and the first-person perspective. (Metzinger 2004, p 628)
Because selves and persons are central to our daily concerns, it is a problem if the truth is counter-intuitive. If we customarily speak and act in ways that are inconsistent with our most deeply considered beliefs, we lose integrity. Unlike quantum physics, which has only marginal implications for human action, our theories about the nature of the self and the justification for self-concern are closely coupled to our behaviour. This makes it important to get it right—to eliminate the problem by reconciling our intuitions with the truth.
Metzinger emphasizes that the self-model is a product of evolution:
Biological evolution is not something to be glorified. It is blind, driven by chance, and it has no mercy. … For millions of years, Mother Nature has talked to us, through our reward system and through the emotional layers of our PSM. We have to learn to take a critical stance toward this process, and to view our own phenomenal experience as a direct result of it. (Metzinger 2004, p 633)
In a footnote, Metzinger describes “the emotional self-model” as an evolutionary product “that has arisen from a fundamentally competitive situation.”
And the emotional self-model, including all its beautiful aspects, is what drives us. It is a virtual organ ultimately developed to spread genes more efficiently, and not a tool for maximizing self-knowledge. It may therefore make it difficult for us to grasp the true state of affairs, or put already existing insights into action. Any theory about consciousness and the phenomenal self that was maladaptive would immediately be intuitively implausible and emotionally unattractive. (Metzinger 2004, p 633n)
That, I think, is close to the heart of the matter. The truth about ourselves is not inconceivable, but to grasp it is an uphill battle. As in so many other aspects of civilized life, we must fight our biological heritage.
The Ethics of Being No One
Metzinger offers a third potential reading of “being no one” from an ethical perspective. He asks whether:
…in the long run, we want to use our new insights into the nature of consciousness, the phenomenal self, and the first-person perspective to change our own minds. Is it better to be someone or is it better to be no one? (Metzinger 2004, p 630)
For example, we might choose to decrease phenomenal transparency.
This normative option, which M calls “minimization of transparency”:
…would consist in making the fundamentally representational character of conscious experience globally available. We could attempt to make more information about earlier processing stages available for introspective attention, thereby also gradually making more and more layers in our own self-model phenomenally opaque. This type of strategy would certainly create an additional computational load for attentional systems in the brain, but it could at the same time serve to weaken the naïve-realistic self-misunderstanding characterizing our present state of consciousness. (Metzinger 2004, p 631)
I think the representational character of conscious experience is already globally available to most people—not always top of mind, certainly, but available on demand. And I’m not much drawn to the idea of “making more information about earlier processing stages available for introspective attention”. Metzinger is surely right that it would increase our computational loads, and to what end? That is what is meant by “navel-gazing.” The world has many urgent problems demanding our attention; I don’t want to spend mine admiring my “earlier processing stages.” And as the examples of Larry and the rubber hand show, we do not need to be aware of earlier processing stages in order to understand that an experience is illusory.
On the other hand, weakening “the naïve-realistic self-misunderstanding characterizing our present state of consciousness” could be very worthwhile, if it results in a change of motivation. If weakening that misunderstanding also weakens self-concern—and I think it does—that could form the basis of a very worthwhile moral reform of our species. Imagine what the world would be like if no one felt tempted to take a bribe. To take unfair advantage of others. To enrich himself at the cost of causing suffering to others. To sacrifice a long-term public good for a short-term private advantage. That possibility makes it worthwhile to persists in our investigations, until we achieve a thorough understanding of how our self-models really work. Metzinger’s valuable work takes us a good distance down that path. When we stop believing that we are the content of our self-models:
…a new dimension opens. At least in principle, one can wake up from one’s biological history. One can grow up, define one’s own goals, and become autonomous. And one can start talking back to Mother Nature, elevating her self-conversation to a new level. (Metzinger 2004, p 634)
Metzinger, T. (2004), Being No One. MIT Press.