Metzinger: Being No One

Being No One is a substantial work by German philosopher Thomas Metzinger about “consciousness, the phenomenal self, and the first-person perspective.”  Its main thesis “is that no such things as selves exist in the world.  Nobody ever was or had a self.”

I have spent some time with the book, making, from its 634 densely-printed pages, 104 pages of notes.  After all that, I still question Metzinger’s ‘main thesis.’  But I have no doubts about the value of the book.  It irrevocably raises the standard for what philosophy of mind must explain.  In its early pages, Metzinger echoes Paul Churchland’s complaint, that “theoretical approaches to the mental, still intuitively rooted in folk psychology, have generated very little growth of knowledge in the last twenty-five centuries.”  Being No One goes a long way towards burying that era.

Metzinger’s theory of mind centres on models.  The brain of a conscious organism supports a model of the real world it inhabits.  The reality-model of more advanced conscious organisms includes a model of the organism itself – a self-model.  On this central idea, Metzinger builds an wide-reaching theory with great explanatory power.  One of several important requirements Metzinger imposed on his theory is that it must explain more than just paradigmatic examples of consciousness – nonpathological waking states of human adults.  It must also be compatible with, and advance our understanding of, pathological neurological conditions – ones that give rise to psychological states that healthy people find hard to understand.  Metzinger’s theory also sheds scientific light on nonpathological, but non-standard varieties of consciousness and self-awareness such as are found in dreaming (including ‘lucid’ dreaming, in which the dreamer knows that he dreams), distortions of body image (such as phantom limbs, and the ‘rubber-hand illusion’), so-called ‘out-of-body experiences,’  and meditative states.

Being No One helps to explain why the traditional reliance on philosophical intuition has illuminated this subject so little.  Human intuitions about consciousness and the self are shaped by consciousness and the self-model, as they are experienced by normal human adults.  Both consciousness and the self-model are products of evolution, a natural process which favours gene reproduction.  Evolution has no fundamental bias towards the truth.  Although it is often adaptive for organisms to form accurate models of reality, it is by no means always so.  Robert Trivers has shown that self-deception can be adaptive (because it helps us to deceive others) ^{(Trivers, 2000)}; and so can ego-boosting misperceptions such as superiority bias.  Another highly adaptive evolutionary product is what Metzinger calls the Phenomenal Self-Model (PSM).  It has conferred such great advantages on our species, that Metzinger describes it as not only a tool, but (acknowledging Andy Clark for the metaphor):

…a weapon, developed in a cognitive arms race. Conscious selves are like instruments or abstract organs, invented and constantly optimized by biological systems.  ^{(Metzinger, 2004, p. 273)}

To describe the model as effective is not to say that it is accurate.  In fact, it is inaccurate in important respects.  We are not the sort of beings that, influenced by our own self-models, we intuitively think we are.  Discipline is required to overcome our biases, to arrive at truths about ourselves that are scientifically justified.

A strength of Metzinger’s methodology stems from rigorously framing his claims as claims about systems.  This discipline avoids the trap of relying on intuitions which only apply to minds, and have no straightforward application to artificial systems.  Given  the basic materialist assumption that minds are nothing over and above bodies and their states, it should become possible to create artificial systems that fully emulate natural ones.  To rely on premisses which could only expressed in terms of living systems would be to build on shaky ground.

Structure and Methodology

Chapters 2, 3 and 4 of Being No One are about consciousness; chapters 5 to 8 are about the self.  Metzinger holds that there can be consciousness without self-consciousness, and without the self-model.  Consciousness should not be thought of as all-or-nothing.  There are different levels of consciousness.  Some of the more advanced levels are in fact only found in organisms – or ‘systems,’ as we should get used to saying – in which a self-model is also found.

Throughout, Metzinger addresses his subject systematically through five different “levels of description:” phenomenological, representationalist, informational-computational, functionalist, and neurobiological.  He identifies eleven characteristics, or ‘constraints,’ of consciousness, and examines each one through the five levels of description.  This makes for a very thorough, systematic treatment; but the massive theoretical structure sometimes, I think, impedes progress.  Constant switching from one ‘level of description’ to another generates a flood of disparate ideas which tend to bog the reader down.   I can’t help thinking that the scientific study of the mind has not yet reached the level of maturity that warrants such systematization—this is not the periodic table of the elements.  As Metzinger admits, “Consciousness research is still in a preparadigmatic stage.” ^{(Metzinger, 2004, p. 116)} I’m fairly sure the mind sciences will undergo conceptual upheavals in which many of Metzinger’s categories will be revised and replaced.  But this first stab at it is valuable.  The book is full of insights.  Being No One throws light on aspects of the mind and the self that have been murky mysteries for too long.

In his opening chapter, Metzinger lists questions which he promises to address.  Most are questions about consciousness and the self that have been posed in philosophy classes and journals for a long time.  They include:

What does it mean to say of a mental state that it is conscious?  Alternatively, what does it mean of a conscious system—a person, a biological organism, or an artificial system—if taken as a whole to say that it is conscious?

What is phenomenal selfhood?  What, precisely, is the nonconceptual sense of ownership that goes along with the phenomenal experience of selfhood or of “being someone?”

Is artificial subjectivity possible?  Could there be nonbiological phenomenal selves?

These questions are repeated 500 pages later, near the end of Being No One, where Metzinger offers answers in light of, and with reference to, the theory laid out in the intervening chapters.  To the extent to which his theory succeeds in ridding readers of the blinkers of philosophical intuition, a fresh and illuminating perspective is gained.

Chapters 2 and 5 are devoted to developing conceptual tools.  Chapters 3 and 6 describe the “Representational Deep Structure” of  phenomenal experience and the first-person perspective, respectively.  Chapters 4 and 7 test the theory against non-standard cases, both pathological and nonpathological.  In chapter 4. Metzinger applies his theory of consciousness to agnosia, hemineglect, blindsight, hallucinations and dreams.  Chapter 7 explores deviant models of the self: anosognosia (deficit denial), identity disorders (including Cotard’s Syndrome, in which patients sincerely deny that they are alive, or even exist), hallucinated selves (including phantom limbs), “out-of-body experiences,” hallucinated agency, and dissociative identity disorder.  The degree to which Metzinger’s theory is able to explain these surprising, baffling, and frankly weird variants of human experience is a measure of its power.

Another laudable goal Metzinger adopts for his toolset is that it should be usable across disciplines, not just in philosophy.  The contributions of psychology, neuroscience, and computational science are also needed in order to understand the self and the conscious mind.  For the various mind-sciences to benefit from one another’s work, they must employ a shared set of concepts expressed in a common language.

Being No One is not for everyone.  Its rewards come only with a considerable expenditure of effort.  For readers with less time, or a more casual interest in the subject, there is an alternative.  Metzinger’s The Ego Tunnel is a popularization of the subject which covers the ground more readably.  If you are not sure you want to take on all the detail of Metzinger’s Big Theory, try The Ego Tunnel first.

Consciousness

Natural organisms and artificial systems can represent both their own states and properties of the external world.  Metzinger distinguishes between conscious and unconscious representation.  Unconscious representations include the informational content of self-regulatory biological processes, such as those that maintain heart rate and fight off infections.

The large majority of cases in which properties of the world are represented by generating specific internal states, in principle, take place without any instantiation of phenomenal properties or subjective awareness. ^{(Metzinger, 2004, p. 17)}

Metzinger distinguishes phenomenal from representational content. Representation is a three-place relation that holds between the system doing the representing, the state of affairs that is represented, and the system state that is the representation. What the latter actually represents, if anything, depends on what actually exists in the environment.  Right now my visual experience includes a representation of the blue coffee cup on my desk.  If I had the same experience and there were no cup on the desk, my experience would not represent anything.

Phenomenal experience, by contrast, is a two-place relation between the system and its phenomenal content.    Phenomenal content is “determined by internal…properties of the conscious system.”  ^{(Metzinger, 2004, p. 13n)} The phenomenal content of my present experience exists regardless of whether I am seeing a real coffee cup or having a visual hallucination.

In Chapter 3, “The Representational Deep Structure of Phenomenal Experience,” Metzinger offers five “levels of description” that are important for understanding consciousness.

• Phenomenological descriptions are based on experience accessed introspectively.
• Representationalist descriptions portray a three-place relation between the system, some aspect of the world, and the system’s internal representation of that aspect of the world.
• Informational-computational descriptions are concerned with computational goals and processing efficiency.
• Functional descriptions of subjective experience are about the causal roles which the neural correlates of consciousness must play, in order to generate subjective experience.
• Physical-neurobiological descriptions.  For biological systems, these are descriptions of the neural correlates of conscious experience.   For artificial systems, they are physical descriptions of processes that satisfy the constraints of conscious processes.

Metzinger proceeds to outline eleven criteria, which he calls “constraints,” that will “allow us to decide if a certain representational state may also be a conscious state.” ^{(Metzinger, 2004, p. 117)} Of the eleven, only two or three are necessary conditions for consciousness as minimally defined.  The others, when added to the necessary conditions, characterize specific types of consciousness – what can be described as higher forms of consciousness.  That is, consciousness comes in degrees; alternatively, consciousness is a ‘cluster concept.’  Metzinger regards the list of constraints as a preliminary one, which he expects “to be continuously enriched and updated by new empirical discoveries.”

Constraint 1, Global Availability is primarily a functional constraint.  Information is consciously represented in a system if and only if it is “globally available for deliberately guided attention, cognitive reference, and control of action.”  Global availability provides for the maxmum flexibility in behavioural response.  The contents of consciousness coexist in a common arena – the system’s world-model – where the system can use them to devise an appropriate response.

Constraint 2, Activation Within a Window of Presence, is primarily phenomenological rather than functional.  It is arguably “the most general and the strongest candidate” for a necessary condition for first-person ascriptions of phenomenality.

Without exception, it is true of all my phenomenal states, because whatever I experience, I always experience it now.   ^{(Metzinger, 2004, p. 126)}

On the face of it, this doesn’t seem like much of a constraint.  What would my experience be like if it were false?  Like nothing at all, says Metzinger.

If we subtract the global characteristic of presence from the phenomenal world-model, then we simply subtract its existence.

But there’s more to it: the “window of presence” is not an infinitely thin time-slice, but a specious present which includes events taking place within a “moving window”.

A psychological moment is not an extensionless point, but for beings like us it possesses a culturally invariant duration of, maximally, three seconds. ^{(Metzinger, 2004, p 127)}

One functional benefit of integration within this short time frame is to allow us to perceive motion of objects in our environment.  Metzinger describes its role in shaping our experience of simultaneity and sequence of events.

Constraint 3, Integration into a Global Coherent State, is closely linked to Constraint 1, Global Availability; Metzinger describes the latter as a “possible functionalist interpretation” of the former, which emphasizes phenomenology more than function, referring to the experience of being in a world.

If and only if a person is conscious, a world exists for her, and if and only if she is conscious can she make the fact of actually living in a world available for herself, cognitively and as an agent.  ^{(Metzinger, 2004,  p. 131)}

To elucidate this, Metzinger introduces the concept of “autoepistemic closure,” which he defines as being:

…constituted by (a) the existence of a comprehensive representation of reality as a whole, and (b) the fact that this representation cannot be represented as a representation by the system itself (this, as we will soon see, is an epistemological reading of constraint 7, the transparency of phenomenal representata.)  ^{(Metzinger, 2004, p. 131)}

Transparency (constraint 7) is a key concept.  Our world-models are transparent, because we look “through” them.  Right now, I have the experience of typing on a keyboard.  I am not aware of my mental model of the keyboard.  Mostly, we are conscious of real objects and events in the world with which we interact; only rarely are we conscious of our conscious experience as such.

Metzinger defines phenomenal transparency this way:

For any phenomenal state, the degree of phenomenal transparency is inversely proportional to the introspective degree of attentional availability of earlier processing stages. ^{(Metzinger, 2004,  p. 165)}

If the process by which the phenomenal state came about is concealed, we ‘look right through’ it.  “We don’t see the window, but only the bird flying by.”

The instruments of representation themselves cannot be represented as such anymore, and hence the experiencing system, by necessity, is entangled in a naive realism.  … What is inaccessible to conscious experience is the simple fact of this experience taking place in a medium.  Therefore, transparency of phenomenal content leads to a further characteristic of conscious experience, namely, the subjective experience of immediacy. ^{(Metzinger, 2004,  pp. 169-170)}

Transparency “minimizes computational load,” and “creates the most important ‘architectural’ precondition for planning processes.”  Transparency helps keep us alive by making the reality of reality salient.

Conscious experience, for a biological system, generates a simply structured user surface for its own nervous system.  Naïve realism…is an [evolutionary] advantage.  It only confronts the system with the final results of its own processing activity, and it does so by making them available for the control of action while simultaneously protecting the system from losing contact with external reality by getting lost in an introspective exploration of the underlying mechanisms.  ^{(Metzinger, 2004, p. 175)}

When walking in the woods, I don’t become aware of a bear-experience; I see or hear the bear.  (And my response is well underway by the time I see it consciously.) One functional advantage of transparency is to make the experiencing system a realist.  “The assumption of the actual presence of a world becomes causally effective.”

Not all phenomenal states are transparent; some are opaque.  We experience them as internal states.  Opacity brings a second functional advantage: “The appearance-reality distinction can now be represented.”  ^{(Metzinger, 2004,  p. 176)} This allows the system to evaluate its own representations and flag some of them as non-veridical: misleading, illusory, or outright hallucinatory.  The system acquires the ability to overcome first impressions in order to correct its model of reality.

Global availability/integration, the window of presence, and transparency are the central attributes of consciousness, according to Metzinger.  Since some phenomenal states are opaque, transparency is not a strictly necessary attribute of conscious experience; the other two are.  Metzinger describes contraints 2 and 3 as “necessary conditions for any kind of conscious experience”,  which together define:

…the weakest possible notion of consciousness: consciousness is the activation of an integrated model of the world within a window of presence. ^{(Metzinger, 2004,  p. 136)}

Elsewhere Metzinger describes constraints 1, 2 and 7 as the only ones that can “count as candidates for necessary conditions in the ascription of conscious experience.”  ^{(Metzinger, 2004,  p. 126)} The apparent inconsistency can probably be attributed to the close coupling between constraints 1 and 3, and to the central role that Metzinger attributes to transparency.  I think he would agree that although transparency is not an attribute of every conscious experience, it is a necessary attribute of consciousness,  even a minimal level of consciousness.  In other words, every conscious system has some transparent experience – which is to say that it experiences something as real.

I have trouble with the way Metzinger uses “transparent” and “opaque.”  The terms are mutually exclusive: one object cannot be both transparent and opaque.  But in the case of phenomenal experience, both attributes can and frequently do apply.  I look around for the phone; I see it, on top of a copy of Metzinger’s The Ego Tunnel.  I see through my visual sensation of the phone, to the phone itself.  But I am simultaneously aware – when I care to be – of my visual sensation, as visual sensation.  At any time, I can judge how clear and sharp my internal image is, or is not.  I can notice the perspective of the image, my viewpoint, and imaginatively project it onto a two-dimensional surface, as though I were framing a photograph or planning to paint a picture of the phone.  Human beings are capable of simultaneously seeing the bird through the window and looking at the window.  I will return to this issue in my next post.

Metzinger’s other constraints – convolved holism, dynamicity, perspectivalness, offline activation, representation of intensities, and “ultrasmoothness” – characterize normal human consciousness, but not all consciousness; and not all non-standard human conscious states (pathological or otherwise).  Constraint 11, Adaptivity, characterizes most conscious functions of living organisms.  It is the survival and reproductive advantage that consciousness confers on species that have it.  Looking at consciousness from the point of view of evolutionary function is a recurring source of illumination in Metzinger’s work.

From Consciousness to Selfhood

Two new concepts dominate the second half of the book: the Phenomenal Self-Model (PSM) and the Phenomenal Model of the Intentionality Relation (PMIR).  The difference between a non-person and a person (or ‘self’), according to Metzinger is:

…a very special sort of PSM, plus a PMIR:  You become a person by possessing a transparent self-model plus a conscious model of the “arrow of intentionality” linking you to the world. ^{(Metzinger, 2004, p. 10)}

The PSM is the self-model, which is a proper part of the world-model.  The existence of a PSM gives rise to the important self-other distinction, and demarks a target for the emotions of self-concern.  Metzinger describes the PSM as “the central necessary condition for a conscious first-person perspective to emerge.”

The content of the PSM is the content of your conscious self: your current bodily sensations, your present emotional situation, plus all the contents of your phenomenally experienced cognitive processing.  They are the constituents of your PSM.  Intuitively, and in a certain metaphorical sense, one could even say that you are the content of your PSM. … Your self-directed thoughts operate on the current contents of your PSM: they cannot operate on anything else. ^{(Metzinger, 2004, p. 299)}

Having a PSM is also necessary for self-directed agency.

If you want to initiate a goal-directed action aimed at some aspect of yourself – for example, brushing your hair or shaving yourself – you need a conscious self-model to deliberately initiate these actions.

That passage can be read in more than one way.  I take it that the subject of “deliberately initiate” is “you”, not “a conscious self-model.”  My self-model doesn’t initiate actions; I do.  But why does Metzinger refer to the “conscious self-model”?   It seems wrong to describe the model as conscious.  Perhaps the intended reading is that I can only perform a self-directed action if my conscious experience includes a self-model.

Metzinger draws the distinction between emulation and simulation.  Software used for meteorological forecasting simulates, but does not emulate, the weather.  A Macintosh computer running Virtual PC software emulates a Windows computer.  Simulation models the behaviour of another system, but emulation replicates behaviour.  Metzinger describes the PSM as self-emulation. ^{(Metzinger, 2004, p. 300)} What does this mean?

Self-modeling is that special case, in which the target system and the simulating-emulating system are identical: A self-modeling information-processing system internally and continuously simulates its own observable output as well as it emulates abstract properties of its own internal information processing—and it does so for itself.  Using a term from computer science, we could therefore metaphorically describe self-modeling in a conscious agent as “internal user modeling.”  ^{(Metzinger, 2004, p. 301)}

The ability to self-model goes hand-in-glove with the extremely valuable social ability to model the internal states of others.  We understand other human beings as subjects of experience, and as agents; in our dealings with others, it is often crucial to be able to model their cognitive, emotional, and motivational states.  If we could not do that, we could not predict their behaviour with any assurance; we would inhabit the puzzling, haphazard world of autism.

The PSM of a normal human adult contains both transparent and opaque elements.  Part of the PSM is the body-image, which contains a variety of data types – visceral, proprioceptive, sensory, appetitive, all integrated into “a supramodel, conscious body image”.  The body-image is normally transparent; we experience the body as real.  But parts of the body-image can become opaque, as when the amputee becomes aware of his phantom limb as a phantom, or when my body-image invades an object which is obviously not part of my body, as in the “rubber-hand experiment” described by Ramachandran and others.  Other contents of the PSM that represent mental processes such as memories, consciously imagined plans, fantasies, daydreams, and some thoughts, are experienced as phenomenally opaque; we know that they are mental phenomena, not part of an independent world.   That is true if we are sane; in some pathological conditions, human beings mistake such phenomena for external reality.

Metzinger offers an intriguing account of our deep-felt conviction of persisting through time.  Although the “coherence and duration of the phenomenal self” is regularly interrupted by deep and dream sleep:

It is the invariance of bodily self-awareness, of agency, and autobiographical memory which constitutes the conscious experience of an enduring self. The conceptual reification of what actually is a very unstable and episodic process is then reiterated by the phenomenological fallacy pervading almost all folk-psychological and a large portion of philosophical discourse on self-consciousness.  ^{(Metzinger, 2004, p. 325)}

Of the eleven contraints, transparency, Metzinger claims, is key to understanding how a  self-model comes to be.  Many machines today have system models which are not self-models.  Applying the transparency constraint “to the concept of a conscious self-model is the decisive step in understanding how…the conscious experience of selfhood …can be reductively explained….” ^{(Metzinger, 2004, p. 330)} We’ll return to this central topic in the next post.

Selfhood is comprised of the PSM and another model, which Metzinger calls the Phenomenal Model of the Intentionality Relation (PMIR).  He characterizes the PMIR as a conscious mental model of  “an ongoing, episodic subject-object relation.”  Examples include (1) “I am someone currently hearing the sound of the refrigerator behind me.”  (2) “I am someone currently grasping the content of the sentence I am reading.”  The PMIR is another subject I must defer to the next post for fuller discussion.

No such things as selves exist in the world.

Metzinger’s ‘main thesis’ remained puzzling to me until the last ten pages of the book.  On its face, it seems an overstatement.  Metzinger admits the existence of the world-model, and of the objective world it models.  He also admits the existence of the self-model, so why not of the self?  The self-model includes a bodily self-image; Metzinger certainly does not deny that human bodies exist.  Getting clear on Metzinger’s claim is another large topic which I must defer.

Needless to say, it is a topic that interests me.  Throughout the Phantom Self project, I’ve been looking for solid grounds on which to base the claim that self-concern is not rationally justified.  If there is no such thing as a self, that would be grounds enough.  I am also attracted by the idea that the self is a model.  A model is an informational entity; Metzinger’s theory is compatible with my own view that persons are informational entities who tend to mistake themselves for biological organisms.  Metzinger’s theory of the PSM may help ‘flesh out’ these ideas – so to speak.

References

Metzinger, T. (2004),  Being No One

Metzinger, T. (2009),  The Ego Tunnel

Trivers, R. (2000), The Elements of a Scientific Theory of Self-Deception.  Annals of the New York Academy of Sciences.

Return to the Phantom Self home page.