The Phantom Self

Parfit’s Glass Tunnel

Dali Phantom selfIn the Introduction to this project, I said:

It’s the strong – and I believe, irrational – hold that the idea of the self has over us, and particularly its role in motivating action, that led me to characterize it as the ‘phantom self’.  Like the Phantom of the Opera, the self has a powerful voice that demands to be obeyed.  Like an amputee’s phantom limb, it is a vividly felt presence – but there is nothing really there.

It is time to flesh out that characterization.

No contemporary philosopher – perhaps no philosopher ever, in the West – has done more to break the phantom’s grip than Derek Parfit.  In Reasons and Persons, Parfit argues persuasively that, although we are strongly inclined to believe that our continued existence is “a deep further fact, distinct from physical and psychological continuity”, that belief is not true.  He goes on to describe the difference this philosophical conclusion made to his own life.

Is the truth depressing?  Some may find it so.  But I find it liberating, and consoling.  When I believed that my existence was such a further fact, I seemed imprisoned in myself.  My life seemed like a glass tunnel, through which I was moving faster every year, and at the end of which there was darkness.  When I changed my view, the walls of my glass tunnel disappeared.  I now live in the open air.  There is still a difference between my life and the lives of other people.  But the difference is less.  I am less concerned about the rest of my own life, and more concerned about the lives of others. (RP p 281)

Oppressive imagery of one’s progress through life, like the glass tunnel, seems to be a mainstay of human experience.  Examples are not hard to come by.  A quick internet search yields:

“Although I’m only 23 I dread getting old.”

“I was around sick people my whole life, and I can’t deal with the loss of independence and begging people to do everything for me.  I gotta stop–I am scaring myself.”

“Can we ever really find a true kinship with a group of other people, or in another person find a ‘soul mate’? We all have our problems, issues, demons, regrets, insecurities, so can anyone ever really break past these and reach our essential being? If another person ever does do this, are we giving ourselves over to them and hence losing a part of ourselves? We are born alone and we die alone.”

“A fellow in a TV program about marriage last night made a point that struck me as quite profound.  He pointed out that we are all basically alone in our journey through this life.  He said, ‘We are born alone.  We die alone.  And we are alone most of the time in between.’”

“I’m looking at you through the glass…
Don’t know how much time has passed
And all I know is that it feels like forever
When no one ever tells you that forever
Feels like home, sitting all alone inside your head”
– “Through Glass” lyrics by Stone Sour

In an earlier post, I quoted Ian Brown’s eloquent expression of his own feelings:

Looking at it from the age of 55…getting older looks like a discouraging journey into loneliness.  … I can’t imagine getting older, therefore weaker and lonelier, without resenting it.  The slightest health scare makes me anxious and the anxiety makes me cranky and the crankiness makes me feel bitter, even mistreated.

It appears from these comments that when people think about their lives, especially about their futures, they often feel afraid, discouraged, lonely, ‘cut off’.  I doubt that is true for everyone; and for most of us it is not true all the time.  But it is a significant, recurring theme.  Why?  Is Parfit right that such feelings are caused by believing that our continued existence is a ‘deep further fact’?

The ‘deep further fact’ is that same elusive relationship which is thought to provide a reason for self-concern.   But if there is no compelling reason for self-concern, and what’s more, self-concern is more unpleasant than pleasant, as the above reports suggest, then self-concern may be bad for us.  This is doubly true if self-concern is not only unpleasant but has other bad effects, motivating us to do things that run counter to other deeply-held values.  If so, then we may have good reason to weaken our self-concern if we can.

How would we go about weakening self-concern?  Developing a clear understanding that there is no rational justification for self-concern is a first step.  If we achieve that, we can choose more freely between selfish values and other values with which the selfish values compete.  If we choose non-selfish values, we can then overcome the emotions of self-concern by applying the methods of self-discipline and self-guided maturation that we use to combat other emotions, urges, and addictions which hinder us from achieving the goals we value most deeply.

Phantom Anatomy – the Psychology of Self-Concern

I have been reading psychology and neuroscience in an effort to better understand the phenomenon of self-concern.   There is evidence that a number of different abilities related to the distinction we make between what is, and is not, ourselves, appear at different developmental stages.

Animals that move are able to distinguish their own bodies from the rest of the world  they move through.  This ability may be grounded in proprioception – the ‘inner sense’ that allows us to feel attitudes and movements of our body.  A study by John Watson indicates that people have a built-in capacity to correlate seen body movements with felt ones, and through that correlation, to distinguish self from other.   Watson’s work shows that an infant has a different level of interest in live video images of his own legs kicking than in similar images of another baby’s legs.  (Babies older than three months pay more attention to images of legs that are not their own – perhaps to avoid being kicked.)

This innate tendency to notice correlations across sensory modalities could be the origin of the basic body awareness I mentioned in the previous post.  Michael Lewis claims that the ability to differentiate oneself from everything that is not oneself comes early in life, “certainly by three months, and most likely from birth”.  It is not a distinctly human capability. “The rat does not run into a wall but knows to run around it.”

Self-awareness and self-concern undoubtedly depend on a robust body-image and the ability to distinguish between self and other.  But there must be more to it, because those attributes are characteristic of many animals which show no sign of self-awareness.  Mirror neurons, which are widely studied in primates, may be another piece of the puzzle.   Ramachandran emphasizes their importance in human development.

Stick your tongue out at a newborn baby and the baby will stick its tongue out too….  To do this it must create an internal model of your action and then re-enact it in its own brain.  An astonishing ability, given that it cannot even see its own tongue, and so must match the visual appearance of your tongue with the felt position of its own.  We now know that this is carried out by a specific group of neurons, in the frontal lobes, called the mirror neurons. (BTHC, p 106)

Notice that Ramachandran is talking about newborns!  His remarks on creating an internal model and re-enacting it should not be thought of as implying that the baby has any intellectual concept of itself or the adult face-maker.  Mirroring behaviour is a basic, innate ability which is thought to underly our capacity to learn by imitation and interpret the behaviour of` others.  Ramachandran writes:

Mirror neurons…permit a sort of “virtual reality” simulation of other people’s actions and intentions, which would explain why we humans are the “Machiavellian” primate – so good a constructing a “theory of other minds” in order to predict their behaviour.  This is indispensable for sophisticated social interactions, and some of our recent studies have show that this system may be flawed in autistic children, which would explain their extreme social awkwardness. (BTHC pp 38-9)

Mirror neurons are also found in macaques, which do not show evidence of human-like understanding of the mental states of others, or of self-consciousness.  Mirror neurons may be necessary, but are presumably not sufficient, for self-awareness and self-concern.

A more sophisticated ability than innate mirroring behaviour is mirror self-recognition,  which is found only in humans and a few other ‘higher’ mammals.  It is manifested in human babies at fifteen to eighteen months of age.

Gordon Gallup developed the standard test for mirror self-recognition, applying a visible mark to a part of the body which the experimental subject cannot see, then introducing a mirror which makes the mark visible to the subject in its reflection.  An animal passes the mirror self-recognition test if it clearly demonstrates a new awareness of the mark on its own body, for example, by scrutinizing the marked area in the mirror, then touching the mark on its own body, trying to rub it off, etc.  Controls involving marked subjects without mirrors, unmarked subjects with mirrors, and subjects ‘marked’ with invisible substances, both with and without mirrors, are used to eliminate competing explanations of the mark-oriented behaviour.

Mirror self-recognition has only been foundd in a handful of species: humans, orangutans, chimpanzees, bonobos, and probably dolphins.  Even gorillas show only weak evidence of mirror self-recognition, with the exception of gorillas that have been taught sign language.  Most monkey species do not have it, nor do dogs or cats.  Recent experiments with Indian elephants make them a possible candidate for membership in this exclusive club.  Species that display mirror self-recognition have large brains, undergo long periods of developmental immaturity, and are highly developed socially.

Mirror self-recognition does not require the subject to understand how mirrors work.  Butterworth, cited by Lewis, says:

There is ample evidence that although children are able to produce self-referential behaviour through the use of the mirror mark technique, they do not know many of the properties of reflected surfaces; for example, they cannot use the mirror to find an object reflected in its surface. (SAAH p 27)

Conversely, the ability to locate objects in mirrors can be demonstrated in species which do not exhibit mirror self-recognition.  Gergely cites studies showing that rhesus macaques and elephants can use mirror-guided search effectively, but do not recognize themselves in mirrors.  Lewis says, “Self-referential behaviour in mirrors is related to the idea of me and not to any elaborate knowledge of reflective surfaces.”  “The idea of me”, for Lewis, is objective self-awareness, or consciousness of oneself as an entity among others in the world.  In recognizing herself in a mirror, a child looks at a visual image and thinks, ‘That’s me.’  Mitchell suggests that mirror self-recognition appears at the same stage of a child’s development in which he begins to imagine representations of himself.  The ability to picture oneself in imagination is key to being able to plan effectively for the future and to learn from one’s memories.

‘The idea of me’, in turn, appears closely correlated with the rest of the gestalt of self-concern.  According to Lewis, several new emotions appear around the time children start to use the personal pronouns “me” and “mine”.  He says:

Although the child exhibits the primary emotions including joy, anger, sadness, interest, disgust, and fear, it is not until the acquisition of self-awareness that the child acquires such emotions as embarrassment, envy, empathy, pride, guilt, and shame. … These latter emotions, often called secondary emotions…should be relabeled as self-conscious emotions. (SAAH p 31)

Embarrassment, envy, empathy, pride, guilt and shame: the emotions of self-consciousness.  I suggest that anticipation (eager or otherwise) of future experience should be added to this list.  Self-consciousness is clearly necessary for contemplating how life may be for oneself in the future.  I look forward to eating lunch, and to reading Logicomix.  I do not look forward to driving out to the UBC library through the winter rain and Vancouver traffic; I have distaste for this prospect, and am inclined to procrastinate.

Anticipation is something like memory.  Both involve mental images.  Both may be pleasant or painful.  But anticipation of future experience is directly motivational in a way that memories are not.  To look forward to an event is to be motivated to do the things that will bring that event about.  I am also motivated to avoid, or delay, doing things that will cause experiences I expect to dislike.  Anticipation of future experiences directly engages my planning and preparatory functions in a way that memories do not.

We have anticipatory feelings about future events which we are certain will happen, and about events which may or may not happen.  In the latter category are outcomes over which we have complete control, outcomes over which we have no control, and outcomes we can influence (or think we can) but cannot fully control.

If I know that something will happen, and I have no control over it, I may nevertheless have strong feelings about it.  If it’s dreadful, I dread it; if it’s delightful, I look forward to it eagerly.  Such emotions seem counterproductive.  If I know something is inevitable, it serves no purpose to dread or long for it.  Insofar as emotions are motivational,  it seems they should be reserved for outcomes over which I have at least some control.  But of course, we have emotions about the past too, and we can do nothing to change the past.  We can only make amends, or apologize, or boast about our glory days. Do these examples show that the function of emotion is not just motivational?   Is there an evolutionary explanation for our feelings towards things we cannot change?

It may be that emotions are simply blunt instruments.  Emotions evolved because they have survival value; but evolution has not (yet) refined our emotional responses to distinguish between things we can change and things we cannot.  In any case, future outcomes are rarely certain; and that may explain why, as Samuel Johnson observed, the prospect of being hanged “concentrates the mind wonderfully”.  We are motivated to bend our imaginations to exhaust all avenues which hold out even the slimmest prospect of escape from impending death.

When trying to understand our feelings towards remembered events, it is helpful to consider what memory is useful for.  Memory certainly aids us in planning for the future.  Those who do not remember the past are doomed to repeat it.  A painful memory furnishes an example of similar future situations to be avoided.

Because we are self-conscious – each of us has the idea of herself as an entity among other entities in the world – we can engage effectively with the future and the past.  The emotional affect we automatically attach to images of ourselves strongly influences our behaviour.  To believe that an imagined person is oneself is – for most of us, most of the time – to have an earnest and powerful desire for that person’s success and happiness.  Not only do we desire it, we feel we have no choice; the act of identifying with someone makes it seem rationally compulsory to value that person’s well-being.  My belief that I really am the person I imagine in the future, and that therefore I need to care about that person, seems almost as entrenched and intractable as my body-image. 

We can learn from amputees about the intractability of body-image.  Most amputees retain a vivid proprioceptive sense of their missing limb, often with associated pain.  This is the well-known ‘phantom limb’ phenomenon.  Phantom limbs persist despite the amputee’s clear awareness, and in the face of constant visual evidence, that the limb is missing.  You cannot banish the illusion of a phantom arm by looking at your stump and telling yourself that the arm does not exist.  Paradoxically, Ramachandran has succeeded in relieving phantom arm syndrome by providing the amputee with the experience of ‘seeing’ his missing arm (a mirror reflection of his good arm) and ‘moving’ it – not only is the phantom arm’s pain and cramp instantly relieved, but in some cases, the phantom actually vanishes after several treatments!  Ramachandran’s ‘trick with mirrors’ accomplishes what intellectual comprehension and visual evidence fail to achieve – correcting the amputee’s body-image to conform to reality.

There is more to the phantom-limb story.  Amputees who wear prosthetic limbs actually find the phantom limb useful, even indispensable.  Oliver Sacks quotes Michael Kremer as follows:

[The phantom’s] value to the amputee is enormous.  I am quite certain that no amputee with an artificial lower limb can walk on it satisfactorily until the body-image, in other words the phantom, is incorporated into it. (MWMHWFH p 64)

Some amputees suffer phantom pain and spasm, such as toes painfully curled, only when the prosthetic leg is off or not being used; they can relieve these symptoms by attaching the leg, getting up and walking.  The prosthetic limb is used most naturally and successfully when the phantom limb conforms exactly to the artificial one, like a hand in a glove.

Neuroscience shows that, despite its intractability to conscious manipulation, body-image is plastic.  In addition to the phantom limb cases, Ramachandran describes simple, amusing experiments which result in distortions of body image.

Consider the ‘phantom nose’ illusion that we recently discovered in our laboratory (Ramachandran & Hirstein 1997). The subject sits in a chair blindfolded, with an accomplice sitting in front of him, facing the same direction. The experimenter then stands near the subject, and with his left hand takes hold of the subject’s left index finger and uses it to tap and stroke the nose of the accomplice repeatedly and randomly, while at the same time, using his right hand, he taps and strokes the subject’s nose in precisely the same manner, and in perfect synchrony. After a few seconds of this procedure, the subject develops the uncanny illusion that his nose has either been dislocated, or has been stretched out several feet forwards, demonstrating the striking plasticity or malleability of our body image. The more random and unpredictable the tapping sequence the more striking the illusion. We suggest that the subject’s brain regards it as highly improbable that the tapping sequence on his finger and the one on his nose are identical simply by chance and therefore `assumes’ that the nose has been displaced – applying  a universal Bayesian logic that is common to all sensory systems (Ramachandran & Hirstein 1997). The illusion is a very striking one, and we were able to replicate it in 12 out of 18 naive subjects. (TRS1998 p 1855)

A similar experiment will cause a normal subject to experience a rubber hand (obviously not attached to his body) or even a tabletop or shoe, as part of his own body.  Ramachandran devised a clever test to verify that the subjective reports of this experience were not overstated:

But how can we be sure that the subjects are not simply using a figure of speech when they say `I feel that the sensations are arising from the shoe’? To rule out this possibility, we waited until the subjects started `projecting’ their sensations onto the shoe and then simply hit the shoe with a giant rubber hammer as they watched. Remarkably, the subjects not only winced visibly but also registered a strong increase in skin conductance when we measured their galvanic skin response. (TRS1998 p 1855)

This, by the way, also exemplifies our strong, automatic, emotional response to a threat to self.

The Glass Tunnel Explained?

Does the psychological account of self-concern help to explain the feelings of isolation, of loneliness, being ‘cut off’, which Parfit experienced when he believed his identity was a ‘further fact’, and which are echoed by many others?  Maybe – but this is speculative.  The association between self-image and the emotions of self-concern gives a special quality to any mental representation of oneself, which is absent from our thoughts and experiences of other people.  If we feel a special kind of connection to ourselves, we thereby feel less well connected to others.  This could be experienced as being ‘cut off’, or isolated.

Consider Capgras syndrome, a rare condition in which patients with specific brain damage are convinced that their loved ones are imposters. The deficit is very limited.  Capgras patients are not ‘crazy’.  Ramachandran reports:

The remarkable thing about these patients is that they are relatively intact in other respects: they are mentally lucid, their memory is normal, and other aspects of their visual perception are completely unaffected. (TRS1998 p 1856)

Ramachandran’s theory of what is going on is interesting.  The amygdala assesses the emotional significance of sensory input and transmits this information to other limbic systems where these emotions are ‘experienced’.  Ramachandran surmises that the face-recognition area of the temporal lobes and the amygdala have been disconnected in these patients.

When he looks at his mother…even though he realizes that she resembles his mother, he does not experience the appropriate warmth, and therefore says ‘…This must be some strange person’. (TRS1998 p 1856)

The Capgras effect even extends to the patient’s experience of himself.

On one occasion when shown an old photograph of himself he said it was ‘a different person . . . see, he has a moustache and I don’t’. Or sometimes, during conversation, he would refer to ‘the other David’ (he once accused his employer of sending the cheque to the other David). (TRS1998 p 1857)

When DS is shown a photograph of himself, he is convinced that it is an imposter.  On Ramachandran’s theory, that is because the face-recognition centres in his brain do not trigger the normal emotional response DS expects if he is looking at himself.

Capgras syndrome is specific to vision.  When DS speaks to his mother on the telephone, he recognizes her unfailingly.  Ramachandran points out that this is strong evidence that the Capgras phenomenon is not a psychological problem demanding a psychoanalytic explanation.  DS is not alienated from the idea of his mother, only from the sight of her.

It is startling that the sense of strangeness when his mother is in the room continues even when she speaks to him.  The strangeness arising from the sight of his mother overcomes the familiarity of her voice, that he readily recognizes over the telephone.

If Ramachandran’s theory is correct, the Capgras cases show that our beliefs about personal identity are profoundly influenced by the way we respond emotionally to experience.  Emotional response, it seems, can easily trump evidence and reason.  As a general statement, this is not surprising, when we consider the tenuous grounds people have for many strongly held opinions.  Yet it remains surprising in this case, because it is very unlike what most of us have experienced.

I dwell on the Capgras cases because they point to a neurological underpinning for the felt difference between familiarity and strangeness.  Regarding the feelings of alienation from others – even from close loved ones – reported so commonly by ‘ordinary people’, I can offer only a speculative theory.  Self-concern is primarily an emotional response towards anything one believes to represent oneself.  Most people feel well ‘connected’ to themselves – indeed, perhaps rather too well connected.  It follows that they will feel comparatively less well ‘connected’ to others.  The felt contrast between these two emotional responses may be the story behind the feeling of alienation from others, and from the rest of the ‘non-self’ world – on the other side of the glass.

A Philosophical ‘Mirror Trick’?

Self-concern is characterized by:

  • strong, automatic emotional responses to perceived threats and potential rewards to persons, real or imagined, whom we believe to be ourselves
  • special emotions such as embarrassment, pride, shame, guilt, dread and eager anticipation, which we extend only to ourselves and (in some cases) to groups we belong to
  • the belief that we have no choice about the object of self-concern – my tendency to believe that there must be a fact of the matter as to whether or not a person I imagine in the future is myself, and if he is, that I am rationally required to be concerned about him.

The lack of mirror self-recognition in most animals strongly suggests that self-awareness and self-concern are recent evolutionary developments.   Other species get along well without either.  The backstory from psychology and neuroscience, although not fully developed, is evidence that self-image, to which self-concern attaches, is a psychological creation.  Although self-image resists conscious attempts to change it, it does show plasticity when subjected to psychological manipulations.

Self-image also shows signs of plasticity when subjected to philosophical manipulations.  Contemplation of fission cases like the one described in Forking can be used to weaken the sense of self the way Ramachandran uses mirrors to amputate phantom limbs.  In his paper,  Personal Identity Through Time, Robert Nozick jokingly remarked:

…when I contemplate my entering a situation of temporal overlap, my notion of self begins to dissolve.  Is temporal overlap a koan for philosophers?

That’s exactly what it is.


Gallup, Gordon “Self-recognition: Research strategies and experimental design”, in Parker. Mitchell, and Boca, ed. (1994) Self-Awareness in Animals and Humans (SAAH)

Gergely, Gyorgy “From self-recognition to theory of mind”, in Parker. Mitchell, and Boca, ed. (1994) Self-Awareness in Animals and Humans

Lewis, Michael “Myself and Me”, in Parker. Mitchell, and Boca, ed. (1994) Self-Awareness in Animals and Humans

Mitchell, Robert W. “Multiplicities of Self”, in Parker. Mitchell, and Boca, ed. (1994) Self-Awareness in Animals and Humans

Nozick, Robert (1981) “Personal Identity Through Time”, in Martin and Barressi, ed. (2003) Personal Identity

Parfit, Derek (1984) Reasons and Persons

Ramachandran, V.S. (1998) “Consciousness and body image: lessons from phantom limbs…”, The Royal Society (TRS1998)

Ramachandran, V.S. (2004)  A Brief Tour of Human Consciousness (BTHC)

Plotnik, de Waal, and Reiss (2006) “Self-Recognition in an Asian Elephant”, PNAS

Sacks, Oliver (1985) The Man Who Mistook His Wife for a Hat (MWHWFH)

Watson, John S “Detection of self: The perfect algorithm”, in Parker. Mitchell, and Boca, ed. (1994) Self-Awareness in Animals and Humans


Return to the Phantom Self home page.


One Response to “The Phantom Self”

  1. Michael O'Donnell says:

    Thanks for the article. One of the lord high mucky mucks in philosophy before Parfit was Bernard Williams, a nasty, nihilistic, fool. Parfit seems his opposite, a gentle man who understands the need, which Williams didn’t, to escape nihilism. In my opinion, however, he fails in this quest. First, the idea that sense of self is confounded by damage to the body is hardly profound. Secondly, one of Parfit’s colleagues in the Oriental Religions department pointed out the similarity between Parfit’s view of self and Buddhist descriptions of self. But the sort of self Parfit describes seems to me akin to the Abidharma, atomistic, self which, for my money, was successfully destroyed by Nagarjuna. But it’s all maddeningly complex, and what do ‘I’ know?

Leave a Reply