The Self Illusion


To find the right answers, ask the right questions.  I have skated around the question, “Is there a rational justification for self concern?” without coming up with a solid argument that settles it one way or the other.  But there is a related question which can be answered.

Two Views of Teleportation

As we have seen, teleportation by means of information transfer can be viewed in two ways.  The facts of the case are: I am scanned in North Vancouver and my information is sent to Omaha, where it is used to construct a living replica of me.  Meanwhile, the original in North Vancouver is destroyed.  Two views of these events are:

SURVIVE: I am transported from North Vancouver to Omaha.

DIE: I am killed in North Vancouver, and someone else – my replica – is constructed in Omaha.

People who think about teleportation disagree about whether SURVIVE or DIE is an accurate description of the case.  How can it be settled which view is true – or whether neither is true?

People who disagree about SURVIVE and DIE do not disagree about the facts of the case.  The facts are not in question.  No scientific experiment can be devised to settle which of SURVIVE and DIE is an accurate description of the facts.

Most attempts to settle this question approach it on logical grounds.  They attempt to defeat either SURVIVE or DIE by showing that it is inconsistent with other statements about persons which are uncontroversially true.

I do not think that either view can be defeated this way.  Both express simple positions which appear defensible against all comers.

DIE would be accepted by someone who thinks that persons are identically biological organisms.  This view is consistent with available scientific evidence.  It also seems logically coherent.  A reason to reject it is that it denies the intuitions of those who think that teleportation is about as good as ordinary survival.  Proponents of DIE are prepared to deny that.

Proponents of DIE must also be prepared to deny the claims of the replicas who insist, commonsensically, that they are the persons they think they are, whose memories they have inherited.  This may be hard to do, especially if society takes the opposite view and accords replicas the full status of the original persons they replace.

Proponents of DIE must also maintain their belief in the face of friendly invitations from the replicas of friends and loved ones who were foolhardy enough to ignore their advice against teleportation.  This may be very hard to do.  Imagine that you sincerely believe DIE.  The phone rings; you recognize the voice of your best friend.  Several weeks have past since you spoke.  He explains that he and his family just got back from the Burgundy region, where they took a holiday on an impulse.  They went by itravel, he tells you excitedly – “So much better than flying!  No line-ups, no security hassles –  I had a Swiss Army knife in my pocket!  The chair is comfortable.  And it’s cheap!”  Your stomach churns as it hits you that your friend is dead.  You are speaking, not to him, but to someone else, a replica.  “Didn’t I warn you not to use itravel?” you blurt out.  Your friend laughs.  “It works perfectly!” he says reassuringly.  “I was nervous at first, but after my wife came back from her business trip I knew there was nothing to worry about.  Carol’s her same old self, my friend!”   Then he invites you to shoot some pool.

You can still maintain your belief in DIE, but you have your friend to deal with – or rather, his replica.  You might reject his friendly invitation, and risk hurting his feelings.  Rejecting him would be a loss to both of you.  You risk loneliness, if you are prepared to reject your friends and family who might jump on board the new technology.  Alternatively, you might accept his invitation and make the best of it.  Your friend’s replica is better than no friend at all.  You might be troubled by cognitive dissonance, as the replica reminisces about things that you and ‘he’ did together in former times.  Stirred by recollection and fellow feeling, it would be hard not to regard him as your original friend.  But if you were determined to believe DIE, you could do it.  And I don’t think you would be making a logical mistake in doing so.  Nor, I repeat, would you be making a factual error.

SURVIVE would be accepted by someone who thinks persons are informational entities that can span spatio-temporal gaps, and be copied without loss of identity.  Such entities include books, recipes, musical works, dances, software programs, genes, and species.  The claim that such entities exist is logically coherent, because it is true.  The claim that persons are such entities is also logically coherent.  It could nevertheless be defeated if it could be shown to be inconsistent with other statements about persons which are incontrovertibly true.

I don’t think any such attempt will succeed.  But there is still room for doubt.  I have argued for the theory of persons as informational entities.   But the theory is not yet fleshed out.  The job will not be done until this theory is shown to be adequate to express all facts about persons.  Demonstrating that the theory of persons as informational entities is adequate for scientific and, indeed, moral, discourse will be the main task of the remainder of the Phantom Self project.

Many philosophers, including Parfit, would object to SURVIVE on grounds that it assumes a questionable account of personal identity.  I have argued that this account of personal identity is sound, and I recommend its adoption.  But it is not essential to my point.  The two views of teleportation can be restated as:

EXPECT-EXPERIENCE: As I wait to be scanned in North Vancouver, I can reasonably expect to experience the sights and smells of Omaha later that day.  From a self-interested point of view, having a replica is about as good as ordinary survival.

EXPECT-THE-END: I cannot reasonably expect to experience anything after my body is destroyed in North Vancouver.  From a self-interested point of view, having a replica is about as bad as ordinary death.

Because there is still room for doubt about my recommended reform of the concept of a person, I will use the weaker statements EXPECT-EXPERIENCE and EXPECT-THE-END, which do not presuppose any particular theory of persons and personal identity, for the argument of this post.

Considering EXPECT-EXPERIENCE and EXPECT-THE-END, it again seems clear that the difference they express is not a factual difference.  The facts of the case are undisputed.  No scientific experiment will help to settle the matter.  And I do not think either statement is vulnerable to logical attack.  If we allow their shared presupposition that the idea of a self-interested point of view makes sense, then neither view appears vulnerable to logical attack.

For the remainder of this post, I will assume there is no logical vulnerability in either view.  Whether or not my theory of persons as informational entities can be successfully defended, I think some theory of personhood that is consistent with EXPECT-EXPERIENCE can be.  I also think that a determined proponent of EXPECT-THE-END could rebuff all logical attacks.

On the assumption that both views are logically coherent, the fact that no test can decide between EXPECT-EXPERIENCE and EXPECT-THE-END is extremely interesting.  It is interesting because it calls into question whether I, or anyone, can ever reasonably expect to have experiences in the future.

As a corollary, it also calls into question whether one can reasonably regard past experiences as one’s own.  EXPECT-EXPERIENCE and EXPECT-THE-END are echoed by a corresponding pair of statements that could be made by the survivor in Omaha:

HAVE-A-PAST: Having dinner with colleagues in Omaha, I can justly boast of the accomplishments of the North Vancouver resident who was scanned this morning.  Being his replica is about as good as being him.

HAVE-NO-PAST: I cannot justly take credit for any work done in North Vancouver.  Being a replica is about as bad as being a freshly-minted university graduate with nothing on my CV.  Worse, in fact, because I haven’t even graduated.  My expertise is the result of someone else’s efforts, as though I have plagiarized his work.

Two views of ordinary survival

The only difference between EXPECT-EXPERIENCE and EXPECT-THE-END is whether I can, or cannot, reasonably anticipate the experiences of my replica in Omaha.  If there is no real difference between these two cases, then there is no difference in the ordinary case either.  No experiment can settle which of the following statements is true:

EXPERIENCE-AS-USUAL: It is 10:40 AM in North Vancouver.  I am feeling well, and the neighbourhood is calm.  I can reasonably anticipate having experiences this afternoon.

THE-END-AS-USUAL: It is 10:40 AM in North Vancouver.  I am feeling well, and the neighbourhood is calm.  Although someone in my house will have experiences this afternoon, he will be no better connected to me now than a replica would be.  Therefore I cannot reasonably anticipate having any experiences this afternoon.

I find it difficult to believe THE-END-AS-USUAL.  In fact I do not believe it; but since I also disbelieve EXPECT-THE-END and HAVE-NO-PAST, my skepticism about THE-END-AS-USUAL does not count for much.  The evolutionary account of self-concern explains why we find it difficult to believe THE-END-AS-USUAL.  We are genetically predisposed to believe EXPERIENCE-AS-USUAL.

I can imagine myself believing THE-END-AS-USUAL.  I could believe it if I believed that the important relation that would be required for me to reasonably anticipate future experiences never holds.  I might believe THE-END-AS-USUAL if I believed the Buddhist doctrine of no-self.

Again, the point about EXPERIENCE-AS-USUAL and THE-END-AS-USUAL is that no scientific experiment can be devised to settle which of them is true.  The fact that we are overwhelmingly inclined to believe EXPERIENCE-AS-USUAL is not evidence that it is true.  As Parfit has pointed out, the evolutionary account of self-concern explains why we would believe EXPERIENCE-AS-USUAL even if it were false.

Since there is this [evolutionary] explanation, we would all have this attitude even if it were not justified.  [RP p. 308]

I conclude that the difference expressed by EXPERIENCE-AS-USUAL and THE-END-AS-USUAL has no counterpart in reality.

Parfit expressed this idea in a comparison between teleportation and ordinary survival.  He said having a replica is about as good as ordinary survival – which is the same as saying that “ordinary survival is about as bad as being destroyed and having a replica.” [RP p. 280]

What does it mean to say the self is an illusion?

It is sometimes said that the self is an illusion.  This statement has often seemed unclear to me.  I thump my chest and conclude I’m as real as anything else.  Where is the illusion?  We are now in a position to answer that question.  The illusion is that there is a real difference between EXPECT-EXPERIENCE and EXPECT-THE-END, and between EXPERIENCE-AS-USUAL and THE-END-AS-USUAL – between having, and not having, reason to anticipate having the experiences that a man going by my name will have this afternoon.

There is no real difference.  Therefore, the difference between surviving in the ordinary way and being survived by a faithful replica is unimportant.

Consideration of other sorts of illusions may help in understanding the illusion of the self.  Here is a variant of Rubin’s famous ambiguous figure:


Is it a picture of a vase, or of two faces?   There is no right answer.  The reality is just  black and white shapes.  Even the question whether black or white is the foreground colour cannot be answered objectively. We perceive the picture in two different ways.  The difference between the experience of seeing it as two faces and the experience of seeing it as a vase is vivid.  There is no difference in reality.

Something similar is going on when we contemplate the difference between SURVIVE and DIE, or between EXPERIENCE-AS-USUAL and THE-END-AS-USUAL.   I imagine the experiences that someone resembling myself will have later today.  It being sunny, I have reason to believe that person will take a pleasant walk in Capilano Canyon.

I can regard that imagined person as myself.  When I do so, I anticipate his enjoyment of the walk – his delight in the sunshine (rare, here, in the winter), the fresh scent of damp cedar, and sights of swelling buds.  I look forward to the walk.

Or, I can regard that person as someone else.  Numerous people will be out today in Capilano Canyon enjoying the weather and the promise of Spring.  Their enjoyment of their walks will be not unlike my own.  I can imagine the experience of one of them, someone not unlike myself.  My attitude towards that person’s experiences on his walk is not one of eager anticipation, because I do not regard him as myself.  But I can.  All I need to do is add a few details to this imaginative picture – make him look like me, sketch an imaginative spatio-temporal path between me, now, and him, then, and presto – I find myself looking forward to his walk!

Anticipation of experience is an attitude and emotional colouring we project onto our lives, not the perception of a real relation.  This is hard to believe in the ordinary case.  It is very difficult to contemplate an actual walk, taking place this afternoon, in two different ways: both with, and without, anticipation of experience.  We are strongly inclined to believe there must be a fact of the matter: the person walking is either myself, or someone else.

When thinking about teleportation, it is much easier to see there is no fact of the matter.  If I am about to be teleported, I can regard that as travel, or as death.  The perceived difference is vivid, much more emotionally charged than the difference between the vase and faces.  Yet it is no more real.

Importance of the Illusion

Self-concern is strengthened by the belief that we have a reason for it.  This belief makes the demands of self-concern hard to resist.  It makes them seem rationally compulsory.

To believe that EXPERIENCE-AS-USUAL and THE-END-AS-USUAL express a real difference is to believe there is a reason for self-concern.  But the difference is not real; it is illusory – a perceived difference with no real counterpart, like the difference between the vase and the faces.

This is important.  One reason it is important is that we should recognize illusions for what they are.  Recognition of psychological bias in ourselves helps us understand reality more accurately.  When we see through illusion, we see more truth.

A second reason it is important is that self-concern motivates so much of what we do.  If self-concern is strengthened by an irrational bias, or illusion, then most human action is influenced and shaped by illusion.  This may not be a good thing.  All selfish actions are motivated by self-concern.  “Selfish” is a word we use to describe actions motivated by self-concern when we wish to impugn them.

Venturing further into moral territory, it seems fair to say that most immoral actions are selfishly motivated.  By weakening the illusion that there is a reason for self-concern – a reason to favour one’s own interests over others’, when they conflict, because they are one’s own – we also weaken the motivation to act immorally.  This could be a very good thing.  If the illusion that there is a reason for self concern can be dispelled, the world can become a better place.

That is not to say that as soon as we see through the illusion, we will no longer be motivated by self-concern.  Although I think I see through the illusion, the emotions of self-concern still affect me daily.  But their effect is weaker than it would be if I believed I had a reason to anticipate future pleasures and pains.  Self-concern is too deeply ingrained in us, over millions of years of evolution, for us not to feel its effects.  But we can improve our understanding of self-concern, and in so doing, diminish its power over us.

Weakening the motivational force of self-concern makes room for other motives to strengthen, and affect us more.   Besides self-concern, what other motives are there?  One is love.  Love and self-concern frequently compete for influence over our behaviour.  Self-concern will always have the advantage if we believe it is rationally justified in a way that love is not.  By destroying that illusion, we level the playing field.

Two centuries ago, William Hazlitt thought about this subject and reached a similar conclusion.

The only reason for my preferring my future interest to that of others must arise from my anticipating it with greater warmth of present imagination.  It is this greater liveliness and force with which I can enter into my future feelings, that in a manner identifies them with my present being: and this notion of identity being once formed, the mind makes use of it to strengthen its habitual propensity, by giving to personal motives a reality and absolute truth which they can never have.  (OPHA, p 74)

Hazlitt was a highly original thinker, underrated in his time, who is now undergoing a well-deserved revival.  We will come back to Hazlitt in future posts, when I will explore in more detail the moral implications of the theory of persons as informational entities.   Before doing that, I want to get clearer on the descriptive side of the theory.  What are we, really, if we are informational entities?  What are the important relations that hold between different times in a single person’s life?  Do those important relations admit of degrees, or are they ‘all or nothing’?  How do they differ from our relations to other people? When do our lives begin and end?


Hazlitt, William (1835) Essays on the Principles of Human Action (OPHA)

Parfit, Derek (1984) Reasons and Persons (RP)

2 Responses to “The Self Illusion”

  1. David Nyman says:

    Another way, consistent with appearances, to intuit why there is no difference between SURVIVE and DIE, or between EXPERIENCE-AS-USUAL and THE-END-AS-USUAL, is that consciousness associates to each of the alternatives “according to its season”. In this view, first-personal experience is a qualitative integration, within each mutually-exclusive here-and-now instance, of a particular stage in the informational history of “how I got here”. These seemingly different pairs are actually framed so as to re-constitute equivalent first-personal histories; consequently they are indistinguishable. Looked at in this way, it is an analytic truth that the “self” referenced in each instance of consciousness is the same.

  2. […] or just waking up tomorrow morning), there is no such fact; no scientific test can ever settle the question of whether or not you will survive. What we have instead are two sets of facts, loosely related: facts about continuity and […]

Leave a Reply