Overview of the Argument

This post is an overview of the Phantom Self project, as I conceive it at the outset.  It presents an overall structure of my case for conceptual reform.  It does not state the entire argument, as most supporting detail will be left to subsequent posts.  It is a plan and a rough map, not the journey itself, in the course of which there will certainly be unscheduled detours.

I will recommend reform of the concepts of self, person, and related ideas.  These concepts are important; they play a prominent role in how we think and feel about the world, in our decision-making and action.  There are dramatic differences in emotional colouring between things I think of as mine or not mine, as self or other.  The concept of person is central to morality – to responsibility for actions – and to the ideas of justice, reward and punishment which underlie our laws.  The reforms I will suggest are not trivial, and should not be dismissed as ‘just semantic’.

I plan to start by researching the progress of technology towards a capability of replicating complex objects, including living organisms.  This capability will allow us to capture everything important about a human being as information, store it in files, and use it to create a replica.  I will make the case that this capability is not only consistent with known laws of physics, but is likely to become a reality in the not-too-distant future, because of the considerable benefits of some of its applications – notably in transportation and life insurance.

My argument does not depend on this technology becoming real.  To make the case for conceptual reform, it is enough that the process is clearly imaginable.  However, I think it is important to include evidence of its likelihood in order to engage ‘practical’ people who tend to dismiss fiction as irrelevant to their lives, unworthy of their consideration.  If I can make a convincing case that the capability of human replication is coming within a generation or two, I have a better chance of being taken seriously when pointing out the difficult decisions that this capability will force us to make, and the ways in which it will force us to refine and change our ideas of who we are.

I will discuss the uses of replication technology on different kinds of things: inanimate objects, dead organisms, living organisms which are not human, and living human beings.

I will raise questions about how much information is required in order to replicate various kinds of objects successfully.  I will argue that replication need not be exact.  Replication technology need not preserve all attributes of an object.  That may not even be possible.  Replication technology will succeed if it captures and replicates all attributes of an object that matter to us.  The list of attributes that matter will vary depending on the kind of object – the list is quite different if the object is a lamb chop or a living animal.   ‘Lossy’ replication that preserves the attributes that matter to us, but not all attributes, will certainly come sooner than ‘lossless’ replication, if indeed the latter is possible at all.

Replication technology inevitably raises questions about identity.  If A is scanned and then destroyed, and the information captured from A is used to make a replica, B, we may ask whether A and B are identical.  If we are inclined to think they are, there is a problem.  The process which allowed A to be replicated as B would allow a second replica C to be made.  B and C are not identical (they are two), yet B and C each have the same relationship to A.  Therefore A cannot be identical to B or C.  A relationship which supports replication cannot be identity.  Note that this difficulty applies to identity of objects in general, not just persons. A, B and C could be thumbtacks as easily as human beings.

Applied to thumbtacks, the above case is not terribly interesting.  In replicating manufactured things like thumbtacks (or more realistically, computer chips) what matters to us is not numerical identity; what matters is that they work.  For clarity, we would probably say that none of A, B and C are identical to each other.  But identity is not what matters, for thumbtacks.  Qualitatively similar thumbtacks are interchangeable for practical purposes, and we typically don’t bother to keep track of individuals.

It’s a different matter when we think about people.  Numerical identity of persons seems important.  Choosing a future for myself seems importantly different from choosing a future for someone else like me – no matter how similar.  The difference between these choices is, intuitively, ‘what matters’ about personal identity.

But maybe intuition is wrong on this score.  If I think about facing an important life choice, it’s quite a different experience than if I think about someone else facing such a choice.  Let’s call the ‘someone else’ Bernard, and let’s make the choice $5M.  The money has been left in Bernard’s trust by an eccentric uncle.  Under the terms of the will, Bernard can choose to confer all this money on himself or on a second cousin he has never met before, but who (Bernard discovers) greatly resembles Bernard.  When I consider Bernard’s choice, I don’t feel that it matters much which way he chooses.  But if the choice is mine – if I can either have the money or give it to a stranger, even a stranger who is ‘just like me’ – I feel that it matters a lot.  I think about what I could do with the money, how it would change my life – and that’s somehow different from thinking about what another lucky guy could do with it, even if he is very like myself, to the point of supporting the same charities, dreaming of a vacation on the same French Polynesian island, and working diligently on a project just like this one.  The difference is not subtle, not nuanced; it’s vivid, and powerfully motivating.  It doesn’t matter much to me what Bernard chooses, but it matters greatly how I choose.  How Bernard chooses matters to Bernard.

But does it really matter to Bernard, or does he just think so?  Bernard thinks that if he’ll do himself a favour by choosing the money for himself – it will be a benefit ‘to him’.  But what does this mean, and is it true?  I want to explore the question whether, in such a case,  there is a benefit ‘to Bernard’.  Why, if at all, is it better for him to receive the money than for the money to go to someone just like him, who will use it for similar purposes?   Considered objectively – dispassionately, the way you or I would consider Bernard’s case – it doesn’t seem particularly better or worse.  Perhaps Bernard is under an illusion that it matters so much to him.

The prospect of replication technology forces us to think about what matters in the continuity of a person’s life – to consider the difference between ceasing and continuing to exist, between dying and surviving.  Our answers will determine not only how we use that technology, but other choices, including many we face today and have faced before.

Whatever the future of replication technology, society is unlikely to allow it to be used to produce multiple copies of the same person at the same time.  Such an application would cause too many problems, of which personal identity is only the start.  The social and legal implications are staggering.  Consider ownership: If B and C are A’s replicas, and A has been destroyed, who owns A’s stuff?   Do they have an equal claim?  What if B was replicated first?  Or should the law confer A’s property on A’s estate?  That would seem wrong – not only unjust but impractical – in a case which is so very unlike death.  Here are B and C, alive, both claiming A’s identity, property, and relationships, both in every way like A.  Either one would have an excellent claim to be A, except for the inconvenient fact of the other’s existence.  Should the law treat this as a death, and leave B and C without property?  Without citizenship too?   What about obligations entered into by A?  Perhaps the law would enforce using A’s property to pay A’s debts.  But what about A’s contractual obligations?  A non-competition agreement, for example?  And what about any‘obligations to society’?  If A had killed someone, should B and C be let off the hook?  No – allowing replication technology to be used for multiplying individual human beings is altogether too problematic for society to condone.  That application of the technology would have to be outlawed (with strict penalties for the replicas if it were violated).  But if the technology were effectively regulated so that simultaneous replication did not occur, it would have substantial benefits.  One beneficial application, discussed in the Introduction, is transportation.  Another is life insurance.

If replication technology is feasible, and has applications that are both desirable and cost-effective, and if can be effectively regulated so that undesirable applications are avoided, then it is likely to be implemented and used.

If the technology is used for serial (not simultaneous) replication of persons, then our concept of what a person is will adapt to this reality.  Concepts are human artifacts that serve human needs.  A person is, among other things, a moral and legal entity that can persist over a period of many years.  If the advantages of faster, cheaper transportation, or of life insurance which can actually restore a life, persuade people to use this technology in sufficient numbers, our concept of who we are will adapt accordingly.  Adaptation of the concept is likely to occur as a parallel process, overlapping adoption of the technology.

The idea that concepts are human artifacts shaped by human purposes suggests that, in some sense, there is no fact of the matter as to what are the ‘correct’ identity criteria for persons.  The identity criteria which will be generally accepted by society will serve society’s legal and moral purposes, and – importantly – will be compatible with human psychology.  The identity criteria must be acceptable to most people when contemplating their personal futures and pasts.  If I am considering a vacation in the Marquesas, with information transfer as my means of transportation, I must be able to identify with that bearded man lying on the beach.   If I did not look forward to his experience as I look forward to my personal future, I would not go.  Yet if I am apprised of all the facts of the case, it still seems to me that I can rationally view it in either of two ways: I can look forward to that man’s future as my own, or I can regard his future as someone else’s.  I can hold either position without making a factual error.

This takes us to the threshold of the idea of conceptual reform.  The commonly-accepted rules for answering the question, “When is A the same person as B?” – the identity criteria for personhood – are part of the concept of a person.   In order to be widely accepted, identity criteria must answer the needs of society, the law, and human psychology, including self-interested motivation.  Concepts routinely adapt in response to changing circumstances, including technological change.   When faced with the disruptive technology of human replication, the concept of a person could go either way.  It could go the way of denial that a person can survive replication.  If it did, the technology would not be widely adopted.  Or the concept could be refined so as to allow a person to be the type of entity that can survive as information, and cross between two organisms separated by a spatio-temporal gap.  In that case, the technology might be embraced, and its benefits would flow to humanity.  If we want those benefits, we should reform the concept.  We may as well be proactive, and begin now.

There are additional reasons to do so.  The convenience of travelling as information and the peace of mind stemming from life insurance that can actually restore life are not the only potential benefits of this reform.  The reform entails a shift in how we think about ourselves, and our continued existence.  If I can travel as information, or be restored from a backup file, then I am information – I am my attributes.  This has two main consequences.

One consequence is that I will never be changed too radically.  If radical change occurs, the person (or thing) that survives will not be me.  If I am diagnosed with progressive Alzheimer’s, I can appropriately fear losing my life – for if the Alzheimer’s is severe, my life will be lost – but I should not fear that I will be condemned to live the life of that Alzheimer’s patient.  That patient, who no more remembers my life than Adam’s, who cannot find his way to my house or recognize my children, will not be me.  With most of the attributes that matter lost, my relationship to him is not much different than my relationship to another Alzheimer’s patient.  There is a relationship of responsibility, such as a child might bear to a parent; I can today take steps that will affect his quality of life in the future, by choosing a good nursing home, for example, setting money aside, and choosing someone kind to bear Power of Attorney.  Except for the Power of Attorney, I could do the same for another Alzheimer’s case, with much the same net benefit.  Neither outcome would be a benefit to me.

Another consequence of conceptual reform is that the difference between self and other is diminished – the distinction is blurred.  If I am attributes, then as I change gradually over time, I may indeed become (that is, be gradually replaced by) a ‘different person’.  Moreover, others who are like me really are me, to the degree that they are like me.  If I can effectively communicate ideas to others,  if I can inspire others to take up a cause I consider important, then they become a little more like me; and becoming like me is no different from becoming me.  This leads to the conclusion that we have little reason to fear the death of our living organisms; rather, we should fear isolation.  The concept of a person as attributes promotes communication and caring.  It makes us more likely to choose what’s good for society and the planet over a benefit to a single biological organism.  It frees us from the grip of the idea that only ‘selfish’, or self-interested action is truly rational.

In the course of developing this argument, I intend to explore the work of others in a variety of fields: the philosophy of identity and personal identity; the history of how the concepts of self and person have changed in response to a changing world; the history, current progress, and prospects for replicative technology; the neurophysiological basis of the ‘sense of self’; evolutionary explanations of self-interest and altruism; and probably more.  I am sure that on this journey, I will hear of many more interesting destinations than I am now aware of, which will prompt unplanned side-excursions.

Part of this exploration will take the form of fiction.  A story can give life to a thought-experiment.  If the characters are engaging, they can draw the reader in to a world that is somewhat different from our own, and make it believable, in a way that is less likely to happen in the sterile laboratory of academic prose.  Also, stories often end up telling themselves, leading the author in directions, and to insights, he did not anticipate.

That’s my trip plan.  I am deliberately not pre-booking hotels or promising to show up anywhere at a set date.  I want the freedom of the open road, to follow questions wherever they may lead.  I hope to find company along the way, travellers interested in visiting some of the same places.  For much of the journey, no doubt I will travel alone.