Thomas Metzinger’s work describes the self-model in which our ideas about ourselves are rooted. The model is (usually) transparent, in that we operate through it without (usually) any awareness of a distinction between the model and the underlying reality. It is a model to which we have a profound emotional attachment—most of us care, deeply, about ourselves in the past, present, and future. As a result, our self-models are motivational. They spur and shape our actions. We evaluate possible courses of action by putting our self-models through various simulations, and responding emotionally to the different outcomes we imagine. The research of Antonio Damasio has begun to show how our emotions must inform our executive decision-making processes in order for us to make what are commonly recognized as ‘rational’ decisions.
Most of what Metzinger and Damasio have to say about the self is as true of isolated individuals as of human beings immersed in society. But a case can be made that the concept of the self could only have emerged in a social context. I have argued that our concepts, particularly the entities recognized by our ontology, reflect what is important to us. The spatio-temporal boundaries between ‘things’ are artificial, not natural; they do not exist in nature, but are imposed upon nature by human beings. A person is an entity whose boundaries roughly coincide with those of a human biological organism. A person is commonly considered to begin sometime around birth; sooner in some traditions, later in others. The person is usually thought to persist until biological death; but many people believe that it continues much longer than that; and some believe that if the organism is sufficiently damaged, then the person may cease to exist before its organism dies.
Among other things, a person is a unit of moral and legal responsibility—a bearer of enduring rights and privileges, duties and obligations, merits and demerits, assets and liabilities, debts and credits. Those attributes of individual persons result from, and depend on, the fact that individuals are members of a larger society. If a human being is isolated for a long time from other human beings, legal obligation disappears from his life, and moral obligation, if it does not entirely disappear, is vastly curtailed. I would not go so far as to say that an isolated human being ceases to be a person; only that certain central and important aspects of personhood simply disappear from his or her life. Having moral and legal rights and obligations is a central and important aspect of personhood.
Laws are, for the most part, a carefully worked-out distillation of widely-shared principles and values. (At least that is true in democracies.) A society’s laws tend to reflect what is important to its members. And one of the most central ideas in law is the concept of a person. If we want to understand personhood, we cannot ignore the legal concept of a person. It would be quixotic to recommend reforming the concept of a person without considering the implications of such conceptual reform for the law.
In Western law, identity of persons matters a great deal. It matters because persons have rights and obligations; because they own property, and incur debts; because they enter into contracts, and are held accountable for their past actions. The deliberations of juries are largely devoted to settling questions of personal identity. Is the prisoner in the dock the same person who committed the crime? The answer, which settles whether the prisoner is found guilty or not, has great significance for him and for society.
When we consider persons as members of society, the idea that personal identity is disrupted by transformations such as teleportation becomes harder to defend. In Western law, the existence of persons is considered to extend from some time near birth until biological death. The legal and moral unity of a human lifetime spans many acknowledged changes which persons undergo in the course of their lives.
A society which used information-based teleportation would certainly recognize that the legal status of persons survives that transformation. Ownership of property, for example, would survive; if it did not, no one would agree to be teleported. Contracts would remain in force, rights and duties would persist, including legal ‘obligations to society’ for criminal acts.
Teleportation and the Law
Derek Parfit, in Reasons and Persons and subsequent papers, argues that transformations like information-based teleportation preserve ‘what matters in survival,’ but not personal identity. The reason they do not preserve identity is that such transformations can take a branching form. Information-based teleportation makes a copy of a person and destroys the original. A process that can make one copy can make more than one. Two coexisting copies could not be the same person, and therefore cannot both be identical to the original. Since both have the same relationship to the original, neither can be identical to the original.
Parfit holds that legal and moral rights and obligations survive transformations like teleportation, even though personal identity does not. Parfit goes further, saying that rights and obligations even survive transformations which in fact take a branching form. He quotes Wiggins: ‘A malefactor could scarcely evade responsibility by contriving his own fission.’ (Parfit, 1984, p 271)
Surely Parfit is correct in holding that preserving legal and moral rights and obligations is an important part of ‘what matters in survival.’ But to accept his proposal verbatim is to create a major problem for the law. As our laws stand, when a person ceases to exist, his property passes to his estate and is disposed of by his executor. His debts are settled from the estate. His social insurance number is retired, and his driver’s licence becomes void. If he has committed a crime for which he has not been held to account, no one can subsequently be held to account for it. To leave the law unchanged while accepting Parfit’s proposal that teleportation puts an end to one person’s existence and begins another, would be to recommend a world in which users of what is intended to be a transportation system become instantly bankrupt, and are absolved of all responsibility for past crimes.
Obviously, that won’t do (and Parfit would agree). As the law stands, if I can be held to account for a murder that happened ten years ago, or even for something as minor as a speeding ticket, it is because I am the person who committed that murder, or drove my car at 68 kph up Highland Boulevard. If there is reasonable doubt that I was the person who did those things, I am off the hook. A legal system which allowed me to escape consequences simply by arranging to be teleported to the next room would be laughably ineffective.
It’s worth examining why it would be ineffective. The reason is that I do not regard my teleportation as death. When I buy a teleportation ticket, I ‘identify’ with the survivor of my teleportation in the same ways, and to the same extent, as I ‘identify’ with myself at any time in the future. ‘Identifying’ with the person means that I have self-concern for him; I care about him just as I care about myself. I am as strongly motivated by the prospect of him evading punishment for my crimes as if I could evade punishment myself. If I have such an easy out, the law loses its deterrent effect.
If Parfit is right about personal identity, and I am right that information-based teleportation is a coming technology, then clearly the law will have to change. A simple and obvious recommendation would be that the legal status of persons should be preserved when ‘what matters in survival’ is preserved—that the law should be revised by a global search and replace, substituting the concept of survivorship wherever it now relies on the concept of same person.
However, that simple fix gives wrong answers in cases where the technology fails to neatly replace an original person with a single copy. When human replication technology comes into common use, even if it is regulated with a view to preventing the duplication of persons, there will likely be cases in which, by accident or design, the rules are violated and human beings are duplicated. The law will be required to respond to human duplication, and the response, “Legal status is preserved intact,” will not be adequate. Before he was accidentally duplicated, Elliot, the main character in the story “Forking,” owned property. His right of exclusive ownership of his car could not logically be inherited by both his copies; if it could, it wouldn’t be exclusive ownership. The rights of joint owners are different from the rights of exclusive owners. (To say that Elliot’s exclusive ownership rights are inherited by both copies is to say that both post-split copies are the same person, a position that is hard to reconcile with the real relationship between them, which is competitive.) If Elliot’s status as Marjory’s spouse were inherited by both post-split Elliots, the effect would be to legalize bigamy. If his status as employee of the Waterloo Institute of Transportation Studies were inherited, would the Institute find itself under a new obligation to write the contractually-specified monthly cheque to each copy, thereby doubling its payroll outlays with respect to Elliot? Or would the Institute’s obligations be limited to one monthly cheque as before? If the latter, then would the post-split Elliots have a shared claim to the cheque, and a shared obligation to fulfill the pre-split Elliot’s teaching duties? Such matters should be decided by law. Because shared claims and obligations are not the same as unshared ones, a blanket substitution of “is a survivor of” for “is the same person as” throughout the law will not work. Between those two relations are real differences with which the law will have to come to grips.
This is true whether or not we eventually decide that personal identity survives teleportation in its normal, non-branching form. In either case, to cope with branching cases, the law will require substantial changes.
Human Branching and the Law
These considerations are not restricted to transformations like teleportation, which preserves psychological and physical similarity without continuity of physical substance. Parfit imagines a case in which his brain is removed from his body, and each hemisphere is transplanted to the body of one of his two “fatally brain-damaged brothers, Jack and Bill.” (Parfit, 1984, p 269) For the sake of argument, Parfit supposes that the hemispheres are substantially the same, and that each, on its own, is capable of sustaining the personality and other psychological attributes distinctive of Derek Parfit. The plausibility of the example is bolstered by the fact that human patients have survived hemispherectomies, and are considered, by society and themselves, to be the same persons they were before the operation. Real cases include left-hemisphere removal and right-hemisphere removal, indicating that neither one is critical to survival; either hemisphere will do. In Parfit’s example, if both operations are successful, the pre-operation Derek Parfit will ‘survive’ as both post-op patients. But if one operation fails, there will be only one survivor with a claim to inherit the rights and obligations of the pre-op Derek.
Again, it is the branching case that causes trouble for the law. If the second transplant fails (because someone accidentally dropped one of Derek’s hemispheres on the floor), Derek’s legal status can be transferred quite unproblematically to the sole survivor, simply by stipulating that the original Derek is the survivor. But real problems crop up if both survive. Each survivor has a moral claim to Derek’s property. Existing law has no provision to settle those conflicting claims. Since the claim is equal, there may be an easy solution; the law could be extended to provide for an equal division of the property of a person who branches. But not all rights and obligations could be dealt with so easily, as Elliot’s case illustrates.
These examples show that it is a mistake to think that moral and legal status is preserved by all transformations that preserve psychological continuity. When such transformations take a branching (or merging) form, moral and legal rights and obligations are substantially altered. But they need be altered only when branching (or merging) actually takes place. If the transformation is in fact one-to-one, moral and legal status can happily ‘flow through’ to the survivor intact.
The fact of whether or not branching occurs clearly has great moral and legal significance. But there is no obvious significance to a transformation which is in fact one-to-one, such as simple teleportation without branching
I don’t agree with Parfit that because information-based teleportation can be used to make multiple replicas of persons, it disrupts personal identity. The fact of branching is of paramount importance; the mere possibility of branching is of no importance whatever. Much less violence is done to the law, to our moral precepts, and to ordinary ideas about personal identity, by maintaining that personal identity persists through transformations that preserve ‘what matters in survival’ as long as the transformations are in fact one-to-one.
I suggest that is exactly the direction the law will take when it is confronted by the reality of information-based teleportation. It is the path of least resistance. The law will have to be extended to cover the new, exceptional cases where human branching occurs, but there will be no need to retrofit it with wholesale changes to cover the unexceptional cases where branching does not occur—in which teleportation is just a cheap and convenient means of travelling from LA to Omaha.
If the law does in fact adapt to the reality of teleportation by following this path of least resistance, some philosophers may continue to argue that it does not reflect the truth about personal identity and what matters in survival. But the path of least resistance is often taken as a signpost to truth, and we should not reject the signpost as misleading if no stronger evidence points in a different direction. I am not aware of any such strongly countervailing evidence.
Life Insurance and Personal Responsibility
The problems that human replication technology poses for the law are not confined to branching examples. Another application of the technology, restorative life insurance, challenges both the law and generally-received principles of moral responsibility for actions.
Rather than paying out money to beneficiaries in the event of an insured person’s death, the new, improved form of life insurance restores his or her life. It works like this: when I take out a policy, I am scanned by the same sort of scanner that is used in teleportation, but instead of a copy being constructed right away, my information is stored away on what would be considered, by today’s standards, a very large hard disk. Periodically, perhaps every month or six, I return to the scanning facility to record a new ‘backup’ file. In the event of my death, my replica is reconstructed on the basis of the stored information. Thus I ‘get my life back’. A drawback to this procedure is that I lose all memories from the time between my latest backup and my death, along with any other changes I may have undergone, including normal aging. But that loss is minor compared to the loss my family, friends and myself would suffer if I were to die permanently. Recovering from a few months of lost memories is a relatively minor nuisance, especially if I’ve kept good records.
Restorative life insurance has the potential to undo tragedies. But it poses thorny problems for morality and the law. If I am scanned for insurance purposes in 2012, commit a crime in 2013, then die in 2014 and am restored from the insurance backup as I was in 2012, should I be then held responsible for the crime? Weighty reasons can be brought to bear on both sides of that question. At the time I am restored from backup, my personal history—that chain of psychologically continuous and causally connected stages that culminates in my present self—does not include committing the crime. I can sincerely claim, “I didn’t do it. I wasn’t even alive in 2013.” In such a case, it is tempting to adopt Parfit’s terminology of railway lines.(Parfit, 1984, p 287) If my life following restoration is the main line, then the criminal turn my life took in 2013 was on a branch line. I—the main-line survivor—should not be held responsible for what happened on the branch line. I didn’t commit that crime. In Parfit’s terms, I am not even the direct ‘survivor’ of the person who did. There is a significant psychological (and physical) discontinuity between the criminal at the time he committed the crime, and myself after being restored from backup.
But we must be careful, in too-quickly exonerating me from responsibility. Restorative life insurance will catch on for two reasons. First, it will be popular with the survivors, who will get their loved one back—or a facsimile that is, in every perceivable and testable way, satisfactory. Secondly, it will meet the approval of the insured persons themselves. Their experience will be much like that of someone who suffers temporary amnesia—a gap, or lacuna, between the time at which they made their backup and the time at which they were rebuilt from the stored information. They will be met, embraced by their families, told of the circumstances of their deaths, filled in on the events that they missed, including (most importantly) the events of their own lives. They will remember deciding to take out their life insurance policies, and making their regular backups. Most will consider themselves both wise and lucky to have cheated an untimely death, at the small cost of the loss of memory of a few weeks or months. (Those who lose a few years may regret not having followed a regular backup schedule.) Persons whose lives have been saved by their insurance policies are likely to be enthusiastic about renewing them. They’ll think, If it happened to me once it could happen again. They will remember their earlier lives as their own, and will not hesitate to believe that their own lives will continue after they are restored, should they be unlucky enough to die again. That is, they will invest their personal hopes for the future—full-blown self-concern—in their restored replicas. And others, who hear grateful testimonials from those restored replicas and their relieved families, friends, and employers, will be similarly inspired to insure their own lives in expectation of surviving death, should death occur.
I am making a psychological claim about human beings—that if restorative life insurance were available, if it worked reliably and was widely used, it would attract new users who would be sold on the expectation of surviving their own deaths, at the relatively minor inconvenience of losing their memories of the period between their latest backup and the time of death,(an inconvenience partially offset by the fact that an equivalent period of physical aging would be reversed).
That natural tendency to invest self-concern in one’s restored replica creates a strong reason for holding revived policy-holders responsible for acts committed after their latest back-ups and before their deaths. The contrary position, that they should be exonerated, would create a motive for the selfishly inclined to evade the law and reap the benefits of crime by taking out a policy and then killing themselves. (The new kind of life insurance would not have a clause excluding payouts in case of suicide.) If people take personal comfort in knowing that they are insured, and can face death with indifference because they are confident they’ll be back, they should expect moral and legal responsibility for their actions to be restored along with their lives.
To hold someone responsible for a crime that not only can he not remember, but is not even part of his personal history, runs counter to a widely-felt sense of natural justice. Yet I think it is easier to justify than the alternative. There will be fewer objections to transferring some other aspects of a person’s legal and moral status. If I earned money since my latest backup, and bought a car, few would argue that that property should not remain my own. If I entered contracts, the question becomes murkier, with reasons on both sides. I suppose that if I took a job during that time, I would be entitled to the usual protections accorded to employees against summary termination. But what if I entered a contract of marriage? Should I be bound by the vows made by my ‘branch-line’ self? And would my wife still be bound by her vows, made “till death do us part?” It seems the answer must be “yes,” because the alternative would be too chaotic. But much remains to ponder. Consideration of the moral and legal implications of restorative life insurance makes me think that the main task of of the Phantom Self project—to reform the concept of a person—has barely begun.
Damasio, A.R. (1994), Descartes’ Error
Metzinger, T. (2004), Being No One
Parfit, Derek (1984), Reasons and Persons