[an error occurred while processing this directive] [an error occurred while processing this directive]

Resurrection: Coping with Information Loss By Mike Perry
Reprinted from Physical Immortality, 3rd Quarter 2004

The idea of resurrecting or "reactivating" the dead has a long history, for most of which it was assumed that supernatural means must be employed. Finally a few began to consider whether the problem might be amenable to a scientific solution, using advanced methods to be developed at a future date. Among these was the Russian philosopher Nikolai Federov (1829-1903), who offered that eventually it might be possible to trace the motions of individual atoms back in time to determine their placement in the bodies of persons now deceased. Using additional methods to be developed it should then be feasible to reposition these particles and thereby restore the dead to life-assuming that this sort of reconstruction would, in fact, recreate the original persons and not just copies. At least the critical recovery of information for positioning the atoms was apparently allowed by Newtonian physics, much as, on a larger scale, planetary motions over centuries could be traced and accurately predicted or retrodicted.

But the advent of quantum physics has seen hopes fade for tracing atomic motions into the distant past. More generally, a straightforward recovery of information now obliterated seems impossible to many. Meanwhile a new practice, cryonics, has become established on a small scale, that attempts to avoid the necessity of tracking the wayward atoms by keeping them in their original positions as far as possible. Persons are cryogenically preserved immediately after clinical death in hopes that eventually methods will be developed to both reverse the damage of deep cooling and cure the ailments that felled them, not excepting the aging process.

The case for cryonics is strengthened by the high degree of structural preservation seen in cryopreserved tissue, despite some evident injury. Nanotechnology, the controlled manipulation of matter at the atomic scale, is a developing field that may in time permit the necessary fine control for repair of this tissue at the cellular level and at low temperature, so a preserved organism can then be rewarmed and restored to a healthy, functioning state. Yet the problem of information loss is still with us. Indeed, imperfect preservation is arguably the most serious and difficult problem we must confront in cryonics, not excepting the various challenges we could face from opponents in the world at large. It is a philosophical as well as a scientific and technological problem. We cannot fully address it yet, nor shall we be able to for some time, but the issues connected with imperfect preservation are too important to ignore. Addressing the problem in cryonics will also have implications for other forms of preservation or its absence.

At best, all biological preservation is presently imperfect; cryostasis is no exception. Even though dramatic improvements in cryonics protocols appear to have been achieved in recent years, we are still some distance away from demonstrated, reversible suspended animation. Until that point arrives, it may require such yet-to-be-developed means as a mature nanotechnology to restore cryopreserved patients to a functioning state. This in turn is only part of the problem, the other, possibly more serious part being the condition of the patient prior to cryopreservation in the first place. Even a perfect preservation would, if performed under today's legally allowable conditions and then simply reversed, leave us with a patient in no better shape than just past the point of clinical death, with medical problems needing heroic treatment, and fast!

In the future, if our methods are brought to perfection or laws change for other reasons, it may be possible to initiate preservation when the patient is in much better shape both physically and mentally, so that reactivation can proceed more straightforwardly when means of treating the still-serious medical problems are available. (Henceforth I will use reactivation as a catchall for any of possibly many ways and varieties of restoring the patient to functionality, using whatever technological means may be developed.) For now, though, we must confront the fact that many of our patients are mentally impaired at the time of preservation. Prospects are unknown for inferring identity-critical brain structure and restoring full functionality with memories intact, even if advanced methods of the future are presumed.

On the plus side, we do expect that a great deal of damage of a certain sort can be sustained without significant loss of the necessary structural information. DNA might be used to recreate the entire body minus the brain, for example, and much of the brain's information seems redundantly encoded so that it might be inferred even after substantial injury. When there is enough information to restore the healthy state, the problem of reactivation should be easy to deal with in principle. That is not the main focus of this article.

Instead I will consider the other cases, where some significant, irretrievable loss of identity-critical information has occurred. These I will call Information Deficient (ID) cases. For these it is less clear what ought to be done, even in principle. Opinions will vary; definitive answers are not to be expected. Here I offer some of my opinions, recognizing that much of what I say is speculative and that not everyone will agree. I'll try to cover several points of view, though naturally with emphasis on what personally seems most significant and relevant.

One issue that arises at the start is that there are different views on what would constitute reactivation and survival. For some, original material-the body or at least the brain or some crucial component of the brain-must be restored to a functioning state: what "you" really are is dependent on a specific physical object or "token." Others would impose less stringent requirements-a duplicate of the original would suffice, or even an upload in which the personality is reduced to a body of information, then transferred and "run" on an advanced information-processing device. In this case the token is not of primary significance but only the information or pattern. Whether one is a tokenist or patternist-or something in-between or otherwise entirely-will affect one's judgment about the efficacy of a proposed reactivation. As it happens I take the patternist view, and see the problem of reactivation as by and large one of information retrieval and/or recreation. Once the necessary information is available, the rest, reinstating the person in some physical and functioning form, should be relatively straightforward and unremarkable. This assumes, of course, that advanced technology of the future is available to carry out what would be needed-a general-purpose nanotechnological assembler should be adequate, and possibly devices of much less capability; such devices seem consistent with physics as we know it. For patternists the result, a construct which can be considered to embody or instantiate the very person we would like to reactivate, need only exhibit sufficient similarity to the original to qualify. Tokenists will object and ask for more but may also see value in what has been accomplished, particularly if it is the best that could be done.

So, proceeding with the assumption that the information or pattern is what is really important, I will also assume that some information about the past cannot be retrievable, a viewpoint that seems well supported by modern physics. Otherwise there would be no ID cases to worry about, even if no part of the patient remained! I'll start by outlining some reasons why we should be concerned about the ID cases now.

First, to summarize and elaborate the concerns raised above, it is possible that very many cryopreserved cases are ID, not excepting some recent cases where advanced protocols were used but the patients had substantial brain injury at the time of preservation. This situation could continue for some time to come, and could also occur unexpectedly, as in the case of fatal accidents, or a period of warm ischemia following clinical death. The latter is especially common with last-minute cases, but clearly we are all at risk. That rather unsettling thought means, for instance, that we should not overlook other ways of storing information about ourselves besides cryopreservation, such as videotapes, diaries, and other memorabilia.

Second, it never hurts to start thinking about the tough questions early, even if there is much that cannot be done yet. This could have repercussions in such matters as how cryonics is presented and how it is perceived by those we'd like to get signed up, as well as orienting us to take advantage of new technologies and methods when they do come along.

Third, there are choices that will have to be made in ID reactivations that we, the potential patients, should start thinking about. If our memories were partly or wholly erased, for instance, would we want to be reactivated amnesiac, or have some memories "reinvented"?

Fourth, cryonics is only one possible means to the end of preserving a person's remains for possible reactivation. Confronting imperfect preservation in cryonics will make it easier for us to consider alternatives such as chemical fixation, where the preservation might be even more imperfect but still not negligible.

Finally, confronting the ID cases raises fundamental issues about the nature of personality and survival that ought to be addressed: for instance, what do we accomplish if we do preserve a given chunk of information about ourselves, and why is that better than just losing the information?

Let us now consider a reactivation scenario for a seriously ID case. It is often said, "the worst thing you'll suffer is some amnesia." Certainly that is reasonable, at least if an intact copy of the genome can be recovered. (For reasons of brevity I'll mainly consider the DNA-recoverable cases here; these are nearly universal in cryonics, though the principal arguments will apply more generally.) With the DNA it should be possible to repair or replace every organ of the body; only memories or other nongenetic information acquired in the course of life cannot be reinstated this way. Even in the more extreme cases, though, clearly we'd not be limited to a person with just the information that can be deduced from the biological remains.

Suppose the brain is entirely missing: all we have is a cell sample. It is sometimes assumed that the best that can be done then is to create an infant, identical twin of the original, who would then have to grow up all over again. Not so, I submit, far from it. A great deal should be inferable about the original person from other sources. We should know: her name, what native language she spoke, and probably much about her various talents, interests, preferences, where she lived, visited, attended school, and what subjects were studied; plus information about friends, loved ones, and acquaintances. We could create someone who, as far as anyone could tell, was the original person down to the smallest detail, and who was fully convinced of this herself. To do so, however, would require inventing some details, and particularly the sort of important details, such as memories, that would be most vulnerable to loss through destruction of the original brain structure.

At this point we have entered a realm of controversy, where opinions will vary as to the "authenticity" of the reactivation as well as the propriety of carrying it out. Some might favor including only details we could be sure about. However, this could result in someone with obvious defects and also major unhappiness, in addition to being an unnatural and improbable reconstruction. On the other hand, if the missing details were to be filled in by imaginative guesswork, many would view the result as a historical fantasy, possibly much like the original, but certainly not the same person, nor any real person who ever lived. So what is to be done?

One important factor, certainly, would be the wishes of the patient in question. It may be that these were spelled out in detail, so that it is clear what desired course was to be followed. Thus, in a hypothetical scenario, perhaps the patient (an adult we'll assume), died in a plane crash, but after first leaving behind a DNA sample. Perhaps the instruction for this contingency was simply to create an infant from the DNA and raise that person to maturity-or perhaps it was something more challenging. In any case, if the instructions are clear, the future choice may be much clearer too, and people accordingly should be encouraged to express their preferences. But we must also keep in mind that future wisdom may dictate other procedures or policies than what were naively requested, particularly in more difficult cases. Rather than narrowly attempting to micromanage one's reactivation, then, it is probably better to focus on providing pertinent information. Videos, diaries, and other personal memorabilia will both serve as backups in case one's remains are compromised, as we have noted, and also can indicate preferences.

I'd now like to look in more detail at the problem of authenticity in reactivations. Here we enter a still more speculative realm, where far-ranging scientific issues become important.

It might seem that, if a person were reactivated from incomplete remains, so that filling in of details by guesswork became necessary, the result must be an unhistorical fantasy, a person who never really lived. This is based on the world view that history has a single, authentic timeline, and that information about what happened can be lost so that purported information about the past can be false, but not verifiably so. (Actually, even in this case there is a small chance that all the billions of bits we would expect to have to guess could be correctly arrived at by accident, but this likelihood is so small I'm ignoring it.) On this basis, then, it becomes impossible to reactivate the original person in the ID cases. Some subscribers to this viewpoint, however, may still have interest in a person much like themselves returning to consciousness, and will express their wishes accordingly.

But there is another possibility, which is that history does not have a unique timeline, but that alternate histories are unfolding in parallel, though almost all of this plural effusion is hidden from our observation. This in particular is the viewpoint of the many-worlds formulation of quantum mechanics, which in addition argues that each single, historical timeline is continually splitting into multiples. This means that any observer is constantly dividing into copies, each of which then acquires a separate consciousness and pursues a separate existence. In fact, essentially every possible history is unfolding somewhere. (The overall creative explosion arguably dwarfs anything yet imagined about supernatural powers!) Timelines also can fuse, when the information necessary to distinguish them is erased, so that loss of information can be said to make the past ambiguous.

The metaphysical claims of many-worlds have left many physicists more than a little skeptical. Yet there is interesting evidence for this strange and wonderful theory, which is simpler in some ways than its rivals. Its logic seems compelling to some thoughtful and knowledgeable scientists and acceptance seems to be growing. In particular, cosmologists, who study physics on the deepest and most far-reaching levels, generally favor many-worlds. (Though it is not yet possible to distinguish, experimentally, between many-worlds and some rival theories, there are ways in principle that distinguishing tests could be carried out; hopefully this will happen when our technology is more advanced.)

If true, many-worlds would have profound consequences at a deep level of reality, and a major impact for the issues considered here. In particular, the outlook for an ID case becomes immediately more hopeful. Assume, for instance, that an attempt will be made to produce the best possible reconstruction of that person, using educated guesswork where necessary, and suppose that it does produce a functioning person who resembles the original in all known details. We'll also assume, as usual, that many details had to be filled in by guesswork, so that our reconstruction is-by conventional intuition-unlikely to be identical to the (seemingly) single original that once existed.

This does not mean that this one original is not reactivated, because at this point parallelism comes into play. In keeping with many-worlds, any historical process is flanked by many similar processes, which together fill out a vast space of possibilities. In our hypothetical case, then, there would be not just one reactivation attempt but very many in parallel, carried out by near copies of ourselves in near copies of our world. Somewhere, among all the ongoing reactivations, there would be those that, however unlikely, just happened to create exact copies of our original. This person, then, could continue his/her existence in much the same way as other people, who under many-worlds are constantly splitting into copies anyway. Conversely, the inexact copies, produced in other parallel worlds, would be exact copies of other persons who had really lived, in other, parallel histories. Every attempted ID reactivation, then, if it produced a person that, as far as one could tell, was authentic, would in fact be authentic, though from a past that was not uniquely tied to our present.

An additional point worth making is that an individual does not belong uniquely to any one of the parallel worlds, but, in the form of identical (or sufficiently close) copies, is simultaneously present in all the differing timelines whose histories agree with his/her own memories. (The interchangeable copies can thus be regarded as "instantiations" together comprising only a single individual, this being consistent with, and actually a necessary consequence of, the patternist position.) The reactivated individual, then, never ends up in the "wrong" universe, but was actually present before, so long as consistency is maintained between the memories and the external, historical record.

Many-worlds, then, appears to yield a hopeful scenario even when substantial information is not preserved. Other theories involving multiple universes, that is to say, a multiverse, and many copies and near-copies of every individual, could, with some adjustments, also allow for similar scenarios. One can even make the argument that, given that conditions seem to have to be very special for intelligent life to appear, there must be some proliferation of this sort to account for our presence here.1

But the issue will then be raised of why preserve one's information in the first place, if it is not necessary for an eventual "coming back" in some form. I think there are good reasons to opt for preservation, even though the arguments take on a certain subtlety. (And extreme situations could call for the noble sacrifice of one's chances of preservation-but this interesting, tangential subject will have to be deferred.) A case for the preservation option can be summarized as follows.

A major consideration is simply coping with the unknown. What may follow our clinical death cannot be predicted with confidence, but preservation should offer a better means of coping, reducing the extent of the unknown. Without well-preserved remains the task of reactivation, which may be hard enough in any case, could burgeon up to planetary proportions, even with future, far-advanced capabilities. As a preliminary, a sizable portion of the inferable history of the earth extending back to before one's birth may have to be recovered, analyzing all relevant materials down to the molecular level, merely to ensure that one's reactivation will, in fact, be consistent with this history. By a reasonable refinement, mutual consistencies with contemporaries and other reactivated IDs will need to be established, which will further complicate matters. It is hard to say how long such operations may take or what features they may assume, but we can imagine armies of archaeological, nanoscopic robots working on and in the earth's crust for decades, centuries, millennia-who knows?-assisted by tidal waves of computation. The difficulties of these and other necessary steps will probably be "passed on to the consumer"-the patient-in one way or another, including the dislocation and disorientation caused by one's reappearance in strange circumstances and possibly a distant future. Efforts will no doubt be made to "smooth the path" and make the whole enterprise bearable and even enjoyable, but at least some significant hardship and adverse reaction can be expected.

A word should be said about those who would carry out or supervise the anticipated reactivations. I am assuming, of course, that advanced, benevolent former humans or other beings will eventually appear who willingly undertake all the necessary, technology-assisted labors, which might develop into a project of scope and duration well beyond even the levels suggested above. (Concern would naturally extend to all other persons of the historical past, and more generally to sentient life-forms which have made their appearance at one point or another since remote prehistoric times.) I think it is a reasonable assumption, all things considered, that benign reactivators will actually emerge and engage themselves as I have suggested, assuming civilization endures-though admittedly this is not self-evident. (Actually I would like to be among their number myself someday, and probably many others, upon serious reflection, would find this idea appealing.) Yet the reactivators too must have their motives, which will probably accord with their refined notions of enlightened self-interest, and may involve some deliberately induced hardships for the ultimate benefit of those reactivated. Preservation, by comparison, should bypass much of this whole problem. If nothing else, those who choose it should need less in the way of "parenting" from their rescuers. These in turn should be much closer to comrades and peers of the patients they are helping, and should bring them into a world less removed in time and circumstance, and more like the one from which they departed.

Additional arguments suggest ways that both oneself and others will benefit through one's preservation. We just noted that the well-preserved should come back sooner with less in the way of difficult adjustments. They might relatively soon join their rescuers and others and advance along with the new civilization of the newly-enduring. So there should be an increased facility for meaningful interaction with fellow beings, which should translate to a better, more meaningful life all around. This could be significant in ways hard to appreciate today. The sort of enduring entities we presently transient humans hope to become would arguably have highly refined survival instincts. Historical ties would then assume a deeper significance than is possible now. Preservation, it should be noted, will prevent a loss that would not be fully reversed even if the same person is recreated by guesswork, since that person's role as a defining element of recorded history will have been irreparably compromised. History that is recreated must retain an ersatz character that will sharply distinguish it from the recorded, recovered variety, even though it too is "real," in a suitable sense, given the multiverse ontology. At the human level, even now it can be seen that the remains, particularly the brain, are a kind of diary whose contents arguably ought to be saved and transmitted to posterity. Of course, here we are not simply trying to preserve historical data but even more important, to save and extend a person's life. (And it is also worth noting that the brain stores information about other people, including those who will not be preserved themselves, and thus could benefit these others by strengthening their own historical ties.)

There is one additional way that the well-preserved should benefit, that is not so hard to understand for us humans today. It involves the issue of status, and it hinges on the earlier start that, as we noted, should be open to those whose reactivation will be the simplest. With this head start they should gain a position of rank and privilege long and well surpassing those who must be reactivated by a slow and laborious alternative. Every privilege carries a responsibility, as the old saying goes, so this must be approached with gravity-yet it is something to think about.

Finally, a moral argument can be invoked to link the other arguments together and remind us that we must bear responsibility for our choices on life-and-death issues. The upshot: though reactivation could happen without prior preservation, it would place the recipient at a disadvantage that he or she, along with others, would come to recognize as such.

Once we accept the idea that preservation for one's eventual reactivation is a good thing, we can plausibly extend this to conclude that, if some identity-critical material must be lost, it is better to lose less than more. (Actually some argument is called for, but the case can be made that those who are more mildly ID could be brought back much sooner than the more compromised, with the fill-in of missing details, relatively minor but still significant, possibly deferred to a later date.) If forced to choose an imperfect preservation, then, we still want to choose as well as possible. Even a cell sample should be better than no DNA, burial should be better than cremation, a first-rate chemical fixation better than more conventional embalming, and so on.

But best of all is preservation that is adequate, so that no information must be conjured by guesswork to avoid a compromising amnesia or worse. This is what we must seek, while not losing hope for the more difficult cases.

1. See, for example, R. Michael Perry, Forever for All, Parkland, Fl.: Universal Publishers, 2000, ch. 6.
[an error occurred while processing this directive]