The Duplicates Paradox (The Duplicates Problem)

by Ben Best

I would be glad to know your Lordship's opinion whether when my brain has lost its original structure, and when some hundred years after the same materials are fabricated so curiously as to become an intelligent being, whether, I say that being will be me; or, if, two or three such beings should be formed out of my brain; whether they will all be me, and consequently one and the same intelligent being.

                         — Thomas Reid letter to Lord Kames, 1775

The idea that personal identity (mind, self, will) is entirely contained in the molecular and biological structure of the brain (ie, is entirely material) implies that a duplicate identity could be created that is identical in every way except physical location. Not only would every element, molecule, neuron and synapse be identical — and in identical physical relationship to the others — but every neuron and synapse would be in an identical state of depolarization at the instant of duplication. That is, both the static condition and the dynamic states would be duplicated.

The duplicates question has practical implications for cryonics because if a cryopreserved brain can serve as a template for atom-by-atom reconstruction of a new brain (restoring the original person's identity) there is no reason why the same reconstruction could not be done more than once. Would each reconstructed brain have the same mind/self/personal identity?

It is said that most molecules in the body are replaced every few years, presumably also including the molecules of the brain. Whether that claim is true or false, the exact atoms and molecules of which our identity is constituted are not essential to that identity. Thus, lives and minds are much like a candle flame — life, consciousness and selfhood continues with a material basis despite the fact that the exact material (atoms and molecules) change.

Not only can molecules be changed, but we can acquire new memories and change our interests & tastes without compromising our personal identity — ie, we still feel ourselves to be the "same person". What determines what can change without compromising personal identity and what cannot change? Do we have thoughts, emotions & motives or are we thoughts, emotions & motives? If we do not always know our true thoughts, emotions & motives, does that make them separate from our identity? Who introspects? And is the change of thoughts, emotions & motives over time a symptom of a changing self? Memories may change less than thoughts, emotions & motives.

Personal identity is perceived as continuous through time. Yet this perception cannot be instantaneous, and must be based on memory. Given the fact that memories can be forgotten, altered or even fabricated, the question arises as to whether memories are essential for personal identity. Certainly no specific memory seems necessary for identity, but a perception of a continuity of the memory process is often believed to be. Subjective experience involves not just memory, but thoughts, desires, feelings and personality. Even when subjectivity is focused on the "outside world", this focus necessarily has a point of view. Any attempt to describe personal identity impersonally will lose an essential element. A self has both sensation and will.

It is facile to say that duplicates are identical people with identical memories, desires and personalities — who will diverge from the moment of duplication onward as circumstances force them to have different experiences. Subjectively this description is unsatisfying because it does not address the personal experience of being in one location in one moment of time — and it does not address the issue of the personal qualia of selfhood, ie, the issue of survival of personal identity.

It does not matter to me that a duplicate can only be a "true duplicate" for a brief instant — any more than it matters to me that I change from moment to moment — because for me change is compatible with survival. I not only survive from moment-to-moment, I survive from day-to-day and from year-to-year. My subjective identity survives. And I want to have it continue to survive.

It follows that a duplicate of me would not only be "essentially me" at the moment of duplication — but would continue to be "essentially me" for a long time afterwards. I do not believe that decades of living have altered my personal identity. Only in the sense that I cannot be in two places at once can I say that the duplicated person is no more me than any other person is me..

People often say "duplicate" without really meaning "duplicate". It makes no sense to describe one copy as an "original" and the other as a "duplicate". Both are originals and/or both are duplicates. I could as easily be him, and I could as easily be called "the duplicate" as he — if we are true duplicates. If a duplicate is created in a new location, it still makes no sense to use location as a justification for distinguishing between a "duplicate" and an "original", because location is not a component of personal identity. Moving from one location to another does not alter identity — we do it all the time! Which leaves the essential paradox: (1) a self can be materially duplicated, (2) a self cannot be in two locations at once — and (1) & (2) are totally incompatible.

"Which duplicate is me?" cannot be answered with the word "both" because one person cannot be in two locations at the same moment. It cannot be that my physical location is an essential ingredient of my personal identity because I frequently change my location without losing my identity. It does seem, however, that an essential ingredient of personal identity is the inability to be in two different locations at the same moment. If someone punches one duplicate in the nose, only that duplicate will feel pain and only that duplicate will bleed. If the duplicate is not me, I won't bleed. If I am killed and my duplicate is not killed, I do not survive (I lose the "qualia" of life).

From a completely objective point of view, the technical possibility of creating an identically duplicate person is an obvious consequence of materialism and in no way a problem. It is no problem for me to imagine two Robert Ettingers standing side-by-side, identical in every respect. From a subjective point of view, however, the possibility of duplicates poses a serious problem. Specifically, it is a very serious problem for me to imagine two Ben Bests standing side-by-side, one facing North, the other facing South. After duplication, would I see the view to the North or the view to the South? I cannot answer this question with the counter-question "which 'I'?" There can only be one I to me, I can only look in one direction at once. If the duplicate facing North is destroyed, I will be either dead or alive. I cannot be in two minds, two bodies and two locations at the same time. I cannot speak of one "I" facing North and one "I" facing South. That is an objective description of "I" — and is utterly inappropriate. The experience of being me is unitary, solitary and subjective.

It is easy to take the Objective View of the Duplicates Problem if you only think in terms of memories. If a person is suddenly duplicated to become person1 and person2, then it is not hard to see that although the two persons shared the same memories before the duplication, their memories will diverge and become different after the duplication — because they cannot possibly share the same location. However, the problem is not so simple if a distinction is made between memories and identities. If identities diverge, the survival of the identity of the original person one cannot be guaranteed — cannot exist in two places, even if all of the original memories now exist in two places, supplemented by new memories. New identities do not necessarily equate with survival of the old identity. The old identity does not necessarily survive if one of the new identities is killed.

Many people, including many cryonicists, can not (or will not) take anything but the Objective View of the Duplicates Problem. Such people deny that there is a Duplicates Problem and speak of duplicates of themselves with the same detachment with which they discuss the duplicates of others — holding that duplicating a person is no different than duplicating a toaster or a ballpoint pen. Such cryonicists speak of making "backup" duplicates of themselves as if this is a way to ensure survival. Subjectively this description is unsatisfying for other cryonicists because it does not address the personal experience of being in more than one location in one moment of time. But, more important, it does not address the issue of the personal qualia of selfhood and the issue of survival of personal identity. Which is not to say that a "backup" could not contribute to survival — any more or less than multiple atom-by-atom reconstructions of a new brains from a digitized template. "Backup" information could at least be of use in the reconstruction of the identity of a person who had been cryonically preserved. But that fact does not resolve the Problem/Paradox.

A Continuity Criterion has been proposed as a resolution to the Duplicates Problem. By this Criterion there can only be one original, and a duplicate cannot be the original because it cannot be in the original physical location and cannot possess the continuity-of-being of the original. A person who is "beamed" or "transported" through space by vaporization & reassembly of atoms has in fact been destroyed, according to the Continuity Criterion. The reassembled person may think he/she is the original, but he/she is not. Any attempt to create duplicates in such a way as to make them perfectly equal (as by splitting the original to create two duplicates, neither of which occupy the physical location of the original) destroys the original, and therefore destroys the original personal identity. No duplicate can ever be created in such a way as to have as much continuity-of-the-original as the original, therefore the original will always retain the original identity — and a duplicate can never acquire the original identity.

I argue against a "Continuity Criterion" which is based on the idea that the same molecules and the continuity of activity of those molecules are necessary for continuity of a personal identity. At least one life-extensionist believes that the Continuity Criterion renders cryonics impossible in principle. This man even believes that loss of consciousness means loss of personal identity — ie, if a person is revived from an unconscious state he/she will not be the same person who went unconscious. Unconsciousness is not the same as sleep. In sleep, brain electrical activity and metabolism can be nearly as great as in the waking state, unlike unconsciousness wherein electrical activity and metabolism may actually stop. A problem with his viewpoint is that it is not based on evidence. If a person recovering from unconsciousness claims to be the same person, how can we justifiably assert that this "new" individual is deluded? You can just as easily deny that a person awakening from sleep is the same person who previously fell asleep. The logical positivists would call such a claim "metaphysical" rather than scientific on the grounds that it is unfalsifiable in principle.

The Continuity Criterion attempts to resolve the Duplicates Paradox by saying that duplication is impossible in practice — an original can only be continuous with an original, and a duplicate will always be a new creation which cannot be continuous with the original in the same way that the original was. Either an original is distinguishable from a duplicate by virtue of the original location (thus possessing a "continuity" which cannot be claimed by the duplicate) — or two duplicates can be created in equally new locations, in which case neither can be said to be continuous with the original (which is deemed to have been destroyed). A technical problem becomes a philosophical problem. Thus, despite the fact that the duplicate may be constructed with every one of the duplicate's atoms having identical elemental form, relative position and relative motion to every other atom (as compared with the original) — the duplicate is distinct by virtue of discontinuity. Thus, the original self resides in the original.

The problem with this argument is that it claims that the duplicate is materially identical but that the original self can only reside in the materially identical original. But if the original and duplicate are identical in every material way, while at the same time differing in the constituent self, this implies that the self has a non-material basis. Materially identical means identical relative location of every atom and identical dynamic state of every atom (including state of electrical activity of the brain). (Some might argue that copies in different locations cannot have identical dynamic states due to different gravitational fields — but this difference cannot be critical to identity because location is easily changed without changing identity.) Thus, the Continuity Criterion of Identity is implicitly & covertly non-materialist, even if it claims to be a materialist argument.

The question of most practical significance to me is: is the creation of a duplicate a means by which I could survive? In particular, there are scenarios of cryonic reanimation in which the remains of the original cryopreserved person are destructively analyzed, and the results of this destructive analysis used for reconstruction from new materials. If one molecule is identical to another with the same composition of atoms, why should it not be possible to reconstruct the original person? This scenario is grounds for hope (and, perhaps, rationalization) for cryonicists. But, of course, many new people could be built using the same template. Why should any new person differ from the old one if only the particular atoms and locations are different?

The paradoxical intuitive answer to the question of whether duplication could be a survival mechanism is: "yes, but only if the original does not exist." This implies the Continuity Criterion of Identity — and cannot be regarded as an unproblematic solution. I do accept that survival implies continuity (continuity is a necessary condition for survival), but what I mean by this is continuity of the essential material factors of personhood, not continuity of non-essential factors like location.

The Continuity Criterion of identity is nothing more than an attempt to arrive at a definitive resolution to the Duplicates Problem in the absence of a good solution. But the existence of the problem does not prove that this is the solution, even if no other solution seems tenable. This is particularly true because the Continuity Criterion does not seem tenable. The Continuity Criterion is not materialist. If personal identity consists of the material molecular organization of the brain and the material dynamic states of the neurons, then there can be no material difference between duplicates.

It has been suggested that the Continuity Criterion renders all possibility of cryonic or chemical preservation impossible, even if suspended animation by these means could be achieved with no structural damage. This has also been suggested concerning the Dynamic States Criterion (ie, the belief that identity resides in active neural circuits that cannot be stopped and re-started without loss of identity) of personal identity. But if suspended animation does no structural damage, why would that break in continuity be more fatal than that of a good night's sleep? And continuity of what? In the suspended animation case, continuity of physical location is retained, whereas in the duplication example the dynamic states of neurons are perfectly duplicated. Thus, the Continuity Criterion for personal identity and survival of identity fails to give a material basis for what is essential for personal identity and personal survival.

It may, however, be true that it is not possible to restore the dynamic states of a person to what it was prior to suspended animation by low temperature vitrification. If this is true, then this is a technical problem involving a failure to produce true duplicates. It would not be necessary to invoke the Continuity Criterion to explain a failure to restore personal identity. What remains is a technical question of whether the dynamic states constituting personal identity can be restored. Why could science not "jump-start" these circuits once the brain is restored? In a sense, the reticular activating system "jump-starts" the cerebral cortex with bootstrapping positive feedback every time a person awakens.

Critics who object to the Dynamic States Criterion point to the fact that victims of barbiturate poisoning have recovered after a state of flat EEG (ElectroEncephaloGraph) recordings. People experiencing low-temperature surgery for brain aneurysms also show a flat EEG. Contrary arguments assert either that electrical activity continued which was too faint for the EEG to detect or that the recovered victim was a distinct new identity, even if the victim did not believe this.

A flat EEG is associated with the failure of axons to depolarize and conduct signals to an extent detectable at a gross level. But the capacity of axons to conduct signals should be retained. Analogously, hearts have been taken to low temperatures at which they are electrically passive — and they regain spontaneous automatic electrical activity upon rewarming. The existence of voltage potentials across cell membranes would also qualify as a brain electrical phenomenon, but there is no rationale for claiming that identity resides in voltage potentials across membranes. Restoration of sodium pump activity will ensure restoration of voltage potentials.

It is probably true that very short-term memory (the working memory — also called "scratch-pad memory" — of the frontal lobe of the brain's cerebral cortex, which allows a person to remember telephone numbers long enough to dial them) has a dynamic basis. But long-term memory has its probable basis in synaptic connections and the strength of those connections. It seems unlikely that personal identity would have a basis in the brain which is more like working memory than like long-term memory because the sense of personal identity lasts a lifetime and withstands the discontinuities of sleep. Moreover, if memories were based on dynamic activity of the brain, then the dynamic activity would increase with each new memory. This would be a highly energetically inefficient way to run a brain (which already consumes 20% of the body's energy, while constituting only 1% of the body's weight. But cumulative brain dynamic activity does not increase appreciably with the acquisition of new memories. Therefore, preserving the "hardware" (the neurons, fibers and synapses) should preserve the self.

When an impulse reaches a synapse, it is the structural features of the synapse which determines whether the next neuron will fire — ie, the number of vesicles of neurotransmitter in the synapse, the size of those vesicles, the number of similar pre-synaptic connections, etc. Learning in the brain occurs by alteration of these structural features. Thus, the dynamic states of the brain are determined by the static states — ie, function is determined by structure. This implies that the new dynamic states following the stasis of low-temperature cryopreservation must not be like the old static states.

Because personal identity is retained with the passage of time, while static states change, the question arises as to how personal identity could be retained in the face of these changes. The neurological basis of personal identity must be changeless in some way. Possibly either personal identity is encoded in neural circuits that are always active — or personal identity is distributed throughout the brain — or both. The former supposition comes uncomfortably close to the homunculus paradox — ie, the supposition that a "little person" somewhere in the brain "sees" information from the visual cortex, "hears" information from the auditory cortex, etc. This model implies another little person inside the first little person — an infinite regress that begs the question of identity.

It is conceivable that a person could be created possessing the same memories, desires and personality as another person, and yet still have a distinct personal identity. But if this is true, there must be material differences in the brains. What those material differences would be, certainly cannot be specified by contemporary neuroscience. If a brain imaging procedure existed which could show these material differences, then there would be objective criteria for personal identity. Otherwise, a person with the same memories, desires and personality of another person will believe he or she is that other person and (if the two have identical appearances) others will also be unable to distinguish. Nonetheless, this state of affairs is only a possibility given current knowledge of personal identity and neuroscience.

Caution should be used in discussing cases wherein it is claimed that people showing identical memories are distinct identities, but do not believe that they are distinct identities. To assume that the Continuity Criterion must be true and then conclude that loss of consciousness means death of personal identity makes the Continuity Criterion an unfalsifiable hypothesis. If one denies assertions of a person who has recovered from unconsciousness that he/she is the same person who existed before the loss of consciousness, then one can as easily deny claims that a person remains the same after a full night's sleep or even a change of physical location.

An alternative to the Continuity Criterion as a solution to the Duplicates Problem is the idea that some Random Event determines that the original identity will reside in one duplicate or another. Although this satisfies the subjective paradox, it raises other questions. What materialist grounds are there for the original identity residing in one duplicate or another? What is the material basis of the original identity? Why cannot the original identity be duplicated if it has a material basis (in fact, it should have been duplicated by the original statement of the problem). This "random" solution contains the covert assumption of some unique non-material basis of identity which is not being duplicated. Moreover, like the Continuity Criterion, the Random Event solution is an assumed solution to a thought experiment.

The same can be said of the Idealist Interpretation of the Duplicates Problem. The Idealist claims that any number of instances of a material mind will still correspond to only one instance of mind, presumably in an "ideal form". The claim here is that only one "ideal triangle" or "ideal chair" exists, whereas many instances can exist (drawn on paper, on sand, a physical chair, etc.). Since a mind is composed of ideas, the ideal form is said to constitute its real essence. This is not a materialist interpretation — which implies that it is a spiritualist one. There are extreme practical problems with the idea that many instances of one mind could be singular (divorced from any physical location) and yet subject to experiential input from many different bodies in different locations.

I do not see a solution to the Duplicates Problem, but I reject the Objective View, the Continuity Criterion, the Dynamic States Criterion, the Idealist Interpretation and the Random Event approaches as being unsatisfactory. If I am ever duplicated, then "I" will have more evidence with which to study the problem. At present, I must simply add the Duplicates Problem to my long list of other problems for which I have no immediate solution. For others, it seems to be unacceptable to acknowledge that a philosophical problem exists without a reasonable solution (although such persons can readily acknowledge technical problems with no immediate solution). And it is easy to settle on a solution when there is little possibility of concrete evidence to support or refute it.

(For an entertaining video that deals with the duplicates problem, see John Weldon's "To Be".)