From Cryonics, 4th Quarter 1994 and 1st Quarter 1995
We refer to persons in cryonic suspension as "patients" and not as bodies. Most people will insist we are wrong, that cryonauts are dead – in the sense implying they can never return to life. This essay, which forms Chapter 2 of my Ph.D. dissertation (The Diachroic Self: Identity, Continuity, Transformation), refutes the nay-sayers, proposing an alternative conception of death to that held by most people, including medical ethicists.
This essay assumes an earlier explanation of a position known as "psychological reductionism". Since this essay appears without the earlier chapter, let me briefly explain what a psychological reductionist claims about personal identity (or continuity, or survival):
Psychological reductionism expresses essentially the same view as that which Mike Perry and Ralph Merkle have called the informational view of identity. Most cryonicists who have expressed an opinion appear to hold this view. Ignoring many fine points, we can say a psychological reductionist holds that what makes you the same person as you were ten years ago or fifty years from now is a relation of psychological connectedness and continuity between these stages or phases of you. An earlier and a later stage form stages of one continuing person if sufficient psychological conditions exist between them. Psychological connections include memories, intentions, dispositions, abilities, values, principles, and projects. Your earlier and your later stages will be connected to the degree that they share these attributes. Even if the person has changed a great deal, leaving only weak psychological connectedness, the person will exist continuously so long as we can trace a chain of overlapping psychological connections.
On this view, the continuity (or continued identity) of a person depends on continuity and appropriate kinds of transformation of character or personality. Identity does not depend on the later stage having the same "soul" as the earlier person. Nor does it require that the later person-stage has the same body or the same atoms as the earlier. Nor need we require a continuous, uninterrupted stream of consciousness.
You can find most of the other unfamiliar terms in the Glossary. I apologize for any difficulty encountered as a result of the academic style of writing. I have tried to express myself clearly while also satisfying the meticulous requirements of my professors.
Concepts of and criteria for death have changed throughout history. A concept of death purports to tell us what death is, while a criterion for the occurrence of death is the sign or signs by which we determine that a person has reached a state of death. The traditional concept of death was left vague until recently, its criteria involving cardiac and/or respiratory function but without explicit distinction between criteria and concept. More recently, the concept underlying the cardio-pulmonary criterion has been defined as the loss of integrated organic functioning of the body (Lamb, 1985). In the current century there has been a partial shift to a brain criterion, though this is usually added to the cardio-pulmonary criterion rather than replacing it. Death is now often supposed to be determined by the death of the whole brain, or the brainstem, which is responsible for maintaining integrated organic functioning.
A more recent proposal for a brain-criterion is the neocortical criterion (Gervais, 1986). As Karen Gervais argues, the neocortical criterion marks a clear shift to a different concept of death. The neocortical criterion (and, more controversially, other brain criteria) moves us from a cardiac-centered to a consciousness-centered concept. According to this new definition of death, a person is dead only when their capacity for conscious thought, for the functioning and expression of their personality, has been lost. Defined this way, it will be clear that the death of part or all of the body is only instrumental and not intrinsic to the death of the person. I accept this new conception of death and will analyze it further, but will modify the neocortical criterion proposed by Gervais. I will argue that this may not always be an adequate criterion even today, and may frequently be insufficient in the reasonably foreseeable future. The need for the category of deanimate will emerge when I consider cases for which the neocortical criterion is inadequate. Some of these cases will be familiar from the personal identity literature.
Proponents of the neocortical criterion and I agree that a proper understanding of death requires a distinction between the human organism and the person. Death of the body or parts of the body concern us only in so far as they bring about the death of our selves. Our selves die when we lose the capacity for conscious thought, i.e., when we can no longer think, feel and express emotions, have desires, form plans, and further our projects. I won't argue this point further here, since it surely follows straightforwardly from a psychological reductionist view of the self.
An initial example of how personal death can diverge from bodily death is in order here: There are cases where a human body is alive and functioning, meeting the criterion of integrated organic functioning, but where the person is dead. This is the situation in which the neurons of the higher brain have been destroyed (neocortical death) thereby removing the possibility of conscious thought, while the brainstem and perhaps the cerebellum, thalamus, and basal ganglia continue to function along with the rest of the body.
Where I differ from Gervais is in my understanding of the conditions essential for the loss of the capacity for consciousness and personality. Gervais expresses her particular conception this way: "[H]uman death, understood as the death of a person, is a state in which the function of consciousness has been irreversibly lost as a result of one of several possible combinations of damage to the brain substratum" . A second statement contends that "[T]he individual's essence consists in the possession of a conscious, yet not necessarily continuous, mental life; if all mental life ceases, the person ceases to exist; when the person ceases to exist, the person has died" [157-58]. While I agree with these statements, I don't think that they lead us to a neocortical criterion for all cases. This is because I understand the irreversible loss of the capacity for consciousness to require the irretrievable loss of personal identity-critical information, and this need not follow from irreversible loss of neocortical function. I will explain this claim in detail below, but will first try to head off possible confusion by exposing an ambiguity in the meaning of 'dead'.
The everyday concept of death, as well as some more refined theoretical concepts, harbor an ambiguity. Failure to recognize this ambiguity leads both to indeterminate concepts of death and mistaken criteria for the occurrence of death. Whether applied to a person or a biological organism as a whole, or to a part, we can distinguish dead1 (functionally dead, non-functional) from dead2 (irreversibly dead). Dead1 means "absence of function" and an assertion that X is dead1 is equivalent to the denial that X is 'alive' or 'living'. Life is a certain kind of dynamic, functional process; if that process ceases then the entity is dead1. This is the straightforward sense in which you might say "My car is dead," or "My computer just died." This has no implication that your car will never work again. Perhaps a spark plug needs replacing, or a connection to your computer's power source is loose.
Dead in the second sense, dead2, has a stronger implication: It requires irreversible loss of function. Suppose that at about the time your car stops running you become rich. You might decide to junk the car rather than have it repaired. You watch as your ex-vehicle is crushed into a thin slab of metal. Now, if you declare "My car is dead," you mean that it is dead2, i.e., the same car cannot be returned to you. There is not enough left of the structure of the car to repair it and make it functional. At best, some of the metal could be used to build a new car of the same model. But that would a different car.
Since most people are religious, the common conception of death has been influenced by religious myths of an afterlife. This is another source of ambiguity. The very term 'afterlife' hints at the ambiguous notions of life and death inherent in religious dogma. 'Afterlife' suggests a time or place that is not life, yet neither is it death. In the Christian tradition, stories are told of Jesus resurrecting the dead, and reincarnationists talk of people dying and then being reincarnated. These examples support the idea that the common notion of death, of being dead, is not entirely a notion of irreversible cessation of consciousness. If religious people understood death as irreversible loss of consciousness, they would describe paradigmatic cases of death as continuing life in another form, and would deny that the person had really died.
I don't want to rely too heavily on religious usage in making a case for the ambiguity of the common concept of death. We could rescue the religious notion (if not its users) from ambiguity if we took statements such as "He's dead" to mean only "He's dead to this world," or "He's physically dead, but truly lives on in Heaven." On the other hand, most religious people have not thought this through to the point of disambiguating their usage. To the extent that the meaning of a term is determined by usage, the unreflective religious use of 'death' does contribute to the ambiguity of the common conception of death. The concept's use even by many non-religious people will reflect the same ambiguity. While the finality of death is reflected in the use of phrases like "Dead and buried," we also see pervasive use of the idea of people dying and then being brought back to life, especially in fiction, whether it be traditional tales such as Dracula, or recent movies such as Flatliners.
Dead2 - irreversible loss of function - is synonymous with the final state of death. Due to the contexts in which it is used, 'death' apparently lacks some of the ambiguity inherent in 'dead'. Death has more of a ring of finality and irreversibility. Some thing, or part of a thing, can be dead1 but not have reached a state of death. When it is dead2 then we can also say that it has reached death. Throughout this paper, when I use the term dead without a qualifier I will mean dead2 - irreversibly or informationally dead. In a later section I will propose a term for dead1 so that the distinction will be clearly reflected in the terms we use. This distinction has been recognized by some writers on death, such as Byrne et al (1979) who objects to brain-related criteria on the ground that loss of brain function is not synonymous with destruction of the brain. While destruction is irreversible, loss of function may sometimes be reversible:
There is no evident contradiction in supposing the existence of permanent synaptic barriers, permanent analogs of botulinus toxin or morphine, or yet other mechanisms that would block all brain-functioning while leaving the brain's neuronal structure intact and ready for action (at least until such time as the effects of this non-function on the rest of the body might react back on the brain in a destructive manner). Therefore there is no reason to think that cessation of function, whether reversible or irreversible, necessarily implies total or even partial destruction of the brain; still less death of the person. [p.1987]
These two ideas have not always been clearly separated. Distinguishing the two senses of 'dead' will allow me to reveal a vagueness in the notion of neocortical death. But I will argue that even a clarified notion of neocortical death is inadequate, since it fails essentially to mark the boundary of irreversible loss of personality. Supporting this claim requires a precise explication of various types of continuity and of the relevant notion of irreversible loss of continuity that will follow. First, I will reveal a second indeterminacy in the usual notion of death, and distinguish this notion from my stricter, stipulative notion of death. My claims about the correct view of continuity in a condition of irreversibility will apply to the stipulative definition, but not necessarily to the indeterminate notion.
In addition to failing to distinguish loss of function from irreversible loss of consciousness, standard conceptions of death contain a further indeterminacy: A failure to disentangle the idea of permanent absence or loss of consciousness from the idea of irreversible loss of consciousness. Permanence and irreversibility are distinct and separable since cessation of consciousness might be permanent yet reversible. Every day, patients are "no-coded" by doctors and declared dead. In no-coding a patient, the attending physician is saying that though the patient could be resuscitated (by CPR or defibrillation), this is not to be done, since the patient's restored life will be brief and unpleasant. Where a no-code instruction has been issued, cardiac arrest entails permanent loss of consciousness. Yet it might be quite easy to resuscitate the patient, at least temporarily. Since the standard notion does not sharply distinguish permanence and irreversibility, we can set out a disjunctive conception of the occurrence of death:
Death of a person occurs when
(a) Irreversibility condition: There is a sufficient degree of destruction or dissolution of the brain (or other medium for support of consciousness);
(b) Permanence condition: The capacity for consciousness is lost and no attempt will ever be made to revive or repair the patient.
We need not hold that only the irreversibility condition is correct, although I will argue that it is more fundamental. Instead we can distinguish two senses of the term 'death' and give them each labels. This will allow me to provide a theory of the type of continuity necessarily involved in the irreversibility condition while granting a role in the standard concept to the permanence condition.
A person is theoretically dead if they meet the irreversibility condition.
A person is permanently dead if they meet the permanence condition, whether or not they also meet the irreversibility condition.
Permanent death occurs when a person is permanently lacking in consciousness. Such an assessment need not coincide with irreversible loss of consciousness. In many instances the two do not coincide, as in the no-coding case above, as well as in more speculative cases such as biostasis. Some individuals, following clinical death, have been placed into biostasis (specifically) cryonic suspension in the belief that they might be resuscitated in the future by more advanced medical technologies. Suppose that this procedure does preserve a person sufficiently well, and that the necessary future repair technologies will be developed. Now, suppose someone had a heart attack and became clinically dead, i.e., their cardiac and respiratory functions ceased, but the decision had been made not to place them into biostasis. Then we could say, immediately after the coronary, that the person had permanently lost consciousness even though they had not lost consciousness irreversibly.
So, there are a range of situations in which permanent and irreversible loss of consciousness are not identical. Permanent death is partly determined by the decisions we make and the actions we perform. This means that permanent death is not fully objective in the way that theoretical death is objective (i.e., independent of decisions and actions). Irreversibility, in the sense I am using it, refers to loss of the capacity for consciousness that cannot be reversed even in principle. No matter how much technology may advance, and no matter how different the medium for support of consciousness may become (embodied in computers, for example), theoretical death refers to a state beyond any possible capability to reverse.
Permanent loss of the capacity for consciousness may appear to be an objective matter also: Either someone will be returned to consciousness at some point or they will not. Our beliefs regarding the probability of resuscitation are subjective, just as are our beliefs about reversibility, but surely the permanence of lack of awareness is an objective matter? Not if by 'objective' we mean "independent of human action and decision." Suppose cardiac and respiratory function cease in Smith. To an uninvolved third party the permanence or lack of permanence may be a factual, objective matter. Either Smith's bodily functions will restart spontaneously or they will not, and either someone else will successfully intervene with CPR or defibrillation or they will not.
However, from the point of view of someone in a position to medically intervene (call her Robinson), the permanence of Smith's condition cannot be regarded as determined independently of the intervener's decisions and actions. (This is assuming that intervention has a more than zero probability of success, otherwise intervention is futile even if Robinson believes otherwise.) Robinson cannot regard her own actions as already determined; she has a genuine decision to make. To Robinson then, if not to an uninvolved observer, the permanence or transience of Smith's loss of consciousness is not fully objective. The same point can be made using Byrne's example above, in which the brain's capacity to function has been blocked. In that case, the permanence of the condition may depend on the decisions and actions of researchers and medical personnel to take steps to reverse the condition. Finally, in the cases of persons placed in biostasis for possible future repair and resuscitation, the permanence or transience of their condition may depend on the actions of those maintaining them in suspension, on researchers seeking to develop repair technologies, and on legislators who may choose to prevent such research or the revival of the biostatic persons.
Permanent death and theoretical death both may involve a shift in our attitudes toward the person. A belief in the person's permanent or irreversible loss means that we will no longer think of the person interacting with us in the future, or having further experiences. We will no longer include them in our plans. This shift in attitudes will be reflected in our customs and in the law. The rights and status of the deceased person will change: They can no longer be rewarded or punished, cannot make contracts, and will not be considered in our plans for the future. However, these attitudinal changes are more closely tied to permanent than theoretical death since they will occur once we have decided that absence of consciousness is permanent even though we may know that consciousness could be restored. That is, attitudes will not shift if we do not believe them to be permanently gone, but attitudes will shift if we believe them to be permanently though not irreversibly gone. The fact of theoretical reversibility will not affect our attitudes once we are certain that the theoretical possibility will never be acted upon.
Irreversibility has (or should have) priority over permanence in determining our attitudes: An assessment of irreversible loss of consciousness ought to lead us to an assessment of permanent loss of consciousness. But the reverse is not true; a belief in someone's permanent lack of consciousness need not require us to believe that they irreversibly lack consciousness. We would only believe "if permanently then irreversibly nonconscious" if the two were identical in a particular instance, as when a person's brain is instantly destroyed. Since decisions to allow reversible cessation of consciousness to continue should be founded on our beliefs about possible reversibility of nonconsciousness, the reversibility condition is the more fundamental to us as agents. This makes it important to be very clear about the limits on reversibility. If we misunderstand these limits, and believe loss of consciousness to be irreversible when it is not, then we risk acting, or failing to act, in such a way that we cause someone to become permanently dead unnecessarily (by burying or burning them rather then treating or maintaining them).
Having distinguished the permanence and irreversibility conditions, I can now focus on the universally applicable irreversibility condition. Various types of irreversible cessation of consciousness might be thought essential to theoretical death. I will argue that the correct condition is irreversible loss of informational continuity. In defending this condition I will deny the universal applicability of the neocortical criterion, even if it embodies the irreversibility condition rather than the permanence condition, and even if it embodies destruction of the neocortex rather than loss of function.
Gervais' reason for proffering the neocortical criterion for death is clear enough: "destruction of the neocortex has been shown to produce permanent unconsciousness and to be an empirically verifiable pattern of brain destruction prior to the failure of the organism as a whole. Since human death is the death of the person, and the death of the person occurs with permanent loss of consciousness, neocortical death is an adequate criterion for declaring death" [150-51]. And, a few pages later: "[T]he individual's essence consists in the possession of a conscious, yet not necessarily continuous, mental life; if all mental life ceases, the person ceases to exist; when the person ceases to exist, the person has died. Upper brain death destroys all capacity for a conscious mental life, and it is therefore the death of the person." (pp.157-58.) I will agree that the neocortical criterion, when carefully stated, is an adequate criterion for present day conditions, but will argue that it will not serve as a universally valid criterion. To establish this, I need to show that persons can continue to exist despite being neocortically dead (in either sense). To this end I will distinguish different types of continuity and evaluate their relative importance for the continuation of the self.
The neocortical criterion is not a universally applicable criterion of death. Gervais would probably agree with this, since she is open to further refinements in our criteria,3 and she accepts that a person embodied in something other than a human body could be the same person. As I will show below, this means a person might survive the destruction of their brain. The neocortical criterion is merely a normally reliable criterionin 1995for diagnosing death or for making a prognosis of death (depending on whether neocortical death is taken to mean loss of function or destruction). A presumption of universal applicability would be acceptable if it were impossible for a person to survive neocortical death. However this is not necessarily the case. Whether we are to understand "neocortically dead" to mean permanent loss of neocortical function or destruction of the neocortex, selves may perdure regardless. I will examine both senses in which someone might be said to be neocortically dead and show that neither are acceptable criteria.
Loss of neocortical function: To say that someone is neocortically dead, or that they have lost the capacity for consciousness, might mean that the neocortex has ceased functioning and it cannot be restarted with available technology; or it might mean that the neurons of the neocortex and their patterns of interconnectivity have either decayed or been destroyed so that no empirically possible future technology could repair them. Which of these Gervais is using is hard to determine since she never actually gives her own definition of neocortical death; she cites definitions found in the literature, without pointing out that they are not equivalent. (She also appears to use permanent cessation and irreversible cessation interchangeably.) Two prominent definitions cited on pages 11-12 involve destruction of neocortical or apallic neurons; but the definition quoted from J.B. Brierley (p.13) is a function-based definition. Brierley states that neocortical death "implies a persistently isoelectric EEG and the absence of sensory evoked responses in the neocortex, together with the resumption of spontaneous respiration and of certain brainstem reflexes."
Of course, given today's standard practices, a patient who is neocortically non-functional will eventually become neocortically destroyed, though this may take hours or days, even without cooling to slow enzymatic degradation. But this is no reason to conflate the two, as Gervais clearly recognizes in the context of a parallel criticism of Lamb: Lamb claims that "the death of the brainstem is the necessary and sufficient condition for the death of the brain as a wholeand that brainstem death is therefore itself synonymous with the death of the individual." Gervais correctly argues against Lamb that "To say that the loss of integration becomes irreversible is not to say that the loss has occurred." (Gervais, p.148) Irreversibility of a function that leads to brain death is prognostic, not diagnostic of brain death. Now, under standard conditions (in the past and present) cessation of neocortical function is prognostic of personal death, but it is not diagnostic unless accompanied by neocortical destruction. If death is an irreversible state, then cessation of neocortical function that is irreversible by current medical technology is no more the point at which death occurs than was cessation of heart beat in the past. If we were to find a way of restoring a "dead" (non-functional) neocortex to function, then we would have to say that the person had not been dead.
So long as the necessary neuronal structures persist we cannot say that the capacity for consciousness is irreversibly gone. Cessation of neocortical function need not imply loss of critical structure or information: Sufficient structural and chemical clues may remain to allow restoration and revitalization of neocortical function and neuronal interconnections. Full structural continuity of the cells is unnecessary for the possibility of repair of the neocortex since the desired structure and function of the neurons may be inferred from residual chemical clues, or it may only be necessary to repair membranes, open ion channels, restore synapses, or replace organelles such as ribosomes.
An objection might be raised to the effect that "capacity for a conscious, but not necessarily continuous, mental life" means that the neocortex can support consciousness given the appropriate stimuli and that these stimuli should be defined in terms of current technology. An analogy might be given as follows: If we say that a car has the capacity to move at 110 mph, we mean that it is currently in a state such that, given appropriate stimuli (such as gas, a foot on the accelerator, etc.) it will achieve 110 mph. The objection claims that we do not mean that the car could achieve 110 mph given available technology, and we do not mean that, given some empirically possible but non-actual technology, the car could achieve 110 mph. The problem with the objection lies in the fuzziness of the terms 'capacity' and 'appropriate stimuli.' Does the car have the capacity to move at 110 mph if a wire has been loosened? In that case it does not have the capacity immediately, given only the normal stimuli. However, there is a perfectly reasonable sense in which it does have such a capacity: The car has the capacity to move at 110 mph if we reconnect the wire. If someone were to say, before reconnecting the wire, that the car could not go 110 mph, the statement would be misleading in that we might be led to think that this kind of car does not have that capacity. The car will not function normally without that repair, but so long as the repair can be effected there is an important sense in which the car does have that capacity.
Suppose the car has suffered some form of damage to its components so that it cannot move, and currently no way exists to replace or repair the components to restore function. Further suppose that the manufacturer tells you that they are working on a new repair process that will restore function, a process that should be available a month from now. We will probably want to say that the car does not have the capacity to move at 110 mph, but that it can potentially regain that capacity. If we say this, it follows that loss of (current) capacity to function does not imply irreversible loss of function. If asked whether our car is dead, in the sense that it can never function normally again, we should reply in the negative. The car analogy, then, supports rather than undermines the case for basing a criterion for death on irreversible loss of capacity rather than currently irreversible loss of capacity.
In the neocortical case, if Gervais's criterion for death is loss of the capacity for consciousness (due to loss of neocortical function), then we can see that her criterion is not equivalent to the irreversible loss of the capacity for consciousness. Loss of neocortical function may be currently irreversible, just as loss of cardiac output once was irreversible with existing technology, but death does not occur (at least) until the neocortex has been destroyed, or degenerated beyond any empirically possible means of repair. For Gervais to deny this would also require her, contrary to her stated view, to claim someone to be dead as soon as their heart has stopped beating (and consciousness has been lost) if there is no available (or known) means of restoring cardiac function. If the neocortical criterion is to serve as an accurate criterion in the present, it must therefore be interpreted as neocortical destruction rather than currently irreversible loss of neocortical function. With this condition specified, and with the exceptions discussed below (in the Deanimate section), I can accept the neocortical criterion as an adequate criterion for death in 1994 and the near future.
The period between cessation of neocortical function and true neocortical death (loss of structure) might seem to be of merely theoretical interest but of no contemporary practical significance since we cannot now restore neocortical function, just as pre-20th Century it might have been claimed that there was no practical significance to the fact that a person with a still heart was not yet dead. Such a claim would be mistaken. Attending to the difference between currently irreversible loss of function and true neocortical death will encourage the search for means of preventing neocortical decay by preserving the neocortex in an unchanging state, and for means of repairing the neocortex and restoring its function. (Again, see the section on Deanimate below.)
Neocortical destruction: By neocortical death, Gervais might mean not loss of function but decay or destruction of the neurons of the neocortex and their patterns of interconnectivity so that no empirically possible future technology could repair them. (This is unlikely to be Gervais' intended meaning, if we interpret "the capacity for consciousness" to mean that consciousness can be restored with currently available means only.) This is less parochial than the loss of function definition and is acceptable as a historically temporary criterion (i.e. given current technology), but it still fails to provide a transhistorical, universally applicable criterion. Locking the criterion of death into neocortical destruction is mistaken since, as I have argued in earlier chapters, our continuity is essentially psychological continuity and connectedness (the R-relation of Oxford philosopher Derek Parfit) and not physical continuity. We might say that we are software and not hardware; the psychological relations that are me are currently instantiated in this neocortex, but I am not essentially this neocortex nor even (more controversially) any neocortex. We can conceive of personal continuation despite neocortical death, and this may even become technologically possible in the future. Here are a couple of ways in which neocortical death and personal death could come apart:
Brain Scanning and Replacement: Suppose that, at some time in the future, some extremely powerful scanner were available, the descendants of today's MRI, NMR, PET, SQUID, SPECT and CAT scanners. These scanners might be used to scan a brain so completely that the resulting data specified the entire neuronal structure, including neuronal interconnections, electrical charges, spiking potentials, and levels of all neurotransmitters and hormones. Suppose that your brain was then destroyed (or is destroyed layer by layer as the scanner does its work). A new brain is then built according to the information gathered from the scan, it is implanted in the original body, and all necessary connections are restored. We should say that this brain is a new brain, for a brain is a physical object and spatiotemporal continuity is a necessary condition for physical objects. (If we were to destroy and replace only a small fraction of the original brain at any one time we would probably say that the same brain remained throughout.)
Despite the spatiotemporal discontinuity and the destruction of one brain and its replacement by another, the same person remains throughout the procedure. Though there is an interval during which there is no structural or functional continuity, there is always informational continuity. The new brain is structured the way it is, and functions the way it does, because of the structure and function of the original brain. The same person persists through this procedure because of the non-accidental causal connection between the structure and function of the new brain and that of the old. We may be puzzled by how to describe the condition of the person during the interval between the destruction of their original brain and their revival in the new brain. They are not dead, but we may not want to say that they are alive. I will return to this issue in the next section.
Uploading: In the second kind of case, I can survive the loss of my brain even though it is never replaced by another biological brain. If what matters in my survival is my psychological continuity, then I will continue to exist so long as my consciousness, my psychological features, are maintained in hardware that is functionally equivalent at the necessary level. This hardware may be nonbiological, perhaps an appropriately-configured parallel-processing computer constructed according to the information gained from the destructive scanning of my brain. The transfer of a person's consciousness from their brain to a computer is referred to as "uploading" and has been described both in fiction and nonfiction. This informationalist conception of personal continuity is expressed by Daniel Dennett:
If you think of yourself as a center of narrative gravity, on the other hand, your existence depends on the persistence of that narrative (rather like the Thousand and One Arabian Nights, but all a single tale), which could theoretically survive indefinitely many switches of medium, be teleported as readily (in principle) as the evening news, and stored indefinitely as sheer information. If what you are is that organization of information that has structured your body's control system (or, to put it in its more usual provocative form, if what you are is the program that runs on your brain's computer), then you could in principle survive the death of your body as intact as a program that can survive the destruction of the computer on which it was created and first run. [Dennett (1991) p.430]
In the brain scanning and replacement and uploading cases the later person-stage is psychologically continuous with the earlier person-stage. If personal continuity is psychological continuity then we cannot say, in these cases, that someone has died and been replaced by another person. The very same person remains throughout, despite the discarding of the hardware that previously embodied their capacity for consciousness. To someone who wasn't aware that our subject had received a new brain, or had a silicon or optical brain-replacement in their skull, no difference would be detectable, making it absurd to think that the original person had died.
We have seen that a person is not dead immediately following cessation of whole brain or neocortical function. Now I will argue that a person is not properly described as dead, at least in the theoretical sense, while they are incapable of consciousness but where their brain (or replacement hardware) retains the structure allowing possible restoration of that capacity. Going a step further, I will show that a person is not dead while they exist only in the form of information lacking a functioning embodiment. Though they are not dead, we may not be comfortable describing them as alive. Indeed, it would be inaccurate to describe them as alive. We need a new category to describe these cases – that of deanimate; I will divide this category into the subcategories deteriorating, biostatic and inactivate.
On the theoretical view, and more ambiguously on the standard view, absence of life is not co-extensive with death. We can see this by attending to the differing ideas of continuity defined above and by a better understanding of the nature of life and death. By 'death' I mean the end point of the gradual dissolution of a living system. Dying is the process that takes an organism from life to death, from a process of living to a state of death. We should reserve the term 'death' for the result of the process of dying, otherwise we will have no term to make that reference. If we were to say that death had claimed someone as soon as their heart stopped beating, or their brain ceased functioning, we would no longer be able to clearly differentiate very different conditions – those of various kinds of biological and neurological cessation of function from those of loss of any present or future possibilities for restoring function. Death is therefore the state a person is in when they are theoretically dead.
Almost no one seems to willing to venture a definition of life. According to Prigogine and Stengers (1984), a living organism is an open system in which matter and energy are exchanged with the environment. A living organism is a dissipative structure: a dynamic state of matter which originates in far-from equilibrium conditions and involves a close association between order and structure on one hand and dissipation or waste on the other. Living creatures keep internal entropy at bay. When they fail to do this they cease coherent functioning and proceed down the necrotic path of increasing dissolution. Death is the end result of this entropic process. This dying process may be arrested and reversed at some stages; how far along the process arrest and reversal can be achieved depends both on how much critical information remains or can be reconstructed and on the level of technology.
Since there is a large gap between cessation of function (whether cardiac or neurological) and loss of all structure and structure-critical information, we need a term to refer to that part of the spectrum in which the person potentially is fully or mostly recoverable. Deanimate can fulfill this need. It connotes absence of movement, cessation of life, leaving us the term "dead" (in our theoretical sense) to exclusively refer to an organism that has reached a state of death, without connoting further decay. When someone's heart stops beating, or their brainstem ceases to integrate bodily functions, that person becomes deanimate. They lose consciousness and cannot spontaneously recover. When they become deanimate, the dying process will continue until they are dead, unless other persons intervene. Such intervention may consist of cardiopulmonary resuscitation, if deanimation has occurred in the last few seconds up to something under an hour, or it may take the form of biostasis (cryonics, vitrification, or advanced suspended animation).
In the case of biostasis, the deanimate person's physical embodiment is stabilized, by locking their constituent molecules in place at extremely low temperatures or by chemical fixation. Biostasis is thus a sub-category of deanimate; biostatic persons are deanimate persons whose dissolution has been arrested (before reaching a critical stage). If the technology becomes available to repair the life-threatening condition causing deanimation, and to reverse the changes caused by the biostasis technique (which may itself add further injury), the biostatic-deanimate person may be restored to life.
Both biostatic and inactivate fall under the category of deanimate but they may usefully be distinguished. In the previous section I described a hypothetical scanning procedure that gathers thorough structural information about the brain. This information might be stored on an inactive, non-biological medium for a period prior to reembodiment of the information in a working brain. In that case, the person would be deanimate and inactivate but not biostatic. Inactivate persons will be thought of as further from being alive than will biostatic persons. There are two related reasons for this: First, while both biostatic and inactivate persons lack the capacity for consciousness, the former exist in their standard embodiment, though in static form. But inactivate persons persist only as information instantiated in a form radically different from that of their standard body. Technology capable of resuscitating biostatic individuals probably is less remote than technology capable of re-embodying an individual from their identity-critical information. Second, these differences in form and temporal remoteness are likely to generate differing attitudes in us. On the permanence view of death, we are more likely to regard an inactivate person as dead than a biostatic person. We will find it harder to regard someone as not dead who persists only as nonbiologically-embodied information, and whose revival we think only remotely likely. This different makes it worthwhile distinguishing biostatic from inactivate within the general category of deanimate.
We can apply the category of deanimate on both the theoretical or the permanence views of death, though its extension will differ between the two. On the theoretical view, when a person ceases living and loses the spontaneous capacity for consciousness, they become deanimate. They will continue to die until they are dead, unless they are put into biostasis or otherwise preserved. So, on this view the person is first alive, then deanimate and degenerating, then deanimate and static, and later, possibly, alive again. The permanence view differs in that when life ends, if we believe the person will never be returned to life, then we will say that the person goes immediately from being alive to being dead. This allows no room for a period describable as deanimate. The permanence view will agree with the theoretical view in allowing for a period of deanimation only where we believe that the person will eventually be restored to life. So, deanimation is compatible with the permanence view but only where cessation of life is thought to be temporary.
For some, it will be difficult to shake the feeling that someone who remains only as inactively stored information is dead. I think the feeling of oddness felt by some in thinking of an inactivate person as not dead is largely generated by the entrenched standard belief that not-dead = alive. Since an inactivate person, even more than a biostatic person, is far from alive, the temptation to think of them as dead can be strong. The feeling of oddness should be dispelled by keeping the third category of deanimate in mind. With that category in place, it will seem even more odd to think of inactivate persons as dead. Inactivate persons and dead (ex-) persons differ in important ways. In the former case, a great deal about the living person is being preserved; all the information specifying their psychology and its physical embodiment persists, and ex hypothesi that information could eventually be used to restore the person to life. All the knowledge and experience of the original person remains, though as potentiality rather than actuality. The fact that the person experiences a break in continuity of consciousness is not a reason to say that they died and will be replaced by a different person if the information is re-embodied. If we were to say this, we would also have to say that persons who go into a coma and later are revived have died and been replaced.
Why not think of the situation as the person's death and replacement? First, we need to remember that if we believe someone to be inactivate it means that we believe it to be empirically possible, now or in the future, to restore them to consciousness. Regarding them as dead has a number of effects: We will cease to think of them as involved in our lives in the future. We will make no efforts to return them to life, instead treating their inactivate form as equivalent to the ashes of cremation. We will withdraw all their rights and disregard their interests in continuing life in the future. The pre-inactivate person may have made plans for the post-reactivation stage of their life, but our regarding them as dead will destroy their plans far more completely than if someone were to destroy all of another person's assets. There is no obvious difference between treating an inactivate person this way and disconnecting a comatose patient from life support when it is believed that they could be brought out of the coma with treatments to become available in the future.
If we believe that the inactivate person will never be restored to life, perhaps because they did not provide the funding to do so and no one else will be willing to bear the expense, then we will correctly regard the person as permanently dead, even though their condition is potentially reversible and so they are not theoretically dead. This kind of situation helps to explain intuitions to the effect that inactivate persons are dead. Inactivate persons who will never again be alive can indeed be regarded as dead – as permanently but not theoretically dead. Maintaining the distinction between the theoretical/reversible and permanence conditions will distinguish the two classes of inactivate persons, those who are dead in the permanent sense and those who are not dead in either sense, and thereby helps dissolve any intuitions about the deadness of inactivate persons as a class.
Although the concept of death has now been defined more precisely by distinguishing it from deanimation, an indeterminacy remains due to the psychological reductionist theory underlying these distinctions. We may expect it always to be simple to tell when someone is dead, at least in principle. If death is put in the context of psychological reductionism, this clarity evaporates as soon as we realize that such continuity is a matter of degree. The R-relation includes both psychological continuity and connectedness. The same person continues only if a person-stage has enough psychological connections to the preceding person-stage. There is psychological continuity between person-stages, i.e., the person survives, if those stages are strongly connected. Parfit stipulates that strong connectedness requires the persistence of at least 50% of the usual number of psychological connections over the course of a day.
However, as Parfit notes, this limit is entirely arbitrary, adopted only for the sake of convenience. We could just as easily claim that 60% connectedness over a day (or a week) meant that the original person has been destroyed and replaced by a new person. The fact is that the previous person-stage has 60% survived over the specified interval. Beyond that fact it is a matter of decision and linguistic convention whether we say the same person continues, or whether the original person retains their identity. Our decision will be influenced by the personal, legal, and cultural consequences of placing the strong connectedness requirement at a particular level. This indeterminacy is worsened by the problems involved in trying to weigh the relative contributions of different components of psychological connectedness. Are memories more or less important than skills? Are dispositions more or less important than intentions? Exactly what kinds and degrees of psychological changes would add up to a 40% or 50% reduction in connectedness?
We can accurately say that someone – some person-stage – suffering significant brain damage has partly died. This is just to say that the remaining person-stage has a significantly reduced degree of psychological connectedness with the previous person-stage associated with that body. Speaking of a person's partial death is not merely a manner of speaking; it is a reflection of a real weakening in the strands that constitute the continuity of the person. The more of these strands that are removed at a time, the less of the pre-existing person who continues. We may usefully think of a person's survival or continuity this way: If my friend came out of a car accident with brain damage resulting in loss of some of his personality or with alteration in personality, I would mourn for the loss of part of the person he was. I would miss the person-stage whom I knew and would have to acquaint myself with the new person-stage. At the same time, I am sorry for what has happened to my friend, the same person who continues to exist. So long as the damage done to him does not exceed the critical threshold beyond which strong psychological connectedness is lost (wherever we set that threshold) he is the same person, not qualitatively, but in the logical sense of personal identity. The same person persists, though a new person-stage has peremptorily replaced the former stage of the person.
Though we can think of death as a matter of degree, the legal system is not good at handling spectra. The law regards persons as either dead or alive – or at least alive or dead or deanimate – just as it regards people as either above or below an age of consent. It may therefore be practically necessary to decide on an line – arbitrary within broad limits – sorting cases of damage to a person that result in their death from those that do not. If someone is too damaged then, although their body may continue to function and some behaviors or responses may persist, we can say that the person has died, changing their legal status, and setting in motion activities such as disposal of the body and distribution of the estate. (The body would not be disposed of if the person to whom it had belonged had signed a living will directing that their body be maintained in such a situation, and had provided funding for this purpose.) A plausible candidate for the standard of being too damaged is the point (or range) where someone loses the characteristics of personhood. These characteristics include the capacity for consciousness and self-awareness, rationality, responsibility, and an ability to communicate. These characteristics are, to be sure, a matter of degree, but there will be some cases where it is clear that personhood has been lost.
Let us imagine a man, Jones, who is a normal, adult human being, possessing all the familiar characteristics of personhood, displaying a robust intelligence, and enjoying diverse relationships. One unfortunate day, Jones is involved in a serious auto accident. His friends and family, gathering at the hospital, are told that Jones has suffered a devastating head injury, but that he will live. However, though he breathes unaided and shows some awareness of the environment - moving away from loud noises and towards the smell of food - he cannot recognize anyone. Despite repeated attempts no one is able to evoke any sign of recognition from him. Let us call this individual Jones-B, and the pre-accident person Jones-A. Jones-B has not only lost his memories of Jones-A's life, he displays none of Jones-A's wittiness, insightfulness, ability to form complex and satisfying relationships, nor can he even carry out a conversation. His brain has been damaged to such an extent that he has to be fed, he does not recognize friends and relatives of the past, and he cannot engage in any of the activities that Jones-A could. On any remotely plausible measure of connectedness, the remaining psychological links between Jones-B and Jones-A are too tenuous to amount to strong connectedness: Jones-A and Jones-B are not psychologically continuous. I hold that Jones-B is not the same person as Jones-A; the two together do not form one longer lived individual, though they consecutively share a single continuous body; Jones-A no longer exists; Jones-A is dead.
In cases like that of Jones, we can say that a normal person, who is suddenly damaged so badly that they lose the qualities of personhood, ceases to exist. The massive and discontinuous loss of psychological connectedness means that personal continuity has been broken, and that person's life has been terminated. If the damage is less, we will say that the same person lives on, though diminished, even if many of their memories are gone, their personality undergoes some changes, and their cognitive and communicative abilities are lessened. The point here is not that personhood is essential to a person's remaining alive; personhood is simply a plausible marker for the minimum retention of connectedness required for a person's survival. My underlying claim, based on psychological reductionism, is that death of a person necessarily results from an excessive severing of connectedness; it is not that loss of personhood necessarily results in the death of a human in all cases. We can see the distinctness of these claims by considering a different kind of case.
Smith is a healthy woman, her cognitive and communicative abilities, her capacity to form relationships, and her ability to reason and to take responsibility being comparable to those of Jones-A in my former example. Smith starts off in a condition much like that of Jones-A, and she ends up as diminished as Jones-B, but she goes from one state to the other differently. Whereas Jones-A was replaced discontinuously and immediately by Jones-B, Smith undergoes a gradual, years-long deterioration resulting from (let us posit) the growth of inoperable cancer in her brain. Comparing Smith at an advanced stage of her deterioration with her former healthy self, we see few psychological connections – as few as connected Jones-A and Jones-B. Smith differs from Jones in that at no time do we find a discontinuity remotely like that resulting from Jones' accident. Comparing her condition between any two adjacent days, or even months, we find only minor changes. Month by month her memories slowly fade, her concentration weakens and dissolves, her confusion increases, and her capacity to reason, communicate, and engage in complex behavior attenuates. Since Smith's loss of connectedness is cumulative rather than acute, psychological continuity is maintained. Even though the deterioration eventually robs her of the characteristics of personhood, the gradualness of the loss means that, unlike Jones-A, she has not perished.
As it stands, the law holds that a person, after falling below the threshold for personhood, has not died regardless of whether this occurred gradually or discontinuously. Despite her defense of the neocortical criterion, Gervais appears to agree with the current legal view. According to Gervais, a person (and not just a biological organism) continues to exist – personal identity is maintained – despite total loss of personality, so long as the capacity for consciousness remains. We can agree in Smith's case, where I also hold that Smith survives past the loss of personhood. We can also agree in the case of Jones-B; though not a person, Jones-B possesses a capacity for consciousness of a limited kind and survives so long as that continues. However, we differ in the case of Jones-A where, according to Gervais, Jones-A continues to live after the accident (being identical with Jones-B).
In criticizing the view of Green and Wikler who, like me, employ a psychological continuity conception of death, Gervais raises anencephalic infants as supposedly being a problem. Gervais thinks that the cases of anencephalic infants and cases like Jones should be treated alike: "If the anencephalic are obviously not dead, then Jones is obviously not dead either."  If Jones, lacking sufficient neocortical function to support personhood, is dead even though he has a living body, then a baby born without neocortical function must also be dead.
Having distinguished Gervais' Jones into two individuals, Jones-A and Jones-B, my response to Gervais is unproblematic: It is true that the anencephalic are obviously not dead. But neither is Jones-B dead. On the psychological continuity view, Jones-A is dead, but there now exists a distinct individual, Jones-B, who lives. Jones-A, in losing personhood, lost so much connectedness that he ceased to exist. His body lived on, forming, along with the remaining mentality, the new individual, Jones-B. Jones-B has suffered no loss of connectedness, and so is clearly alive. His case is therefore parallel with that of anencephalic infants, for they – born without the capacity for consciousness – have never lost connectedness.
In considering a possible response to her objection from Green and Wikler, Gervais complains that granting that the anencephalic baby is alive while the brain-damaged ex-person is dead will lead to two identity criteria: "In the anencephalic case, identity criteria and conditions of life and existence do not overlap. It is confusing to speak of alive bodies and dead persons, since a similar distinction could be drawn across the board, even in brain-death cases." (141) Her response is unconvincing; I see no reason not to make a similar distinction across the board, including brain-death cases. Refraining from distinguishing conditions for life and death of bodies and persons, Gervais will be forced to say that someone survives even if practically everything that made them who they were is destroyed. She will have to claim that the bare capacity for consciousness is most of what makes someone who they are. Furthermore, Gervais' charge of inconsistency against Green and Wikler can be applied to her own position because her own criterion of upper brain death makes it possible to talk of alive bodies and dead persons. In attacking the whole-brain death criterion, Gervais made it clear that a person dies along with their upper brain, even if the lower brain continues to maintain the body.
When should a declaration of deanimation be made? A person should be declared deanimate (or be said to have deanimated) when it is judged impossible or pointless to revive them given the available medical technology; that is, when conscious activity: (i) has ceased, (ii) will not restart spontaneously, and (iii) cannot immediately be restored artificially. The third condition is optional: I include it because if the capacity for consciousness can (and will) be restored immediately (e.g., by defibrillation) probably we will want to regard them as alive rather than deanimate in order to maintain all their rights as living persons. We could refer to persons who meet the first two conditions, but not the third, as dormant. Deanimation might be declared on the same grounds that death is now declared: Cardiac and respiratory arrest, or lower brainstem dysfunction or destruction, or loss of neocortical function.
When should a declaration of death be made? An individual will be theoretically dead and can be so declared if their neocortex has been destroyed. On the theoretical view, death should not be declared on cessation of cardiac or respiratory function, nor upon brain dysfunction. However, though irreversibility provides a deeper criterion of death, the permanence view may better suited to declaring death. A declaration of death is as much a legal and social act as a medical act, marking a point where our attitudes toward the person shift. We will no longer expect to see them again; we will not think of them acting and living in the future; we will not make plans for them; we will no longer take into account their interests (except residual interests in the disposition of their former property).
The patient might be declared dead when heart-beat and respiration or brain functions cease. A declaration of death would then mean that the (theoretically) deanimate person was to be allowed to continue moving toward death without intervention. This would amount to a declaration that those involved regard the person as having permanently lost consciousness even if they are not irreversibly nonconscious. The patient may have stated that they wish to be regarded as dead when they deanimate, either because they believe their reanimated quality of life will be unacceptable, or because they believe that reversal of their deanimation is not now and never will be possible.
Embodying the category of deanimate in the law would have several beneficial results. Recognizing the category of deanimate persons, and the potential for expanding that category, would spur research in neocortical preservation and repair. Common practice would gradually move from disposal and dissolution of persons when they are not theoretically dead to a situation where the possibilities for maintenance and revival were affirmed and acted upon.
Understanding the difference between deanimate and dead will help us to clarify the moral status of comatose persons. I cannot examine this issue here, except to note that the differential condition of deanimate and dead persons should be accompanied by differing rights and obligations. Deanimate persons should have rights more extensive than dead persons but weaker than those of living persons. When a person deanimates, they will lose rights which depend on awareness, such as being bound by new contracts, but will maintain rights against being harmed in various ways (including being killed), and some control over their finances (through agents chosen pre-deanimation or else appointed post-deanimation).
As matters stand now, families suffer distress and heavy costs because persons who are either truly dead or else irretrievable with current techniques must be maintained on life support equipment. On the informational criterion for death, more patients will be recognized as dead and so released from pointless support. In addition, some of those who would not be dead on the new criterion could be disconnected from life support in order to place them in biostasis instead of leaving them on a downward spiral. Setting conditions for transferring patients from mechanical functional support to stable biostasis is already needed due to the practice of cryonics and would be facilitated by a broader understanding of the deanimate/dead distinction.
Law and medical practice should allow a patient to be declared deanimate before the patient is fully brain dead (or neocortically dead), in order to allow biostatic preservation of the brain in good condition. This will give the patient more choice over what happens to them: Estimates of the possibility, probability, and desirability of eventual reanimation will vary greatly between persons. Recognizing this as rightfully the patient's choice (perhaps with consultation) would be another benefit of introducing the category of deanimate.
Brockman, John. (1988/91). Doing Science: The Reality Club 2. New York: Prentice Hall Press.
Byrne, P., O'Reilly, S., and Quay, P.M. (1979). Brain death: An opposing viewpoint. JAMA, 2, Nov, 242 (18), 1985-90.
Dawkins, Richard. (1976; 2nd ed. 1989). The Selfish Gene. Oxford: Oxford University Press.
Dennett, Daniel. (1991). Consciousness Explained. Boston: Little, Brown and Company.
Dyson, Freeman, J. (1988). Infinite in All Directions. Harper and Row: New York.
Gervais, Karen. (1986). Defining Death. New Haven: Yale University Press.
Green, Michael, and Wikler, Daniel. (1980). Brain death and personal identity. Philosophy and Public Affairs, 9 (2), Winter, 105-33.
Kelly, Kevin. (1991). Designing perpetual novelty: Selected notes from the Second Artificial Life Conference. In Brockman (1988/91): 1-44
Lamb, David. (1985). Death, Brain Death and Ethics. Albany: State University of New York Press.
Christopher G. Langton, ed. (1989). Artificial Life. Addison-Wesley, Reading, MA.
Lycan, William G. (1987). Consciousness. Cambridge, Mass: Bradford Books, MIT Press.
Merkle, Ralph. (1993). Uploading: Transferring consciousness from brain to computer. Extropy: The Journal of Transhumanist Thought 5, No. 1 (#11).
Moravec, Hans. (1988). Mind Children: The Future of Robot and Human Intelligence. Cambridge, Mass: Harvard University Press.
Prigogine, Ilya and Isabelle Stengers. (1984). Order Out of Chaos: Man's New Dialogue with Nature. New York: Bantam.
Ross, David. (1992). Persons, Programs, and Uploading Consciousness. Extropy: The Journal of Transhumanist Thought 4, No. 1 (#9).
Rucker, Rudy. (1982). Software. Ace Books.
Schwarcz, R. (1981). Effects of tissue storage and freezing on brain glutamate uptake. Life Sci 28: 1147-1154.
Tipler, Frank J. (1994). The Physics of Immortality. New York, Doubleday.
Tower, D.B., and Young, O.M. (1973a). Interspecies correlations of cerebral cortical oxygen consumption, acetylcholinesterase activity and chloride content: Studies on the brains of the fin whale (Balaenoptera Physalus) and the sperm whale (Physeter Catodon). J. Neurochem 20: 253-267.
Tower, D.B., and Young, O.M. (1973b). The activities of butyrylcholinesterase and carbonic anhydrase, the rate of anaerobic glycolysis, and the question of a constant density of glial cells in cerebral cortices of various mammalian species from mouse to whale. J. Neurochem 20: 269-278.
Tower, D.B., Goldman, S.S., and Young, O.M. (1976). Oxygen consumption by frozen and thawed cerebrocortical slices from warm-adapted or hibernating hamsters: The protective effects of hibernation. J. Neurochem 27: 285-287.
Veatch, Robert M. (1975). The whole-brain-oriented concept of death: An outmoded philosophical foundation. Journal of Thanatology 3:13-30.