4The Enigma of DeathThe Gift of LifeThe goal of this chapter is to consider whether it is possible to formu-late a satisfactory philosophical analysis of the concept of death. Ifthere is such an analysis, it seems likely that the concept of life playsa central role in it. Since any obscurity in the concept of life wouldapparently carry over into the analysis of the concept of death, Ihave devoted the previous two chapters to a discussion of the analy-sis of the concept of life. Unfortunately, it appears that life is enig-matic. I have been unable to say precisely what 'x is alive at t' means.Insofar as the concept of life is obscure, it will import obscurity intoany analysis of death in which it appears.Since I have not been able to analyze the concept of life, Ipropose to make use of the concept of life as an unanalyzed primi-tive. Thus, for present purposes, I simply give myself the gift oflife. That is, I assume that our grasp of the concept of life issufficiently firm to permit it to be used in the analysis of the con-cept of death. I recognize that any obscurity in the concept of lifemay make what follows somewhat less than crystal clear. So be it.The Biological Concept of DeathBefore turning to a consideration of some proposed analyses of theconcept of death, it may be useful to try to be a bit more specificabout the target concept. That is, I should try somehow to identifythe concept whose analysis is here in question. As I remarked in56The Enigma of Death 57Chapter 1, some of the literature on death seems to be based onthe assumption that the most interesting concept of death is oneapplicable only to people. Such analyses mention brain function,consciousness, or some other concept that has no application tolowly organisms. Thus, some writers seem to be trying to analyze aconcept of death that could not possibly apply to any lower animalor plant.1I also mentioned in Chapter 1 that this view seems quite odd tome. I see no reason to suppose that the word 'died' has differentmeanings in these sentences:1. JFK died in November of 1963.2. The last dodo died in April of 1681.3. My Baldwin apple tree died during January of 1986.I acknowledge that I cannot prove that 'died' is univocal in thesesentences. Nor can I prove that there is no essentially \"personal\"sense of 'died'. However, I shall proceed here on the assumptionthat there is a concept of death that has application throughout thebiological realm. As I see it, just as there is a single concept of lifethat applies to every living plant and animal, so there is a singleconcept of death that potentially applies to every organism fromthe lowliest plant all the way up to the most complex mammal. Icall this \"the biological concept of death.\" Our topic here is theanalysis of this concept.It may also be important to recall another point I made in Chap-ter 1. This concerns the distinction between two fundamentallydifferent projects. One project is the attempt to discover a correctanalysis of the concept of death. The other project is the attempt toformulate a useful criterion of human death.As a sample criterion of death, we might consider:4. A person, S, dies at a time, t, if and only if S's brain ceases toemit z-waves at t.As a sample analysis of death, we have this:Dl: x dies at t =df. x ceases to be alive at t.58 THE NATURE OF DEATHA criterion is something that is proposed for acceptance; thatmight be adopted if enough people think it would be useful; thatmight later be rejected if it proves inadequate or becomes obsoleteas a result of technological advances. A criterion is better if it ismore useful—easier to apply; more practical; more decisive. Itwould apply only to human beings. Thus, if enough morticians,prosecutors, judges, transplant teams, etc., agreed at a conventionto adopt (4), it would be the criterion of death in human beings(until replaced).A philosophical analysis, on the other hand, is supposed to re-port a necessary truth about the construction of a concept. If Dl istrue, then the concept of death is constructed out of the conceptsof life and cessation in the indicated way. No amount of voting oradopting could make Dl true if in fact it is false; nor could anyhuman activity make it false if in fact it is true. (Of course, wecould conventionally adopt some other meaning for the word'dies'. But this would have no bearing on the analysis of the con-cept that word formerly expressed.)Furthermore, the existence of possible counterexamples showsthat a proposed analysis is simply false. It shows that the conceptexpressed by the term to the left of the ' = df.' sign is not the sameas the concept expressed by the term to the right. When this hap-pens, the analysis fails. Counterexamples are not decisive againstproposed criteria in this way. The excellence of a criterion is useful-ness, not truth. It's not entirely clear what a counterexample wouldbe. But the mere possibility of a few counterexamples to (4) has nobearing whatsoever on its suitability as a criterion of death forhuman beings. If actual counterexamples are sufficiently rare, (4)might remain perfectly acceptable as a criterion of death in humanbeings.The topic under consideration here is the analysis of the biologi-cal concept of death. Let us now turn to an analysis that hasrecently been proposed.Perrett's AnalysisIn his book Death and Immortality, Roy Perrett seems to be discuss-ing the nature of death itself. He distinguishes between personsThe Enigma of Death 59and biological organisms and says that he is focusing on the ques-tion about what we mean when we speak of the death of a biologi-cal organism.2 Thus, it appears that Perrett's target in his book isidentical to my target in this chapter. He calls it \"the concept ofdeath that is neutral to all deaths.\"3 He goes on to say: \"My pro-posal is that death be identified with [the destruction of a function-ing biological organism].\"4If Perrett said no more on this topic, the reader would surelyassume that he meant to defend this analysis of the concept ofdeath:D2: x dies at t =df. x is a living biological organism up to t, but at t,x is destroyed.However, Perrett has more to say.In a passage just a sentence later than one already cited, Perrettsays that \"death is the annihilation of a functioning biological organ-ism.\"5 And only a page after that, he says that death is 'the disinte-gration of a living organism . . .'6 These remarks seem to confusematters, since it now appears that Perrett has committed himself tothree different analyses of the biological concept of death. Accord-ing to the first, death is the destruction of a living biological organ-ism; according to the second, it is the annihilation of such an organ-ism; according to the third, it is the disintegration of the organism.The three proposals are distinct. We could investigate them inde-pendently. However, it seems to me that none is correct. Oneexample suffices to show that each of Perrett's proposals is false.Suppose a butterfly collector captures a rare specimen. Supposeshe carefully places it in the killing jar. Surely it is possible that shemight kill it without breaking off any legs and without dislodgingeven so much as a single scale from the wings. The specimen mightbe \"perfect.\" In such a case, I think, even though the butterfly haddied it would be wrong to say that it has been \"destroyed.\" It iseven more obvious that it would be wrong to say that it has disinte-grated or that it has been annihilated. Thus, Perrett's proposal, nomatter how interpreted, is false.In another passage, Perrett notes that death \"marks the transi-tion from being alive to being dead.\"7 This hints at a much morepopular conception of death—the conception according to which a60 THE NATURE OF DEATHthing dies at a time if and only if it then ceases to be alive. Sincethis conception is so popular, I refer to it as \"the standard analy-sis.\" Let us turn to it.The Standard AnalysisI think it is fair to say that something like the analysis formulatedabove in Dl is almost universally assumed to be correct. Somewould accept Dl as it stands. Thus, in his recent book ThinkingClearly about Death, Jay Rosenberg says, \"to die is to cease to live,to cease to be in the condition of life.\"8 Others would modify thisby adding that death occurs only if life permanently ceases. Rosen-berg cites a dictionary definition that suggests this idea. Accordingto this dictionary, death is \"the total and permanent cessation of allthe vital functions of an animal or plant.\"9 Still others would mod-ify it in a slightly different way by adding that death occurs whenlife irreversibly ceases. Rosenberg himself suggests this view whenhe says elsewhere in his book that \". . .an organism dies when itloses its power to preserve and sustain its self-organizing organiza-tion permanently and irreversibly.\"10 Since Rosenberg takes life tobe the cited power, this is tantamount to saying that an organismdies when it irreversibly ceases to live. The variations may seemtrivial. Let us say that each of the proposed analyses, and anyothers relevantly like them, are instances of \"the standard analy-sis.\" According to this view, death is the (perhaps permanent,perhaps irreversible) cessation of life.Puzzles About Suspended AnimationI think the three above-mentioned versions of the standard analy-sis are genuinely distinct and mutually incompatible. Furthermore,I think none of them is true. Although each of them is open toseveral sorts of objection, I want to discuss two main sorts ofdifficulty. The first difficulty is that the standard analysis is incom-patible with some facts concerning suspended animation. Let usthen consider this phenomenon.11As a rough first approximation, we may say that an organismThe Enigma of Death 61undergoes suspended animation when it temporarily ceases to bealive. The most familiar type of suspended animation involvesfreezing. It takes place every day in biology laboratories all overthe world. In a typical case, some sort of microorganism has beengrown in a culture. The culture is then flooded with glycerol orsome other suitable cryoprotectant, and the whole thing is gradu-ally cooled until frozen solid. Subsequently, the frozen culture isplaced for storage in liquid nitrogen at a temperature of -196°C.The glycerol prevents crystalization within the cells, which other-wise would rupture.Later on, when there is need for the microorganisms, a labtechnician can remove the culture from the freezer and allow itgradually to warm up. If the culture has been properly handled,the microorganisms will return to life merely as a result of beingreturned to room temperature.This sort of procedure can be applied to all sorts of microorgan-isms, as well as to isolated cell cultures. It is also an important stepin certain reproductive techniques. For example, consider in vitrofertilization as applied to cows. Sperm and eggs can be removedfrom adult animals and then mixed in a dish. Fertilized eggs canthen be removed and allowed to undergo a relatively small numberof cell divisions. The blastulas are then soaked in glycerol andfrozen in liquid nitrogen. They cease to be alive.12 Later, the fro-zen blastulas may be thawed and implanted in the reproductivetracts of suitable cows. They resume growth and eventually areborn just like old-fashioned calves.In vitro fertilization followed by fetal implantation is very com-mon in cows and horses. It is much less common in human beings,but it has been used on hundreds of occasions during the pasttwenty-five years or so. All these cases illustrate suspended anima-tion, since in every case a living organism (or cell culture, orblastula) temporarily ceases to be alive and then lives again.I realize that it is currently impossible to freeze adult humanbeings (or any other large mammals) and subsequently revivethem. Freezing destroys too many cells. However, it is reasonableto believe that the problems are all merely technical. Just as we cannow freeze and later reanimate a day-old human blastula, so some-day we will be able to freeze and then later reanimate an adulthuman being. Let us imagine that the technology has in fact been62 THE NATURE OF DEATHdeveloped, and that an adult human being can be frozen and laterreanimated. I prefer to proceed in this way primarily for dramaticeffect, even though my argument could just as well be formulatedby appeal to an example involving organisms that currently can befrozen and revived.13To see how facts about suspended animation bear on the stan-dard analysis of death, let us consider a case.Case One. A man has a bad disease. There is currently nocure. Unless some way can be found to stop the disease, hewill die in a few days. There is good reason to believe that acure will be found in a dozen years or so. Cryogenics, Inc.,offers to inject some specially formulated glycerol and tofreeze the man solid. Then, when the cure has been per-fected, they will thaw him out, reanimate him, and see to itthat he is cured of the disease. The man accepts the offer andis injected and frozen. Ten years later, a cure for the diseaseis found. The body is thawed, reanimated, and subjected tothe cure. The man goes on with his life.In one of the passages cited above, Rosenberg said that to die isto cease to live.14 This surely suggests the following version of thestandard analysis of the concept of death:Dl: x dies at t =df. x ceases to be alive at t.Case One refutes Dl. For in Case One the man ceased to bealive when he was frozen. Without an accepted analysis of theconcept of life, this point is hard to prove. But it seems reasonableto say that the man ceased to live when he was frozen. After all, hethen ceased to engage in metabolism, growth, motion, and theother life functions. Dl therefore implies that the man died whenhe was frozen. But the implication is false—the man did not diewhen frozen. He went into suspended animation. Unless some-thing went wrong with the procedures, and it became impossible toreanimate him, no one would want to say that Cryogenics, Inc.,killed its client.The mere possibility of suspended animation shows that deathcannot be denned as the cessation of life. When an organism entersThe Enigma of Death 63suspended animation it ceases to live, but it does not then die. Wemust alter the standard analysis to accommodate this fact. Theanalysis of death must be consistent with the fact that not all organ-isms that enter suspended animation die.Perhaps we can think of suspended animation as the temporarycessation of life. Then perhaps we will want to say that death is thepermanent cessation of life. This suggests another version of thestandard analysis of death:D2: x dies at t =df. x ceases permanently to be alive at t.Notice first that D2 yields a different result in Case One. Sincethe frozen man did not cease permanently to live when he wasfrozen, D2 (unlike Dl) entails that he did not then die. This mayseem to be an improvement, since it seems to be consistent withthe facts about suspended animation. But another example showsthat there is something implausible about D2.Case Two. Each of two identical twins has the same currentlyincurable disease. Both are frozen and go into suspendedanimation. Unfortunately, about one year later, one frozenbody is damaged. The damage is so severe that it would beimpossible ever to reanimate the body. It is then thawed outand buried. That twin never lives again. The other twin re-mains frozen until a cure is found. He is then thawed, reani-mated, and cured. The second twin goes on with his life.D2 yields strange results in Case Two. Since the first twin in factnever lives again after being frozen, D2 entails that he died whenhe was frozen. His loss of life was permanent. On the other hand,even though there is no discernible difference between the twinsduring their first year on ice, D2 entails that the second twin didnot die when he was frozen. This follows from the stipulated factthat the second twin comes back to life later. His loss of life wasonly temporary. But it seems to me that until the accident occurs,the twins are in relevantly similar conditions. We can imagine that,cell-for-cell, they are indiscernible. So either they are both dead,or they are both alive, or they are both neither dead nor alive. (Myown view is that they are both neither dead nor alive. As I see it,64 THE NATURE OF DEATHsuspended animation is a state that excludes both life and death.But the point of the example does not depend on my intuition.15)D2 entails that the twins are in different \"vital states\" during thefirst year on ice—one is dead, the other not. Since the twins are infact not in different vital states during that period of time, D2 iswrong.Consider yet a third analysis of death suggested by Rosenberg'sremark16 about permanence and irreversibility:D3: x dies at t =df. x ceases permanently and irreversibly to bealive at t.D3 has truly bizarre implications in Case Two. Consider the twinwhose body is damaged. According to D3, this twin never dies. Hedoes not die when frozen, because at the time of freezing laterreanimation is still possible. Though he then ceases permanently tobe alive, he does not cease irreversibly to be alive. If he had beenhandled properly, his condition would have been reversed. Hewould have come back to life. Nor according to D3 does he diewhen the body is damaged in handling, for he does not cease in anyway to be alive at that time. That is a time at which he is not aliveto start with. So there is no time at which the twin \"ceases perma-nently and irreversibly to be alive.\" Surely this is wrong; surelythere is some time at which that twin dies.Let us consider a variant of D3:D4: x dies at t =df. (i) x ceases permanently to be alive at or beforet, and (ii) at t, it becomes physically impossible for x ever tolive again.The idea behind D4 is that the time of death is the time at whichthe loss of life becomes irreversible. The loss of life may haveoccurred years before. D4 implies that there is a time of death forthe damaged twin. The time of his death, according to D4, is notthe time when he was frozen but the time when the body is dam-aged beyond repair, for this is the time at which the loss of lifebecomes irreversible.The Enigma of Death 65In more mundane cases, D4 implies that death occurs approxi-mately when life ceases. For in more mundane cases, when lifeceases, it is almost immediately impossible for it to return.Some would reject D4 because it implies that it is impossible foran organism to live again after it dies. The comedian Jerry Lewisclaims that he died several times while undergoing open-heart sur-gery. If D4 is correct, Lewis must be wrong. We can offer a some-what less striking claim for Lewis. Perhaps he would be satisfied tosay instead that he ceased to live several times while undergoingopen-heart surgery. D4 permits that. Maybe that's all Lewismeans.A more serious problem with D4 can be brought about by consid-eration of a distinction. In some cases, later reanimation becomesimpossible because of changes that take place within the body.Thus, for example, if the body is damaged beyond repair, internalchanges make later reanimation impossible. In other cases, how-ever, changes that take place outside the body may make laterreanimation impossible. Perhaps the body is moved to a placewhere it cannot be reached; perhaps a crucial reanimation chemi-cal is irretrievably lost; perhaps the atmosphere of the earth be-comes so choked with pollution or radioactivity that it would beimpossible to reanimate the frozen corpse (even if there were sometechnicans to try!).Suppose a body is in suspended animation, and some such exter-nal change takes place, thereby making later reanimation impossi-ble. Provided that the body is internally unchanged, I would beuneasy about saying that it had just died. I would rather say thatthe body remains undead until internal changes occur that wouldindependently make subsequent reanimation impossible. Thus Ipropose:D5: x dies at t =df. (i) x ceases to be alive at or before t, and (ii) att, internal changes occur in x that make it physically impossiblefor x ever to live again.While I think that D5 comes pretty close to solving the problemof suspended animation, I still have my doubts. I am troubled bythe obscurity of the concepts of internality, physical impossibility,66 THE NATURE OF DEATHand life. But let us assume that we have come close enough. I wantto turn to another problem for the standard analysis.Problems Concerning Fission and FusionIn spite of its plausibility and in spite of the fact that Rosenbergseems to endorse something quite like it, D5 is inconsistent withcertain other plausible views Rosenberg maintains. In an interest-ing passage, Rosenberg asserts that death is not the only route outof life. To illustrate his point, he describes the case of an amoeba,Alvin.17 He tells us that Alvin was a fat and healthy amoeba.According to the story, Alvin was so fat and healthy that at pre-cisely midnight on Tuesday night/Wednesday morning, Alvin un-derwent fission and became two amoebas. According to Rosen-berg, Alvin no longer existed on Wednesday. Apparently, Alvinwas \"replaced\" by his two descendants, Amos and Ambrose. Ro-senberg claims that Alvin's example shows that \"there are otherways for a life to come to an end besides death.\"18 So while Alvin isno longer among the living on Wednesday, it is \". . . clear that hedid not die.\"19 My own intuitive sense of the situation is identical toRosenberg's. I would not say that Alvin died.Fission is not the only biological process that may seem to pro-vide a deathless exit from life. Rosenberg apparently thinks thatmetamorphosis does the same thing. As he sees it, when a caterpil-lar turns into a butterfly, the caterpillar ceases to exist but does notdie.20The point that Rosenberg seems to have missed is this: if Alvinceased to be alive at midnight, but did not die at midnight, thendeath cannot be the cessation of life. When we say that a thingdied, we cannot mean just that it ceased to live. For Alvin ceasedto live without dying. If we think that the caterpillar gets out of lifewithout dying, we will have to say that its case also refutes the ideathat death is the cessation of life.A natural \"fix\" would be based on a crucial feature that is com-mon to division and metamorphosis. In each case, an organismseems to go out of existence, but the stuff of which it is madecontinues to exist—and this stuff continues to support life. We canThe Enigma of Death 67make use of this common feature in a relatively economical newanalysis of death:D6: x dies at t =df. (i) x ceases to be alive at or before t, and (ii) att, x undergoes internal changes that make it physically impossi-ble for x ever to live again, and (iii) it's not the case that x turnsinto another living thing or a bunch of other living things at t.It may be useful to say a few words about a phrase—'turnsinto'—that appears in D6. This phrase is intended to express whatis traditionally called 'substantial change'. Some would say that thecaterpillar undergoes substantial change when it turns into a butter-fly. As I understand it, the crucial elements in a pure example ofsuch a change are these: the first entity (the caterpillar) is a con-crete individual substance—a \"thing.\" It is made of some \"stuff\"—a certain parcel of protoplasm, perhaps. During the substantialchange, the first entity goes out of existence, and a new concreteindividual substance (in this case, the butterfly) comes into exis-tence. The new entity is diverse from the old entity, but they aremade of the same parcel of stuff (or \"matter\"). In such a case, wecan say that the first entity \"turned into\" the second.21D6 gets the fission example right. At the moment of division,Alvin turns into Amos and Ambrose. Each of these is a livingthing. So, according to D6, Alvin does not die. Furthermore, in anordinary case, in which some organism ceases to live and simplyrots, D6 still yields the correct result. Since, in such cases, theorganism does not turn into living things, D6 entails that it dies. D6also preserves Rosenberg's intuitions concerning the caterpillarexample.22 At the moment of metamorphosis, the caterpillar alleg-edly turns into the butterfly, which is a living thing. So, although itceases to be alive, it doesn't die.Reflections such as these on cases of fission invite reflections oncorresponding cases of fusion. Are there examples in which organ-isms go out of existence by fusing with others? Would we want tosay that such organisms die when they fuse? Let us look into this.Under certain environmental conditions, certain types of single-celled green algae engage in a sort of fusion. These creatures,called chlamydomonas, are flagellated, chlorophyll-bearing plants.In their normal state, each individual is haploid. Although they areDO THE NATURE OF DEATHall of approximately the same size, they come in two differentmating types. When conditions are favorable, large clusters ofindividuals form.Eventually the clustered cells move apart in pairs. The members of apair are positioned end to end, with their flagella, which bearspecies-specific and mating-type-specific attractant sites at theirtips, in close contact. The cells then shed their walls, and theircytoplasms slowly fuse. Finally, their nuclei unite in the process offertilization, which produces a single diploid cell, the zygote.23Suppose two chlamydomonas, cl and c2, fuse to form a newzygote, c3. It seems reasonably clear that, in this process, cl and c2go out of their existence. Furthermore, it seems reasonably clearthat neither one of them turns into any new living individual. Noliving part of the resultant individual, c3, can be identified as thepart such that cl turned into it. The stuff from which cl and c2were made is thoroughly blended in c3.In this case, we must say that cl ceases irreversibly to live anddoes not turn into another living thing or even into a bunch ofliving things. D6 then legislates that cl dies at the moment offusion. Yet I would hesitate to say that cl dies at the moment offusion. I would say that the example of the chlamydomonas showsthat there are still more ways of getting out of life. In addition todeath, suspended animation, and deathless fission, there is also acertain sort of deathless fusion that sometimes does the trick. SoD6 is wrong.24We can revise D6 in such a way as to take account of fusion, too.We merely add a clause specifying that if an organism engages indeathless fusion, then it does not die. In other words, if it is amember of a set of living things that fuses into a new living thing,then it does not die:D7: x dies at t =df. (i) x ceases to be alive at or before t, and (ii) att, x undergoes internal changes that make it physically impossi-ble for x ever to live again, and (iii) it is not the case that xturns into a living thing, or a bunch of living things, at t, and(iv) it is not the case that x is a member of a set of living thingswhose members fuse and turn into a living thing at t.The Enigma of Death 69The fundamental idea behind D7 is reasonably simple: a thingdies if and only if it ceases irreversibly to live without making useof one of the deathless exits; the deathless exits are metamorphosis(turning into another living thing); a certain sort of fission (turninginto a bunch of living things) and a certain sort of fusion (being amember of a set of living things that fuse into a living thing). Ithink D7 gets a wide variety of cases right. In simple cases, inwhich an organism ceases to live and simply rots, D7 says that theorganism dies. In cases of fission like the one illustrated by Alvin,D7 says that the organism does not die, even though it ceases tolive. Similarly for the chlamydomonas—they cease to live withoutdying because they make use of one of the deathless exits.Nevertheless, it seems to me that D7 still fails. One problem isthat there are forms of division that mimic deathless fission butthat seem to involve the death of the divided organism. Consideran example. Imagine a device for use in biology laboratories—a\"cell separator.\" This is a machine that grinds up mice and thenemits a puree of mouse cells. The machine is constructed in such away that all the mouse cells come out alive. Each cell can be placedin a suitable medium and kept alive indefinitely.Suppose some mouse is placed in the cell separator and isground up into a puree of living mouse cells. In this case, themouse goes out of existence, and hence ceases to be alive. How-ever, it turns into a bunch of living things. As a result, the mousefails to satisfy the right-hand side of D7. D7 then legislates that themouse does not die. It seems to me, however, that the cell separa-tor kills the mouse.Another example involves not cells, but bodily organs. Reason-ably sane medical personnel sometimes want to harvest living or-gans from dying patients. Suppose a mad scientist wants to harvestall the organs from some perfectly healthy victim. Suppose hecaptures his victim, knocks him out, and then carefully dissects thevictim's body in such a way as to waste nothing. Every organ ispreserved alive. (If need be, we can imagine that each organ istransplanted into some needy body, where it remains alive foryears to come.) In this case, it would appear that the poor victimgoes out of existence and is replaced by a complete set of livingbodily organs. If D7 were correct, we would have to say that thevictim did not die. This seems wrong.2570 THE NATURE OF DEATHIf we allow ourselves to make use of another rather obscureconcept, we may be able to revise D7 in such a way as to accommo-date these examples. Let us assume that we have sufficient under-standing of what we mean when we say that something is an organ-ism. Now notice that when an amoeba deathlessly divides, it turnsinto living organisms, but that when a mouse is killed in the cellseparator, it does not turn into living organisms. It turns into livingcells. Similarly, in the case of the Mad Organ Harvester, the victimdoes not turn into living organisms. He turns into the members of aset of living organs. Perhaps this marks the distinction betweendeathless and deadly division.We can revise D7 as follows:D8: x dies at t =df. (i) x ceases to be alive at or before t, and (ii) att, x undergoes internal changes that make it physically impossi-ble for x ever to live again, and (iii) it is not the case that xturns into a living organism or a bunch of living organisms at t,and (iv) it is not the case that x is a member of a set of livingorganisms whose members fuse and turn into a living organismat t.It seems to me that the introduction of talk about organisms inD8 is a fundamental mistake. It is a mistake, as I see it, because theconcepts of life and death apply univocally to biological entities,whether organisms or not. The difficulty can be brought out byreflection on a variant of the example concerning Alvin theamoeba.Suppose a researcher has removed a single cell from a frog and iskeeping it alive in a suitable medium. Suppose the researcher isinterested in cell division. She treats the cell in a special way.Subsequently, the cell divides, giving rise to two \"daughter cells.\"Since neither daughter cell is an organism, the original frog celldoes not turn into a bunch of living organisms. Thus, D8 entailsthat the frog cell dies at the moment of division. But it seems to methat the frog cell is relevantly like Alvin the amoeba. Since wedon't want to say that Alvin dies when he divides, we should notsay that the frog cell dies when it divides. Each of them gets out oflife deathlessly. So D8 is wrong.The Enigma of Death 71Fission and fusion are puzzling. I find that I cannot explain thedifference between their deathless forms and their deadly forms.The Mystery of DeathI think there is a single concept of death that applies across thebiological board. When we say of some plant or animal, or of somecell or tissue, or of some organ, that it has died, we may be express-ing this concept. I call this the biological concept of death.Roughly, what we seem to mean in such cases is that the biologicalentity has ceased to live but has not entered suspended animationand has not engaged in one of the deathless forms of fission orfusion. Explaining death in such rough (and circular) terms is nottoo difficult. The difficulty arises when we try to clarify the con-cepts of suspended animation and deathless fission and fusion. It isthen that the enigma of death begins to reveal itself.My main point is that when we say that some biological entityhas died, we do not invariably mean that it has ceased to live. I aminclined to suspect that we never mean just this. If there is somesingle thing that we do mean, then it is hard to say precisely what itis. So, though death looms large in our emotional lives, though wehate it, and fear it, and are dismayed by the thought that it willsomeday overtake us and those we love, we really don't knowprecisely what death is. The Reaper remains mysterious.