This is the second essay presented. It was a joint effort between myself and John Martin Fischer, Professor of Philosophy at the University of California, Riverside. The essay was presented at the Fourteenth Annual J. Lloyd Eaton Conference on Science Fiction and Fantasy, April 1992. It is published in Immortal Engines, University of Georgia Press, 1996.

Philosophical Models of Immortality in Science Fiction
by John Martin Fischer and Ruth Curl

Science fiction is often described as a literary genre well suited to philosophical speculation. SF and philosophy share a common interest in the question of immortality, and comparisons and contrasts can be made regarding their respective treatments of the theme. We propose here a sketchy taxonomy of different models or pictures of immortality offered by philosophers and SF writers. After noting important difference in these models, we shall suggest that some problems and concerns expressed by philosophers and SF writers alike are the result of conflating different models. It is our hope that these comparisons will provide a preliminary sense of the way SF can be said to function as philosophical discourse.

Our discussion will use as its base the analytical framework presented in Bernard Williams's influential discussion of immortality, The Makropulos Case: Reflections on the Tedium of Immortality.⁹   This simple and natural framework involves two criteria to make immortality truly appealing: first, there must be a future in which an individual can recognize himself or herself--someone genuinely identical to the individual, not just qualitatively similar or with several identical properties. Second, the future life of the individual must be appealing [in some way] to that individual; it cannot involve constant torture, hard labor, tedium, or the like. These conditions can be dubbed the identity condition and the attractiveness condition. With these, we can construct a taxonomy of different models of immortality [see table I].

Table I    A Taxonomy of Immortality

Although our focus will be the immortality of sentient creatures or constructs, another treatment of immortality in science fiction is also possible: universe immortality, in which there is an attempt to overcome laws of entropy to create an immortal world, forever self-perpetuating. The center of attention here is not the immortality of sentient creatures but rather the immortality of the physical universe.²   Only science fiction seems to deal with universe immortality; and while it is not our focus, this vision of immortality merits attention because it has no corollary in other fields of literature or philosophy.

We turn now to depictions of immortality pertaining to sentient entities, beginning with a distinction between nonatomistic and atomistic concepts of immortality. The former involves a kind of fusion of different individuals into a type of immortal entity; the later involves the immortality of individuals. The nonatomistic model usually involves the merging of various individuals into some sort of superorganism. The individual's stream of consciousness may either be retained, as in Greg Bear's Blood Music, or lost, as in Arthur C. Clarke's Childhood's End and one episode of Robert A. Heinlein's Methuselah's Children.³

In Blood Music, a brilliant researcher, after losing his job, injects himself with lymphocytes he has genetically manipulated so that he can smuggle them out of the lab and continue his research. The altered lymphocytes then invade the biosphere and trigger the mutation of humanity into a new organism composed of individually intelligent cells. Eventually the cells unite to form a superintelligent being. Each cell can either function separately or compartmentalize with other cells, which can then isolate themselves to work on various problems. Bear's vision of the mutation and transformation of humanity is best expressed in the novel's last lines: "Nothing is lost. Nothing is forgotten. It was in the blood, the flesh. And now it is forever" [M 247].

In Childhood's End, children and adolescents transform beyond the comprehension of the rest of humanity. Clarke's vision clearly shows a complete but unintentional and uncontrollable break with human characteristics, memories, and emotions. In Methuselah's Children, members of the Howard Families [immortals in a mortal world] fleeing persecution on Earth and searching for a hospitable planet encounter the Little People, who "in an utterly basic sense . . . differed from humans in kind. They were not individuals. No single body of a native housed a discrete individual. There individuals were multi-bodied, they had group "souls." The basic unit of their society was a telepathic rapport group of many parts. The number of bodies and brains housing one individual ran as high as ninety or more and was never less than thirty-odd" [MC 134-35].

Clearly, there can be different versions of both nonatomistic concepts, including differences in the nature of the transition from individuals to composites, which can be a genetic mutation [as in Blood Music, ignoring for the moment the manipulation by Vergil] or a nonmutational evolutionary transformation [as in Childhood's End]. And there can be differences in the nature of the composites: for example, there may be one or many composites, and the existence of the composites might be relatively desirable or undesirable.

But any sort of nonatomistic immortality--even one in which the nature of the composite's existence is relatively attractive--appears to run afoul of Williams's first criterion: the identity criterion. Arguably, the types of fusion envisaged in nonatomistic models [even ones that somehow preserve individual streams of consciousness] do not allow individuals to look forward to their own future existence. As such, these nonatomistic models are not very appealing models of immortality.

Is it then appropriate to eliminate nonatomistic models? When we look to the future we seem to care about the welfare of our communities and friends, the planet and it nations. We might care about the continued development of the arts, the preservation of natural beauty, and the attainment of human rights and distributive justice; but we care especially about how we ourselves will fare--we especially look forward to future pleasurable states of ourselves and particularly reject prospective unpleasant future states of ourselves. So, for example, if we are told that some future individuals will be tortured horrible for days, we can genuinely regret this; however, if we are then told that those people will be ourselves, we are horrified--we especially regret this. Thus, individuals might care to some extent about a future in which individuals have become group entities of certain sorts; indeed, it might even be desirable in some sense. But we do not and cannot look at such a prospect in the special and especially vivid way we look at future scenarios in which we exist as individuals. This special sense in which we care particularly about what happens to us is not engaged by nonatomistic models of immortality.

Since nonatomistic models seem to run afoul of the identity condition, let us instead turn to atomistic models of immortality. In this class there are serial and nonserial models. In serial models of immortality, the individual in question in some ways lives a series of lives; in nonserial models, the individual simply leads an indefinitely long single life.

The atomistic serial model of immortality comes in at least two versions: the disjoint-lives serial model and the connected-lives serial model. In the disjoint-lives model, one individual lives an indefinitely long series of lives without internal psychological connections: there are no significant continuities or connections of memory or other psychological states, such as values, beliefs, desires, and intentions, from one life to the next. In this view, the self is some sort of soul or bare particular without any essential mental contents. When the soul enters a new body, the person itself persists, even if there are no remaining memories, beliefs, preferences, values, or intentions. This model recalls the Hindu model of reincarnation. A possible metaphor is the tulip bulb--the different lives correspond to the different plants and flowers that spring from the bulb from one year to the next, whereas the persisting self corresponds to the essential bulb.

But, like the nonatomistic model, the disjoint-lives serial model runs afoul of the identity condition. It is unclear how an individual could recognize a future individual as genuinely identical to himself or herself if there is no psychological connection between the two [including connections of memory]. We do not know if it is metaphysically coherent to suppose that persistence of personal identity means the persistence of a bare, psychologically empty soul; there are deep perplexities here into which we cannot go. Even if the model is metaphysically coherent, the identity condition does not seem to be satisfied in the relevant way, a way that makes it possible for us to recognize ourselves in the future scenario. That is, even if there is no insuperable ontological problem with the disjoint-lives serial picture, there is an epistemic problem: presented with a description of a future scenario, there is no way individuals can recognize or identify themselves. And, if the relevant future person has no psychological connection to the current individual, why should the individual care especially [in the way one cares especially about oneself] about this future person? Given this problem, the disjoint-lives model is unappealing--it cannot capture the sense in which we might value especially our own immortality.

Unfortunately, the connected-lives serial model fares no better. Imagine, if you can, what it would be like to lead one life--to go through childhood, adolescence, and all the stages of life--accumulate memories and associated values, and then begin again: go through a second childhood [but with memories of the previous life], a second adolescence [but with memories of the previous life plus the new childhood], and so forth. What would it be like to be a small child carrying memories of adolescence, marriage, raising a family, seeing one's children grow up, and so forth? The model of full or robust psychological connections within serial lives seems either entirely incoherent or entirely unattractive; in any case, it surely does not meet the two criteria. Nor does it seem possible to weaken the psychological connections in any natural or appealing way; in particular, it does not seem plausible that one could have only certain memories [just enough to be able to recognize oneself as a persisting entity] at certain stages of life. This would involve "blackouts" of parts of memory at some stages and not others--an almost stroboscopic and bizarre picture of memory. SF often puts a skeptical valance on this sort of connected-lives serial model. Consider Clarke's The City and The Stars [1956]. In this novel people are reborn into new bodies without memories of previous lives; then, as they near adulthood, they gradually remember their previous lives. From the perspective o0f the "new" individual, it would surely be disconcerting to be suddenly flooded by a vast set of old memories of earlier lives; and--more relevant to the issue of immortality--from the perspective of the "old" individual, it would not be pleasant to stop being conscious one day; then reawaken with a new set of memories involving a new childhood and adolescence.

Thus, even though SF novels may claim that a character leads many lives, our ruminations above lead us to call this possibility into question. On closer scrutiny, these novels do not depict characters who themselves lead different lives in the senses required by the serial model. A particular character [say, Lazarus Long] does not lead many lives; rather, he leads one extended life in which many other people play roles. Many lives become part of his life when they intersect it--but Long himself does not genuinely lead a series of lives. [A life with a series of people need not be a series of lives.]

Our taxonomic trail leads finally to atomistic nonserial conceptions of immortality, of which there are various versions. Primarily, these involve different ways of generating or maintaining the indefinitely long life, and different ways of viewing the nature of the life [pattern and distribution of experiences, relationships to other lives, and so forth].

Let us first consider the different ways of generating or maintaining a nonserial atomistic form of immortality. There are a number of horror stories in which a vampire draws on the life force of another to continue existence. Literature and films have produced many variations on this theme; the vampire does not always follow the count Dracula formula. Numerous films, for example, depict beautiful young women who seduce young men to feed off their energy; and Oscar Wilde's The Picture of Dorian Gray [1897] features a protagonist who remains young while his portrait ages. In E. E. "Doc" Smith's Lensmen series the Overlords live off the life force of the Velantians. In one Balzac story, an old man lives off young girls. A recent Stephen King film, Sleepwalker [1992], depicts a young man who gains vitality from the innocence or purity of young virgins, whom he kills in order to devour their life force, or souls. Obviously, the topic continues to generate discussion.

In Anne McCaffrey's series comprising Crystal Singer [1982], Killashandra [1985], and Crystal Line [1992], humans have developed a symbiosis with a spore, which makes them extremely long-lived. Unfortunately, they must periodically return to the planet of the spores to avoid a terrible death. [This somehow resembles the need to visit one's parents regularly--at least in our families!] In McCaffery and Jody Lynn Nye's The Death of Sleep [1990], the protagonist, who engages in cryogenic sleep, ages only four or five years in seventy-two. Another type of life prolongation is envisioned in Hugo Gernsback's Ralph 124C41+: A Romance of the Year 2660 [1925], in which a man reacts to a scientist's revival of a dead dog by exclaiming, "I only regret for myself that you had not lived and conducted this experiment when I was a young man, that I might have, from time to time, lived in suspended animation from century to century, and from generation to generation as it will now be possible for human beings to do."⁴   This would not be continued conscious existence with stroboscopic memory, but rather stroboscopic consciousness of a certain sort. In some novels, cloning gives characters a form of immortality. For example, in Heinlein's Time Enough for Love [1973], Lazarus Long is more or less cloned as his own daughters.

Another biological method of achieving immortality consists in so-called body transfers, which presupposes the falsity of the "bodily identity" criterion of personal identity. In The World of Null-A [1948] and The Players of Null-A [1956], by A. E. van Vogt, Gosseyn's consciousness transfers from one body [when it is destroyed] to another. As long as there are bodies, he can exist forever. Of course, conceptually one can distinguish between various sorts of body transfers. In some cases the brain is transferred to a different body. In others the brain itself is not transferred, but the mental state is, as in the film Invasion of the Body Snatchers [1956]; in some of the latter sorts of cases, there can be teleportation as well as mental transfer.⁵

There are also rather less exotic [thought by no means mundane] biological methods of generating and maintaining immortality, as portrayed in Mary Shelley's "The Mortal Immortal"[1810] and Larry Niven's Ringworld [1970] Ringworld Engineers [1979], and Protector [1973], all of which feature immortal beings. In Shelley's story, a young apprentice drinks the creation of his master and becomes immortal. In Niven's books, individuals live for centuries by using a drug especially tailored for their chemistry; without the drug, they die. Any human who eats the Spice of Life becomes a Pak Protector. Human Protectors undergo a physical change that makes them almost unrecognizable as human and a mental change that makes them protect whatever society they are in at the moment. These beings seem to be biological analogues to Isaac Asimov's robots, and they follow laws [instincts, in this case] similar to his Laws of Robotics.

There are also nonbiological methods of generating and maintaining immortality. In Neuromancer [1984 William Gibson creates a kind of human immortality by allowing the transfer of human mental states to computers. In Gregory Benford's Great Sky River [1987] the transfer is accomplished through the insertion of computer chips into the human, resulting in a combination of biological and mechanical capabilities: though the body may die, the "mind" continues. These procedures involve mental transfer ["downloading" of the "mind"] not accompanied by actual brain transfer.⁶

Other SF authors increase human longevity or create immortality by augmenting or supplanting normal human biological capacities through mechanical means. In McCaffrey's The Ship Who Sang [1969], a future society trains deformed by mentally functional babies to work in cyborg-type bodies if the parents so choose. This falls under the rubric of cyborg-type models of generating and maintaining atomistic nonserial immortality. In other works robots are created and then allegedly made sentient. Their mechanical nature makes them more or less immortal. Thus, in Asimov's Robot and Foundation series, Daneel first acquires a feel for human phenomena and then leans more and more toward the human, becoming telepathic to get a better insight into human nature and reasoning. Finally, he makes plans to transfer his knowledge and memories [which in a robot are also his essence] to the brain of a Solarian child, thereby becoming mortal. But, it is a prolonged mortality because Solarians, like all Spacers, live three or four hundred years. Further, if he can perform the operation once, he can do it again, particularly because this child is a hermaphrodite who will produce at least on offspring that is for all intents and purposes "itself." So, unlike Andrew Martin in "The Bicentenial Man," Daneel leaves his "option" of immortality open.

One other nonbiological way to produce immortality [one might call it "relativistic" immortality] involves time travel, as in Joe Haldeman's The Forever War [1974], in which time travel paradoxes are manipulated to achieve a sort of immortality.⁷

Having briefly surveyed the methods of generating and maintaining immortality [in particular, atomistic nonserial immortality], we not turn to the nature of immortal lives--their relationship to other lives and the pattern and distribution of their experiences.

First, consider a kind of solipsistic model. Heinlein's "'-All You Zombies-'" [1958] features an endless temporal loop in which the main character is a man who travels in time. But the pattern of his time travels indicates that he is in fact his own father, mother, and baby. There are other, nonsolipsistic, conceptions of the nature of atomistic nonserial immortality. One posits the "lone immortal" who lives among other individuals, all mortals. There are at least two versions of the lone immortal model--;one in which the lone immortal is known by [certain] others to be immortal, and another in which the immortality is a secret. such models appear, respectively, in Asimov's The End of Eternity [1955] and Shelley's "The Mortal Immortal." In another conception the immortal is not alone--perhaps others are immortal, as in Methuselah's Children, or perhaps everyone is immortal.

We have found problems with all models of immortality except for atomistic nonserial approaches. Nonatomistic models do not seem to meet the identity criterion, and some atomistic models--run afoul of the attractiveness criterion [if not also the identity condition]. but what about the atomistic nonserial models of immortality? Surely some methods of generating and maintaining immortality [such as feeding off the blood and vitality of others] make the resulting immortality less attractive. And some pictures of the nature of such immortality [such as the solopsistic model] make immortality unappealing. but not all methods of generating immortality are similarly problematic, and not all concepts of the nature of such immortality are straightforwardly problematic. Is there something about the nature of atomistic nonserial immortality that renders it, on reflections, necessarily undesirable?

Though some philosophers argue for this undesirability,⁸ some SF models are not so pessimistic. A common trait in science fiction is its positivistic faith in the ability of technology to accelerate the moment in the process of history when desirable immortality can be experienced. And today, there is already the hope that the human life span can be extended [through cryonics, for example] long enough to allow us to outlive the immediate causes of death and in a sense live to see the dawn of immortality. Yet, SF has negative models, too, and can be every bit as critical of positivist aspirations as are many philosophers.⁹   One brief example: though some SF novels depict efforts to achieve immortality through transformation into robots or mechanical beings, perhaps an equal number offer the opposite maneuver: a reverse immortality, or "Pinocchio Syndrome," in which an immortal strives to become mortal [not to die, but to become "subject to mortality"]. Somehow, even facing the prospect of immortal existence, human [mortal] qualities still retain such value that they are worth the reversal.

Despite a certain symbiosis between models in SF and philosophy, science fiction may be the more open to the possibility of transformation of the human body and life span. But in the end, is SF any more willing to abandon human limits? That vast and intriguing question is, unfortunately, beyond the scope of this essay.

NOTES