Living Value Theory holds that life is recursive mediation in a mesocosm. This article asks what death is by holding that proposition against its most severe test, an animal that can apparently leave the condition of being alive and return to it. The tardigrade, entering anhydrobiosis under desiccation, forces a distinction between three states rather than two: realised recursive mediation, an organised suspension that preserves the capacity for its return, and the irreversible loss of that capacity.
The article argues that most binaries through which humans classify the mesocosm are classificatory achievements built by suppressing continuous variation, but that life and death form an exception: a categorical difference between recursive and nonrecursive organisation persists beneath the empirical difficulty of locating its boundary in particular cases, once that boundary is specified for a given organisation at a given scale. Death is defined as the irreversible loss of a living organisation's capacity to be remediated into recursive mesocosmic coordination, a capacity this article argues concerns only the three zoetic mediations, embodiment, being-with, and dwelling, since the tardigrade, like the overwhelming majority of life, never engages the two poietic mediations at all. The definition is tested against hibernation, seed dormancy, organ transplantation, and, as a deliberately uncomfortable final case, a powered-down computational system, in order to show why recoverability alone cannot distinguish life from its absence and why living recursivity must be distinguished from recursive form.
I. The One Real Binary: Why Death Is Different
Among the innumerable binaries through which humans classify the mesocosm, most turn out on inspection to be classificatory achievements rather than direct transcriptions of how the world is divided. Nature and culture, male and female, simple and complex, human and animal, sacred and profane: each of these oppositions selects certain differences as decisive, suppresses others that cut across it, and stabilises a continuum into two portable classes. None of this makes such binaries useless. Protective and mechanical L4 abstraction of exactly this kind is often indispensable to coordination at scale. But it does mean that the appearance of a binary's self-evidence is itself something that had to be produced, sustained, and, in many documented cases, later reopened once the coordination it served began to strain against cases it had not been built to hold.
Life and death look, at first glance, like one more instance of this pattern.
Difficult cases abound. Clinicians disagree about the moment of death in particular patients. Legal definitions have shifted across jurisdictions and decades, from cardiopulmonary criteria to whole-brain and higher-brain criteria and back again under public pressure (Wijdicks, 2001). Cultural traditions attribute continuing presence, hearing, and even agency to the dead long after any biological criterion would locate life's end. If a theory of recursivity is committed to showing how binaries dissolve into classificatory labour, life and death would seem an obvious next target.
This article argues the opposite. Life and death are not one more L4 binary awaiting the deconstructive treatment given to nature and culture or male and female. Difficulty in locating a boundary is not the same thing as absence of a boundary, and the argument that follows turns on keeping these two questions rigorously apart. Epistemological uncertainty about exactly where a line falls does not, by itself, establish that the line is a mere artefact of classification rather than a mark of something real. There really is a categorical difference between a living animal and its corpse, and the difficulty many practitioners have in identifying the precise moment at which one becomes the other is a fact about the limits of observation and instrumentation, not a fact about the underlying organisation of the animal.
The clearest way to see this is to place a living animal and its corpse side by side in careful, deliberately reductive detail. Consider an animal in the instant immediately before death and the same animal in the instant immediately after. Its mass is unchanged, or as near to unchanged as any measurement could register. Its DNA is intact in essentially every cell. Its organs are, for some interval, structurally continuous with what they were a moment before. Its chemical constituents have not been replaced. Its shape, its location, its relations to the objects immediately surrounding it: none of these have shifted in the interval that separates the two states. Many of its cells, in fact, remain alive for minutes, hours, or in some tissues considerably longer, still carrying out the metabolic processes that defined them as living components a moment before. And yet something categorical has happened.
Something has disappeared that no inventory of the animal's remaining materials will locate.
One qualification has to be entered before the argument proceeds further, because otherwise the claim just made invites an objection the rest of the article is built to answer rather than to evade. The categorical difference asserted here cannot yet be stated as a claim about a single material object being simply alive or simply dead, because a corpse, as later sections will show in detail, continues to contain a great deal of unambiguously living material at smaller scales: individual cells, resident microorganisms, in some cases whole organs. The categorical difference is real, but it will turn out to hold of a specified recursive organisation at a specified scale, not of an undifferentiated lump of matter considered as a whole. Holding this qualification in view without resolving it immediately keeps the argument honest about a complication the tardigrade and the corpse will both be used to confront directly.
This is the question the rest of the article works to answer with precision. What has disappeared?
The question already rules out two of the most tempting shortcuts. It cannot be materiality, because the corpse remains entirely materially present, indeed multimaterially entangled with everything around it in ways that a living body, always partly resistant, partly absorbing, partly transforming its surroundings, was not. A corpse can be moved, weighed, buried, embalmed, or left to decompose, and every one of these is a material relation the corpse sustains perfectly well. Nor can it be embodiment as such, because the corpse is still a body: it has the shape, the organs, the surfaces of one. What has disappeared is not a substance called life sitting somewhere inside the tissue, but a particular organisation of that same matter, an organisation through which the animal's own prior states were altering the conditions of its own subsequent states.
This is precisely the phenomenon Living Value Theory names recursivity, and the proposition that life is recursive mediation in a mesocosm can now be brought to bear on death directly rather than only on life. Recursivity names the capacity of a process for the consequences of its own prior mediation to alter the conditions of its subsequent mediation. Mediation insists that this capacity never exists as an isolated interior property of an organism considered apart from its surroundings, but always operates through some differentiated relation: a body coordinating its own continuation, a living being coordinating another, a found or made ground within which either occurs. Mesocosm names the field of availability, dependency, and consequence within which these mediations become possible at all.
Selfrecursive, interrecursive, and nonrecursive name not three different strengths of recursivity but three different kinds of counterpart a recursive process can be coordinating with: its own continuing state, another recursively responsive being, or a ground that does not itself respond in turn. Recursivity itself is not a matter of degree; what varies is only which of these counterparts is in play, and how many mediations are engaged in coordinating them. An organism does not become recursive in the abstract. It is recursively mediating a mesocosm, or it has ceased to be able to.
Death, on this initial formulation, is not the disappearance of a substance but the collapse of an organised capacity for recursive mediation. This formulation, however, runs immediately into the problem around which the rest of the article turns. Living organisms regularly and routinely suspend enormous ranges of recursive activity without dying. They sleep, losing most conscious responsiveness for hours at a stretch. They are anaesthetised, losing nearly all of it. They hibernate for months. Seeds remain dormant for years, in some documented cases for millennia. Spores persist indefinitely. And certain animals, of which the tardigrade is the most severe and most instructive case, can apparently cease almost every recursive activity by which an organism is ordinarily recognised as presently alive, and later resume them as though nothing categorical had occurred in between.
If present recursive activity is what distinguishes the living animal from its corpse, these cases become impossible to explain. If a tardigrade in cryptobiosis is, for the whole duration of its dormancy, indistinguishable in its measurable activity from a dead tardigrade, and yet one of them is alive and the other is not, then the distinction between recursive and nonrecursive organisation cannot rest on what is presently, observably occurring. It must rest on something else, and locating that something else is the task the tardigrade will be used to perform with unusual severity.
II. The Animal That Disappears from Life: Entering Cryptobiosis
Tardigrades are microscopic, eight-legged invertebrates found in marine sediments, freshwater habitats, and, most relevantly here, the thin films of water that coat mosses, lichens, and soil particles in terrestrial environments. Terrestrial tardigrades generally require this film of surrounding water to remain in ordinary recursive activity: to move, feed, sense their surroundings, and reproduce. Remove the water, and most animals of comparable size and physiology would simply die of desiccation within a short span. Many tardigrade species instead enter anhydrobiosis, one instance of the broader phenomenon of cryptobiosis, in which the animal survives the loss of nearly all its body water by entering a profoundly altered morphological and physiological state (Wełnicz et al., 2011).
The transition is visible and dramatic. As desiccation proceeds, the animal retracts its legs and head, contracts along its body axis, and forms a compact, barrel-shaped structure known as the tun. In this state the tardigrade's water content, ordinarily comprising the great majority of its body mass, falls to a small fraction of what it was, and the tun can persist for extended periods before rehydration is attempted. Precision matters more than drama here. Systematic investigation of specimens kept dry for as long as nine years confirmed revival in tardigrade eggs, and popular claims of survival across many decades or even centuries, though frequently repeated, have not survived close scrutiny of the underlying evidence and should be treated with real caution (Guidetti and Jönsson, 2002; Wełnicz et al., 2011).
Care is needed with a claim frequently repeated in popular accounts of the tun, namely that the tardigrade's metabolism during cryptobiosis reaches literally zero.
The empirical literature is considerably more careful than this. Reviews of anhydrobiosis in tardigrades generally describe metabolic activity during the ametabolic tun state as arrested, virtually undetectable, or reversibly halted by ordinary measurement, rather than proven to be absolutely and instrumentally zero (Wełnicz et al., 2011). Nothing in the argument that follows requires the stronger, unproven claim. What matters is that cryptobiosis pushes observable biological activity toward the limit at which the ordinary criteria by which an organism is recognised as presently alive, movement, feeding, sensory responsiveness, measurable metabolic turnover, cease to function as reliable evidence one way or the other. This is precisely what gives the tardigrade its philosophical usefulness. It does not need to achieve a metaphysically pure zero to pose the problem. It only needs to fall far enough that ordinary observation goes silent.
This creates the central comparison around which the rest of the argument is built: a living, active tardigrade going about its ordinary business among mosses and soil particles; a cryptobiotic tardigrade contracted into its tun; and a dead tardigrade, its tissues no longer capable of any renewed activity under any subsequent condition. The cryptobiotic and the dead tardigrade confound almost every intuitive criterion an observer might reach for. Neither moves. Neither feeds.
Neither reproduces. Neither displays the ordinary sensory responsiveness by which a living animal is recognised as engaging its surroundings. Neither is, at the moment of observation, presently coordinating anything interrecursively with another animal. Whatever metabolic measurement can be performed on either specimen is likely to return results so faint as to offer little immediately obvious evidence of the difference between them.
And yet one is dead and the other is not.
The biology available makes this puzzle sharper rather than resolving it by decree. Entry into cryptobiosis is not something that merely happens to a passive tardigrade once its surrounding water disappears, the way a puddle dries or a plant wilts. It is an organised, actively produced transition. Tun formation involves substantial morphological reorganisation of the animal's own body, and molecular research over the past decade has increasingly clarified the mechanisms by which tardigrade tissues are protected from the damage that desiccation would otherwise inflict on ordinary cellular structures. Boothby and colleagues showed that several tardigrade species express a family of tardigrade-specific, intrinsically disordered proteins, generally known by the abbreviation CAHS for cytoplasmic abundant heat-soluble, that are strongly upregulated during the drying process and that appear, both inside tardigrade cells and when introduced into other organisms entirely, to confer substantial protection against desiccation damage (Boothby et al., 2017). Different tardigrade lineages rely on somewhat different combinations of protective mechanisms, so there is no need, and no empirical warrant, for treating any single molecule as the sole secret of tardigrade survival. What matters for the present argument is the general shape of the finding: the protection that allows the tun to survive desiccation is manufactured by the living animal before its ordinary recursive activity becomes suspended, not imposed on an inert body afterward by the desiccating environment itself.
This distinction is conceptually decisive, and it is stated here with unusual precision because everything that follows depends on it. Cryptobiosis is not simply something that happens to the tardigrade once its environment disappears. The transition into the tun is produced by the still-active, still-recursively-responsive organism in advance of the interruption it is preparing to survive. Life actively participates in producing the very material organisation through which its own interruption becomes survivable rather than fatal.
This licenses a formulation that will sound paradoxical until its temporal structure is made explicit: the tardigrade recursively prepares for nonrecursivity.
While still recursively responsive to the deteriorating conditions around it, sensing falling humidity, initiating the expression of protective proteins, reorganising its own tissues, the animal changes itself in ways calculated, in the evolutionary rather than deliberative sense, to preserve the possibility of later reactivation. The active animal produces the conditions under which active animality can disappear from observation without thereby becoming death. What is suspended is the ongoing exercise of recursive mediation across embodiment, being-with, and dwelling, the three zoetic mediations, which are the only mediations a tardigrade's life has ever engaged. What is not suspended, because it has already been discharged in advance, is the organised capacity for that same zoetic mediation to resume once appropriate conditions return.
This restriction to the three zoetic mediations holds well beyond the tardigrade, and it identifies something general about the whole question this article is pursuing.
Multimateriality and multisymbolization, the two poietic mediations, require a sustained, open-ended capacity to deposit recursive coordination into durable material and symbolic form, a capacity found, so far as the evidence goes, only in a small number of lineages and developed to an extraordinary degree in exactly one of them. A tardigrade does not build, does not mark, and does not inherit anything resembling a tool, a path, or a symbol. Its entire life, active or suspended, is conducted through embodiment, being-with, and dwelling alone, at L1 and L2, without ever touching the poietic mediations at all. The same is true of the bacterium, the plant, and the overwhelming majority of animal life across the whole of evolutionary history. If the tardigrade nonetheless poses the life and death question with full force, and it plainly does, then whatever life and death turn out to mean cannot depend on multimateriality or multisymbolization in any way. The entire question this article is asking, framed at its most general, is a question about zoetic sufficiency: the three zoetic mediations, coordinated through L1 and L2 recursivity alone, are enough to constitute a complete living mesocosm, and death, whatever else it requires, must be specifiable entirely in these zoetic terms. Poiesis, when it eventually arrives in certain lineages, transforms what a life can do. It does not touch the more basic question of whether that life is alive at all.
This gives cryptobiosis a theoretical status quite different from ordinary biological adaptation to a harsh environment. Adaptation to cold, to low oxygen, or to nutrient scarcity typically involves the organism continuing to operate, often at reduced tempo, within an ongoing coordination with its surroundings. Cryptobiosis is something more radical: an evolved capacity for surviving a complete and, for the observer, indistinguishable-from-fatal breakdown in present mesocosmic coordination, by converting the very capacities that made that coordination possible into a durable, dormant configuration capable of being returned to activity when the mesocosm becomes viable again. Whether that return occurs, and what exactly must be restored for it to occur, is the question the next part takes to its most extreme test.
III. A Tardigrade in Space: Life outside the Mesocosm?
The tardigrade's capacity for cryptobiosis has attracted a great deal of popular attention under a formulation that repays close examination because it is precisely wrong in a theoretically productive way: the claim that tardigrades can live in space.
In 2007, desiccated tardigrades were carried aboard the European Space Agency's FOTON-M3 mission and exposed, for a period of roughly ten days, to the environment of low Earth orbit. Jönsson and colleagues found that tardigrades exposed to the vacuum of space alone survived at rates comparable to unexposed control specimens once rehydrated, and that even under the considerably more damaging condition of combined exposure to space vacuum and unfiltered solar radiation, some individuals survived and were subsequently capable of reproduction (Jönsson et al., 2008). This finding, rightly, made tardigrades famous.
It did not, however, show what popular retellings routinely claim it showed. The tardigrades in question did nothing remotely resembling ordinary tardigrade life during their time in orbit. They did not move, feed, sense, or reproduce while exposed. What the experiment demonstrated is that, under particular cryptobiotic conditions, a tardigrade's tun can survive an interval during which ordinary mesocosmic coordination has been entirely suspended, and can subsequently be returned to active life once appropriate conditions are restored.
Surviving outside viable mesocosmic coordination is not the same thing as living without a mesocosm, and the distinction corrects an error that is easy to fall into whenever an organism's resilience is described in dramatic terms. Tardigrade cryptobiosis does not demonstrate that an organism can dispense with a mesocosm and continue living regardless. It demonstrates almost the opposite. Remove liquid water and the other conditions that active tardigrade life depends upon, and recursive engagement across the five mediations collapses toward suspension.
Restore those conditions, and the capacity for recursive mediation, which has been preserved rather than exercised throughout the interval, can resume.
The space experiment can therefore be read as an almost literal experiment in mediation rather than merely an experiment in toughness. A tun can persist in an environment where an active tardigrade could not survive for an instant. But the tun cannot become an active tardigrade while still in that environment. It must be returned, physically and chemically, to conditions under which its living organisation can once again coordinate itself across embodiment, dwelling, and the relevant relations to other beings and materials, before anything recognisable as tardigrade life resumes.
This licenses the rejection of a picture the popular framing implicitly invites, in which some portion of life itself sits latent inside the tun, sealed away from the world and untouched by it, waiting to be switched back on. There is no such portion:
what persists through the interval of suspension is not a fragment of life held in reserve independently of any world, but a highly organised material configuration, produced in advance by the living animal, that is capable of becoming recursively active again once it is appropriately remediated by conditions outside itself.
The concept of remediation carries real theoretical weight at exactly this point, rather than serving as a loose synonym for restoring conditions. Rehydration of a tun is not simply an external stimulus that causes an otherwise inert mechanism to restart, in the way that pressing a switch restarts a machine that has been sitting powered off. Rehydration restores a condition under which the animal's own organised capacity for recursive mediation can once again actually operate. Water, in this transition, is not merely a chemical input but a medium through which the animal's relation to its dwelling becomes viable again. Temperature becomes, once more, part of a habitable dwelling rather than merely a physical variable recorded on an instrument. The animal's immediate material surroundings again become navigable rather than simply present. Sensory differences in that surrounding world again acquire relevance to the animal's own continuation. Food, once again, can become food rather than merely biomass in the vicinity of an inert tun. Other organisms can once again become prey, competitors, or simply co-present neighbours in a shared field of coordination. Every one of these restorations is zoetic. Water and temperature restore dwelling. Prey, competitors, and neighbours restore being-with. The reactivated tissue restores embodiment. Nothing in this remediation involves multimateriality or multisymbolization returning to activity, because the tardigrade never engaged them to begin with. Remediation, for this animal, is remediation into the three zoetic mediations and nothing further.
What this shows is that the mesocosm cannot be understood as a fixed container that exists independently, waiting to receive an organism that is inserted into or withdrawn from it. Mesocosmic coordination is achieved, continuously and specifically, between the organised capacities of a living being and the structured conditions that surround it. When the tun sits in low Earth orbit, there is, for the duration of that exposure, no tardigrade mesocosm actually being coordinated at all, because there is no ongoing relation between organised living capacity and surrounding conditions through which coordination could occur. What exists instead is a durable, protected material configuration awaiting the restoration of a relation it is not presently engaged in.
This gives the article its sharpest formulation so far, and one worth holding onto through everything that follows: the tardigrade can leave the conditions of active life without thereby becoming dead, because it retains, in fully organised and materially specific form, the possibility of returning to recursive mediation once a viable mesocosm is restored to it. Death, whatever else it will turn out to require, must be distinguished from this kind of departure. It cannot simply be the absence of present mesocosmic engagement, because the tun in orbit has exactly that absence and remains, by every criterion that matters, alive.
IV. Dormancy Is Not Death: The Temporality of Recursive Potential
The tardigrade generalises immediately once its lesson is made explicit. Seeds remaining dormant in soil for years, spores persisting through conditions that would kill the active organism outright, mammals hibernating through winters during which heart rate, respiration, and body temperature fall to a fraction of their active values (Carey, Andrews, and Martin, 2003), and human beings under general anaesthesia, in which most ordinary recursive responsiveness is deliberately and reversibly suppressed by pharmacological means (Brown, Lydic, and Schiff, 2010), are all instances of the same underlying structure the tardigrade makes visible with unusual severity. In every one of these cases, an organism can, for a substantial interval, display almost none of the behaviours through which humans ordinarily recognise living agency, and nevertheless remain unambiguously alive throughout.
The central argument this licenses is that life cannot be identified from a snapshot. A single instant of observation, however carefully instrumented, cannot by itself distinguish a hibernating animal from a dead one, a dormant seed from a dead one, or a cryptobiotic tardigrade from a desiccated corpse, because the observable difference between these pairs at any single moment can be vanishingly small or entirely absent. The converse failure is equally instructive and equally important to state. Something can display behaviour that is spectacularly life-like, responsive, adaptive, even apparently purposive, while remaining entirely nonliving, a possibility the article returns to directly in its confrontation with artificial systems. Neither direction of error, mistaking dormant life for death nor mistaking convincing behaviour for life, can be corrected by looking harder at a single moment. Both require attention to something that a snapshot cannot capture.
This makes temporality constitutive of the distinction between life and death rather than a supplementary detail added afterward. The difference between a dead tardigrade and a cryptobiotic one does not lie in anything currently observable about either specimen in isolation. It lies partly in their different futures under comparable conditions of remediation. Return both specimens to appropriate humidity and temperature. The cryptobiotic animal may, over a period ranging from minutes to days depending on the duration and severity of its dormancy, resume active recursive coordination: legs extending, movement resuming, feeding recommencing. The dead animal does nothing of the kind under any duration of the same treatment.
A caution is necessary here, because it would be easy to overcorrect into treating recoverability as though it were simply a fixed property the organism carries with it, independent of anything outside it, the way mass or volume might be treated.
This would be a mistake. A tardigrade does not recover in the abstract, considered apart from any relation to a world. Recovery requires the actual restoration of a viable relation between the organism's own organised capacities and surrounding conditions capable of sustaining renewed coordination. The capacity for recovery is real and it is a fact about the organism's own organisation, but its exercise is never something the organism achieves alone.
The tardigrade forces a distinction the argument has been assembling without yet stating outright, and it belongs among this article's central results rather than remaining implicit. Recursive mediation in a mesocosm, the proposition this whole architecture begins from, cannot mean only presently occurring recursive mediation, on pain of making the cryptobiotic tardigrade count as dead while it sits in its tun. Three states, not two, need to be distinguished. Active life is realised recursive mediation in a mesocosm: the coordination is actually, presently occurring. Dormant life is an organised suspension that preserves the capacity for renewed recursive mediation in a mesocosm: the coordination is not occurring, but the organisation capable of resuming it remains intact and materially specific.
Death is the irreversible loss of that capacity: no organisation remains through which recursive mediation could resume, however favourable the conditions subsequently offered to it. This is not a retreat from the founding proposition but that proposition's temporal completion, and the tardigrade is what forces it into view, because an ordinary continuously active organism never presents a case in which the three states need to be told apart. Being alive, the tardigrade shows, cannot be identical with presently performing life.
Dormancy, in the vocabulary just introduced, is the middle state: not a deeper or more prolonged version of death, but a different kind of thing altogether, categorically closer to active life than to death despite resembling death by every external measure available at the time.
This also clarifies why metabolism cannot bear the entire conceptual burden that popular accounts sometimes place on it. Metabolic activity is, for the great majority of terrestrial life, an enormously important and usually reliable indicator of ongoing recursive coordination, and nothing in this argument diminishes its importance in the overwhelming run of ordinary cases. But the tardigrade demonstrates decisively that the rate of presently measurable metabolism does not map cleanly onto the life or death distinction at its outer limit. Reviews of tardigrade anhydrobiosis describe metabolic activity during the ametabolic tun state as falling to levels near or below the threshold of ordinary detection, while the same animal remains fully capable of revival once rehydrated under appropriate conditions (Wełnicz et al., 2011). If metabolism alone defined life, this pairing of near-undetectable metabolic activity with intact capacity for revival would be paradoxical. Once life is understood instead as an organised capacity for recursive mediation, rather than as a currently running metabolic process, the pairing ceases to be paradoxical and becomes exactly what the theory predicts.
The seed makes the same point in a different register. A dormant seed and a dead seed can be materially extraordinarily similar: comparable mass, comparable chemical composition, comparable structural integrity under casual inspection.
What distinguishes them becomes apparent only through their different relations to subsequent conditions. Soil, moisture, appropriate temperature, and in many species specific cues such as fire, cold stratification, or passage through a digestive tract, do not merely act upon a dormant seed the way heat acts upon any inert chemical mixture. Under conditions the seed's own dormant organisation is prepared to respond to, these same environmental factors permit the re-emergence of a recursively organised, actively growing plant, in a way that no comparable treatment of a dead seed will ever produce (Finch-Savage and Leubner-Metzger, 2006).
The conceptual advance this licenses is that potentiality itself must be specified mesocosmically rather than treated as an abstract capacity floating free of any particular relation to a world. There is no such thing as an abstract potential for life sitting inside a seed, a spore, or a tun, but rather, in each case, an organised capacity for renewed recursive coordination under some specific and empirically determinable range of environmental conditions, a capacity that is real, that varies from species to species and even from individual to individual, and that can itself be lost even while the material substrate that once carried it remains largely intact.
One further caution belongs here before the argument moves to matters of scale. Nothing in this article should be read as claiming that a living organisation must be recursively active across all five mediations at once in order to count as alive. The tardigrade, even fully active, barely touches being-with in any developed sense, and, as the previous part argued, never touches multimateriality or multisymbolization at all. The five mediations describe the differentiated dimensions into which recursive mediation can develop, not five simultaneous requirements every living thing must satisfy. Different organisms have radically different mediational profiles, from the bacterium coordinating almost nothing beyond its own embodiment and its immediate dwelling, to the human being whose life is saturated with all five at once. What life and death require, at the most general level this article has been assembling, is only that some recursive mediation across some subset of the zoetic mediations remains, in the dormant case, capable of resuming, or, in the active case, is actually occurring.
Dormancy can be understood as an extreme form of closure imposed on active recursive engagement, a closure that remains, crucially, reopenable given the right conditions. Death is not merely another instance of this closure, deeper or more prolonged than the others. It is the loss of reopenability for that particular recursive organisation altogether.
This yields the article's first serious working definition. Death is the irreversible loss of the capacity of a living organisation to be remediated into recursive mesocosmic coordination.
The word irreversible in this formulation will immediately require qualification, and the qualification matters enough to state clearly before the definition is put to further use. Irreversibility is not a fixed, timeless property but depends on the interventions, the scale of restoration attempted, and the point in history at which the question is asked. Advances in resuscitation medicine have repeatedly made recoverable, by later standards, conditions that earlier medical practice would have classified, in good faith and using the best available criteria, as irreversible death.
The definition offered here cannot therefore identify death with whatever a particular historical medical procedure happens to be capable of reversing at a given moment, on pain of making death itself a moving target defined by the state of human technology rather than by anything about the organism in question.
Even this refinement is not yet precise enough, because appealing to what remains recoverable in principle invites a thought experiment that threatens to make death unrecoverable even from itself. Imagine a future technology capable of reconstructing a dead tardigrade molecule by molecule, restoring every protein, every membrane, every structural detail to its exact pre-mortem configuration.
Would this reverse the animal's death, or would it manufacture a second tardigrade that merely resembles the first down to the last atom? The same question arises, at a much more modest level of speculation, for any sufficiently aggressive programme of repairing a recently dead organism until organismic activity resumes: at what point does remediation of a persisting organisation become reconstruction of a new one from a destroyed original?
The relevant concept is organisational continuity rather than mere physical similarity, however exact. Rehydrating a tun reactivates an organisation that has persisted continuously throughout the interval of dormancy, materially unbroken from the active animal that produced it to the active animal that emerges from it.
Resuscitating a person whose heart has stopped restores a disrupted but still continuously existing organisation, one whose integration has been interrupted rather than destroyed, and whose components have not yet crossed whatever threshold makes their coordinated resumption impossible. Reconstructing an organism after its organisation has actually been destroyed, however faithfully, produces a new instance of that organisation rather than a return of the original, in exactly the way a perfect replica of a burned manuscript is a new manuscript rather than the reappearance of the one that burned. Continuity of organisation, not mere similarity of eventual configuration, is what remediation requires and what reconstruction cannot supply. What must remain invariant across historical shifts in medical capability, then, is not what humans are presently capable of reversing, but whether the organisation capable of sustaining that particular living process remains, whatever the available technology, continuous enough for recursive mediation to resume within it, rather than merely be reconstructed alongside it.
V. What Exactly Dies? Organisms, Cells, Organs, and Recursive
Organisation The argument so far has treated death somewhat loosely as something that happens to an organism as a whole. This part confronts the hardest consequence of taking recursive organisation, rather than a single fixed spatial boundary, as the relevant unit of analysis.
A dead organism, by any ordinary criterion of death applied to the whole animal, continues to contain enormous quantities of living material. Individual cells throughout the body remain metabolically active for a considerable interval after the organism as a whole has died, in some tissues for many hours. An organ removed from a body that has, by every relevant clinical and legal criterion, died, can continue to function, sometimes for a considerable interval under appropriate perfusion, and can subsequently be transplanted into another body where it resumes functioning as part of an entirely different organism's ongoing life, a possibility organised, in practice, around carefully specified circulatory criteria for the donor's own death (Manara, Murphy, and O'Callaghan, 2012). Microorganisms already resident in and on the body continue their own living processes uninterrupted by the death of their host, and are shortly joined by others colonising the newly available resource the corpse represents. Molecular processes of many kinds continue for an extended period. Decomposition itself is not the absence of biological activity but an extraordinary proliferation of it, as an entire ecology of bacteria, fungi, and invertebrates takes up residence in and metabolises what remains, releasing a concentrated pulse of carbon and nutrients into the surrounding soil (Carter, Yellowlees, and Tibbett, 2007).
Death, therefore, cannot mean that matter becomes nonliving in any absolute or total sense at the moment an organism dies. This was already implicit in the corpse comparison that opened the article, but it must be confronted directly rather than left as an unresolved tension, because it looks, at first glance, like a serious objection to everything argued so far. If living cells persist throughout and after the moment of death, in what sense has anything died at all?
The answer requires the distinction between the five mediations and the recursive counterparts they coordinate to do real analytic work rather than remaining a piece of background vocabulary. What dies, in the case that matters for the ordinary meaning of the word, is not every constituent living process at every scale simultaneously, but a particular recursive organisation at a particular scale:
the organisation through which processes occurring at the level of cells, tissues, and organs were being coordinated together as the continuing existence of one organism, capable of mediating its own embodiment, its relations to other living beings, and its dwelling as a single, integrated locus of recursive activity. An organism is not alive merely because intracellular biochemical reactions continue somewhere within its tissues. Organismal life consists in the recursively coordinated integration through which processes occurring across these different scales contribute, in a mutually adjusted and ongoing way, to the continued existence and self-modification of that particular being as a whole.
Death, on this account, is a collapse of integration rather than an instantaneous cessation of every constituent living process the organism's body happens to contain. This reformulation removes the apparent paradox of organ transplantation entirely. The donor, by the relevant clinical and legal criteria, has died: the recursive organisation that integrated the donor's cells, tissues, and organs into one coordinating whole has irreversibly collapsed. This is entirely compatible with certain subordinate living organisations within the donor's body, an organ maintained under perfusion, for instance, remaining viable for a further interval and capable of being incorporated into the recursive organisation of a different, living organism. Nothing mysterious has occurred in this transition. The organism-level organisation that constituted the donor as a single integrated living being has ended. Subordinate organisations that were part of it, and that do not themselves require the whole organism's continued integration in order to remain viable for a time, have not necessarily ended at the same moment.
The corpse itself becomes, on this view, theoretically revealing rather than merely a leftover to be disposed of once the interesting question has been settled.
After death, the body remains present and highly consequential within the mesocosms of others. It becomes food for microorganisms and, depending on circumstance, for larger scavengers. It becomes an object of mourning, of ritual preparation, of legal classification, of forensic examination. It may remain, for a long time and in ways ethnography has documented across an enormous range of societies, central to the ongoing selfrecursive and interrecursive lives of the survivors who continue to relate to it, to its memory, and to whatever social position it continues to occupy.
Yet every one of these forms of continuing relevance is relevance for other living beings, not evidence of continuing recursive mediation on the part of the corpse itself. The corpse no longer recursively mediates its own relations. It has become what this architecture calls a nonrecursive counterpart: something that can matter enormously to the living, that can organise mourning, inheritance, and memory, and that can even be treated, in many documented traditions, as though it retained some form of ongoing responsiveness, without the corpse itself thereby regaining the recursive organisation that has, in fact, irreversibly ended.
This creates a clean bridge to comparative ethnography. Human societies very frequently attribute continuing recursive presence to their dead. Ancestors may be understood to hear petitions, to intervene in the affairs of the living, to demand particular forms of remembrance, to protect or to punish, to remain active members of an ongoing social and moral world in ways that are neither vague nor marginal to the societies that hold them, and mortuary practice across a wide range of societies is organised precisely around the passage from one such status to another (Bloch and Parry, 1982). These ethnographic phenomena are real, they are consequential, and they deserve to be analysed with the same seriousness given to any other form of recursive attribution: the mesocosmic, contestable, and historically variable process through which the living settle on an answer to the question of what kind of responsive being a particular dead person, or the dead more generally, is being coordinated as. Analysing these attributions with this seriousness does not require Living Value Theory to conclude, on either side, that the ontological status of the dead is exhausted by whatever a given culture happens to attribute to them. The distinction between recursive attribution, which is always mesocosmic and always contestable, and recursive status, which the present argument treats as a categorical fact about organisation rather than a settled cultural verdict, allows both questions to be investigated without collapsing one into the other.
The larger methodological point this section is building toward is that the existence of difficult cases, transplantation, decomposition, ancestor cults, does not demonstrate that life and death form an overburdened L4 category straining under cases it cannot hold, in the way many familiar binaries do. It demonstrates instead that a real binary can nevertheless be difficult to apply in particular cases, because recursive organisation exists simultaneously at multiple nested scales, from the cell to the tissue to the organ to the whole organism, and death at one scale need not coincide in time with death, or the absence of life, at another. The question that should replace the deceptively simple is this alive, in exactly the cases that make the simple question hardest to answer, is a more precise one: which organisation are we asking about, and is that organisation, at the scale in question, capable of recursive mediation in a mesocosm? The hard binary this article has been defending, in other words, was never simply a material object being alive or dead, but always a specified recursive organisation, at a specified scale, being one or the other, a formulation considerably more precise than the one with which the article opened and one that protects the binary against every objection that living cells inside a corpse could otherwise be used to mount.
VI. The Tardigrade and the Server: Why Restarting Is Not Revival
The sixth part introduces a deliberately uncomfortable comparison, because comfort at this point would be a sign that the argument has not yet been tested against its hardest case.
Place beside the cryptobiotic tardigrade a powered-down computational server, the kind of machine on which a large artificial intelligence system might run. Both are materially organised in complex and specific ways. Neither, in its present dormant state, presently behaves. Neither presently communicates with anything in its surroundings. Neither presently performs any observable information processing. Both require the restoration of appropriate external conditions before activity resumes: the tardigrade requires water, appropriate temperature, and the other conditions of a viable dwelling; the server requires electricity, cooling, and, for most of what it does, network connectivity. Restore these conditions to either, and both can, in the relevant sense, come back.
If recoverability alone were sufficient to define life, the server would appear to pass the test as readily as the tardigrade does. This should be treated as a serious problem for the argument as developed so far, not brushed aside, because it shows directly that recoverability by itself cannot be the criterion that finally distinguishes the living dormant organism from the merely switched-off machine.
Hubert Dreyfus's long insistence that intelligence requires embodiment is relevant here (Dreyfus, 1992), but it turns out, on close examination, to be insufficient on its own to resolve the comparison. The server is materially embodied in the ordinary sense: it has a physical structure, occupies space, and is causally entangled with its surroundings through heat, electricity, and data. A sufficiently elaborate robot could possess a considerably richer body still, equipped with cameras functioning as something like vision, microphones functioning as something like hearing, pressure sensors functioning as something like touch, and manipulators and locomotor systems functioning as something like limbs. None of this, however elaborate the engineering, by itself establishes that the resulting system possesses living multisensorial embodiment in the sense this architecture has developed. Attaching more interfaces to a computational system multiplies its channels of interaction with its surroundings without thereby establishing that those channels participate in an organised capacity for recursive mediation of the kind a living body exercises.
The relevant question, in other words, is not whether something possesses a body, a criterion the server and any sufficiently equipped robot can satisfy trivially, but what kind of organisation that body participates in.
The tardigrade, on rehydration, returns to recursive mediation in the full sense this architecture has developed throughout. Differences in its immediate surroundings become, once again, relevant to the continuation of its own living process: a threat must be avoided, food must be located and ingested, damaged tissue must be repaired, favourable conditions must be moved toward and unfavourable ones away from. It senses, moves, feeds, avoids, grows, and repairs itself, and it does all of this by entering selfrecursive, interrecursive, and nonrecursive relations across embodiment, being-with, and dwelling, the same three zoetic mediations discussed throughout, each modifying the conditions of the others in the way characteristic of living organisation generally.
The server, on being powered up, resumes computation. This is not a dismissive description intended to belittle what contemporary computational systems can do, but a precise one, and the precision matters because the comparison is only useful if it is not allowed to be won or lost by rhetorical framing on either side.
This distinction cannot be secured merely by choosing more flattering or more deflationary vocabulary for one side of the comparison, and the argument must be honest about the difficulty this creates for itself. Computational systems, including contemporary artificial intelligence systems, genuinely contain feedback loops, recurrent architectures in the technical, mathematical sense, self-modification through training and fine-tuning, error correction, extensive memory, predictive modelling of their own future outputs, and complex interactions among multiple interacting components or agents. If all such processes were simply called recursive in the sense this architecture has developed, the theory would collapse into a restatement of ordinary cybernetics. The distinction between recursive form, an operation mathematically or computationally defined in terms of its own prior states, and living recursivity has to be insisted upon here with full force, not merely asserted as a definitional stipulation convenient to the argument. The distinction cannot rest solely on the observation that a living process's own prior states alter the conditions under which its continuation can go better or worse, because a sufficiently sophisticated autonomous machine can also be built so that certain states advance its continued operation and others threaten it. A computer scientist presented with only that formulation could reasonably object that the machine now has things going better or worse for it too, in whatever sense a stored objective function can be said to register such a thing. The deeper difference has to be mesocosmic and mediational rather than a matter of optimisation alone. The tardigrade is not executing a stored objective function that represents survival as a target state toward which its behaviour is directed. Its own continuing organisation exists only through recursively changing relations with a world that is never represented, by the animal, as a fixed set of variables upon which it subsequently acts from the outside. The living process and the field of what is relevant to it emerge together, each altering the other, rather than one computing over a model of the other held constant. A machine optimising against a programmed objective function, however adaptively, is still coordinating a representation of a world rather than being recursively constituted by its relation to one, and this is the distinction the vocabulary of mediation and mesocosm exists to make precise where the vocabulary of goals and objectives cannot.
A further complication should not be smoothed over. An artificial system can participate enormously, and increasingly, in human mesocosms without thereby possessing a mesocosm of its own. It can change what human beings think, decide, and do. It can mediate symbolic relations between people who never interact with each other directly. It can control material systems of considerable consequence. It can become incorporated into institutions in ways that reorganise how those institutions coordinate the people who depend on them. None of this causal participation in living recursive mediation is equivalent to being, itself, a recursively living participant in that mediation. A letter, a law, and an architectural plan all participate enormously in the mesocosms of the living without themselves being alive, and an artificial system, however much more dynamically it participates, does not automatically cross that line simply by virtue of participating more actively or more responsively than a letter does.
The server's own material existence, moreover, is entirely poietic rather than zoetic. Its casing, its circuitry, and its physical plant are multimateriality: made nonrecursive ground, deposited by human labour and engineering, no different in kind from a building or a bridge. Its trained parameters, its code, and its outputs are multisymbolization: made nonrecursive symbolic ground, however dynamically it gets recombined at each invocation. The server has no embodiment of its own in the sense this article has developed throughout, no being-with of its own, and no dwelling of its own: it is, in its entirety, a deposit, an extraordinarily elaborate one, produced by and remaining dependent on the zoetic mesocosms of the human beings who built it, maintain it, and address it. This is the precise sense in which it participates in a mesocosm without having one: what it participates in is always someone else's zoetic coordination, mediated through poietic deposits it does not itself live through.
This produces one of the article's most compressed and, for that reason, most useful propositions: something can participate in a mesocosm without having a mesocosm of its own.
The question this leaves for the future of artificial systems is consequently not whether machines will eventually become sophisticated enough, as though sophistication alone were the missing ingredient and life simply the far end of a continuum of computational complexity. Nor should this architecture declare, as a matter of principle rather than evidence, that carbon-based chemistry is metaphysically indispensable to life, a claim it has no grounds to make and every reason to avoid making. The stronger and more useful criterion is substrate-neutral:
an artificial entity would have to become an organisation for which recursive mediation through a mesocosm constituted its own mode of continued existence, an organisation for which its own prior states altered the conditions under which its own continuation could go better or worse for it, rather than merely for the humans and institutions that depend on its outputs. Whether such an organisation could ever be produced by engineering remains an open question this architecture is not equipped to settle in advance. What this architecture does reject, on the grounds developed throughout this section, is the inference from computational complexity, however impressive, directly to life.
The tardigrade and the server therefore perform, in this final comparison, exactly opposite conceptual functions. The tardigrade can appear, by every available observational and metabolic measure, almost completely nonliving, and yet remains, throughout its dormancy, unambiguously on the living side of the boundary this article has been drawing. The artificial system can appear, by every available behavioural measure, extraordinarily alive, responsive, adaptive, even conversationally convincing, and yet remains, on the argument developed here, on the nonliving side of that same boundary. Relevant comparison, rigorously applied, overturns superficial resemblance in both directions at once, and it is this double reversal, rather than either case taken alone, that gives the comparison its force.
VII. What Is Death? From the Tardigrade Back to the Mystery of
Life The final part returns to a mystery the article has clarified rather than solved, and the difference between the two must be stated plainly.
Living Value Theory cannot explain how life originated. It cannot explain how carbon, water, membranes, proteins, nucleic acids, and the other nonrecursive materials out of which every known living organism is built first became organised into a process capable of recursive mediation at all. That is simultaneously a problem for origin-of-life research, for evolutionary biology, for chemistry, and for physics. Living Value Theory makes no attempt to fill this explanatory gap with metaphysics dressed up as theory.
What this architecture contributes instead is conceptual rather than explanatory in the causal sense. It identifies, with a precision the tardigrade has been used throughout to sharpen, exactly what requires explanation.
The mystery is not merely how sufficiently complicated molecules came to exist, though that is itself an enormous question. It is not merely how a capacity for replication began, though that too remains substantially unresolved. It is not merely how metabolism became possible, though metabolism is indispensable to nearly every form of life known. The deeper question, the one this architecture has been assembling the vocabulary to state precisely, is this: how did nonrecursive matter become organised such that differences in its surroundings could acquire relevance to the continuation of that very organisation itself?
Once this transition has occurred, whatever its causal history, something radically new exists in the world that did not exist before it. Previous encounters can now modify subsequent engagements. Disturbances can matter to the organisation they disturb, rather than merely altering it the way any causal impact alters any object. Remediation becomes possible, and with it the entire architecture of dormancy, suspension, and recovery this article has traced through the tardigrade in such detail. The organisation can coordinate itself differently because of what has happened to it before, rather than simply persisting or being destroyed by what happens to it. Relations with other living beings can become interrecursive, each modifying the other in ways neither could achieve with a nonresponsive object.
Nonliving materials can become recursively relevant grounds, coordinated with rather than merely collided with. And across the whole subsequent history of evolution, the five mediations differentiate and elaborate into the extraordinary range of forms life takes, from the simplest self-maintaining cell to the human capacity for language, law, and theory.
The tardigrade allows this mystery to be watched disappearing and returning with a clarity no other organism offers so cleanly. The active tardigrade is recursively coordinated within a viable mesocosm, sensing, moving, feeding, avoiding, growing. As desiccation proceeds, it actively transforms itself, while still recursively responsive to its deteriorating surroundings, into the tun. Ordinary recursive activity becomes radically attenuated, in some respects apparently absent by any measure available to an observer. The resulting organisation can persist, protected by mechanisms it produced in advance of its own suspension, in conditions under which active tardigrade life would be immediately impossible, including, as the FOTON-M3 experiment demonstrated, the vacuum of low Earth orbit itself (Jönsson et al., 2008). When appropriate mesocosmic conditions are restored, whether by a researcher's pipette or by rainfall on a moss patch, recursive organisation resumes, and the tardigrade that emerges is, in every sense that matters, the same recursively coordinating being that entered the tun.
The dead tardigrade supplies the decisive contrast that makes the whole comparison work. Restore precisely the same conditions to a dead specimen, exactly the water, the temperature, the time, and nothing comparable occurs. The material configuration that would need to resume its organised activity is no longer capable of doing so, irreversibly, regardless of how faithfully the external conditions are reproduced.
This yields the article's final definition, sharpened by everything the tardigrade has been made to demonstrate along the way. Death is the irreversible loss, for a particular living organisation, of the capacity to resume recursive mediation through a viable mesocosmic coordination.
This definition does real explanatory work across every case the article has examined. It explains why inactivity, however extreme, is not death, and why sleep, anaesthesia, and hibernation pose no threat to the distinction between the living and the dead. It explains why the life and death binary was never quite the distinction between present recursivity and present nonrecursivity, a distinction the cryptobiotic tardigrade satisfies on the wrong side while remaining obviously alive, but the distinction between an organisation that retains a mesocosmically real possibility of recursive return and an organisation for which that possibility has irreversibly disappeared. It explains why cryptobiosis specifically is not death, despite pushing observable activity to a limit at which ordinary criteria fall silent. It explains why metabolism, indispensable as it is to the great majority of ordinary cases, cannot by itself define death, since the tardigrade's metabolism during dormancy falls to levels the available literature describes as virtually undetectable while its capacity for revival remains fully intact. It explains why the persistence of individually living cells within a dead organism does not mean the organism itself remains alive, because what has been lost is the integration of those cells into one coordinating whole, not every constituent process at every scale. It explains why a corpse can remain, and often does remain, enormously consequential within the mesocosms of the living without itself being a recursively living participant in those mesocosms. And it explains, in the comparison this article has treated as its hardest test, why restarting a powered-down computer is not equivalent to reviving a tardigrade, however similar the two transitions might look from the outside to an observer attending only to behaviour before and after.
The conclusion returns, finally, to the proposition with which the article began.
Most binaries encountered across the mesocosm are classificatory achievements:
selective, historically produced, and, under sufficient pressure, reopenable, in the way other apparent oppositions built on continua of difference generally are. Life and death are exceptional. Beneath the very real and very extensive empirical ambiguity that surrounds particular cases, a categorical difference between recursive and nonrecursive organisation persists, and difficulty in locating the boundary in a given instance is not evidence against the boundary's reality.
Attending carefully to the classificatory labour that goes into other binaries does not dissolve this one. It makes possible, instead, a considerably more rigorous investigation of exactly which differences are relevant to recognising it correctly in the hardest cases, of which the tardigrade may be the most severe available to biology.
The tardigrade earns this role because it strips away almost every obvious sign by which humans ordinarily recognise life in another organism. Movement disappears. Feeding disappears. Ordinary sensory responsiveness disappears.
Metabolism falls to a level the literature can describe only as nearly or effectively undetectable. The animal contracts into a tun capable of enduring conditions that would destroy nearly any other animal of comparable size, including, for a limited experimental period, the vacuum and radiation of space itself (Jönsson et al., 2008). And yet appropriate remediation, water, warmth, time, is sufficient to bring the animal back into full, active, recursively mediated coordination with a living mesocosm.
The final paragraphs of this argument therefore reverse the question with which the article opened. The question was what death is. The answer that has emerged, across every case this argument has examined, is that death turns out to be conceptually easier than life, once the two are properly distinguished. Nonrecursive matter is everywhere, and its existence requires no special ontological principle to explain. What remains genuinely astonishing, and what the tardigrade has been used throughout this article to bring into the sharpest possible relief, is that particular organisations of that same ordinary matter become capable, however this capacity first arose, of recursively mediating their own continued existence at all.
The deepest mystery this article leaves standing is consequently not why living beings eventually die. It is why there was ever anything in the universe capable of dying in the first place. Living Value Theory cannot solve that mystery. What it can offer, and what this article has tried to demonstrate through the single, unusually severe case of the tardigrade, is a more precise language for identifying exactly where that mystery lies.
Death is where recursive mediation can no longer return. Life is the mystery that it ever could.
References Bloch, Maurice, and Jonathan Parry, eds. 1982. Death and the Regeneration of Life. Cambridge: Cambridge University Press.
Boothby, Thomas C., Hugo Tapia, Alexandra H. Brozena, Samantha Piszkiewicz, Austin E. Smith, Ilaria Giovannini, Lorena Rebecchi, Gary J. Pielak, Doug Koshland, and Bob Goldstein. 2017. “Tardigrades Use Intrinsically Disordered Proteins to Survive Desiccation.” Molecular Cell 65 (6): 975-984.
Brown, Emery N., Ralph Lydic, and Nicholas D. Schiff. 2010. “General Anesthesia, Sleep, and Coma.” New England Journal of Medicine 363 (27): 2638-2650.
Carey, Hannah V., Matthew T. Andrews, and Sandra L. Martin. 2003. “Mammalian Hibernation: Cellular and Molecular Responses to Depressed Metabolism and Low Temperature.” Physiological Reviews 83 (4): 1153-1181.
Carter, David O., David Yellowlees, and Mark Tibbett. 2007. “Cadaver Decomposition in Terrestrial Ecosystems.” Naturwissenschaften 94 (1): 12-24.
Dreyfus, Hubert L. 1992. What Computers Still Can’t Do: A Critique of Artificial Reason. Cambridge, MA: MIT Press.
Finch-Savage, William E., and Gerhard Leubner-Metzger. 2006. “Seed Dormancy and the Control of Germination.” New Phytologist 171 (3): 501-523.
Guidetti, Roberto, and K. Ingemar Jönsson. 2002. “Long-Term Anhydrobiotic Survival in Semi-Terrestrial Micrometazoans.” Journal of Zoology 257 (2): 181-187.
Jönsson, K. Ingemar, Elke Rabbow, Ralph O. Schill, Mats Harms-Ringdahl, and Petra Rettberg. 2008. “Tardigrades Survive Exposure to Space in Low Earth Orbit.” Current Biology 18 (17): R729-R731.
Manara, Alex R., Paul G. Murphy, and Gail O’Callaghan. 2012. “Donation after Circulatory Death.” British Journal of Anaesthesia 108 (Suppl 1): i108-i121.
Wełnicz, Weronika, Markus A. Grohme, Łukasz Kaczmarek, Ralph O. Schill, and Marcus Frohme. 2011. “Anhydrobiosis in Tardigrades: The Last Decade.” Journal of Insect Physiology 57 (5): 577-583.
Wijdicks, Eelco F. M. 2001. “The Diagnosis of Brain Death.” New England Journal of Medicine 344 (16): 1215-1221.