In debates about criteria for human death, several camps have emerged, the main two focusing on either loss of the “organism as a whole” (the mainstream view) or loss of consciousness or “personhood.” Controversies also rage over the proper definition of “irreversible” in criteria for death. The situation is reminiscent of the proverbial blind men palpating an elephant; each describes the creature according to the part he can touch. Similarly, each camp grasps some aspect of the complex reality of death. The personhood camp, in contrast to the mainstream “organism” camp, recognizes that a human organism can still be a biological living whole even without brain function. The mainstream camp, in contrast to the personhood camp, recognizes that a person can be permanently, even irreversibly unconscious, and still be a living person so long as his/her body is alive. The author proposes that hylomorphic dualism incorporates both these key insights. But to complete the picture of the entire “death elephant,” a fundamental paradigm shift is needed to make sense of other seemingly conflicting insights. The author proposes a “semantic bisection” of the concept of death, analogous to the traditional distinction at the beginning of life between “conception” and “birth.” To avoid the semantic baggage associated with the term “death,” the two new death-related concepts are referred to as “passing away” (or “deceased”) and “deanimation,” corresponding, respectively, to sociolegal ceasing-to-be (mirror image of birth) and ontological/theological ceasing-to-be of the bodily organism (mirror image of conception). Regarding criteria, the distinguishing feature is whether the cessation of function is permanent (passing away) or irreversible (deanimation). If the “dead donor rule” were renamed the “deceased donor rule” (both acronyms felicitously being “DDR”), the ethics of organ transplantation from non–heart-beating donors could, in principle, be validly governed by the DDR, even though the donors are not yet ontologically “deanimated.” Thus, the paradigm shift satisfies both those who insist on maintaining the DDR and those who claim that it has all along been receiving only lip service and should be explicitly loosened to include those who are “as good as dead.” Even so, a number of practical caveats remain to be worked out for non–heart-beating protocols.
The debate surrounding the concept and criterion of human death is in some ways like the proverbial blind men and the elephant. Each position is based on some important insight, and irresolvable disputes arise when each claims to be the whole story.
In the first half of this paper, I shall examine two contrasting positions and try to construct the part of the elephant between them. The first is what could be called (up to recently) the “mainstream” position, advocated by the President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research (1981), the Swedish Committee on Defining Death (1984), the Pontifical Academy of Sciences (White, Angstwurm, de Paula, 1992), and many others. It was eloquently defended by Bernat, Culver, and Gert (BCG; 1981), is still maintained by Bernat (2008), and expounded in slightly modified form by Gert, Culver, and Clouser (GCC; 2006). (These four authors together will herein be referred to as “BCGC.”) In honor of the President’s Commission, I shall refer to this as the “PC” position (pardon the pun), personified in BCGC. The second position is generally known as “higher brain death.” It has also been advocated by many (Youngner and Bartlett, 1983; Cranford and Smith, 1987; Zaner, 1988; Veatch, 1993; Machado, 2000) and championed recently in particular by philosopher Lizza (2006), whom I shall treat here as its spokesperson. BCGC recognize the importance of bodily integrative unity in human death but do not see that it is essentially an emergent phenomenon deriving from the mutual interaction among all the parts of the body (not a top-down micromanaging by some master integrator organ, the brain). By contrast, Lizza sees that human (and other vertebrate) organisms without brain function can still be “organisms as a whole”; but he does not see that this has any relevance for human death. In explaining and defending their respective positions, both make heuristic use of thought experiments involving organism bisection. I shall evaluate the utility of such thought experiments for understanding the ontological status of “total brain failure” (TBF)1 and shall conclude that, entertaining as they may be, they fail to shed conceptual light on TBF and in fact require actual TBF cases to shed empirical light on them. I then suggest a bridge between the body-centrist and mind-centrist views of human death based on the hybrid biological-psychological nature of human beings.
In the second half of the paper, I propose a view of the entire death elephant, achieved by stepping back from the terms, concepts, and questions of the traditional debate and reexamining them from the perspective of a different paradigm. This entails abandoning the assumption that death is a single concept—an assumption reinforced by the fact that there is only one word for it in our vocabulary—and rethinking the issues in terms of two equally real and valid death-related phenomena. Some of the irresolvable controversies can be traced to definitional competition between permanence and irreversibility as the essential temporal feature of “death” (with “irreversibility” itself also interpreted in various competing ways). I propose that each is the distinguishing feature of one of the two death-related phenomena. Moreover, all the competing interpretations of irreversibility will have some valid role to play, whether in the definition of one of the death-related phenomena or in the ethics of actions during the time interval between them.
CONSCIOUS NONORGANISMS, UNCONSCIOUS PERSONS, AND BISECTED PERSON-ORGANISMS
Organism as a Whole
Bernat, Culver, and Gert (1981) defined death as the “permanent cessation of functioning of the organism as a whole”, a definition that Bernat (2008) continues to hold. Gert, Culver, and Clouser (2006, 290) subsequently felt obliged to expand the definition to “the permanent cessation of all observable natural functioning of the organism as a whole and the permanent absence of consciousness in the organism as a whole and in any part of that organism.”2 For the sake of argument, I am happy to accept a slightly modified version of that with “observable” omitted and “natural functioning” understood as not being invalidated by the fact that one or a few particular vital functions require artificial support. Advocates of the PC rationale assert that fulfillment of such a definition is guaranteed by the criterion of “total and permanent loss of functioning of the whole brain” (Gert, Culver, and Clouser, 2006, 297). Given that permanent loss of all brain function obviously entails permanent absence of consciousness, their justification for the neurological criterion critically hinges on whether (for postembryonic vertebrates) brain function is necessary for being an “organism as a whole.” They claim that it is, offering as the main (if not only) reason a sampling of bodily dysfunctions resulting from lack of brain regulation. In an earlier publication (Shewmon, 2001), I offered several counterarguments, two of which are, to my mind, conclusive, and I have yet to receive serious rejoinders to them.
First, rather than trying to conclude anything from comparing lists of somatic functions and dysfunctions, that entire line of argumentation can simply be bypassed and trumped by the following syllogism.
A functionally brain-disconnected patient on a ventilator in an intensive care unit (ICU) (e.g., from high spinal cord transection or extreme Guillain-Barré syndrome [Hassan and Mumford, 1991; Marti-Masso et al., 1993; Vargas et al., 2000]) is a severely disabled organism as a whole, not just a conscious head connected to an unintegrated collection of organs and tissues enclosed in a bag of skin.
The somatic effects of brain nonfunction are necessarily identical to those of brain disconnection.
Therefore, a patient without brain function is also a severely disabled organism as a whole (merely an unconscious one).
The major premise requires some clarification and justification. For the sake of brevity, suffice it to say that the somatic physiological equivalence between TBF and high spinal cord transection can be made exact for purposes of the argument.3 The somatic equivalence between brain disconnection and brain nonfunction is not merely theoretical; it can be confirmed by a clinical comparison of the two conditions (Shewmon, 1999, 2004a).
A second fatal flaw in the PC thesis is that there exists no measuring scale or methodology for “degree of integration” of a complex organism such as a vertebrate. And even if (hypothetically) degree of integration could be meaningfully measured, there would be no point along that continuum that could reasonably nonarbitrarily constitute the dividing line between extremely sick, dying organisms, and just-dead (non-)organisms. BCGC simply list a few integrative functions mediated by the brain, followed by “etc.,” omit mention of any integrative functions not mediated by the brain, and perfunctorily declare that the set of lost functions lands squarely on the “disunified” side of the dividing line.
But even without a system for measuring degree of integration, it can easily be shown with the following syllogism that many bodies with TBF lie on the “integrated whole” side of the dividing line:
Dying patients in ICUs, with multisystem dysfunction and a rapid downhill spiral, by virtue of being still alive, are (by definition) necessarily on the “whole” side of the hypothesized dividing line along the continuum of degrees of integration.
Many patients with TBF in ICUs are as stable as, and some are more stable than, such dying patients.
Therefore, such TBF patients are also on the “whole” side of the dividing line.
Other arguments can also be made for the organism-as-a-whole status of many bodies with TBF, but to my mind, these two are the strongest. The TBF patient has the same ontological status as a permanent vegetative state patient who, for whatever reason, has lost the capacity to breathe. I am hardly alone in judging the integrative unity rationale for a neurological criterion of death to be inadequate. Most promoters of higher brain death also reject it, as do the growing number of critics of all brain-based criteria for death (Rix, 1990; McCullagh, 1993; Evans, 1994; Jones, 1995; Watanabe, 1997; Cranford, 1998; Potts et al., 2000; Taylor, 1997; Reuter, 2001; Lock, 2002; Byrne and Weaver, 2004; Zamperetti et al., 2004; de Mattei, 2006; Joffe, 2007b; Truog, 2007; Karakatsanis, 2008; Verheijde et al., 2009). Even the President's Council on Bioethics (2008), in its recently released white paper, has just lent its voice to the swelling chorus of rejecters of this hitherto mainstream rationale. (I shall not elaborate here on the Council's proposed novel alternative based on self-preserving “commerce with the environment,” a critique of which has been published elsewhere [Shewmon, 2009b].)
Gert, Culver, and Clouser (2006, 307) state that four criteria of death have been put forward: “permanent loss of natural functioning of (1) the cortex, (2) the brain stem, (3) the whole brain, and (4) heart and lungs” and that “Only the third criterion … is perfectly correlated both with loss of all observable natural functioning of the organism as a whole, and with total absence of consciousness.” In my opinion, they are right in rejecting #1, #2, and #4, wrong in accepting #3, and they overlook an important fifth criterion that has in fact also been put forward, namely “a critical degree of molecular-level damage … throughout the body, beyond a thermodynamical ‘point of no return,’” a reliable test for which is sufficiently “sustained cessation of circulation of oxygenated blood” (Shewmon, 1998). Ironically, this test was articulated long before me as a criterion by the President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research (1981, 73) in the first arm of its Uniform Determination of Death Act (which Gert, Culver, and Clouser [2006, 304] even quote but strangely do not include in their list of proposed criteria).4 A “circulatory-respiratory” criterion is not at all synonymous with a “cardiopulmonary” one, which GCC rightly reject. For reasons developed elsewhere (Shewmon, 1998), if not by sheer process of elimination, I believe circulatory-respiratory to be the criterion that best corresponds (pre-paradigm shift) to the concept of “cessation of the organism as a whole.” What is truly necessary for the life of a higher organism is not the functioning of heart and lungs, but the circulation of oxygenated blood and exchange of gases at the cellular level throughout the organism.
Organism-Bisection Thought Experiments
It is interesting that both BCGC and Lizza (and many others including my past self) make use of decapitation as a heuristic for understanding whether TBF is death (Thurston, 1975; Tendler, 1978, 395; Green and Wikler, 1980; President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research, 1981, 36, 39; Youngner and Bartlett, 1983, 265; Shewmon, 1985; Wikler, 1988; Gert, 1995, 25–6; Pallis and Harley, 1996, 4; Capron, 1999, 125; Rosner, 1999; Machado, 2000, 206–8; Spittler, 2003, 110; Z. H. Rappaport and I. T. Rappaport, 2004; Gert, Culver, and Clouser, 2006, 292, 294; Lizza, 2006, 28, 107; Bernat, 2008, 253). GCC felt obliged to amend the original version of Bernat, Culver, and Gert (BCG; 1981) of the death concept by adding “and the permanent absence of consciousness in the organism as a whole and in any part of that organism” (Gert, Culver, and Clouser, 2006, 290) in order to accommodate both hypothetical and actual cases of decapitation and artificial maintenance of the living head (e.g., the gruesome experiments carried out on decapitated criminals in the late 1800s and on dogs in the early 1900s, reviewed by neurosurgeon White, 1968, and on monkeys in the 1960s by himself [White, Albin, Verdura, 1964; White et al., 1965, 1971; White, 1968]). GCC regard the isolated head as clearly alive, by virtue of being conscious, but not a living organism—only part of a former organism. The artificially maintained decapitated body, in their view, is of course also not an organism; for them (and many others), the functional similarity to a body with TBF constitutes strong additional proof that the TBF body cannot be a living organism either.
Whether the conscious head is a severely mutilated organism or a nonorganism (which used to be part of a former organism) depends of course on the definition of “organism.” GCC offer no such definition but simply assert that the maintained head is not an organism. Lizza seems implicitly to agree on this point when he argues, contra GCC, that the maintained headless body has more claim to organism status than the head (Lizza, 2009, 541). As far as I am aware, there has not been much development of a “philosophy of organism” that would provide a reasonable nonarbitrary dividing line between mutilated organism and nonorganism along the continuum of imaginable mutilations.
But what is relevant to the debate over TBF is not whether the conscious head is an organism, but whether the maintained headless body is. To answer this question, one would need to examine the actual biological properties of an artificially maintained headless body and see if there are any emergent, holistic properties that must be attributed to a whole that is greater than the sum of its parts. One opportunity to examine such a body was with White’s monkey experiments, but he was interested in the heads and did not maintain the headless bodies chronically or study their properties. There is one report of a pregnant sheep that was decapitated and artificially maintained for 30 min until Cesarean section delivery of a healthy lamb (Steinberg and Hersch, 1995). Evidently, the experimenters had no interest in maintaining the mother sheep any longer than that to search for possible holistic properties in a chronic headless state. May no such experiments ever be carried out on humans even by the most terrible future Nazi-style doctors (or ever again on our animal cousins)! The only way to (ethically) know what the physiological properties of a technologically maintained headless human body might actually be is by inference from analogous “experiments of nature,” that is, actual bodies with TBF. Many of these, as demonstrated above, are rightly classified as living organisms as a whole, albeit severely disabled organisms. So, far from the hypothetical headless body proving that a body with TBF is a nonorganism (on the same ontological level as an amputated limb), actual bodies with TBF prove that the hypothetically maintained headless body would be a living organism!
Now, in the thought experiment, we have on the one hand a conscious head (whether one wants to call it an organism or not), and on the other hand, a distinct unconscious (perhaps better stated nonconscious) human organism. Actually, the head’s consciousness should not simply be taken for granted. In the aforementioned dog and monkey experiments, the heads reacted in grossly canine or simian fashion to stimuli (mainly noxious ones), but that is hardly proof that a maintained human head would be subjectively conscious and would communicate its thoughts to us. But, assuming for the sake of argument that a maintained human head would exhibit conscious behaviors, there are several conceivable ways of interpreting the thought experiment philosophically. Let us call the pre-decapitation person “John.”
Regarding the maintained head:
The head is John’s; when it communicates with us, John is communicating with us.
The head is not John’s anymore. The behavior and communications of the head only resemble John’s former personality because of the retained neural circuitry, but they are coming from an automaton or “zombie” rather than from John.
Regarding the maintained headless body:
3. The headless body is, or is part of, John’s body. If #1, then both head and headless body are part of John’s ongoing body, and we would have to posit his person maintaining some sort of transspatial relationship with both. If #2, the headless body is John’s body, equivalent to John with TBF.
4. The headless body is a nonpersonal human organism. (This assumes a concept of “person” that accommodates actual dissociability between “human person” and “human organism.”)
5. The headless body is a new human person-organism, a kind of identical twin of John that came into existence not from embryo splitting but from a very special kind of adult splitting (that does not occur spontaneously in nature). The new person (call him Fred) is unable to exercise cognitive functions due to lack of a brain—something like an anencephalic infant suddenly coming into existence except with an adult body. (This assumes a concept of person that is not reducible to mind or to brain activity, one for which “permanently unconscious person” is not an oxymoron.)
Even within the imaginary context of the thought experiment, no “thought-empirical” observation or test will help distinguish #1 from #2. Nor is there a way to thought-empirically distinguish among #3, #4, and #5. The difference between #4 and #5 depends entirely on the philosophy of personhood that one brings to the thought experiment; it is not decided by means of the thought experiment. Lizza would subscribe to the philosophy of personhood in #4, whereas I and presumably BCGC (prescinding from our differences regarding the organism aspect) would subscribe to that in #5. I am not sure whether anyone subscribes to #3, but it is at least a logical possibility.
Entertaining as such mind games may be, they shed no light on the relationship between brain function and human death. Suppose the bisection were done in the midline sagittal plane rather than across the neck. The midline raw surfaces could be covered over through plastic surgery, and the half with a missing heart could be sustained by a transplanted or artificial heart (with either the other unpaired vital organs replaced through transplantation or the thought experiment simply terminated by death from liver failure or whatever).5 This thought experiment can be understood as simply taking split-brain experiments to the extreme. From one person-organism (John), the bisection generates two (John and his identical twin Fred), one of whom can communicate with us verbally (at least through eye movements) and the other nonverbally. If the bisection were done in childhood, both could be verbal. Now, if we surgically reunite the two halves, including microanastomosis of every severed axon in the nervous system, resulting in one person-organism again (physically identical to before the bisection), is it John or Fred? By repairing the mutilation, have we killed someone? At least one hard-core split-brain enthusiast has gone so far as to maintain that in each of us there are actually two minds (one associated with each cerebral hemisphere), who think and act in parallel and are completely unaware of each other’s existence, with observational uncertainties vaguely analogous to quantum mechanics (Bogen, 1986).
Similar thought experiments based on cases of dicephaly, Siamese twins, brain transplants, and so on can be devised ad infinitum. What do they really prove about persons, organisms, and TBF? Nothing whatsoever. Why? Because they all are about questions of personal enumeration and personal identity in relation to two or more distinct pieces of living matter. But in actual clinical TBF, there is only one piece of living matter and only one person at issue. The question is not whether the body with TBF is John’s or Fred’s but whether it is John’s or nobody’s. The question is really two-fold: (1) whether it is a living human organism at all and (2) whether human personhood is actually separable from human organismhood (not just logically distinct). The first question, I maintain, has already been convincingly answered in the affirmative. The second question goes to the very core of philosophical anthropology and has been debated intensely by philosophers for a very long time without any signs of rapprochement. Neither question is answered by thought experiments; rather, conversely, the interpretation of the thought experiments depends on the answers to these questions. (For a more extensive analysis of the decapitation analogy, see Shewmon [2007, 2009a].)
Personhood and Body-Brain Dualism
Regarding the concept of personhood, BCGC and I6 join many others in the camp that regards a human person as essentially a biological-psychological hybrid being (Lee and George, 2008). If either the biological or the psychological dimension is lost, but not both (e.g., conscious head or permanent vegetative state), we would say that there is a severely disabled person. We understand permanent unconsciousness in a living human organism as the extreme along the spectrum of mental disabilities. By contrast, Lizza (2006) joins many others in the camp that regards human person as essentially a human mind, in principle and sometimes in actuality separable from human organism or “human being.” Even if some in this camp accept that a human person is a biological-psychological hybrid and not just a mind, they would argue that if only one of these two dimensions were missing, there would be no hybrid and therefore no human person. But they apply this asymmetrically, accepting that the conscious head (perhaps even just a conscious brain in a vat) is a disabled human person by virtue of consciousness despite lacking the biological component (not being an organism) but denying that the vegetative state patient is a disabled human person by virtue of being an organism despite lacking the psychological component. This asymmetry shows that the purported acceptance of the notion of biological-psychological hybrid is superficial, ultimately contradicted by a more fundamental stance that a human person is essentially a mind.
The two positions seem irreconcilable at the most fundamental level of philosophical worldview. Returning to the parable of the elephant, such disagreement over the nature of personhood does not stem from the blind men feeling different parts of the elephant but from feeling the same part with different hand coverings: those with bare hands claim that the skin is firm and leathery, whereas those wearing wool mittens claim that the same area of skin is soft and wooly and those wearing coats of mail claim that it is hard and metallic. Being blind, no one can see what he himself or the others are wearing, and everyone has been dressed that way for so long that they know no other kind of hand-sensation. Each personhood camp will claim that it is the bare-handed one in the parable and that all the others are impeded from a true understanding by a wrong fundamental worldview cloaking their minds.
A particularly dense “hand covering,” which underlies much of the person-mind reductionism in the debates over TBF, is body-brain dualism, an offspring from the marriage of body-self-dualism and mind-brain reductionism. Let me briefly discuss the two parent concepts first and then their child. Body-self-dualism is manifested when the referent of “I” (“me,” “you,” “he,” “she,” etc.) is understood (whether explicitly or implicitly) as either the body or the self, but not an integral union of the two. Someone who focuses on one of these aspects to the exclusion of the other may not be a dualist in the sense of asserting two fundamental principles of existence; nevertheless, the two complementary monisms stem from the same Cartesian dualistic root, and the prevalence of both in contemporary society reflects a common tacit body-self-dualism. Those who consider “I” to be the animal organism focus on the body’s sensorimotor functions and behavioral patterns (the classical example being behaviorists, but many neuroscientists who regard the mind as a mere epiphenomenon of brain activity also fall into this category). Those who consider “I” to be the self focus on subjective experiences, qualia, thoughts, desires, pleasures, and pains.
Body-self-dualism has been thoroughly and critically analyzed by Lee and George (2008) in a recent book under that title. The alternative they propose, which I think is correct, is to acknowledge that the referent of “I,” the human person, is a unity at once both animal and mental. Moreover, the mental is not simply reducible to the animal level, as mind-brain reductionism maintains. As Lee and George (2008, 50–65) explain, reflective self-awareness, universal concept formation, abstract reasoning, and free will all have properties that transcend spatiality and cannot in principle “emerge” from a complex electrochemical network. They therefore derive from an immaterial principle, but nevertheless, one profoundly oriented to operate in and through a body. If substance dualism is (rightly) rejected as incapable of explaining the intimate, codependent relationship between body and mind, and if both materialism and idealism are rejected as simplistic and incapable of explaining those aspects of reality that the complementary “ism” rightly recognizes, then the only position left is hylomorphic dualism. And it is worthy of serious consideration not simply by process of elimination but by virtue of its explanatory power. In this view, the soul is not a spiritual thing like a ghost or an angel that inhabits or is somehow extrinsically related to an essentially mechanical body (substance dualism) but is both the immaterial principle of the intellectual and volitional powers (which operate through the brain but are not reducible to brain activity) and the vital principle, or substantial form, of the body, making it to be precisely a living body.
Hylomorphism is admittedly a hard sell in this day and age, when reductionistic materialism is so prevalent among intellectuals, particularly scientists. But I would point out that the a priori conviction that reflective self-awareness, conceptual understanding, and free will are ultimately nothing more than physical “emergent phenomena” from electrochemical brain activity has more the character of a religious faith than of philosophical reasoning (to say nothing of empirical evidence). The phrase “emergent property” has become the standard explanatory cop-out for every immaterial aspect of mental phenomena. In religious debates, the defender of an unprovable position, when backed up against the evidentiary wall, ends up saying, “Well, it’s ultimately a matter of faith.” So, too, in debates about mind-brain relationship, the material reductionist always brings the discussion to an abrupt end with the non sequitur: “Well, it’s an emergent phenomenon. Perhaps someday science will discover the mechanism for it, but even if not, that’s only because of its exceeding complexity.” Mind-brain reductionists, to be consistent, must assert (as many do) that they themselves are a mere illusion of “self” created by complex electrochemical activity, that their own reasoning processes (including their belief in mind-brain reductionism) are nothing but programmed computations, and that their willed acts (such as trying to convince everyone else of mind-brain reductionism) are either macrodetermined or microundetermined (anything but self-determined, i.e., free), thereby rendering morality baseless and good and evil meaningless. They would rather hold these implausible and self-undermining positions than entertain the possibility that such mental phenomena are essentially immaterial, thereby requiring an immaterial principle in their subject. A whole article could be written on the ersatz-religious character of contemporary mind-brain reductionism (complete with its own dogmas, repression of heresies, high priests, and liturgies). I merely allude to it here for the sake of pointing out, by contrast, that hylomorphic dualism is not really so unreasonable a philosophical theory by comparison.
Someday I would like to write a book paralleling Lee and George’s, entitled “Body-Brain Dualism,” which would explain how neo-Cartesianism, wedded to mind-brain reductionism, has profoundly influenced the neurosciences. Symptoms of it crop up everywhere. Hardly a grant application or manuscript submitted for publication in the biological sciences gets accepted unless it elucidates “basic mechanisms.” On March 29, 2009, a search of PubMed (a literature database for the medical sciences) on the word “mechanism” in either title or abstract generated 568,155 “hits” (of which mechanism was in the title of 97,715). For the single year 2008, the numbers were 39,406 and 4,129, respectively. Evidently, Descartes’ view of animals as essentially machines is still alive and well.
Brains are equally considered machines. A PubMed search for both “mechanism” and “brain” in the title or abstract identified 27,935 articles, including 2,202 for the single year 2008. Moreover, the brain is often tacitly understood as an entity unto itself, not a part of the body as a whole. Neurologists often jest that the only purpose of the body is to keep the brain alive and carry it around. But such quips are also more than half serious. The history of neurology reinforces the view that the brain is much more related to the mind than to the body. Neurology and psychiatry began as one field, went their separate ways around the turn of the 20th century, and have been gradually reconverging. In contradistinction to all other organ-system subspecialties such as cardiology, gastroenterology, nephrology, and so on, neurology departments are rarely structured in medical school and hospital hierarchies as a subdivision of internal medicine departments, but as departments of their own, often more closely related to psychiatry than to medicine.
Naturally, brain-body dualism strongly impacts the debate over TBF. A frequent way of describing chronic TBF cases is “brain death with prolonged somatic survival,” as though the brain and the body (soma) were distinct entities. The traditional emphasis on the brain as (putative) integrator of the body ignores the anatomical and functional continuity between the brain and the spinal cord as well as the integrative functions of the cord, creating an artificial division of the central nervous system at the cervicomedullary junction—as though the spinal cord were part of the “body,” whereas the brain is not. In contrast to this view of the brain as extrinsic master integrator of the body (conceived as all organs and tissues except the brain), Lee and George aptly state:
The various cells, tissues, organs, and so on, must be organized or unified so as to make up one being. The ultimate principle of unity cannot be a material organ, since this would only give rise to an aggregate of this organ with other bodies (and so the unity of this organ with others would remain unexplained), but it must be a form or order determining the components to be one substantial entity [I would add: “including the brain”—DAS]. In a living being, this form or principle of unity can be referred to as a soul, since the soul is (philosophically) defined as the first principle of life in an organized body. (Lee and George, 2008, 66 [emphases in original])
All this boils down to whether, in the final consequence, persons use their brains as an instrument for mental functioning or, rather, brains generate their persons (as epitomized in the very title of a recent book, The Brain and Its Self [Knoll, 2005]). Stated differently, it boils down to whether I think (see, feel, desire, intend, etc.) by means of my brain or my brain thinks (sees, feels, desires, intends, etc.). If the former, then the “I” is both the subject of immaterial/material hybrid mental acts and the possessor of “my brain”; and since the brain is a part of the whole animal organism, then “I” must be both immaterial and an animal organism. And if we reject substance dualism, the immaterial aspect and the animal organism must constitute a single entity, the former being the formal principle of the latter (i.e., the hylomorphic view).
This is precisely why, if my animal organism dies (ceases to be an organism as a whole), I die. But if my mental functions are impeded, even if permanently, by a defect in the material organ required for their actualization, I do not die because I am something more than just those functions (contra functionalism) and also something more than just “my brain” (contra mind-brain reductionism, and as evidenced by the very phrase my brain, which even mind-brain reductionists obliviously and inconsistently use). As long as the animal organism that is me is alive, I am alive, even if I may be unconscious. Again in the words of Lee and George,
… psychological continuity cannot be the criterion for determining when we come to be and cease to be. Rather, we are human animals and we came to be when the animal organisms that we are came to be, and we will cease to be only when these animal organisms cease to be. (Lee and George, 2008, 38)
The following objection might be raised, a seeming flip side of the asymmetry objection that I raised earlier against higher brain death advocates who acknowledge the biological-psychological hybrid nature of human persons. One way of expressing the PC view of human death is the breakdown of the psychophysical unity that constitutes the human person.7 It seems (according to the objection) asymmetrical and therefore illogical for me to admit, on the one hand, that if the psychophysical unity is rent by destruction of the physical component, whereas the “psycho” component lives on as the soul (by virtue of its immateriality), then the person dies; but, on the other hand, if the unity is rent by elimination of the psycho component only, whereas the human organism lives on, the person does not die. The answer to the objection is as follows. First, I am not so sure that the stated formulation of human death is a good one to begin with. But even accepting it for the sake of argument, the asymmetry objection involves a semantic problem. The psycho in “psychophysical unity” refers not so much to psychological functioning per se as to the ultimate source of psychological functioning, that is, the immaterial soul. Now, if body and soul were understood in the manner of substance dualism in which soul equals conscious mind, then the objection would stand. But with a hylomorphic understanding of these terms, there is no mind-body symmetry to begin with, and consequently, no illogicality in death being asymmetrical in this regard. In the hylomorphic view, “mind” (or “psyche”) and “soul” are not synonymous, but the soul is both the principle of the immaterial aspects of mind and the substantial form of the body, making it precisely an organism as a whole. So long as the organism is present, the soul is necessarily present (with its immaterial powers inseparable from its function as substantial form); therefore, the ultimate basis of intellectual and volitional functions is still present even if a brain lesion impedes their actualization. Moreover, even if the immaterial soul survives death of the human person/organism, according to the hylomorphic view, it would be dubiously conscious, and, if conscious, then not in any properly human manner. If there is “life after death,” such “life” will necessarily be very different from how the term applies on earth, and it surely will not negate the reality of the death of the human person/organism on earth.
Capacity, Radical Capacity, and Substantial Change
Before leaving the topic of personhood, I want to address a challenge to my rejection of a brain-based criterion for death from within the framework of hylomorphism: namely, that with total or nearly total brain destruction (or irreversible nonfunction), the organism lacks essential properties that previously made it the kind of organism it was; a substantial change has taken place to some lower form of organism (although not a naturally occurring kind). Such a substantial change would be analogous to that which would take place on destruction of an entire human body except for one cell that remains alive in a culture dish—except instead of a unicellular organism, what remains is a multicellular one (which also happens to require some technological support to stay alive). I myself proposed this kind of argument in my earlier defense of higher brain death, referring to the post-substantial-change organism as “humanoid” (Shewmon, 1985), a term also employed by Lizza (2006).
Before presenting the argument in syllogistic form, some terminology needs to be established. In saying that a human being is a rational animal or that a mammal is a sentient organism, by “rational” or “sentient,” we do not mean that the human or mammal is always thinking or sensing, nor do we even mean that it necessarily has an immediately exercisable capacity for thinking or sensing (as during sleep). A human or a dog can be comatose for months and then recover consciousness; during the coma, it was still a living human or dog and therefore still a rational or sentient organism in some proper sense of those terms. This sense has been referred to as the “radical capacity” (or “basic natural capacity” or “second-order capacity” or “active potency”) for some property (Lee, 2010). The “radical capacity for X” can be understood as the “capacity to acquire the capacity for X.” Thus, human embryos lack the capacity to see or to think, but they have the radical capacity for these operations. The syllogism also accepts, for the sake of argument, that brain function is not necessary for a higher organism to be a living whole.
The syllogism is expressed here in terms of mammals and sentience (adapted from May [2008, 352–3], in turn based on an unpublished argument from Germain Grisez), but it could just as well be rephrased in terms of humans and rationality. It focuses on mammals rather than animals for the sake of logical strength, since mammals are more indisputably sentient than some lower forms of “animal” and since humans are a subset of mammals. It goes like this:
All mammals, by definition, have a radical capacity for sentience.
Destruction of the entire brain eliminates the radical capacity for sentience in postembryonic mammals.
Therefore, destruction of the entire brain in a postembryonic mammal produces a substantial change to a nonmammalian organism.
It may be that destruction of less than the entire brain would serve just as well, but the above formulation avoids distracting empirical controversies over precisely what parts of the brain are critical for this purpose. I am sympathetic to this line of reasoning, as I myself employed it in the 1980s with respect to humans and rationality (Shewmon, 1985). I now reject it for two main reasons, the first related to the major premise and the second related to the minor premise.
First, the definition in the major premise, while posing as an essential definition, is actually a stipulative one, therefore begging the question. In a nutshell, the syllogism as a whole reduces in essence to the single statement, “brain-destroyed organisms are not mammals by definition.” Let me explain. For centuries, the definition of animal as a sentient living thing seemed adequate for the broad classifications of observable nature. But when more was learned about embryology, classifiers of nature were faced with a dilemma: early animal embryos are not sentient, yet they clearly should be classified as animals because they are an integral part of the life cycle of animals. Rather than insist that developing embryos were not animals “by definition,” they improved the definition to accommodate a category of living thing that they knew through empirical evidence was the same natural kind as animals in later stages of development. The definition of animal was therefore tweaked (in concept if not in terminology), replacing sentient with “having a radical capacity for sentience.”
Just as developing embryos were not part of the observable nature that gave rise to the ancient definition of animals as sentient beings, severely brain-damaged animals on ventilators in veterinary ICUs were not part of the observable nature that occasioned the revision to “having a radical capacity for sentience.” The latter definition was for purposes of distinguishing animals (including their embryologic forms) from plants, not for distinguishing moribund animals in ICUs from dead animals.
Now, even if we accept, for the sake of argument, that brain-destroyed mammals lack a radical capacity for sentience (a risky assumption—vide infra), it is not logically valid to take a definition developed for a certain context of observable nature and slap it onto a totally different context for which it was never intended. Rather than argue that brain-destroyed mammals are no longer mammals simply “by definition (Q.E.D.),” we need to examine their properties and decide whether (given their acceptance as living organisms) (1) they are at one end of the continuum of mammalian life cycle, the opposite end of which is embryos, or (2) they have undergone a substantial change to the vegetative level. If their properties cry out for categorizing them as moribund, permanently comatose, and technologically dependent animals, then what is called for is to fine-tune the definition once again. Such a revised definition could be something like, “an animal is a living being with sentience or a radical capacity for sentience, or a sick or disabled or dying living being that had a radical capacity for sentience.”
By rejecting the possibility of fine-tuning the definition to accommodate an expanded context, which is now important to us but for which the previous version of definition was not intended, the syllogism begs the question at its very first step. If we are trying to decide whether brain-destroyed mammals are alive or dead (as mammals), the syllogism does not help because the entire syllogism is equivalent to a mere assertion of one of the possible answers, not a proof of it.
My second reason for rejecting the substantial change syllogism has to do with its minor premise: the assumption that the radical capacity for sentience resides in the brain. It surely seems that brain destruction entails loss of the “capacity to acquire the capacity” for sentience, as contrasted with (reversible) coma. But since the radical capacity for sentience exists in the brainless embryo, it is clear that the brain is not the sole and necessary locus of that radical capacity in an absolute sense. In the embryo, the radical capacity is grounded in the combined genetic and epigenetic factors distributed throughout the embryo. Could it be that even in the adult body, brain-based capacities are still not “radical” enough to serve as the basis for mammalhood (or for humanhood, if we substitute rationality for sentience in the syllogism)?
Adult stem cells can now be induced to develop into various functioning tissues and even parts of organs, a feat imaginable only as science fiction just 10 years ago (Cogle et al., 2004; Hoffmann et al., 2006; Warnke et al., 2006; Yacoub and Nerem, 2007; Di Girolamo et al., 2009). Moreover, the capacity to regrow an amputated limb (previously thought to be possessed only by salamanders) seems to be latently present even in higher animals, including humans. Actualization of this potential requires not even stem cells but merely release of its epigenetic suppression in the cells involved in normal wound healing. One research team speculates that by “tapping into and activating developmental programs necessary for regeneration” (present in the human genome but inactive postembryonically), “we may be only a decade or two away from a day when we can regenerate human body parts” (Muneoka et al., 2008, 63).
So, we should not be too hasty to assume that the radical capacity for regenerating body parts, present in human embryos, is absent in mature human bodies. Of course regenerating a brain is a far cry from regenerating a limb. But a radical capacity is something intrinsic to an organism; how long it takes medical technology to become sophisticated enough to remove the impediment to its actualization is ontologically irrelevant. Moreover, for purposes of minimal sentience, a regenerated brain would not have to be normal or very large. Even hydranencephalic children, with only a brainstem and diencephalon, can be sentient and have rudimentary consciousness (Shewmon et al., 1999).
Lizza, following Feinberg (1974), has objected that such a notion of capacity (or potentiality) is so “promiscuous” as to be virtually useless (Lizza, 2005, and J. P. Lizza, personal communications). I am not convinced, however, that this is so. What is proposed is not a notion of potentiality by which virtually anything can “potentially” become anything else. The question is whether a damaged organism has the radical capacity (potential) to repair itself. Although not discussed by Lee (2010) or May (2008, 352–3), the notion of “capacity to develop the capacity for X” should be understood in a not overly restrictive sense. “To develop” should be understood not as necessarily entirely spontaneous, but as also accommodating instances where external elimination of some impediment is required for the development to proceed. So long as the assistance is truly the removal of an impediment and not a frank replacement, such a “capacity to develop the capacity” should be ontologically relevant. One can distinguish potential for X without assistance, potential for X with assistance, and performance of X by an external agent. The first is too restrictive a concept, and the third is not a potential of the organism at all. The second sense is the appropriate one for the question at issue.
For example, we should not adopt an interpretation of radical capacity so strict that the facts that the healing of large wounds requires sutures, the healing of compound fractures requires plates and screws, and recovery from serious infections requires antibiotics would negate the body’s radical capacity for self-healing. Similarly, the radical potential for sight should be understood as present in someone with dense cataracts, even though ophthalmologic surgery is necessary to actualize that potential. The hidden regenerative potentials of mammals, including humans, that are actualizable through epigenetic desuppression are now beginning to come to light. If a certain kind of self-repair requires assistance to become actualized, that should not negate the existence of a radical capacity for self-repair, any more than requiring external assistance to survive negates being alive.
The syllogism fails because the minor premise cannot be simply assumed. Nay more: until proven otherwise, it is philosophically safer to assume the opposite—that the truly radical capacity for sentience lies not in the brain but in the genetic and epigenetic information throughout the living organism. To assume that the brain, rather than the whole organism, is the seat of that radical capacity smacks of yet another manifestation of latent body-brain dualism.
Suppose that someday a new brain could be grown in a (human) brain-destroyed body. Would that result in a new person or the same person merely with a new personality and set of memories? The answer depends on which theory of personal identity is correct and, again, whether persons use their brains as an instrument of mental functioning or, rather, brains generate their persons. In either case, one could at least say that there was all along some human person by virtue of the continuity of the human organism, which ipso facto possesses the radical, if not always actualizable, capacity for consciousness.
In summary, only the hylomorphic vision of human nature can coherently connect the two parts of the death elephant represented by the PC and higher brain death positions: on the one hand, recognizing the essential importance of both consciousness and the bodily organism for human personhood, whereas on the other hand, accommodating both permanently unconscious persons and human organisms as a whole without brain function. But there are still more parts of the elephant to account for … .
A SYNTHETIC PARADIGM SHIFT
A Catalyzing Experience
A small but significant shift in paradigm was catalyzed by the recent experience of the euthanasia of my family’s beloved dog Soran, who had cancer and had reached the point where continued existence was no longer appropriate. As a physician, I have witnessed deaths in hospitals, but this canine death made a particular impact because it was my first occasion to witness the death of a close family member, and it occurred right after participating in two conferences on the moment of death. The whole family was present. When we felt ready, the vet injected a general anesthetic, followed quickly by the lethal drugs. First, his breathing stopped, and very soon after, the heartbeat, which we could feel with our hands. From that moment on, we had a very clear sense that what lay before us was no longer Soran, but the remains of Soran. If death is an event (which I take it to be), it was obvious that the state-discontinuous moment separating the end of his dying process from the beginning of the decomposing process of his remains was the end of that last heartbeat. There was nothing dubious, nebulous, or mysterious about it. Any moment afterward would have been an arbitrary and vaguely identified point along a continuum of physical changes, not an instantaneous radical change. Our grieving process quite appropriately began at that moment.
As I reflected on the experience, I realized how preposterous it would have been to wait to begin grieving until some other, theoretically hypothesized, moment of death, such as the point beyond which autoresuscitation would have been impossible (had quick-acting antidotes for the drugs been administered), or the point beyond which anoxia-ischemia rendered his brain totally and irreversibly nonfunctional, or the point beyond which respiration and circulation could not be reestablished even with technological assistance, or any number of other putative moments of death that have been proposed over the last four decades. The more I thought about it, the more conceptually coherent, and the more true to the physically observable phenomena it seemed, to place the moment of Soran’s death at the cessation of his last heartbeat, even if that meant abandoning my previous conviction that a correct concept death had to include irreversibility and the idea that death is a single phenomenon.
If assimilation of good things, elimination of bad things, growth, and development are signs of biological life, so are they also signs of intellectual life. Rigid repetition of established patterns and lack of adaptability to the new and unforeseen are signs of senescence, both biologically and intellectually. Therefore, if we are to construct the whole death elephant, (1) we must expect that it will not correspond in every detail to any of the proposals to date (my own included), (2) we must be willing to think “outside the box” of the concepts and terminology that have characterized the (unresolvable) debates surrounding death for the past 40+ years, (3) we must incorporate all the valid insights made by everyone who has contributed to the debate, including those with whom we disagree on important points, and (4) we must be willing to abandon aspects of our own long standing and strongly held position that do not fit the emerging big picture.
Contributions that ought to be incorporated include (in addition to the points made earlier in this paper): (1) death is an event that demarcates the process of dying from the process of decaying (D. A. Shewmon and E. S. Shewmon, 2004); (2) our concepts in general, including death, are unconsciously influenced by the language we grow up in; just because English has the one word death, that does not necessarily mean there has to be a single, all-encompassing death concept (D. A. Shewmon and E.S. Shewmon, 2004; Chiong, 2005); (3) various types of “death behaviors” may appropriately correspond to different moments along the physically continuous process of dying and decomposing (Veatch, 1976, 26; Brody, 1999; Fost, 1999; Halevy, 2001; Youngner and Arnold, 2001); (4) statutory definitions of death should enshrine a philosophically and biologically valid criterion for a correct concept of death; they should not be mere utilitarian legal fictions; (5) Soran really died the moment his heart stopped; (6) death is irreversible (barring miracles); (7) perhaps death should be reconceptualized as a state, which can be entered and exited before becoming irreversible (DuBois, 2002). Some of these parts of the death elephant seem at first glance irreconcilable, for example, #1 with #2, #5 with #6, and #6 with #7. When there is apparent incompatibility between certain features of a complex reality, each of which seems valid in itself, that is a sure indicator that a fundamental paradigm shift is required for further progress (Kuhn, 1970).
One major problem with the traditional formulation of death is the logical disconnect between a dichotomous, all-or-none concept (an organism either is a whole or does not exist at all; it is either alive or dead8) and a temporally nebulous criterion (irreversible cessation of X). At best, such criteria and corresponding tests can establish that death has happened sometime in the past but cannot pinpoint the moment of death; at worst, they are mere legal fictions, setting the legal “time of death” as arbitrarily as setting speed limits.
For a long time I, along with many others, believed that death must properly be understood as a single phenomenon, that is, that there should be no distinction in reality (only logical distinctions) between what might variously be called legal, medical, philosophical, and religious deaths (Shewmon, 1992). In 2004, my view was broadened through enriching discussions with my linguist wife, resulting in two publications on the semiotics of death in which we explored how concepts and language mutually influence each other’s evolution within a particular culture, and the possibility that our own concept of death might be artificially constrained by peculiarities of the language in which we speak, write, and think about death (Shewmon, 2004b; D. A. Shewmon and E. S. Shewmon, 2004). These two papers have been misunderstood by some as proposing a completely relativistic notion of death, but they were intended merely as an invitation to step back from the confines of our lexicon and reexamine the actual phenomena to consider whether there might actually be more than one real death-related moment, each deserving of its own distinct death-word. The paradigm shift to be described below can be understood as a next step in the development of that thought.
Winston Chiong proposed something similar based on Wittgenstein’s observation that some words cannot be properly defined but have intrinsically vague semantic boundaries (Chiong, 2005). From the perspective of the following proposal, I would say that this observation, valid as it is with respect to some words and their corresponding concepts, need not apply to “death.” The semantic nebulosity surrounding “death” is a modern chapter in the history of language, resulting from attempts to gerrymander under the one term “death” disparate conditions that are considered desirable to conceptualize and legally define as “death” for the sake of organ transplantation. The paradigm shift that I shall now propose recognizes two fundamental death concepts, each of which has perfectly sharp semantic boundaries; and one yields criteria and tests that identify the exact moment of transition from life to (that type of) “death.”
The Proposal—A Semantic Bisection
I suggest that real progress in understanding death might be achieved not by hypothetically bisecting organisms in thought experiments but by semantically bisecting the concept death. The proposed paradigm shift involves not one but two fundamental death-related concepts, for which we need to assign names. The motivating force behind the distinction is the recognition, on the one hand, that Soran and all other higher animals and humans really do undergo death (in some appropriate sense of the term) at the onset of permanence of cessation of all vital signs and, on the other hand, that it is impossible (barring miraculous intervention) to return to life from death (in some appropriate sense of the term). Since the two considerations are not mutually implicating, the two appropriate senses of the term cannot be the same. It is necessary to distinguish two corresponding concepts of death: one indicating that an organism does not (in fact) return to life after death and the other indicating that an organism cannot return to life after death. (For convenience, I shall omit the phrase “barring miraculous intervention” every time irreversibility is referred to, but this is always to be understood.)
For naming these concepts, terms etymologically unrelated to “death” are preferred to reinforce that they are not two subtypes of one death phenomenon but two distinct phenomena. Rather than invent neologisms, I have chosen terminology that evokes the desired meanings: passing away and deanimation.9 For the remainder of this paper, I shall avoid the term “death” to the extent possible, on account of its semantic baggage, which would interfere with getting the new proposal across.
This pair of concepts forms a mirror-image analogy with the beginning of life, which also features two distinct concepts and terms etymologically unrelated to “life.” One is entry into society as a new member, which is called “birth”; the other is the coming to be of a new organism, which is called “conception.”10 That the distinction at both ends of life should be precisely two-fold is appropriate for several reasons. Regarding relatedness, there are two fundamental possibilities: an organism can be related either to itself (i.e., what it is in itself) or to others. Conception and deanimation mark the beginning and end of an organism as it is “in itself,” whereas birth and passing away mark the beginning and end of an organism in relation to the rest of the world. The former could also be thought of as an organism’s “metaphysical” beginning and end and the latter as its “civil” beginning and end.11 A two-fold distinction is additionally appropriate at the end of life because the difference between passing away and deanimation hinges on a fundamentally two-fold distinction in the temporal domain: permanence versus irreversibility. The first has to do with when someone exits life; the second has to do with when they can no longer be revived. There is no logical reason why these two moments should necessarily be the same. (Note that a successful resuscitation would not be a reentry into life following passing away; the fact that it was successful ipso facto contradicts the permanence and that the person ever passed away in the first place.)
It should be emphasized that these two death-related concepts are not the same as the two kinds of definitions that Gómez-Lobo distinguished:
ordinary language definitions and specialized language definitions. Most people understand ‘water’ to mean, roughly, ‘a transparent liquid that flows from the kitchen or bathroom faucet, and is safe to drink.’ However, people with some knowledge of chemistry define it as ‘a liquid whose basic molecule is composed of two atoms of hydrogen and one of oxygen.’
Likewise, it is reasonable to expect that there will be two kinds of definition for the term ‘death.’ First, ‘death’ as ordinarily understood means ‘the irreversible cessation of life’ and applies to all things that have been alive. There is no separate definition that applies, say, only to humans, to the exclusion of animals or plants. Nor can life irreversibly cease more than once. Hence, there is only one death for each organism. Death, furthermore, is a natural, biological event with social consequences, not a moral, legal, or political decision on the part of those observing it. Death itself should not be confused with the ruling that death has occurred. (Gómez-Lobo, 2008, 96)
The proposal here is more radical. Rather than two ways of defining the same phenomenon, the two definitions correspond to two conceptually (and usually temporally) distinct moments, neither of which is a mere “moral, legal, or political decision on the part of those observing it,” and each of which is “a natural, biological event with social consequences.” I agree with Gómez-Lobo that life cannot “irreversibly cease more than once”; what I am proposing is that it permanently ceases once and it irreversibly ceases once and that these two “ceasings” are different phenomena that typically occur at different times.
I shall follow the classical tripartite distinction of levels of consideration, initially proposed by Bernat, Culver, and Gert (1981) and virtually universally accepted, namely: definition (concept), criterion, and tests. As aptly expressed by GCC, we must distinguish:
(1) the definition of death, which should be determined so as to capture most accurately the ordinary use of the term “dead” and related terms, (2) the medical criterion for determining that death has occurred, (which must stay current with changes in our scientific understanding of the organism), and (3) the tests to prove that the criterion has been satisfied [in a particular case], which often change with improvements in medical technology. (Gert, Culver, and Clouser, 2006, 289)
Gómez-Lobo (2008, 96) further clarifies, “If the contemporary dispute about death is to be intelligible, the definition of ‘death’ must remain stable.” He describes a “criterion” as an “instrument” for “separation or distinction” (giving the example of “a sieve that separates liquids from solids”). As applied to death, the criterion is “an instrument for setting apart the living from the dead” (Gómez-Lobo, 2008, 97). I emphasize these interpretations of “definition (concept), criterion, and tests” because some things called “definitions” of death (e.g., “statutory definitions”) are actually criteria, and some things called “criteria” (e.g., “standard diagnostic criteria”) are actually tests. I shall employ the terms according to their classical BCGC meanings.
It is a mistake to imagine that the definition (whether of one or two forms of death) must be applicable to every conceivable scenario. To the extent that the entity to which the definition applies is the organism, the definition of its death will have the same semantic scope as “organism.” Some of the confusion in the death debate results from the lack of a developed philosophy and clear definition of “organism.” Challenges have been made to proposed criteria for death on the basis of rare or hypothetical cases where the organism concept itself is ambiguously applicable (dicephaly, conjoined twins, extreme amputation/mutilation, cyborgs, etc.). Such philosophically interesting but clinically irrelevant scenarios were considered in the first half of this paper. The remainder of the paper will be exclusively concerned with cases where the organism concept applies unambiguously, which account for 99.9999 per cent of actual deaths and ethically laden death-related situations. It may be that totally different definitions and corresponding criteria need to be developed for the bizarre contexts—a task best left to philosophers.
Passing away (noun), to pass away (verb), passed away or deceased (adjective)
This is the death-related-concept for most practical purposes. It is the moment that loved ones instinctively and appropriately begin to grieve. It marks the annihilation of that individual as related to the world and as a member of society; it should therefore be the basis for the legal “time of death” (or better expressed, post-paradigm shift, as the legal “time of passing”).
Passing away is defined as the permanent cessation of the organism as a whole. It occurs at the moment of onset of the cessation (not when the permanence is declared).
A sufficient, but not necessary, criterion is the permanent absence of both consciousness and circulation of oxygenated blood. A criterion both sufficient and necessary is the absence of both consciousness and circulation of oxygenated blood for a period of time at least up to the moment of deanimation.12 The moment of passing away is the beginning of that joint absence. In the case of an unsuccessful attempt at cardiopulmonary resuscitation (CPR), the moment of passing away corresponds to the cessation of resuscitation effort, not the onset of the cardiopulmonary arrest (given that some circulation of at least partially oxygenated blood occurs during CPR). (N.b., the redistribution of blood across the capillary bed as the arteriovenous pressure gradient disappears during asystole does not count as “circulation.” It does, however, form the basis for the capacity for autoresuscitation and the “Lazarus phenomenon”—vide infra.)
Appropriate tests for the criterion will vary tremendously according to the specifics of individual cases. The focus is on circulation since consciousness cannot be sustained more than a few seconds following cessation of circulation. In the context of a do-not-resuscitate order, a sufficient test would be the retrospective observation that, in fact, circulation of oxygenated blood did not resume during the period of its potential for spontaneous resumption. In the context of a “full code,” the observation time would have to be extended to the point of loss of potential for spontaneous resumption of circulation following the stopping of unsuccessful resuscitative efforts. The precise observation times are a matter of empirical determination, a topic beyond the scope of this paper. Exactly how the absence of circulation can be reliably determined will also vary from case to case. In many situations, simple palpation of the precordial and peripheral pulses should suffice. With obese people, however, the pulse may be difficult to palpate, complicating testing for the criterion. In the context of shock, the pulse can be very weak and an unreliable indicator of the presence or absence of circulation. In an ICU setting, an indwelling arterial catheter can reliably reveal cessation of circulation, as would an electrocardiogram showing asystole or ventricular fibrillation. Pulseless electrical activity, which is sometimes seen as a preterminal electrocardiographic phenomenon, does not count as circulation.
A peculiarity of the paradigm shift—in fact a key element—is that one of the death-related concepts, for the first few minutes or perhaps even tens of minutes, is physically though not logically reversible. (If the patient were successfully resuscitated during that interval, then by definition, the criterion for passing away would not have been fulfilled because the cessation was not, in fact, permanent.) There are several advantages to defining the “civil end,” so to speak, in terms of permanence rather than irreversibility. For one thing, it places the moment of civil end at the time and occasion when people intuitively understand it to occur. My intuition that Soran really “passed away” at the end of his last heartbeat is no doubt similar to the experiences of those witnessing most deaths throughout human history. For another thing, it permits the concept and criterion to correspond in degree of temporal precision. Up to now, experts on death have grappled (unsuccessfully) with the discrepancy between the all-or-nothing life/death dichotomy and the imprecise timing of its putative criterion based on irreversibility. The irreversibility of cessation of X (whether circulation, brain function, or anti-entropic capacity) may in principle occur at a moment, but it is intrinsically unobservable; it is an inference based on our understanding of physiology. No matter how much is learned about the biochemical pathophysiology of anoxia-ischemia, there will always be an estimated range of time when the damaging effects on a critical number of organs and tissues throughout the body become irreversible, and this range will always have a degree of practical uncertainty about it as applied to a given case. Not so with permanence, the onset of which is directly observable and precisely identifiable.
Another peculiarity of passing away is that during the first few minutes or perhaps tens of minutes following the onset of asystole, two bodies could be in identical physical states, but one would have passed away and the other not, based on a future contingency. That is, if the cessation of circulation remains permanent, then the patient passed away at the onset of combined asystole and unconsciousness, whereas if the patient were successfully resuscitated, then the patient did not pass away at the onset of combined asystole and unconsciousness. In the traditional way of thinking about death, if two bodies were in an identical physical state, then they necessarily had the same life/death status. That this is not the case with passing away is not a defect in the concept but a virtue, insofar as it allows incorporation of disparate insights that seemed incompatible under the traditional notion of death. It allows the whole death elephant finally to be pieced together.
What I am suggesting is in some ways similar to the proposal of James Dubois that death be regarded not as an event but as a state, which begins with the onset of asystole and unconsciousness but can also be exited, returning to life during the period of potential for reversibility (citing near-death experiences as particular support) (DuBois, 2002). But passing away differs significantly from Dubois’ death state, insofar as it is by definition permanent, thereby maintaining an important feature generally attributed to traditional death, namely that (barring miraculous intervention) the dead do not return to life.
In the above-stated criterion for passing away, I intentionally avoided mention of heartbeat and breathing, or “cardiac and pulmonary functions,” or even “respiration,” so as to accommodate hospital-based situations such as the intraoperative death of someone disconnected from a cardiopulmonary bypass machine, whose passing away would occur at the moment of disconnection (when circulation of oxygenated blood ceased), not at the moment the patient’s heart and lungs ceased functioning some hours earlier.
The criterion is similar in intent to proposals that explicitly require absence of the three major signs of life: consciousness, circulation, and respiration, mediated (under usual circumstances), respectively, by the brain, cardiovascular system, and respiratory system (including lungs, muscles of respiration, and the medullary centers that drive them—see President’s Council on Bioethics, 2008, 21–30). The focus, however, should not be on organ systems but on their function with respect to the organism as a whole: rather than “cardiac function,” “circulation of blood”; rather than “pulmonary function,” “oxygenated.”
In actual cases of human death, the order in which these three critical functions are lost can occur in every possible permutation, depending on the cause of death. For example, with a primary cardiac arrest, first circulation stops, then consciousness, then breathing. With drowning, first breathing stops, then consciousness, then circulation. With most neurological catastrophes, first consciousness is lost, then breathing, then circulation. The point of the criterion is that passing away occurs at the moment all three of these critical functions permanently cease. If one of them returns (whether spontaneously or through some intervention) during the ensuing period of reversibility, then the criterion would not have been fulfilled and the concept of permanent cessation of the organism not instantiated.
A paradoxical consequence of the paradigm shift is that someone can be morally and legally responsible for causing someone to pass away, even though the culpable act or omission may have taken place after the passing away. The backward causality is only apparent, however, because the relevant causality is not physical but logical. Let me explain by way of example. Suppose a patient is hospitalized for minor surgery; she is not terminally ill and is “full code” status. Suppose further that, entirely unexpectedly, she suffers a cardiac arrest and that her nurse notes the monitor alarm but intentionally does nothing. That culpable omission did not cause the cardiac arrest, but it did cause the cessation of circulation of oxygenated blood to be permanent. Therefore, that nurse would be just as guilty of negligent homicide under the concept of passing away as under an irreversibility-based concept of death. The flip side of this coin is that certain otherwise-death-causing acts during the period of reversibility (such as removing unpaired vital organs) can be ethical and legal, precisely because they do not cause passing away, nor do they cause or hasten deanimation. But more on this later.
Deanimation (noun), to deanimate (verb), deanimated (adjective)
This is the death-related-concept relevant to ontology and some religious purposes. It marks the annihilation of the organism as such, the substantial change from organism to “collection of parts of a former organism”—or expressed in hylomorphic terms, the cessation of the soul “informing” and unifying the body as its animating principle.
Deanimation is defined as the irreversible cessation of the organism as a whole. It occurs at the moment of onset of irreversibility (not onset of the cessation). “Irreversible” is here understood in the strongest sense, that is, “impossible to reverse by any natural means” (whether presently available or not). Weaker senses of “irreversible” become relevant for the ethics of transplantation, as will be explained below, but do not enter into the definition of deanimation.
A sufficient and necessary criterion is the irreversible cessation of anti-entropic exchange of substances with the environment.13 I have chosen not to express the criterion in terms of irreversible cessation of circulation of oxygenated blood, but to move circulation to the level of tests, because cessation of circulation is not truly irreversible until the entire vascular system becomes plugged by either diffuse coagulation of stagnant blood or breakdown of the entire capillary bed, and this may not occur until well after the irreversible cessation of anti-entropic dynamics. (A mechanical pump-oxygenator could in principle be attached to a postdeanimation corpse and force oxygenated blood to flow through the vascular tree. In such a case, a circulatory criterion would fail, in contrast to an anti-entropic criterion.)
A sufficient (but not necessary) test for this criterion is some minimal duration of absence of circulation (to be empirically determined) beyond which diffuse anoxic-ischemic cellular damage has extinguished the body’s anti-entropic capacity, and the body (strictly speaking, its remains) gives way to entropy (i.e., inexorable biochemical breakdown). Another sufficient (but not necessary) test could be some form of high-tech imaging (e.g., something like whole-body magnetic resonance spectroscopy) that reveals widespread biochemical changes throughout the body pathognomonic for the loss of antientropic capacity. Other sufficient tests could be prescribed passages of time without circulation (beyond antientropic irreversibility) particular to various (nonhylomorphic) religious beliefs regarding the soul. Of all bodily functions, circulation is singled out as key for testing for deanimation because it is the only means of metabolic communication among the parts of the body (mediating their unifying mutual interaction) and between them and the environment, and it is a prerequisite for the only other form of coordination within the body and between it and the environment, namely, nervous system functioning.
Although deanimation has relevance for some religious purposes, it is essentially an ontological concept, referring to the ceasing-to-be of a living organism as such, considered in itself and apart from relationship with the world. From the perspective of hylomorphic dualism, deanimation corresponds to (and literally means) the separation of the soul from the body (or more accurately expressed: cessation of the soul “informing” the matter of the body as its substantial form and vital principle). This applies not only to human souls, which (according to hylomorphic dualism) have a spiritual dimension that allows them to persist after death (though in an unnatural incomplete manner), but also to animal and plant “souls” (which is just another term for their substantial form). I subscribe to this view of animate nature and human nature in particular, but one need not be a hylomorphic dualist to appreciate the conceptual advantages of defining a “metaphysical end,” so to speak, in contradistinction to a “civil end.”
What Pope John Paul II said about death in his 2000 address to the Transplantation Society was an excellent description of what I am calling “deanimation”:
[T]he death of the person is a single event, consisting in the total disintegration of that unitary and integrated whole that is the personal self. It results from the separation of the life-principle (or soul) from the corporal reality of the person. The death of the person, understood in this primary sense, is an event which no scientific technique or empirical method can identify directly. Yet human experience shows that once death occurs certain biological signs inevitably follow, which medicine has learnt to recognize with increasing precision. In this sense, the ‘criteria’ for ascertaining death used by medicine today should not be understood as the technical-scientific determination of the exact moment of a person’s death, but as a scientifically secure means of identifying the biological signs that a person has indeed died. (John Paul II, 2000 [emphases in original])
The definition of deanimation hinges on the notion of irreversibility. Four different interpretations of irreversibility appear in the literature on death. After presenting them, I shall offer reasons why only the temporally last (and “strongest”) interpretation is applicable to deanimation. They are in chronological order (and also more or less in the order of what has been called the “weakest” to “strongest” sense):
Impossibility of autoresuscitation. The “Lazarus phenomenon” refers to spontaneous resumption of heartbeat some minutes after what seemed to be its permanent cessation. It is believed to be due to a coronary arteriovenous blood pressure gradient (the “coronary perfusion pressure”), which takes some minutes to decay asymptotically to zero during asystole (Verheijde et al., 2009). The critical perfusion pressure above which cardiac tissue remains viable and below which necrosis begins to set in is around 15 mmHg (Paradis et al., 1990). In clinical practice, the coronary perfusion pressure is never measured, and its rate of decay to zero after asystole surely varies across patients, depending on the particular cause of asystole and multiple other factors. All we have to go on for estimating the point of impossibility of autoresuscitation are statistical data on the latencies to spontaneous resumption of heartbeat that have been reported anecdotally. The distribution of such latencies is presently unknown except that they cannot be normally distributed. The literature related to non-heart–beating organ donation (NHBOD) suggests that the bulk of spontaneous resumptions after asystole occur within the first minute, fewer during the next minute, and very few (but a still ethically significant number) afterward, with an upper limit yet to be determined. Surely, 75 s (Boucek et al., 2008) is too soon, as is the Pittsburgh protocol’s 2 min (Lynn, 1993), as is the 5 min that many other, more conservative, protocols employ (Verheijde et al., 2009). The longest reported latency to autoresuscitation that I know of is 10 min (Adhiyaman et al., 2007; Joffe, 2007a), and we should be prepared for future reports of even longer latencies. Of course, low body temperature will raise the upper time limit for autoresuscitation. Little data exist on this, and the aforementioned times assume normothermia.
Decisional irreversibility. This is the interpretation that Lizza has proposed as particularly relevant to death (Lizza, 2005, 2007), although he is currently in the process of rethinking it (J. H. Lizza, personal communication). Such a concept of irreversibility applies after the potential for autoresuscitation has passed. It is based not on an intrinsic property of the dying organism but on a decision of someone else not to intervene to restore circulation. There is still an intrinsic potential for resuscitation with external assistance, but in the particular case, resuscitation would (presumably) be ethically inappropriate and a decision has been made not to attempt it. (There could also be unethical and criminal decisions not to resuscitate.) If the decision was in place prior to the cardiac arrest, then the timing of decisional irreversibility coincides with autoresuscitative irreversibility. If the decision is made after the onset of autoresuscitative irreversibility but before current interventional irreversibility, then decisional irreversibility takes place at the time of the decision.
Current interventional irreversibility. This means that the condition cannot be reversed by any means available at the moment. It is contingent not only on the progress of medical technology (what is reversible today in an ICU might well have been irreversible a century ago) but also on the circumstances of the case (what is reversible today in an ICU might well be irreversible today on a hiking trip). The moment of onset of current interventional irreversibility surely varies tremendously across cases, depending on multiple, mostly unknowable, factors; the variation is surely much greater than for autoresuscitative irreversibility. In some terminally ill patients, the onset of current interventional irreversibility might occur only seconds after the onset of autoresuscitative irreversibility. In healthy patients with cardiac arrest due to extrinsic factors, such as electrocution or near drowning, the point of current interventional irreversibility is affected by multiple factors, including blood glucose concentration (Müllner et al., 1997) and especially body temperature. At normothermia, the success rate for CPR drops off asymptotically for arrests longer than 2 min, but the tail of the curve extends at least up to 20 min (Myerburg and Castellanos, 2001, 910–11) With cold water near-drowning in which body temperature quickly plummets to protectively low levels, successful resuscitative efforts have been reported after immersion as long as 66 min (Hughes et al., 2002). Controlled intraoperative circulatory arrest under deep hypothermia as long as 91 min has been reported with complete recovery (Knezevic et al., 2008) (and as long as 120 min in a swine model; Alam et al., 2008). We should distinguish current interventional irreversibility of cessation of heartbeat from current interventional irreversibility of cessation of the organism’s antientropic capacity, which probably occurs some time after the former and has a closer relationship with the criterion of deanimation than the former.
Absolute irreversibility. This means that the condition cannot be reversed by any natural means whatsoever, whether presently available or possibly available in the future. This point is reached when sustained anoxia-ischemia has caused a critical number of cells in a critical number of organs and tissues to begin the biochemical cascade inextricably leading to autolysis (i.e., the point beyond which the organism’s antientropic dynamism definitively fails and disintegration proceeds unabated). The number and identity of organs and tissues and the percentage of cells within each organ or tissue required for such “criticality” is unknown and perhaps unknowable. One should distinguish between the impossibility of restoring the body’s antientropic dynamics and the impossibility of restoring spontaneous heartbeat and circulation; only the former is relevant to deanimation. As with current interventional irreversibility, the duration of circulatory arrest required to produce absolute irreversibility is no doubt dramatically affected by body temperature. It surely occurs prior to livor mortis or rigor mortis, which at normothermia typically begin around 30 min to 2 h and 2–4 h, respectively, after circulatory arrest (D. J. DiMaio and V. J. M. DiMaio, 2001, 23, 26). How much prior is anybody’s guess.
It is important to realize that all four types of irreversibility are really in the epistemic rather than ontological domain. The literature almost always treats these ethically significant milestones as single time-latencies applicable to all cases, usually taken to be the longest reported latency to X (whether autoresuscitation or successful interventional resuscitation, etc.), with the value being raised each time a new clinical record is set. But such durations are really upper ends of statistical distributions (based on a sample population of similar cases) rather than potentialities of individual organisms. Whether John’s heart is going to autoresuscitate following asystole, and if so, when, is determined by myriad properties of John’s body at the time, almost all of which are unknown. If the confluence of these factors destines him not to autoresuscitate, then he has no potential for autoresuscitation. If they destine him to autoresuscitate at 1 min, then he has the potential to autoresuscitate at 1 min. Since it is impossible to know what John’s potential for autoresuscitation really is, the best we can do is to estimate the probability that John will autoresuscitate within elapsed time T, based on experience with many cases like John. That probability is an expression more of our ignorance about John’s potential than of the potential itself. Therefore, when we say that autoresuscitative irreversibility occurs at M minutes, what we really mean is that the probability that John will autoresuscitate becomes essentially zero at M minutes, not that John has the potential for autoresuscitation up to M minutes. The same is the case for the other three types of irreversibility, even to some extent decisional irreversibility, which either coincides with autoresuscitative irreversibility or occurs after it (depending on when the decision not to intervene was made). By realizing that these various irreversibility times are statistical upper limits, not intrinsic potentials of a particular organism, one understands better why they might serve as conservative markers of “ethical safety,” and for the one that corresponds to deanimation, why it inherently cannot identify when deanimation occurs but only that it has occurred.
Of these four kinds of irreversibility, the only one applicable to the concept of deanimation is the “strongest” kind, that is, absolute irreversibility. This is because deanimation has to do with an intrinsic property of the body at the transition between organism and remains of an organism, between an entity and a collectivity. What the organism (or remains) is in itself cannot be contingent on extrinsic vicissitudes such as someone else’s decision to resuscitate or not or the state or availability of medical technology. This strongest form of irreversibility has been rightly criticized as a criterion for death on the grounds of both its epistemic temporal vagueness and its impracticality for many death-related purposes (particularly transplantation). Nevertheless, it is the only form of irreversibility that makes any sense in relation to ontology. This is one more reason why the field of thanatology stands in need of a paradigm shift. With the introduction of the concept of passing away, the criticisms leveled against a strong-irreversibility interpretation of death no longer apply, and no theoretical or practical disadvantages are occasioned by the unknowability of the moment of deanimation. Practicality and ontology no longer have to stand in competition. Unlike with passing away, there is no need to know exactly when deanimation occurs; it is sufficient to know simply that it has occurred.
The practical consequences of deanimation have largely to do with religious practices (apart from Judaism, in which the concept of death corresponds more to passing away than to deanimation [Jakobovits, 1975; Weiss, 1988; Tendler, 1989; Steinberg, 1996; Rosner, 1999; Z. H. Rappaport and I. T. Rappaport, 2004)—accommodation of all religions being yet another positive consequence of the paradigm shift). Catholicism, for example, officially endorses hylomorphic dualism according to the Aristotelian-Thomistic tradition (Council of Vienne , 1957). The biochemical-entropic underpinnings of deanimation fit perfectly with the understanding of death as the cessation of the soul’s “informing” the matter of the body as its substantial form and vital principle. Because of the uncertainty of timing of deanimation (not because the soul is an invisible entity that invisibly “departs”—which would be a Cartesian caricature—but because of the empirical uncertainty when absolute irreversibility occurs), Catholic priests are permitted and encouraged to administer the last sacrament conditionally to an apparently already dead body (in the terminology of the paradigm shift to a “passed away” or “deceased” body) prior to indisputable evidence that the soul is no longer present (in the terminology of the paradigm shift: “indisputable evidence that deanimation has taken place”), such as livor mortis or rigor mortis. (By contrast to hylomorphic dualism, substance dualism would have a very hard time offering a nonarbitrary criterion and test for deanimation and providing a coherent explanation why the soul departs when it is claimed to in relation to the cascade of bodily changes taking place after asystole.)
Some non-Christian religions have a nonhylomorphic notion of the human spirit for which the word soul (anima) should probably not be considered synonymous. For example, a traditional Japanese belief, deriving from Buddhism and Shinto, is that after biological death, the spirit of the deceased is still in some way related to the bodily remains for days afterward; during that time, one must not mutilate the corpse (e.g., for autopsy or organ donation) for the sake of the spirit’s happiness and eventual harmonious transition to the definitive afterlife (Feldman, 1988; Lock, 1996, 1999, 2001, 211–5). Even in such a framework, the concept, criterion, and tests of passing away and deanimation as proposed here could still apply (with deanimation understood in its entropic sense, not as definitive departure of the spirit), with the religious duties owed to the bodily remains of the deceased continuing postdeanimation for as long as custom prescribes.14
The differences of opinion across religions regarding the criterion and timing of deanimation are of course irreconcilable, but this is not a bad thing. So long as everyone can agree on passing away as the permanent (not irreversible) cessation of the organism as a whole (and I believe that everyone can, once they realize that the paradigm shift allows them to accept this without threat to their understanding of deanimation), then there is no problem with a diversity of criteria and tests for deanimation. It is worth pointing out, however, that the Catholic view on this matter is perhaps unique among religions because it is essentially philosophical in nature. By basing its understanding of death (deanimation) on the philosophical framework of hylomorphism rather than on, say, scripture, revelation, faith, or tradition, Catholicism understands the empirical aspects of human deanimation in essentially the same way it understands animal deanimation. Only what happens to the human soul afterward becomes a matter for theological speculation.
Implications for NHBOD
The weaker kinds of irreversibility do not enter into the definition of either passing away or deanimation, but they become highly relevant for ethical decision making during the interval between passing away and deanimation, that is, between the onset of permanent cessation and the onset of absolute irreversibility of cessation of the organism as a whole. I shall focus here on the implications of the paradigm shift for the ethics of NHBOD.15
I want to make the following disclaimer up front. The ensuing discussion, whether in principle vital organs can be ethically removed soon enough after asystole that they are still viable for transplantation, is to be understood as purely theoretical. I have previously argued that they can be (Shewmon, 2004b; D. A. Shewmon and E. S. Shewmon, 2004), and I continue to hold that position. This does not necessarily mean, however, that I advocate NHBOD as public policy at the present time. In particular, I am not convinced that the requirements of truly informed and truly free consent, absence of conflict of interest, certainty of apnea and cardiac arrest within a short time following disconnection from the ventilator, and respect for the dignity of the donor’s body are presently fulfilled in any current NHBOD enterprise. Nor do I suspect that they are likely to be met in the near future, as the perceived “need” for organs and the pressure to obtain them relentlessly increase. In this regard, I share many of the practical concerns expressed by the President’s Council on Bioethics (2008, 79–88) regarding NHBOD. Nevertheless, it is still worthwhile to examine how NHBOD might be understood in relation to the proposed paradigm shift, assuming for the sake of argument that all the ethical prerequisites were in place.
Up to now, the fundamental question surrounding NHBOD has been whether the dead donor rule (DDR) was respected or not. The DDR has generally been regarded as sacrosanct, and NHBOD advocates have gone through all kinds of conceptual contortions to justify that death has taken place by the time period specified in their particular protocol (whether 75 s, 2 min, 5 min, or whatever). These attempts (largely based on weak interpretations of irreversibility in the definition of death) have never been very convincing. I have argued, along with others, that the prohibition against intentional killing does not necessarily require the DDR and that under certain clearly defined conditions resembling NHBOD protocols, even unpaired vital organs can be removed without causing or hastening death (Shewmon, 2004b; D. A. Shewmon and E. S. Shewmon, 2004).16 That is, because, once circulation has permanently ceased, those organs are no longer truly “vital” in the sense that they were prior to asytole. Without blood flow to or from them, they are nonfunctional (although not irreversibly so yet); nor does their unactualized potential for functioning contribute to maintenance of the organism as a whole. Thus, these are not vital organs anymore for the donor, although they have the potential for becoming vital again in a recipient.
From the perspective of the paradigm shift, the NHBOD donor actually passed away at the moment of onset of asystole, not 2 or 5 min or some other time interval later. That is the time that should be entered on the “decedent certificate.” That does not imply, however, that it would be ethical to remove organs immediately upon asystole, on the (purported) grounds that the DDR is respected because the donor is already deceased. Recall the example mentioned above of negligent homicide, and how an intentional omission after asystole can cause the permanence and therefore be the (retroactive) cause of passing away. The same reasoning applies to the removal of vital organs during the initial period of potential for autoresuscitation following asystole. Such removal either eliminates the very basis of spontaneous reversibility (if what is removed is the heart) or risks guaranteeing death days later (if what is removed is some vital organ or organs other than the heart, and the heart were to autoresuscitate). Therefore, such removal would be the proximate cause, in an ethically and legally meaningful sense, of the passing away that took place prior to it (because it caused the permanence). What is ethically important is not that the donor be dead (in some irreversible sense) at the time of removal or organs but that the act of removal not cause the donor’s passing away. The DDR, as traditionally understood, fallaciously focuses on the former to the exclusion of the latter consideration. However, with the concept of passing away, combined with the requirement to wait until after the period of spontaneous reversibility before beginning organ retrieval, both aspects are achieved at once: the donor is in fact legally (and really) deceased at the time of organ retrieval and the removal of organs did not cause the passing away (because what established the permanence was the appropriate waiting period, not the removal of organs). By replacing “dead” in “dead donor rule” with “deceased,” the DDR is still fulfilled (even the acronym is preserved), and the timing of passing away becomes very simple and straightforward, in contrast to the logically incoherent and biologically problematic attempts to place “death” at some time after the onset of asystole. Put another way, a “deceased donor rule” combined with the existing laws against homicide fulfills the intent of the “dead donor rule” without its associated conceptual problems.
The removal of heart, lungs, or the heart-lung complex entails unique ethical issues compared with removal of kidney, liver, or other organs because of the intimate connection between heart-lung and “circulation of oxygenated blood,” which is at the core of the criterion for passing away. I know of one ethicist who has proposed placing the donor on a cardiopulmonary bypass machine while heart and lungs (and perhaps other organs as well) were explanted, so that the removal would not in fact cause cessation of circulation of oxygenated blood (Dr. Manfred Lütz, personal communication). Afterward the bypass machine is recognized to be an extraordinary, disproportionate means of life support, so it can be ethically discontinued for the same reasons that it would have been ethical to discontinue the ventilator from the patient in the first place apart from organ donation (an ethical prerequisite for all NHBOD approaches). (The bypass machine was an extraordinary, disproportionate means also before the completion of organ retrieval, but one that the patient accepted and consented to in this hypothetical scenario.) According to the new paradigm, the patient would pass away on discontinuation of the bypass machine after the completion of organ harvesting (in contrast to all other NHBOD scenarios, where the patient passes away at onset of permanent asystole, prior to the commencement of organ harvesting).
The ethical propriety or impropriety of this approach to heart-lung harvesting seems to require a different analysis from the typical NHBOD scenario, which does not involve a bypass machine. Intuitively, the proposal seems like a mere ethical sleight-of-hand to accomplish the equivalence of a contemporary Aztec-style human sacrifice. But perhaps if all the ethical requirements of NHBOD were (hypothetically) in place, that would be enough to ground a radical ethical difference. However, even if in the final analysis that might be deemed so (which I doubt), I shall go on record as not endorsing this approach. If hearts and lungs cannot be harvested in a viable condition after waiting for the potential for autoresuscitation to pass, then perhaps they simply should not be harvested.
I shall refrain from digressing onto the ethical controversies surrounding administration of anticoagulants and other medications prior to passing away, controlled versus uncontrolled deaths, presumed consent, and many other variations inspired by the desire to maximize the number of transplanted organs and their viability. This section is not intended to be a comprehensive treatise on NHBOD, but only an illustration of how the proposed paradigm shift applies to transplantation.
“Brain death” began as a legal construct without a coherent philosophical or even factual biological basis (Shewmon, 2009b). It remains a legal fiction, and a substantial proportion of health-care professionals and society at large understand it as such (Youngner et al., 1985, 1989; Siminoff et al., 2004; Joffe and Anton, 2006; Joffe et al., 2007). Thought experiments involving decapitation, dicephalic twins, and myriad variations on these themes do not shed light on the nature of TBF, but rather require empirical data from actual TBF cases for their proper interpretation. The radical capacity for specifically human or even specifically mammalian properties may reside not so much in the brain as in the genetic and epigenetic information distributed throughout the living organism.
The vast literature on the definition and diagnosis of death represents an increasingly complex compendium of valid observations and insights mixed with false assumptions, loose terminology, and fallacious inferences. Attempts to construct the entire “death elephant” from all the disparate perspectives have been stymied by the apparent incompatibility between certain features, each of which seems valid in itself. That is a sure sign that a fundamental paradigm shift is required to move ahead.
The proposed paradigm shift for the concept, criterion, and tests for death involves reconceptualizing as dual what had previously been considered a single phenomenon, a duality based on the distinction between permanence and irreversibility. Everyone recognizes that these two notions are of key importance for a proper understanding of death, but they have never been reconciled or brought together in a coherent way. Instead of defining and diagnosing “death,” I propose that we define and diagnose two death-related events, labeled here as “passing away” and “deanimation.” The former represents an organism’s relational or civil end and is defined as the permanent cessation of the organism as a whole; the latter represents its ontological end and is defined as the irreversible cessation of the organism as a whole. By recognizing each as equally “real” and valid death-related phenomena, the seemingly intractable paradoxes and incompatibilities disappear, and the death elephant can finally be constructed coherently. By no means, do I imagine this proposal to be the last word in the death debate, but perhaps it might be the first word in a new chapter of it.
This paper represents a modified and substantially expanded version of one prepared for a panel session on brain death at the APA meeting in Vancouver, BC, April 10, 2009, subsequently published in the APA Newsletter (Shewmon, 2009c, forthcoming).