Journal of Scientific Exploration, Vol. 6, No. 2, pp. 127-144, 1992 0892-33 10192 O 1992 Society for Scientific Exploration Survival or Super-psi? STEPHEN E. BRAUDE Philosophy Department, University of Maryland Baltimore County, Baltimore, MD 21228 Abstract-Even the most sophisticated discussions of the evidence for survival underestimate the conceptual difficulties facing the survival hypothesis. Per- haps the major challenge is posed by the rival "super-psi" hypothesis, which most writers fail to confront in its most plausible and potent form. Once the super-psi hypothesis is taken seriously, two major weaknesses in discussions of survival stand out clearly. First, analyses of apparently anomalous knowledge that tend to be fatally superficial in their treatment of subject psychodynamics. And second, analyses of apparently anomalous abilities and skills trade on an impoverished and naive conception of the nature of human abilities. Introduction Two questions have dominated parapsychology since the founding of the Soci- ety for Psychical Research. The first is: "Do human beings have psychic (psi) abilities (ESP or PK)?". The second is: "Does human consciousness survive bodily death?" The first question, in my opinion, has been answered satisfactori- ly in the affirmative, and although the road to that answer has been bumpy and somewhat circuitous, the issues are relatively straightforward-at least as com- pared to the second question. But the question about survival is almost transcen- dentally recalcitrant; it may, in fact, be impossible to rule out alternative hypotheses. Unfortunately, however, most writers on survival have failed to appreciate the point. Indeed, even the most sophisticated authors tend to underestimate the enormity of the theoretical challenges facing them. The goal of this paper is to bring some of their more serious shortcomings into sharper focus. Any satisfactory discussion of survival must proceed along two fronts, one of which is empirical and the other of which is philosophical. In this respect, the study of survival is no different from any scientific investigation, except that in most areas of science fundamental philosophical assumptions usually form part of the received conceptual background. But in the case of survival research, deep philosophical issues dominate the foreground. For example, many have wondered whether the very concept of survival is intelligible. And as the reader may realize, some have decided that it is not, and have therefore argued that we should reject the survival hypothesis a priori (see, e.g., Flew, 1976). Now I am by no means opposed to a priori arguments against scientific claims. In many cases, ostensibly empirical claims rest on thoroughly indefensi- ble philosophical presuppositions, and those presuppositions often blind 128 S. E. Braude researchers to alternative ways of interpreting the data. That is perhaps most conspicuously the case in the so-called cognitive sciences, a great deal of which is no more than bad philosophy couched in (and obscured by) the imposing tech- nical vocabulary of the electrical engineer. Nevertheless, some a priori argu- ments are more persuasive and profound than others. And I consider the usual arguments against the intelligibility of survival to be quite shallow. Flew's argu- ments, for example, trade on superficial features of language use and suspicious thought-experiments whose plausibility rests on our meager imaginative capaci- ties. In my view, it is irrelevant how much difficulty we have imagining what survival might be like. Similarly, it is hardly surprising that we have trouble describing an after-life in terms designed for dealing with an ordinary embodied existence. What really matters is that it is relatively easy to construct hypotheti- cally ideal cases so coercive that we would have no choice but to admit (or at least to entertain seriously) that survival of some sort is a fact, no matter how much of a challenge that poses for our familiar conceptual framework. Our abili- ty to formulate such ideal cases shows that the evidence for survival cannot be rejected for the sorts of facile reasons provided by Flew. The more pressing question is to what extent actual cases approach the theoretical ideal. But even the best real cases-and possibly also the best ideal cases-face cer- tain purely conceptual obstacles. These have to do with the challenge from an alternative psi-hypothesis, usually called the "super-psi" hypothesis. As I see it, the most serious obstacle to taking even the best evidence for survival at face value is the possibility that the data can be explained in terms of highly-refined psi among the living. What I shall argue below is that no case so far investigated resists explanation along these lines, and that the usual arguments against super- psi explanations are seriously inadequate. Super-psi and the Ideal Case for Survival Parapsychologists and others tend to approach the study of psi with a standard set of indefensible tacit assumptions. First, they often assume that if psi occurs, it occurs only to .a very modest degree. Second, they usually assume that when observable (and not merely statistically demonstrable) psi effects occur, those effects will be sufficiently overt or unusual to be identified as psi events. Both assumptions, however, are intolerably naive. Indeed, the former is methodologically egregious. Given our current degree of ignorance concerning 1 psychic functioning, we simply have no grounds for placing antecedent limits of any kind on its scope or refinement. To accept this point, it does not matter whether we believe that psychic functioning occurs or whether we are simply open-minded skeptics entertaining the mere hypothesis that psi occurs. If psi can occur at all, then until we have evidence to the contrary, we must assume that it can occur at any level of magnitude or sophistication. In fact, not only do we have no evidence against the possibility of unlimited psi, certain bodies of data actually support it. I have argued elsewhere that probably the best evidence for psi of any kind is the evidence of physical mediumship (Braude, 1986). But the Survival or Super-psi? 129 best of those cases shows that psi occurs in forms far more subtle, complex and extensive than would be suggested by any of the results reported from laboratory experiments. A sober appraisal of the mediumistic evidence can only make us more open to the possibility of even more elaborate or sweeping effects. The second unwarranted assumption usually appears in attempts to argue that there is simply no evidence for super-psi. Many protest that if super-psi occurred, it would make itself known to us; but (they would argue), we have no evidence that people are able to do such things as psychokinetically affect the weather or make planes crash, or carry out sophisticated and detailed psychic spying. The assumption lurking beneath this argument is that occurrences of super-psi in everyday life will generally be conspicuous or easily identifiable as such, and that they will not simply blend in with or be masked by the extensive network of surrounding normal events. But that assumption is clearly defective (for a detailed discussion of the issues, see Braude, 1989). As far as physical phenomena are concerned, there need be no observable difference between (say) a normal heart attack or car crash and one caused by PK. The only difference may be in their unobservable causal histories. Similarly, there is no reason to suppose that information gathered by ESP has to hit us over the head with its obviousness. It needn't carry some marker-a phenomenological analogue to a flourish of trumpets-which identifies it as paranormally derived rather than ran- dom or internally generated normally. When we think about the possible operation of super-psi, we must be careful not to suppose that it functions in total isolation from the full range of human needs and organic capacities. Quite the contrary; it is more reasonable to sup- pose that psi plays some role in life, that it may be driven by our deepest needs and fears (rather than those of which we are immediately or consciously aware), and that it does not occur only when parapsychologists set out to look for it. Moreover, if psychic functioning may be a component of everyday life, we must be open to the possibility that, like manifestations of other organic capacities, occurrences of psi will range from the dramatic and conspicuous to the mundane and inconspicuous. In fact, it may be in our psychological interest for most psi manifestations to be covert. In the case of PK, that would help to preserve the useful illusion that we are not responsible for events we might have brought about. And in fact, most of us are acutely uncomfortable about accepting the possibility that we might play an active (but covert) role in the misery and calamities that occur around us (Braude, 1986, 1989; Eisenbud, 1982). That concession would force us to accept a magical world view we somewhat condescendingly associate only with so-called "primitive" societies, a world view in which we need to take seri- ously the possibility of hexing (or the "evil eye") and other uses of psychic func- tioning for malevolent and lethal purposes. Equally terrifying for most people is the fear that super-ESP might confer something like near-omniscience, or at the very least a far greater range of knowledge than we would be able to handle emotionally. Let us suppose, then (if only to see where it leads), that we are not justified in ruling out the possibility of large scale or extremely refined psychic functioning. 130 S. E. Braude And let us consider how that complicates our assessment of the evidence for sur- vival. The Survival Hypothesis: Some Preliminaries Before we can get to the heart of the issues before us, we must make some rather elementary, but very important, observations. First, we must resolve an important ambiguity in the claim that consciousness may survive bodily death. This concerns the distinction between survival of bodily death and mere life after death. To be a case of survival, properly speaking, there would have to be a relation of identity between a post-mortem individual and an ante-mortem indi- vidual. In some sense, the two deserve to be considered the same. No doubt it will be difficult to state precise or generalizable criteria in virtue of which that sort of identity holds. Nevertheless, one might think that this sort of identity is not too different from the relatively familiar respects in which other sorts of things remain the same through change. My body now, for example, is quite dif- ferent from the body I had as an infant. Millions of cells have been lost and only some of them have been replaced; others have appeared for the first time as part of the normal growth process, and many of them have likewise been lost and occasionally replaced. Moreover, my thoughts and other mental states have evolved and changed deeply over time. In short, a person may remain the same in some important sense, despite undergoing a vast number of physical and men- tal changes. Similarly, there is a sense in which an acorn is the same as the oak into which it evolves. But even if we grant that two things can be identical with- out being strictly identical-that is, without satisfying Leibniz's Law that every property of one is a property of the other, we must be open to the possibility that after bodily death there is life without survival. Suppose, for instance, that after death one's consciousness gets absorbed in a great universal soup of conscious- ness (or something of the sort). I'm not sure I understand that hypothesis; but I encounter claims like it all the time. In that case, we would say that although no ante-mortem individual is even loosely identical to (or survives as) a post- mortem individual, nevertheless there is some sort of post-mortem existence. Another vital preliminary distinction concerns two different sorts of knowl- edge. Generally speaking, a case suggestive of survival is one in which one or more living persons display knowledge closely (if not uniquely) associated with a deceased person, and which we have good reason to believe could not have been obtained by ordinary means. But that knowledge tends to fall into two broad categories: knowledge that (information or propositional knowledge) and knowledge how (abilities or skills). For example, some cases hinge on pieces of information displayed by a living person, which could not have been obtained normally, but which would have been known to a deceased person now ostensi- bly communicating that information. In other cases, however, a living person (say, a professional medium or a child) displays an ability or skill she never dis- played before (say, the ability to speak a certain language, or write music), or perhaps an ability or skill quite idiosyncratic to a deceased person (say, a certain Survival or Super-psi? 131 distinctive style of humor or musical composition). And of course, in an impres- sive case of this type, we would have good reason to believe either that these abilities could not have been acquired by normal means at all (say, if they are likely to be organic endowments which only a few enjoy), or that they could only be acquired normally after a period of practice which we are quite certain never occurred. Many would say that the cases most strongly suggestive of survival are those of the latter sort, displaying the persistence of knowledge how. I tend to agree. But many mistakenly think that cases of both sorts easily resist alternative expla- nations in terms of super-psi. Let us, therefore, consider the issues surrounding each type of knowledge. This should help us to appreciate, first, why the appar- ent persistence of abilities or skills is more impressive than the mere display of propositional knowledge, and second, why not even the manifestation of abili- ties is as coercive as it might first appear. Super-psi and Propositional Knowledge To simplify matters, let us assume that the cases under discussion are always well-authenticated. In other words, let us assume first, that the hypothesis of fraud is highly improbable, and second, that we have good reason to believe that events occurred as reported. By assuming that testimony is honest and that observation and testimony are reliable, we may therefore concentrate on explaining the phenomena observed and reported. And let us assume, further, that there is good reason to believe that no explanation in terms of currently understood processes is wholly satisfactory. In that case, we may jump irnrnedi- ately to the question of which alternative account, in terms of presumably para- normal processes, we should embrace. Now in a number of cases of ostensible mediumship and reincarnation, per- sons exhibit unusual and surprising pieces of propositional knowledge. Some of the best examples of this sort are cases featuring what Stevenson has dubbed "drop-in" communicators. As the name suggests, "drop-in" communicators appear without invitation, and are usually unknown to both medium and sitters. In the best of these cases, the "drop-in" communicators make various statements about themselves which are later verified and which nobody present at the sitting knew to be true through any normal means. As Gauld (1982, pp. 58ff) has properly observed, such cases would seem to discourage super-psi explanations for two main reasons. The first concerns the identity and apparent purpose of the communicator. One would have to explain why the medium (or someone else present at the sitting) used their ESP to obtain information about an individual unknown to those present and at best only tan- gentially connected to one of them. One would also have to explain why the communicator supplies information of no apparent interest to the sitters but of understandably serious concern to the communicator. (A good example is the case of Runki's leg, reported in Haraldsson and Stevenson, 1975.) By contrast, the survival hypothesis seems rather straightforward. To put it simply, a S. E. Braude deceased individual needs to deal with a matter of importance to him (e.g., to console a grieving relative or take care of some unfinished business), and so he seizes the opportunity of the sitting to get his message across. The second obstacle for super-psi explanations concerns the obscurity and dif- fuseness of the information which the medium (or someone else) would have to acquire paranormally. To begin with, in many cases that information is quite arcane and apparently irrelevant to sitters' concerns-for example, the fact that the ostensible communicator was buried without one of his legs, which he lost in the accident that killed him. And although in most of those cases the information would need to come from only a single source (such as a written record, or one living person's memory), in others the information would have to be assembled from separate and equally obscure sources (e.g., different written records and memories). By contrast, on the survival hypothesis, the necessary information may all be reasonably attributed to the communicator. Both of these alleged problems strike me as overrated. In fact, the second may be dispensed with rather quickly. Since we presently have no grounds for impos- ing any antecedent limits on the scope or refinement of psychic functioning, we are simply in no position to assert that accessing multiple sources of obscure information is any more imposing than accessing one. In evaluating the super- psi hypothesis, we must be careful not to treat the processes involved as if they were simply a collection of really good psi, of the kind we apparently see in lim- ited forms in some lab experiments. When we do that, it is all too easy to think that psi functioning involves an efSort of some kind, and that if one psi perfor- mance is difficult, several ought to be out of the question. But in fact, in all its intimidating richness, the super-psi hypothesis should perhaps be called the magic wand hypothesis. It asserts that (as far as we know) anything at all can happen, given the relevant need for it to happen. For example, we needn't sup- pose that refined PK must be accompanied by constant ESP vigilance of the results of one's activities, in the way that driving a car requires sensory feed- back. It may be enough simply to wish for something to happen, and then it does. Task complexity is simply not an issue. (Ironically, the irrelevance of task complexity has been emphasized even in laboratory experiments on random number generators; see Braude, 1979; Schmidt, 1975, 1976.) The first problem, that of explaining the identity of the communicator, raises a rather different set of issues. Gauld notes correctly that the survival hypothesis has obvious advantages when it comes to explaining why the medium selects one unknown deceased person rather than another unknown deceased person as the subject for her extrasensory researches. The deceased person selects himself. (1982, p. 61) Indeed, as Stevenson once remarked, "Some 'drop-in' communicators have explained their presence very well" (1970, p. 63). But according to Gauld, on the super-psi hypothesis "we seem reduced . . . to supposing that selection of communicator depends upon the random operation of wholly unknown factors." Survival or Super-psi? 133 (1982, p. 59). Stevenson concurs, and his way of stating the point brings its weakness squarely into the open. He writes, since the [super-psi] theory assumes that discarnate personalities do not exist, it has to attribute motive for a particular mediumistic communication or apparitional experience to the subject. But evidence of such a motive is not always available, and we should not assume that one exists in the absence of such evidence. (1984, p. 159) The proper reply to this has two parts: first, that we should not assume such evidence is absent unless we look for it, and second, that hardly anyone looks for it, except in the naive way we expect questionnaires or casual conversations to reveal the deepest secrets of one's soul, or the half-hearted or superficial way searches are usually conducted by the lazy or frightened. If the motives in ques- tion exist, they are unlikely to reveal themselves to the sorts of surface investiga- tions Stevenson and others conduct. Without an extensive and penetrating exam- ination into the lives of clearly relevant (and perhaps even seemingly peripheral) personnel, we are simply in no position to reject explanations in terms of moti- vated super-psi. Some might feel that this criticism is unfair. After all, our goals and interests are often unconscious and difficult to discern, and in actual case investigations we may have no real prospect of ferreting out potentially relevant deep needs and concerns. That is especially true for the older cases, where we are no longer able to interrogate medium and sitters. Now I grant that in many (if not most) cases, we may never get a clear feel for the pertinent underlying psychodynam- ics, no matter how hard we try. But that is no reason for not trying, and often we do not have to probe very far to glimpse some of the significant psychological activity simmering beneath the surface. Indeed, some case studies reveal clearly how much we stand to learn from depth-psychological detective work. Probably the best example is Eisenbud's brilliant analysis of Mrs. Chenoweth's Cagliostro persona (1983). The case is complex and deserves to be read in its entirety. But in a nutshell, here is what it is about. In 1914 James Hyslop held a series of sittings with one of his favorite medi- ums, Mrs. Chenoweth. Also present was Doris Fischer, whom W.F. Prince described exhaustively in his monumental study of her multiple personalities (W.F. Prince, 1915116) and in whom Hyslop was interested because of his suspi- cion that multiple personality might be a disguised form of mediumship (Hys- lop, 1917). One of the most interesting of Mrs. Chenoweth's trance personalities was a "drop-in" who emerged on several occasions over the series of sittings. He claimed to be Count Alessandro Cagliostro, the notorious eighteenth-century mystic, healer, and (as some have alleged) con-artist. His behavior at the sittings was vivid and flamboyantly salacious, but nevertheless rather one-dimensional. Cagliostro came through as a vigorous defender of sexual freedom, including that of women, and as a severe critic of Christianity-indeed, "as a reckless blasphemer who wouldn't have lasted forty-eight hours in the Church-dominated Europe of the time" (p. 230). More importantly, however, the behavior of this 134 S. E. Braude trance personality corresponded to nothing in any of the accounts of his life pub- lished up to the time of the sittings. The real Cagliostro was arrested in Rome in 1789 and brought to trial by the Holy Inquisition. Charged with freemasonry, heresy, and promulgating magic and superstition, he was condemned to death. That sentence was later reduced to life imprisonment. But it was not until 1972 that an account of Cagliostro appeared which presented the Vatican's version of the trial (the account was translated into English in 1974). That characterization was based on a digest of charges obtained in 1855 by the National Victor Emmanuel Library of Rome. The original was still held in secret by the Vatican. For various reasons surveyed by Eisenbud, it is highly unlikely that any of the sitters had normal access to the material in the National Library, and Eisenbud's sleuthing turned up no publica- tions citing the Vatican's version prior to 1972. But again, perhaps the most striking feature of the trance persona was that it did not correspond to the picture of Cagliostro painted by all the reliable sources available not only at the time of the sittings but until the present time as well. Not even critical accounts of Cagliostro accused him of being lascivious or reli- giously cynical. Indeed, there is reason to think that Cagliostro's trial was rigged, and that it was simply expedient for the Vatican to charge him with blas- phemy and rampaging licentiousness. Hence, there is no good reason to regard this case as presenting evidence for the survival of Cagliostro. But in that case, what was the function, psychody- namically speaking, of the Cagliostro persona? Why should a colorful but histor- ically inaccurate trance personality emerge who was so flagrantly sexual and religiously cynical? Eisenbud offers numerous intriguing reasons for thinking that the Cagliostro persona had a great deal to do with, among other things, the sitters' sexual repressions and religious upbringing. For example, Hyslop, who "apparently devoted much of his life to spiritual and moral development7' (pp. 2334), predictably found the Count to be a deplorable figure. Hence, it seems both interesting and significant that, by his own admission, Hyslop repeatedly encountered non-spiritual "sensuous" charac- ters in sittings he conducted. Moreover, Mrs. Chenoweth displayed a surprising- ly intense attachment to the Count when it looked as if other communicators might banish him from the scene. Crying to the other ostensible communicators who tried to exorcise him, she said, "You give him back [Pause] You give him back . . . Give the Count back to me." Hyslop asked who wanted the Count, and Mrs. Chenoweth replied, "We all do. We are lost. We are lost, we are lost, we are lost [Pause] Oh, Devils, to take him away from us. [Distress and crying] . . . I won't stand it [Pause] I don't want your old God. I want the Count." Further- more, Doris (like Hyslop) was a model of moral propriety. In fact, she seemed almost to be a caricature of naive virtue. According to W.F. Prince, "A purer, more guileless soul it was never the writer's good fortune to know." Prince also notes that Doris had a "notable lack of sex-instinct." Now it apparently never occurred to Prince (or, apparently, Hyslop) that Doris' lack of sexuality may have indicated an inhibition of powerful sexual Survival or Super-psi? 135 desires. And in 1914 nobody considered seriously the possibility of experi- menter-influence, or more generally the possibility that persons other than the medium might play an active role in shaping the material presented by a medi- um. Eisenbud proposes, for these and many other reasons, that the Cagliostro persona might have been a composite "dream figure omnibus for the repressed unconscious hankerings of all the principals at the sittings" (pp. 237-8). And considering some of the startlingly close correspondences between the trance personality and the behavior attributed to Cagliostro by the Church, it appears as if one or more of Mrs. Chenoweth, Hyslop, and Doris Fischer psychically raided an extremely obscure portrayal of Count Cagliostro to provide some material for the sitting. For present purposes, it does not matter particularly whether Eisenbud's psy- choanalytic conjectures are correct. What matters is the level at which he attempts to evaluate the data. Still, Eisenbud's proposals do make very good sense of the evidence, including peculiar and otherwise unexplained bits of behavior on the part of the sitters. Hence, his analysis demonstrates clearly the potential benefits of depth-psychological probings. And, of course, since the Cagliostro persona corresponds only to a false characterization of the Count apparently cooked up by Vatican officials, the case is not even a remotely plausi- ble candidate for the survival hypothesis. On the contrary, it strongly suggests the operation of a high level of dramatic creativity and some pretty dandy psy- chic functioning. But with this sort of evidence staring us in the face, it is both presumptuous and naive to rule out super-psi conjectures in cases where no comparable depth-psychological study has been conducted. Regrettably, howev- er, by comparison to Eisenbud's standard of analysis, most other case investiga- tions are unacceptably superficial. Consider, for example, Stevenson's treatment of the Sharada case of ostensi- ble reincarnation (Stevenson, 1984). Since the most striking feature of this case is the evidence it offers for responsive xenoglossy-hence, the persistence of an ability or form of knowledge how, it does not exemplify the sort of case we are currently considering. On the other hand, since we are presently focusing on the need for depth-psychological investigation in connection with the evidence for survival, discussion of the Sharada case is quite apt. The subject, a woman named Uttara, began to have apparent memories of an earlier life when she was in her thirties. These memories occurred during dramatic changes in Uttara's personality, at which time she spoke a language (Bengali) which she apparently neither spoke nor understood in her normal state, and which she apparently never had an opportunity to learn. Stevenson and his associates did a great deal of valuable and careful inves- tigative work to rule out explanations in terms of normal processes. Most of that investigation was devoted to uncovering the extent to which Uttara might have learned normally about Bengali history and customs, and whether she was ever exposed to the Bengali language in a way that would explain her apparent facili- ty in speaking it. But despite all the detail Stevenson provides, the reader gets no feel whatever for Uttara and other relevant individuals as persons. We have no 136 S. E. Braude idea what moved them or what their needs and desires were. We get no sense of the profound personal issues that shaped their lives and actions. In fact, it is quite remarkable how little effort Stevenson apparently made to dig beneath the surface of their concerns, either in the actual course of investigation or in his subsequent evaluation of the case material. A couple of examples should make this clear. One of the most glaring con- cerns Uttara's relationship with a homeopathic physician which began prior to Sharada's first appearance. The physician, Dr. Joshi (a pseudonym), had been treating Uttara for a variety of physical ailments. For several years he treated her as an outpatient, and then (since her condition was not improving) he admitted Uttara to his private hospital. Uttara clearly felt strongly attracted to Dr. Joshi, and at times she behaved toward him like a jealous spouse rather than a patient. Eventually, Uttara's behavior became so annoying that Dr. Joshi had her removed from the hospital. Moreover, when Sharada finally appeared she claimed that the doctor was her husband from her "previous" life in Bengal. Although Stevenson notes that Uttara's relationship with Dr. Joshi may have contributed to the appearance of Sharada, he has strikingly little to say about the nature of Uttara's interest in the doctor, or in the doctor's reaction to her affec- tion. This is especially regrettable, since there are good reasons to think that there is more to their relationship than meets the eye. For example, Stevenson claims that Uttara had been "strangely moved (1984, p. 105) when she first met Dr. Joshi. Now to put it bluntly, Stevenson offers no reason to think that there was anything strange about it, and in the absence of serious probing into Shara- da's feelings toward the doctor in particular and toward men in general, that is a surprising choice of words. Indeed, it is out of character in a prose style that oth- erwise aims at being quite neutral (it may effectively reveal Stevenson's antecedent inclination to treat the case as indicative of reincarnation). Perhaps Uttara's feelings could justifiably be termed strange if they were really those of a deceased individual who had not yet clearly manifested in Uttara's behavior. But Stevenson has given us no reason to think they were anything more than the common sort of attraction one feels suddenly for another, which we often later leam has quite pedestrian origins in our various hidden needs and agendas. Perhaps most notably, Stevenson tells us that despite Uttara's annoying dis- plays of affection, Dr. Joshi visited her at home "a few times" (p. 105) after she had been discharged from his hospital. Stevenson attributes these visits to Dr. Joshi7s "interest and perhaps compassion," but he says the doctor "indicated no deeper attachment to either Uttara or Sharada" (p. 105). But Stevenson also notes that Dr. Joshi was evasive and unrevealing during attempts to examine the nature of his relationship with Uttara. The only explanation Stevenson offers for that evasiveness was that the doctor found Sharada's attentions embar- rassing. But that explanation is hardly compelling, and Stevenson offers nothing to support it in the face of rather obvious sorts of doubts. To begin with, why should Uttara's affection and attention toward the doctor be embarrassing? Patients often fall in love with their doctors. Why would Dr. Joshi not simply