Is Xenoglossy Evidence of Survival?

Occasionally in reincarnation and mediumship cases, the subject speaks words in a language he or she had never learned and could have no normal means of knowing – a phenomenon called ‘xenoglossy’. Some researchers of the phenomenon have considered such cases to be good evidence for survival. Philosopher Stephen Braude argues that this view is too simplistic, and that the ability to utter words in an unknown language, while certainly exceptional, is possible in certain non-paranormal circumstances.

This article is adapted from Braude’s book Immortal Remains (2003). A description of reincarnation xenoglossy cases can be found here.


A common belief about survival cases, reincarnation cases in particular, is that subjects sometimes display abilities characteristic of the deceased. In other words, those subjects display what is often called ‘knowledge-how’, as opposed to mere ‘knowledge-that’, that is, propositional knowledge or knowledge of facts or pieces of information. And one of the most dramatic abilities apparently manifested in these cases is the ability to speak responsively in an unknown language. That ability is called responsive xenoglossy.

However, the body of evidence for responsive xenoglossy is quite controversial, for several reasons. First, critics can argue, without much difficulty, that the linguistic competence displayed is not as impressive as survivalists maintain. Typically, subjects do not even speak responsively in complete sentences. Moreover, proponents of a survivalist interpretation of xenoglossy uniformly ignore a wide-ranging body of evidence demonstrating that the relevant empirical landscape is extremely complex, and complex in ways that pose major obstacles to the survivalist.

In fact, to make a compelling case for survival based on the evidence for responsive xenoglossy, one must first examine and be clear about the relevance of the remarkable abilities of savants, prodigies and dissociative virtuosos (for instance, in cases of multiple personality), all of whom can display abilities – sometimes quite astonishing abilities – in the absence of any practice or training beforehand, and sometimes despite physical or mental handicaps which we would ordinarily expect to rule out any possibility of developing the ability in question. The relevant bodies of evidence here are typically ignored altogether – or at most treated quite superficially – by writers on survival. Similarly, one could argue that discussions of evidence for genuine xenoglossy rest on superficial treatments of the specific notion of linguistic competence and also the more general notion of a human ability. Thus, critics could argue that the evidence can more plausibly be construed as manifestations of abnormal but well-documented forms of living human creativity, as well as the notoriously hard-to-dismiss living-agent psi.

This article will survey both the evidence for xenoglossy – some of it undeniably intriguing – as well as the arguments pro and con. However, before examining particular cases, let us first consider the conceptual and empirical background against which any putative evidence for xenoglossy must be assessed.

The Challenge of Knowledge-How

We should first consider a general, and seemingly under-appreciated, background issue, concerning any form of knowledge-how displayed in survival cases – not just linguistic abilities, but, say, musical or artistic abilities as well. Even when writers on survival concede that it is difficult to rule out, a priori, explanations of the evidence in terms of living-agent psi (LAP), they typically maintain that the appeal to LAP works best – or only – in cases of anomalous knowledge-that. In other words, even if we grant that people can para­normal­ly ‘perceive’ or acquire remote bits of information, many argue that it is implausi­ble to suppose that one can paranormally acquire someone’s abilities or skills, especially in forms that are as idiosyncratic as a person’s finger­prints.

Fortunately, in order to evaluate this position, we may ignore the ques­tions arising in connection with the persistence of a deceased person’s idiosyncratic abilities, such as a distinctively quirky sense of humor or highly specialized technical expertise. We may safely focus instead on more general abilities, such as the ability to write or speak in a foreign language, play a musical instrument, compose music, discuss theoretical physics, or solve mathematical problems, never mind the singular forms the abili­ties might take. If a non-survivalist hypothesis cannot account for these general competencies, we needn’t worry about more highly specialized forms.

Now the general line of reasoning behind the position mentioned above is this. Mere information or propositional knowledge is the sort of thing we can acquire simply through a process of communication, normal or paranormal. But skills, such as playing a musical instrument or speaking a language, cannot be accounted for in this way. Granted, obtaining information is often a necessary part of skill development, but it is hardly sufficient. That is because skills are things people develop only after a period of practice. But since the subjects in survival cases who display anom­alous skills have had no opportunity to practice them first, it is reasonable to reject explanations in terms of living-agent psi and resort to survival­ist explanations instead.

Although this argument is not outlandish, it is nevertheless highly problematical. In one of his discussions, Ian Stevenson claimed that the above line of reasoning rests on the usually tacit princi­ple that ‘if skills are incommunicable nor­mally ... they are also incommunicable paranormally’.1 According to Stevenson, it was philosopher CJ Ducasse who first applied this prin­ciple to the evidence for survival,2 and Stevenson appar­ently considered it to be self-evident, or at least not worthy of a defense.

However, it is not clear that the argument above does rest on this prin­ciple, because it is not clear that communicating skills is at issue. All we know is that some individuals manifest anomalous abilities. How they got them remains a mystery, and, as we will observe below, non-survival­ist explana­tions need not appeal to a process of transmission or communication. In any case, Du­casse’s principle is not nearly as obvious as Steven­son sug­gests, and if it is really an essential step in the survival­ist argument, it may be more a liability than a virtue.

Consider: if Ducasse’s principle is true, that is not because it is an in­stance of the more general principle (K): If any bit of knowledge x is incommunicable normally, then x is incommunicable paranormally. No matter whether principle K is supposed to be a conceptual truth or an empirical generalization, it must be judged false by anyone who accepts the evi­dence for ESP. In ESP people can have access to data which at that time is inaccessible through all known channels of information. So if we ac­cepted principle K, we would have to conclude that ESP is impossible. There­fore, it is reasonable to assume that Stevenson and others do not accept this more general principle.

But then if Ducasse’s principle is true, presumably it is true only of skills, not propositional knowledge or knowledge-that. But why think it holds even for skills? There are at least four crucial topics to consider here: (1) the extent to which we can express and develop skills by side­stepping our customary resistance-laden modes of cognition; (2) wheth­er it is question-begging in this context to talk of acquiring skills; (3) the relationship between skills and practice; and (4) the difficulty in generalizing about skills or abilities, including the ability to speak a lan­guage. These different issues overlap considerably, as we will see.

Survival and Human Abilities

Consider, first, the sorts of things that can interfere with skill devel­opment, even when we have opportunities to practice. For one thing, when we learn a new skill we usually do a certain amount of unlearning, if only of acquired motor and cognitive habits which would interfere with manifesting that skill. For example, a piano student might have to unlearn long-ingrained habits of fingering and pedaling in order to advance to the higher level of expertise required by a difficult new piece. Moreover, learning of any kind, whether of skills or information, is often highly resist­ance-laden. It can be hampered by an endless number of interfering beliefs, insecu­rities, and other fears.

One might think that these barriers to learning a new skill only strengthen the survivalist position. After all, they only increase the num­ber of challenges facing a medium or apparent reincarnation subject who manifests a deceased person’s skills without benefit of practice. However, these physical, cogni­tive, and emotional obstacles can be overcome relatively easily in hyp­notic or other profoundly altered states. For example, under the influ­ence of stage hypnotists, good hypnotic subjects do things they have never done before – for example, dance the tango, accurately imitate their boss or various farm animals, behave in an overtly seductive manner – and more relevantly, display dramatic and creative abilities they might otherwise be too inhibited to express.

More generally, it is plausible that manifesting a skill might be facilitated if the process bypasses the normal states in which our inhibi­tions and other constraints are strongest. We know that people can ex­hibit unexpected abilities or perform at unexpectedly high levels under certain unusual conditions. What is at issue is whether subjects in sur­vival cases fall into this class – that is, whether they find them­selves in situations that are conducive to surprising levels of performance or the manifestation of surprising skills.

In order to resolve that issue, one must consider not simply spo­radic instances of dramatic dissociation such as the performances elic­ited from good hypnotic subjects by experimenters or stage hypnotists. It is important also to consider recurrent or chronic forms of dissociation, especially multiple personality/dissociative identity disorder (MPD/DID). It seems clear that dissociation facilitates the emer­gence and devel­opment of personality traits and skills which might never be cultivated or displayed under normal conditions. For example, alternate personalities/identities exhibit behavioral and cognitive styles which are not explainable simply in terms of propositional knowledge, and which we would have judged highly unlikely in light of the multi­ple’s previously observed repertoire of skills and level of achievement.3 In addition to changes in handedness and handwriting, an alter’s cognitive style may encompass, for example, artistic and literary ability, mechanical apti­tude, and the skills of drawing, sculpting, and writing poetry. But be­cause alters appear suddenly and sometimes evolve quickly, their dis­tinctive skills apparently emerge with­out any practice.

So it seems clear that good dissocia­tors, at the very least, can develop or manifest novel abilities and skills without benefit of practice or a normal process of learning. But then, since en­tranced mediums gener­ally and at least some subjects in CORT cases probably experience periods of dissociation, we are hardly in a position to assert the improbability, much less the im­possibility, of a person suddenly manifesting new or latent abilities or skills in cases sug­gesting survival.

Another possibly deeper set of problems concerns the way even so­phisticated writers on survival such as Ian Stevenson and Alan Gauld generalize about skills. For example, Stevenson asserts, ‘Practice does not just make perfect; it is indispensable for the acquisi­tion of any skill’.4 There are at least two related problems with that claim. First, skills differ dramatically in many respects, one of which is the importance of practice in skill development. ­The second problem is that the acquisition of skills may not be the issue. All we are entitled to discuss, strict­ly speaking, is the manifestation of skills. We have no idea wheth­er or to what extent new skills have been acquired by mediums or by subjects of rein­carnation investigations. This is not a triv­ial distinction, because although practice seems essential to perfecting a skill, it is not always needed to manifest skills for the first time.

To see that, we need only to consider child prodigies and cases of savantism. In fact, musical prodigies such as Mozart, Mendelssohn, and Schubert, and mathemat­ical prodigies such as Gauss, usually manifest exceptional skills prior to perfecting or developing them through prac­tice. It is not simply that prodigious skills are rudimentary at first and then evolve with amazing rapidity. The skills of prodigies can be amaz­ing even at the beginning. The same is true of savants. For example, one fascinating musical savant was reportedly able to read music without ever receiving instruction. She was also able to improvise in the styles of various composers at the piano the first time this feat had ever been requested. In fact, she found that she could play in different composers’ styles at the same time, the right hand playing in one style and the left hand playing in the other.5 But more important, we have no reason to think that the subjects in survival cases demonstrate levels of expertise more im­pressive than, say, Mendelssohn’s initial displays of musician­ship. Quite the contrary: the suddenly emerging skills of savants and child prodigies often far exceed anything displayed by the subjects investigated in xenoglos­sy cases or other cases suggest­ing survival. But then, we must concede that we don’t know to what extent certain conditions such as dissocia­tion may unleash impressive, if not prodigious, capacities latent in many or all of us. This last observation will be amplified further when we consider, below, the related but non-xenoglossy cases of Patience Worth and Hélène Smith.

Of course, we don’t need to consider prodigies and savants to appre­ciate this point. Ordinary people demonstrate it all the time. Consider, for example, the skill of play­ing tennis. Many people are naturally athletic, even though they may not be prodigiously gifted. And to the occasional consternation of those who are more athletically challenged, natural athletes can, on their first try, play a game of tennis reasonably well – or at least without looking hopelessly foolish. In fact, on their first try they might even play as well or better than others who have played for years and taken lessons. But more important, the natural athlete’s beginning level of tennis skill would arguably be at least as good, and quite possibly bet­ter, than the language skills exhibited in vast majority of responsive xenoglossy cases, as we will see below.

Behind these various considerations lurks a more sweeping problem, one that stands in the way of ever concluding confidently that mastering one skill is more difficult than mastering another. First of all, we have, at best, only a rudimentary understanding of what skills are. For exam­ple, we don’t know whether the various things we call skills are similar enough to permit useful generalizations. In fact, we don’t even know to what extent we can generalize about individual skills.6 The things we iden­tify as specific skills – for example, the skill of speaking a language, or com­posing music – typically consist of other skills and capacities. But those subsidiary skills and capacities, may also be organized collections of other skills and capacities, and at no point along the way is there some preferred set or ar­rangement of lower-level endowments necessary for exhibiting the more general capacity.

For example, people who can compose music have various other musi­cal abilities that make their skill in composition possible. But composition­al skill can be expressed in a great variety of ways. Many composers notate their compositions; others lack that ability. Some com­posers have absolute pitch, some only good relative pitch, some nei­ther. Some composers can compose directly onto paper without the aid of a piano or some other instrument; many others cannot. Some compos­ers work well with large forms; others do not. Most composers write par­ticularly well or idiomatically only for certain instruments, and only some composers demonstrate a keen ability to set words to music. Some composers are particularly skilled in harmony, rhythm, melody, or in­strumental color, but those secondary skills take different forms and exist in different degrees and combinations with different composers. So there is no reason to assume that the skill of musical composition will allow many useful generalizations. And perhaps most important, there is no reason to think that this is a unique feature of that particular skill. Indeed, it seems to be the rule rather than exception. People who possess a gen­eral skill may exhibit it in various ways and to varying degrees, depend­ing on which subsidiary skills they possess and the manner in which they possess them. Presumably, then, there is no reason to assume that what we identify as a skill enjoys a deeper theoretical unity.

But in that case, certain familiar arguments in the survival literature seem fatally simplistic. For example, when Stevenson argues that skills cannot be communicated or manifested without practice, he men­tions riding a bicycle, dancing, and speaking a foreign language as ex­amples. Similarly, Gauld writes,

The ability to play bridge well is not simply a matter of learning (whether normally or by ESP) the rules (considered as a set of facts together with the precepts given in some manual). It can only be acquired by practising intelli­gently until things fall into place. And it is the same with learning a language.7

However, if there are serious disanalogies between linguistic competence and these other skills, they may be deep enough to prevent us from general­izing usefully across abilities. And if we cannot say how difficult or easy it is, generally, to learn or de­velop a new skill, including learning a new language, then this sort of survivalist argument is simply a non-starter.

Xenoglossy and Linguistic Competence

To pursue the issues further, let us consider some aspects of language learning. Language use, like musical composition, encompasses a vari­ety of other capacities and manifests in quite different sorts of contexts. Not surprisingly, then, we cannot say, in general, how difficult it is to learn a new language. The de­gree of difficulty seems to depend on many things, including linguistic aptitude, a good ‘ear’, the context in which the language is learned, and how different the language is from one’s native tongue. The key issues here are explored in depth in an interesting and sensible review of the research on second-language acquisition by Bialystok and Hakuta, which should probably be required reading for students of survival.8

Bialystok and Hakuta observe, first of all, that new lan­guages are learned in many different contexts, some more demanding than others, and some more conducive than others to general linguistic proficiency. For example, the wife of an Amer­ican businessman in Japan might, with the help of her new Japanese neighbors, learn enough Japanese to do her shop­ping and banking and also to have conversations over tea. De­spite her many grammatical errors, she speaks well enough to be under­stood. The family’s children might seem to have a good working knowl­edge of Japanese, despite attending an Amer­ican school. They can play with their Japanese neighbors, discuss comic books, order their own food in restaurants, and do these things in ways that strike their father, who has not learned the language, as very fluent. A Vietnamese carpen­ter’s assistant in Toronto might learn English in a way suitable to his work needs. So he might learn the appropriate technical terms, and pos­sibly even words unknown to many native speakers of English. But his command of English may be largely receptive; he may know the lan­guage well enough to understand orders and carry them out, but not well enough to assume responsibilities as a foreman. A student at Yale might earn top marks in a course in Russian, know his vocabulary and rules of grammar, but be unable to converse with or even understand a Russian exchange student he meets.

Although each of these scenarios is a legitimate and familiar example of second-language learning, the language users have acquired different skills. 

The American businessman’s children learned conversational skills adap­ted for interaction with other children in play situations. The mother learned a mode of speech particular to the interaction of women in Japan and the kind of conversation used in shopping. The Vietnamese carpenter mostly learned receptive vocabulary specialized to the routines of his daily work. The Yale student learned a lot of grammar and vocabulary.9

Bialystok & Hakuta conclude, 

When we learn a new language, we invariably gain exposure to that lan­guage in a more limited range of contexts than those in which we regularly use our first language ... Therefore, the aspects of language proficiency that we need to master or even have the opportunity to learn depend on the par­ticulars of these circumstances.10

They also note that there is no clear, single, or privileged standard of linguistic proficiency. What counts as linguistic proficiency varies wide­ly with context. And they observe that there is no single set of abilities in virtue of which people are able to learn and speak a new language. Different people draw on different aptitudes and skills, which they pos­sess in distinctive combinations and in varying degrees.

Along the same lines – and of particular relevance in this context, Bialystok and Hakuta also caution against general­izing about the difficulties of learning a new language. In most cases we cannot make reliable predictions on the basis of age, personality, or apti­tude. Out of what they call ‘methodological necessity’, formal studies of second-language acquisition focus on only a small range of contexts in which people learn a new language. In fact, the usual ‘tests of overall ability seem to correspond only to classroom success’.11 But once we look beyond the ‘limited circumstances that provided the data for study’,12 we find, as we might expect, that 

not all people learn well in all instructional (treatment) settings, and any given learning situation may be good for some people, but not for others. Putting learners who are highly successful in one context into a situation that requires a different set of skills could well reveal the limitations in that learner’s achievement.13

Nevertheless, Bialystok and Hakuta hazard a few generalizations about situations either conducive or resistant to second-language acqui­sition. And appropriately, they seem to recognize that these are, at best, useful statistical generalizations that can accommodate a variety of ex­ceptions. They suggest that a second language will be most dif­ficult to learn in those respects in which it differs significantly from the first language.14 The problems are most likely to concern grammar or vo­cabulary, but presumably they could also concern the language’s general descriptive categories, and their embedded meta­physics. Moreover, Bialystok and Hakuta concede that ‘exposure to the language and prac­tice in its production seem to be essential to phonological mastery’, and ‘there is no shortcut for learning words. They need to be stud­ied, memorized, encountered, and reflected upon’.15

So, how does all this help us get a handle on xenoglossy generally? First of all, it seems clear that learn­ing a second language is a significantly different process from learn­ing a language for the first time. It also seems clear that many people can easily attain various kinds of minimal second-language competence, even when the new language is grammatically and semantically novel, but especially when the new language isn’t radically different from their own. And since formal tests of linguistic aptitude don’t measure real-life linguistic adaptability, we cannot expect to be enlightened by administer­ing such tests to subjects in xenoglos­sy cases, for instance as Stevenson did in the Jensen case, which we consider below.16 Contrary to what Stevenson claims, it seems false that the best tests measure ‘the ability to learn a modern language easily’.17

Moreover, even if there is no shortcut for learn­ing words, we cannot specify, in gen­eral, how much and what sort of exposure to a new lan­guage is necessary for the kind of low-level linguistic proficiency demonstrated in the vast majority of xenoglossy cases. That seems to vary widely from person to person and context to context. In fact, it seems reasonable to assume that, as in many other areas of life, some people learn much more quickly than others. Given the right combination of needs and natural aptitudes, some people might require only a very brief exposure to elements of a language, while others might require repeated exposure over a long pe­riod. And as cases of MPD/DID demonstrate dramatically, it may only be under very special circumstances that we exceed our ordinary capaci­ties or demonstrate otherwise latent natural gifts.

But in that case, the relatively crude linguistic competence displayed in most cases of xenoglossy may not be all that impressive. The context of an­swering simple questions put to a medium (probably in a dissociated state) seems, in many respects, to be a low-pressure situation, far less demanding – and arguably less likely to elicit our most impressive responses – than real-life social situations in which people must master a new language, and where important personal and professional issues and relationships are at stake. Mediumistic xenoglossy may require little more than some native – and possibly la­tent – linguistic aptitude, and also rudimentary knowledge-that of vo­cabulary and grammar, at least some of which could be acquired paranormal­ly. In fact, since we are considering exotic explanations, we cannot rule out the possibility that subjects gain the needed exposure to the new language unconsciously and psychically. And of course, if those subjects have a knack for this sort of thing, they might be able to learn a surprising amount with only the most meager information.


Consider, for example, the Gretchen case,18 in which a Methodist minister, CJ, hypnotized his wife DJ in order to help relieve her pain from a backache. When CJ then asked his wife if her back hurt, to his surprise she responded, ‘Nein’. In another hypnosis session soon thereafter, his wife said ‘Ich bin Gretchen’. Then in follow-up sessions over the next few months, the hypnotized DJ introduced at least 237 German words before they were spoken to her, and 120 of these were uttered by ‘Gretch­en’ before any of her in­terlocutors addressed her in German.

However, according to one skeptical estimate, less than 20 percent of Gretchen’s German comments (28 of them) were appropriate to the questions asked in German. Now we can agree with Robert Almeder that a person who can do this in some sense knows German, and that we need to ex­plain ‘how somebody who has never been taught German can success­fully understand the language enough to respond successfully to unre­hearsed questions 28 times’.19 But Almeder claims that this level of proficiency cannot be explained, for example, ‘by appeal to World War II movies or casual glances at German books, because one would need to know what was being said in such movies or books’.20 The prob­lem with this claim is that some people might be able to learn a great deal from such limited material, and it is often not too difficult to discern the meaning of words or phrases in foreign movies, especially if the movies have subtitles.

Still, we can agree that the Gretchen case is in­triguing and perhaps not easily dismissed. But given the complexities, discussed above, in generalizing about second-language acquisition, and considering the actual achievements of good dissociators, savants, and the more common skills of the linguistically gifted, we need to be more circumspect about this case. (This conclusion can only be reinforced by considering the astonishing, although somewhat different, case of Hélène Smith discussed below.)


In the Jensen xenoglossy case, a Philadelphia doctor discovered that his wife TE was a good hypnotic subject who ‘could enter deep trances readily’.21 To explore that capacity further, he began con­ducting hypnotic age-regression experiments on TE, during which she began speaking in Swedish, manifesting a personality named ‘Jensen’. The Jensen persona spoke in a somewhat archaic language, which, together with the details provided about his life, suggested a pre­vious existence in Sweden during the seventeenth century. But this case offered little, if any, non-linguistic evidence that Jensen corresponded to a real previous personality.

Moreover, during the first five sessions, no one present spoke a Scandinavian language, and at those sessions Jensen uttered only occasional Scandinavian-sounding words or phrases. Session four was recorded, and at that session Jensen spoke two phrases that were later ‘identified clearly from the tape recordings’.22 However, once Swedish speakers began attending the sessions, Jensen spoke a great deal of Swedish, or perhaps a mixture of Swedish and Norwegian. (Of course, this is precisely the sort of scenario that raises the spectre of telepathic sitter-influence.) Jensen’s pronunciation and grammar were good, but he ‘rarely responded in full sentences, and when he did his sentences were short’.23 Several Swedish speak­ers listened to the tapes or interviewed Jensen themselves, and they concurred that Jensen introduced words into the conversations that had not previously been used by the interviewers in TE’s presence.

TE’s command of Swedish or Norwegian seems clearly superior to DJ’s mastery of German. But perhaps neither is out­landish for an adult with previously untapped linguistic aptitude, who is a good dissociative subject, and who might have been exposed to elements of those lan­guages unconsciously, and even psychically. Interestingly, Stevenson himself seems to make a crucial concession on this point. Citing a case reported by Dreifuss,24 he says that it shows ‘that an ability to speak intelligibly (not merely to recite) a foreign language may remain dormant and emerge later in life’.25 Granted, the Sharada reincarnation case (which we examine below) confronts us with a kind of linguistic fluency far beyond that demonstrated in the Jensen or Gretchen cases. But­ we should observe now that the linguistic competence of the ostensible previous personality, Sharada, would have been more of a feat had her language been radically different from that of the subject. Moreover, the subject had already demonstrated a facility in learning new languages, and, perhaps most important, she already had learned some Bengali—the language of the previous personality.

Apparently, then, the research on second-language acquisition tends to undermine, rather than support, the survivalist position. It dis­credits the usual, rather glib, survivalist generalizations about linguistic proficiency and second-language acquisition; it reinforces the commonsense view that some people can accomplish a great deal with relatively little effort, input, or support; and it reminds us that impressive, if not prodigious, abilities may lurk under nearly anyone’s surface, awaiting an appropriately fertile ground for expression. That is not to say that we can clearly or justifiably reject a survival­ist interpretation of the good xenoglossy cases. But one can easily argue that the evidence is not nearly as persuasive as some seem to think.

Empirical Digressions

A similarly cautious attitude seems appropriate to another type of case. Strictly speaking, cases in this group would not qualify as examples of xenoglos­sy, but they raise issues analogous to those we are currently considering. First of all, the parapsychological literature contains scattered reports of children producing automatic scripts, even though they haven’t yet learned the alphabet. For example, FWH Myers mentions two cases.26 The first concerns a girl of five who wrote a few words, ostensibly in a lady’s (not a child’s) handwriting. But the case is poorly described, and many details cry out for further explana­tion and probing. Quite apart from the issues, noted above, of latent abilities, cryptomne­sia, and psychic influence from persons nearby, it is interesting that the girl had been watching her older sister produce au­tomatic writing. So first, we need to know more about the possibly com­petitive relationship between the sisters. And second, we need to know whether the young sister’s family life was such that she might have picked up rudiments of writing by age five, independently of formal in­struction, simply through exposure to the usual activities of normally literate people. That would require only a modest degree of linguistic precocity.

In the second case, a girl of four who had never been taught the alphabet or even how to hold a pencil, scribbled ‘your Aunt Emma’, a tracing of which Richard Hodgson described as ‘resembling the planchette-writing of an adult rather than the first effort of a child’.27 This case, too, is not thoroughly described. It may be that the child had never at­tended school or been taught the alphabet. But it is unlikely she had never seen a written word or observed the act of writing. And we are given no information about her visual acuity, manual dexterity, and abil­ity to draw or copy what she had seen. So we have no idea what the child might have learned or accomplished on her own. In fact, in both cases, it would be good to know how quickly the girls exhibited linguis­tic mastery once formal instruction began. Moreover, Hodgson’s state­ment in the second case is misleading. The four year old’s writing, the production of which he observed, was the last of several attempts to write the name ‘Emma’. So, as so often happens in connection with cases suggesting postmortem survival, those cases are considerably under-described, particularly in connection with psychodynamics relevant to evaluating LAP alternative interpretations.28

It is also worth noting some additional puzzling features of xenoglossy cases, which in turn lead us to well-documented concerns about hypnotic regression.

In his second book on xenoglossy, Stevenson dis­cussed several potentially important, and certainly intriguing, linguistic features of xenoglossy.29 One feature is that when communi­cators respond in their native language to questions posed to them, the questions are not always in those languages. Sometimes, the language is that of the person asking the question, and in some of those cases this different language is one that the communicator was not supposed to know. For example, the trance personalities Jensen and Gretch­en re­sponded in their alleged native languages to questions posed in English. Along the same lines, some communicators speak their alleged native languages with the distinctive accent of someone whose original tongue is that of the medium, and occasionally they speak with the characteris­tically stilted or botched grammar of someone trying to master a second language.

We can agree with Stevenson that these features are fascinating. But perhaps Stevenson did not ask the right questions about them, or perhaps he did not ask enough of the important questions. Stevenson seemed con­cerned only with making sense of these phenomena on the assumption that the communicators or trance personalities are what they purport to be. Now granted, we cannot seriously entertain the survival hypothesis unless we address that issue squarely. And that means we must con­sider, as sympathetically as we can, (a) what the experience of commu­nicating might be like from the communicator’s point of view, (b) to what extent there might be translation problems between different lan­guages, and (c) what other factors might aid or hinder the process of communication. But those topics need to be discussed as part of a broad­er inquiry. After all, Stevenson (and, of course, others) consider xenoglossy cases in or­der to decide whether there is any compelling evidence of survival. Un­der the circumstances, then, the more fundamental question is: Do the peculiar linguistic phenomena discussed by Stevenson make more sense from the survivalist or non-survival­ist point of view? And the reason that question is especially important is that those features of xenoglossy seem strongly to support a non-survivalist interpretation.

Stevenson imagined that the hypnotic subjects, TE and DJ, en­gaged in what he called ‘layering’, a kind of subsurface interaction be­tween those subjects and the discarnate minds or personalities of Jensen and Gretchen, respectively. He proposed (a) that interlocutors’ English words evoke certain images (or other causally efficacious mental states) in the subjects’ minds, and (b) that those states then trigger appropriate mental states in, and eventually verbal responses from, the communica­tors. Of course, it is far from clear how (a) would work, if the languages of the previous personality and the subject are different. Translations cannot be automatic, and meanings are not abstract Platonic entities. So would we have to posit a translation ‘program’ to make the appropriate transfor­mations? And if so, where would it come from?

In any case, if the com­municators are what they purport to be, and so long as the apparently required translation process is possible, which is doubtful,30 we must concede that something similar to what Steven­son proposes might in fact occur. So perhaps Stevenson has at least speci­fied a process in what we could call ‘logical space’. But we need to con­sider whether there is any reason to regard Stevenson’s proposed process as actual, rather than mere­ly possible. And curiously, Stevenson illus­trates the process by citing research on a dissociative phenome­non – that is, a phenomenon we can naturally and fairly easily explain in terms of just one living agent. But that suggests strongly that Steven­son’s proposed survivalist explanation is gratuitous.

Stevenson mentions a case of hypnotic age regression reported by Spiegel and Spiegel.31 The subject was a twenty-five-year-old man who learned English only after emigrating from Austria at age thirteen. When he was regressed to any age younger than thirteen, he could apparently speak no English and required the hypnotist to communicate through a German-speaking interpreter. How­ever, the subject was still able to respond correctly to some instruc­tions given in English, even though he had been regressed to age ten.

Sim­ilarly, Martin Orne reports that a subject 

who spoke only German at age six and who was age regressed to that time answered when asked whether he can understand English, ‘Nein’. When this question was rephrased to him 10 times in English, he indicated each time in German that he was unable to comprehend English, explaining in childlike German such details as that his parents speak English in order that he not understand. While professing his inability to comprehend En­glish, he continued responding appropriately in German to the hypnotist’s complex English questions.32

Contrary to what Stevenson seemed to think, these examples pose a problem for the survivalist interpretation of xenoglossy, because they seem to show clearly that the subjects did not actually regress to a previ­ous stage in their mental life. Instead, their behavior seems obviously to be ­dissocia­tive and to presuppose their knowing both German and En­glish. In fact, the subjects’ behavior seems continuous with similar be­havior reported frequently throughout the history of hypnosis. Hypnotic suggestion can produce significant alterations in the experience and thought processes of good hypnotic subjects, and often those changes make it difficult for the subjects to comply with apparently simple re­quests, for example, pronouncing words containing the letter ‘r’. Moreover, as researchers in hypnosis and dissociation are well aware, although disso­ciation can erect perceptual or cognitive barriers, those barriers tend to affect only some aspects or levels of a person’s awareness and perfor­mance. In fact, researchers know that dissociated systems are never com­pletely independent of one another, no matter how isolated they might appear in some contexts.

Apparently, the subjects just mentioned were doing something simi­lar to what has been reported in cases of negative hallucination. For example, subjects in some recent studies were hypnotized not to see the chair in front of them, and good hypnotic subjects behaved differently from those asked merely to simulate hypnosis in the same situations. When the hypnotized subjects were asked to walk around the room, some walked into the chair and evinced surprise that something touched them, and others avoided contact with the chair by walking or stumbling around it.33 When asked to explain their surprise or curious chair-avoid­ance behavior, the subjects seemed genu­inely puzzled and offered transparently lame excuses, frequently de­scribed as examples of ‘trance logic’.

Similar behavior has been re­ported in recent studies of hypnotic blindness and visual conversion disorder, or hysterical blindness, in which subjects seem to be influ­enced by objects or information of which they are apparently unaware.34 These cases raise a number of interesting issues.35 But for now, the crucial point about them is that they are paradigmatic ex­amples of conflicting dissociated systems within a single subject. And as such, they discourage the sorts of survivalist explanations Stevenson of­fered for the analogously odd linguistic features of xenoglossy.

In fact, the examples cited by the Spiegels and Orne resemble an­other famous example of ostensible age regression.36 Orne re­gressed a subject to age six, and in that state the subject’s handwriting changed to the immature style of a young child. Nevertheless, when the experimenter asked the subject to write the sentence ‘I am conducting an experiment which will assess my psychological capacities’, the sub­ject complied exactly, even correctly spelling the polysyllabic words no six-year-old would know. Again, this makes clear sense so long as we assume that the subject is not genuinely regressed, but is drawing on creative capacities perhaps manifested most easily in a hypnotic or dis­sociated state.37

Similar observations apply to the aforementioned peculiarities in com­municators’ accent and grammar. Stevenson proposes an underlying tension between two linguistic systems: the native language of the me­dium or subject and that of the communicator. He writes, 

A secondary personality such as we may call Jensen, Gretchen, and Sharada—who tries to speak his native language, must nevertheless express it through the linguistic apparatus (mental, cerebral, and vocal) of the pri­mary personality. The conflicting pulls of the two different phonemic sys­tems give the impression of a nonnative speaker.38

However, even if we accept the reality of survival, it is not clear why communicators ‘must’ express their native languages through all three of the mental, cerebral, and vocal systems of the host, or why com­municators would be equally dependent on each of them. And in fact, Stevenson noted, with an example from the case of Mrs Leonard, how the tug between aspects of the host and communicator’s linguistic systems seems to vary. But if communicators can be more or less free of the host’s linguistic system, it is unclear why they cannot sometimes be entirely free of it, or free enough for the tug to be negligible. In any case, let us grant tentatively that Stevenson is correct and that communi­cators cannot liberate themselves from the mental, cerebral, and vocal systems of the host. The question still remains: Are the peculiarities in communicators’ accent and grammar most easily viewed as dissociative phenomena?

What Stevenson did not mention is the similarity between these lin­guistic aspects of xenoglossy cases and certain common features of MPD/DID – in particular, a phenomenon often called ‘co-presence’.39 Co-presence is a condition in which alters share execu­tive control of the body. During periods of co-presence, those alters seem to blend or partially integrate, even if only temporarily, so that it is difficult at those times to decide whether alters A and B are two distinct centers of consciousness or just one. And multiples themselves seem to experience co-presence as a state of variable blending. Braude reported a memorable conversa­tion he had with a multiple whose personalities were vying for executive control of the body, and in which she said to him, ‘I’m mostly Karen right now’. So the tug Stevenson postulated between host and communicator linguistic systems resembles the tug and interference between personality or iden­tity systems in MPD/DID, and the latter often exhibits the varying de­grees of blending and interference noted by Stevenson in the Leonard case.

To repeat, we cannot rule out Stevenson’s explanation if we are taking the survival hypothesis seriously. Nevertheless, the phenomena in question seem more parsimoniously explainable in terms of relatively common dissociative processes in a single subject.

The ‘Martian’ Language of Hélène Smith

Earlier, we considered why one should not underestimate the possibility of impressive latent human abilities in xenoglossy cases. To reinforce that point, we need to consider a famous case of automatism – that of Hélène Smith and her ‘Martian’ scripts.40 Here we find a subject who exhibited a novel and quite striking degree of linguistic proficiency and creativity. And unlike the mediumistic and reincarnation cases at issue here, the question of survival, or of genuine com­munication or identification with a deceased individual, simply does not arise. In the present instance, there is no doubt that a remarkable set of automatic produc­tions issued from the subject’s subconscious.

The case in question is extremely complex, and fortunately it has been documented in a careful, penetrating, and detailed book by the Swiss psychologist Théodore Flournoy. The automatist described in the book is Élise Müller (1861–1929), a woman from Gene­va to whom Flournoy assigned the pseudonym Hélène Smith. Smith’s father was a Hungarian merchant with a facility for lan­guage, but apparently his daughter had no comparable proficiency or interest in foreign languages. The linguistic ability manifested during her medium­ship occurred only for a relatively brief period of time.

The case divides neatly into several interesting phases, which Flour­noy described and whose psychogenesis he traced in great detail. Smith’s mediumship began in the winter of 1891–92, and communica­tions took the form of visual and auditory messages, automatic writing, and also table tipping. Her original spirit control claimed to be Victor Hugo, but after about six months he was replaced, following a struggle for supremacy, by a control calling himself Leopold. Flournoy began sitting with Smith in the winter of 1894–95, and to his surprise the medium gave him accurate information about his family life during the period before he was born. Flournoy concluded, therefore, that Hélène’s apparent psychic talents merited further scrutiny. After making friends with Leopold, he persuaded the control to reveal his real iden­tity, which Leopold said was Guiseppe Balsamo – also known as Count Caglios­tro. Evidently Flournoy was the only member of Smith’s circle who did not believe that Leopold was actually a discarnate spirit.

In Flournoy’s view, the Leopold persona originated in a traumatic experience suffered during an attack by a big dog when Hélène was ten years old. In any case, although Leopold’s personality was quite different from Hélène’s, Flour­noy found no evidence of survival in Smith’s control per­sona. Indeed, he noted serious discrepancies between Leopold’s behav­ior and knowledge and what was known about Cagliostro. It remains unclear whether Hélène suffered from a dissociative disorder, much less MPD/DID, and Flournoy wisely resisted that conclusion. Never­theless, Leopold’s origin and function are strikingly similar to those in classic cases of alternate personalities, and those similarities seem to justify Flournoy’s conjecture that Leopold was akin to a secondary personality or self, created through a process of autosuggestion.

What matters for present purposes is that Smith’s mediumship underwent a series of what Flournoy called ‘romances’, the first of which he dubbed the ‘Royal Cycle’. Ac­cording to spirit messages, Hélène was the reincarnation of Marie An­toi­nette, and in trance, Hélène’s behavior as Marie Antoinette was vivid and dramatically appropriate. But as with the Leopold/Cagliostro control, there were persuasive reasons for believing that Hélène’s trance persona was constructed subconsciously by the medium. Nevertheless, Flournoy re­mained open to the possibility that Hélène used ESP to acquire infor­mation incorporated into her séances.

The next phase of Smith’s mediumship, called the ‘Hindu Cy­cle’, began in October 1894. In this romance, Hélène assumed the role of a fifteenth-century Indian princess, Simandini, who was burned alive on her husband’s funeral pyre. During her trance states, Hélène spoke and wrote a kind of ersatz Hindu language, which Flournoy described as ‘a mixture of improvised articulations and of veritable Sanscrit words adap­ted to the situation’.41 Here, too, Flour­noy found good reasons for thinking that Hélène’s medium­ship was a subconscious production. For example, Sanskrit experts con­curred that Indian women did not speak Sanskrit at the time alleged or at any other time, and also that the language spoken in what was ostensi­bly Simandini’s home was the quite different language, Dravidian.42 Nevertheless, Flournoy considered the historical and linguistic knowledge displayed by Hélène in this phase of her medium­ship to be a ‘psychological enigma’. He thought it might at least partially be ex­plained as an example of cryptomnesia, or subcon­scious memory of material learned at an earlier time.

Unquestionably the most important and interesting phase of Hélène’s medium­ship is the ‘Martian Cycle’, which began in November 1894, apparently in response to an inadvertent remark by Professor Lemaître, who had said during a séance that it would be interesting to know about activities on other planets. During this phase of her mediumship, Hélène’s spirit was reportedly transported to Mars. While there, she described the human, animal, and plant life of the planet, and she spoke and wrote fluently in an apparently subconsciously-invented Martian language.

Because Smith’s spirit guide provided word-by-word transla­tions of the written messages, Flournoy and colleagues were able to ex­amine the structure of the language carefully. Like the Hindu ‘lan­guage’ invented earlier, this, too, was more intricate and creative than a collection of sense­less or random phrases. But unlike its predecessor it exhibited a very high degree of syntactic and semantic consistency. The corresponding written alphabet was novel and quite beautifully ornate, and the spoken sound of the language was apparently distinctive as well. However, grammatically and phonetically the language was clearly modeled on Hélène’s native French. Flournoy concluded that ‘the Martian language is only French metamorphosed and carried to a higher diapason’.43 As an intellectual achieve­ment, he considered it to be as ‘infantile and puer­ile’44 as other features of the Martian romance. Nevertheless, Flour­noy regarded the language as an impressive feat of memory and subconscious creativity. Reflecting on the linguistic facility of Hélène’s father, he wrote, ‘the question naturally arises whether in the Martian we are not in the presence of an awakening and momentary display of a hereditary faculty, dormant under the normal personality of Hélène’.45

For present purposes, the important point is that in fluency, thoroughness, and the originality of its written alphabet, Hélène’s Martian language is still something of a creative tour de force. It reminds us to be on the lookout for related sorts of creative eruptions in the altered states of mediums and in the subjects of reincarnation cases.

Of course, there is always the danger, when appealing to creative capaci­ties of a subliminal self, of ‘entering a land of darkness where all analo­gies fail us and where anything may happen’.46 Indeed, we should heed William McDougall’s admonition that ‘the phrase ‘the subliminal self’ may prove detrimental ... if we do not sternly resist the tendency to use it as a mere cloak for our ignorance’.47 (And obviously, the same may be said about easy re­course to the spirit hypothesis.) But the history of psychology provides considerable empirical grounding for speculation about latent capacities. And the history of automatism in particular suggests strongly that ‘the advantage of relegating voluntary ends to automatic execu­tion... [is] getting the needed thing done... with a verve and completeness which conscious effort finds it hard to rival’.48 So the Hélène Smith case seems to help us focus more constructively on the empirical question: What are the limits of our subconscious, latent creativity?


Hopefully fortified by the foregoing considerations, we may now profitably examine what is probably the strongest case of responsive xenoglossy. That is not to say the case is air-tight, and indeed, in its psychological superficiality it is seriously flawed. However, the distinctive relevance of the case lies in the fact that the subject spoke an allegedly unlearned language with remarkable fluency. Moreover, the previous personality, Sharada, made several verified statements about a family that lived at the appropriate time and loca­tion. Nevertheless, the case’s most compelling features are linguistic. Other examples of ostensible reincarnation have offered more impressive – that is, finer-grained and more specific – evidence for the knowledge of a former life.

The psychological shortcomings of the case cannot be examined here in detail, although later they will be mentioned briefly.49 At any rate, it is certainly true that the Sharada case is impressive on its face. The subject in the case, a Marathi-speaking woman named Uttara Huddar, was born in 1941 and lived and worked part-time as a lecturer in public administration in Nagpur, India. At the age of 32 she began to manifest a personality named Sharada,50 who spoke fluent and somewhat archaic Bengali, and who claimed to be and acted as if she were a Ben­gali woman of the early nineteenth century. Sharada claimed to have died at age 22, after a cobra bit her on the toe. When she ‘awoke’ in 1974 she did not recognize Uttara’s family and friends, and she appar­ently did not understand them when they spoke in Marathi, Hindi, or English. (However, she did eventually learn a few words and phrases in Marathi.) Uttara never married, and as Braude explains, she appears to have been profoundly disappointed and frustrat­ed in affairs of the heart. But Sharada dressed and behaved like a married Bengali woman. She spent much of her time in – sometimes old-fash­ioned – Bengali religious prac­tices, and she appeared perplexed by modern ways and somewhat re­pelled by Marathi customs.

When Uttara’s mother was pregnant with Uttara, she often dreamed of being bitten on the toe by a snake. Those dreams ceased when Uttara was born, and her mother claims to have forgotten them until Sharada appeared and mentioned that she had died of a snakebite on the toe. How­ever, the mother’s claim to have forgotten the dream may not be entirely credible. Both parents report that Uttara had a severe phobia of snakes throughout much of her childhood, and that after the age of six­teen her attitude toward snakes changed to one of attraction. So there is reason to believe that the topics of snakes and Uttara’s fear of snakes would have been fairly common in the household, at least until Uttara’s phobia disappeared.

As investigators have recognized, it is important to deter­mine the extent of Uttara’s normal exposure to the Bengali language and to Bengali customs. And initially at least, it looks as if one should be skepti­cal, because there is no doubt that Uttara had studied Bengali and that she had at least a modest ability to read the lan­guage. On the other hand, certain features of the case lend support to a survivalist interpre­tation of the evidence. For one thing, it is not clear whether Uttara had demonstrated the somewhat independent ability to speak Bengali. And for another, Sharada’s spoken Bengali differed in various ways from the modern Bengali Uttara presumably learned while in school.           

Nevertheless, the evidence is not as ‘clean’ as one might like, and so a survivalist explanation of Sharada’s proficiency in Bengali faces serious obstacles. Since Uttara had learned some modern Bengali, it is reasonable to think that it provided a foundation for Shara­da’s proficiency. ­Moreover, as we observed earlier, learning a second lan­guage is a distinct process from learning a language for the first time. And when the second language is not radically different from one’s native tongue – or from a second language one has already learned – the process may be relatively easy, especially for someone proficient in lan­guage. There is no question that Uttara was reasonably sophisticated linguisti­cally and that she had the ability to learn new languages. She spoke English, and she had also stud­ied Sanskrit in high school. In fact, since Sanskrit is the language from which North Indian dialects evolved – just as Spanish, French, and Italian evolved from Latin – Uttara’s proficiency in Bengali does not seem particularly mysterious, if we allow that additional exposure to Bengali could have occurred nor­mally – but unconsciously – and also possibly through ESP. It might also be relevant that approximately ten thousand Bengalis live in Nagpur. So although the city in which Sharada claimed to live was five hundred kilometers from Nagpur, there may well have been nu­merous opportunities closer to home for exposure to crucial information about the Bengali language and customs.

We should also note that Uttara seemed to be deeply interested in Bengal and the Bengalis, and she even ‘claims that she had a strong desire to learn Bengali.51 Beginning in her teens, Uttara became quite attached to her father, who was ‘a great admirer of Bengali revolution­aries and leaders’.52 at least one of whom had stayed with him in his home. Moreover, some of Utta­ra’s relatives spoke Bengali, and Uttara had read Bengali novels trans­lated into Marathi. According to Stevenson, Uttara ‘complained that Marathi literature displayed no real heroines; in contrast, she thought that Bengali women were more courageous and also more feminine than other Indian women’.53 Furthermore, as Anderson properly observes, both Akolkar and Ste­venson ‘include infor­mation on the linguistic features of Sharada’s Bengali suggesting that her command of the language, while impressive, is not that of a native’.54

Braude’s dissection of the reports by Stevenson and Akolkar provides considerable report for the conclusion that Uttara suffered from a dissociative pathology, and that the Sharada persona was a dissociative defense against – among several things – frustrations and disappointments in affairs of the heart. We must also remember that this case is quite weak evidentially; Uttara provided little if any evidence for the previous existence of a person corresponding to the Sharada persona. Apart from Uttara’s xenoglossy, the case very strong­ly resembles many relatively humdrum cases of dissociative pa­thology for which survivalist conjectures are not even tempting. Were it not for the xenoglossy, we would not seriously consider this case as providing anything but superficial evidence for survival. We would take it no more literally than we do cases of MPD/DID in which psychologi­cally useful alter personalities are clear­ly modeled after childhood icons or images such as Snow White, or Turkish cases in which alters claim to be the jinns or genies of Turkish folklore.55

An additional fact supporting an anti-survivalist interpretation of the case is that Uttara engaged in automatic writing, which relatively few people can do, but which some dissociators and other hypnotically gifted individuals do quite well. Also, when Professor Kini, a consultant on yoga, touched Uttara’s fore­head with his index finger, Uttara went immediately into Sharada.56 That, too, looks like the behavior of a highly hypnotiz­able individual. And again, Akolkar reports that Uttara would ‘sort of see’ another image behind her own in the mirror.57 That, too, is similar to a phenomenon reported by many multiples, who tend not only to be gifted hallucina­tors, but who even hallucinate their alters at distinct locations in a room.


Although cases of apparent responsive xenoglossy are undeniably interesting, they seem to do little to strengthen the case for postmortem survival. First, the degree of linguistic proficiency demonstrated is not clearly superior to other surprising eruptions of human competence demonstrated by savants, prodigies, and even more ordinary people in dissociative or other altered states, or else in conditions that enable them to draw on latent resources they would otherwise be unable to access. Those are all cases which seem clearly and most parsimoniously explicable in terms of a single living subject. Moreover the literature on xenoglossy shows a surprising and lamentable lack of familiarity with the relevant empirical and theoretical literature on hypnosis, dissociation generally, and second-language acquisition. This is not to say that one cannot mount a compelling case in favor of postmortem survival. It means only that if one wants to do so in a way that convincingly eliminates troublesome appeals to living-agent psi , or even just appeals to The Unusual Suspects – dissociation, latent capacities, prodigious memories58 – one must look elsewhere.

Stephen Braude


Akolkar, V.V. (1992). Search for Sharada: Report of a case and its investigation. Journal of the American Society for Psychical Research 86, 209-47.

Almeder, R. (1992). Death and Personal Survival. Lanham, Maryland, USA: Rowman & Littlefield.

Anderson, R.I. (1992). Commentary on the Akolkar and Stevenson reports. Journal of the American Society for Psychical Research 86, 249-56.

Bialystok, E., & Hakuta, K. (1994). In Other Words: The Science and Psychology of Second-Language Acquisition. New York: Basic Books.

Braude, S.E. (1995). First Person Plural: Multiple Personality and the Philosophy of Mind (Rev. ed.). Lanham, Maryland, USA: Rowman & Littlefield.

Braude, S.E. (2003). Immortal Remains: The Evidence for Life after Death. Lanham, Maryland, USA: Rowman & Littlefield.

Braude, S.E. (2014). Crimes of Reason: On Mind, Nature & the Paranormal. Lanham, Maryland, USA: Rowman & Littlefield.

Bryant, R.A., & McConkey, K.M. (1989a). Hypnotic blindness: A behavioral and experiential Analysis. Journal of Abnormal Psychology 98, 71-77.

Bryant, R.A., & McConkey, K.M. (1989b). Hypnotic blindness: Awareness and attribution. Journal of Abnormal Psychology 98, 443-47.

Bryant, R.A., & McConkey, K.M. (1989c). Visual conversion disorder: A Case analysis of the influence of visual information. Journal of Abnormal Psychology 98, 326-29.

Dreifuss, F.E. (1961). Observations on aphasia in a polyglot poet. Acta Psychiatrica et Neurologica Scandinavica 36, 91-97.

Ducasse, C.J. (1962). What would constitute conclusive evidence of survival after death? Journal of the Society for Psychical Research 41, 401-06.

Flournoy, T. (1900). From India to the Planet Mars: A Case of Multple Personality with Imaginary Languages. Princeton, New Jersey, USA: Princeton University Press.

Flournoy, T. (1902). Nouvelles observations sur un cas de somnambullisme avec glossolalie. Archives de Psychologie 1, 101-255.

Gauld, A. (1982). Mediumship and Survival. London: Heinemann.

Goldberg, B. (1982). Mechanism and meaning. In Knowledge and Mind, ed. by C. Ginet & S. Shoemaker, 191-210. Oxford: Oxford University Press.

McDougall, W. (1906). The case of Sally Beauchamp. Proceedings of the Society for Psychical Research 19, 410-31.

Myers, F.W.H. (1900). Pseudo-possession. Proceedings of the Society for Psychical Research 15, 384-415.

Myers, F.W.H. (1903). Human Personality and its Survival of Bodily Death. London: Longmans, Green, & Co.

O'Connell, D.N., Shor, R.E., & Orne, M.T. (1970). Hypnotic age regression: An empirical and methodological analysis. Journal of Abnormal Psychology, 76 (Monograph Supplement No. 3), 1-32.

Oakley, D. (1999). Hypnosis and conversion hysteria: A unifying model. Cognitive Neuropsychiatry 4, 243-65.

Orne, M.T. (1951). The mechanisms of hypnotic age regression: An experimental study. Journal of Abnormal and Social Psychology 46, 213-25.

Orne, M.T. (1959). The nature of hypnosis: Artifact and essence.’Journal of Abnormal and Social Psychology 58, 277-99.

Orne, M.T. (1962). Hypnotically induced hallucinations. In Hallucinations, ed. by L.J. West, 211–19. New York: Grune & Stratton.

Orne, M.T. (1972). On the simulating subject as a quasi-control group in hypnosis research: What, why, and how. In Hypnosis: Developments in Research and New Perspectives, ed. by  E. Fromm & R.E. Shor, 399-443. Chicago: Aldine-Atherton.

Putnam, F.W. (1989). Diagnosis and Treatment of Multiple Personality Disorder. New York: Guilford.

Putnam, F.W. (1997). Dissociation in Children and Adolescents: A Developmental Perspective. New York: Guilford.

Ross, C.A. (1997). Dissociative Identity Disorder: Diagnosis, Clinical Features, and Treatment of Multiple Personality. New York: John Wiley & Sons.

Sar, V., Yargic, L.I., & Tutkun, H. (1996). Structured interview data on 35 cases of dissociative identity disorder in Turkey. American Journal of Psychiatry 153, 1329-33.

Schiller, F.C.S. (1902). Professor Flournoy's 'Nouvelles observations sur un cas de somnambullisme avec glossolalie'.’Proceedings of the Society for Psychical Research 17, 245-51.

Spiegel, H., & Spiegel, D. (1978). Trance and Treatment: Clinical Uses of Hypnosis. New York: Basic Books.

Stevenson, I. (1974). Xenoglossy: A Review and Report of a Case. Charlottesville, Virginia, USA: University Press of Virginia.

Stevenson, I. (1984). Unlearned Language: New Studies in Xenoglossy.Charlottesville, Virginia, USA: University Press of Virginia.

Viscott, D.S. (1969). A musical idiot savant. Psychiatry, 32: 494-515.

Zoroglu, S., Yargic, L.I., Tutkun, H., et al. (1996). Dissociative identity disorder in childhood: Five Turkish cases. Dissociation 9, 253-60.


  • 1. Stevenson (1984), 160.
  • 2. Ducasse (1962).
  • 3. Braude (1995); Putnam (1989, 1997); Ross (1997).
  • 4. Stevenson (1984), 160.
  • 5. Viscott (1969).
  • 6. For a more detailed exploration of this topic, see Braude (2014), chap. 6.
  • 7. Gauld (1982), 102.
  • 8. Bialystok & Hakuta (1994).
  • 9. Bialystok & Hakuta (1994), 205-6.
  • 10. Bialystok & Hakuta (1994), 206.
  • 11. Bialystok & Hakuta (1994), 158.
  • 12. Bialystok & Hakuta (1994), 206.
  • 13. Bialystok & Hakuta (1994), 207.
  • 14. Bialystok & Hakuta (1994), 213.
  • 15. Bialystok & Hakuta (1994), 210.
  • 16. Stevenson (1974).
  • 17. Stevenson (1974), 50. Stevenson seems more cautious when he observes that the Modern Language Aptitude Test ‘predicts ... which students will perform well in modern language courses’ (1974, 51). But he incorrectly infers from this that the test predicts whether one can learn a foreign language ‘more or less easily’ (1974, 51).
  • 18. Stevenson (1984).
  • 19. Almeder (1992), 30.
  • 20. Almeder (1992), 30.
  • 21. Stevenson (1974), 25.
  • 22. Stevenson (1974), 26.
  • 23. Stevenson (1974), 28-29.
  • 24. Dreifuss (1961).
  • 25. Stevenson (1974), 53. Curiously, Stevenson doesn’t cite this report in his later opus on xenoglossy (Steven­son, 1984), although it would have been relevant there as well.
  • 26. Myers (1903), vol. 2, 484-86.
  • 27. Myers (1903), vol. 2, 486.
  • 28. For a detailed exploration of that deficiency in the survival literature, see Braude (2003).
  • 29. Stevenson (1984), 161ff.
  • 30. See, e.g., Goldberg (1982).
  • 31. Spiegel & Spiegel (1978).
  • 32. Orne (1972), 427. See also Orne (1951), 219, 222-23.
  • 33. Orne (1962), 218.
  • 34. Bryant & McConkey (1989a, 1989b, 1989c); Oakley (1999).
  • 35. See Orne (1959, 1972).
  • 36. Orne (1951).
  • 37. See also O'Connell, Shor, & Orne (1970).
  • 38. Stevenson (1984), 164.
  • 39. Braude (1995).
  • 40. Flournoy (1900).
  • 41. Flournoy (1900), 195.
  • 42. Flournoy (1902); Schiller (1902).
  • 43. Flournoy (1900), 156.
  • 44. Flournoy (1900), 156.
  • 45. Flournoy (1900), 163.
  • 46. Schiller (1902), 248-49.
  • 47. McDougall (1906), 431.
  • 48. Myers (1900), 415.
  • 49. See Braude (2003), chap. 4, for those details.
  • 50. Note that this is much later than the typical reincarnation case, in which subjects are young children. Arguably, this reinforces suspicion that dissociation, linguistic proficiency, and ESP together are sufficient to accommodate the facts of the case. See Braude (2003) for details.
  • 51. Akolkar (1992), 214.
  • 52. Akolkar (1992), 214.
  • 53. Stevenson (1984), 81.
  • 54. Anderson (1992), 252.
  • 55. Sar, Yargic, & Tutkun (1996); Zoroglu, Yargic, Tutkun, Ozturk, & Sar (1996).
  • 56. Akolkar (1992), 220.
  • 57. Akolkar (1992), 223.
  • 58. Braude (2003).