PreprintPDF Available
Preprints and early-stage research may not have been peer reviewed yet.


Reports of alleged cases of xenoglossy analyzed by professional linguists.
The Xenoglossy Analyzed by Linguists
Octavio da Cunha Botelho
The practice of speaking, reading and writing in a foreign language has
always required a lot of effort from the learner that is why so many enroll in language
schools and university courses. In addition, living in a foreign language country is a
way of learning. Even more difficult is the learning of ancient languages, since
vocabulary, grammar, and syntactic construction are very different from contemporary
languages. Even after learning a foreign language, the speaker must practice
constantly, because in the same way that we learn a foreign language, the lack of
practice leads to forgetting it.
So, with so much effort to learn and preserve the learning of a foreign
language, imagine the possibility of the occurrence of someone understanding,
speaking or writing in a foreign language that he or she has never learned. This is the
case of xenoglossy, word derived from the Greek ξένος (xenos) "strange" and γλ σσα
(glōssa) "language", therefore, “strange language". It is the controversial ability of
someone to speak or write in a language they have never learned through the
incorporation of a disembodied spirit or the regressive hypnosis of past lives.
Obviously, its occurrence is controversial, as we shall see below.
Understood by most Christians as an inspiration of the Holy Spirit, the most
popular case is the episode in Acts of the Apostles, 2.04-13, when the disciples of
Jesus spoke in different languages ( τέραις γλώσσαις - etherais glossais), an event
known as Miracle of Pentecost, before Jews from different parts of the world, who
claimed to understand in their own tongues what the apostles said. According to those
who believe in this phenomenon, it is the fact of speaking in an existing language,
which can be understood by a same speaker of that language, which differentiates
xenoglossy from glossolalia. The latter being speech through a cluster of
incomprehensible sounds that cannot be understood by another speaker because it is
a non-existent language commonly practiced by followers of the charismatic church
and Pentecostal churches in states of delirium. The Dictionary of Linguistics, by Jean
Dubois et al, defines glossolalia as follows: "The term glossolalia (...) designates the
verbal delusions of certain persons mentally ill. It is characterized by the voluntary
creation of deformed words, systematically associated with the same meaning and
resulting in a language incomprehensible to those who do not know their vocabulary ..."
(Dubois, 1973: 311). Glossolalia is also defined as "speaking in tongues", which is the
assertion of meaningless syllables, in a way that resembles a real language (Matlock,
2017). Unlike glossolalia, Ian Stevenson (1918-2007) defined xenoglossy as "speaking
in a real language wholly unknown to the speaker in his normal state" (Stevenson,
1974: 01). The written xenoglossia is also called xenography.
Recitative Xenoglossy and Responsive Xenoglossy
The term xenoglossy was coined by the French physiologist and
parapsychologist Charles Richet in the early twentieth century. In 1974, well-known
psychiatrist and parapsychologist Ian Stevenson divided xenoglossy into two
modalities: recitative xenoglossy and responsive xenoglossy. Recitative xenoglossy, as
its name implies, is one in which the hypnotized or the incorporated only recites or
repeats phrases, in a real foreign language, but does not understand its meaning, so it
is not able to converse, nor to ask questions or answer questions. It is simply a matter
of mechanically repeating memorized words from a real foreign language. In general,
even parapsychologists or reincarnation sympathizers do not recognize it, as true
xenoglossy, since it is the most frauds. In responsive xenoglossy, on the other hand,
the medium or the hypnotized is able to elaborate questions and answers on his own,
so he has a command of the grammar and vocabulary of the foreign language he is
uttering, that is, he is able to converse comprehending the questions and knowing how
to construct grammatically correct answers. This last modality of xenoglossy is what is
considered as real xenoglossy by the adherents of this belief and so is the one that is
investigated by the linguists of this supposed phenomenon. (Stevenson, 1974: 05,
Samarin, 1976: 271, Thomason, 1995: 01, McClelland, 2010: 279 and Matlock, 2017).
James Matlock recently introduced a third category, passive xenoglossy, to
explain the unconscious influence of a language not learned in pronunciation and in
other aspects of the production of speech, reading or writing, of those mediums
undergoing incorporation or of those patients during regressive hypnosis. Skeptics,
however, suspect that this third modality was introduced to justify pronouncement
failures and defects in grammatical constructions of those who utter tongues that spoke
in the past incarnation, since all the tests done by linguists so far have proven failures
in pronunciation, comprehension, accent and grammatical construction of the
languages spoken by the embodied or the hypnotized, as we shall see below. That is to
say, for these skeptics it is the mind of the medium who creates the fraud of a language
never learned by the embodied one, or the mind of the hypnotized one that creates the
false impression of being speaking in a tongue, also never learned, of an earlier
A curious difference between xenoglossy and the other experiences of
previous incarnations is its objectivity and concreteness, which makes it possible to test
the foreign speech never learned by the embodied or the hypnotized by speakers or
linguists. That is, language is a concrete tool to verify, through professional linguists,
whether the person investigated is correctly speaking a language they have never
learned. Some tests have already been done, so that we will then report and analyze
the most rigorously tested, therefore more cited in the literature on the subject.
The Xenoglossy Cases Analyzed by Linguists
If we believe the accounts of believers in this matter, the number of cases is
uncountable. However, the vast majority do not have enough data to enable them to
carry out detailed tests and rigorous analyzes by linguists in order to verify the truth of
the event, that is, whether or not the investigated is actually speaking in a language
that he or she has never learned. James Matlock's Xenoglossy in Reincarnation
Cases, published in the Psi Encyclopedia (Online), summarized 17 cases of responsive
xenoglossy (Matlock, 2017), three of which, most closely analyzed by linguists William
J. Samarin of the University of Toronto and Sarah G. Thomason of the University of
Pittsburgh will be analyzed next (the cases of Jensen, Gretchen and Sharada).
The Jensen Jacoby Case
This was a Swedish peasant, who manifested himself in an American
housewife, referred to as TE, under regressive hypnosis, conducted by her husband in
1955-6 for eight sessions. This case was investigated by Ian Stevenson and reported
in his 1974 book. TE was born and raised in Philadelphia; his immigrant parents spoke
English, Polish, Yiddish and Russian at home. The only foreign language she learned
at school was French. She has never had contact with the Swedish language, or any
other Scandinavian language. Stevenson confirmed that she never learned Swedish at
school, nor secretly without the knowledge of her parents. He concluded that, "in this
case, the ability to speak Swedish, as investigated, was not acquired through normal
learning" (Stevenson, 1974: 71; Samarin, 1976: 271; Rogo, 1985: 158-61 and
Thomason, 1995: 04). However, linguist Sarah G. Thomason remarked that, "his
(Stevenson's) demonstration that there was no fraud in the case is convincing, but his
claim that Jensen had the ability to speak Swedish is not convincing" (Thomason,
1995: 04).
Stevenson believed in the TE case and justified his investigation thus, "I
report the following case because I believe it provides an example of genuine
responsive xenoglossy. I think it is almost certain that the medium could not learn
Swedish, the foreign language here investigated, through normal means. In addition,
under hypnosis, she underwent a transformation into a male personality named
Jensen, who spoke and understood Swedish in an understandable way. This
personality was not simply reciting meaningless phrases: meaningful exchanges were
exchanged with Swedish-speaking people" (Stevenson, 1974: 23 and Samarin, 1976:
In contrast, the University of Pittsburgh linguist, Sarah G. Thomason, studied
and analyzed the TE / Jensen case and came to different conclusions. According to
her, "the Jensen's Swedish is, as Stevenson himself admits, less than perfectly fluent".
She then pleaded three reasons to mistrust Jensen's ability to speak Swedish correctly.
"First, Jensen uses only about sixty words spontaneously (i.e., in front of Swedish
speaking speakers) and, according to Stevenson's consultants, eliminating the words
correlated with English, German, and Yiddish, this number is reduced to thirty-one
understandable words. Secondly, Jensen has a total Swedish vocabulary of about a
hundred words; this is not very impressive when compared to thousands of words
known by any native speaker of any natural language, even taking into account the
limited context that Jensen spoke Swedish. Thirdly, he rarely answers questions with
complete sentences (...) the vast majority of Jensen's answers are of one or two words,
without context sentences" (Thomason, 1995: 04).
Other researchers who know the Swedish and Norwegian language
interviewed TE while she was manifesting Jensen's personality during regressive
hypnosis, and they agreed that Jensen's Swedish was mixed with Norwegian.
Stevenson justified this by claiming that Jensen's mother was Norwegian. In addition,
Jensen, who lived in Sweden, spoke English. Stevenson, in like manner, justified that
Jensen lived in the seventeenth century and immigrated to New Sweden in North
On the other hand, according to S. G. Thomason's conclusion, Jensen's
Swedish is so precarious that she even dismissed the possibility of fraud because, she
said, someone who had studied Swedish secretly with the intention of defrauding would
surely know more than Jensen knows of Swedish (Thomason, 1995: 12). In other
words, Jensen's Swedish does not even serve to deceive, because it is so defective.
The University of Toronto linguist, William J. Samarin, who also analyzed the TE /
Jensen case, commented that, "Jensen, who manifested through TE, was a very
passive participant. Always tired, sleepy and lazy, he answered in a few words what
was said to him, only rarely in full, but with very short sentences. Even more, they (the
answers) combined very precariously, in terms of normal dialogue with questions put to
him”. He then presented a few examples of the inconsistent answers to the questions
asked in Swedish.
Question: What do you do for living?
Answer: Em bonde [a farmer]
Question: How often do you go to Haverö? Do you go there often?
Answer: Ja, ja. Här torv [yes, yes, here market]" (Samarin, 1976: 272).
W. J. Samarin also added that, "Jensen's functional vocabulary was little
more than a hundred words, but of these, only 60 were used by Jensen before they
were used by the interviewers. In many instances, it was difficult to understand what
Jensen was saying. There were occasional grammatical errors”. Stevenson considered
Jensen's accent to be excellent, but Samarin noted that, "the accent was excellent only
on a few occasions that some words were poorly pronounced and others had an
American quality" (Samarin, 1976: 272).
Sarah G. Thomason noted that when Jensen was asked how much he used
to pay for the purchase of an item in the market, he replied "my wife" (Thomason,
1995: 13), confirming that he did not understand the question. She also added that "TE
had a little experience with Swedish, and that many of the 60 Swedish words Jensen
uses spontaneously are very similar to the words in French, English, Yiddish, or
Russian, all of which TE had studied or heard at home as a child" (ibid: 13).
The Gretchen Case
This was a German-speaking personality who manifested during four
sessions in a North American housewife, Dolores Jay, also through regressive
hypnosis by her husband. Stevenson, who had a certain knowledge of German along
with two native Germans, confirmed Gretchen's German proficiency and that she had
never learned German before. For them, therefore, it was an authentic case of
responsive xenoglossy, although parapsychologist and xenoglossy sympathizer James
Matlock himself recognized "lack of fluency and grammatical failures" (Matlock, 2017).
In the general evaluation of S. G. Thomason, "Gretchen's linguistic
performance is qualitatively similar to that of Jensen. Like TE, Dolores Jay's earlier
acquaintance with German was confined to television programs and a glance at a
German book, and she studied a German dictionary" (Thomason, 1995: 05). According
to this linguist's analysis, Gretchen's responses "are largely confined to one-word or
two-words statements, and many of them are simply repetitions of the interviewer's
question. Gretchen's vocabulary is minimal, and her pronunciation is spotty. For
example, it pronounces the blue word in German as "blü", which is the English word
"blue" with the German vowel "ü", but the blue word in German is "blau", which
pronounces the vowels as the English word "cow" (ibid: 05).
Another suspicion about Gretchen's German is that she seems to be more
influenced by spelling than by pronunciation itself. For example, she pronounced the
German word "schön" (beautiful) as the English pronunciation "shown", whose
pronunciation is not correct (ibid: 05-6). Another example of Gretchen's failure to
understand the German questions happened when she was asked what she ate for
breakfast, she replied with the word "bettzimmer", which is a literal translation of the
English word "bedroom", but not the correct German for bedroom, which is
"schlafzimmer", literally "sleeping room" (ibid: 13). Therefore, in addition to not
understanding the question in German, she answered by a non-existent word, that is,
two errors in a single answer of a single word.
Stevenson explained that Gretchen was "illiterate and was a neglected child
who spent most of her time in a kitchen with a maid who was poorly educated"
(Stevenson, 1984: 46). Sarah G. Thomason contended that even so, people with little
education have a vocabulary of thousands of words and grammar as complex as the
spoken language of an educated person, so Stevenson's explanation is not promising
(Thomason 1995: 13- 4). In addition, this linguist argued that, although Gretchen
reported that she was illiterate, she at one point wrote about 40 words (some of them
repetitions) in German, with misspellings, which resemble someone who learned only a
little German (ibid: 06).
The Sharada Case (Uttara Huddar)
Different from the cases of Jensen and Gretchen, when speech in foreign
language happened through regressive hypnosis, the case of Sharada (Uttara Huddar)
happened through spontaneous incorporation or reincarnation. Uttara Huddar was born
in 1941 in the city of Nagpur, Maharasthra state, India. Their native language is
Marathi, one of the languages derived from Sanskrit. This city is hundreds of miles
away from Bengal region, which speaks Bengali (the language of Indian poet and writer
Rabindranath Tagore [1861-1941], winner of the Nobel Prize for Literature in 1913),
another language derived from Sanskrit . His parents spoke almost exclusively Marathi,
I say almost exclusively, since in India, because of the diversity of languages, it is
common for someone to speak more than one language, or speak a language by
adding words from another language, especially English. Almost all the languages of
India include Sanskrit words in their vocabulary, especially the proper names.
About 10,000 immigrants from Bengal lived in his city, but Uttara had no
contact with these people. In her youth, she was not an exceptional student, but
nonetheless developed an interest in dance, theater, and Sanskrit, attended a college
and graduated in biology. After graduation, he taught at Nagpur University. She was
also interested in spirituality, her father introduced her to yoga and meditation practice
in 1965, after some time she tried the practice of awakening kundalinī. However, these
experiences had a bad effect on her mind in the year 1973, so she had to be
hospitalized for frequent headaches and blackouts.
The earliest memories of her past life began in 1973 at the hospital. She was
already interested in the culture of Bengal, the eastern region of India rich in literature,
art, religion, music and dance. For example, one of the most internationally known
Indian religious movements, the Hare Krishna Movement, has its roots in Bengal.
Therefore, it was during the internment period that Uttara had the first experience of
incorporating Sharada. She used to experience Uttara's personality one day, and the
next day she would wake up with Sharada's personality, speaking in Bengali, saying
that she did not belong to that family and to that place. Between February 1974 and
April 1977, for example, there were 23 cases of Sharada's possession of Uttara's body.
When she awoke to Sharada's personality, she spoke in Bengali, a language far from
where she lived in India, so her family and hospital staff did not understand her.
In one of her psychotic outbreaks, she witnessed her director's dinner with a
woman in the private rooms of this doctor, so she had a jealous crisis because she had
developed a strong attraction to the hospital director, who did not correspond to her;
she thought he was her husband.
Uttara went on to say that, it was Sharada, a Bangali woman, who lived in the
early nineteenth century, then proceeded to wear the sari according to the custom of
that region. Her family allowed eight speakers of the Bengali language to question
Uttara, who confirmed that she, when possessed, spoke the Bengali language.
Sharada provided the names of his relatives in Bengal, so Ian Stevenson and his
assistants conducted a survey in that state and confirmed that the relatives mentioned
by Sharada actually existed at that time, but only the male genealogy was found, so
that the confirmation of existence Sharada's continued to be a doubt.
The intriguing thing about this genealogy of only male names is that when
Sharada, in a state of trance, was asked to mention the names of her relatives, she
mentioned ten names only of men and reported the kinship with six of them. She
reported that her father was Brajanath Chattopadhaya, her mother Renukha Devi and
her grandfather Ramnath, both male names confirmed in the genealogy. However, five
names mentioned by Sharada do not appear in the genealogy. Since the genealogy
mentions only male family names, it was not possible to confirm the existence of
Sharada and her mother Renukha Devi. Ian Stevenson asserted that Uttara could not
have knowledge of such a genealogy (Rogo, 1985: 152-3). However, the fact that
Sharada provided almost only male names of his relatives and, coincidentally, the
genealogy presents only male names leaves the suspicion that, secretly, Uttara may
have had, in a very casual way, access to this genealogy, without the knowledge of
others. What might have made this access possible was the fact that the genealogy
was published in a Bengali magazine in 1907 (Rogo, 1985: 153).
Ian Stevenson researched Uttara's past and discovered that she received
some Bengali classes when she was studying Sanskrit, he added that she learned only
to read a few words and nothing more, not being able to read even a complete
sentence. Nor could she speak Bengali. Stevenson's conclusion is that Sharada's case
is a genuine example of paranormal xenoglossy (Rogo, 1985: 153-4). Another
researcher, Akolkar, came up with a different result, claiming that Uttara may have had
some contact with Bengali at her school and that she would have more command of
the language than was expected (ibid: 154). The situation was complicated when the
researchers found a former classmate of Uttara, who claimed that they both studied
Bengali together. He assured that they had learned enough to do an elementary
reading in Bengali. These revelations may be the signal that Stevenson exaggerated in
his assessment of Sharada's case as genuine responsive xenoglossy (ibid: 154).
The linguist Sarah G. Thomason, who also analyzed the Sharada case, noted
that "Bengali knowledge of Uttara may have been obtained by normal means: she read
translated Bengali novels" (Thomason, 1995: 07, see also: Stevenson: 1984: 153).
Specialists in the Bengali language who investigated Sharada's language skills
disagreed over her proficiency. For example, one Dr. Roy said that 'Sharada
demonstrated a complete mastery of the Bengali language', and one Professor Pal
agreed. In contrast, M. C. Bhattacharya said that 'although Sharada could speak
Bengali understandably, she did not speak fluently and sometimes had to seek for the
words', this assessment was repeated by Ranjan Borra, who added that her Bengali
accent was not that of a speaker native Bengali, was more to that of a non-Bengali
speaker who learned to speak Bengali after childhood. Dr. Roy himself commented that
her Bengali pronunciation was not good” (Thomason, 1995: 05).
The most important evaluation, however, was that of Tagore Bengali
professor at the University of Delhi, Sisir Kumar Das, the only linguist trained among all
the native speakers who studied Sharada's Bengali. He concluded that her Bengali
was neither natural nor fluent, that her accent was foreign, that her Bengali represented
a dialect inferior to the standard of West Bengal, that she spoke a non-native variety of
Bengali of the twentieth century, not a variety of the nineteenth century and, in short,
that her Bengali resembled that of someone who had learned Bengali as a second
language, though not quite correctly (ibid: 07-8). Stevenson reproduced Professor
Das's testimony in full, but suggested that since the conversations between Professor
Das and Sharada were brief, it could be that Sharada did not have time to "warm up"
when speaking to him, and so did not demonstrate his abilities in Bengali, he also
argued that Sharada's influence on Marathi might have explained his need to speak
through the mouth of Uttara (Stevenson, 1984: 137). On this, S. Thomason remarked
that, "not even Professor Das's testimony nor Uttara's deep interest in Bengali shook
Stevenson's belief in the paranormal nature of Sharada's Bengali" (Thomason, 1995:
08). This confirms the revelations of one of his former assistants that, "Stevenson used
to ask ‘selective questions' to the investigated in such a way that they tended to get
answers he wanted", and also: "there was a tendency on the part of Stevenson to
unconsciously 'fill' a history investigated in order to make it more complete"
(McClelland, 2010: 262 and Botelho, 2018: 08).
S. G. Thomason explained that, unlike the cases of Jensen and Gretchen,
when Stevenson published abundant transcripts of the interactions between the
investigators and the interviewers, the transcripts of the conversations with Sharada
are scarce, "only a few excerpts from the English translations of the interviews
(Stevenson, 1984: 206-9 and Thomason, 1995: 08), then she concluded that, "the sum
total of the linguistic proof provided by Stevenson is thus inconclusive" (Thomason,
1995: 08).
The General Assessments of Linguists
Linguist William J. Samarin, in addition to being wary of Ian Stevenson's
findings in the TE / Jensen case, also discredited his co-workers, claiming that "one
cannot help but suspect that all or most of Stevenson's co-workers were , to some
degree or other, 'fellow believers' inclined to begin with the premise that good faith
xenoglossy is possible". Since they were not horrified when they showed a seed to TE /
Jensen and asked her, what was that, Jensen could not answer (Samarin, 1976: 273).
Well, it is inconceivable that a speaker does not know the word seed on his own,
because, faced with a failure like this; no one needs to be a graduate linguist to
convince himself that Jensen could not be a real Swedish speaker.
Broadly speaking, Sarah G. Thomason evaluated that "despite Stevenson's
efforts to provide genuine proof in support of his paranormal claims, his linguistic proof
is completely incapable of convincing a professional linguist. There are two main
problems with this. First, his notion of 'responsive xenoglossy' is fatally flawed as a
methodological criterion for determining a person's ability to speak a language. And
second, most of the explanations he suggests for the obvious failings of the Swedish
and German of his investigated fits his paranormal purpose squarely into the realm of
pseudoscience" (Thomason, 1991: 86).
Against arguing, James G. Matlock remarked, "that Thomason and other
linguists intend to see signs that Jensen and Gretchen are capable of performing (a
linguistic proficiency) in the same way as mature native speakers, while Stevenson is
looking for evidence that they have only some degree of language proficiency.
Linguists understand that a language handed down by successive lives will be
structured like the language spoken in any past life, but Stevenson is open to a wider
horizon of possibilities” (Matlock, 2017). What Matlock means is that an individual
under regressive hypnosis or incorporation will not necessarily speak in another
language in exactly the same way or with the same performance as someone in a
normal state of consciousness. That is, there will always be an interference or an
obstacle, due to regression or incorporation that will deform the proficiency in the use
of the foreign language, so the linguistic criterion must be different from the
parapsychological criterion.
Now if Matlock's argument is sustainable, then responsive xenoglossy can never be
tested by scientific and linguistic methods, so that it will invariably remain, as S. G.
Thomason has pointed out, as a pseudoscience.
Sarah G. Thomason evaluated the three investigations as follows: "All three
investigated by Stevenson made mistakes in pronunciation and foreign accents.
Sharada made grammatical errors in his Bengali, while Jensen and Gretchen were so
laconic that their statements exhibit very few examples of grammatical constructions"
(Thomason, 1995: 13). Moreover, "the level of comprehensiveness of Jensen and
Gretchen was too low to convince a linguist that they had any significant degree of
knowledge of a language. Contrary to Stevenson's beliefs, these individuals
investigated showed no skill in languages beyond knowing a handful of words and
some grammatical traits. Also, their passive knowledge of Swedish and German, their
ability to understand the which was said to them, was, if any, weaker than their active
knowledge of words and phrases, for real speakers of real tongues, including second
language speakers, have a much less passive knowledge than the active knowledge of
the tongue" (ibid, 15). In addition, he concluded: "Stevenson's responsive xenoglossy is
flawed as a criterion for proving the knowledge of a language, at least at the low level
of understanding demonstrated by Jensen and Gretchen. So while one might agree
with Stevenson that a genuine case of xenoglossy would be impressive for a case of
paranormal phenomena, it is also true that no convincing case has happened so far"
(ibid: 15).
BOTELHO, Octavio da Cunha. Reincarnation under Scrutiny, August 2018.
DOI: 10.13140/RG.2.2.32255.92323
DUBOIS, Jean et al. Dicionário de Lingüística. São Paulo: Editora Cultrix, 1973.
MATLOCK, James G. Xenoglossy in Reincarnation Cases in Psi Encyclopedia, Society
for Psychical Research, 2017, Online Edition.
MCCLELLAND, Norman C. Encyclopedia of Reincarnation and Karma. Jefferson:
McFarland & Company, Inc. Publishers, 2010, p. 279.
ROGO, D. Scott. The Search for Yesterday: A Critical Examination of the Evidence for
Reincarnation. Englewood Cliffs: Prentice Hall, 1985, p. 137-61.
SAMARIN, William J. Xenoglossy: a Review and Report of a Case by Ian Stevenson in
Language, vol. 52, number 01, 1976, p. 270-4.
STEVENSON, Ian. Xenoglossy: A review and report of a case. (Proceedings of the
American Society for Psychical Research, 31). New York: American Society for
Psychical Research. (Republished, Charlottesville: University Press of Virginia, 1974).
_____________ Unlearned language: New studies in xenoglossy. Charlottesville:
University Press of Virginia, 1984.
THOMASON, Sarah G. Past Tongues Remembered? in The Hundredth Monkey and
Other Paradigms of the Paranormal. Kendrick Frazier (ed.). Buffalo: Prometheus
Books, 1991, p. 85-94.
___________________ Xenoglossy. January 1995. Electronic Edition.
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.