Journal of Scientifi c Exploration, Vol. 24, No. 1, pp. 101-108, 2010




How To Improve the Study and Documentation
of Cases of the Reincarnation Type?
A Reappraisal of the Case of Kemal Atasoy


Rua Professor Manoel Ferreira, 144, Apto 608 Rio de Janeiro,
Rio de Janeiro 22451-030, Brazil

Abstract—This article tries to analyze the problems related to the methodology of investigating cases of the reincarnation type, starting from the description of the case of Kemal Atasoy as carried out by investigators Jim Tucker and Jurgen Keil. Also, it tries to identify problems in the reports presented by these researchers. It suggests possible solutions to the problems presented, namely the use of tape recorders and video cameras to enable a much more reliable recording of the interviews, and of tests of recognition of persons and objects, which should be evaluated by an expert in cold reading. The importance of psychological testing is stressed, both for the child and for the family. Constructing a genealogical tree of people involved is recommended and so is the use of double-blind questionnaires.

Keywords: reincarnation—type case—field methodology—records—field tests



In 2005, H. H. Jurgen Keil (JK) and Jim B. Tucker described the case of Kemal Atasoy (KA), a boy from Turkey who claimed to have memories of a previous life. They published the case in the Journal of Scientific Exploration (Keil and Tucker, 2005) and in the introduction of the book Life before Life by Jim Tucker (2005), apparently due to its strength as a case that points to an anomalous phenomenon and is suggestive of reincarnation.

The very moment I fi rst read about the case, I was struck by the impression that it could have been described in a better way for the public. Moreover, I believe that the case could also have been better investigated. Therefore, I will try here to point out in an objective way what weaknesses I've discovered in the documentation and presentation of the case, and to suggest improvements.


The Participation of the Interpreter

The article mentions the use of an interpreter:

JK's interpreter was acquainted with KA's parents, who agreed to arrange this fi rst visit. The interpreter had mentioned JK's interest in children who speak about previous lives to KA's parents, who told him that KA was doing this. (Keil and Tucker, 2005:92)

Such investigations have already suffered enough from the accusation of fraud on the part of the interpreters. In his autobiography, the Canadian-born physician and psychiatrist Ian Stevenson, who recently passed away, discusses the delay that his material suffered due to accusations about one of his interpreters:

The report of my fi rst studies in Asia was in press when unexpectedly a man who had helped me with some cases was accused of cheating. Although the allegation applied to experiments with which I had nothing to do, suspicion spread to the work the accused man had done for me, and the editor stopped the printing of my report. I had had other interpreters besides the man accused of cheating, and, believing that the man had not cheated when working with me, I proposed to return to India and study the cases anew. Yet this entailed great additional expense, and I asked Chester Carlson's advice. He encouraged me to return to India. I did this, and, with new interpreters, showed the authenticity of the cases. The printing of my report was then resumed, and it was duly published as Twenty Cases Suggestive of Reincarnation (Stevenson, 1966/1974). (Stevenson, 2006:16-17)

The interpreter was Hemendra Nath Banerjee, who published various books about cases suggestive of reincarnation which he had studied himself.

Furthermore, because the researchers were not fl uent in the language or familiar with the culture, they could not immediately evaluate the reliability of the interview carried out by the interpreter.

In the case of KA, the interpreter was known to the family and had relatives in Istanbul (which is the same location wherein the previous personality used to live), which increased the possibility of fraud.

To minimize such problems, I would like to suggest that at least the interviews with the child should be recorded on audio and video, which would later on enable any person who is fl uent in the language to verify what is being said. Below, I will address this more specifi cally, responding to the reasons presented by the researchers as to why they did not use such methods. It would be interesting also if the investigators prepared a genealogical tree of all the people involved, thereby demonstrating that there are no family ties—or that there are such ties—that might facilitate the transmission of information.


The Interview with the Child

According to the article, the interviewing of the boy took place as follows:

His mother was present for the entire interview, and his father joined the interview partway through it. The interpreter translated what was said after short intervals, and JK took notes and asked questions. Audio recordings were not made since JK has found that a cassette recorder often creates a distraction. At this fi rst meeting, JK was impressed with the clarity of KA's statements and the confi dence with which he made them. (Keil and Tucker, 2005:92-93)

No video or audio recordings of the interview were made, which is a pity. Such a recording would enable other investigators to verify the exact words of the child, which could have been misunderstood by the translator, especially regarding names. Antonia Mills (2003:76-77) describes the case of Sakte Lal, in which the child would mention the names of Amlapur, Vishnu, and Avari, whereas in reality the child would be referring to Jamlapur, Kishnu, and Itwari, respectively. Mills also states that the father of the child never mentioned the initial distortions of the names uttered by the child, and that this was only done by the mother.

The reason the researchers give for not using recording equipment, that such use would distract the child, is in my view unacceptable. Microcameras hidden in the clothing worn by the researchers are often used by detectives, journalists, and police offi cers, and they enable excellent recordings, i.e. both audio recordings and video recordings. Such a device should at least be used when interviewing the child, after getting permission from the parents, to avoid any distraction for the child. With respect to the child's relatives, current methods, though less sophisticated, may continue to be used, and a range of justifi ed motives can be given for this. In a review of a former draft of this article, Erlendur Haraldsson mentioned several such motives, including that one should try to minimize the problems related to using an interpreter.

Audio recordings could be used more often but as Keil and Tucker state in their paper on Kemal Atasoy . . . such recordings are often a distraction to the child and other witnesses, and may even make the interviewees more reluctant to speak freely. Using video equipment is cumbersome. It has to be set up, it causes a lot of questions and curiosity, and it may make the witnesses reluctant to take part (“Are you going to take this interview to the television?”). Investigating CORT [a case of the reincarnation type] often involves a balancing act between the optimally desirable and the practically possible. Usually we end up taking handwritten notes and trying to make the best use of the interpreters who are never professional interpreters and most of whom tend to give us summarized reports rather than verbatim translation, however much they are encouraged to do the latter. (Haraldsson, 2009)

The constant presence of the mother forms another problem. It is possible that she might help the child, in a subtle way, to answer the questions. One example of this phenomenon was recently established by Antonia Mills, in the case of Sakte Lal (Mills, 2003:79-80). In India people don't accept money from relatives by marriage. Based on this, a test was carried out by the stepfather of the daughter of the previous personality with the subject of the case, in order to confirm his identity. Thus, the stepfather offered money to Sakte Lal, as he wanted to know if he would return the money or not. The child returned the money, which at the time was considered a positive proof of his reincarnation. Fortunately, Mills made an audio and video recording of this specific moment, and after listening to the tape at least three times and editing it for a presentation, Mills heard how Sakte Lal's sister told him to return the money! Thus, the "positive proof" was considered invalid. The words of the sister had escaped the attention of both the interpreters that Mills had used. This incident not only demonstrates that recording is necessary, but also that it is important to interview all the persons involved individually.

Erlendur Haraldsson adds two more interesting criticisms:

In the case of Kemal Atasoy the child and the mother seem to have been interviewed only once as it is not explicitly stated otherwise. (See though on page 92: "In April ...JA met KA for the fi rst time", which indicates that there was more than one interview, but no further details are given.) Stevenson—and so has this reviewer—made it a rule always to interview a child on more than one occasion particularly when the case was of special interest or was going to be published. This would involve another or more visits several months or a year later to test the consistency of the child's and other witnesses' statements.

Another very important point is to find as many witnesses to the child's statements as possible, not only the mother, but grandparents, siblings, playmates, etc. In the paper on Kemal Atasoy, it is not mentioned that other witness were interviewed (or found?), and that justifi es some criticism of the case and its investigation. It should though be borne in mind that there may be mitigating circumstances, such as lack of funds (always hard to fi nd) for many visits and the time the researchers have for their investigation. (Haraldsson, 2009)


The Interview with Mr. Toran Togar

Once the statements of the child had been noted down, Keil tried to verify them with a local historian, Toran Togar.

The following is a description of how the events took place, according to Tucker:

When JK interviewed Mr. Togar, he was careful not to prompt any answers or to make any suggestions. They met near the previous personality's house, and JK asked Mr. Togar what he remembered about the people who had lived in this house. JK told him he wanted to fi nd out some details for a family that might have had a connection with the house, but he did not say that a child had provided information about it. (Keil and Tucker, 2005:94)

It is good to know that Keil was cautious enough not to lead the interview, but, as I understand it, in order to make an investigation stronger, the interview should not have been carried out by Keil, since he already possessed the statements of the child, but by another independent investigator, who would lack all knowledge of the statements. The procedure was single-blind, whereas it could have been double-blind.

And, once more, an audio or video recording of the interview should have been carried out. An example of the importance of such a procedure was cited by Matlock (1990) concerning the case of Mounzer Haidar.

When he studied this case, researcher Ian Stevenson fi rst made a drawing of the location of a birthmark on the abdomen of the subject. When he, later on, interviewed the mother of the previous personality, he asked her where the shot wound had been located, and she pointed at the right side of her own belly. Stevenson then showed her his drawing, and the woman told him that the wound had been on the marked spot. This amounts to leading the testimony. The ideal procedure would be that Stevenson would first have asked the woman to make a drawing of the location where the bullet had entered the body, before showing her his own drawing.

Thus, the recordings of the interview would constitute a guarantee that such errors would not occur. However, if the procedure had been double-blind, I believe that we could manage without the recording of the interviews to avoid the possible problems of distraction and inhibition that the cameras could provoke in the persons interviewed, as in such a case there would be no way in which the investigator could infl uence them.


Tests of Recognition

The child was not taken to the house of the previous personality to fi nd out if he would be capable of recognizing objects and making more statements. Keil simply took a photograph of the house with him for Kemal, who apparently was highly impressed by it. I consider this a mistake of Keil. Keil did not even fi lm the reaction of the child to the photograph, and missed the chance to mix the photo of the house of Karakas with the photographs of other houses, so that he could have asked the child to select the photo that matched his previous life.

Still, an even more serious question remains: Why was the child not taken to the location where tests of recognition could be carried out? Or, better still, why did they not arrange for the child to show the way to the house, while driving in a car, i.e. via a route that he (the previous personality) would have been used to?

Evidently, in order to avoid the possibility that the child would pick up cues about the location of the house from other persons, in the car there should only be persons present who would not know the route nor possess any other important information concerning the case. The route should be fi lmed from the beginning until the end, so that we could be certain that the child would not have any normal means of obtaining information (for example, signs with street names could amount to cues).

Once they have reached the house, tests of recognition of objects and persons could be carried out. In the case of Kemal, the house was abandoned, but this point of advice is important for other CORTs. The ideal procedure would be that such tests would be carried out by an expert in cold reading, a magician or a mentalist, in such a way that he would make sure that the child would not recognize his relatives or friends from the previous life by subconsciously—or consciously—picking up clues, from themselves or even from the investigators, who would already know who such friends or relatives would be. Naturally, everything should be recorded.


Psychological Tests

The article does not state anything about carrying out psychological tests on the child or the parents. Could the child be easily infl uenced? Does the family possess some history of psychological disorders? All this must be addressed in the article in order to make the investigation more robust and to reduce the number of possible normal explanations of the case.

Psychological tests with children who claim to recall a previous life were already carried out by Haraldsson (1997) and Haraldsson et al. (2000), and the results of those tests indicated rather strongly, at least regarding solved cases, that children claiming memories of past lives were not more suggestible than children of the same age who live in comparable circumstances.

However, if we analyze a particular case, it is good to look for all possible guarantees that the case in question does not amount to an exception to the rule. We may compare this to CORTs based on fraud: They are very rare, but they do exist (Stevenson et al., 1988).

In my view, the fact that cases of fraud or cases wherein children are highly suggestible are extremely rare among CORTs does not imply that we should not take the necessary precautions to avoid them. Furthermore, replication is part of the scientific process.

However, once more, there are mitigating factors, to which Haraldsson (2009) alludes in his review:

There exist tests of interrogative suggestibility but with no norms for various countries or age groups. Hence they cannot be used for individual children.


Description of the Case in the
Journal of Scientific Exploration and in the Book Life before Life

Keil states that Kemal was very clear in his statements, but neither in the article published in the Journal of Scientifi c Exploration (Keil and Tucker, 2005) nor in the book by Jim Tucker (2005) do we fi nd any transcription of the exact words of the child. Moreover, the questions posed to Kemal and Mr. Togar are not transcribed either, although in the latter case (Tucker, 2005) we do get a good idea of what the questions were like. The transcription of the interviews, with the questions and answers, is essential for the reader to be certain that the informants were not infl uenced by any leading questions from the interviewer. That material could be published separately from the main text, as an appendix, or else online on a webpage of the authors, so that it would not affect the reading of the main article.


Is It Still a Strong Case?

After we have pointed at various weaknesses and the methodology of the researchers, an important question that can be asked is whether the case remains "strong", in the sense that an explanation based on normal means is extremely difficult.

In my view, the answer is positive, because I think it is improbable that all the possibilities I have referred to would have existed all at the same time in this case: a fraudulent interpreter, a relative who in a subtle way helps the child to respond correctly to questions posed by the investigators, induction of a desired answer from the persons interviewed by leading questions from the investigators. As far as I know, such errors were found only separately in other cases, meaning that they were never found in combination. In fact, the impression I get from examining a case such as that of Kemal is that it is an extremely difficult task to fi nd a somewhat plausible naturalistic explanation for it.


Earlier Criticisms

Edelmann and Bernet (2007) criticized contemporary studies of CORTs addressing various improvements that I've suggested here, but they created a model of investigation that, in my view, is too complicated and unnecessary. I believe that my recommendations are suffi cient and much better suited to be adopted by researchers.



There remains a lot to be improved in studies of CORTs, but one should realize that there are also several mitigating factors involved in many of the errors found. It would be interesting to add a presentation of the genealogical tree of the people involved, as well as the use of double-blind questionnaires during the interviews. The collaboration of magicians and/or experts in cold reading is recommended when tests of recognition of persons and objects are applied. Whenever possible, psychological tests of all the persons involved should be carried out. I recommend audio and video recording of all the interviews carried out, and such interviews should be completely transcribed in published articles in an appendix, to avoid the papers becoming practically unreadable. In order to generate more knowledge and interest among the scientific community concerning research of CORTs, I suggest that as much material as possible—such as audio and video recordings—should be placed on the Internet, so that a larger number of scholars can get access to the studies, and would thus be able to identify more easily the strengths and weaknesses of the cases. Such measures, if adopted, would certainly considerably heighten the level of evidence for the occurrence of some paranormal phenomenon in CORTs.



I'm grateful to biologist Júlio César de Siqueira Barros for his advice and ideas which have greatly improved my paper. The researcher Titus Rivas was also of great help in translation. I would also like to thank the reviewers of this article, especially Erlendur Haraldsson, for their constructive criticisms.



Edelmann, J., & Bernet, W. (2007). Setting criteria for ideal reincarnation research. Journal of Consciousness Studies, 14, 92-101.

Haraldsson, E. (1997). A psychological comparison between ordinary children and those who claim previous-life memories. Journal of Scientifi c Exploration, 11, 323-335.

Haraldsson, E. (2009, 24 September). Personal communication.

Haraldsson, E., Fowler, P., & Perriyannanpillai, V. (2000). Psychological characteristics of children who speak of a previous life. A further fi eld study in Sri Lanka. Transcultural Psychiatry, 37, 525-544.

Keil, H. H. J., & Tucker, J. B. (2005). Children who claim to remember previous lives: Cases with written records made before the previous personality was identifi ed. Journal of Scientifi c Exploration, 19, 91-101.

Matlock, J. G. (1990). Past life memory case studies. In Krippner, S. (Ed.), Advances in Parapsychological Research, 6, 184-267. Jefferson, NC: McFarland.

Mills, A. (2003). Are children with imaginary playmates and children said to remember previous lives cross-culturally comparable categories? Transcultural Psychiatry, 40, 62-90.

Stevenson, I. (2006). Half a career with the paranormal. Journal of Scientifi c Exploration, 20, 13-21.

Stevenson, I., Pasricha, S., & Samararatne, G. (1988). Deception and self-deception in cases of the reincarnation type: Seven illustrative cases in Asia. Journal of the American Society for Psychical Research, 82, 1-31.

Tucker, J. B. (2005). Life before Life: A Scientifi c Investigation of Children's Memories of Previous Lives. St. Martin's Press.