Journal of Scientific Exploration, Volume 5, Issue 1

@JOURNAL OF SCIENTIFIC EXPLORATION A Publication of the Society for Scientific Exploration

Volume 5, Number 1 1991

CONTENTS

Page

Research Articles

1 The Cydonian Hypothesis

27 New Cases in Burma, Thailand, and Turkey: A Limited Field Study Replication of Some Aspects of Ian Stevenson's Research

John E. Brandenburg Vincent DiPietro Gregory Molenaar

Jurgen Keil

6 1 Effects of Consciousness on the Fall of Dice: Dean I. Radin A Meta-Analysis Diane C. Ferrari

85 The Wasgo or Sisiutl: A Cryptozoological Sea- Michael D. Swords Animal of the Pacific Northwest Coast of the Americas

103 The Extraterrestrial Hypothesis Is Not That Bad Robert M. Wood

1 13 Toward a Second-Degree Extraterrestrial Theory of Jacques Vallee UFOs: A Response to Dr. Wood and Prof. Bozhich

1 2 1 Low-Frequency Seismogenic Electromagnetic Takeo Yoshino Emissions as Precursors to Earthquakes and Volcanic Eruptions in Japan

Indexed in CABS ISSN 0892-3310 (836)

@JOURNAL OF SCIENTIFIC EXPLORATION A Publication of the Society for Scientific Exploration

Editor Bernhard Haisch*

ERL 306 Stanford University Stanford, CA 94305

*Mail should be addressed to the Editorial Office (see below).

Associate Editors R. G. Fowler Roy B. Mackal Dean I. Radin Department of Physics Department of Safety Energy AI/MMI Laboratory University of Oklahoma University of Chicago Contel Technology Center Norman, OK 73069 Chicago, 11,60637 Chantilly, VA 2202 1-3808

David M. Jacobs Earl F. McBride Ron Westrum Department of Hisfory Department of Geological Sciences Department of Sociology Temple University University of Texas Eastern Michigan University Philadelphia, PA 1 9 1 22 Austin, TX 787 12 Ypsilanti, MI 48 197

Editorial Board Richard C. Henry, Ronald A. Howard, Robert Jahn

Assistant Editor: Marsha Sims.

I Editorial Ofice: Journal of ScientiJic Exploration. ERL 306, Stanford University, Stanford, CA 94305-4055. Telephone/ Telefax Message Center: 41 5-593-858 1

Publishing, Subscription and Advertising Ofices: Pergamon Press, Inc., 395 Saw Mill River Road, Elmsford, New York 10523, USA, INTERNET "[email protected]"; and Pergamon Press plc, Headington Hill Hall, Oxford OX3 OBW, England

Published Two Per Annum. Annual institutional subscription rate ( 199 1 ): US$120.00; Professional subscription rate (1991); US$40.00. Prices are subject to change without notice. Members of the Society of Scientific Exploration receive personal subscriptions as part oftheir dues; details of membership are available upon request. Notify 8 weeks in advance of address change with a copy of the subscription mailing label. Back issues; Back issues of all previously published volumes, in both hard copy and on microform, are available direct from Pergamon Press offices. Subscription rates for Japan include despatch by air and prices are available on application.

Copyright 63 199 1 Society for Scientific Exploration

It is a condition of publication that manuscripts submitted to this journal have not been published and will not be simultaneously submitted or published elsewhere. By submitting a manuscript, the authors agree that the copyright for their article is transferred to the publisher if and when the article is accepted for publication. The copyright covers the exclusive rights to reproduce and distribute the article, including reprints, photographic reproductions, microform or any other reproductions of similar nature and translations. No pan of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by an means, electronic. electrostatic, magnetic tape, mechanical, photocopying, recording or otherwise, without permission in writing from the copyright holder.

Photocopying informationfor users in the U.S.A. The Item-Fee Code for this publication indicates that authorization to photocopy items for internal or personal use is granted by the copyright holder for libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service provided the stated fee for copying, beyond that permitted by Section 107 or 108 of the United States Copyright Law, is paid. The appropriate remittance of53.00 [$0.50 for review journals] per copy per article is paid directly to the Copyright Clearance Center Inc., 27 Congress Street, Salem, MA 01970.

Permission for other use. The copyright owner's consent does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific written permission must be obtained from the publisher for copying. Please contact the Subsidiary Rights Manager at either Pergamon Press, Inc. or Pergamon Press pic.

rw The paper used in this publication meets the minimum requirements of American National Standard for lnformation @ Sciences-Permanence of Paper for Printed Library Materials. ANSI 219.48-1984.

The Ifem-Fee Codefor this publication is: 0892-3310/91 $3.00 + .00

Printed in the U.S.A.

Journal of Scientific Exploration. Vol. 5 , No. 1, pp. 1-25, 1991 Pergamon Press plc. Printed in the USA.

0892-331 019 1 $3.00+.00 0 1991 Society for Scientific Exploration

The Cydonian Hypothesis

JOHN E. BRANDENBURG, VINCENT DIPIETRO, and GREWRY MOLENAAR

Mars Research, P.O. Box 284, Glenn Dale, MD 20769

Abstract-Evidence suggesting a past humanoid civilization has been found at several sites on Mars. In particular, what appear to be large carved faces, with similar details, have been found at two separate sites. Together with geochemical and geological evidence that suggests Mars was once more Earth-like in climate, the images of the objects support the Cydonian Hy- pothesis: That Mars once lived as the Earth now lives, and that it was once the home of an indigenous humanoid intelligence.

Introduction

Mars is a planet whose past is a mystery, but like all great mysteries, the keys to solving the puzzle may hide in plain sight. Mars, with its somber red surface, its vast and winding canyon systems displaying deep deposits of sedimentary rock, its great scablands telling of past awesome floods ofwater, and its windswept plains cratered like the Moon, is a planet that appears to tell two conflicting stories. One, of a small planet that was struggling for a long, indefinite period to have a climate like Earth; and another, of a planet battered by asteroids that lost all but a thin atmosphere early in its history and has been barren and Moon-like ever since. It has been found that in the middle of this planetscape of past desolation and destruction, there appears to be a carved humanoid face. What do all of these clues mean? In particular, is the face a spurious occurrence, irrelevant to solving the puzzle of Mars' past, or is it perhaps the key piece?

After many years of studying the whole array of data concerning Mars, we have decided to advance the hypothesis that the Face on Mars is in fact the "Rosetta stone," the key piece of evidence for understanding Mars' past climate. This hypothesis is termed the Cydonian Hypothesis since it hypothesizes that the face is in fact an image of a Cydonian, an extinct humanoid race indigenous to Mars who, before they perished, carved the face and created other objects found in Viking images.

The Cydonian Hypothesis

The hypothesis, that Mars, the home of a long-lived biosphere similar in scope and diversity to that of Earth, and that out of this biosphere evolved an indigenous humanoid race, termed Cydonians, that constructed large mon-

2 J. E. Brandenburg, V. DiPietro, and G. Molenaar

uments similar to those constructed by Old Kingdom Egypt, is based on three main points:

1. The Assumption of Mediocrity (Sagan & Shklovskii, 1960). The As- sumption that neither the Earth nor its biosphere nor its humanoid deni- zens nor the civilization and artifacts they have produced are unique or even remarkable in the Cosmos. By this assumption, the discovery of a dead civilization on an Earth-like planet such as Mars would not be surprising.

2. Images of the surface of Mars showing, at several sites what appear to be three carved humanoid Faces (Brandenburg & DiPietro, 1986), of kilometer scale, and having similar anatomical and ornamental details between all three. Appearing with these objects are numerous other objects and surface features that resemble Earth-like archaeological ruins, of a Bronze Age culture, with no evidence of advanced technology or civilization.

3. Geological and geochemical data that are consistent with past conditions on Mars that were favorable to Earth-like life forms: Abundant liquid water (Masursky, Boyce, Dial, Selaber, & Strobell, 1977), and an atmosphere that was dense and warm, and possibly rich in oxygen (Toul- min I11 et al., 1977).

Modern Searches For Civilizations On Mars Because of its close proximity and similarity to Earth, Mars has always

been the subject of speculation about its capacity to support life and intelligence. Sagan and Fox (1 975; Sagan & Wallace, 197 1) examined the first high-resolution photographs of the Martian surface acquired by Mariner 9 for signs of a civilization of our technological level and extent. They were guided in this search by images of Earth at similar resolution (Sagan & Wal- lace, 197 l). These pictures of Earth showed human civilization clearly. How- ever, in the images of Mars, no signs of a civilization of our technological level and extent were found. Furthermore, in the Mariner 9 images, no objects that were strongly suggesting civilization of any known type were found. Other investigators reported objects resembling ancient pyramids of large size (Figure 1) in Elysium Planitia. When the Viking probes obtained high-resolution pictures of Mars in 1976, however, an object was found by Dr. Tobias Owen which resembles a mammoth carved head; this object was dubbed the "Face of Mars" (Soffen, 1980), (Figure 2).

Two of the authors (DiPietro and Molenaar) studied the original image from the Viking frame 35A72 and discovered a second image on frame 70A 13. They enhanced the images using digital methods and copies of the raw data tapes, which were supplied by the Jet Propulsion Laboratory in Pasadena, California. The results of their investigations were published at a Mars science conference (Oberg, 1983) and in a series of pamphlets (DiPie- tro & Molenaar, 1 9 80; DiPietro, Molenaar, & Brandenburg, 1 9 88). A subse-

The Cydonian hypothesis 3

Fig. 1 I bc pyramids of Elysium inlaged by Manner 9 In 197 1 arc indicated by arrows at sunset on DAS 07794853 From C'nzr\zrtrl l lar~lul l S1crflzc.c F(;(rttrri~c Edrrion /bT(L',ifkYF4), p. 46, (DiP~etro, Molenaar, & Brandenburg, 1988). Copyrrght 1988 by Molenaar, Tnc. Re- prrnted 'b! p c ~ m14sion

Jig. Z. The face of Mars discovery kame 35A72 found by r o b ~ a s Owen. From (IMLSF4, p. I?. (DiPletro, Molenirar, & Brandenburg, 1988). Copy]-ight 1988 by Molenaar, Inc. Re- printed by pernlission.

4 J. E. Brandenburg, V. DiPietro, and 6. Molenaar

quent investigation. called the Independent Mars Investigation Team (IrVIIT) (Pozos, 1986), confirmed the work of DiPietro and Molenaar and studied the geochemical data pertaining to Mars to determine the likelihood of a past Earth-like biosphere on Mars (Beatty et al., 1984). Faces and other strange objects at other sites on Mars were found (Figure 3) , and their similarity of sire and detail to the Face of Mars was discovered and presented at a scientific conkrence (Brandenburg & DiPielro, 1986). Finally, other investigators have studied the Cydonia site and published work concerning it (Car- lotto, 1988; Carlotto & Stein, 1990; O'l,eary, 1990).

Overview

In the reminder of this reporl, the basic imaging data will be presented, and a brief overview of the geochemical issues will be given. We will discuss

Fig, 3. A map of Man showing the locatlonc ofobjects of ~ntcrest lnvestrgated by the a ~ i t h o r ~ . ( 1 ) Cydon~a Mcnsa (2) Deutert~nilus Mensa (3) Hecates Tholu\ (1 Jtopia) (4) The pyramids of t l y s ~ u m


the formation of the Cydonian Hypothesis and its testability, and alternative hypotheses.

T/ze CTyj?donia Sitc~. 'The Cydonia Mensa region of Mars is marked by a range of small mountains and mesas that runs Norlh at approximately the low plain-highlands boundary, and then makes an abrupt nght turn out into Acidalia Planitla (Figure 4). The Cydonia Mensa has been extensively photo mapped (Frey & l,owry, 1979; Guest, Buttenvo~h, & Creely, 1977) and shows signs of ab~indant water in the past. The site of interest lies Just above the Zero Kilometer elevation line of Mars (Bahson, Bridges, & Inge, 1979). At the corner of the range of nlesas, at 9 degrees W, 41 degrees N lies the object known as the "Face of Mars." ""The Face of Mars" here referred to as "the Face in Cydonia," or simply the ""Face" appears in the Viking frame 35A72 taken at a range of 1,873 km, and 70A 13 taken at a range of 1724 km. Both of these frames were taken during the course of photo mapping in the Cydonia region in the afternoon. DiPietro and Molenaar used digital techniques to enhance the images, and used a version of bilinear interpolation, which they developed and call the Starburst Pixel Interleaving Technique (SPIT) process, to enlarge and smooth the images, A thorough discussion of

I lg 4. A rnap of the Mare Acadallurn reg~on showing the location of the face of Mars. Note the location of the 0 krn elevation contour (Martian 5ea level) marked by an arrow.

6 J. E. Brandenburg, V. DiPietro, and C. Molenaar

Fig. 5 . The enhanced images of the Face produced by Dr. Mark Carlotto from both Viking frames 35A72 and 70A i 3, From UMSF4, pp. 90-9 1, (DiPietro, Molenaar, & Branden- burg, 1988). Copyright 1988 by Molenaar, Inc. Reprinted by permission.

these techniques is given in a booklet called U P Z I ~ S U Q ~ MurLs Surface Feutz4rt.s (4th ed.).'

Further enhancement of the images was done by Dr. Mark Carlotto of The Analytic Sciences Corporation (TASC) in Reading, Massachusetts (Carlotto, 1988), using a technique that be has developed. His images appear to show teeth in the mouth area of the Face (Figure 5). Mark Carlotto and Brian O'Leary have also published further analyses of t k Face (Carlotto & Stein, 1990; O'Leary, 1990).

- - - -

This pamphlet is available for $15 by writing to: Mars Research, P.C>. Box 284, Cilenn Dale, MD 20769.

'The Cydonian hypothesis 7

The basic dimensions of the Face are 3 km from the chin to the top of the headdress or helmet, and 2 krn fronl one side of the headdress to the other. The Face appears to have numerous anatomical and ornamemital details, including eyes, nose, nlouth, a helmet or headdress, as well as cheek ornaments, and ari illdentation over the ~ g h t eye. Several other images of the Face in Cydonia were discovered (Pozos, 1986), all at ~nuch lower resolution: these are 673B56 and 673B54 at nearly the same sun angle as 70A 13, at the local afternoon time; and 753A33 and 753A34 both taken at the local morning time. The latter two images are remarkable in that they show the Face illuminated fiom the right side rather than the left. These in~ages support the premise that the Face is basically symmetrical. That is, it is a dome-shaped object, although the poor resolution and picture quality of the images limit their usefulness In the study of any details.

Fig. 6. (A) Images of the pyramid in Cydon~a from three different Viking frames. (B) An enhanced version of frame 70A 13 by Dipietro and Molenaar showing relative locations of the face and pyramid. From UllfSF4, pp. 39 & 44, (DiPietro, Molenaar, & Brandenburg, 1988). Copyright 1988 by Molenaar, Inc. Reprinted by pemission.

8 J. E. Brandenburg, V. DiPietro, and C. Molenaar

DiPietro and Molenaar also discovered an object appeafing like a five- sided pyramid. The pyramidal object appears in 3 different Viking frames, and lies 10 km from the Face (Figure 6). The dimensions ofthis p y r a ~ ~ ~ i d are 3 km on each facet, and 1 km high. These dimensions are similar to the large pyramids at Elysium. The exact symmetry of the object, if it has a syrnmetfy, is dificult to assess due to a landslide or1 one parl of the object.

Thc~ Ut(1p1~. SIIO. The Utopia site was originally found as part of the IMIT investigation by looking for a site similar to Cydonia Mensa. The site in Cydonia Metlsa is found at the boundary of the low nor-thern plain system of Mars and the rugged highlands, It was located at slightly a h v e the Lero kilometer elevation contour that follows this plains-highlands boundary in the northern hemisphere. The Cydonia site was locally prominent and shows evidence of abundant water in the past. If the objects found at the Cydonia site were signs of a dead civilization, such a civilization might have created similar objects at other sites on Mars that shared charactensties with the Cydonia site. Following this "archeological site n~odel," Brandenburg chose a site on the edge of Utopia Planitla fbr investigation.

The Utopia site was found by following an ancient water channel known as Hrad Vallis up to its "head waters9' at the base of Heccates Tholus (Figure 7). T11e site is characterized by a range of clliN\s and outlying rnlesas bordering

Fig. 7 A map of the Ut~pia reglon of Marc, rhtrwing the locatron of the face\ In CJtopta. An ar rot& ~ ~ i a r k s the course of Ilrad Vallis, atid old water chdnnel.

what may I~ave bcen a large bay or lake. 'This area has been sttrdied by Mars geologists because it shows signs of extensit~c permafrost (Can & Schaber, 1977)-

'I'wo object5 having ;I detailed resen-rblanee to the Face in Cydonia were h u n d at the Utopia site on Viking fl-ame 86A I0 (Izigure 8). This frame was haken at a range of 1,576 k n ~ , a closer range than in Cydonia, 'The objects are slightly smaller thar-r l l ~ e objects in CCgidorlia at rrpproxirnately 2 k n ~ from the chin to the top of "shc head and 1.5 km across. As shown in Figure 9, the

Fig 8 A porqlon of kame 86A10 on whrch two objects re5cmhllng facts were found Note the prewnce ol'~ndent,~trons orr the eheehc, and rrbovc the left ele of the Fdces.

10 J. E, Brandenburg, V. DiPietro, and G. Molenaar

Fig. 9, Comparisons of the faces in Utopia with the Face in Cydonia imaged on frames 70A13 and 35A72. Similarities noted: (A) cheek notches (B) indentation above left eye (C) helmet or headdress.

objects appear to share details with the Face in Cydonia shown in Figure 5. While the objects are not as startling as the face in Cydonia, we have included them in this discussion because they resemble the Face, are of similar size, and lie in close proximity to each other, suggesting similar forces may have

The Cydonian hypothesis 1 I

Fig. 10. A digitally smoothed, enlarged, and contrast stretched version of the upper Utopia "Face" seen In Figure 8A. This image was obtained from a NASA supplied data tape of Viking frarne 86A 10. White and black starbursts are produced by static in the original data. This object, though not as startling In appearance as the Face in Cydonia, appears to share many of its characteristic"; Note the presence of dark central spots in the eye sockets, as well as the helmut structure that appears to surround the head. Note also the apparent indentation over the left eye and cheek mark below it.

shaped all three ob-jects. The existence of the Utopia Faces is thus relevant to discussions of whether the Cydonia Face is artificial, and who might have carved it.

Images of the two Faces have been enlarged digitally, smoothed, and contrast stretched by the authors. These processed images reveal what seem to be hnimanolid anatomical details (Figures 10 and 1 l ) such as eye pupils and a jowl line. Also, cheek ornaments and an indentation over the left eye appear to be shared by the face in Cydonia. (Figure 9). 'The sun angle of the Utopia picture, close to noon, and the objects smaller sizes make it difficult to improve on the NASA versions of the images, since contrast stretch seems already near optimum in the NASA shots and the pixel size is larger compared to the face size, relative to the Cydonia Face. The Faces appear less distinct, which could mean they are eitber an intermediate form of an erosion formed face or also are true artifacts that are simply smaller and more eroded than the Face in Cydonia.

The lower Face is imaged at moderate resolution with light coming from the bottom on Viking frame 243S01, and both Faces are imaged on Viking frame 54 1A14 at moderate resolution. These images confirm our under-

12 J. E. Brandenburg, V. DiPietrs, and C. Molenaar

FIR 1 1. A digitally smoothed, enlarged, and contrast stretched verslorl cjT the lower litopia Face seen in Figure 8. Wh~te and black starbur-rt are produced by stat~c in the onglnal data. This iinage was obtained from a NASA supplied data tape of V~krng Game 86A10. Note the apparent jowl Ime, prominent cheekbone or mark and an ~ndenlation over the left eye. Note also the apparent dark central areas of the eye sockets.

standing of the basic shape and symmetry of the Faces. Additional low-resolution images are found on 844A 13 and 846A 14.

Additional surface features, appearing to be the result of intelligent activity, are found in the irnn~ediate vicinity of the Faces (Figure 12).

The Deliferonilus Sitt'. Deuteronilus Mensa forms a series of mesas bordering Vastitas Borealis. Near the mouth of the Mamers Vallis is a very flat region that is above the zero kilometer elevation contour (Figure 13). On this plain are many so-called "pedestal craters" thought to be formed by meteorite impacts into water-logged soil (Carr, Crumpler, Cutts, Creely, & Ma- sursky, 1977). One pedestal crater attracts parlicular attention in this area because it is associated with an object that is higher than any landform for a 100 krn radius. This object is located at 353 degrees W and 46 degrees N. The object is imaged on three Viking frames at high resolution, 43AOI, 43R03, 43A04 (Figure 14). The ranges at which these frames were taken are listed as 2,109 km, 2,108 km, and 2,093 km, respectively. This makes the resolution of these fi-ames slightly poorer than those taken in Cydonia and Utopia, Low-resolution images of the object are found on Viking frames 673B38 and 675B53 under much diKerent lighting conditions and are very important because they allow cross-checking of models of the objects9 geometric structure from shadowing. The suggestion that this object might owe its appear-

'l'he Cydot~ian hypothesis 13

Erg. 13. Objects imaged on 86A08 near the Urop~a Faces. What appears to he a ramp lead~ng up to the top ofa mesa 1s seen with large rubble at the edge ofthe mesa. A long linear feature that resembles an aqueduct or wall is also seen on the ground ~learby. Dark clrcle IS

merely a water spot on lens.

ance tu intelligent actiivity was first made 'by Richard Hoagland dur-ing the IMIT investigation (Pozos, 1986). The object is geologically anomalous and would completely dominate the landscape as seen by ground observers. The object is on, or i m e r s e d in, the debris apron of a large crater, yet it does not appear lo have deflected or disturbed the flow of debr~s (bigure i 2) . l his would suggest that the object was mplaced AFTER the crater-forming impact. NternaLlvely, the debris could have fallen in a blanket rather than jetting outward from the impact. Tlzis would leave only the base of the object burled and would allow the object to predate the impact. No similar feature is seen on other debris aprons of pedestal craters in the neighborhood, although such craters are commonplace in this region. The only object even remotely similar to this feature, and known to the authors, is in the Cydonia Mensa region 30 km northeast of the Face in Cydonia. This object is called simply the "wall" (Figure 16). Like the Deuteronilus object, it was found by Richard Hoagland, and appears to have been ernplaced after the cratering impact. It is at righmwles to the debris flow, yet does not disturb it.


Fig. 13. A map of the Isrnenius Lacus reglon of Mars showing the locatlon of the object at Deuteronilus, also callcd 'Thc Crater Pyramid." 7 ' 1 ~ object 1s found to be the tallest feature for a radius of 100 km. From CrZISI*'4, p 29. (Batson, Bridge\, Rr Inge. 1979). Copyright 1979 by Molenaar, Inc, Reprinted by permlsslon.

The Deuteronilus object's true shape is dificult to discern from its appearance alone, because its albedo (reflectivity) apparently varies strongly across its surface. The highest point on the object, approximately 0.6 km higher than the surrounding plain, is not the vertex of the triangular brightly lit region. The highest point on the object is known Gom shadow rneasure- ments to be in the low albedo region as is shown in Figure 15. The object is approximately 3 km in its lateral dimensions, making it rather flat and dome shaped. The pointed appearance of the shadow is due to the extremely low sun angle (84.5 degrees from verlical). A refined understanding of the object's shape is elusive because of the low resolution of the images.

In a manner similar to the other sites, other surface features at the Deuter- onilus site suggesting intelligent activity are found in the nearby vicinity of the object: A large area of crosshatched walls or embankments is seen on the debris apron of a nearby crater as is shown in Figure 15. Such lines are not found on other crater debris aprons in the area. Also, a series of embankments or albedo variations ("Turnowed Ground," Figure 14) is found on a nearby region of higher ground that resembles agricultural terracing. Purely geological explanations such as permafrost features (Can- & Schaber, 1977) are also possible.


Flg. 14. Viking Images showing the object In Deuteronilus and its surroundings. Arrows mark the object and a nearbv feature called ""Furrowed Ciround" that resembles agricultural terracing. Z3icture I 7 3 ~ ~ 1s from tne 331% series ot irnages i.akcn on a airleicat orbrt. From L'lfSF4, p. 107, (D~Pietro, Molcnaar. 8L Rrandenburg, 1988). Copyright 1988 by Molenaar, Inc. Reprinted by permirslon.

The firmation qf7a Hjy~otllzc~sis To be useful, a hypothesis must satisfy Karl Popper's requirement of falsi-

fiability. That is, it must be testable. Given the data presented in the previous sections, three hypotheses seem admissable:

( 1 ) The Nzdl Ifl~potlrt~sis: The objects discussed are the result of random geological and erosional

forces. The apparent resemblance of the objects to carved humanoid Faces, and other archeological monuments found on Earth, is both fortuitous and a

16 J. E. Brandenburg, V. DiPietro, and 6. Molenaar

Flg 15. I hc "Crater Pyramid" rs reen, rn thl\ ~ ~ n \ a r g c n ~ e n t f r o i n 43A04, as in object ca'it~nh; ;I long shadow Shadow ;rn,ilyar\ S I I O ~ V G C elldt the object 15 .rciu,ill\ dornc 5En;apcd Sun angle I \ n~arked by an asaow. A n iir1o-w SO ~~a~trk",II?c 10cLifro11 01 the ''cross h;atchrng'- ncal the object

sign of the human tendency to find flnailiar patterns in clata even when they are rrot precent and tc-, reject data that dcres not fit the familiar yattcrn. Phis rs also known as the fallacy of the enumeration of favorable circurmctances. The hct that the oh-iects are similar in size and geometry is due cprlly to the fact that all are fornnd in thc plains-highlands bransititan rcglon of Mars where mesa outlyers and kr~obs oftlais size and shape are ubiquitous. -l'his hypothesis is not strictly falsifiable but is complementary to the other hypotheses.

( 2 ) 7-/1c C\*cr"r~t~rlx.n H j -~~o f / 1091 '~ : T'he Face of Mars 1s a portr~iit o f a Cydonian, that is, the oh-jects discussed

are a product o f a humanoid civiltiation indigenous to the Cydonia region of Mars, Mars was once the home s f an indigenous race of intelligent Inurnan- oid beings which cons"lructcd rnc,nuxnents similar to those found in old Kingdon1 Egypt. The motivation for this corastrluctic~n of large Faces and Pyramids was sln~llar to the God-King worship of ancient Egypt. 'The Faces thus resemble the Cydonlans tlaemselves. U ncJer the Cydonian Hypothesis, the objects look Fdmiliar to us kccause tiley were constructed by a race of beings similar in appearance and psychology to ourselves, Such a hypothesis is totally in keeping with thc Assun~ption of Mediocrity. 'The similar-ity in

'Pke Cydoniam hypothesis 17

detail a r~d size found between the Faccs is merely a reflection of extensive communicatian and cultural equiiibration between the centers of tIae hypothetical civilization in the northern hemisphere of Mars. This hypothesis requires Mars to have had a long-lived Eadh-like biosphere to allow the formation and evolution of indigenous intelligence. Such a hypothesis also requires the death of a planetary biosphere, since Mars i s presently hostile to Earth life. The hypothesis is testable by a closer inspection of the objects found at the sites, arsd by a more complete understanding of Mafiran climatic history.


(3) The Prior Colonization Hypothesis: The objects in question owe their appearance to a culture that was not

indigenous to Mars. Such a hypothesis does not require a long-lived Earth- like biosphere to have been present on Mars nor its subsequent death. Since the civilization would possess capabilities that we do not or have not pos- sessed in recorded history, i.e., colonization of other planets and interstellar travel, its nature and cultural forms would be unknown, as would its motivations for construction large humanoid Faces. Such a hypothesis is very difficult to test, since it involves so many unknowns. In particular, reimaging of the Face in Cydonia at higher resolution might not reveal more detail if the object was only crudely "bulldozed" into the Martian surface to be viewed from space. This hypothesis is not favored by the authors, however, it has been proposed by other investigators (Hoagland, 1987). However, this hypothesis does not require a long-lived Mars biosphere and thus would be supported if the objects appear to be artificial but no sign of such a past biosphere is found.

Discussion

The hypotheses listed in the previous section have the virtues that they explain what is seen on Mars in a plausible manner, and do not contradict known facts. However, in the opinion of the authors, only the first two hypotheses are truly testable. For this reason, we will restrict our discussion primarily to the Cydonian Hypothesis and its complement-the Null Hy- pothesis. The Cydonian Hypothesis is testable because it states that processes that occurred on Mars are similar to those that have occurred on Earth. Because of this, the next generation of space probes to Mars may gather enough data to provide significant support for either the Null or Cydonian Hypothesis. In particular, the Mars Observer Spacecraft could reimage the Face in Cydonia at much higher resolution than the Viking images, and perhaps reimage the objects at other sites as well. Archaeological monuments found on Earth almost always display more detail at higher resolution, even when eroded. This is because the objects of Earth were meant to be viewed from close range on the ground. If the Face in Cydonia does not display greater detail in images at higher resolution than the Viking images, then the Cydonian Hypothesis would be considerably weakened, if not re- futed. Other hypotheses would then have to be considered more likely, such as the Null Hypothesis, or the possibility that the objects where constructed to be viewed from space, but not highly finished so as to give an impressive appearance from the ground, such as under the Prior Colonization Hypothe- sis. It has been pointed out by other investigators that the fact that the Face faces upward suggests it was meant to be viewed from above. This might indicate it is a sign of an advanced civilization capable of flight. However, construction of such a large face is only possible if it faces upward, regardless of technological ability of the constructing civilization and on Earth the


presence of patterns at Nasca shows that even low-tech civilizations will construct objects that can only be fully appreciated from above. Therefore, the fact that the Face faces upward does not weaken the Cydonian Hypothe- sis, but lack of new detail in high-resolution images would weaken it. If, on the other hand, brickwork, stairways, or writing on the objects became visible in higher resolution images, then the Cydonian Hypothesis would be greatly strengthened. If, in addition, evidence of a long period of favorable climate and biosphere on Mars, such as coal or petroleum deposits, deep sedimentary formations, and fossils were to be detected by the Mars observer or other probes, then the Cydonian Hypothesis would also be strengthened.

The fact that, at present, Earth provides the only known example of a civilization in space, limits our ability to form testable hypotheses concerning possible civilizations on other planets. All present searches for signs of extraterrestrial civilization rely on an Earth reference. That is, the Assump- tion of Mediocrity is always implicit even if it is not stated. For this reason, any statement that an object looks like an artifact really means that it resembles artifacts found on Earth. There is, therefore, no truly "generic" test for intelligent origin of an object at this time. Since the civilization of Earth is indigenous and artifacts found here represent only the products of low to present technological levels, products of a more advanced or truly alien culture might not be recognizable to us. Someday, data on extraterrestrial civilizations will be available to assist us in the search for signs of civilization on other planets. Data from several civilizations, including our own, could be used to create a generic test for intelligent origin of an object seen on a new planet. However, testable hypotheses concerning extraterrestrial artifacts are, presently, restricted to processes that are known, and this means they must involve processes that are terrestrial.

The Null Hypothesis says that the objects on Mars were shaped by geologic processes similar to those known on Earth. Wind and water erosion, faulting, and meteor impacts are known to shape landforms on Earth and some of these landforms resemble human artifacts even though they are natural. The objects in Cydonia are found in an area that abounds in landforms of roughly similar size and shape. These objects are called "knobs" and mesas. The details of the "Face" that distinguish it from its neighboring landforms could have been formed by a series of meteor impacts, landslides, and faulting events that produced somewhat symmetrical facial features by chance. The probabilities of this occurring seem remote and are difficult to model, however, the number of knobs and mesas is large on Mars. There- fore, it seems reasonable that out of all the pictures taken of this and similar regions of Mars, the probability of finding one object resembling a face would be high. The same can be said for the nearby "Fortress" and "Pyra- mid." However, the geologic forces that would create these objects are non- local so one would not expect them to produce an object as different from a face as a pyramid, yet so near to it. Similarly, the geologic forces are blind, so


tions, and anatomical detail. For these reasons, the Null Hypothesis has difficulty explaining the association of the "Face" and "Pyramid" in Cy- donia and their degree of detail.

Accordingly, the occurrence of other "Faces" in Utopia with similar size and details seems an unlikely event if only geologic forces are involved. Like the Cydonia site, the Faces in Utopia are associated with other objects that look like artifacts, such as the linear feature found in an adjoining image. The object found in Deuteronilus also looks like an artifact and similarly to the objects in Cydonia and Utopia, is associated with other features, the "Furrowed Ground," whose appearance leads one to believe they might be artificial. Therefore, it is not just the objects themselves, but their similarities of form and association which require alternatives to the Null Hypothesis to be considered. This leads us to consider the Cydonian Hypothesis, which proposes that the objects are artifacts.

Given evidence of Earth-like conditions in Mars past, the presence of ancient archeological monuments on Earth that resemble the Mars objects, and the lack of signs of advanced civilization similar to Earth's at the sites, the Cydonian Hypothesis seems the simplest possible hypothesis involving an artificial origin of the "Face" and other objects on Mars. Like the Null Hypothesis, it involves only processes demonstrated on Earth. Like the processes involved in the Null Hypothesis, however, it is difficult to calculate the probabilities of the processes involved, or the exact way by which these processes produced the objects in question. That is, the processes are poorly understood but they are known. The main virtue of the Cydonian Hypothe- sis is that it can be tested, since it involves only known processes.

The Cydonian Hypothesis states that Mars was once Earth-like and remained so for a long period-long being enough time, billions of years-for something like humanity to appear. This would necessitate a planet with large amounts of liquid water and, at some point, an oxygen rich atmosphere sustained for a prolonged period of time by photosynthesis. Such an environment on Mars would leave abundant, though perhaps subtle, clues to its past existence. The evidence both for and against this aspect of the Cydonian Hypothesis is worth discussion. Mars has an abundance of water channels indicating past conditions of a warm dense atmosphere (Masursky, Boyce, Dial, Selaber, & Strobell, 1977). Mars is red; this redness is due to highly oxidized iron in its soil (Toulmin 111 et al., 1977). It has been suggested by Huguenin that this oxidation was due to oxygen from water released by the action of ultraviolet light on the water (Huguenin, 1974). An earlier suggestion by Carl Sagan was that this high oxidation state was produced by an oxygen atmosphere produced by photosynthesis (Sagan, Phanouf, & Ihnat, 1965). The apparent presence of superoxides in the soil (Oyama & Berdahl, 1977) tested by the Viking landing indicates that some process of oxidation is occurring now on Mars, although such a process cannot preclude any earlier period of photosynthesis.

Mars shows evidence of having an ocean in its past (Brandenburg, 1986; Parker et al., 1986). This ocean would have filled the northern plains region

'The Cydonian hypothesis 2 1

Fig. 17 Mars as tt may once have looked. An ocean nnged the north portlon of the planet with smallcr 9eas in the southern reglon. The nolqhern region probably had a molst, temper- ate climate whlle the south was mostly decert. The sltes of possible archeological re- maln5 appear to have been on the shorel~ne of the northern ocean, which was probably near the 0 km cle\ ;it1011 contoilr

of Mars and formed a rlrlg around the northern polar cap (Figure 17). The approxir11ate s2lorelinc ol'the ocean appears to have been the zero kilometer elevation contour. 'l'he sltes ofapparel-rt civilization appear to have been on the coasts of this ocean. 'The past exiistence of an ocean on Mars argues for a Mars that held some form of life since life began in the oceans of Earth very quickly afier they formed. Therefore, there is geochemical evidence that supports the Cvdonian I-lypothesis; however, there is also evidence that seems to argue against it.

The inost powerful argument against the Cydonian Hypothesis is the estimates of the age of the Marlian surface by crater count dating (literally counting meteorite craters). Estimates made using this method indicate Mars surf'ace ages from 3 billion--5 billion years old (Mazursky, Boyce, Dial, Selaber, Br. Strobel, 1977). Tllis method conlpares the number of craters on a given area of Mars wit11 that on a given area of the Moon. Our astronauts brought back sainples of rock from areas of the Moon and we have radioac- tively dated them. So we know that a cerlain number of craters per given area on the Moon translates to a given age, There are fewer craters on Mars than on the Moon because of erosion, but we can still get an estimated age for a place on Mars by counting craters and comparing it to the Moon.

2 2 J. E. Brandenburg, V. DiPietro, and G. Molenaar

A three billion year estimated age for Mars surface would mean that any period of Earth-like conditions would have lasted only a billion years. This is far too short by Earth standards for any advanced life to appear. If the cratering derived ages on Mars could be calibrated by sample returns from Mars and verified at this three billion-year-old value, then the Cydonian Hypothe- sis would be clearly weakened. One would have to suggest a highly acceler- ated evolution on Mars relative to Earth, and the Cydonian Hypothesis argues for a similar process. Thus, commonly suggested interpretations of cratering data suggest that any past-life bearing period on Mars would be short. Mars looks heavily cratered like the Moon, so how could it have once lived?

Methods of crater counting to determine the age of the Martian surface are built on a crucial number: an estimated magnitude of the rate of cratering on Mars relative to that on the Moon. This estimated magnitude is the key to the whole scheme, and it is full of uncertainties. Mars is near the asteroid belt, the source of most meteorites, much closer to it than the Moon, and so one would expect that the rate of cratering impacts by asteroidal rubble would be somewhat higher-but how much higher? A higher estimated rate of cratering means a younger Martian surface and thus a longer period of erosion and Earth-like conditions. Such a younger Martian surface would support the Cydonian Hypothesis, since a long-lived biosphere on Mars could support an Earth-like evolutionary time scale of three billion years. A recent model "Neukum and Hiller 111" (Neukum & Greely, 1988) indicates Earth-like conditions may have existed on Mars until .5 billion years ago. Such an estimated age strongly supports the Cydonian Hypothesis and is based on estimates that the Martian cratering rate is high relative to the Moon.

Two pieces of evidence suggest a higher cratering rate and thus suggest a younger Martian surface. The first is the discovery of pieces of Mars that have fallen to Earth as meteorites. These meteorites are termed the Shergot- tite, Nakhlite, and Chassigny (SNC) meteorites (Vickery & Melosh, 1987) and have been found along with pieces of lunar material recovered as meteorites. In both cases, these meteorites have come to Earth because they are secondary fragments blasted off of the Moon or Mars by meteorite impacts on those bodies. The fact that Mars has a much higher escape velocity and is further away from Earth than the Moon, plus the fact that much more Mar- tian material than Lunar is recovered, argues that the Martian cratering rate must be much higher than that on the Moon. This would mean Mars surface is younger than it looks, it is simply being hammered by meteorites so it looks old.

The second piece of evidence that supports high-cratering rates and, hence, younger Martian surface estimates is the discovery of deep sediments on Mars in the Martian canyon system (Nedel & Squyres, 1986). These sediments are 5 km thick. This is very thick by Earth standards. Sediments exposed in the Grand Canyon are only 2 km thick and go down to rock formed at the origin of Earth's oxygen atmosphere. This would again argue

The Cydonian hypothesis 2 3

that long periods of erosion on Mars, similar in scope and duration to those seen on Earth, must have occurred. However, these arguments concerning the SNC meteorites and the sedimentary layers that support erosive history on Mars are presently quite controversial, although not as controversial as the objects in Cydonia and Utopia.

In any case, the view of Mars' past as Earth-like for long periods, that is implicit in the Cydonian Hypothesis, can be accommodated within our present understanding of Mars,.given its present uncertainties. However, it can be accommodated only with difficulty. In general, the issues central to the validity of the Cydonian Hypothesis are also central to our understanding of the past climate and atmosphere on Mars. For this reason, the Cydonian Hypothesis is testable in the near term, as the exploration of Mars continues and answers to the riddle of its past climate are answered.

Summary

The Cydonian Hypothesis and a brief summary of the data that supports it has been made. The Cydonian Hypothesis seems bold and perhaps even speculative at this time, for it hypothesizes something unprecedented: that a race similar to humanity once lived on a nearby planet. However, the fact that the hypothesis seems bold is only an accident of this present time. At some point in the future, we will know of many other civilized species besides humanity occupying the Cosmos either past or present. In that future time, responding to the sight of an object resembling a large carved face on some distant planet with hypotheses of its intelligent origin, will not seem bold, but obligatory. Looked at from that future perspective, the Cydonian Hypothesis will be viewed as merely part of an early period of human igno- rance about the universe we dwell in. This period will be marked by many successful and unsuccessful hypothesis. In this sense, the Cydonian Hypoth- esis is like all hypotheses, it is a question.

The Cydonian Hypothesis is a response to the facts that objects resembling Earthly archeological movements have been found on Mars surface and past Earth-like conditions on Mars cannot be ruled out at this time. Based on these facts, the Cydonian Hypothesis seems the simplest hypothesis that can be presently formulated. It is the simplest because it hypothesizes processes known from Earth occurred on Mars, rather than unknown processes. Thus, it is only the new locale of the processes not the processes themselves which are being hypothesized. Humanity will soon send new probes to Mars and reimage the Face on Cydonia and other objects and gather other data relevant to these discussions. Hopefully, the data gathered by these probes will provide strong evidence supporting one of the hypotheses discussed.

In anticipation of the eventual arrival of such new data, the Cydonian Hypothesis has been articulated; that is, the reason Mars is red, and covered with old water channels and deep sediments, and in some places shows what appear to be large carved humanoid faces, is that Mars once lived. It lived even as the Earth lives now and it perished even as the Earth could perish


now if we are not better stewards of it. We believe this hypothesis to be the simplest explanation for the appearance of these features of Mars that we have found. This hypothesis is falsifiable. Therefore, let anyone who can, put it to the test.

References

Batson, R. M., Bridges, P. M., & Inge, J. L. (1979). Atlas ofMars: The 1:5,000,000 map series, p. 25-38. Washington, D.C.: U. S. Government Printing Office.

Beatty, W., Brandenburg, J. E., DiPietro, V., Dolphin, L., Hoagland, R. C., Pozos, R. R., & Rautenberg, T. (1 984, July). The preliminary findings of the independent mars investigation team, new thoughts on unusual surface features. Proceedings of the Case for Mars 11 Confer- en

c

e. pp. 33-34. Boulder, Colorado: University of Colorado at Boulder. Brandenburg, J. E. (1 986). The paleo-ocean of Mars. In Symposium on Mars: Evolution of Its

Climate and Atmosphere (V. Baker et al., eds., pp. 20-23). LPI Tech Rpt. 87-0 1. Houston, Texas: Lunar and Planetary Institute.

Brandenburg, J. E., & DiPietro, V. D. (1986). Observations concerning some unusual surface features in the Cydonia, Deuternonilus, and Utopia regions of mars. EOS, 67, (16), 402.

Carlotto, M. (1 988, May). Digital image analysis of unusual Martian surface features. Journal of Applied Optics, 27, 1926-1933.

Carlotto, M., & Stein, M. C. (1990). A method for searching for artificial objects on planetary surfaces. Journal of the British Planetary Society, 43, 209-2 16.

Carr, M. H., Crumpler, L. S., Cutts, J. A., Greely, R., Guest, J. E., & Masursky, H. (1977). Martian craters and emplacement of ejecta by surface flow. Journal of Geophysical Re- search, 82, 4055-4076.

Carr, M. H., & Schaber, G. G. (1977). Martian permafrost features. Journal of Geophysical Research, 82, 4039-4053.

DiPietro, V., & Molenaar, G. (1982). Unusual martian surface features, 3rd ed. Willmar, Min- nesota: Molenaar Inc. Press.

DiPietro, V., Molenaar, G., & Brandenburg, J. (1988). Unusualmartian surface features, 4th ed. Willmar, Minnesota: Molenaar Inc. Press.

Frey, H., Lowery, B., & Chase, S. A. (1979). Pseudocraters on Mars. Journal of Geophysical Research, 84, 8075-8086.

Guest, J. E., Butterworth, P. S., & Greeley, R. (1977). Geological observations in the Cydonia region of mars from Viking. Journal of Geophysical Research, 82, 4 1 1 1-4 120.

Hoagland, R. C. (1987). The monuments of Mars, A city on the edge of forever Berkeley, California: North Atlantic Books.

Huguenin, R. L. ( 1974). The formation of geothite and hydrated clay materials on Mars. Jour- nal of Geophysical Research, 79, 3895-3905.

Masursky, H., Boyce, J. M., Dial, A. L., Selaber, G. G., & Strobel, M. E. (1977). Classification and time of formation of martian channels based on Viking data. Journal of Geophysical Research, 82, 401 6-4038.

Nedell, S. S., & Squyres, S. W. (1986). Formation of the layered deposits in the Vallis Marineris, Mars. In Symposium on Mars: Evolution ofIts Climate andAtmosphere (V. Baker et al. eds.) pp. 20-23. LPI Tech Rpt. 87-01, Houston, Texas: Lunar and Planetary Institute.

Nuekum, G., & Greely, R. (1 988). Mars sample return and cratering chronology models: Conse- quences for the Martian climatic history and landing site selection. In Proceedings of the 19th Lunar and Planetary Science Conference, (Tanaka, K. L., ed., pp. 852-853.) Houston, Texas:Lunar and Planetary Institute.

Oberg, J. E. (1983). Mission to mars, pp. 108. First Meridian Printing, New York: New Ameri- can Library.

O'Leary, B. (1990). Analysis of images of the "Face" on Mars and possible intelligent origin. Journal of the British Planetary Society, 43, 203-208.

Oyama, V. I., & Berdahl, B. J. (1977). The Viking gas exchange experiments from Chryse and Utopia surface samples. Journal of Geophysical Research, 82, 4669-4676.

Parker, T. J., Schneeberger, D. M., Pieri, D. C., & Saunders, R. S. (1986). Geomorphic evidence for ancient seas on Mars. In Symposium on Mars: Evolution of Its Climate and Atmosphere (V. Baker et al. eds., pp. 96-97.) LPI Tech. Rpt. 87-0 1. Houston, Texas: Lunar and Planetary Institute.


Pozos, R. R. (1986). The Face on Murs: Evidence.fi,r a lost civilization? Chicago, IL: Chicago Review Press.

Sagan, C., & Fox, P. ( 1975). The canals of Mars: An assessment after Mariner 9. Icarus, 25, 602-6 1 2.

Sagan, C., & Shkolovskii, I. S. ( 1960). Intelligent Ilfe in the universe. San Francisco, CA: Holden- Day.

Sagan, C., Phanouf, J. P., & Ihnaut, M. (1965). Total reflection spectrophotometry and thermo- gravimetric analysis of simulated Martian surface materials. Icarus, 4, 43-6 1.

Sagan, C., & Wallace, D. (197 1). A search for life on Earth at 100 meter resolution. Icarus, 15, 5 15-554.

Soffen, G. (1982). Life on Mars. In The New Solar System, 2nd Edition, (Beatty, J . , O'Leary, B., & Chaikin, H., eds., pp. 93-96.) Cambridge, MA: Cambridge University Press and Sky Publishing Company.

Toulmin, 111, P., Baird, A. K., Clark, B. C., Keil, K., Rose, H. J., Jr., Christian, R. P., Evans, P. H., & Kelliher, W. C. (1977, Sep.). Journal cfGeophysica1 Research, 82, 4625-4634.

Vickery, A. M., & Melosh, H. J . (1987). The large crater origin of SNC meteriorites. Science, 237, 738-743.

Journal ofscient~fic Exploration, Vol. 5, No. I , pp. 27-59, 1991 Pergamon Press plc. Printed in the USA.

0892-3310/91 $3.00+.00 0 1991 Society for Scientific Exploration

New Cases in Burma, Thailand, and Turkey: A Limited Field Study Replication of Some Aspects of Ian Stevenson's Research

JURGEN KEIL

Psychology Department, University of Tasmania, Hobart, Tasmania 7001, Australia

Abstract-For about 25 years, Ian Stevenson has presented detailed accounts of his first hand investigations which suggest a substantial basis for the widespread acceptance of rebirth or reincarnation cases in some communities. This investigation was limited to two main questions: Do some children and their families report that information and other indicators emerge which provide connections to other persons who have previously died? Is it necessary to assume that some of the information and that some of the indicators are due to paranormal processes because they cannot be reasonably accounted for by more orthodox means? The results of this investigation are based on a preliminary evaluation of all 23 cases studied in 1988. The 23 cases can be regarded as relatively unselected samples from three countries, namely Burma, Thailand, and Turkey. The answer to the first question is an unqualified "yes." The answer to the second question is also affirmative but due to the limitations of field study conditions, the affirma- tion can only be justified in somewhat subjective probability terms, nevertheless, in terms comparable to what in quantitative research is called significant. Although the question whether the suggested paranormal results support the reincarnation hypothesis was not pursued, some aspects of this investigation which seem relevant to this question are briefly discussed.

Connections Between Children and Persons Who Had Previously Died

Devoting a major part of his life to this research Ian Stevenson (IS), hasar- argued in numerous publications (1966, 1975, 1977, 1980, 1983, and 1987), that in various communities, some children have memories, behavioral characteristics, as well as physiological peculiarities which seem to uniquely connect them with particular individuals who have previously died. These cases can only be regarded as potentially important with respect to reincarnation or rebirth hypotheses if these connections to previous lives (PLs) include some paranormal aspects. Two publications by IS (1 988, 1990) in the JSE, although primarily concerned with unusual aspects of the cases studied, provide examples of IS'S investigations.

Acknowledgments. This work was partly supported from the Division of Personality Studies, Health Sciences Center, University of Virginia. Thanks are due to the interpreters referred to in this report, to Abdurrahman Karaali, Edip Ediboglu, and Semsettin Duan for their help with contacting cases in Turkey and to Sue Ross for typing the manuscript.

2 8 J. Keil

In 1988, 1 carried out a field investigation which attempted to establish:

1. whether some children and their families report that connections exist which are based on memories and other characteristics of persons who had previously died. These characteristics may include behavior patterns, such as skills and aptitudes, health-related indicators and, other physiological peculiarities such as birthmarks and/or defects and racial characteristics.

2. whether on the basis of a qualitative analysis strong indications emerge which suggest that, in addition to normal aspects of such connections, paranormal ones must have also played a part.

3. whether frameworks and methods-in addition to those already employed by IS-can be introduced which might assist in estimating the probability of (2) above.

Prior to the investigation, it was agreed that if the conclusion was reached that paranormal aspects played a part, it would not be the purpose of this report to pursue the question as to whether such results can be best accommodated within the framework of a rebirth or reincarnation hypothesis. Nevertheless, at the conclusion some aspects of this question will be briefly discussed.

For this investigation, it seems appropriate to define the term connection as an identifiable similarity or equivalence between, on the one hand, actions or mental and/or physiological states and characteristics of a previous personality (PP), i.e., of a person who had previously died and, on the other hand, those of a subject (S), usually a child, who is alive. It is assumed that all these connections contain some normal elements, including coincidental similarities and not necessarily any paranormal aspects at all. Nevertheless, paranormal aspects or processes may manifest themselves in these connections, and may be the most striking feature of scme of them. Rebirth or reincarnation as a possible and perhaps even as the lrnost appropriate interpretation of such paranormal aspects, could not be reasonably maintained as more than a symbolic concept, if it was assumed that paranormal aspects do not play a part at least in some of these connections.

Since the field work was carried out along similar lines previously employed by IS, other terms already introduced by IS (1987, p. 20) will be used without suggesting that any interpretative conclusions which may be implied by these terms should necessarily be adopted. For the summaries and discussions of actual cases, terms like rebirth, reincarnation, and previous life (PL) will frequently be used because these were the terms which were used by the people concerned.

Independence of the Study

This relatively independent field work was encouraged and supported by IS. Some comments, therefore, are appropriate which clearly indicate the

New cases in Burma, Thailand, and Turkey 2 9

limitations of this independence as well as other aspects of my interest in this area which may have had a bearing on the fieldwork itself as well as on this written account of it.

More than 20 years ago, I visited IS for the first time at the University of Virginia. At that stage, I mainly worked in association with the late J. G. Pratt, carrying out experiments quite unrelated to Stevenson's reincarnation studies. During several further visits to the Division of Personality Studies, some of them supported by IS, I have for some time developed a high regard for IS which could best be summarized by such terms as professional appreci- ation and personal friendship. Some readers may question whether this is a good basis for an independent study. However, my high regard for IS leaves me in no doubt that he would welcome any results in his field of interest which are based on sound research, whether they support his point of view or not. Indeed, I am sure that he would regard it as quite improper if I reported any research findings in such a way that they did not fully reflect my own views based on the research which I conducted.

Prior to this field work, I have not been involved in survival, reincarnation, or rebirth research. It is fair to say that my interest in this area has always been rather limited because I regard a belief in the finality of death as reasonable. I regard the normal influence human beings can exercise beyond their lifespan (on account of their work and involvement with others when they are alive) as a sufficiently humbling, as well as an elating experience which to me does not suggest a need for any additional survival possibilities. I also realize that this apparently quite personal position in my belief system is nevertheless influenced by cultural and other factors and that in a different community setting I might have developed quite a different point of view.

It is also for this reason that I preferred not to be involved in the interpretation of any paranormal results. Although I have no difficulty in considering rebirth or reincarnation intellectually, I am not sure to what extent such considerations can be kept free of unconscious distortions when these considerations begin to have a bearing on my personal belief system. Perhaps more importantly, by avoiding any involvement in IS'S interpretation that the rebirth hypothesis provides the best fit for such data, my own investigations remained relatively straightforward and do not require any special expertise with respect to philosophical and religious questions which cannot be avoided when the rebirth or reincarnation hypothesis is included.

Procedure

I investigated all the cases by interviewing the persons concerned with the help of an interpreter. Notes and occasionally tape recordings were made during these interviews.

There are some general differences between the way cases were investigated in this study compared to most cases reported by IS. A large number of IS'S cases were based on an initial investigation and frequently on several

30 J. Keil

follow-up interviews at a later stage after all the details of the initial investigation had been examined. Although such follow-up interviews would obviously be desirable in practically all of my cases, under the time and travel restrictions this was only possible to a limited extent.

Since the second question-whether paranormal processes are involved- very soon became the major consideration in this study, I tried to pursue those aspects which might throw some light on this question and I avoided time consuming, but in the long run, highly desirable tasks such as finding postmortem records, birth certificates, and similar documents. I can, to some extent, justify this choice by pointing out that if there is no disagreement about the existence of a subject and of a PP, official documentation frequently does not have a bearing on the paranormality question. Postmor- tem reports may be most valuable if birthmarks are a major and potentially paranormal feature of a case, but my cases-to the extent that postmortem reports might have been obtained at all-would probably not have been improved (as far as the paranormality question is concerned) on account of such records.

Indigenous Acceptance of Cases

Readers in countries in which reincarnation cases are rarely reported may need to be assured that in other communities such cases are, in fact, reported and often accepted. Based on field observations and qualitative evidence, I would argue that the essential aspects of these cases are truthfully reported from the point of view of those who experience them directly or indirectly. Although deliberate deceptions may occur from time to time, such occurrences are quite rare and do not have any real bearing on this first question. My finding that with rare exceptions those who report cases speak the truth as they see it is also based on the following more specific observations:

1. Practically all interviews conducted in Burmese villages (and to a lesser degree, this was also happening in Thailand and Turkey) took place in public. An individual could not make statements without the rest of the village being aware of it. In villages, particularly in Burma, belief in reincarnation is an important aspect of religious and community life. False claims would generally be regarded as major transgressions of the behavioral codes of these communities. In small villages most of the people seem to be aware of existing cases and intentional misrepresenta- tions made by one individual would be obvious to most of those present, who, during interviews, frequently express approval and agreement with statements which correspond to their memories and experiences. I have no doubt that in most of these small communities disapproval would be clearly voiced if false or misleading statements were made.

2. In Thailand and Turkey, parents of children who appear to have connections to PLs, frequently do not welcome these connections. Particu-


larly, some fathers are most upset if their sons tell them something like: "You are not my father, my father's name is . . ." Some parents in Thailand give special food to such children which is supposed to help them to forget their PLs. Particularly in Turkey, at times, some such children are beaten by their fathers when these children try to talk about their PLs.

3. Many parents-whether they strongly oppose their children's connections to PLs or not-make an effort to investigate their children's claim and visit the villages and houses that these children point out to them. Such visits are noticed by others who are not directly related to the case.

4. A study of the details of the cases investigated suggests agreement with the assumption that most of the people involved provided information in good faith and considerable disagreement with deception and conspir- acy hypotheses.

5. On one occasion, a professional psychologist, Duansmorn Likitsawat, from the hospital in Surat Thani in Thailand participated in the investigation of the Ourarat Srineel case. She stated at the beginning that she did not believe in reincarnation and she had not changed her mind after our investigation. Nevertheless, she came to the conclusion that the people involved spoke the truth as they saw it.

In general, I am confident that any reasonable person-no matter how opposed to reincarnation or to paranormal aspects of such connections- would come to a similar conclusion about most of the cases. Once the conclusion is reached that children and their relatives do indeed report connections between themselves and PPs, the main question that remains is whether these connections can be accounted for by normal means, including coincidences, or whether paranormal ones appear to be involved as well.

Assessment of Paranormality

Analyzing the cases in order to estimate to what extent paranormal aspects or processes are involved probably invokes in most readers some negative reaction to the term paranormal and perhaps in some a complete rejection of paranormal possibilities. Unfortunately, in the popular literature, the term paranormal is frequently distorted and sensationalized to a point which jus- tifies such a complete rejection. For the purpose of this analysis, paranormal simply indicates that some information was not obtained by any normal means that can presently be suggested, such as a child overhearing conversa- tions, etc., and later repeating details while the child is no longer aware of the source. Normal information may be mixed up with paranormal information. Similarly birthmarks in themselves may be due to normal physiological processes, but in some cases the location, shape, appearance, color, and

3 2 J. Keil

Although it is never possible to rule out all hypothetical normal explanations with absolute certainty, the degree to which this is attempted provides some measure of the likelihood that a paranormal process is involved. Case studies are relatively complex summaries of statements and observations relating to even more complex behavioral responses and experiences, personality characteristics, and physiological peculiarities. Consequently, it is not possible to list conditions which help to exclude normal aspects, in some kind of hierarchical order. Nevertheless, the following conditions are likely to strengthen a claim for the presence of paranormal aspects.

1. S and S's family do not know and have no contact with PP's family until S has made definite and potentially verifiable verbal statements.

2. S makes numerous unambiguous statements which are relatively independent of each other about a PP and a PL which can be verified.

3. S provides information about something not known to anybody-except, in the past, to PP-which can be verified; e.g., some item hidden by PP and recovered by S.

4. Statements made by S-preferably before PP's family is involved-are noted by more than one member of S's family and preferably by others (not belonging to the family) as well.

5. Without any opportunity to learn or imitate, S is able to do something which corresponds to some activity PP was able to carry out. For example, S is able to speak a language or dialect which is not used in S's family, and which is not spoken by persons with whom S had contact.

6. Similar to ( 5 ) , but S has some opportunity to learn and imitate. Never- theless, S seems far more proficient than would normally be expected.

7. S has birthmarks and or deformaties which correspond to injuries or other peculiarities of PP. (Several birthmarks which correspond to specific injuries sustained by PP are more impressive than one birthmark which corresponds to a large injury which has no compelling feature in agreement with the location of the birthmark.)

8. Unusual preferences and phobias of S at an early age which do not make sense in terms of S's experiences in his or her life, but which correspond to some important preferences or traumas in PP's life.

9. Although more difficult to evaluate, the intensity and spontaneity with which statements are made and with which emotions are expressed, also have a bearing on the assessment.

This list could be extended. It can also be integrated, to some extent, if it is kept in mind that there are mainly two criteria in operation here on the basis of which paranormality can be assessed. On the one hand, the number and the complexity of relatively independent statements and other characteristics which correspond to verifiable statements, features, and facts associated with PP must be taken into account because a small number of statements based on fantasies could, by chance, be in agreement with some aspects of

New cases in Burma, Thailand, and Turkey 33

PP's life. On the other hand, the question should be raised, to what extent did barriers exist that made it impossible or unlikely that seemingly paranormal connections between PP and S occurred by normal means.

As I indicated earlier in my assessment, deliberate deceptions do not play an important role, but the probability of self-deception errors and distortions must be seriously considered and, as far as possible, eliminated. Witnesses who are not part of the immediate family circle of S may not be completely free from influences which maintain self-deception but the risk of this occurring is reduced. Motivation also has a bearing on this question. Parents who do not welcome reincarnation cases are less likely to interact with S in ways that might lead to self-deception. However, it would be unfair to conclude that parents who are less opposed to reincarnation necessarily encourage such interactions. Children who seem to remember a PL under more favorable circumstances than their present life are more likely to be motivated by wishful thinking but such socio-economic differences are not a sufficient reason to reject such claims. Similarly, memories of a less favorable PL cannot, on socio-economic grounds, be regarded as a sufficient reason to accept such claims.

In order to assess the cases in a way that has some similarity to a quantitative evaluation, I am suggesting the following procedure and criteria. Each case is first assumed to be based on normal processes only and an attempt is made to fit the data into a framework of normal processes. Items which cannot be readily included, and which seem to demand some paranormal aspects can be combined such that each case can be assessed as significantly positive, strongly positive, or moderately positive. A case may also be judged to be neutral or moderately negative, strongly negative, or significantly negative. However, negative cases (in contrast to some quantitative research) cannot be combined with positive ones in order to support (along similar lines as in two-tailed tests) the hypothesis that paranormal processes are involved. In the overall assessment of cases, negative ones necessarily reduce or eliminate the total positive findings. For the purpose of this study, negative cases are defined as those which are regarded as reincarnation cases in the community in which they are reported, but which, after some analysis, appear to be due to normal processes. Most of these negative cases could probably be called self-deception cases. Numbers suggest a degree of accuracy which is not justified for this assessment. Nevertheless, the categories can be visualized by keeping in mind that "neutral" attempts to indicate a .5 and "significantly positive" a greater than .95 probability that paranormality is involved. It is assumed (in a somewhat similar way as in a heads/tails distribution of coins) that seemingly paranormal aspects of cases have a .5 probability given that normal causes and/or chance are responsible.

In order to make the selection processes explicit, all 23 cases pursued in this study were included in a preliminary evaluation. Brief summaries are provided for about half of the cases which are sufficient to illustrate the evaluation. One case, as a kind of anchorage point, is included with full

34 J. Keil

details. Some of the cases (7) were then.excluded from further consideration because, under the existing time limitations, they could not be investigated to a point where a reasonable assessment could be made. This exclusion was carried out predominantly against the hypothesis that paranormality is involved, because such incomplete cases typically look positive in the first place but, without further checks and information, this assessment does not yet seem justified.

In the following list of cases, a call name which refers to a particular feature of each case is added, and these names will be used when cases are compared and discussed. These names are introduced for convenience only, since most of the real names of the subjects concerned are unfamiliar to many readers, and somewhat difficult to remember.

Cases in Burma

8 August 1988: Ma Aye Aye Maw (5)-Shy girl case (S at first would not talk to us). 8 August 1988: Ma Nyo (36)-Sex change, dacoit case (PP died as a dacoit). 8 August 1988: U Tun Kyi (50)-Firewood chestmark case (PP's chest was burned by a piece of firewood). 9 August 1988: Ma Tin Sein (1 7)-Pierced corpse case (PP's body was pierced when it was pushed into a grave). 9 August 1988: Ma Myint Myint (22)-Discarnate terror case (S claims PP died of fright by a discarnate). 10 August 1988: Mg Myint (26)-Old age case (PP died at age 73).

Cases in Thailand

18 August 1988: Ourarat Srineel (8)-Hospital case (S was taken to a hospital when S started to talk about her PL). 4 November 1988: Sripha Ruken (16)-Buffalo case (PP died when looking after a buffalo). 4 November 1988: Scrimon Manokhan (34)-Retribution case (S believes that her foot deformity is due to misconduct in the PL of the PP). 6 November 1988: Tutkhorn Chitpricha (6)-Bangkok, long distance case (S lives in Bangkok, PP lived and died about 500 kilometers away). (Full details are provided). 9 November 1988: Todsac Gowana (1 1)-Great grandmother case (PP is claimed to be the great grandmother of S).

Cases in Turkey

19 November 1988: Amed Biiyukasik (56)-Execution case (PP was executed by mistake). 19 November 1988: Gomol Biiyukasik (10)-Suicide case (S claims that PP committed suicide). 20 November 1988: Mehmet Arikdal(5)-Adana case (S claims to have lived in Adana during his PL).


2 1 November 1988: Demet Kiziltan (8)-Murdered wife case (S claims PP was murdered by PP's husband). 2 1 November 1988: Bedia Gulbahar (24)-First Istanbul case (S claims as PP she lived in Istanbul). 22 November 1988: Izzettin Genc (35)-Sweet shop man case (S has a sweetshop at the highway to Antakya). 23 November 1988: Yildiz Octay (1 9)-Second Istanbul case (S claims as PP she lived in Istanbul). 25 November 1988: Zeynep Sonmez (6)-Doctor's case (investigated with the help of an MD from Antakya). 26 November 1988: Aziz Guzel(30)-Reluctant case (PP's family gave information, but S was unwilling to talk). 26 November 1988: Emire Koc (7)-Turkish shy girl case (S would not talk to us). 29 November 1988: Fatma Altinoz (30)-Pin case (S found pin from PP's life). 30 November 1988: Semsettin Menken (2 1)-Motorbike case (PP died in a motorbike accident).

In the following brief summaries of about half the cases which I studied, only a few aspects will be summarized. All available details of these cases are on file at the Division of Personality Studies, Department of Behavioral Medicine and Psychiatry, School of Medicine, University of Virginia. In addition to the brief summaries of about half the cases, one case which I regard as relatively important from the point of view of paranormality will be presented in more detail. This example may also be of help in clarifying the categorization and overall assessment of all the cases.

Brief Summaries of the Cases Studied

Cases in Burma

IS had studied cases in Burma previously and made arrangements for Daw Hnin Aye (DHA) to prepare preliminary information about new cases in different localities. Visits to Burma in 1988 were restricted to seven days, and several visits (up to five) had been planned. Even in five weeks it would not have been possible to study all the cases DHA had located. In consultation with IS, I selected a number of cases that seemed particularly interesting because of birth marks, sex change, age of S, etc.

As it happened I arrived in Pyawbwe (a town on the railway line between Rangoon and Mandalay, about 1; hours by train from Mandalay) on 8 August 1988. This was the day when major disturbances started in Rangoon, and soon spread throughout the country. On the 8th I visited Pattar village (about 3 hours drive by jeep from Pyawbwe) and investigated the case of Ma Aye Aye Maw (the Burmese shy girl case), the case of Ma Nyo (the dacoit, sex

3 6 J . Keil

change case), and the case of U Tun Kyi (the firewood chestmark case). Because of the limited time during that day only one of the preselected cases, the dacoit sex change case, was investigated.

On account of the disturbances, only three additional cases near Pyawbwe could be pursued. Among these six Burmese cases, only one was preselected. DHA participated as an interpreter for all of them.

The Shy Girl Case. As a three-year-old, the subject (S) said she was the previous personality (PP), who had died at the age of 19. In my presence, S, then 5-years-old, initially would not talk at all. We obtained statements from S's elder sister and S's mother. After much prompting, S later spoke a few sentences. I could not really direct any questions to her. It is difficult to judge how far S was influenced in what she said by those who were present. Ac- cording to DHA, the contents of S's sentences were unrelated to the encouragement which she received. S spoke only briefly and what she said was limited to claiming to be PP and to a short account of how, as a discarnate, she followed a cow into the village. I do not feel confident that S's statements were her own, but it is possible they were. I had no opportunity on that day to talk to members of PP's family who lived in the same village some distance from S. S's mother claimed that PP was not mentioned in the presence of S until after S had started to talk about her PL at the age of three. I am confident that S's mother and other present members of the village believe that S is a reincarnation of PP. Apparently neighbors had heard S talk about her PL.

Evaluation: As the case stands I would regard it as neutral. However, since a good deal of additional information could be obtained in the future, I regard this case as sufficiently incomplete to be excluded from the overall evaluation of this study.

The Dacoit Sex Change Case. PP's family lives in the same village as S. I interviewed S (aged 36 in 1988) and S's father. DHA had previously spoken to S's mother and to Mg Aye Myint from the same village who remembers various details about PP. We had no opportunity to talk to a member of PP's family. PP, as a dacoit, was remembered by various people in the village including by villagers who were present during our interviews with S and S's father.

S at 36 seemed to remember a good deal about her PL which is unusual at that age. Her explanation was that every day her deformed feet reminded her of her PL. S's statements agreed with those made by her parents. S's behavior was clearly moderately masculine. S was noticeably independent, i.e., she did not seem to modify or change her statements in any way in order to accommodate expectations by others. Her birthmarks agree with PP's injuries, but not to a significant extent because PP received multiple injuries at death. S apparently was able to indicate where she had hidden some money in her PL. The money was found, but on the day of my visit, I could not contact anybody who could independently verify these events.


There was a suggestion of which I had not become aware until after we had left the village, namely, that S's deformities were due to leprosy and that S tries to conceal this by referring to the injuries PP suffered at the time of death.

Evaluation: The case is rich in detail and seems to receive support from various sources, and although additional information is needed, it seems likely that such information will strengthen rather than weaken the case. Although leprosy was mentioned as a possible motive for deception, it seems unlikely that leprosy was actually involved. A dacoit in general, and this PP in particular, on account of his misdeeds, cannot be regarded as a suitable basis for deception which S's parents might have started. Since S started to talk about her PL at the age of three, it is unlikely that S used normal information about PP which may have been available in the village at that time. In summary, I regard this case as moderately to strongly positive.

The Firewood Chestmark Case. The subject (aged 50 in 1988) can still remember, but added that members of his family reminded him of PP and that this is the reason. As PP, he fell into a lake, and in this life he fell into a river. S is afraid of water. S has a birthmark which is still visible on his chest which is supposed to be due to a piece of firewood which burnt PP on the same spot when PP was four-years-old. The firewood event was unrelated to PP's death. PP had fished without permission, was beaten by monks, and died three days later.

We had no opportunity to talk to other members of S's or PP's families. None of the statements could be verified. S is a village elder and there is no good reason why he should wish to deceive us or others. The birthmark on his chest, although visible, is not disfiguring and does not seem to provide a good reason for deception which might have been initiated by his parents. Nevertheles~, the amount of information available is very limited at present.

Evaluation: On the basis of the available information at present I would rate this case as neutral to moderately positive. The case is sufficiently incomplete to be excluded from the overall evaluation of this study.

The Pierced Corpse Case (Ma Tin Sein). We interviewed S, (aged 17 in 1988), her father, PP's sister, a witness (who had heard S talk about a PL), as well as a witness who was present during the preparation of PP's funeral.

PP died when a lorry overturned outside his home village. According to local customs, the corpse must be kept outside the village, and the corpse had started to bloat before the funeral could be arranged. PP was buried in a tomb but the corpse could not be easily fitted into the tomb. Sticks were used to move the corpse and one thigh was pierced. This account of PP's death was given by PP's sister and later confirmed by a witness who was present during the funeral preparations. S has a birthmark on one thigh which is still clearly visible. S's parents did not know PP who lived in a different village about 35 kilometers away. S's parents had no direct contact with PP's village, but a sister of S's father lives in PP's village. S had started to talk about a PL

38 J. Keil

when she was two- or three-years-old. S spoke spontaneously about a PL until she was 5. This was confirmed by a neighbor who had heard S at that time. S gave PP's name and details about PP's accident. When PP's sister became aware that S made these statements, S was taken to PP's village. S recognized PP's clothes. S also inquired about a tree that PP had planted but which was no longer there.

S is a girl, PP was a man. S, as a child, had a preference for boys clothes and participated in some games which are usually only played by boys. However, S cried easily.

Evaluation: S said that she could still remember some aspects of her PL, but since the statements which she had made as a child were known by various people, and since contact had been established with PP's relatives for some time, it is difficult to judge to what extent S's statements can be regarded as independent memories. Nevertheless, it seems clear that S's parents did not know the PP, and that S gave information about a PP without any apparent contact with persons who could have known PP. The witness (not related to S) who had heard S's statements about a PL some years ago, was intelligent and not inclined to accept reincarnation stories easily. How- ever, she accepted S as such a case. Factual details provided by S agreed with statements made by PP's sister and the person who participated in the preparation of PP's funeral. If it had been possible to investigate the case when S was still younger, the case might have been judged to be significant in its own right. As it is, I rate it as moderately to strongly positive.

The Discarnate Terror Case. The subject (aged 22 in 1988) can still remember her PL, but she could not give us many details. Apparently, as a child, she recognized and pointed out objects that belonged to PP. During my visit, we were only able to talk to S and her grandmother-who could give us some information about PP. Other members of S's and PP's families could be contacted but were not available during my visit.

Evaluation: On the basis of the information available at present, I would rate this case as neutral to moderately positive. The case is sufficiently incomplete to be excluded from the overall evaluation of this study.

The Old Age Case. The subject (aged 26 in 1988) can still remember his PL. He gave a fair number of details. Apparently, he spontaneously recognized some items that belonged to PP.

S also remembered that, as PP, he gave some scriptures to his second son (in his PL) who still has these scriptures. Apart from S, we were only able to talk to S's mother, who confirmed his statements.

Evaluation: Although the information available is limited and based on only two persons, some of the items and events appear to be substantial. Further inquiries should be pursued. Nevertheless, on the basis of the information available to date, I feel sufficiently confident to include this case in


Cases in Thailand

As indicated in the introduction to the Burmese cases, the problems which occurred in August 1988, prevented the investigation of most of the cases which I had planned to pursue there. On short notice, I had to turn to Thailand for the continuation of my research. IS had studied cases in Thai- land some years ago but, unlike in Burma, there was only one case to which I could turn immediately. Although I had most useful assistance from a number of Thais who had been associated with IS'S research before, generally it was not possible for any of them to participate in the actual investigations during my visit. It took more time to find and pursue cases in Thailand. On ths other hand, it was possible to follow various leads and to travel freely throughout Thailand for this purpose.

Although Thailand is predominantly a Buddhist country and most people accept reincarnation in an abstract kind of way, few Thais expect to experience these strange connections which some children claim to have. Most Thais are decidedly uneasy when they are personally confronted with such events in their own families. Many believe that these memories will create problems for their sons and daughters, and usually children are discouraged from expressing their alleged memories. In some regions it is even customary to prepare special food which is supposed to help these children to forget their PLs.

This relatively negative attitude toward cases probably developed fairly recently, and may not have been equally predominant 20, or even 10, years ago. Relatives of the great grandmother case told me that the custom to mark dead bodies in order to recognize corresponding birthmarks on reincarnation cases was abandoned about 30 years ago. A number of senior monks could not remember having heard of any cases in their districts during recent years. Some had no knowledge of any cases at all. Although this indifferent attitude toward cases appears to be fairly widespread, there are regional differences. It must also be mentioned that professional and academic groups expressed a good deal of positive interest in these case studies. I would not be surprised, though, if most of the people in these groups would be equally uneasy if it concerned their own families.

For the investigation of the retribution case and the buffalo case, Dr. Boon Nilakesh from Chiang Mai University participated as interpreter. Dr. Boon also translated when we contacted PP's father, brother, and former wife in connection with the Bangkok case. PP's sister speaks English and except for a few words which Dr. Boon translated, no interpretation was required. Both Dr. Chieng Siriyananda from the Siri clinic in Bangkok and his son, Mr. Chutinun, translated for me when we interviewed the parents of the Bang- kok case (i.e., s's parents).

Nathapong Muadtip, from the Nakorn Sri Thammarat teacher's college, participated as an interpreter during all the interviews associated with the

40 J. Keil

Ourarat case. The great grandmother sex change case I initially contacted with the assistance of a monk, Mr. Putsupat Karahong from Wat Tha ton, Tumbon Tha Ton, Amphur Mae-Ai, who speaks a limited amount of En- glish. Mr. Jirapon Kurenak, an architect from Bangkok who was temporarily in Tha Ton, kindly volunteered to translate during a second visit during which we obtained the information for this case.

The Ourarat Srineel Case (Hospital Case). We met S, S's mother, and PP's father. S was seven-years-old in 1988. PP had died suddenly when she was eight. S's parents have a shop about one kilometer from PP's house. Until S was two-years-old, S's parents lived somewhere else, and there was no contact between s's parents and PP's relatives. There was little contact later, but PP's father occasionally visited S's parents' shop.

S started to talk about a PL when she was two. S cried a lot, and particularly wanted to be with "her father" but not with her real father, i.e., S's father. One day, when S saw PP's father in her parents shop, she called him father and was happy to be with him. S's parents were concerned about these events and took S to a hospital. Nevertheless, S's parents also noticed that S was much more content and did not cry after she had met PP's father. S eagerly waited for further visits by PP's father and kept food items for him.

When S visited PP's house for the first time, she pointed out where PP had slept even though this house was no longer used. In the meantime, PP's parents had built a new house. PP recognized photos and other items.

Information was mainly provided by S's mother, who is a thoughtful and reasonable woman. S's parents are in a better socio-economic position than PP's parents. A local psychologist, from the hospital that S's parents had consulted, accompanied us during these interviews. She said, she did not believe in reincarnation but also stated that she believed that the people spoke the truth as they saw it. When we met S she could not really remember details from a PL any longer. S has become involved in religious matters and associated religious beliefs with her views of her PL which seem to be based on what her parents had told her about her earlier behavior. Although we had no opportunity to interview other persons not related to these families, apparently a number of persons had heard S talk about a PL because many people visit the shop.

Evaluation: It is possible that S may have overheard some discussion when she was two, and that she incorporated some information into what seemed to be memories of a PL. Nevertheless, I was positively impressed with this case because S's parents had not welcomed S's behavior, and had tried to modify it by taking her to a hospital. I was unable to obtain records or direct accounts about these consultations from hospital staff, but there is no doubt that they took place. Gradually S's parents accepted S's claims. There was no possible contact between S and PP's relatives until S was two, and it is unlikely that S could have obtained any details when S's parents moved to the shop. It was shortly afterward that S started to talk about a PL. S expressed an


exceptionally strong emotional need for "her father," which seemed to have been satisfied when she met PP's father. The socio-economic position of PP's parents is clearly less favorable, and S's behavior cannot be regarded as wishful thinking. The information provided by S's mother and PP's father are in good agreement, and the events took place fairly recently. S's mother was reluctant to consider reincarnation as a possibility but gradually accepted this interpretation. It would be desirable to contact other people who wit- nessed S's accounts of a PL when S was younger; as it stands, I rate the case a strongly positive, but not as significant in its own right.

The Sripha Ruken Case (The Buflalo Case). The subject (aged 16 in 1988) does not now remember a PL. We talked to S's sister and S's parents. Ac- cording to S's mother, S may have started to talk about a PL before he was 5. When he was 5, S indicated to her that in his PL he looked after a buffalo. S took his mother to PP's place. He told her the name of PP's father. In PP's house, S identified some photos. PP's parents confirmed some statements that apparently were made by S but there were also some discrepancies. The discrepancies do not suggest deliberate misrepresentation.

Evaluation: Although a number of details are available which are in agreement when statements made by s's relatives and PP's parents are compared, most of these details could be explained by normal means, particularly if S did not start to talk about a PL until he was 5, or close to 5. It is difficult though to assume that all the details are due to normal information transfer. Until additional interviews can be conducted I rate the case as neutral.

The Srimon Manokhan Case (The Retribution Case). The subject (aged 34 in 1988) apparently still remembered her PL. S has a clearly deformed foot in agreement with (according to S) more severe but otherwise similar deformities of PP. S believes, that as PP, she could also remember a PL during which, as PP2, she had kicked her father. When PP2 was reborn as PP, the deformity occurred as a retribution for PP2's behavior toward her father. S can remember a number of details related to her PL. Apparently, S recognized people and objects when she went to PP's village for the first time. We met S's parents (S's mother was 82 in 1988) who were still in good health, and who confirmed what S had said. They also added some details. S's mother had known PP and confirmed PP's foot deformity. A sister of S's mother also confirmed S's statements. Several additional relatives who were not available during my visit, could provide information which has a bearing on this case.

Evaluation: Additional interviews are desirable. As it is, a sufficient amount of detail seems available to include this case in the overall evaluation, and I rate it as moderately positive. In this moderately positive assessment, I have not considered the retribution claim, i.e., that PP also remembered a PL.

The Todsac Gowana Case (The Great Grandmother, Sex Change Case).

42 J. Keil

English accompanied me) and a second meeting with a volunteer interpreter, I had no opportunity to talk to S (aged 1 1 in 1988) who was at school on both occasions. S's mother (S's father was at work on both occasions) and S's grandmother were the main informants. here was agreement between the informants, and a number of details emerged. S referred to and recognized personal items of PP. He started to talk about his PL when he was two-years- old. S used phrases characteristic of PP. S's relatives did not encourage S to talk about his PL, but they all accept that he is a rebirth case of PP.

Evaluation: Although PP is S's great grandmother, and information about PP could have been conveyed to S inadvertently, this does not readily account for such details as the names of villages where PP had lived, and which apparently were not mentioned during the time when S might have unconsciously assimilated such information. S started to talk about his PL at 2. S's parents did not encourage S to talk about his PL. On the other hand, S's mother who had a close relationship with her grandmother, i.e., with PP, might have been motivated through dreams to see S as a rebirth of PP. Interviewing S and S's father would be desirable, but this is not likely to add much because S apparently no longer remembers his PL. No other relatives or other persons are available who could provide relevant information. Suffi- cient details seem available to include this case in the overall evaluation, and on the basis of the information received to date, I rate the case as moderately positive.

The Tutkhorn Chitpricha (Bangkok-Long Distance Case). (Full details are presented for this case.) This is the only case for which the families involved requested anonymity. All the original details are on record. Critics may view requests for anonymity with suspicion. All the indications which I have suggest that this request, although perhaps unnecessary, is a reasonable one on account of the socio-economic positions of the families. A reincarnation case would probably be regarded as a sign of irrational views by some for- eigners with whom the families have contact, and even some of their Thai associates may have negative reactions. There were certainly no indications of any kind which might suggest that the people concerned were trying to mislead the investigators.

Knowledge about this case is almost entirely restricted to the circles of the two families. I became aware of it through a Thai academic who knew of my interest in such cases. Without being aware of any details, he had heard something about S from PP's sister. Through him I was introduced to PP's sister, Nit, who speaks English. All other interviews were carried out with the help of Dr. Boon Nilakesh (BN), Dr. Chien, and Mr. Chutinun Siriyananda. Although we met S, a boy who was 6 in 1988, on three occasions, we agreed not to interview him because his parents were concerned that reviving memories of PP might also bring back fears and nightmares which S apparently experienced when he spoke spontaneously about PP. For these reasons S's


parents had encouraged S to forget about PP and apparently S had not referred to PP during the last 2; years. S's parents believe that S no longer remembers PP. We interviewed PP's father and PP's brother before we had an opportunity to speak to S's parents. We had no opportunity to return to PP's relatives in order to check out some of the statements made by S's parents.

S's father Tosarot Chitpricha is a professional man in a government position. S's mother Chongdee is a high school teacher in a distinguished school. S's mother has a good knowledge of English but she preferred to communicate with the help of an interpreter. S, Tutkhorn Chitpricha, is their only son. PP, Dang Chitpricha, was Tosarot's brother. Until Tosarot went to Univer- sity and later married, he grew up as one of PP's siblings in a town (A) approximately 500 kilometers from Bangkok. PP's father was, and is, an important person in A. He has one of the major business enterprises in that district. Nit now has lived for sometime in another town (B) several hundred kilometers from A. She has followed a professional career and has advanced qualifications. PP's brother Korn who was present when we interviewed PP's father, works in his father's business.

In November 1988, PP's father (of Chinese descent) was 66. PP's mother is of Thai-Chinese origin. Their first child, Nit, was then about 4 1. Their second child, Korn, participated in our interview. Their fourth child, Dang, was PP who died in 198 1 at the age of 26. He would have been 33 in 1988. Approximately one year after PP, Tosarot (S's father) was born. Tosarot has moved to Bangkok, got married, and Tutkhorn (S), who was 6 in 1988, is their only child.

The following information about PP was supplied by PP's father and PP's brother, Korn, in a joint interview, and by Nit in a separate interview. In the joint interview, PP's father was mainly speaking; BN translated. The local abbot was also present. He had kindly agreed to accompany us because, without him, PP's father may not have agreed to talk to us.

According to PP's father, PP, as a child, was self-interested and selfish. He liked meat and rice the way it is prepared locally. As a child he was afraid of spirits. He was also afraid of death particularly when he was not well. PP completed class 10, aged about 18. At first, he worked with his father. He had a good relationship with most other people, especially with ladies. His relationship with ladies was a little bit too good. He especially liked cars and ladies. He was a man of high tastes. He liked music as well. All kinds of music. He had a good gift to play almost any kind of instrument. He was very good at entertaining others. He would stop his business and entertain first. He was very good with anything connected with entertainment. He was regarded as a leader among the young with respect to anything connected with entertainment. When there was a festival in the district, he would be entertaining young people. He was not hot-tempered but good-humored.

44 J. Keil

friends with almost everyone. Before marriage, a young person should become a monk first. He did not. Maybe he wanted to become a monk later but he liked to be married first. He had a good family life. After marriage he worked for his own family. With his business he was successful. He supplied equipment for building constructions to another province.

PP died when a large 20-wheel truck hit the pickup truck in which PP was traveling. The accident occurred 97 kilometers from PP's and PP's father's house. The street and district in which the accident occurred was given by PP's brother, Korn. PP's father agreed. The accident occurred in a curve. The pickup truck was on the correct side of the road. The oncoming truck was too far in the middle of the road with no room to avoid it.

PP's father gave us the name of the surviving driver. We were unable to locate him during our visit. PP's father said he cannot reject that S is PP, but he cannot really accept it either. Korn agreed with this. He also agreed with the statements his father had made and could not think of anything to add. (There was no opportunity to meet PP's mother, but this might be of interest in the future.)

We met Nit in a town several hundred kilometers from A. Nit speaks English well and BN did not need to translate. He occasionally helped with a particular word. She was clear but not overconfident. According to Nit, PP and Tosarot (S's father) were very close. The age difference is only about one year. As children, they slept together in the same room and prayed together. On the other hand, Nit is about eight years older than PP, and theirs was not such a close relationship. According to Nit, PP was married, had one boy, and his wife was pregnant when he died. PP was happy with his family and children. PP's sister gave further details about PP's life and death which agreed with the statements made by PP's brother and father. BN was present during this meeting.

We then visited S's parents. Dr. Chien and Mr. Chutinon Siriyananda translated. Both parents are intelligent, clear, reluctant about what happened, but also amused by it, particularly now after the events have stopped. They both talked together, there was no disagreement between them. When things happened, S's father was more concerned than S's mother. The following statements were made by both.

Before S's mother was pregnant, she dreamt a bad dream about PP's home town. There was a funeral ceremony. She asked who had died and then woke up. S's mother told her husband not to go to PP's home district. The next morning, the accident with PP happened. S's father was supposed to have been there. Perhaps the dream saved him, and perhaps PP died instead of S's father. They also heard of a dream that Korn (PP's brother) had. Korn dreamt about the accident: PP came and told him that this is not the time to die, that he did not want to die, rather to go back, but his body was injured. In the dream, Korn told PP to meet Tosarot and ask for his advice. Then PP left and went to the temple. Korn woke up. When S's mother was about three


months pregnant, she dreamt that PP wanted to live with them. She was frightened because she knew that PP had died. Nearer the time when she expected the baby, she dreamt that PP came to her with a smiling face. Also PP came with luggage to stay with her. She was frightened and dreamt this "luggage dream" more than once. She may have agreed, in a dream, that PP could stay with them. She was also frightened of these dreams.

There were no complications during birth or pregnancy. S has a birth mark (dark pigmentation about three centimeters in length

and 1i centimeters wide) on his lower left arm. There is no clear relationship to PP's injuries, which were mainly chest injuries, although arm injuries may have been present as well.

S started to talk when he was about nine-months-old. When he was 18 months-old, he started to talk about a PL. He said he was afraid to die. His parents told him that it is difficult to die at a young age. S spoke about an accident. S said there would be an accident and that he would die. S said that it is a road accident. S had these memories and talked about it about once a week until he was about 27-months-old. (S's parents looked at family photos and checked the age in relation to other events depicted in the photos.)

When S was about 2, he went to PP's place for the first time in a car with his aunts. S's parents were not in the same car. When S came to the point where PP had died, he said to stop the car because an accident has happened there. The driver of the car in which S and his aunts traveled knew where the accident had happened, and confirmed that S was correct. S's aunts apparently did not know the locality but S's parents were not quite certain about this. S's parents are not sure now who the aunts were who traveled with S. It would also be difficult to pick out the driver now. There are quite a number of drivers who worked for PP or PP's father's business.

In Bangkok, PP had used S's father's address for an account he had there. When a letter came addressed to PP, and the postman asked the maid whether someone by the name of PP is living here, the maid said no. S heard this and claimed the letter, saying it was his letter. S was then about 21- months-old.

About five months before PP died, he had bought a new earth-moving machine which was probably an excavator. PP liked this machine very much. When S went to PP's home for the first time, he asked for it. S asked his mother every morning to bring him to this machine. He would sit on it. He stayed for 4 or 5 days during this first visit. Every day, he wanted to sit on this machine, every morning and every evening. When S was 2f years-old, he went to PP's home a second time. S was no longer interested in the machine. They stayed three to four days during the second visit. Nothing related to the PP happened during this second visit.

Two former drivers for PP came to visit S's parents in Bangkok. S's parents knew their nicknames, but not their real names. S called them by their nicknames and addressed them as inferiors. This is done through a particular

46 J . Keil

language form. It would have been the way PP had spoken to the drivers. S also postured himself as a kind of boss putting his hands behind his back. (S's father demonstrated). According to S's parents this would have been a typical posture of PP in that sort of situation. At other times, other drivers who had not previously worked for PP, came for a visit, but S did not recognize or react to them. During this period ( I f to 2:)' S could understand a different dialect, quite different from the way his parents speak, and the way Thai is spoken in Bangkok. This dialect from PP's home district is difficult to understand. S also spoke this dialect but not many words. S does not use these words any longer. When S met the two drivers who worked for PP, he used the eastern dialect. He could also understand the drivers and was able to reply but his vocabulary was limited.

When S spoke of his PL-although quite young then-he spoke like an adult. He usually talked about it before going to sleep. His voice was sad and he looked sad almost to the point of crying but not actually crying. Later, he stopped talking about it. S's family now lives in a new house in Bangkok. S has not talked about a PL in this new home. When S spoke of his PL he expressed fear of death. S sometimes had nightmares which S's parents believe were associated with S's memories of a PL. S did not call his grandpar- ents parents. PP liked sticky rice that is popular in the east. S still likes it, but S's mother does not encourage it. This rice is quite differently prepared and can be clearly distinguished from the more common rice dishes. S's parents did not like S to talk about a PL. Particularly, S's father did not like it. PP's father met S five years ago. S did not say anything to them but immediately came to his own car (i.e., PP's car) and tried to drive it himself. There were other cars around. This happened when S's mother walked with S. As soon as he saw "his" car, he immediately went for it. S did not like to go back to Bangkok. This was a problem for PP's father. When PP was alive, he and his father were not always on the best of terms. PP's father felt that his relationship with S was similar.

Evaluation and Discussion: Points which may be raised against the assumption that paranormal processes are involved:

1. S's father and PP are closely related, and were close friends for many years.

2. S's mother and paternal uncle had an announcing dream, and may have unintentionally influenced S to identify with PP.

3. All information revealed by S was known by S's parents, and might have been unintentionally conveyed to S.

4. Similarly behavioral characteristics of PP which S displayed could have been unintentionally conveyed to S by S's parents.

5. Food preferences by S in agreement with PP could be regarded as coincidences.

New cases in B~rma, Thailand, and Turkey

The involvement of paranormal processes is supported because:

1. S made specific and correct statements about PP's death. 2. S apparently recognized the locality (a section of road) where PP died. 3. S changed his behavior when he spoke about his PL and, in specific

ways, when he spoke to two visiting drivers who had worked for PP. 4. S spoke a dialect which was not used by S's parents when he addressed

the two drivers. This dialect is not spoken in the environment in which S lived.

5. S has food preferences (sticky rice) in agreement with PP. 6. S's parents did not favor the idea that S is a rebirth case and tried to

discourage S from talking about a PL.

It seems unlikely that paranormal processes can be completely rejected. S's recognition of the accident locality 90 kilometers from PP's home is difficult to explain by normal means. S spoke about a PL at an early age. It is unlikely that he would have absorbed enough information from PP's parents-assuming they provided information unintentionally-to speak coherently about PP's death. Also, S's mood changed when he talked about his PL. Although S might have noticed some behavior patterns, it is unlikely that this explains, in a satisfactory way, S's behavior when he met the two drivers who had worked for PP. S differentiated between these drivers and others who visited Bangkok. S's reaction to the delivery of a letter addressed to PP seems more comprehensive than could be assumed on the basis of information which S's parents may have unintentionally conveyed. On the basis of the data available to date, I would rate this case as moderately to strongly positive, but not as significant in its own right.

Cases in Turkey

My investigation of cases in Turkey was planned within a short time span of a few weeks, when it became clear that I could not return to Burma during 1988. IS had carried out research in Turkey approximately ten years ago (Stevenson, 1980) and had obtained the cooperation from Dr. Can Polat from Istanbul. In recent years, Dr. Polat has also completed investigations of some cases on his own. Dr. Polat was unable to disengage himself from other commitments during the time of my visit to Turkey, and he kindly arranged for Murat Senova to participate as an interpreter (TurkishIEnglish) during the major part of my stay. Murat Senova had not been previously involved in reincarnation research. During a period of 2; weeks, all interviews were conducted in the Hatay region of Turkey. All the cases are situated in communities in which Arabic is predominantly spoken as the first language. The people are known as Alevis. They follow a particular version of Islam that, among other variations (from the Sunni interpretation of Islam), also accepts

4 8 J. Keil

reincarnation. Alevis are a minority in Turkey but probably not in the Hatay region where some villages are entirely populated by Alevis. Rebirth cases are not as frequently reported as in Burma, but are more common than in Thailand. Alevis believe that if a PP has died in an accident, prematurely or through illness, the reborn child is more likely to remember the PP's life. Sex change cases had not been reported by IS before, and did not come to our notice during 1988. Most adult Alevis speak Turkish fluently, but during interviews with some elderly people and preschool age children, we had to rely on relatives to translate from Arabic to Turkish before Murat could continue. With one exception, all first interviews were conducted with Murat assisting as an interpreter. One case, the doctor's case, was investigated with the help of Dr. Abdurrahman Karaali from Antakya. After Murat returned to Istanbul because of professional commitments, I conducted some follow-up interviews with the help of Naim Bilgin and Cemil Karaca- when I had time for a few more days in Turkey. Both have a limited knowledge of German.

Analyzing the Turkish cases in a similar way, four were excluded because of insufficient details. Among the remaining eight cases, I evaluated one as moderately negative to neutral, two as neutral to moderately positive, one as moderately positive, three as moderately to strongly positive, and one as strongly positive.

As can be seen from the Results section, my evaluations of Turkish cases show a similar distribution over the various categories compared to the cases in Thailand and Burma. In order to avoid excessive length and repetition, brief summaries of Turkish cases are omitted except for the following case which I rated as moderately negative to neutral.

The Mehmet Arikdal (Adana) Case. This case was studied in a preliminary way by Dr. Can Polat from Istanbul who rated it positively with respect to paranormal aspects. Dr. Polat also obtained information about PP's home in Adana. I had no opportunity to investigate PP's background. My evaluation is based on two interviews with two different interpreters. In 1988, S was five-years-old. S had started to talk about a PL when he was two, or younger. He gave the names of PP's parents and details about PP's home in Adana. Some of these details agree with information Dr. Polat was able to obtain. Various details were difficult to verify. When S first started to speak, he spoke Turkish in spite of the fact that his siblings started to speak Arabic first, which is typical for this region of Turkey. S's grandfather whom we met, and who lives with S's family, can only speak Arabic. When S started to speak about his PL he spoke in a deliberate way, which differed from his normal way of speaking.

Most of the information was provided by S's father. S and S's mother confirmed S's father's accounts. When S spoke, I had a strong impression that the information which he provided was meant to be in agreement with


his father's views and that, to some extent, S's accounts of his PL may have been modified or even induced by his father. I do not believe that S was deliberately misleading us, but it seemed that information provided by S was strongly influenced by his father, and I cannot accept S's accounts as relatively independent statements. Similarly, S's mother who only provided minor details, was strongly influenced by her husband. Although PP's background is viewed as superior to their own by s's parents it is difficult to see what S's father is trying to achieve if he was deliberately misleading us. S's father's behavior has a kind of theatrical quality, which may be entirely due to completely innocent personality characteristics. I have no direct evidence that S's father was trying to deceive us, but because of some suggestion of this I have presented this case with pseudonyms.

S may well have started to provide paranormal information about a PP at an early age, but for some reason, at some stage, S's father apparently became very interested and now appears to control the case. It is difficult to see how S's father could have started to deliberately induce memories of a PL in his son but he may have misinterpreted and modified them at a later stage. Perhaps, this does not justify a negative evaluation but I am sufficiently uncertain to opt for a somewhat negative category. The two interpreters who assisted us came to a similar conclusion without being particularly adamant about it.

S does not have much contact with children from other families, and it was not possible to obtain confirmation from persons outside S's family of the way S provided information at an earlier age.

Evaluation: Although the case has several features which could lead to a strong positive evaluation, such as detailed information by S, a substantial separation between S's and PP's families and apparently no contact between the families to date, for the reasons outlined above I rate the case as neutral to moderately negative.

Results

Twenty-three cases were investigated in 1988. Seven were excluded because of insufficient details. None of these seven cases appeared to be negative. The remaining 16 cases were evaluated as indicated below:

Burma Thailand Turkey

Moderately negative to neutral 1 Neutral 1

Neutral to moderately positive 2 Moderately positive 2 2 1

Moderately to strongly positive 1 1 3 Strongly positive 1 1

50 J. Keil

Although there are, in my evaluation, no particular cases which I rated as significantly positive in their own right, the distribution of positive cases strongly suggests that the total sample contains paranormal aspects which, in a comparable quantitative analysis, would be regarded as significant.

Discussion

If, on the basis of an experimental study, statistically significant results are reported which are unexpected and unusual, the question arises whether the apparently significant results are really due to chance. Perhaps, many similar experiments which did not yield significant results were never reported.

Particularly, if there is evidence that some unreported experiments did not produce significant results, the reported findings may be dismissed completely. Such dismissals are justified if it can be shown that the results from all the experiments carried out (including the "significant" one) follow a distribution which is in reasonable agreement with what is expected by chance.

It is important to realize, though, that the distribution of all the results could also turn out to be even more significant (than the single significant experiment reported in the first place), if a substantial proportion of the insignificant experiments show a common trend. Indeed, a number of separate experiments without a single significant result may reach a high significance if analyzed together. It is certainly possible to think of experimental conditions (e.g., involving small numbers of particular subjects who cannot repeatedly participate in the experimentation) which do not produce statistically significant results and which, in spite of apparent trends in the results, cannot be improved to a point where statistical significance can be demonstrated under the existing limitations. Nevertheless, if similar trends occurred in a number of comparable, separate investigations, the combined results may well be highly significant in statistical terms.

Unfortunately, when the significance of unusual results is questioned, it is practically impossible to obtain the necessary information about all relevant experiments which may have been carried out elsewhere. At any rate, evidence that another comparable experiment produced statistically insignificant results needs to be scrutinized in order to see whether, on account of these insignificant results, the overall findings are strengthened or weakened. Without such scrutiny it is certainly not justified to dismiss the first set of results.

In the Results section, I have summarized my evaluation of the paranormality question by suggesting that the overall result is equivalent to what, in a statistical evaluation of quantitative experimental data, would be called significant. In a way, this conclusion is not very different from what IS has stated in qualitative terms on a number of occasions. Nevertheless, my at-

I


to clarify one aspect which previously has, perhaps, not been specifically discussed; and that is: how far the selection and subsequent reporting of particular cases may have been misleading.

IS has never claimed that any one case can be regarded as paranormal with absolute certainty, and I would agree with IS's assessment. It could then be argued that, if even the best cases have weaknesses which cannot be completely eliminated, and if these cases had been selected from a very large number of otherwise extremely weak or negative ones, then the apparent significantly positive cases can be readily explained as chance coincidences which must be expected when all unreported cases are taken into account. Indeed, IS's careful selection of the best cases and his rejection of doubtful cases could have contributed to the emergence of a false significance. By considering all the cases which I have studied in 1988, and from the brief summaries of half of them, and from the detailed evaluation of one of them, it should be clear though, that what I call "significant indications of paranormality," cannot be easily rejected by claiming that the apparently paranormal features are just unrecognized coincidences.

The cases which I studied strongly suggest that with most subjects who are seriously regarded as rebirth or reincarnation cases in the various communities, some paranormal elements are associated, and that if the best cases are left in a larger sample rather than evaluated separately, the overall significance is increased rather than diminished.

The Question of Bias

In the fieldwork carried out in Burma, DHA participated as an interpreter. She had previously located the cases which I investigated there. Although only briefly, DHA had also previously worked with IS. It seems unlikely that DHA's previous work with IS could have introduced any bias in my findings. The question of how far an interpreter who had previously worked for IS might consciously or unconsciously distort information in order to present a particular point of view is only relevant to the cases in Burma and, perhaps, to a very limited extent, to two cases in Thailand. For all the other cases, the interpreters involved had no direct or indirect contact with IS. Since the cases investigated with the help of the latter do not differ from the former I cannot see any justification to reject any of them on account of the interpreters. I am not suggesting here that original statements are always translated without distortions, but it seems clear that the essential aspects of the cases were not substantially modified through the interpretation process. I had more direct evidence of this when on two occasions I had the opportunity to compare translations between two interpreters.

I have excluded three cases which came to my attention from consideration because the reincarnation or rebirth claims are based entirely on meditation experiences with no opportunity for investigations. Some Thais are

52 J. Keil

satisfied with findings entirely based on meditation, particularly if made by a revered monk or abbot.

Although all the cases which I investigated in 1988 and for which I could obtain a reasonable amount of information, are included in the evaluation (and brief summaries are provided for about half the cases), the question must also be considered whether these cases can be regarded as relatively representative samples of the various regions or whether they had already been preselected from larger pools. Prior to my visit to Burma, DHA had collected preliminary information about more than 30 cases. Initially, I had intended to select from this group cases which appeared to be particularly interesting, not necessarily because they looked potentially strong as far as the question of paranormality is concerned but in order to include, for instance, sex change and birthmark cases. From my introduction to the Bur- mese cases, it will be apparent that, due to external events over which I had no control, my investigations in that country were limited to only six cases. Among these six cases, only one happened to be among those that I had previously preselected.

Because of the paucity of cases in Thailand and the relatively short time available for preparations in Turkey, in both of these countries, I studied all the cases that we could find during the time at my disposal. Therefore, it seems justified to present the cases included in this evaluation as unselected and relatively representative samples of the countries and regions that I visited.

It is now also sufficiently clear that the positive results obtained by IS cannot be readily dismissed by suggesting that IS may have intentionally or unintentionally misrepresented some aspects of his research. Apart from my investigation, Mills has recently carried out an independent study which agrees with IS findings (Mills, 1989).

The Question of Intentional or Unintentional Deception

Rejecting selection (and any peculiarities in IS'S research) as a possible explanation for paranormality still leaves the question open as to how far deliberate fraud or unconscious or unintentional distortions and errors in the reports of our informants may have been responsible for the apparently overall significant indication of paranormality. Although deliberate fraud may occur from time to time, such occurrences are rare, and I have earlier pointed out why I reject fraud as a relevant explanation when a group of cases is considered. In relative terms, the weakest section in the structure on which the suggested paranormality is based, consists of possible unintentional and unconscious distortions and errors introduced by the informants. Such distortions and errors may add up to provide examples of apparently consistent and meaningful cases. Particularly, if it is kept in mind that young children may be fairly easily influenced by suggestions and behavior patterns


of which those in contact with these children are not aware, the risk of unintentional leakage and subsequent absorption of information which is later interpreted as paranormal, cannot be ruled out with absolute certainty.

Lack of contact between the S's family and the PP's family, as well as other factors discussed earlier, help to reduce this problem but cannot entirely eliminate it. Indeed, it is in this area where I can appreciate that critics may be less impressed with those aspects of the cases which I rated positively, i.e., in favor of paranormality. It is also in this area where I would later like to suggest some new approaches.

IS has, for some time, paid particular attention to birthmarks and birth defects in children which seem to be connected to PPs. I agree that such physiological indicators can carry more weight than verbal ones because physiological changes cannot be easily misrepresented without awareness, and may provide a better way of assessing the paranormality of a particular case. IS is preparing a major publication about birthmark and birth defect cases, and IS'S findings are bound to be discussed in some detail in the not too distant future.

With respect to the cases which I have studied, birthmarks did not play a major role in the assessment of paranormality, although a number of my Ss had birthmarks and/or defects. In cases where PP had multiple injuries, a relatively large number of birthmarks on S could be regarded as being in agreement with PP's injuries, and consequently one mark on S would not carry much weight. Other problems which arose were the difficulty of obtaining clear descriptions or pictures of apparently corresponding injuries, and defects which PPs were reported to have had. These problems do not suggest that birthmarks and defects could not be very strong and significant features in connection with some cases, but that among the cases which I studied this did not occur.

A question which must be raised, but which cannot be readily answered at present, is whether there are any particular personality characteristics which distinguish persons who seem to have paranormal connections with a previous life from the general population or more specifically from other siblings. For the distinctly different cultural settings, the more familiar personality assessment procedures would first have to be carefully adapted, and then administered by qualified persons who at least can easily communicate without the assistance of an interpreter.

There are additional problems which make comparisons difficult. In some communities, for instance, in Turkey, children with such memories are apparently readily accepted by their peers when they are at school. This is not the case in some areas of Thailand. If an individual child is compared with its siblings, it is also difficult to assess how far the child was affected by those assumed memories, by expectations of its parents and/or by the reactions of others to these memories and how far the child has particular dispositions

5 4 J. Keil

tions also need to be made between children who are recognized as rebirth cases from the time they are born, and children who had not been regarded as possible reincarnation cases until they started to speak about a PL.

Finally Ss with birth defects and/or birthmarks which have a substantial bearing on the physiological development and appearance of these Ss, must be compared with children who are similarly afflicted, but do not claim to have memories of a previous life.

Keeping these problems in mind, there were, on the basis of my subjective observations, no particular personality characteristics that I can point out with any degree of confidence.

Some Comments on Future Research

One of the aims of this study was to look for methods and procedures beyond those already employed by IS which may help in the assessment of the paranormality of cases. In a moderate way, this was pursued by considering all the cases and by evaluating the qualitative findings in a semiquantita- tive framework.

I found that a control method which occurred to me independently had already been initiated by IS; that is, an attempt to obtain photographic records of corpses which in some communities are marked (e.g., with char- coal) in order to facilitate identification of rebirth cases. The marks made on the corpses are expected to correspond to birthmarks on rebirth cases. As the discussions with relatives of the great-grandmother case in Thailand suggest, in some communities the opportunities for such controls have already been missed because such practices are no longer continued. Even in Burma, where such identification marks are still made in some communities, I did not get much encouragement for organizing such a scheme because of the small numbers involved quite apart from the logistic problems.

Returning to the problem of unintentional and unconscious influences which by normal means may create, in young children's minds, imagined connections to PPs, it seems to me that the study of further cases by similar methods is not likely to throw much light on this problem. Of course, additional strong cases, which appear to eliminate any possible normal explanations to an even greater extent than before, will contribute to the overall strength of this research. However, it is unlikely that they would eliminate the problem completely, and it is unlikely that they would throw much light on questions which have arisen in this area.

Although, difficult to execute in practice, it seems to me that it is now time to concentrate on two main aspects; one is the cultural background which seems to be responsible for certain details in these case reports. In Burma, the subjects as well as the relatives often mention that memories of previous lives occur most frequently on dark and gloomy nights. A study of families in which rebirth cases occur, as well as of families in which there is no occur-


rence, may help us to understand how rebirth is viewed by adults in a general sense, as well as in a particular sense when a rebirth case occurs in a family. More importantly, the normal references to relatives who have died need to be examined from the point of view of how this is experienced by young children.

Folklores and customs in general, and in relation to aspects of the rebirth cases, also need to be studied; e.g., what is generally associated with dark and gloomy nights in Burma and what relevance can be found with respect to the reincarnation cases. More importantly, how do young children become aware of this? What are the games and stories that have a bearing on these questions? How are birthmarks and deformities interpreted in the communities, and how are they related to community beliefs and customs? Such a study would take some time (probably one or two years), and would require frequent access to two or more families in a village. The investigator would need to speak the local language and would need to be familiar with local customs. He or she would need to be accepted by the village community. All this is not easy to achieve, but it is in my assessment a worthwhile goal and perhaps not that difficult to attain if compared with the considerable difficulties IS had to overcome to date. If the internal conditions in Burma improve sufficiently, such a plan could be carried out in that country.

A second and somewhat related approach would be an attempt to study, in a similar but longitudinal way, an emerging case from the birth of a S. My visit to Turkey in 1988 indicated that cases which are believed to be reincarnation cases occur from time to time under suitable conditions; i.e., occasionally a child is already regarded as a rebirth case at the time of birth or even earlier. One such case was reported during my visit: A senior student (medicine, psychology, anthropology) would again need to observe and record all relevant interactions between the child and any persons with whom the child has contact. If possible, the interaction between S and others should also be compared with those that are experienced by S's closest siblings. Cultural aspects mentioned above should be studied as well, but the main emphasis should be on the child and how information about a PL (if any) manifest themselves and how they compare with information available to the child by normal means. Such an approach would have the additional problem that a particular S under observation may never refer to a PL, and that the time scale for such a study (even if not necessarily continuous) would need to be of the order of four years. Nevertheless, it seems to me that such an approach is a very promising one which may well strengthen the argument against unintentional leakage of information, but which could also contribute some important leads to the understanding of rebirth cases in general.

Additional studies, which would be useful, but less crucial from my point of view, should concentrate on obtaining background information on communities in which cases occur relatively frequently. From my experience,

56 J. Keil

Burmese and perhaps some Turkish villages would be most suitable for such purposes. Background questions which are relevant for a more detailed assessment of particular cases include: (1) How often do birthmarks and defects occur in the general population and in the rebirth population of a village or small community? IS'S body drawing forms could be used to obtain some information about the relative distribution over the body of birthmarks. (2) Are birthmarks of similar size and vividness recognized as such, to a similar extent, in the general population compared to families with rebirth cases?

Some Comments on the Question of Reincarnation

This study did not attempt to evaluate cases for evidence in support of the reincarnation hypothesis but was restricted to an assessment of the paranormality of information and other indicators that seem to connect children and adults who are alive, with persons who have previously died. Within the communities in which these cases were studied, these connections were usually regarded as manifestations of a rebirth or reincarnation process. The meaning that is attached to these terms varies to some extent between individuals and families within the same community, as well as between communities with different customs, traditions, and religious views.

Some Buddhists reject the idea that rebirth indicates the continuation of a previous individual personality. They would argue that the karma of a PP is transferred to a new and different individual. In terms of the practical consequences for a community this view is not really very different from one which does not allow for any notion of rebirth at all but which acknowledges that the good and bad deeds of persons who have lived before influence-by normal means-the present generation. The difference of this Buddhist's view is limited to the assumption that the karma of a particular PP is transferred to a particular S, and that this transfer occurs in a-from a Western point of view-paranormal way.

On the other end of the scale, a degree of continuation of an individual previous personality is accepted; i.e., S is accepted as a continuation of a PP. Although this belief was widely accepted on a somewhat abstract basis among the communities that I had contact with, virtually all the parents and relatives of Ss seem to regard the Ss as distinct new personalities. These observations may be misleading because in some communities, particularly in Thailand and Turkey, parents and relatives do not necessarily welcome indications that a child may be a rebirth case. Consequently, they may try to reject personality patterns outside a certain range of expectations. An acceptance of S as a genuine continuation of a PP would also require that S's parents have detailed memories and/or information about the particular PPs, and that does not frequently happen.

Nevertheless, my observations of an admittedly small number of actual


neutral framework through the use of the term connection. My observations do not seem to provide strong indications that the Ss are necessarily continua- tions of PPs.

Although the S of the sex change dacoit case behaved in a way that may be regarded as unusual for a female in a Burmese village, and although the Bangkok case also displayed unexpected behavior patterns apparently in agreement with the PP, I do not see these examples as clear signs that some continuing aspects of a PP is responsible for this behavior. If it is accepted that both cases contain some paranormal aspects, my observations of the cases would agree with the assumption that a limited paranormal influence which probably occurred during early childhood, could be regarded as the main and perhaps as the only source of the observable verbal and other behavioral indicators which suggest a connection to a PP. The assumed influence may become integrated into a relatively unformed but otherwise distinct personality of S, and may still be recognized in distinct behavioral characteristics many years later.

In a different context, I had suggested earlier (Keil, 1980) that the behavior patterns of most Ss who can be observed, agree with what I would expect from a normal experience or influence. For example, a city child might have had such an experience from a month long visit to a farm. Certain aspects of the child's view of the world and, in a sense, even certain personality aspects may have changed noticeably and almost permanently. We may even use the expression that after an important experience a child or person is not quite the same any more but this does not mean that the child has adopted a different personality. In a similar way, it seems to me that although some of the cases which I studied, apparently developed behavior patterns which had a noticeable and long-lasting bearing on some personality aspects of the Ss-compared to what one might expect from the observations of siblings who are not regarded as rebirth cases-this difference seems neither qualitatively nor quantitatively dissimilar from the country visit example above.

If it is assumed, as I do, that the cases support the view that paranormal aspects played a role in these connections, and if it is further assumed that the paranormal aspects can be accounted for by a temporary influence-in most cases retrospective-during early childhood, the question arises whether ESP or psi could account for this kind of influence. Those aspects of the connections between Ss and PPs which provide the strongest indication that paranormality is involved (but excluding birth marks as far as my cases are concerned) suggest that ESP would have to operate at a remarkably successful level in order to provide the necessary information to the Ss. It could be argued that the best ESP tests do not show a success rate which is in agreement with those requirements and that, consequently, ESP or psi should be rejected as a possible framework to account for the paranormal influences experienced by an S. The term "super-ESP is sometimes used to

5 8 J. Keil

though, that experimental tests suggest a goal directness (Schmidt, 1970) which is quite remarkable. That is, subjects in some tests apparently were able to be successful even though they could not, in any way, understand how the desired outcome had to be achieved. For example, subjects were asked to change the rate at which radioactive decay occurred orland at which this decay was registered. If it is also kept in mind that some spontaneous cases apparently require a great deal of accurate information; e.g., a person becoming aware (without any sensory information) of the detailed circumstances of a nonfatal accident involving a relative-it seems to me that psi cannot be completely ruled out as an alternative suggestion compared to one that maintains that PP continues to exist as S after PP's bodily death. In other words, there is at least a greater than zero probability that thoughts, emotions, and even some aspects of the personality of a person who is still alive but, for example, injured in a serious accident (on account of which death occurs within minutes) are maintained for some time and eventually transmitted, and that the relatively unformed personalities of some children are influenced by these transmissions.

This suggestion may account for apparently paranormal information about some traumatic events which PPs experienced, often just prior to death, but it does not really account for apparently paranormal information which refer to a variety of events over a longer period of a PP's life. It seems, at least, somewhat strange that the thoughts of a dying person reaching out (so to speak) to find some acceptance and resonance should contain such details as to where a pin or other minor items are hidden. It is possible though, that a rapid review of a fading life may contain various relatively unimportant items-perhaps only from the point of view of an observer but perhaps also, because during traumatic conditions, significant and unimportant details may be mixed up in an irrational manner, as in a dream. Indeed, given that some subjects do provide rather detailed information about some important events in a PP's life-as well as apparently unimportant ones-it is often surprising that other, perhaps even more important, events of a PP's life are not remembered by S if S's life is regarded as a continuation of the PP.

The above comments in agreement with my subjective belief system are not offered as an argument against the reincarnation hypothesis, but as a clarification and justification as to why I can suggest paranormal processes in connection with my cases without necessarily accepting reincarnation as a continuation of a PP in a new life of an S.

References

Keil, H. H. J. (1 980). Parapsychology-Searching for substance beyond the shadows. Australian Psychologist, 15, 145- 168.

Mills, A. (1989). A replication study: Three cases of children in North India who are said to remember a previous life. Journal of Scientijic Exploration, 3, 133- 184.

New cases in Burma, Thailand, a n d Turkey 5 9

Schmidt, H. (1 970). A quantum mechanical random number generator for psi tests. Journal of Parapsychology, 34, 2 19-224.

Stevenson, I. (1966). Twenty cases suggestive of reincarnation. Proceedings of the ASPR, 26, 1-362.

Stevenson, I. (1 975a). Cases of the reincarnation type. Vol. 1: Ten cases in India. Charlottesville, VA: University of Virginia Press.

Stevenson, I. (1 975b). Children who rememberprevious lives. A question of reincarnation. Char- lottesville, VA: University of Virginia Press.

Stevenson, I. ( 1977). Cases of the reincarnation type. Vol. 2: Ten cases in Sri Lanka. Charlottes- ville, VA: University of Virginia Press.

Stevenson, I. ( 1 9 80). Cases of the reincarnation type. Vol. 3: Ten cases in Lebanon and Turkey. Charlottesville, VA: University of Virginia Press.

Stevenson, I. (1983). Cases of the reincarnation type. Vol. 4: Ten cases in Thailand and Burma. Charlottesville, VA: University of Virginia Press.

Stevenson, I. (1990). Phobias in children who remember previous lives. Journal of ScientrJic Exploration, 4, 243-254.

Stevenson, I., & Samararatne, G. (1988). Three new cases of the reincarnation type in Sri Lanka with written records made before verifications. Journal of Scientrfic Exploration, 2 ,2 17-238.

Journal of Scientific Exploration, Vol. 5, No. 1, pp. 6 1-83, 1991 Pergamon Press plc. Printed in the USA.

0892-33 10191 $3.00+.00 0 199 1 Society for Scientific Exploration

Effects of Consciousness on the Fall of Dice: A Meta-Analysis

Intelligent Systems Laboratory, Contel Technology Center, 15000 Conference Center Drive, Chantilly, VA 22021

DIANE C. FERRARI Department of Psychology, Princeton University

Abstract-This article presents a meta-analysis of experiments testing the hypothesis that consciousness (in particular, mental intention) can cause tossed dice to land with specified targets face up. Seventy-three English language reports, published from 1935 to 1987, were retrieved. This literature describes 148 studies reported by a total of 52 investigators, involving more than 2 million dice throws contributed by 2,569 subjects. The full database indicates the presence of a physical bias that artifactually inflated hit rates when higher dice faces (e.g., the "6" face) were used as targets. Analysis of a subset of 59 homogeneous studies employing experimental protocols that controlled for these biases suggests that the experimental effect size is independently replicable, significantly positive, and not explain- ~ able as an artifact of selective reporting or differences in methodological quality. The estimated effect size for the full database lies more than 19 standard deviations from chance while the effect size for the subset of bal-

l anced, homogeneous studies lies 2.6 standard deviations from chance. We conclude that this database provides weak cumulative evidence for a genuine relationship between mental intention and the fall of dice.

Introduction

Iacta alea est. (The die is cast). -Julius Caesar (100-44 B.C.)

Forty thousand years ago, primitive peoples believed that destiny could be revealed by casting bones, or influenced by sacrifice and prayer (Radin, 1957; Watson, 1988, p. 13). It appears that astragalomancy (divination by dice) was universally employed in ancient times, with evidence for "casting lots" ranging from African tribes, to the Inuit, to the Mayans. The related concepts of chance and destiny play significant roles in the beliefs of early peoples, as reflected, for example, in Siva, the Hindu god of a thousand

Acknowledgments. We thank two anonymous referees as well as Ephraim Schechter, Ed May, Chuck Honorton, and Roger Nelson for their valuable comments on an earlier version of this article.

62 D. I. Radin and D. C. Ferrari

names, who is portrayed in some statues as determining the fate of mankind by throwing dice (Harvie, 1973).

Today, sophisticated men and women still "roll the bones" in casinos, and still fervently wish for favorable destinies. The language, garb, and technology have changed, but the primeval urge to know and control fate remains the same.

Is it possible to mentally control how "the die is cast"? Many believe that the answer must be "no," for gambling casinos generally enjoy huge profits. However, casino profits are calculated by subtracting gamblers' wins from losses. With gambling wagers running in the hundreds of millions of dollars a year in many casinos, and gamblers losing more than winning on average, the gambling industry can still record a healthy profit. Gamblers who tend to win or lose consistently undoubtedly differ in innate mathematical abilities and memory skills, but one wonders whether some of the consistent winners occasionally violate chance expectation (assuming a fair game).

Beginning in 1935, researchers' have tested the hypothesis that the fall of dice may be influenced by mental intention. Over the next half century, 52 investigators2 published (in English) the results of 148 such studies. The basic experiment is deceptively simple: A die face is prespecified, then a die (or group of dice) is tossed while the subject "wills" that face to turn up. If the subject's intention matches the resulting die face, a "hit" is scored. If more hits are obtained than expected by chance, this is taken as evidence for the hypothesis.

The relevant literature has been criticized and reviewed in detail numerous times (Edge et al., 1986; Girden & Girden, 1985; Girden, 1962, Girden et al., 1964; Murphy, 1962; Rhine, 1944), but in spite of continuing experiments and reviews, no clear consensus has emerged on the state of the hypothesis. Much of the ongoing controversy is undoubtedly the result of poor replicability of the hypothesized effect, and the fact that behind this straightforward task lies a wide array of methodological pitfalls (Barber, 1976), any one of which can legitimately cast doubt on experimental results.

Still another factor perpetuating the controversy is that previous reviews of this literature have generally focused more on the presence of real or imagined methodological flaws than on the empirical data itself. Studies examining how referees judge the adequacy of experimental reports have shown that prior beliefs about a hypothesis correlate highly with assessments of methodological adequacy (e.g., Nisbett & Ross, 1980). Thus, in general, skeptics tend to judge these experiments as methodologically flawed, while proponents tend to see the same set of experiments as methodologically adequate. This interaction between belief and assessment of flaws is important because skeptics characteristically dismiss any experiment they consider to be flawed, and hence, end up dismissing most if not all of the positive evidence. The

Effects of consciousness 63

present report uses the techniques of meta-analysis to address the question of methodological quality in a way that is more explicit and objective than the narrative review (cf. Glass, McGaw, & Smith, 198 1 ; Rosenthal, 1984).

1 Method I

Meta-Analysis

The quantitative literature review, also called meta-analysis, has become a valuable tool in the behavioral and social sciences. Meta-analysis is analo- gous to well-established procedures used in the physical sciences to determine physical parameters and constants. The technique assesses the issue of replication of a hypothesized effect within a body of studies by examining the distribution of "effect sizes" (Bangert-Drowns, 1986). In the present context, the null hypothesis (no mental influence on the fall of dice) specifies an expected mean effect size of zero. A homogeneous distribution of effect sizes with nonzero mean indicates replication of an effect, and the size of the deviation of the mean from its expected value estimates the magnitude of the effect.

Meta-analyses assume that effects being compared are similar across different experiments, that is, that all studies seek to estimate the same population parameters. Thus, the scope of a quantitative review must be strictly delimited to ensure appropriate commonality across the different studies that are combined (Glass, 1978). This can present a nontrivial problem because replication studies typically investigate a number of variables in addition to those studied in the original experiments. In the present case, because different subjects and experimental protocols were employed within the reviewed literature, some heterogeneity attributable to these factors was expected in the obtained distribution of effect sizes. However, the circumscription for the review required that every study in the database have the same primary goal (mental influence of the fall of dice) and, hence, would estimate the same underlying effect.

Circumscription of Meta-Analysis

This meta-analysis examined studies testing whether mental intention could influence the fall of dice and, in particular, cause a prespecified die face to land face up after being tossed. The usual measure of this "influence" is a statistical test that compares the number of obtained hits to chance expectation. Because this test is almost always reported as a standard normal deviate (i.e., z score) in the reviewed literature, we determined efect size as a z score normalized by the square root of the sample size (N), e = z/fi, where N was the total number of individual random events.

Excluded from this analysis were experiments where the goal was to cause dice to fall in specific locations (e.g., "placement" studies), or where the


and similar studies). Given this restrictive circumscription, it should be noted that J. B. Rhine (and well as other parapsychologists) considered that the best evidence for a mental effect on dice were consistent patterns observed in dice data rather than hits on prespecified die faces. Nevertheless, the present study examined only the die face studies in an attempt to evaluate an elementary hypothesis in a group of conceptually simple experiments. If the present meta-analysis showed no evidence for the elementary hypothesis, there would be less motivation to study more complicated hypotheses and experiments.

The Literature

The following journals were surveyed for studies fitting the above circumscription: Journal of Parapsychology, European Journal of Parapsychology, Journal of the American Society for Psychical Research, and Journal of the Society for Psychical Research. In addition, through the use of a specialized bibliographic databa~e,~ we retrieved other relevant studies from the following sources: Research Letter (of the University of Utrecht's Parapsychology Laboratory), Newsletter of the Parapsychology Foundation, Proceedings of the Society for Psychical Research, Proceedings of the First International Conference of Parapsychological Studies, Journal of Experimental Psychol- ogy, Parapsychological Journal of South Africa, and a book, The Algonquin Experiments (see the references for details). Abstracts of the proceedings of the annual conventions of the Parapsychological Association, as reported in the volumes Research in Parapsychology (for the years 1980- 1987), were also e~amined .~

Unit of Study

A "unit of study" was defined as the largest possible aggregation of non- overlapping data collected under a single target (i.e., a predefined die face). For each unit of study, the following data were recorded: The number of subjects, the intended die face target, the aim (i.e., to hit or to miss the target), the number of dice tossed at once, the total number of dice thrown in the study (N), the number of hits (H), the probability of a single event ( P ) , ~ and a series of quality criteria (described below). From H, N, and p, a standard normal deviate ( z score, corrected for continuity) and an effect size ( e = z/ \rN), were calculated. The sign of the z score (and e) was established according to whether the observed result matched the intended aim. Thus, a nega-

Parapsychological Sources of Information Center, 2 Plane Tree Lane, Dix Hills, NY 1 1746. We may have missed a few studies reported in Research in Parapsychology for the years

1972-1979. The usual hit probability was p = 116, but in a few cases the target was more than one die

face, such as faces 4, 5, or 6, in which case the hit probability was adjusted accordingly.

Effects of consciousness 6 5

tive z score obtained under intention to "aim low" was recorded as a positive score.6

Control studies were defined as experiments performed under conditions in which (a) dice were tossed with no specified target aim, and (b) the condition was defined as such in the published report. This definition excluded some studies in which control data was collected by replicating an experiment with a different aim (e.g., aiming low for the same target, or aiming for faces 1 through 6 an equal number of times). These studies were coded as "protocol7' controls.

Mean Eflect Size

Given the per study effect size e, described above, the overall weighted mean effect size was calculated as 2 = C wie,/C wi, where w . = Ni, and i ranges from 1 to Kstudies. The standard error of Cis s, = l / v z . A quality- weighted effect size was determined as 2, = C (Qiwiei)/C (Qiwi), where Qi is the quality score i. The standard error associated with

Quality Criteria

The following quality criteria cover virtually all published criticisms of the present experiments, including the issues raised in Girden's (1962) critical review of dice experiments. The criteria were coded as being either present (I) or absent (0) in each report:

(A) Automatic recording-Die faces automatically recorded onto perma- nent medium (e.g., photographed onto film). This helped eliminate human recording errors, and reduced the possibility of data selection.

(B) Independent recording-Independent (human) data recorder, other than the experimenter. If present, this criterion helped increase the trustworthiness of the recorded data, and provided a separate record for later comparison and double checking.

(C) Data selection prevented-This refers to designs with sequential-frame photographic data recordings, or studies in which data was kept in bound record books, or some other method of ensuring that all data were used in the final analysis.

(D) Data double checked-Data was manually or automatically double checked for accuracy.

(E) Witnesses present-Witnesses were present during data recording to help reduce the possibility of mistakes or fraud.

(F) Control noted-A control study was mentioned, but no details were

Six of 148 ofexperimental studies were low aim, i.e., the prescribed intention was to miss the


published. This criterion reflects the fact that the author(s) recognized the importance of performing a control test.

(G) Local control-Control data obtained under the same conditions as the experiment, using the same subject(s) and same conditions, but with no specific mental effort applied to the dice. A local control is usually conducted at the same time as the experiment, sometimes alternated with experimental conditions.

(H) Protocol control-Study designed in a manner such that controls are inherently a part of the experiment (e.g., controls for dice bias, equal number of throws for each die face, equal number of throws for a single die face in both high- and low-aim conditions, etc.).

(I) Calibration control-Long-term randomness test; usually before and after an experimental series. This differs from a local control by its longer duration, and by not being conducted under the same setup and protocols as the experiment itself.

(J) Fixed run lengths-Optional stopping ruled out by prespecified design. (K) Formal study-Prespecified methodology; analyses prespecified in ad-

vance of experimentation.

In addition to the above factors, the following criteria were coded with (arbitrarily assigned) partial credit:

(L) Dice toss method- 0.00 = dice tossed by hand, 0.50 = dice tossed by hand but bounced against a back wall, 1 .OO = dice tossed by a cup or chute (semimechanical release), 2.00 = tossed automatically by machine.

(M) Subject type- 0.00 = persons selected on the basis of prior performance or individuals

claiming special abilities,' 0.50 = experimenter(s) as sole subject, 0.75 = experimenter participated in experiment along with subjects, 1 .OO = unselected subjects.

The maximum score a study could receive is the sum of the present/absent codes, plus 2 for the dice toss method and 1 for unselected subjects, for a total of 14. This relatively simple additive method has been shown to be an excel- lent predictor of overall methodological quality (Dawes, 1979). Note that this quality measure is dependent on the details that researchers describe in their publications, thus, given that some particulars may be left out of shorter

' This criterion is assigned zero credit to reflect the skeptical suspicion that persons claiming special abilities may be more induced to attempt fraud to back up their claim. Recall that these criteria are designed to reflect a worse-case assessment of experimental methodology. It could be argued that persons selected on the basis of prior performance should be assigned higher credit.


Control Expt Expt / I

Fig. 1 . Effect size means for controls, experiments, experiments weighted by quality, and for 24 experimenters who performed three or fewer studies. Error bars are one standard error.

reports and abstracts, the present quality measure is expected to underesti- mate "true" methodological quality.

Results

Overall Analysis

The literature search retrieved 73 relevant publications, representing the efforts of 52 investigators (40 different first authors) from 1935 to 1987. Over this 53-year period, a total of 2,569 subjects attempted to "influence" 2,592,8 17 dice-casts in 148 experimental studies, and 1 5 3,288 dice-casts in 31 control studies. Sample size (i.e., the number of dice tossed) per study ranged from 60 to 240,000 (median = 5,500); the number of subjects per study ranged from 1 to 393 (median = 3). Of the 148 studies, 44% were individually significant at the 5% level.

Figure 1 summarizes the overall results: a nonzero effect for experimental studies (2 = .0 1220 a .00062),8 and a control effect within chance limits (2 = .00093 k .00255, p = .7 15).9 The overall level of statistical significance

* Weighted mean effect size estimate, plus and minus one standard error. Probabilities are estimates calculated with the "STAT PAC" inverse normal distribution

function on the Hewlett-Packard HP4 1 CX programmable calculator. Probabilities associated with effect sizes are determined from z = else, where se is the standard error.

6 8 D. I. Radin and D. C . Ferrari

(Stouffer z method) for experimental studies is z = 18.2, and for control studies, z = 0.18. Experimental effect size weighted by methodological quality is also nonzero ( F = .00723 + .0007 1). Note that the quality weighted effect size is significantly smaller than the nonweighted effect size. More about this later.

The number of studies conducted per investigator lo ranged from 1 to 2 1, with the majority of investigators (64%) reporting one, two, or three studies. To examine the possibility that a minority of investigators reporting four or more studies may have skewed the mean effect size with exceptionally large scores, we calculated mean effect size for 25 investigators reporting three or fewer studies (n = 42 studies). The result was F = .01640 + .OO 126, thus the overall effect size does not appear to be due to a few exceptional investigators.

Filedrawer Analysis

In the behavioral and social sciences it is well known that experiments with significant results tend to be published more often than nonsignificant studies (Hedges, 1984; Iyengar & Greenhouse, 1987; McNemar, 1960). In a meta-analysis, this reporting bias may inflate the estimate of effect size and overall statistical significance. This is called the "filedrawer effect." One way of assessing the effect of this presumed hidden filedrawer of nonsignificant studies is by calculating a "failsafe" number (see Rosenthal, 1984, p. 108). This is the number required to reduce the observed overall mean z score to a nonsignificant level (i.e., just below z = 1.645).

In the present analysis, the failsafe number is 17,974. This means it would take almost 18,000 additional studies averaging a null effect to reduce the observed results to a nonsignificant level. This is a ratio of 12 1 unretrieved, nonsignificant studies to each study retrieved in the meta-analysis. Ro- senthal (1984, p. 110) suggests that an effect is robust with respect to the filedrawer if this ratio is 5: 1 or greater. A filedrawer of 18,000 studies would require, for example, each of the 52 investigators involved in these dice experiments to have conducted one unpublished, nonsignificant study per month, every month, for 28 years.

There is another way to address the filedrawer problem. If investigators (or journal editors) tend to favor publication of statistically significant reports, we might expect the distribution of reported z scores to show a depressed number of studies reported just before the point of conventional significance, or a discontinuous "spike" starting at z = 1.645. Figure 2 plots the distribution of z scores; a suspected filedrawer "notch" is evident in the range 1.0 I z I 2.0. This notch gives us a way to estimate the size of the "real" filedrawer.

'O "Investigator" here refers to the first author of a publication; there were 40 different first authors in the present database.

Effects of consciousness

i 1 8 expected normal c w e

Z score

Fig. 2. Distribution of observed z scores, expected normal curve based on 148 studies, and best Gaussian fit to the observed distribution.

We begin by assuming that the positive tail of the curve (scores z 2 1.645) does not hide a filedrawer effect. Since this portion is, by definition, 5% of a normal distribution, we can reconstruct the magnitude of the "entire" distribution. The number of studies in the positive tail is 65, thus the total distribution is 20 X 65, or 1,300 studies. Subtracting the 148 observed studies from this figure leaves our estimate for the size of the "true" filedrawer at 1,152, for a ratio of nine filedrawer studies for each observed study. Now we append to the reported z-score distribution an additional 1,152 z scores, selected at random from a normal distribution and bounded within the range - 1.645 I z r 1.645. The Stouffer z score resulting from this filedrawer-adjusted database is 7.557 (p < 1 0-13).

We can also add an additional 1,152 nonsignificant effect sizes to the original distribution to determine a new, filedrawer-weighted effect size. To form the nonsignificant effect sizes, we used the randomly selected z scores generated for the above filedrawer-adjusted database, as well as the same sample sizes observed in the original data. The procedure was to take the sample sizes from the 148 original studies and replicate that set eight times (subtracting 32 extra data points) until 1,152 samples sizes were generated. Then, effect sizes were calculated in the usual way ( Z I P for each study). The resulting overall mean effect size was 2 = -00 165 -t .0002 1 (p < 10-14). This


distribution consisted of an overall sample size of some 23 million (simulated) dice tosses. Thus, it seems that the filedrawer problem cannot ade- quately explain the mean effect size or the overall level of statistical significance in this database.

It must be noted, however, that the above assessment of the filedrawer effect makes the implicit assumption that significantly negative studies are as likely to be reported as significantly positive studies. In fact, only eight studies were reported as significantly negative (compared to 65 positive), implying that significantly negative studies truly occurred some eight times less often than significantly positive studies, or that there was a reporting (or editorial) bias favoring the publication of positive studies. Which is a more plausible conclusion?

One way of addressing this issue is to compare studies published before and after the mid-1970s, for this is when the Journal of Parapsychology adopted the policy of publishing all studies, positive or negative, significant or nonsignificant (Broughton, 1987). The number of studies published after say, 1975, was 16. Of these, one was significantly negative and three were significantly positive. If we assume that reporting of post- 1975 studies is not biased according to the outcome of the study (recognizing the fact that the post-1975 sample size is small), this would imply that a genuine "mental effect" may produce something like a 3:l ratio of significantly positive to negative studies. If we reduce the full database ratio of 8: 1 by three, we are still left with a 5: 1 ratio, implying that the pre- 1975 database indeed contains a reporting bias favoring studies reporting positive significance. Therefore, we must consider the overall ( 1935- 1987) database as suspect with respect to the filedrawer problem.

Another way of examining the potential biasing effects of selective reporting is to examine the relationship between reported z scores and the year of publication. This relationship, shown in Figure 3, reveals that studies reported from the 1940s through the 1960s were substantially more significant than later studies. This suggests that (a) earlier studies were conducted in some significantly different fashion than later studies, (b) that the effect declined over the years, or (c) that the apparently greater magnitude of the earlier studies reflect an artifact of selective reporting. We believe that selective reporting is the more parsimonious explanation of this decline. (How- ever, it is interesting to note that effect size does not significantly decline over time; slope = -.001, t,,, = 1.691, p = .093.)

Homogeneity Analysis

If replication experiments were exact, and if the hypothetical ability were uniform in the population, the distribution of effect sizes might be expected to be homogeneous. However, such distributions are rarely homogeneous because replications are rarely intended to be exact. Instead, most experiments are designed to examine the same or similar phenomena under many


Fig. 3. Reported I? scores declined over time (slope = -.059, t,,, = 3.15 1, p = .002).

different conditions. Thus, to adjust for the heterogeneity introduced by different experimental conditions, and to reduce possibility that the overall mean effect size (and significance levels) may have been spuriously enlarged by extreme values, we deleted potential "outlier" studies as follows: If the homogeneity statistic" for all studies combined was significant (at the p < .05 level), the study that produced the largest reduction in this statistic was deleted. This procedure was repeated, reducing the database one study at a time, until the homogeneity statistic had become nonsignificant.

Results showed that out of the total of 148 studies, it was necessary to delete 52 (35%) to produce a homogeneous distribution of effect sizes. This may be compared with exemplary physical and social science reviews, where it is sometimes necessary to discard as many as 45% of the studies to achieve a homogeneous distribution of effect sizes (Hedges, 1987). Of the discarded studies, 33 had positive effect sizes (22% of the entire database) and 19 had negative effect sizes (1 3%). The mean weighted effect size with the remaining 96 studies is F = 0.00459 + .00087. The combined significance level (Stouffer z) of these 96 studies is z = 5.336 (p < lo-'). This homogeneity analysis suggests that the experimental effect is independently replicable.

Note that a filedrawer failsafe number is based on the assumption that a distribution contains all retrievable studies. When some studies are eliminated as statistical outliers, as in a homogenized distribution, we have by

1 " A test for homogeneity for K estimates of ei is given by H, = C w,(ei - z)~, where HK has a chi-square distribution with K - 1 degrees of freedom.

7 2 D. I. Radin and D. C. Ferrari

definition created a selected subset of studies, so one of the underlying assumptions of the filedrawer effect is violated. We assume that if the filedrawer effect is robust with respect to a full distribution, then it is also robust with respect to the homogenized form of that distribution. (However, in the present case, since the pre-1975 database is uncertain with respect to the filedrawer problem, the homogeneous subset is suspect as well.)

Temporal Analysis

Some critics assert that each new generation of parapsychologists starts from scratch, without acknowledging or benefitting from previous researchers' efforts (e.g., Marks, 1986). The assertion is part of an argument alleging that parapsychology is a pseudoscience, because it lacks a research tradition, like a "real" science. The assertion can be tested by examining assessed experimental quality over time (see Fig. 4). The trend is significantly positive, indicating that later researchers did not take note of earlier methodological criticisms. The critical allegation is not supported.

Quality Analysis

Critics have long maintained that experimental effects in parapsychology will decline as methodological quality improves, thereby reflecting a regression to the true (null hypothesis) mean effect size of zero. The relationship

Fig. 4. Methodological quality improved over time (slope = .07, t,,, = 5.12, p <lo-'). [Some of the points in this and following graphs may overlap, thus the number of plotted points may appear to be fewer than indicated by the degrees of freedom. In all graphs it was confirmed that all data points are in fact plotted.]


between assessed methodological quality and experimental effect size in the full database is nearly significantly negative (slope = -.002, t,,, = 1.845, p = .0672), but this decline is less apparent in the homogeneous database (Fig. 5).

However, the slopes from these linear regressions do not reflect the fact that we should have greater confidence in effect size estimates from studies with larger sample sizes and less confidence in studies with smaller sample sizes. Therefore, a more appropriate test would examine the slope of a weighted linear regression, where the weight reflects our confidence in the individual effect sizes. We use the inverse of the standard error as the weighting factor for each study (i-e., weight = E., where N is the sample size from study i).

The slope (4 1 standard error) of a weighted linear regression for the entire database is -0.0029 + 0.0010 (N = 148, z = slope/se = -2.874, p = .004). Thus, in the overall database, there is indeed evidence that effect size de- creases as quality increases. This is also reflected in the fact that the quality- weighted effect size is about half the unweighted effect size (as shown in Fig. I). However, this relationship is not significant in the homogeneous database: slope = -0.0006 + 0.0007 (N = 96, z = -0.8927,~ = .372). Based upon the assumption that a homogeneous database provides a better basis upon which to assess the magnitude of an effect size, it appears that the effect is not affected by methodological quality. Still, it is prudent to keep in mind that the slope observed in the overall database supports the critics' assertion.

Fig. 5. In the homogeneous database, effect size is essentially constant across methodological quality (slope = - .OO 1, t,, = .802, p = .4244).


Distributional Analysis

Some critics have claimed that some psi studies involving dice are flawed because the a priori probabilities of die faces are rarely equal (Scarne, 1974).12 These biases arise because die faces on most common dice are marked by scooping out small bits of material. For example, the "6" face has six scoops removed which cause it to have slightly less mass than the other die faces. On any random toss, this bias makes the 6 most likely to land face up, followed by the 5, 4, 3, 2, and 1 face. Thus, an experiment exclusively using the 6 face as the target may be flawed because we could not tell whether the reported results were due to a consciousness-mediated influence, or the (potentially) higher probability of a 6.

To see whether this suspected dice bias was present in the current database, we examined all reports where published data allowed the separate calculation of effect size for each die face tossed under experimental and control conditions. The biases (Fig. 6) are found in these studies, as indicated by a high correlation between the experimental and control curves (r = 0.826), as well as a significantly positive slope of a linear regression between the two curves (slope = 0.5 14, t , = 2.935, p = .0426).

However, the observed experimental effect sizes also suggest the presence of an anomaly beyond the artifacts caused by dice biases. For example, in Figure 6 we see that for each die face the experimental effect size is larger than the control; and after a Bonferroni adjustment, we find that die face 6 is significantly larger (z = 3.765, p = .0002, two-tailedI3).

But, regardless of this suggestive difference, the presence of any artifact in this group of studies raises doubts about the overall database. Therefore, we examined a subset of studies that controlled for these dice biases-studies employing design protocols where die faces were equally distributed among the six targets. We refer to such studies as the "balanced protocol subset.

Balanced Protocol Analysis

Sixty-nine experimental studies employed protocols in which targets were evenly balanced among the six die faces. These studies (Table I ) show (a) evidence for the "mental influence" hypothesis, both in terms of effect size and overall level of statistical significance; (b) an effect size that is relatively constant across different measures of methodological quality; and (c) a filedrawer requiring a 20: 1 ratio of unretrieved, nonsignificant studies for each observed study.

However, as mentioned in the analysis of the entire database, it is observed that of these 69 studies, 23 were reported as significantly positive while only

I2 It should be noted that J. B. Rhine and his colleagues were among the first to recognize and take into account the dice bias problem.

l3 The statistic used to determine significance of the difference between means was z = (t.,

- ' > c ) / i m .


-0.02 1 2 3 4 5 6

die face Fig. 6. Relationship between die face and effect size for experimental and control conditions.

The numbers indicate the number of studies contributing to the effect size estimate. Error bars are one standard errors.

one was reported as significantly negative. The large excess of significant positive studies raises the possibility of a reporting bias, so to examine this issue, we calculated the relationship between reported z scores and time of publication. A linear regression results in a virtually flat slope (slope =

-.002, t,, = .136, p = .892); a regression between effect size and time is also nonsignificant (slope = .0003, t = 1.109, p = .272). Therefore, there is reason

TABLE 1 Analysis of subset of studies using balanced target protocols. The quality vs. effect size slope

is calculated using a weighted linear regression, as discussed above in the section on Quality Analysis

Probability (Two-tailed)

Number of studies 69 Effect size .0086 1 4 .00110 < lO-I4 Quality-weighted effect size .00706 + .OO 1 15 < Quality vs. effect size slope -.0008 1 * .00007 .279 Stouffer z 7.6 17 < 1 O F 4 Filedrawer failsafe number 1410


TABLE 2 Analysis of homogeneous subset of studies using balanced target protocols

Probability (Two-tailed)

Number of studies 5 9 Effect size .00315 k .00121 .009 Quality-weighted effect size .00294 & .OO 126 .020 Slope for quality vs. effect size -.OOO 13 k .00024 .603 Stouffer z 3.188 .OO 1

to believe that this subset of the entire database does not appear to be seriously compromised by a selective reporting artifact.

A similar analysis was performed on a homogenized, balanced distribution (achieved after discarding 10, or 14% of the studies). Results summarized in Table 2 show (a) evidence for the hypothesized effect in a homogeneous subset of experiments, suggesting that the effect is replicable; and (b) the effect size is essentially constant across different measures of methodological quality.

Efect of Dzferent Quality Weighting Schemes

One might argue that the method used to assess experimental quality was nonoptimal because it gave the same (unit) weight to the different criteria used to rate quality. For example, one might assert that the "balanced protocol" quality criterion should be weighted more heavily than, say, the "data doubled-checked" criterion, because without a balanced protocol one could not be sure how much of the final effect was due to inherent dice biases. The same goes for the other quality criteria-some are arguably more important than others, and the overall quality assessment should take such differences into account.

To judge the effects of different weighting schemes, we asked 11 scien- tists14 to assign importance weights to each of the quality criteria. A scale of 1 to 5 was used, where 1 indicated that a criterion was of relatively low importance in establishing a credible dice experiment, and 5 indicated critical importance. Table 3 shows the means and standard deviations of their combined responses, listed in descending order by weight.

On observing the values in Table 3, notice that there is relatively good agreement on which criteria are judged to be important in establishing a credible dice experiment, and which criteria are judged as not so important.

l4 This included seven members of the Parapsychological Association, two members of the Society for Scientific Exploration, an experimental physicist, and a computer scientist. All were familiar with standard experimental methodologies.


TABLE 3 Means and standard deviations for quality criteria weighting factors

Criterion Mean Weight SD

Dice-machine Protocol control Fixed run lengths Automatic recording Formal study

No data selection Local control Double check data Independent recording Dice-cup Witnesses present Calibration

Subjects-unselected Subjects-experimenter and unselected Subjects-selected Subjects-experimenter only Dice-bounced

Control noted Dice-by hand

This agreement is reflected in the fact that more extreme weights generally resulted in smaller standard deviations. To reflect these agreements, we created a simplified weighting scheme in which criteria rated with a mean of 4 or above were assigned a 5, and the rest were assigned a 1. These were called extreme weights, whereas the means from the survey were called opinion weights.

To assess the effect of these new weighting schemes on the meta-analysis, we produced three arrays of quality scores for all 148 studies: one array for the original method, and one for each of the new weighting schemes. Then we examined the correlation matrix among these arrays. Table 4 shows that these three methods produced similar overall scores. This confirms the expectation that a simple combination of unit weights on various elementary variables can be as good as (and sometimes better than) so-called expert opinion (Dawes, 1979).

TABLE 4 Correlation matrix for three methods of weighting experimental quality (N = 148)

Unit Opinion Expert

Unit 1 Extreme .872 Opinion .9 16


Potential Moderating Variables

Further analysis revealed a possible moderator variable: effect size in the homogeneous database (N = 96) differed according to the type of subjects tested (Fig. 7). This suggests that unselected subjects did not perform as well as persons selected on the basis of prior testing. Honorton & Ferrari (1989) reported similar findings in a meta-analysis of forced-choice precognition tests.

Discussion

Reviews of dice experiments have raised numerous arguments why the evidence for a "mental influence" hypothesis should not be taken seriously (e.g., Girden, 1962, 1985). Besides raising an assortment of methodological and statistical criticisms that can affect all experiments, several dice-specific criticisms are often mentioned. One is the fact that the a priori probabilities of die faces are rarely equal (Scarne, 1974); a related problem is that die face probabilities tend to change as dice are used (Scarne, 1974). In addition, the effects of environmental variables on dice behavior (such as temperature and humidity) can be difficult to predict. Nevertheless, by controlling for these and other factors that can potentially introduce statistical biases, a rigorous dice experiment can be performed.

0.006 . m m m

unselected E with S E alone selected subject type

Fig. 7. Effect size by the type of subject, with standard error bars. "Unselected" were volunteers; "E with 5"' means that experimenters participated as subjects in the experiment along with other subjects; "E alone" means that experimenters were the sole subjects; and "Selected" were subjects chosen on the basis of prior testing. " N is the number of studies used to determine the effect size.


In analyzing the full database, an overall experimental effect was observed to be more than 19 standard errors from chancel5 while the control effect size was within 1 se. After partialing out those first authors who had reported three or fewer studies, the experimental effect size was found to be 13 se, thus the effect could not be explained as an artifact of a few experimenters who had reported exceptionally significant results. A filedrawer analysis indicated that nearly 18,000 additional nonsignificant studies would be required to nullify the observed results, thus selective reporting would not be a reasonable explanation for the effect.

Next, we addressed the possibility that the results were primarily caused by studies that can be identified as statistical outliers. This was examined by creating and analyzing a homogeneous distribution of effect sizes. This reduced the overall effect size from 13 to 5 se. This 5 se homogeneous result suggested that the anomaly is both genuine and replicable.

The next step was to see whether effect sizes would decline as experimental quality improved-the "regression-to-the-mean" criticism. If such a relationship were observed, it would suggest that the "perfect" experiment would consistently demonstrate null results. Analysis of the full database, using a linear regression weighted by experimental sample size, did lend support to the regression-to-the-mean hypothesis, however analysis of the homogeneous subset did not.

The next analysis examined the data for evidence of biases that can be attributed to slight differences of mass among the different dice faces. Evi- dence for this bias was found, prompting an analysis of a subset of the database in which such biases were controlled for by experimental protocol. In this subset of 69 "balanced protocol" studies, the effect size was some 7 se, the effect did not significantly covary with quality, and the filedrawer effect required 20 nonsignificant studies for each observed study. To achieve a homogenized subset of balanced studies, it was necessary to discard 10 outliers. This lowered the effect to 2.6 se, but no significant relationship with quality was observed.

The last analysis addressed the question of whether a different weighting scheme might have provided quality assessment scores that better reflect expert opinion. This analysis indicated that weights based upon expert opinion, an extreme weighting scheme, and unit weighting all produced similar scores, thus the original quality assessment method was deemed to be an adequate measure of experimental quality.

Conclusion

This meta-analysis examined a straightforward hypothesis: Mental intention is correlated with the fall of specified die faces. Based primarily on the analysis of a homogeneous subset of balanced protocol studies, we conclude

80 D. I . Radin and D. C. Ferrari

that the aggregate evidence suggests the presence of a weak, genuine mental effect.

Many parapsychologists consider the best evidence for a mental intention effect on dice to be consistent patterns observed in the data rather than simple counts of direct hits on die faces. Given that we believe the present meta-analysis provides some evidence for a mental effect on dice, we recom- mend that future meta-analyses of dice studies concentrate on examining these claimed patterns.

The possible mental effect evaluated in this study resembles similar consciousness-related effects observed in both microscopic (Radin & Nelson, 1989; Schmidt, 1987) and macroscopic (Dunne, Nelson, & Jahn, 1988) random physical systems.

References

Bangert-Drowns, R. L. (1 986). Review of developments in meta-analytic method. Psychological Bulletin, 99, 388-399.

Barber, T. X. (1976). Pitfalls in human research: Ten pivotal points. Elmsford, NY: Pergamon. Broughton, R. S. ( 1987). Publication policy and the Journal of Parapsychology. Journal of

Parapsychology, 51, 2 1-32. Dawes, R. M. (1 979). The robust beauty of improper linear models in decision making. Ameri-

can Psychologist, 34, 57 1-582. Dunne, B. J., Nelson, R. D., & Jahn, R. G. (1988). Operator-related anomalies in a random

mechanical cascade. Journal of'Scientijic Exploration, 2, 1 55- 180. Edge, H. L., Morris, R. L., Palmer, J., & Rush, J. H. (1986). Foundations ofparapsychology.

New York: Routledge & Kegan Paul. Girden, E. ( 1962). A review of psychokinesis (PK). Psychological Bulletin, 59, 353-388. Girden, E., & Girden, E. (1985). Psychokinesis: Fifty years afterward. In P. Kurtz (Ed.), A

.skeptic9.s handbook ofparapsychology (pp. 129- 146). Buffalo, NY: Prometheus Press. Girden, E., Murphy, G., Beloff, J., Flew, A., Rush, J. H., Schmeidler, G., & Thouless, R. H.

( 1964). A discussion of "A review of psychokinesis (PK)." International Journal c?fParapsy- chology, 6 , 26- 137.

Glass, G. V. (1978). In defense of generalization. Behavioral and Brain Sciences, 3, 394-395. Glass, G. V., McGaw, B., & Smith, M. L. ( 198 1). Meta-ana1y.si.s in social research. Beverly Hills,

CA: Sage. Harvie, R. (1973). Probability and serendipity. In A. Hardy et al., (Eds.), The challenge of

chance. London: Hutchinson. Hedges, L. V. (1984). Estimation of effect size under nonrandom sampling: The effects of

censoring studies yielding statistically insignificant mean differences. Journal of'Educationa1 Statistics, 9, 6 1-86.

Hedges, L. V. ( 1987). How hard is hard science, how soft is soft science? American Psychologist, 42, 443-455.

Honorton, C., & Ferrari, D. C . (1989). "Future telling": A meta-analysis of forced-choice precognition experiments, 1935- 1987. In J. Palmer (Ed.), Proceedings of Presented Papers of the Parap.sychologica1 Association 32nd Annual Convention, San Diego, CA, pp. 1 10- 12 1.

Iyengar, S., & Greenhouse, J. B. ( 1987). Selection models and the file-drawer problem. Techni- cal Report 394, Department of Statistics, Carnegie-Mellon University.

Marks, D. F. ( 1986). Investigating the paranormal. Nature, 320, 1 19- 124. McNemar, Q. (1960). At random: Sense and nonsense. American Psychologist, 15, 295-300. Nisbett, R., & Ross, L. (1980). Human inference: Strutpgies and shortcomings of social judge-

m

e

nt. Englewood Cliffs, NJ: Prentice-Hall. Murphy, G. ( 1962). Report on paper by Edward Girden on psychokinesis. Psychological Bulle-

tin, 59, 520-528. Radin, D. I., & Nelson, R. D. (1989). Consciousness-related effects in random physical systems.

Foundations o f Physics, 19, 1499- 15 14.


Radin, P. (1957). Primitive religion. New York: Dover Publications. Rhine, J. B. (1 944). "Mind over matter" or the PK effect. Journal of the American Societyfor

Psychical Research, 38, 185-20 1. Rosenthal, R, (1984). Meta-analytic procedures for social research. Beverly Hills, CA: Sage. Scarne, J. (1 974). Scarne on dice. Harrisburg, PA: Stackpole Books. Schmidt, H. (1 987). The strange properties of psychokinesis. Journal of Scientific Exploration,

1, 103-1 18. Utts, J. M. ( 1988). Successful replication vs. statistical significance. Journal o f Parapsychology,

4, 305-320. Watson, L. (1988). Beyond supernature. New York: Bantam Books.

Appendix

Publications Used in the Meta-Analysis

Averill, R. L., & Rhine, J. B. (1945). The effect of alcohol upon performance in PK tests. Journal of Parapsychology, 9, 32-41. [2]16

Breederveld, H. (1976). Towards reproducible experiments in psychokinesis: I. Experiments with dice. Research Letter, 7, 1 -5a. [ 11

Carington, W. (1935). Preliminary experiments in precognitive guessing. Journal of the Society .for Psychical Research, 29, 86-104. [1]

Cox, W. E. (195 1 ). The effect of PK on the placement of falling objects. Journal ofParapsychol- O ~ Y , 15, 40-48. [6]

Dale, L. A. (1946). The psychokinetic effect: The first A.S.P.R. experiment. Journal of the American Society for Psychical Research, 40, 123- 15 1. [ 11

Dale, L. A., & Woodruff, J. L. (1947). The psychokinetic effect: Further A.S.P.R. experiments. Journal of the American Society for Psychical Research, 41, 65-82. [3]

Feather, S. R., & Rhine, L. E. (1 969). PK experiments with same and different targets. Journalof Parapsychology, 33, 2 13-227. [3]

Fisk, G. W., & West, D. J. (1 957). Psychokinetic experiments with a single subject. Newsletter of the Parapsychology Foundation, 4, 3-7. [ l ]

Fisk, G. W., & West, D. J. ( 1958). Dice-casting experiments with a single subject. Journal of the Society for Psychical Research, 39, 277-287. [2]

Fonvald, H. ( 1 96 1). A PK experiment with die faces as targets. Journal of Parapsychology, 25, 1-12. [I]

Forwald, H. ( 1963). An experiment in guessing ESP cards by throwing a die. Journal of Parapsy- chology, 27, 16-22. [2]

Gatling, W., & Rhine, J. B. (1 946). Two groups of PK subjects compared. Journal of Parapsy- chology, 10, 120-125. [ l ]

Gibson, E. P. (1947). Note on an impromptu experiment in psychokinesis. Journal of the American Society for Psychical Research, 41, 22-28. [2]

Gibson, E. P. (1948). An exploratory PK experiment based upon throwing twenty-four dice. Journal of Parapsychology, 12, 289-295. [ l ]

Gibson, E. P., & Rhine, J. B. (1943). The PK effect: 111. Some introductory series. Journal of Parapsychology, 7, 1 18- 134. [4]

Gibson, E. P., Gibson, L. H., & Rhine, J. B. (1943). A large series of PK tests. Journal of Parapsychology, 7, 228-237. [2]

Gibson, E. P., Gibson, L. H., & Rhine, J. B. (1 944). The PK effect: Mechanical throwing ofthree dice. Journal of Parapsychology, 8, 95- 109. [3]

Greene, F. M. (1960). The feeling of luck and its effect on PK. Journal of Parapsychology, 24, 129-141. [2]

Herter, C. J., & Rhine, J. B. (1945). An exploratory investigation of the PK effect. Journal of Parapsychology, 9, 17-25. [ 1 ]

The number of experiments reported in the reference is shown in brackets.


Hilton, H., Baer, G., & Rhine, J. B. (1943). A comparison of three sizes of dice in PK tests. Journal of Parapsychology, 7, 1 72- 190. [3]

Hilton, H., & Rhine, J. B. (1943). A second comparison of three sizes of dice in PK tests. Journal of Parapsychology, 7, 19 1-206. [2]

Humphrey, B. M. (1 947). Help-hinder comparison in PK tests. Journal of Parapsychology, 11, 4-13. [ l ]

Humphrey, B. M. (1947). Simultaneous high and low aim in PK tests. Journal of Parapsychol- ogy, 11, 160-174. [2]

Humphrey, B. M., & Rhine, J. B. (1945). PK tests with two sizes of dice mechanically thrown. Journal ofParup.sychology, 9, 124- 132. [ 1 ]

Hyde, D. H. ( 1945). A report on some English PK trials. Proceedings of the Society for Psychical Resc)arch, 47, 293-296. [ l ]

Kelly, E. F., & Kanthamani, B. K. (1972). A subject's efforts toward voluntary control. Journal ofParapsycholog.v, 36, 185-197. [ l ]

Knowles, E. A. G. (1949). Report on an experiment concerning the influence of mind over matter. Journal ofParapsychology, 13, 186- 196. [2]

Lyndon, R., & Rose, R. (195 1). Psi experiments with Australian Aborigines, Journal ofParapsy- chology, 15, 122-131. [ l ]

Mangan, G. L. (1954). A PK experiment with thirty dice released for high- and low-face targets. Journal of Parapsychology, 18, 209-2 18. [2]

McConnell, R. A. (1 953). Some unpublished evidence regarding the occurrence of psychokinesis. Proceedings ofthe First International Conference of Parapsychological Studies, p. 12- 13. [ l l

McConnell, R. A. (1955). Remote night tests for PK. Journal of the American Society for Psychical Research, 49, 99- 108. [2]

McConnell, R. J., Snowdon, R. J., & Powell, K. F. (1955). Wishing with dice. Journal of Experimental Psychology, 50, 269-275. [2]

Mitchell, A. M. J., & Fisk, G. W. (1953). The application of differential scoring methods to PK tests. Jozirnal ofthe Society for Psychical Research, 37, 45-6 1. [4]

Nash, C. B. (1944). PK tests of a large population. Journal of Parapsychology, 8, 304-3 10. [2] Nash, C. B. ( 1946). Position effects in PK tests with twenty-four dice. Journal ofParapsychology,

10, 51-57. [ l ] Nash, C. B. (1 956). An exploratory analysis for displacement in PK. Journal of the American

Society for Psychical Research, 50, 15 1 - 157. [2] Nash, C. B. (1981). High- and low-dice psychokinesis with hidden targets. Parapsychological

Journal of South Africa, 2, 1 1- 17. [ l ] Nash, C. B. (1987). A blind-target PK die test with future targets. Journal of the Societyfor

Psychical Research, 54, 257-260. [ l ] Nash, C. B., & Richards, A. ( 1947). Comparison of two distances in PK tests. Journal ofParapsy-

chology, 11, 269-282. [ l ] Nicol, J. F., & Carington, W. (1947). Some experiments in willed die-throwing. Proceedings of

the Societj~fbr Psychical Research, 48, 164- 175. [5] Osis, K. ( 1953). A test of the relationship between ESP and PK. Journal of Parapsychology, 17,

298-309. [2] Parsons, D. ( 1945). Experiments on PK with inclined plane and rotating cage. Proceedings of the

Society.for Psychical Research, 47, 296-300. [5] Pegram-Reeves, M., & Rhine, J. B. (1943). The PK effect: I. Journal of Parapsychology, 7,

76-93. [ l ] Pegram-Reeves, M., & Rhine, J. B. (1943). The PK effect: 11. A study of declines. Journal of

Parapsychology, 7, 1 18- 134. [ l ] Pegram-Reeves, M., & Rhine, J. B. (1945). The PK effect: The first doubles experiment, Journal

of Parapsj~chology, 9, 42-5 1. [ 1 ] Pope, D. H. ( 1946). Bailey's comparison of a coin and a die in PK tests, Journal ofParapsycho1-

ogy, 10, 213-215. [ l ] Pratt, J. G., & Woodruff, J. L. (1946). An exploratory investigation of PK position effects.

Journal of Parapsychology, 10, 197-207. [ 11 Price, M. M., & Rhine, J. B. (1944). The subject-experimenter relation in the PK test. Journal of

Parapsychology, 8, 177- 186. [2]


Rae, J. P. ( 1978). The Algonquin experiments. Hazeldean, Ontario: Canadian Institute of Para- psychology. [ 121

Ratte, R. J. ( 1960). Comparison of game and standard PK testing techniques under competitive and noncompetitive conditions. Jolirnal of Parapsychology, 24, 235-244. [6]

Ratte, R. J., & Greene, F. M. (1960). An exploratory investigation of PK in a game situation. Jozlrnal of Parapsychology, 24, 159- 170. [5]

Rhine, J. B. ( 1943). Dice thrown by cup and machine in PK tests. Journal qfParapsychology, 7, 207-2 17. [3]

Rhine, J. B. ( 1944). The PK effect: Early singles tests. Journal qfparapsychology, 8,287-303. [2] Rhine, J. B. (1945). Early PK tests: Sevens and low-dice series. Journal u f Parapsychology, 9,

106-1 15. [ l ] Rhine, J. B. ( 1946). Hypnotic suggestion in PK tests. Journal of Parapsychology, 10, 126- 140.

121 Rhine, J. B. ( 1946). The Schwartz PK experiment. Journal ofParapsychology, 10, 208-2 12. [I] Rhine, J. B., & Humphrey, B. M. ( 1943). The PK effect: The McDougall one-die series. Journal

o f Purap.syc~hoIogy, 7, 252-263. [ 11 Rhine, J. B., & Humphrey, B. M. (1944). PK tests with six, twelve, and twenty-four dice per

throw. Jozlrnal c?fParap.sychology, 8, 139- 157. [4] Rhine, J . B., & Humphrey, B. M. (1945). Position effects in the six-by-six series of PK tests.

Jotlrnul c?fParapsychology, 9, 296-302. [ 11 Rhine. J. B., & Humphrey, B. M. (1945). The PK effect with sixty dice per throw. Journal yf

Parap.s.vchologgv, 9, 203-2 1 8. [ 1 ] Rhine, J. B., Humphrey, B. M., & Averill, R. L. (1945). An exploratory experiment on the effect

of caffeine upon performance in PK tests. Journal c!f'Parapsychology, 9, 80-9 1. [2] Rhine, L. E., & Rhine, J. B. ( 1943). The psychokinetic effect: I. The first experiment. Journal of

Parap.sychology, 7, 20-43. [3] Rose, R. ( 1950). Some notes on a preliminary PK experiment with six dice. Journal ~fParaps~v-

chology, 14, 1 16-126. [ l ] Rose, R. ( 1952). Experiments in ESP and PK with Aboriginal subjects. Journal qfParapsycho1-

ogy, 16, 2 19-220. [2] Rose, R. (1955). A second report on psi experiments with Australian Aborigines. Journal of

Parapsychology, 19, 92-98. [ 1 ] Stanford, R. G. (1969). "Associative activation of the unconscious" and "visualization" as

methods for influencing the PK target. Journal ofAmerican Societyfor Psychical Research, 63, 338-35 1. [1]

Steilberg, B. J. ( 1975). "Conscious concentration" versus "visualization" in PK tests. Journal of Parapsychology, 39, 12-20. [ 11

Thouless, R. H. (195 1). A report on an experiment on psycho-kinesis with dice, and a discussion of psychological factors favouring success. Proceedings of the Society for Psychical Research, 49, 107-1 30. [3]

Van De Castle, R. L. ( 1958). An exploratory study of some personality correlates associated with PK performance. Journal of the American Society.for Psvchical Research. 52, 134- 150. [1]

Vasse, P., & Vasse, C. ( 195 1). A comparison of two subjects in PK. Journal ofParapsychology. 15, 263-270. [I]

Woodruff, J. L., & Rhine, J. B. (1942). An experiment in precognition using dice. Journal of Parap.sychology, 6, 243-262. [2]

Journal of Scientific Exploration, Vol. 5, No. 1, pp. 85-101, 1991 Pergamon Press plc. Printed in the USA.

0892-3310/91 $3.00+.00 01991 Society for Scientific Exploration

The Wasgo or Sisiutl: A Cryptozoological Sea-Animal of the Pacific Northwest Coast of the Americas

MICHAEL D. SWORDS

Western Michigan University, Department of General Studies, Kalamazoo, MI 49008

Abstract-Various lines of soft evidence converge upon the tentative conclusion that an unclassified sea-animal of significant size is living, or at least recently lived, in the ocean waters of British Columbia. This animal has had several names within the various Amerindian cultures of that area, and has had a history among them for many centuries. The animal species may be identical or similar to other reported or historically pictured creatures worldwide.

Introduction

Most people easily assume that the oceans of our planet have many discoveries still hidden from us, and that many of these will be biological. Few scientists would deny the potential for the discovery of species even of very large size, or great antiquity. The coelacanth is, of course, the favorite example, but the truly giant octopus and giant squid seem two species from which we may have received tissue samples or body parts, making them "almost ac- ceptable" even to conservative minds. We have yet to answer the mystery of the famous Japanese fishing boat catch off New Zealand, and we have a world-full of "sea-serpent" reports (Heuvelmans, 1 969; Mackal, 1 986; Wel- fare & Fairley, 1980).

This article documents pursuit of one possible sea-going crypto-beast, called by the Amerinds of the Pacific Northwest, Sisiutl, Wasgo, Sea-Wolf, and Haietlik. It is a "candidate for reality" at least equal to the more famous cryptozoological cousins (Sasquatch, Abominable Snowman, Nessy, Mo- kele Mbembe, Champ, Ogopogo, etc.) due to the variety of pieces of soft evidence pointing to it, and the easy potential of the sea to spawn, nurture, and hide such spectacular forms.

Pacific Northwest Cultures

The native peoples of the Pacific Northwest coast are a major focus of this study in the sense that it is their culture which may best contain the experience and remembrance of this cryptozoological animal. References to what may be a nonclassified "water-beast" occur in many groups in and around the British Columbia coast (see map in Fig. 1). The peoples tend to organize

8 6 M. D. Swords

Fig. 1 . Locations of Pacific Northwest Coast Indian nations with folklore pertaining to Sisiutl- Wasgo. A = Tlingit, B = Haida, C = Tsimshian, D = Kwakiutl, E = Bella Coola, F = Nootka.

their societies in similar ways, broken into what we "Western culture" histo- rians might call clans. Each clan is designated by a "crest," a primary totem- animal which symbolically speaks to the origin of the clan and the special

Wasgo or Sisiutl 8 7

privileges and powers that the animal signifies. There are many possible totem-animals used in crests. The principle animals used in crests are easily recognizable-the Raven, Eagle, and Wolf. Other common family crests feature other well-known forms-"Blackfish" (killer whale), Beaver, Frog, Salmon, Bear, Seal, Shark, Hawk, even the Octopus, along with others. The rights to use the crest figures are passed on through inheritance and mamage, and signify the rights to certain "spirit powers." These crest figures, allegedly, sometimes arise from "significant encounters" with such an animal (Boas, 1976; Drucker, 1965; Olson, 1955).

Cryptozoological interest is readily stirred when one notices, embedded almost casually amid the lists of crest animals, a name with no meaning to modern zoology. The Field Museum of Natural History in Chicago has a plaque which points this out, and seems to "summon" questions. It is enti- tled "Northwest Coast Crest Designs," and pictures ten of the commonest designs. They are: Hawk, Killer Whale, Dogfish Shark, Raven, Frog, Beaver, Bear, Eagle, Seal, and "Sea-Wolf." The description of this latter mystery is: "Like a wolf, with long muzzle, and narrow, pointed teeth, but with curled tail, flippers, and fins. Often shown with whales between ears, in mouth, or in curl of tail."

This sort of enigma should be the sign post for a small quest, and so one was made into the stories and artifacts of the relevant peoples. It turns out, as many readers know, that the Pacific Northwest Coast Amerindians have some of the richest legends and artistry ever produced on our planet. It is a feast of the spirit and the imagination. There are wonderful magical stories of heroism, tricksters, and monsters. In fact, there are so many "monster stories" of wildly different kinds that one begins to feel that maybe there is no use in sifting through the sands of Kwakiutl or Haida imagination to find a bit of cryptozoological gold. And, we have recently been warned not to do so.

Meurger (1988) looks at the tradition of lake serpents in this region and others with the eye of a folklorist and an anthropologist. He says:

It is inevitable that these Indian monsters-which do not have the nerve to present themselves as naked as an academic's discourse before the pale-face scientist-pro- voke hilarity at times, and the desire to cut down, at any cost, the great tree of indigenous beliefs. Alas! those who believe in separating 'facts' from 'superstitions' in order to find an unknown zoological creature under the tree's foliage, risk reducing the tree trunk to the dimensions of a tooth pick before they have finished. But the myth does not allow itself to be so easily done away with. For this reason, we reject the cultivated bush of the positivistic public garden, preferring to study the sequoia of the native culture. This is the choice of ecologists.

Such spirited condemnation almost makes one want to give up on cryptozoology entirely. But, with respect, let us look twice at Meurger's warning and at the "Sea-Wolf." No one with any spiritual sensitivity would think that they had somehow reduced the meaning, the richness, of a cultural legend merely by discovering a real animal upon which some small part of the legend was based. We have known ravens for centuries. Surely the raven-

88 M. D. Swords

trickster of the myth tales is in no way diminished by the fact of the bird's reality. And deliberate nonpursuit of the possible seems a violation of the spirit of many of these exact tales. Finally, the real animal is a wonder, at least as full of richness and meaning as any human story; there would be nothing cheapened by knowing it.

Concerning the "Sea-Wolf" itself it is true that the legends of the Haida, Kwakiutl, Nootkan, etc., are filled with "monsters." The "Sea-Wolf," however, still stands out in this author's opinion. It is often used as a totem- figure, and this is rarely, if ever, true of most of the other "unknowns" (a second exception is the "thunderbird"). It is widely pictured in artifacts, and in a relatively consistent way, given the artistic styles of the peoples. It seems to correspond to rock art drawings still viewable in these and nearby areas. And it has alleged witnesses today.

Kwakiutl and Haida artifacts abound in "Natural History" museums due to their beauty of design. One of the commonest and most striking types of artifact (canoes and canoe-like poles) consistently picture the "Sisiutl," "Sea- Wolf," or "Was-go" (see Fig. 2). The great "serpent" heads dominate both ends of the canoes, leading some to refer to it as "double-headed" (thereby, obviously mythical), but other representations, and stories, make it obvious that the canoe design is the natural expression of the canoe having Sisiutl "power" or protection, regardless of which way it is going.

What are the consistent features of the crypto-animal as represented? Ac- cording to the artifacts, it is:

Fig. 2. Two examples of Kwakiutl representations of the Sisiutl-Sea-Wolf. Drawn by author from artifacts in the Chicago Field Museum of Natural History.

Wasgo or Sisiutl

1. elongated; 2. sea-going; 3. horned or crested on head; 4. snout-nosed, and with large eyes; 5. mouth prominent with sharp teeth, comparable to that of a crocodile; 6. noticeable tongue; 7. possibly finned; 8. flippered, only on the forward end possibly; 9. large and powerful, perhaps more so than the killer-whale; and

10. long tailed for swimming.

The crude combination of these features (author's version) can be seen in Figure 3.

The tribal tales do not add a lot to the picture above, but some bits and pieces are interesting to note. In the classic tale of the son with the nagging mother-in-law, the badgered son kills a lake monster, the Wasgo, and imper- sonates it. The details of possible cryptozoological interest are: the Wasgo is aquatic, extremely powerful (mightier than the killer whale), carnivorous (lured into a trap by trout bait), toothed, tailed, and finned. The animal may have had only forelegs.

Shortly the water of the lake began to churn, and the head and finned forelegs of the Sea-Wolf, which some call the Wasgo, appeared near the surface. As the huge beast rose through the open trap, snapping at the bait, the waiting hunter yanked on the line, dislodging the alderwood pole, and the split cedar snapped shut on the monster, breaking its back. In spite of this injury, the Sea-Wolf snarled and pawed and thrashed.

Later in the tale the Wasgo is credited with "claws and fins and jaws and tail" and with "wide black eyes." (Reid & Bringhurst, 1984).

Another "historical tale" proceeds thus:

Going on a hunting expedition in a canoe, his man suddenly saw a frightful monster near the shore shaped like an alligator. The Indians knew all about it and described it as a long creature with huge mouth and teeth; in every other respect like a serpent. They called it Haietlik and said it was very scarce. They offered twenty sea-otter skins for a specimen for if they should have but the least piece of this animal in their boat they were sure to kill a whale, which among them is deemed one of the greatest honors. Indeed, a piece of this magic animal insures success at all times and on all occasions.

Fig. 3. Speculative conception of the Sisiutl-Wasgo based upon author's interpretation of the relevant artifacts and folklore.

90 M. D. Swords

A quite correlative description of the Sisiutl-Wasgo, and its empowering function painted on the Haida canoes, one would think. And, note the real- ism in this tale: people paying for an actual, albeit rare, animal (Hill & Hill, 1974).

Another hint of the rarity of the animal is in the Tsimshian legend of a hero fleeing from a pursuing clan of bear-men. He succeeds by turning his canoe into the fast-swimming Was-monster. When asked by the recording anthropologist what he meant by that, the "informant-storyteller" said: "The canoe was like the mouth of a crocodile. This animal existed at the time of the former people, but not now" (Boas, 19 12). An unusual artifact, obviously the head of a crocodilian of some sort, appears in the Chicago Museum of Natural History under a Nootkan provenance, showing that these peoples did know of crocodilians, and knew that the Wasgo-Sisiutl was not this animal (Fig. 4).

This Sisiutl-Wasgo "monster," in many ways, reads more like a real (natural) than an imagined animal. In fact, the Kwakiutl people, who have a densely populated mythology of supernatural sea-monsters, specifically say that the Sisiutl is "of the earth," and not one of the myth creatures which live in the kingdom beneath the sea (Hawthorn, 1979). Other realisms, such as the woman's name which referred to the well-known sea-monster of Cousins Inlet near Ocean Falls (Olson, 1955), or the carved face of the sea-monster in honor of the Sisiutl of the Johnstone Strait near Port Neville (Boas, 1934), or the Barclay Sound sea monster from West Vancouver, so large that a full- sized canoe could fit between its claws (Hall, Blackman, & Rickard, 198 1): these bits of concreteness fit the hypothesis of an actual animal. And, right up on the coast into Alaska, the Eskimos speak of the dangerous, carnivorous sea-serpent "Tirichik" or "Mauraa" (Hall, 1979, the "nikaseenithu- looyee," something like an alligator, but gone from the Tanaina area in these times (Vaudorn, 198 I), or the "pal-rai-yuk," drawn as with the Haida on the side of the Umiak canoes for fortune and protection. It is a crocodile-headed creature with long tongue and short horns. Carnivorous and dangerous, it lives in the waters, and was common in old times when the climate was much warmer. It has fine, thick fur (although the commentator thought that this feature may have been an addition). The analysis of the myths seemed to indicate to the commentator that the crocodilian characteristics were "very

Fig. 4. Drawing by author of Nootkan Mask held in collection of the Chicago Field Museum of Natural History. Dominant coloration is pea-soup green.


remarkable" and predated contacts with modern whites. The "problem" with this tale, cryptozoologically speaking, is that "pal-rai-yuk" has an impossible physical feature: six legs. These are very odd legs: long in back, short in front, and rudimentary in the middle-more grist for the mystery mill (Nel- son, 1983). For two versions of the Eskimo canoe-serpent, see Figure 5.

Despite the anatomical peculiarities of pal-rai-yuk, it is the opinion of this author that there is ample reason for cryptozoological interest here. Among a collection of small ivories at Chicago's Field Museum of Natural History is one of Tlingit provenance labeled "sea monster." Indeed, it seems a good candidate for our quarry, and it lies alongside several other artifacts, easily identified, attesting to the realistic craft skills of the people. It is pictured as 1 aquatic, elongated, toothed, tongued, snouted, short-homed, big-eyed, fore- flippered-in fact, a perfectly good Sisiutl-Wasgo in a more zoologically credible artistic rendering than most of the wonderfully exotic artifacts of the

I area. As an amusing, perhaps meaningless, aside, the creature is arched, head

Fig. 5. Two versions of the Eskimo Umiak (canoe) guiding sea-serpent: top version from a canoe from Nunivak Island (Field Museum), and bottom, redrawn after representations in Nelson (1983).

9 2 M. D. Swords

and tail rearing, in a similar manner to Hans Egede's sea serpent of 1734 in the Atlantic arctic (see Fig. 6).

A final, certainly related, portion of evidence from our native Amerindian populations comes in the form of petroglyphs, judged by many to be quite old. Some of the rock carvings have been specifically labeled as referring to Sisiutls, Wasgos, or Haietliks by tribal descendents of the people who made them. Sometimes, as in the Kuleet Bay "rain god," they are referred to other stories. Sometimes, no one remembers what the rocks are telling us. The petroglyphs certainly do not all look the same (see Fig. 7 for a "rogue's gallery" of some of them). But they do seem to refer to an elongated, aquatic animal with some projection on the head, and a swimming tail. The glyphs support what the people say: that the animals have been in these areas for a very long time. The real question, it seems is: Are they still? (Cain, 1950; Hill & Hill, 1974; Meade, 197 1).

Modern Interest

As the existence of the International Society of Cryptozoology attests, there has been much 20th century interest in the general concept of "sea serpents." In the literature, Bernard Heuvelmans' In the Wake of the Sea- Serpents (1968) is, of course, the definitive work, but other writers have contributed substantially (e.g., Dinsdale, 1966; C . Gould, 1989; R. Gould, 1930; Mackal, 1980; Mangiacopra, 1975- 1982). Two "lake monsters" books, in particular, by Costello (1 974) and Meurger (1 988), have broached the subject of the present study, in part. But the most significant work scientifically has been the data (case) gathering research of LeBlond and Sibert (1973). Drs. LeBlond and Sibert of the University of British Columbia's Institute of Oceanography made a "direct eyewitness" survey of people who had felt they had observed an unusual ocean-going animal. The survey was mostly via mail with a few personal interviews. Altogether, they collected 23 eyewitness accounts.

The sightings were spread out along the northern Pacific coast from Ore- gon to Alaska with an understandable, undoubtedly population-related,

Fig. 6. Comparison of the Chicago Field Museum of Natural History's Tlingit "Sea-Monster7' with Hans Egede's rearing sea-serpent of 1734. (redrawn by author).

Wasgo or Sisiutl 93

Fig. 7. Pacific Northwest petroglyphs sampling of sea-monsters: A = Kulleet Bay Rain God; B = Dogfish Bay, Vancouver; C = Nanaimo River; D = Nanaimo River; and E = McLaughlin's Canyon, Washington.

concentration around the southern tip of Vancouver Island. The relevant sightings (27) from Heuvelmans' collection were similarly plotted and compared. LeBlond and Sibert try mightily to make classificatory sense out of the disparate descriptions. The task proves almost impossible, the characters described being too sparse, too vague per case, or simply not correlating well between cases.

Heuvelmans attempted to classify the sea-serpent data in his book by physical type and geographical location. He ended up with nine classifica- tions of which three have very few representations. Reading his tentative descriptions, and reflecting on the tribal concepts of the Sisiutl-Wasgo, the current author rates them as follows (from most numerous world cases to least):

(A) Long-necked: It is not impossible, but unlikely, that this category is describing the Sisiutl. The animal does not seem elongated enough, but the main disparity is its alleged small to invisible eyes. However, the animal is described, elsewise, with a proper head and occasionally two horns. The "dog-like" head could lead to the "Sea-Wolf" image, but the teeth must be more vicious in prominence.

(B) Mer-horse: This is the animal that Heuvelmans locates in the British Columbian region. The description, again, partially fits and does not. The large eyes and mouth are plusses, but the prominent mane is not (unless one

94 M. D. Swords

credits the tufts of hair placed on some of the Sisiutl poles behind the Sea- Wolf head). There is, again, no discussion of the prominent carnivorous teeth (perhaps, sea-observers do not get close enough to encourage that display). A "maybe"; but no horns.

(C) Many-humped: The body of this animal could fit much of the Sisiutl description (long, tailed, single pair of flippers), but the unspectacular head (medium eyes, no horns, no mention of teeth) make evaluation even more guesswork than usual. The projecting fin just behind the head relates especially to many petroglyphs.

(D) Super-otters: This animal seems wholly unacceptable as a Sisiutl; no horns, small eyes, not a very ferocious appearance, unless we say that Hans Egede's sea-serpent was type of these (which Heuvelmans does).

(E) Many-$inned: The nickname gives this one away as a poor candidate for Sisiutl.

(F) Super-eels: Also, of little chance, due to lack of appendages. (G) Of the lesser reported classes, the "Marine Saurian" deserves mention:

very long crocodilian head with large eyes and menacing teeth; pairs of flippers and a long tail. If small horns or a head crest were mentioned, one would say that this rarely described beast would be the prime candidate.

Paul LeBlond and J. R. Sibert looked at their own new data with Heuvel- mans7 categories in mind. They also decided that the schema was not helping them much with their animals. They ended up producing a three-part classification as follows:

1. "A creature with large eyes set laterally on a horse or camel-shaped head mounted at the end of a long neck. This animal has short dark brown fur and no mane. . . . This animal strongly resembles Heuvelmans' Mer- horse, but for the absence of the name." My evaluation: better, but again, no horns and no mention of the ferocious teeth.

2. "A similar animal but with much smaller eyes, sometimes described with horns or a mane." My evaluation: horns plus, eyes minus.

3. "A long serpentine animal, showing loops of its body above the water and swimming fast, but with much thrashing. Its head is described as sheep- like and it has a dorsal fin running along its back." My evaluation: pretty vague for the kind of details derivable from artifacts and folklore.

Once again, everything "sort o f 7 fits, and does not. Maybe that is to be expected, given the poor quality of most of the modern witness observation versus the intense (albeit, rarely achieved) familiarity of native Amerindians who had been face-to-face with live animals and carcasses. Naturally, the latter group would be more interested in eyes, teeth, and horns, and better able to feature them in their tales, glyphs, and artifacts. If we could take LeBlond and Sibert's first category, give it their second category's horns, and make a compromise with Heuvelmans' "Marine Saurian's" head, such

Wasgo or Sisiutl 95

might make a very good Sisiutl indeed. Alas, no obliging reporter with an equally obliging Sisiutl has done this for us recently.

A last note on reports before moving on: when LeBlond and Sibert collected their witnesses they were surprised that none of these witnesses had already made "public reports," which were subsequently collected and described by Heuvelmans. I am not. From the perspective of someone who has a little time in another field of anomalies (UFO studies), it is common knowledge there that fewer than 10% of sightings are estimated to ever be reported. These estimations come from both official formal polls and informal lecture hall canvassing. Even among sympathetic people (e.g., those who come to a UFO lecture), the large majority never report their own experience. The relevance of this, for our current topic, is that if Drs. LeBlond and Sibert were able to dredge out a few tens of willing witnesses, there are probably at least a few hundred out there. One way of mobilizing those witnesses (in UFOlogy) has been to give well-advertised lectures on the subject. If the speaker maintains a respectful, open-minded, exploring attitude during the talk, witnesses inevitably come forward. Of course, if the speaker acts in a derisive or demeaning fashion, we can forget that data forever.

What Goeth There?

In Searching for Hidden Animals, Mackal (1980) reviews the stories of Canadian Lake monsters, and comes up with almost precisely the same description as one derives from artifacts picturing the Sisiutl-Wasgo. My bracketed interjections occasionally interrupt Dr. Mackal's description as follows:

The animals look most like a log, elongated, serpentine, no thickened body cen- trally, about 12 meters (40 feet) long, although a range of smaller sizes has been reported and a few larger, up to say 20 meters (70 feet). The head tapers toward the snout and is somewhat flattened top to bottom. Comparison is most often made to the head of a horse, sheep, or alligator. Eyes are definitely reported large enough to be clearly noted. Very occasionally a pair of protuberances referred to as "ears" or "horns" have been noted. Nostrils have not been reported as such [Sisiutl definitely has them], but "blowing" has been observed, although rarely.

The skin is described as dark green to green black or brown to black and dark brown [these colors are prominent in Kwakiutl artifacts of the Sisiutl]. Occasionally, the color is given as gray to blue-black or even a golden brown. Most often the skin is smooth with no scales, although part of the body must possess a few plates, scales, or similar structures observed by close-up viewers and compared to the lateral scutes of sturgeon [one wonders if this feature is a contamination from misinterpreted large lake sturgeon encounters, a notorious source of bogus lake monster reports]. Most of the back is smooth, although a portion is saw-toothed, ragged-edged, or serrated. Sparse hair or hair-bristle structures are reported around the head, and in a few cases, a mane or comb-like structure has been observed at the back of the neck. The food appears to be fish, since a number of people have observed behavior that can best be attributed to an active fish predator.

96 M. D. Swords

Almost all of which is a very fitting description of our British Columbian sea-monster. Dr. Mackel says that his conception grows from 200 remarkably repetitive accounts.

At that point in his book, he delivers his hypothesis for the nature of the lake monsters: "a small population of primitive whales," archeoceti, Basilo- saurus cetoides, or near cousin. Since these "basilosaurs" or "zeuglodons" have supposedly been extinct for several million years, this would be a fine cryptozoological triumph indeed. I, personally, have no objection to the concept that the zeuglodons, or some very similar elongated mammalian carnivores represent an outstanding candidate for Sisiutl-Wasgo (see Fig. 8 for a possible rendition of a zeuglodon). The folklore and artifact evidence cannot prove anything, of course, but is certainly consistent with such a view. Whether or not ocean-going animals of this size can swim (almost) unnoticed deep into Canadian lakes, in order that they may also be the legendary lake monsters, I'll leave for further "on-the-ground" research. Concerning the objections about Basilosaurus being extinct, however, one really needs to face up to a hushed truth about our fossil record: it is a horribly fragmentary and inadequate basis on which to be making absolutist declarations. Millions of years in certain locations and whole species never make it into our petrified boneyards for analysis. Some species are known only by tracks they have made. How many more left even less? And how many known animals persisted far longer than we can measure? The proper comment might be: the zeuglodons, as far as we can guess on the fragments of data we have currently, were extinct several million years ago.

A minor curiosity, for what it is worth: while musing over another interesting, mysterious place, it seems that this author may have run into the Sisiutl unexpectedly. The place was the Plain of Nazca, and the fascinating drawings thereon. Being no Von Danikenite, but interested in this wonder nevertheless, I sought information at the nearby Museum of Ica for what the drawings were all about. The director pointed out Nazca culture pottery

Fig. 8. Author's rendition of a Zeuglodon or Basilosaur.

Wasgo or Sisiutl 97

displaying the common figures of both the pots and the plain. There were the spiders, geese, hummingbirds, and . . . the Sisiutl? There, on the pottery, was the rearing sea-monster with large eyes, sharp teeth, horns, elongated body with swimming tail, and only forelimb paws. And, on the Plain itself, albeit less distinctly, the same animal writ large (Fig. 9). The relevant cultural period for the pottery, and the likely dating for the drawings on the Plain, is c. 200 B.C. to 400 A.D. (Hawkins, 1974). The greatest expert on the Plain felt that the major drawings were known animals which had achieved an astrolo-

Fig. 9. Comparison of Nazca pottery motif of "Whale-Sea-Monster" (top) with Nazca Plain drawing of the "Whale" by local Ica museum artist (author's collection).

98 M. D. Swords

gical-calenderic status with the peoples of the area (Reiche, 1976). The ring- tailed monkey represented both the Big Dipper and the beginnings of the rainy season, for example. Our "Sisiutl" drawing is spoken of thus:

A small figure, the line of its contours continuing inside, as is the case in the fish. . . . But for its ears, the figure could represent a whale. . . . Its direction, as seen from its centre, pointed to the setting sun on December 2 1 st, which is the summer solstice on the Southern hemisphere.

Perhaps it is possible that our southern ancestors also knew of the great "primitive whale" or Sisiutl, and they, fishermen also like the Haida and Kwakiutl, encountered it chasing the migrating shoals at sea. As an amusing aside, as we talked of Nazca, the director and I were stepping around several messy piles of recently excavated bones which had been laid inconveniently on the Museum floor: the bones of zeuglodons.

With that bit of unscientific romance to inspire us, let us take a couple of more ancillary excursions into our subject, and then let the Sisiutl slide once more into the mysterious depths. Back to Mesopotamia briefly, to the beginnings perhaps of the sea-serpentldragon concept we go. One of if not the earliest representation of the great water-serpent (in Mesopotamian legend the goddess Tiamat) appears on an Assyrian cylinder seal held in the British museum (Brookesmith, 1984; Smith, 19 19) (Fig. 10). The characteristics of the beast: elongated, aquatic (Tiamat is goddess of the ocean), large eyes, large mouth, horns, and only forelimbs like our friends the zeuglodons, and perhaps the Sisiutls. In fact, dragons in lore from the beginning only have two feet and kill with their powerful tails (Benton, 1990). "Dragons" meant 'b~erpent~" in ancient times, and in Vergil, small household snakes were called "dracontuli," or little dragons.

Lastly, via a communication from Professor Glanville Price to Dr. Le- Blond to myself (1 987):

the Picts have a special artistry which expresses itself in symbol-stones. Amid the interesting designs, a common motif is of the "Pictish Beast" [Fig. 1 I]. This wonder-

Fig. 10. Drawing of the figure of the Ocean goddess, Tiamat, from an Assyrian cylinder seal, as pictured in Brookesmith ( 1984).


Fig. 1 1. Pictish artistic motif of the so-called "Pictish Beast," courtesy of Prof. Glanville Price.

ful creature is sometimes referred to as "swimming" and has the following qualities: elongated, scroll-like "paddles"? instead of digited hands or feet, long neck and head, big mouth, big eyes, a hint of a horn, and perhaps a mane. . . .

Perhaps a relative of our water-dragon Sisiutl friends? Professor Price thinks that the beast is aquatic, and notes that it was the Picts to whom St. Columba was going when "something" rose out of the river Ness, "way back when" at the start of this Quixotic quest.

Mentions could be made of many other reported sea creatures, such as Chessie (the serpentine "monster" of Chesapeake Bay on the Atlantic coast), Champ of Lake Champlain, and Ogopogo of Lake Okanagan, but in the interest of preserving space in the journal and time in this author's life, I will refer the avid reader to the works of the likes of Messrs. Mangiacopra (1 976- 1982), Mackal et al. (1980), and to our fine sister journal, Cryptozoology, edited by Richard Greenwell. In my estimation, the details of possible animal characteristics derivable from the Pacific Northwest Coast resources exceed these Atlantic ocean and lake monster materials, and I have chosen to focus my attention in the more limited geographical area.

Conclusion

An adventurous trek through folklore, artifacts, petroglyphs, and anecdo- tal sightings proves nothing, of course. But it remains intriguing how so many lines of soft evidence can point in the same general direction. There is no "best" evidence here. It is all a feeling for the realness of the described animal versus the unnaturalness of many of the other beings inhabiting the same cultural heritages. And it is a feeling about something feasible, some-

100 M. D. Swords

thing which would not dramatically violate any current deep-rooted scientific wisdom: a large ocean-going animal of carnivorous intent, which has interacted with fishermen populations along the coasts of our continents for centuries; always rare, always spectacular, it achieved mythic proportions before being pushed or frightened away by the more powerful technologies of modern man. Perhaps it is still there, as reporters occasionally maintain, or perhaps it finally has gone extinct through the pressures we have placed upon it. Either way, or never at all, there seem to be enough reasons to maintain a curious attitude about our Sisiutl, and even cast an open-minded eye upon the sea in hopes of seeing something wonderful.

References

Benton, J. R. (1990). Antique antecedents of medieval monsters. Presented at 25th Interna- tional Congress on Medieval Studies, May 10- 13, Kalamazoo, MI.

Boas, F. ( 19 12). Tsimshian texts. Leyden: E. J . Brill. Boas, F. (1934). Geographical names o f the Kwukiutl indians. Berkeley, CA: University of

California Press. Boas, F. (1976). Kwakiutl culture. Millwood, NY: Kraus. Brookesmith, P. (1984). Legends ofthe Lost. London: Orbis. Cain, H. T. (1 950). Petroglyphs qf Central Washington. Seattle, WA: University of Washington

Press. Costello, P. (1974). I n Search of'Lake Monsters. New York: Berkley. Dinsdale, T. (1966). The Levialhans. London: Routledge and Kegan Paul. Drucker, P. ( 1965). Cultures ofthe North PaciJic Coast. New York: Harper and Row. Gould, C. (1989). Mythical monsters. New York: Crescent. Gould, R. (1 934). The Loch Ness monster and others. London: Geoffrey Bles. Hall, E. T. ( 1975). The Eskimo storyteller. Knoxville, TN: University of Tennessee. Hawkins, G. ( 1974). Prehistoric Desert Markings in Peru. In P. H. Oehser (ed.), National Geo-

graphic Socicty Research Reports (pp. 1 17- 144), 1967 Projects. Washington, DC: National Geographic Society.

Hawthorn, A. (1979). Kwakiutl Art. Seattle, WA: University of Washington. Heuvelmans, B. (1969). In the Wake oj'the Sea Serpents. New York: Hill and Wang. Hill, B., & Hill, R. (1974). Indian petroglyphs qf'the Pacific Northwest. Saanichton, British

Columbia: Hancock House. LeBlond, P., & Sibert, J. (1973). Observations qflarge unidentijied marine animals in British

Columbia and adjacent waters, manuscript report no. 28. Vancouver: Institute of Oceanogra- phy, University of British Columbia.

Mackal, R. (1 980). Searchingjbr hidden animals. Garden City, NY: Doubleday. Mackal, R. ( 1986). Biochemical analysis of preserved. Octopus giganteus tissue. Cryptozoology,

5, 55-62. Mangiacopra, G. (1976-77). The great unknowns of the 19th Century, part 1: Sea and Shore,

Winter 1976-77: 20 1-205, 228; part 2: Sea and Shore, Spring 1977: 17-24, 58; part 3: Sea and Shore, Summer 1977: 95- 102, 104; part 4: Sea and Shore, Fall 1977: 175-1 78.

Mangiacopra, G. (1978). Lake Champlain: America's Loc Ness, part 1. Sea and Shore, Spring 1978: 2 1-26; part 2: Sea and Shore, Summer 1978: 89-92.

Mangiacopra, G. ( 1980). The great unknowns of the 20th Century, part 1: Sea and Shore, Spring 1980: 1 3-20; part 2: Sea and Shore, Summer 1980: 123- 127, part 3: Sea and Shore, Fall 1980: 193- 196; part 4: Sea and Shore, Winter 1980; 259-26 1.

Mangiacopra, G. (1982). Sharlie: a preliminary report of possible large animals in the Payette Lakes of Idaho. Sea and Shore, 12(1): 43-46.

Wasgo or Sisiutl 101

Mangiacopra, G. (1 982). The two monsters of Flathead Lake, Montana. Sea and Shore, 12(2): 93-96, 114.

Mangiacopra, G. (1 982). Canada's La Bete du Lac: The beast of Lake Pohenegamook. Sea and Shore, 12(3): 138-140, 18 1.

Meade, E. (1 97 1 ). Iudian rock carvings of' the PaciJic Northwest. Sidney, British Columbia: Gray's Publishing.

Meurger, M. (1988). Lake monster traditions. London: Fortean Tomes. Nelson, E. (1983). The Eskimos about the Bering Strait. Washington DC: Smithsonian. Olson, R. (1955). Notes on the Bella Bella Kwakiuitl. Anthropological Records. Berkeley, CA:

University of California. Price, G. ( 1987). Personal communication. Reiche, M. (1976). Mystery on the desert. Stuttgart: Heinrich Fink. Reid, B., & Bringhurst, R. ( 1984). The raven steals the light. Seattle, WA: University of Washing-

ton. Smith, G. E. (1 9 1 9). The evolution of the dragon. Manchester, UK: University Press. Vaudorn, G. (198 1). Tanaina talesfrom Alaska. Norman, OK: University of Oklahoma. Welfare, S., & Fairley, J. (1980). Arthur C. Clarke's mysterious world. New York: A and W

Publishers.

Journal of Scientific Exploration, Vol. 5, No. 1, pp. 103- 1 1 1, 199 1 Pergamon Press plc. Printed in the USA.

0892-33 1019 1 $3.00+.00 01991 Society for Scientific Exploration

The Extraterrestrial Hypothesis Is Not That Bad

McDonnell Douglas Corporation, Huntington Beach, CA 92647

Abstract-The Extraterrestrial Hypothesis (ETH) explanation of Unidenti- fied Flying Objects (UFOs) has not been attractive to many scientists because of the apparent requirement to exceed the speed of light in making such trips. It is postulated that if the basis of gravity control systems were discovered, and that if the speed of light can either be raised or exceeded using such devices, then the pattern of UFO reports is consistent with the ETH. Recently, five arguments against the ETH have been advanced by Vallee in this Journal (1990). Each argument can be countered as follows: (1) thousands of visiting civilizations account for the wide variety and large number of reports; (2) the frequently reported similarity to homo sapiens shape may be due to historical interactions, or due to biological optimization; (3) the primitive treatment during reported abductions may be the exception from some less ethical civilizations; (4) the historical extension of the UFO phenomenon is to be expected; and ( 5 ) the near-magical reports of time, space, and light manipulation is due to high technology. Travel distance and travel time for constant acceleration and deceleration trips are displayed graphically, based on the feasibility of hyperoptic gravity control transportation systems. Convenient astronomical distance reference points are noted.

In a recent issue of this Journal (Volume 4, No. 1, 1990), well-known and respected UFO researcher Jacques Vallee presented five arguments against a simple extraterrestrial origin of unidentified flying objects (UFOs). The purpose of this study is to take polite issue with each of these arguments, based on some different assumptions.

That the universe includes a large number of intelligent civilizations is usually not viewed as controversial. Indeed, the classic book by Shklovskii and Sagan (1966) postulates, in some detail, that these large numbers of civilizations are likely to include some that are more advanced and some less advanced than our own. The basic problem most scientists encounter in projecting this large population into UFO visitations is the assumption that our current science, and all science of the future, will never permit travel faster than the speed of light. This is the linchpin assumption, because it is immediately clear that, if the opposite assumption is made (such as "it is very simple to travel at 10% if you know how to do it"), then it is hardly necessary to go through the details to imagine that one would expect a lot of visitors to Earth all the time.

It cannot be proven at present that it is possible to travel faster than the speed of light. However, what we can say is that all experiments accelerating

104 R. M. Wood

particles to high speeds (but never faster than the speed of light) have always used electromagnetic radiation to do so. This is like trying to use sound waves to propel a vehicle to supersonic speeds. Thus, it is not surprising that we have no hyperoptic records.

Most people would also agree that records of our history of science permit us to say that coherent light, which does not occur naturally, never existed on this planet until the laser was invented. Analogously, our conventionally known civilizations have never discovered how to control gravity. When they do, it may turn out that the interaction with gravity control energy sources permits travel at hyperoptic speeds. As Hynek noted (1 966):

I have begun to feel that there is a tendency in 20th-century science to forget that there will be a 2 1 st-century science, and indeed, a 30th-century science, from which vantage points our knowledge of the universe may appear quite different. We suffer, perhaps, from temporal provincialism, a form of arrogance that has always irritated posterity. (p. 329)

These analogies or quotes do not prove that hyperoptic travel is possible. However, zero-point energy fluctuation physics described by Boyer (1 985), coupled with condensed charge technology studied by Puthoff ( 1 990), along with the interpretation of gravity being merely a Van der Waals force (Pu- thoff, 1989), may be relevant. If these ideas are also blended with the standard Reissner-Nordstrom solution to the Einstein equation (Misner, Thorne, & Wheeler, 1973), which includes the effects of charge on the effec- tive dielectric constant of the vacuum, values of the speed of light would be charge-dependent by virtue of being dielectric-dependent. Let us temporarily assume that travel at speeds well in excess of 3 X lo8 m/sec is easily achievable (such as lo6 times). (We will later examine possible parametric relationships between assumed travel speed, civilization density, and visitation frequency.) With this assumption in mind, let us deal with the five arguments against the ETH. Table 1 summarizes Vallee's five arguments against the extraterrestrial hypothesis, and the principal reason why each argument may not be valid. Vallee's arguments are shown in italics as follows.

Argument One: Close Encounter Frequency

"Unexplained close encounters are far more numerous than required for any physical survey of the earth."

"It is difficult to claim that space explorers would need to land 5,000 times on the surface of a planet to analyze its soil, take samples of the flora and fauna, and produce a complete map" (p. 107). Vallee later argues that the number is perhaps nearer to 50,000 close encounters due to incomplete reporting, 100,000 if correcting for other continents, and perhaps a million if corrected for population density, and another factor of 14 if corrected for time of day, leading to 14 million landings in 40 years. This is a reasonable

Extraterrestrial hypothesis 105

TABLE l Summary of Vallee's arguments against the ETH and Wood's rebuttals

Argument Number Synopsis Rebuttal

1 Too many close encounters for an earth survey

2 "Alien" body structure not alien enough

3 Treatment during abductions is primitive

4 UFOs are not a contemporary phenomenon

5 Manipulation of time and space is awesome

Not if there are thousands of surveying civilizations

"Alien" evolution not required to be independent of "ours"

Some civilizations may not have developed social skills

This is just as easily an argument for the ETH

"Any sufficiently advanced technology is indistinguishable from magicw-Clarke

estimate. If we assume that each civilization is authorizing 1,000 close encounters on the average (perhaps 100 if we, Earth's occupants, appear dull, 10,000 if "we" are interesting, or if "their" bureaucracy is big), we should expect 14,000 other civilization visits in 40 years. This converts to (14,000/ (365)(40) = 0.95 visits per day, which is perhaps higher than one might expect (but which will be treated parametrically later). Furthermore, if Val- lee is a bit high in his estimates, say by a factor of 100, this leads to only one visit every 100 days. Additionally, if the visitors were to use "mother ships," as have been reported extensively, rather much like we use aircraft carriers to take our tactical vehicles to a location efficiently, then one civilization's visit might easily spawn thousands of landings. The enormous variety in UFO shapes is not inconsistent with different civilization origins, and the occa- sional multiple separate reports of identical shapes would also be expected. In addition, motivations other than simple physical surveys can be postulated, such as extraterrestrial vacations, systematic extraction of natural materials, and, of course, the relatively recent reports of genetic attention toward either humans or animals.

If we also assume that our presence is known and that advanced civilizations do not waste much effort visiting uninhabited planets, such visitation rates should not be surprising. Exploration scenarios, which assume that all planets must be explored (Markowitz, 1966), do not take advantage of assuming that "they" know where "we" are, a feat not inconsistent either with alleged paranormal sensing or with the existence of comprehensive civilization maps.

Argument Two: Physiology

"The humanoid body structure of the alleged "aliens" is not likely to have originated on another planet, and is not biologically adapted to space travel."

106 R. M. Wood

Here, Vallee makes the explicit assumption that the ETH states that independent evolution has occurred on other planetary bodies. Independent evolution would seem to be his assumption and is not essential to the ETH at all. I would think that the simplest form of the ETH is as stated by Friedman ( 1979), "In other words, some UFOs are somebody else's spacecraft" (p. 48). Reports of homo sapiens emerging from space ships can mean one of two things: ( I ) the independent evolution of life forms winds up with homo sapiens, because it is very functional; or (2) the development of homo sapiens did not occur independently on different planets. More generally, one can argue that the species that have been reported in close encounters generally reflect the population distribution of the nearby universe. It is not at all clear that different surface gravity, different inclinations to the equator, different day lengths, or different stellar light intensities would guarantee substantially different final life forms. It is possible that 98% of the life forms of the universe occur right here on earth? If space travel can happen easily and often, we would expect a "transpermia" situation (summarized by Gordon, 1966) in which all life forms can readily go everywhere.

Vallee suggests that if aliens did naturally evolve to a humanoid shape, they might modify their bodies using genetic engineering to enhance their ability to work and survive in space. This idea would only have merit if the travel times were long compared to species lifetime. "Short" trips at many times the speed of light would not be important enough to warrant genetic twiddling. The fact that the physiology of some of the "aliens" conforms to ours is totally consistent with an ETH that provides for species interactions in the past or for near-identical evolutionary pathway. In either case, the occupants of the UFOs might simply represent the population distribution of the nearby universe.

Argument Three: Abduction Reports

"The reported behavior in thousands of abduction reports contradicts the hypothesis of genetic or scientijic experimentation on humans by an advanced race."

Vallee assumes that such visitors would know at least as much as we do in the fundamental scientific disciplines such as physics and biology. It is certainly reasonable to assume that if they have reached us here from other planets that they know more about transportation systems than we do. How- ever, each civilization, in principle, could be characterized by a number of attributes such as social skills, medical skills, ethics, religion, biology, physics, chemistry, psychic skills, genetics, electronics, materials sciences, and other subjects perhaps yet to be identified. It is not obvious that all these skills would proceed in the development of an advanced civilization at the same pace. There could be some who are strong on propulsion, but weak on medicine or genetics; some who are strong on paranormal, but weak on hard sciences. Current analogies might include the Australian aborigines, who

Extraterrestrial hypothesis 107

allegedly have psychic skills, but little technology; or street gangs, which have some technology, but little ethics.

The abduction reports such as those described by Hopkins (1 987), if taken as accurate reports of what is actually happening, do indeed show a consider-

I able crudity compared to even our present medical procedures, which include consideration for the patient's comfort and well-being. However, because a wide spectrum of behavior is reported, it is logical to imagine that the abductions performed in the most skillful manner would leave little memory or ill will. Thus, such abductions, if they are done very well, would never be reported and could even be the majority of them.

Argument Four: History

"The extension of the phenomenon throughout recorded human history dem- onstrates that UFOs are not a contemporary phenomenon. "

It is startling for a believer in the ETH to grasp that, in the 9th century, some UFOs reported were vessels in the sky; in 1897, they were dirigibles; in 1946, ghost rockets; and today, the UFOs are high-tech in physics and advanced vehicle performance. One likely explanation for this observation is that there is a consistent plan to covertly mask the visits within the social structure of the civilization being visited. Not meddling with developing civilizations such as ours is a reasonable policy. "Knowledgeable researchers like Vallee, earlier, in emphasizing folklore comparisons with UFO close encounters, left the impression on some people at that time that he thought the whole thing is an illusion (Vallee, 1969). Vallee is now emphasizing the physical evidence and signatures as not imaginary at all" (Vallee, Confronta- tions, 1990). The fact that most cultures on Earth have an ancient tradition of little people who fly through the sky and abduct humans could also be totally consistent with the idea that that is precisely what has been happening. What could be more simple?

Argument Five: Physical Considerations

"The apparent ability of UFOs to manipulute space and time suggest radi- cally diferent and richer alternatives. . . "

While it is true that the mastery of time and space is reflected in reports such as light beams that bend or stop in mid-air, ability to merge with or pass through solid objects, ability to shrink, grow, change shape, multiply, or accelerate enormously, all these reports need to be explained. The simple idea that "any sufficiently advanced technology is indistinguishable from magic," as advanced by Arthur C. Clarke, provides a useful frame of mind. More specifically, if the physics of zero-point vacuum energy is correct, the density of empty space contains the energy equivalent of 1 094 g/cm3 (Misner, Thorne, & Wheeler, 1973; Puthoff, 1990, in press). This would mean that

108 R. M. Wood

what we treat as solid matter is only a tiny fraction of the basic energy density in a vacuum. Thus, the idea of objects merging with each other or passing through solid walls (e.g., Fowler, 1979) is not remarkable, because our "solid" objects may only be a tiny ripples of coherence in an enormous background of much denser energy.

Vallee concludes by imaginatively offering three newer hypotheses, namely earth-light, control system, and wormhole travel. The latter can actually be viewed as a special case of the ETH, providing for time travel as well as space travel. It is certainly true that the time-delay or time-shift reports from the abductions do point toward a requirement to provide for some form of time control. However, with a simple forward-running time, simple three-dimensional geometry, and Einstein's Special Theory effects valid in most circumstances, and with vehicles traveling faster than the speed of light in a vacuum some of the time, an advanced civilization could presumably interweave time and space control as they chose. This would be intellectually similar to Vallee's "wormhole hypothesis."

Expected Visitation Rate

Figure 1 is a simple-minded plot of distance and time on log scales. One line shows the speed of light, with v = c labeled at the lower right side. The solid lines, sloping at about 45 degrees, plot the distance a spacecraft would travel if accelerating at a constant number of g's halfway to its target and decelerating for the other half. Relativistic effects are not included, because there is no evidence that they need to apply to gravity-driven systems any more than the drag rise in aircraft as one approaches mach 1 applies when flying at mach 10. The reason for considering travel at constant acceleration instead of constant speed for assumed gravity control systems is due to the equivalence of acceleration and gravity, and the conjecture is that if there is a limit to gravity control effectiveness, it would express itself as a limit to acceleration value.

The distances of some astronomical reference points, such as the nearest star, the edge of the Milky Way, the galaxy in Andromeda, and the edge of the universe are shown (Kaufmann, 1985). In addition, because Zeta Reti- culi 1 and 2, reported by Dickinson (1974), and The Pleiades, reported by Kinder (1 987), have been identified as possible civilization origins, they are also shown. Thus, if you want to travel from here to Andromeda in this manner, it would take you 1,000 days (approximately 3 X 365 from the chart, interpolating on the log scale) traveling at lo6 g7s and, if you had done it at constant speed, you would have traveled at roughly lo6 times the vacuum speed of light. (Actually, with the constant acceleration-deceleration assumption, acceleration in Earth g's and time in years, the maximum speed is always gT/4 times the speed of light.)

Andromeda Galax

Milky Way dlameter~, .00E+05

Milky Way disk edge

c.

Oort Cloud 1.0.. . . . 1 0 0 . .

voyager In 2190---r1 .00E-02 --

v=c 1 -00E-04 ....

Sun

Time (years)

Fig. 1 . Distance and travel times at constant acceleration, deceleration.

acceleration

* 100 g's.

-x- 10,000 g's

x- 10A6 g's

* 10A8 g's

r>- 1OA10 g's

110 R. M. Wood

Now let us imagine that in our galaxy we have had one visit every 1,000 years from civilizations as far away as the close edge of the Milky Way, 1,000 light years (ly) away. In Figure 1, this is a point at the right side of the line, sloping downward to the right. Now let us assume that the extraterrestrial civilization density is uniform and that, if one examines a sphere with ten times the radius and each civilization knew that we were here to visit, the time between visits would be 1 O3 times more frequent, or once a year. This is shown as a point at (1 yr., lo4 ly) in Figure 1. By interpolating the acceleration lines, it can also be seen that it takes about 1 year at an acceleration and deceleration of 40,000 g's to get here from that distance. This kind of thinking helps establish the possible technical requirements for gravity-control propulsion systems, possibly using vacuum zero-point fluctuation control or control of the space-time metric itself.

Conclusion

Whether any hypothesis is emotionally exciting or dull should not be a basis for selecting the least unlikely hypothesis. The selection or rejection of the ETH is no exception. This author concluded long ago, along with Stan Friedman and James McDonald, that the ETH was the least unlikely hypothesis. McDonald (1968) in his testimony to Congress stated, "I now regard Hypothesis 7 ("extraterrestrial devices of some surveillance nature") as the one most likely to prove correct" (p. 36). Twenty-two more years of evidence would seem to merely strengthen this conclusion.

The major assumptions, to be consistent with the ETH, are (1) it is very simple to travel at many times the speed of light, resulting in very short trip times compared to species lifetime; (2) most travelers have knowledge of the locations of other civilizations; (3) "they" have a common policy of noninterference except for a few mavericks; and (4) previous genetic interactions with homo sapiens during our known history may be responsible for similar biology in some extraterrestrials.

As mind-boggling as it may seem, the extraterrestrial hypothesis is not that bad.

References

Boyer, T. (1985). The classical vacuum. Scientific American, August, p. 70. Dickinson, T. (1 974). The Zeta Reticuli Incident. Astronomy 2, 4. Fowler, R. ( 1979). The Andreasson Aflair (p. 4) Englewood Cliffs, NJ: Prentice-Hall. Friedman, S. (1979). UFOs under wraps. New Realities (p. 48). Gordon, T. (1966). Ideas in conflict. (p. 98) New York: St. Martin's Press. Hopkins, B. (1987). Intruders. New York: Random House. Hynek, J. A. (1966). UFOs merit scientific study. Science, 154. October, p. 329. Kaufmann, W. (1985). Universe. San Francisco: W . H. Freeman and Company. Kinder, G. (1987). Light years. New York: The Atlantic Monthly Press.

Extraterrestrial hypothesis 1 1 1

Kuiper, T., & Morris, M. ( 1977). Searching for extraterrestrial civilizations. Science 196, 6 16. Markowitz, W. ( 1967). The physics and metaphysics of unidentified flying objects. Science, 157,

1274. McDonald, J. ( 1968). Symposium on UnidentlJIed Flying Objects: Hearings Before the Commit-

tee on Sc,ience and Astronazitics. U.S. Government Printing Office, p. 36. Misner, C. W., Thorne, K. S., & Wheeler, J. A. (1973). Gravitation (pp. 840-841 2nd 1202). San

Francisco: W. H. Freeman and Company. Puthoff, H. E. (1989). Gravity as a zero-point fluctuation force. Physics Review A39, 2333. Puthoff, H. E. ( 1990). Condensed Charge Technology and its Potential impact on the electronic

industry. Presented at the June 1990 Meeting of the Society for Scientific Exploration, Palo Alto, CA.

Puthoff, H. E. (in press). The Energetic Vacuum: Implications for Energy Research. Specula- tions in Sci~jnce and Technology 13.

Shklovskii, I . , & Sagan, C. (1966). Intelligent Ilfe in the universe. Oakland, CA: Holden-Day, 409.

Vallee, J. ( 1969). Passport to Magonla. Chicago: Henry Regnery. Vallee, J. (1990). Confrontations. New York: Ballantine. Vallee, J. (1990). Five Arguments against the extraterrestrial origin of unidentified flying ob-

jects. Journal ScientiJic Exploration 4, 105.

Journal of Scientific Exploration, Vol. 5 , No. 1, pp. 1 13- 120, 199 1 Pergamon Press plc. Printed in the USA.

0892-33 10/91 $3.00+.00 01991 Society for Scientific Exploration

Toward a Second-Degree Extraterrestrial Theory of UFOs: A Response to Dr. Wood and Prof. Bozhich

JACQUES VALLEE

1550 California Street, #6L, San Francisco, CA 94109

Abstract-Counter-arguments to the views expressed by this author in a previous article questioning the extraterrestrial nature of unidentified flying objects have been presented by Dr. Robert Wood, an American aerospace expert, and by Prof. Serge Bozhich, a Soviet mathematician. These counter- arguments propose alternative explanations for five major inconsistencies we had noted in the ETH theory. In this response, it is shown that such explanations represent a significant alteration of the basic tenets of the ETH, and that they do introduce useful new insights for future research on UFOs. However, some of the contradictions remain, especially when the matter of "abductions" is brought under close scrutiny.

Introduction

The most common interpretation of UFOs is the extraterrestrial theory or "ETH." Under this hypothesis, such objects are physical devices controlled by intelligent beings from another planet who have been visiting the earth as part of a scientific survey begun at the time of World War Two. In a previous article, the author offered five arguments drawn from the current UFO data that contradicted this theory, suggesting that the ETH had become obsolete, that it should be updated, and that other hypotheses should be considered (Vallee, 1990).

In his thoughtful article responding to these arguments, Dr. Wood has contributed an important clarification to a long-overdue debate which ad- dresses the core issues of UFO research (Wood, 199 1).

In the past, efforts by various researchers to broaden the spectrum of hypotheses have met with strong protests and occasionally strident opposi- tion from most US ufologists. Thus, it is refreshing to read an article that places the discussion firmly on scientific rather than emotional grounds, while restating and clarifying the basic tenets of the extraterrestrial theory.

In a related effort, a Soviet mathematician, Prof. Serge Bozhich, has articulated his own counter-arguments from the perspective of Soviet UFO data. Our purpose here is to examine both of these attempts to broaden the ETH while preserving its basic assumptions.

Wood's Counter-Arguments

It is appropriate to consider Dr. Wood's points one by one, in the manner he has used to discuss my own article, to which the reader is referred. He

114 J. Vallee

approaches the subject equipped with a set of four assumptions, namely (1) the removal of the speed of light as an interplanetary speed limit, (2) alien knowledge of the location of civilizations such as ours, (3) an alien policy of noninterference, and (4) historical genetic interaction with homo sapiens. Since Wood's counter-arguments are based on theoretical considerations rather than on UFO observations and report statistics, it is difficult to refute them, but it is useful to examine their relevance and their possible limitations.

Argument One: Close Encounter Frequency

There is good agreement between Wood and myself on the problem statement to be tackled here, since he does not challenge the estimated number of close encounters, perhaps as large as several million in 40 years. Our differences only have to do with the probability of the proposed explanations. Wood finds it plausible to assume that several thousand different extraterrestrial civilizations have each authorized up to 1,000 visits by their members over that time period. This is already in sharp contrast with the first-degree ETH, which generally assumes that a single race or, possibly, only a few races of aliens, began visiting the earth after World War Two, with the Kenneth Arnold sighting of June 1947 as the often quoted starting point. In other words, the high frequency of close encounters does demand that the commonly held view of visiting aliens as interplanetary scientists be abandoned.

Argument Two: Physiology

The humanoid body structure of the alleged "aliens" can be explained, according to Wood, if the human shape turns out to be a cosmic optimum, or if evolution has not proceeded in independent fashion on other planetary bodies, at least in the solar neighborhood.

While the panspermia theory is supported (or at least, not negated) by recent discoveries about the presence of complex molecules in space and the frequency of meteoritic and cometary impact on our planet, Wood may be underestimating the difficulties of survival for a humanoid in a foreign planetary environment. Many of the points I brought up in this regard are not addressed in his response. In particular, one does have to explain why aliens, even if they do have a humanoid form, would have eyes adapted to the particular set of light frequencies we happen to recognize, why they could breathe our air (especially if the hypothesis of genetic manipulation is set aside, as Wood proposes to do), and why they should happen to be adapted to our gravity and especially to our culture.

One difficulty in this discussion may come from the fact that most of the experts consulted on this point by ufologists have been medical doctors rather than zoologists. Understandably, medical specialists will generally be tempted to take the human body as their standard when discussing evolution

Response to Wood and Bozhich 115

and intelligence. In the last few months, the author had an opportunity to discuss the probability of humanoid shapes among interplanetary beings with French biologist Remy Chauvin, who brought up some neglected points. In particular, he noted that the common assumption that higher intelligence was necessarily vested in bipeds who had opposable thumbs (hence, capable of developing a complex brain in connection with the building and handling of tools), could be seriously challenged.

In a recent publication, Prof. Chauvin has offered two specific counter-examples in this regard. The first one concerns birds such as the orthotomus of Ceylon, which build sophisticated nests by sewing large leaves together using long vegetal fibers. If the female does not find suitable fibers she manufac- tures them by twisting spiderwebs. A precise beak, he proposed, might be as desirable an instrument in some environments as an opposable thumb. In fact, whenever humans need to perform a delicate task in electronics, we use special pliers shaped like bird beaks rather than our fingers, or any other tool (Chauvin, 1990).

The second example concerns the octopus, which is quite capable of grab- bing, examining, and removing the tops of glass jars containing food using his highly sensitive, powerful tentacles in connection with its eyes and brain. As appendages go, tentacles are equally as versatile and better endowed in sensory and prehensile abilities than hands. Such examples abound in zoology, and they should make us pause when we are tempted to assume that the human shape is always the most desirable one for a higher being. Right here on earth, nature had many options, it seems, for the development of a complex brain.

By the same token, if nature has evolved a cosmic form able to survive and adapt in a variety of planetary environments, it is rash to assume that such a form would necessarily follow a humanoid model. As for the human body being the end product of optimization, this assumption is contradicted every day by the observations of modern medicine. The emerging science of gene therapy is predicated on the notion that specific human diseases can best be overcome by altering the genetic makeup of susceptible individuals, and theory now opens up the exciting but frightening temptation to modify the DNA code itself, at least in lower-level species such as plants or insects. We should assume that at least some proportion of Wood's thousands of visiting civilizations would already have gone beyond this level, and would have evolved into much more comfortable, durable, and competent shapes than the humanoid form allows.

Argument Three: Abductions

Here again, there is agreement on the problem statement. Wood and I concur to observe that the reported abductions exhibit a level of "crudity" that would be unnecessary to a scientifically sophisticated civilization of

116 J. Vallee

Wood leans once again toward an explanation in terms of the wide variety he expects to find among aliens, including some with high technology and low ethics. But in this respect he contradicts his own third basic assumption, namely the alien rule of noninterference in lower-level cultures such as ours.

We cannot have it both ways. If the aliens do have such a law, we should never have been able to detect them in the first place, just as the aborigines of Australia are unable to detect the hundreds of surveillance satellites that pass over their location every day, except perhaps as mere points of light. On the other hand, if the aliens do not have such a law, then Wood cannot go on to argue that the variety of UFO shapes throughout history are used to make their craft unobtrusive. The very fact that we have such a large number of published records of close encounters flies in the face of the noninterference assumption. Thousands of books, hundreds of magazines, and many movies have been inspired by such sightings and they have made a deep impact on human cultures. If there is a noninterference law in force in the solar neighborhood it is applied very poorly indeed.

This contradiction in Wood's reasoning leaves the behavior of aliens in alleged abductions unexplained, and I believe my argument still stands. Fur- thermore, a wide variety of alien forms and behaviors is not what the abductees seem to be reporting, according to those who have made a specialty of interviewing such witnesses under hypnosis. If we are to believe the results of this admittedly dubious practice, the typical abductor is a short, bald being with elongated eyes, and the process of abduction always goes through the same standardized phases, in clear contradiction to Wood's expectation of a variety of abducting and contacting civilizations.

Argument Four: History

Here, Dr. Wood may be the victim of a frequent misrepresentation of my views when he speaks of Passport to Magonia as expressing the notion that "the whole thing is an illusion." This was, perhaps, the way my book was misinterpreted by those American ufologists for whom any theory that de- viated from the ETH was automatically heretical. I believe that if Wood read the book again from the perspective of the two decades that have elapsed since its publication he would see that I never argued that the phenomena- including their most ancient forms-were illusionary at all (Vallee, 1969).

Magonia was written in reaction to the fact that in the United States "the material that pertains to the landings of the craft . . . (had) been generally ignored." Using this material, and specifically the "occupant" cases, I sought to demonstrate that modern-day UFO phenomena were merely an extension of a centuries-old historical tradition that blended myth and physical reality. It is not my fault if hardcore ufologists chose to hear only half of that message. The book agrees fully with Wood's conjecture that "the fact that


through the sky and abduct humans" could be "consistent with the idea that that is precisely what has been happening." Wood and I have no difference on this point.

If he follows this line of speculation much further, however, Wood will find that acceptance of such ancient traditions is anathema in the context of the ETH as it is generally understood. Far from being hailed as a champion of the extraterrestrial theory, he may find himself at odds, as I have, with many contemporary American ufologists, who have repeatedly taken violent issue with the parallels that have been drawn between modern-time ufonauts and the elves, angels, and flying demons of old. Indeed, Hopkins has sum- marily denounced such references as "folklore of obviously uncertain au- thenticity" (Hopkins, 1990). Other researchers committed to the ETH, such as David Jacobs, have asserted that abductions were a brand new piece in the UFO puzzle. Jacobs has come to the firm conclusion that "what we have here is a systematic program of exploitation of human beings for sperm and eggs," and that it began relatively recently. He added that "no alternative theory accounts for all the data." In particular, he stressed that invoking folklore and legend would demand that "every single case is going to be qualitatively different" (Jacobs, 1990, p. 9). Other researchers, such as Me- heust in France and Evans in England, have argued the opposite point, namely that it is precisely the qualitative consistency in abduction stories that should make one suspect a strong link to the constant themes of popular imagination and folklore.

That particular controversy aside, what Jacobs and others have articulated is a strongly held belief among ETH proponents today, and it excludes any consideration of "ancient traditions of little people who fly through the sky and abduct humans," to quote Wood's words. Such a position is understandable: If abductions, as we know them today, have been experienced by people not for a few years but for centuries, does it really make sense to assume that the "aliens" are extraterrestrial visitors conducting a scientific survey or taking biological samples to improve their race's survival chances, to overcome degenerative disorders, or to produce hybrids, as the abductionists hold? Wood has not resolved this contradiction.

Argument Five: Physics

The often-quoted remark by Arthur Clarke that "any sufficiently advanced technology is indistinguishable from magic" is invoked here by Wood to account for the extraordinary observations of UFOs that disappear on the spot or materialize inside closed spaces such as witnesses' bedrooms.

It is perhaps in this regard that Wood and I have our greatest differences. It is unfair to argue one minute that observed UFO phenomena are nicely behaved and consistent with modern physics, perhaps assuming a few exten- sions such as the removal of the light speed limit and the use of a zero-point

118 J. Vallee

drive, and to throw all constraints to the wind the next minute by stating that any advanced technology would look magical anyway, hence anything goes. Ifanyth ing goes, then other theories, such as those invoking other dimensions beyond spacetime, the theory of a psychic control system, or the Wormhole travel hypothesis will work just as well or better than the first-degree ETH Wood is defending, as I pointed out in the original article. In fact, they may be more parsimonious, since they also get around some of the other arguments we have reviewed.

Under such an alternate theory UFOs could be physical devices acting upon our spacetime environment without necessarily originating on a different planetary body, and their occupants could be closely related to the human race.

It seems to me that Wood should remain in line with his stated objective, namely demonstrating that "if the speed of light can either be raised or exceeded . . . , then the pattern of UFO reports is consistent with the ETH," and magic should be left aside while we pursue this exercise to its logical, scientific consequences.

The witnesses to whom I speak in the field tell me of a variety of remarkable effects that are not easily explained simply by assuming that UFOs can travel at arbitrarily high speeds through conventional spacetime. For instance, they describe individual beings or entities that appear on the spot, "flat" beings that lack one dimension, and craft that either merge slowly with identical objects or even sink inside the solid earth. While some of the objects are luminous, as a zero-point vacuum energy model would appear to suggest, many others are perfectly dull or even dark. It is the totality of these reports that we should attempt to explain, rather than a selected subset that matches a particular theory.

Bozhich's Counter-Arguments

Another useful line of discussion has recently been added to this debate from the perspective of Soviet UFO data. In a letter dated 9 November 1990, following up on a round-table discussion we both attended in Moscow in January 1990, Soviet mathematician Serge Bozhich has sent me his own set of counter-arguments against the first four of my "Five Arguments" (Boz- hich, 1990). Bozhich's views can be summarized as follows:

1. On the very large number of close encounters: the greatest frequency of visits to the earth may well be made by tourists rather than scientists. In this regard, he matches Wood's opinion that we should relax the assumptions regarding space-traveling aliens.

2. On the absurdity of the humanoid appearance of alleged aliens: Simple speculation on the physiology and possible general appearance of ratio- nal beings from other planets cannot approximate the truth. By this, I


assume Prof. Bozhich means that we should refrain from speculating in the absence of stronger data. I would then answer that, by the same token, we must refrain from assuming that such beings are necessarily extraterrestrials.

3. On the reported behavior of ufonauts during abductions: Abduction stories should not be taken into account, according to Bozhich, because they probably belong to the fantasy world of the witnesses. In support of this, he points out that contactees tend to report that ufonauts are interested in the same topics as the contactees themselves, namely, the fear of nuclear war, ecology, and the price of goods. Bozhich notes further that the descriptions of aliens made by contactees (or abductees) cover the entire range of entities found in human dreams, fantasies, religious ap- paritions, folklore, and in our science fiction literature. Thus, "all these tales are psychological phenomena characteristic of the contactees themselves and not of metal craft generating traces and visible to everyone." Like children, he writes, abductees are not liars because they believe their own stories and "cannot distinguish reality from fantasy." In this respect, he takes the opposite view from that of Wood and of most American abduction researchers. I find myself somewhere in the middle in this argument, since I regard the abduction as a genuine experience which is distorted beyond recognition by the unfortunate lack of standards that characterizes such research.

4. On the historical extension of the alleged craft, Bozhich claims that appearance and kinematics of modern flying saucers are different from those of UFOs seen in the sky before the 20th century. It may be possible to prove that flying saucers have an extraterrestrial origin, but this cannot be proven for ancient UFOs separated from modern ones, he says. Thus, Bozhich sees no factual reason to unite them into a single phenomenon.

This last statement is interesting and provocative, but it is not currently supported by specific references or by clear statistics arguing against a con- tinuum of kinematic and physical properties across time. However, it does suggest an avenue of research that could and should be checked using the existing database.

Discussion

The original article that triggered this discussion was not meant to eliminate the ETH from the list of explanations for UFOs. Indeed, it observed that "until the nature and origin of UFO phenomena can be firmly established it will naturally be possible to hypothesize that extraterrestrial factors, including undiscovered forms of consciousness, are playing a role in its manifestations." But, it sought to clarify the difficulties with such a theory and to advance other, equally attractive hypotheses. It also urged that the idea of

120 J. Vallee

extraterrestrial intervention be "updated to include current theoretical speculation about other models of the physical universe."

In his well-articulated response Dr. Wood has proposed just such an up- date, in terms that are both scientifically sound and intellectually appealing. In the process of saving it, however, he has been forced to stretch the "first- level ETH" to such an extent that it may no longer be recognizable by most ufologists. While it would be a pleasure, indeed, to welcome him into the ranks of the heretics, I doubt that such a distortion was part of Wood's plans when he wrote his rebuttal. But, what are we to make of a model involving 14,000 different civilizations, all of them humanoid in shape, which travel throughout the universe in ships that can pass through solid objects, yet feel it necessary to stop in the remote countryside to terrify human victims, performing crude operations upon their bodies? While this revised ETH model does account for some of the observed facts which the first-level ETH had overlooked, it does not really overcome the five arguments that contradict the extraterrestrial theory.

It is my hope that the debate we have initiated here will continue, and that other scientists will join us with their own contributions. In the process of exploring these issues we may well shed new light on critical factors of the phenomenon that have not been previously recognized.

References

Bozhich, S. (1 990). Personal communication: 9 November 1990. Chauvin, R. ( 1990). La biologie de /'Esprit. Monaco: Rocher. Hopkins, B. (1989). Letter to the Editor. Mu;fi,n Journal 250. Jacobs, D. (1990). Interview with Floyd Murray. Caveat Emptor 23, 9. Vallee, J. (1 969). Passport to Magonia. Chicago: Henry Regnery. Vallee, J. ( 1990). Five arguments against the extraterrestrial origin of unidentified flying objects.

Journal of Scient13c Exploration 4, 105. Wood, R. (1 99 1). The extraterrestrial hypothesis is not that bad. Journal ~J'Scientijic Explora-

tion, 5 .

Journal of Scientific Exploration. Vol. 5 , No. I , pp. 12 1 - 144, 199 1 i Pergamon Press plc. Printed in the USA.

0892-3310191 $3.00+.00 0 199 1 Society for Scientific Exploration

1 Low-Frequency Seismogenic Electromagnetic Emissions as I Precursors to Earthquakes and Volcanic Eruptions in Japan

TAKEO YOSHINO

University of Electro-Communications, 1-5-I Chofugaoka, Chofu-shi, Tokyo 182, Japan

Abstract-A multipoint network was constructed in the Tokyo area for earthquake prediction using seismogenic electromagnetic emissions. The network consists of eight observation points within 50 km of each other. Each point has a digital direction-finding detector with two loop sensors tuned to 82 kHz. The output signals of the receivers are added into a digital vector composition circuit to obtain the direction angle of the source point, and this signal is telemetered to the central computer.

To protect from false alarms caused by local man-made noise interference, the warning is announced only when there is a high cross-correlation between almost all detectors pointing to one small area. The mechanism of these earthquake precursors can be explained as electromagnetic emissions from the rocks around the focus when they are crushed completely by the distortion pressure. These emissions propagate along the fault plane as an EM surface wave mode and radiate from the slit antenna formed by the intersection of the fault plane and ground surface.

In the last five years, we have detected impulsive noise bursts of seismogenic emissions at 82 kHz, 1.525 kHz, and 36 Hz using our multipoint detection network around the Tokyo region and Izu peninsula. This system has recorded EM signals prior to the following events: volcanic eruptions on November 15 and 2 1, 1986 at Mt. Mihara on Ohshima Island, and on July 12, 1989 in Itoh Bay in the Izu peninsula region, and also a minor earthquake on October 14, 1989 at Ohshima Island.

I Introduction

The Japanese-Soviet cooperative project for the study of electromagnetic emission phenomena related to earthquakes was started in 1980. The first emissions were observed at 16:33 JST (UT + 9 hours) on March 3 1, 1980, at Sugadaira Space Radio Observatory, University of Electro-Communica- tions, Sugadaira, Nagano prefecture in Japan. The magnitude of this earthquake was about 7, and the depth of focus was approximately 380 km. The epicenter was located in the Kyoto prefecture and the distance between the Sugadaira observatory and the epicenter was approximately 250 km. The noise level recorder for 8 1 kHz registered an anomalously high change in the background noise level, more than 15 dB over the usual level, beginning 50 minutes before the main shock. The noise dropped sharply back to the previous level exactly at the moment of the earthquake as shown in Figure 1. The VLF whistler recorder at Sugadaira Observatory also showed unusual

12 1

122 T. Yoshino

81.0 KHz SUGADAIRA MARCH, 31,1980

C r MAIN SHOCK 1- *O

-1 0 P

15:OO 15:30 16:OO 17:OO 17:30 LT 16:33 JST M = 7 DEPTH = 480 Krn

Fig. 1. Electromagnetic radiation level at a frequency of 8 1 kHz at 16:33 JST (07:33 UT) on March 3 1, 1980 observed at Sugadaira Space Radio Observatory. (After Gokhberg et al. 1986b.)

impulsive emissions at frequencies below 1.5 kHz before the earthquake as shown in Figure 2. Similar 81-82 kHz emissions were observed prior to other earthquakes of magnitudes 5.5 to 6.5 on September 25, 1980, as shown in Figure 3. These earthquakes were located in the Tokyo suburbs (Gokh- berg et al., 1982).

Since 198 1, the author and colleagues have observed several emission events in the 8 1-82 kHz range, just prior to earthquakes. Based on these measurements, we set up a new multipoint observation network with direction finding capabilities around the Tokyo area. The purpose of this network was to investigate the possibility of immediate warning just prior to earthquakes, and elimination of man-made noise interference to improve the accuracy of any potential prediction. One of our most promising results was in the case of an "under foot" earthquake which occurred in southwestern

MARCH 31, 1980 - 40 40% 2

m m 1 5 ~ 5 0 ~ 5 0 ~ (USUAL NOISE L E V E L ) E ' 30 30 >- k v, z 20 20 W I- 2 - 2 10 10 Q z c3 -

0 0 0 1 2 3 4 5 6 7 8

FREQUENCY (KHZ)

Fig. 2. Spectral amplitude of the sonagram of wide band VLF whistler detector in Sugadaira observatory at 16:20:04.7 JST (about 13 minutes before the earthquake) for the highest noise level of the wave packets and also at 15:50:50 JST (17 minutes after the earthquake) as an example of the usual noise level. (After Gokhberg et al. 1986b.)

~ B P 81.0 KHz SUG INAMI SEPTEMBER, 2 5 , 1 9 8 0 ~ B P

25 25 MAIN SHOCKl P N D SHOCK

20 20

10 0- 1 I

01130 02:54 02:59:48 J S T

35'31 N 35'33N 1 40'13E 140'10E

DEWH 8 0 K ~ DEPTH 7 0 K ~

Fig. 3. Electromagnetic radiation level at 8 1 kHz, 02:54 JST, September 25, 1980, observed at Suginami in the western suburbs of Tokyo. (After Gokhberg et al. 1986b.)

124 T. Yoshino

Ibaragi prefecture at 2 1 : 14 JST on February 27, 1982. The magnitude of this earthquake was 6.3 and the depth of focus was approximately 40 km, as shown in Figure 4a and b. This retrospective prediction of the epicenter was the result of direction finding from the following three observation points: Suginami in Tokyo, Sugito in Saitama prefecture, and Yatsugatake in Na- gano prefecture. The earthquake subsequently occurred in the predicted area, as shown in Figure 4c (Yoshino et al., 1985).

Fig. 4. (a) Locations of observation points and epicenter of the earthquake at 2 1: 14 JST on February 27, 1983. (b) Levels of emissions at Sugito, Suginami, and Yatsugatake observation points before the earthquake. The traces from top to bottom illustrate the varia- tion in signal level for the north-south direction at Sugito, and the east-west and north- south directions at Suginami and Yatsugatake. East-West data from Sugito are missing owing to preamplifier failure. A great number of anomalous emissions with short duration appeared between 12:30 and 2 1:30 JST at Sugito and Suginami. Anomalies at Yatsugatake appeared after local sunset (1 9: 10). These emissions were recorded independently at each point, but a number of occurrence times show good coincidence with each other. (c) Intersection of the bearings of the anomalous electromagnetic emissions. These bearings are calculated from the data of Suginami at 19: 14 and 20:09 JST, and of Yatsu- gatake at 20:OO and 20:09 JST. The ruled area shows the epicenter area predicted by the emission data. "x" indicates the actual epicenter which falls within the predicted area. (After Yoshiro et al. 1985.)

Seismogenic emissions 125

FEB. 27. 1983 21 14 JST

SUGlTO ( N - S)

20' SUGINAMI (E -10- m c o d a Jl-L.Ll'

-20'- 1 I I I I I I I 10- YATSUGATAKE (E - W ) 3'

- I I I I I I I I I

I

13 14 15 16 17 18 19 20 21 i2 2: 24 w JST > W

W !!? SUGITO ( E - W ) 0 - NO DATA

* - -

JST

? 2p 4 f 60 80 1 DISTANCE

YATSUPfiTAKE a - -dg=-- - - - - - - - - - - - - - - - - e - - - - - SUGITo

2 0 . 0 0 ~ ~ ~ - - - - - _ - -

Fig. 4. (continued)

/ / 21114 JST C ~ p i c e n t e j ,

- - /, I I 1W1-r r ~ - - 3 6'N - - _ _

Area

126 T. Yoshino

Since 1984, the authors have set up a multipoint direction finding network with eight observation points around the Tokyo region. The locations of observation points are shown in Figure 5. Now, we are building a data telemetry system for realtime cross-correlations (Yoshino, 1986a, 1986b).

A major volcanic eruption occurred on Mt. Mihara in November 1986, and the 82 kHz direction-finding network operating on Ohshima Island recorded several impulsive noise emissions prior to eruptions. Since 1987, we have added two frequency bands [VLF (1.525 kHz) and ELF (36 Hz)] to our multipoint network to improve S/N ratios, owing to increasing 82-kHz background noise levels in the last decade.

In July 1989, the area around the city of Izu Itoh was struck by a series of locally very strong volcanic earthquakes. Izu Itoh is one of the major cities located on the east coast of the Izu peninsula, and the area is famous for its beautiful volcanic mountains, lakes, seaside landscape, and many hot springs. The city extends along the coast of Ito bay, and is located about 35 km northeast of Izu-Ohshima island. These volcanic earthquakes ended with the eruption from a new volcanic crater on the bottom of Ito bay on

138"E 140°E Fig. 5. Location of observation points around the Kanto area as of January 1988.


July 12, 1989. The author obtained several interesting correlations between LF, VLF, and ELF emissions related to these earthquakes.

Radiation Mechanisms

The study of the radiation mechanism of precursory electromagnetic emissions in Japan and the USSR has been ongoing since 198 1, but these studies have not obtained any clear and reasonable result until today. Gokh- berg and his colleagues in the USSR had tried two different approaches. In the first, they assumed that the source was located in the lower region of the ionosphere and precipitated the plasma instabilities by large gradients of the electric field at the ground surface, dE/dh, and the geomagnetic field intensity at the epicenter region (Gokhberg et al., 1984). In the second, they changed this explanation to electromagnetic emission produced by micro- dislocations in rocks before shallow earthquakes (Gokhberg et al., 1987). The observation systems and the research approach of the Soviet groups shifted toward lower frequency regions, on the order of a few hertz, because these frequencies were observed in the electric field variations in the lower and upper ionosphere by satellite observations since 1985 (Chmyrev et al., 1987; Larkina et al., 1987; Migulin et al., 1987). We have great trust in these estimations, but the results are not sufficient to explain the source mechanism of the emissions, the mechanism of the electromagnetic energy transmission in the soils and rocks from the earthquake focus to the earth surface, or the mechanism of electromagnetic wave radiation at the surface of the ground.

Laboratory experiments performed by our colleagues show that the rocks emit electromagnetic radiation when crushed (Mizutani & Yamada, 1987). Similar emissions were observed by Cress et al. (1987) in laboratory experiments and by the author in the eruption of Mt. Mihara on November 2 1, 1986 (Yoshino & Tomizawa, 1989). During the entire duration of the eruption, the emissions were only observed when the dike of magma intruding into the mountain body caused the crushing of rocks.

To explain the mechanism of transmission of electromagnetic impulses from the source around the focus area to the surface of the ground, we have applied a surface mode transmission model. The conductivity along the fault plane is usually very high compared to values in undisturbed rock, and the conductivity gradient is distributed along the direction of fault lines. Such characteristics of a fault surface promote surface mode propagation along the fault plane from the focus to the ground surface. We estimate that the optimum condition for wave propagation is 25 dB/ 10 km below the usual case for homogeneous soils and rocks at the same distance. To explain the radiation condition at the ground surface, we used the optimum impedance match between the surface and free space and calculated a VSWR of 1.5 using the optimum case when the fault terminates in a slot dipole antenna a t . the ground surface. Based on these results, the author will attempt to explain

128 T. Yoshino

the source mechanisms for the electromagnetic precursor emission phenomena in this article.

Observation Frequencies

Based on very careful searching through the frequency spectrum for several months, the author selected 8 1-82 kHz for observation frequencies for seismogenic emissions during the first measurements in 1980, because this range was the only one sufficiently protected from man-made noise sources. This is because so many radio transmitters and other man-made noise sources are widely spread across the frequency spectrum from ELF to EHF on the Japanese islands. Unfortunately, the natural background noise level during the night time at these frequencies of 8 1-82 kHz is usually 6-10 dB higher than in the daytime, owing to noise signals generated by lightning discharges from tropical thunderstorms. The threshold level of background man-made noise in this frequency range has also been increasing in the last decade. Therefore, the author tried to search for new observation frequency ranges with lower background noise interference. The result was the selection of 1.525 kHz and 36 Hz. The ionospheric propagation of VLF waves is usually a guided mode, and this propagation has a lower cutoff frequency with respect to the dominant frequency. The usual cutoff of dominant modes of the night time ionosphere is approximately 1.8 kHz. Thus, 1.525- kHz waves will not be able to propagate the noise of lightning discharges from the distant tropical regions to middle latitudes. On the other hand, the frequency range from approximately 7-45 Hz is globally very noisy due to the Schumann resonance phenomena. However, the noise spectrum in the Schumann resonance band has characteristics similar to a Gaussian distribution, so the background noise level at 36 Hz will be reduced to a nearly neglegible level. The frequency range below 5 Hz is disturbed by the strong continuous pulsations (PC- 1) and (PI- 1) emission, and it will be very difficult to clearly select natural noise emission or geomagnetic field pulsation phenomena when observations are made in this frequency range.

Equipment for Detection of Seismogenic Emission

Figure 6 is a block diagram of the new standard detection unit with three frequencies for our multipoint network system, showing the flow of the data processing for the prediction of epicenter bearings by means of electromagnetic precursor emissions at each observation point. As shown in this figure, a sensor for 82 kHz consists of two tuned loop antennas set up perpendicular to one another, north-south and east-west, for direction-finding purposes. Each antenna consists of an 85-cm diameter, 50-turn coil with electrostatic shielding, and tuned to 82 kHz. The two sensors at 1.525 kHz and 36 Hz, added to better exclude man-made noise, consist of 5,000-turn coils with single Permalloy cores 1-cm square by 80-cm long. The output signals from


Fig. 6. Block diagram of the data analysis system for electromagnetic earthquake precursor observations.

/ass--

these coils are transmitted through digital telemetry links to the computer of the earthquake prediction and warning center by telephone cables, micro- wave links, and optical fiber cables. Data are also stored on digital tape at each observation point. Our test network around the Tokyo area now consists of eight points separated by approximately 50 km.

Protection From Man-Made Noise Interference

36 Hz Amp

The data observed by all observation points are telemetered to the local earthquake prediction and warning center, and are processed to obtain the bearing and location of the noise emission source. To prevent false alarms, we intend to issue an earthquake prediction only when all the computational results of emission bearings are pointed to a single small area with a high level of cross-correlation of most of the bearing data. If a strong noise signal has been received at one observation point, but the computational results of the cross-correlations between almost all other stations do not intersect in a small area, the alarm signal would not be announced. This technique elimi-

-

mma--

1

A /D

PCM- TO

1525 HZ Amp

N - S A

PM

A ID

Timing

Mod.

. I

82 kHz Amp

TM '

u

-,

A /D -

L.0 -

A

-

- Comp

Time. G. .

E - W 4

Amp

- Amp

82 kHz ID

T. Yoshino

ROCK SPECIMEN

E E * 00

TO AMP & RECEIVER

Fig. 7. Size of rock specimen for electromagnetic emission observation, before being crushed by pressure machines (after Mizutani, 1987).

-I- I I I I I I

Fig. 8. Acoustic emission (AE).(top) and EM emission wave form observed at 1.0,0.5,0.3,0.2, and 0.1 MHz (second from top to bottom). The time difference between the AE and EM onset times correspond to the travel time of AE (after Mizutani, 1987).

Seismogenic emissions

200 r K E Y

z Salida Granite 120 . .--. -. .. . .

Yule Marble ----

0 2 4 6 8 10 FREQUENCY, kHz

Fig. 9. Laboratory experimental results showing frequency characteristics of emissions when the rock is crushed by a pressure machine (after Cress et al., 1987).

nates man-made noise interference because it is not usually radiated beyond 50 km from its sources.

Model of Radiation Mechanism of Seismogenic Emissions

The author has offered a possible mechanism for the seismogenic electromagnetic radiation above. The emission will be induced as one of the kinds of boundary-charge phenomena when the rocks around the focus of an earthquake are crushed under the very strong distortion forces increasing rapidly just prior to the earthquake. Laboratory experiments have been done by Mizutani and Yamada (1987) and Cress et al. (1987). In these experiments, very strong electromagnetic impulses were observed at the instant the specimen of rock was crushed under high pressure. The values of induced electromagnetic emission obtained were different for each kind of rock, wet or dry, and are dependent on the many different conditions at the time of observations. The size of specimen is shown in Figure 7, and one of the examples of experimental curves is shown in Figure 8. The observation results of Cress et al. are given in Figure 9. Compared with the example of natural emissions observed at Sugadaira, which was shown in Figure 2, the traces of natural and experimental data show very good agreement in frequency characteristics with peaks around 1.5 kHz.

The authors have built a model to explain the radiation mechanisms of seismogenic emission, as shown in Figure 1 0. Usually the earthquake focus is located on the fault plane as shown in this figure. If the distortion forces are

132 T. Yoshino

EARTHQUAKE Fig. 10. Model to explain the radiation mechanism of seismogenic emissions.

increased in the fault area, the crushing of pieces of rock is initiated, and the energy of induced electromagnetic impulses is emitted within this region in the fault.

The electrical conductivity inside and directed toward the fault line is usually higher than for the rock outside, this difference being over approximately 20 dB. Also, a sheath structure of high dielectric soils (fault gouge) is often observed on the boundary surface of faults. This condition will be able to support the surface mode of the TEM-electromagnetic wave propagation along the boundary of the fault, and the energy of seismogenic emission will be able to be transmitted from the source depth to the earth's surface with very little attenuation, as compared to the usual plane wave propagation outside the boundary.

The surface mode propagation of electromagnetic waves was developed by Goubou (1950) and Cullen (1954), and today this advanced technique is often applied to microcircuit designs for compact equipment systems in the centimeter- and millimeter-wave bands. The profile of the surface boundary structure is illustrated in Figure 11. As shown in this figure, the boundary surface of a fault has a structure similar to a surface wave transmission line.

The results of the numerical simulation on the estimation of attenuation values for the surface mode wave propagation along the boundary surface of the fault are as follows:

1. resistivity outside of the fault is 10 kL?/m; 2. resistivity inside of and parallel to the direction of the fault is less than

10 Blm;


E L E C T R I C F I E L D LINE

\ f , D I E L E C T R I C

METALIC CONDUCTOR Fig. 1 1 . Structure of a surface wave transmission line and electric field lines of force for the

dominant T M mode.

3. specific dielectric constant at the boundary surface of the fault is 20, and outside the area of the fault is 6.0;

4. frequency is 82 kHz, 1.525 kHz, and 36 Hz; 5. depth of focus is 50 km; and 6. TM mode.

The calculated value of total propagation loss when the depth of focus is 50 km for 82 kHz is approximately 63 dB, for the case of dielectric sheath thickness of 10 m, and 65 dB for 20 m. In the case of 1.525 kHz, it is approximately 56 dB in power ratio. If a large quantity of acid water is contained in the fault, the propagation loss will decrease more than 10 dB from the above calculated values. Ifthe ground surface is covered by homogeneous soil and rocks, the value of transmission loss for surface mode propagation is more than 25 dB below the value of plane wave propagation at the same depth.

The radiation impedance matching between surface mode feed and free space radiation mode by a slot antenna, which consists of the boundary between the top end of the fault and the ground surface, was also simulated for many types of matching systems. One of the best cases of VSWR values that can be obtained is 1.5 for the structure shown in Figure 12. We are continuing estimates and experiments on surface mode propagation and impedance-matching for surface radiation by means of a scale model for higher frequencies.

Observation Results at the Eruption of Mt. Mihara

Before July 1986, the volcanic activity at Mt. Mihara had been quiet for the previous 12 years (Weather Record, 1987a). Our observation equipment

134 T. Yoshino

Fig. 12.

Fig. 1 3.

FAULT

I

E WAVE Structure of a slot antenna on the surface of the ground and a fault excited by a TM mode surface wave.

Map of Izu-Ohshima which illustrates the location of the main eruption, the and the recording stations. (After Yoshiro et al. 1989.)

new craters,


was set up in the Ohshima Volcano Observatory of Tokyo University, east of Motomachi, as shown in Figure 13. A typical example of our usual monthly noise conditions data at 82 kHz, recorded by the equipment mentioned above and located at Ohshima, is illustrated in Figure 14a and b. Figure 14a shows the north-south data, Figure 14b the east-west data, and the data are compensated using the monthly average.natura1 background noise level. The sampling rate is 10 seconds. The natural background noise level in the night- time is usually 8-10 dB higher than the daytime level, owing to the effect of lightning noise from the southern tropical regions and the good propagation conditions of the night E-layer in the ionosphere. The average noise level of the north-south direction is always a few decibels higher than in the east- west direction because Ohshima Island is located south of the large man- made noise emission areas of Tokyo, Yokohama, and the large industrial area in Kanagawa prefecture.

Volcano microvibrations were observed at Ohshima observatory from July 1986 onward (Weather Record, 1987b), but the anomalous impulsive

1 7. -,

ta- - . I S. /

144- *

Fig. 14. Typical example of noise levels during the quiet period 17 months before the eruption. (a) North-south direction; (b) east-west direction. (After Yoshiro et al. 1989.)

136 T. Yoshino

noise emissions at 82 kHz did not appear until after October 20. The observation of several clear burst-like emissions were recorded from November 3-22. The plots of our improved analyzed data for a sampling time of 1 second at Ohshima on October 20 and during all of November are presented in Figure 15a, b, and c. Figure 15a shows the data for the first observed anomalous emission, Figure 15b shows the data from November 1-1 3, the

I I I I I I I I ' I ,

Oshima

Fig. 15. (a) First clear anomalous burst-like emission of an 82-kHz magnetic field observed on October 20 at Ohshima, 26 days before the eruption. (b) Eighty-two kHz data for November 1-1 3 at Ohshima. The upper plots show the north-south directional data, and the lower plots show the east-west directional data. (c) Data for November 14-25 at Ohshima. The arrow on November 15 is the starting time of the eruption at summit crater A and the arrows on the 2 1 st are the starting time of the eruptions in craters B and C. (After Yoshiro et al. 1989.)


two-week period just before the eruption, and Figure 15c shows the data for the eruption period from November 14-23. The upper plots show the noise level of the north-south sensor and the lower plots show the level or the east-west sensor. The noise level during the night is usually 6- 10 dB higher than the daytime level as a result of lightning in the southern tropical region.

As shown in Figure 15a and b, the observation of burst-like emissions occurring four and two weeks prior to the first eruption were identified at 14- 16 JST on November 3, at 09- 1 1 and 16 JST on November 4, at 1 1 :30 and 14- 16 JST on November 8, at 09 JST on November 10, and at 1 1 JST on November 1 1.

At 17:25 JST on November 15, the first major eruption occurred in the main crater. The location of this crater is indicated as A in Figure 13. Anoma- lous burst-like emissions were observed at 10- 16 JST on November 14, one day before this major eruption, but were not observed at the time of eruption. The volcanic activity continued with violent eruptions, and the lava fountain reached heights over 200 m. Lava flowed out from the summit crater A to the caldera at 10:35 JST on November 19, as shown in Figure 1 3. The earthquake and volcanic microvibrations during the eruption in crater A continued violently, and burst-like emissions were only observed at 08-09 JST and 14- 15 JST on November 17, but the average background noise levels at night were 12-20 dB higher than the usual daytime levels. At 23 JST on November 19, the eruption activity of crater A quickly decreased.

At 10- 12 JST on November 2 1, several strong burst-like emissions were observed, as shown in Figure 1 5c, the strongest peaks during these emissions reaching over 12 dB more than background noise level. At the same time, strong local earthquakes started and continued until evening. At 16: 15 JST, 4 hours after the emissions at 10- 12 JST, eruption occurred in 19 new craters, which appeared along a line trending NW-SE in the bottom of the old caldera. The location of this group of craters is labeled B, and shown in Figure 1 3.

At 16: 15 JST, summit crater A (old crater) erupted violently, and at 17:45 JST new craters suddenly appeared in the virgin fields on the northwestern slope of the mountain. The location of this group of craters is labeled C in Figure 13. Lava flowed down rapidly toward the station at upper Motoma- chi, as shown in Figure 13. Motomachi has the largest population on this island. By special order from the Mayor of Tokyo, a state of emergency was declared and all of the 10,000 inhabitants of the island, except for a few scientists, policemen, and firefighters, were evacuated until the early morning of November 22. The major eruptions in crater B and C completely ceased on the morning of November 22. Our data recording also ceased after November 25 because the sensors were destroyed by the lava flow. Later on, our station was moved to a safer area near Nomashi village, 2 km south of Motomachi, and the replacement equipment continues to operate at Noma- shi observatory.

138 T. Yoshino

Observed Results at an Eruption in Izu Itoh Bay in July 1989

Since early April 1989, volcanic microvibrations and earthquakes were observed at a very local area around the city of Izu-Itoh. After early June, the number and amplitude of these volcanic earthquakes increased significantly, and felt earthquakes increased to several hundred per day into early July. The inhabitants of the city were afraid and anxious to know the location of the next eruption, whether on the bottom of Ito Bay or on land. The worst case would be a sudden eruption in an area of high population density, with new craters erupting lava and scoria in the city. In this case, a huge number of lives would be lost, and be the major city of Itoh would be destroyed.

The observed magnetic field emissions increased in proportion to the increase in occurrence of volcanic earthquakes. Observations of magnetic field emissions at 1.525 kHz and 36 Hz at Ohshima Island increased during July 1-26 in 1 989, as shown in Figure 16a and b. Many emissions were recorded from July 1 to noon of July 1 1, as shown in these figures, but after noon of July 1 1, the characteristics and levels of emissions seemed to change. The volcanic activity also changed. A very large number of volcanic earthquakes occurred before noon of July 11, but after this time, the volcanic activity changed suddenly from the volcanic earthquake mode to the strong volcanic vibration mode. Many inhabitants of Itoh city had listened to the loud sound incessantly like the striking of very large drums with many huge hammers, just under Itoh City. However, no clear emissions were observed during these violent volcanic vibrations after noon of July 1 1.

Fortunately, at 16:24 JST on July 13, the eruption started at a new crater in the bottom of Itoh Bay. This point was located only 2.5 km from the Itoh shore, on the eastern side of the Izu peninsula, as shown in Figure 17.

The key point in our observed data for this volcanic eruption is that we received very many emissions at 36 Hz during the time of active volcanic earthquakes, until approximately noon of July 1 1. Then, the mode of vibrations shifted from volcanic earthquakes to volcanic vibrations, and the emission signals ceased, as shown in Figure 16a and b. This result strongly suggests a most interesting point, namely that emission occurred when the rock was crushed in this event. Unfortunately, we could not obtain data in the most important time, from July 6- 1 1, as mentioned above.

Discussion of the Source Mechanism

The observation of anomalous, burst-like emissions at 82 kHz does not directly correspond to the time of the initiation and duration of the most active eruptions, as can be deduced from inspection of Figures 4c, 7a, and 7b. In the case of Ohshima, the observed anomalous emissions at 10- 12 JST on November 21 occurred 4-6 hours before the eruptions from the new craters on the virgin fields inside the caldera at point B and on the mountain slope outside the caldera at point C , as shown in Figure 13.

TlME (JST)

I= -1 3tI:0 fl! P-735.0 198917' $ -140.0 $ -145.0 0 -150.0 J g 0 3 6 9 12 15 18 21 24 U-

TlME (JST) Fig. 16. Intensity of the seismogenic magnetic emissions observed at Izu-Ohshima in July 1989, during the volcanic activity and eruption at Izu Itoh Bay.

1.525 kHz, and (b) at 36 Hz, respectively. On 16:24 JST, July 13, 1989, a new crater erupted in the bottom of Ito Bay, and this point was located 2.5 km from the coast of a major city. Between July 6- 1 1 there are no data owing to trouble with our marine telemetry cable, possibly an effect impending eruption.

MT. FUJI I

OBS ;I Oxx STN u4- MT MIHARA

IZU - OHSIMA

Fig. 17. Location of the new volcano which erupted in Izu Itoh Bay on July 13, 1989.


Geological studies after the eruption of Ohshima indicated that the mixing ratio of silica in the lava and scoria erupted from the summit crater A differed from that of the lava that extruded from locations B and C (Sasai, 1987; Utada, 1987; Weather Record, 1987c; Yoshino et al., 1989). These results suggest that the source of lava extruding from the summit crater A had a different source from the lava flow extruding at locations B and C. A preliminary model to explain the eruption mechanism of September 2 1 was developed by Prof. Aramaki of the Earthquake Research Laboratory, University of Tokyo. We have applied his model (Aramaki, 1987) to explain our observed results. Figure 18 shows the movement of magma based on this model. The mechanism for these anomalous burst-like emissions is explained as follows: (1) the magma flow to the eruption of the summit crater A was supplied directly from the base magma as primary magma Y; and (2) the eruptions in the new craters B and C were created by new magma flows, forced up as dikes inserted into the mountain body on the morning of No- vember 2 1 (Aramaki, 1987).

The quantity of dike material seems to have been approximately equal to the size of magma volume for the total eruption of this volcano (Weather Record, 1987d). The basalt with a high mixture ratio of silica grains was formed under craters B and C by the heating effect of this dike, and this

Mt. MIHARA

O o Y 2 krn

d \Cmmmm BASE MAGMA

Fig. 18. Vertical profile of Mt. Mihara, illustrating the mechanism of magma flow in the mountain body based on the Aramaki (1987) model. The summit crater A erupted as a result of the insertion of the primary magma Y, and the new crater B and C, erupted as a result of basalt Y, which was heated by a dike of primary magma. The starting time ofthe dike insertion corresponds to the time of the burst-like emissions observed during 10- 12 JST on November 21, 1986. (After Yoshiro et al. 1989.)

142 T. Yoshino

molten basalt was extruded 600 m from the mouth of the craters as a result of the high back pressure of the internal gas, which was created as a result of high density of silica grains. The eruption from craters B and C ceased a few hours after it began, and duration of this short but violent eruption fits the Aramaki model well. At the same time, data obtained from other kinds of instruments suggested that the commencement time on the dike insertion was around 10-12 JST on November 2 1 (Sasai, 1987; Yukutake et al., private communication, 1987).

To explain the emissions at dike insertion, we apply the results of laboratory experiments by Mizutani and Yamada (1987) and Cress et al. (1987). We believe that the observed emissions on November 2 1 were produced by the rock crushing due to the insertion of dikes. Emissions recorded during 10-16 JST on November 14, 10-14 JST on November 17, and the other various emissions that were observed before the eruptions from crater A, can be explained as the impulse emission induced during the intervals when cracks propagated through the rocks as a result of the movement of magma or dike intrusion prior to the eruption. These phenomena can be explained and the model developed can be proved by means of the results of the ongoing laboratory experiments of Cress and Mizutani.

Conclusion As shown in this article, the building of a multipoint observation system

for the prediction of earthquakes by use of the seismogenic electromagnetic emission phenomenon is progressing positively. The characteristics of the receiving system and the practical design of the details of facilities are completed, and now we are continuing the telemetry system development for high speed, high accuracy, the highest reliability, and the lowest operational cost for continuous and long-term automated operation. We are developing the computer software to obtain the most reliable detection algorithm for the prediction at the central computer, and also for transmission of data to the local earthquake alarm center.

The most important aspect for the development of this automatic alarm system is protection and discrimination against man-made noise interference. We are developing a noise-reduction method using a network spread over eight observation points, with a point-to-point distance of approximately 55 km. Each observation point consists of a LF (82 kHz) direction finding system. The bearing and signal intensities are converted to 16-bit digital values and transferred through a telemetry system to the computer in the warning center. When the cross-correlation appears to be high and the bearing data of all or several points are directed to one area, the alarm signal will transmit an earthquake warning and show the predicted epicenter area automatically. If the signal intensity is strong at only one or a few points and the calculated cross correlations are very low, this increase of signal intensity will be ignored as man-made noise interference. Although the author had only one case where we could clearly have predicted an epicenter location


observational data retrospectively. We believe strongly that this will be one of the most reliable systems for man-made noise reduction. And we have started the search for new observation frequencies in the VLF range, below the cut-off frequency of the VLF ionospheric guided-mode propagation (1.525 kHz) and the higher frequencies of ELF (36 Hz) in the Schumann resonance band. Attempts to eliminate natural noise interference from lightning discharges in tropical regions were started by our research groups.

In our investigations of the noise source and propagation path of seismogenic electromagnetic emissions as a precursor of earthquakes, we utilized the experimental results of electromagnetic emissions due to the crushing of rocks by Mizutani and Yamada (1 987) and Cress et al. (1 987) to explain the mechanism of emission around the focus of the earthquakes. We also applied the theory of surface mode propagation of electromagnetic waves along the boundary surface of a fault to explain why the energy transmission of seismogenic emission from the focus to the ground surface is less attenuated in comparison with the case of usual plane wave transmission. On the problem of the radiation mechanism of waves at the surface of the ground, the authors applied the radiation mechanism of slit antennas which consists of the top of the fault at the ground surface. However, the theoretical explanation of the impedance-matching between a surface mode transmission line and a slit antenna is very complicated, and still requires a more detailed investigation.

We roughly estimated the transmission loss by use of an optimum value, and averaged measurements for a typical fault in the Kanto area. We then concluded that the total attentuation loss for a specific case of the surface mode transmission and slit antenna radiation was 26 dB lower than the case of usual plane wave transmission without a fault. The precise investigations for this problem have to be continued for each earthquake observed from now on.

In this study, we presented the results of observed electromagnetic emissions related to the volcanic eruptions at Mt. Mihara on Izu-Ohshima Island during November, 1986, and at Izu Itoh Bay during July, 1989. These recordings are believed to represent the first observations of electromagnetic emissions during a volcanic eruption measured anywhere in the world. Since 1987, the data recorded at Ohshima have used one of the 82-kHz direction- finding detectors and two new frequencies at 1.525 kHz and 36 Hz in a multipoint receiving network around the Tokyo area which was designed for the detection of electromagnetic emissions prior to earthquakes. This earthquake prediction experiment was started in 1982.

A large number of data sets were obtained by other kinds of instruments during the time interval mentioned above, including: DC, ELF, and VLF conductivity, telluric currents, total magnetic-flux density, gravity, seismol- ogy, volcanic microvibrations, etc. (Sasai, 1987; Utada, 1987; Weather Rec- ord, 1987d; Yoshino & Tomizawa, 1989; Yukutake, private communication, 1987). These data support the theory that dikes branched off from the

144 T. Yoshino

These observations suggest that the source mechanism of emissions at Mt. Mihara can be explained by the Aramaki (1987) model, as well as the more interesting observational results obtained from the volcanic eruption in the Izu Itoh Bay on July 13, 1989, and the experimental results of the ongoing laboratory experiments by Cress et al. (1987) and Mizutani and Ya- mada ( 1987).

Author's note. A portion of this research was financially supported by Japan IBM Corporation in 1986.

References

Aramaki, S. (1987). A suggestion of possibilities of dike insertion at the eruption on November 2 1, 1985. S.vmposiurn on the Eruption of the Izu-Ohshimu Volcano in 1986. Tokyo Univer- sity, 1 1 - 15, April (extended abstract).

Cress, G. O., Brady, B. T., & Rowell, G. A. (1987). Source of electromagnetic radiation from fracture of rock samples in the laboratory. G~oph.vsics Rc~search Letters, 14, 33 1-334.

Cullen, A. L. (1954). The excitation of plane surface waves. IEE Part IT/, 101, 225. Chmyrev, V. M. et al. (1987). Electric fields and hydromagnetic waves in the ionosphere over

the earthquake centre. IUGG Abstracts, 1, 384. Gokhberg, M. B. ( 1984). The models of electromagnetic processes related to seismotectonics.

Terra Cogniru, 4, 369. Gokhberg, M. B. et al. (1 987). The earthquake preparation processes and earth's crust electro-

magnetic emission connecting modeling. IUGG Abstracts, 1, 382. Gokhberg, M. B., Morgounov, V. A., Yoshino, T., & Tomizawa, I. (1982). Experimental mea-

surement of electromagnetic emissions possibly related to earthquakes in Japan. Journal of Geophysics Research. 87. 7824-7827.

Goubou, G. (1950). Surface waves and their application to transmission lines. Journal qfAp- plied Physics, 21, 1 1 19.

Larkina, V. I. et al. (1987). Low frequency emission data in upper ionosphere over the earthquake center. IUGG Abstracts, 1, 384.

Migulin, V. V. et al. (1987). Earthquake electromagnetic forerunner diagnostic problems by wave experiment low latitude satellite data. IUGG Abstracts, 1, 384.

Mizutani, H., & Yamada, I. (1987). Electromagnetic emission and acoustic emission associated with rock deformation. XIX General Assembly of the IUGG, Vancouver. Abstracts, 1, 384.

Sasai, Y. ( 1987). Geomagnetic field variations before and after the eruption of the Izu-Ohshima volcano in 1986. Tokyo University 23-25, April (extended abstract).

Utada, H. (1987). Variations in the electrical resistance before and after the eruption of the Izu-Ohshima volcano in 1986. Tokyo University, 26-28, April (extended abstract).

Weather Record (Japanese Meteorological Agency-JMA), Volcano Research Section of JMA (1987a). Special Issue on the eruption of the Izu-Ohshima volcano in 1986. Kisho, 31(1), 8-25 (in Japanese).

Weather Record (1987b). Kisho, 31(1), 35-44. Weather Record (1987~). Kisho, 31(1), 26-33. Weather Record ( 1987d). Kisho, 31( I), 74-82. Yoshino, T. (1986a). On the study of electromagnetic emissions related to earthquakes and

application to earthquake prediction, JIEICE Technical Report, 85, 19-24, 3 13 (in Japa- nese).

Yoshino, T. (1986b). The EM emission phenomena as a precursor of earthquakes and the possibility of epicenter location prediction. Proceedings of the 8th International Wroclaw Symposium EMC, 8, 5- 14.

Yoshino, T., & Tomizawa, I. (1989). Observation results of low-frequency electromagnetic emissions as precursors of the volcano eruption at Mt. Mihara during November 1986. Physics of the Earth and Planetary Interiors, 57, 32-39.

Yoshino, T., Tomizawa, I., & Shibata, T. (1985). The possibility of using a direction finding technique to locate earthquake epicenters from electromagnetic precursor radiation. An- nales Geophysicae, 3, 727-730.

Society for Scientific Exploration

Officers

Peter A. Sturrock, President Laurence W. Fredrick, Secretary Center for Space Science and Astrophysics, Department of Astronomy

ERL 306 P.O. Box 38 18 Stanford University University of Virginia Stanford, CA 94305 Charlottesville, VA 22903

Yervant Terzian, Vice President Charles R. Tolbert, Treasurer Department of Astronomy Department of Astronomy Cornell University P.O. Box 38 18 Ithaca, NY 14853 University of Virginia

Charlottesville, VA 22903

Council

John S. Derr U.S. Geological Survey Albuquerque Seismological Laboratory Building 10002, Kirtland A.F.B.-East Albuquerque, NM 87 1 15

Brenda Dunne Engineering Anomalies Research School of Engineering & Applied Science Princeton University Princeton, NJ 08544

Robert Jahn School of Engineering & Applied Sciences D-334 Engineering Quadrangle Princeton University Princeton, NJ 08544

Hal E. Puthoff B- 12 1, Institute for Advanced Studies Barnes-Connally Center 130 1 Capital of Texas Highway, S. Austin, TX 78746

Dean I. Radin AI/MMI Laboratory Contel Technology Center Chantilly, VA 2202 1-3808

Ian Stevenson Dept. of Behavioral Medicine and Psychiatry University of Virginia Charlottesville, VA 22908

Marcello Truzzi Center for Scientific Anomalies P.O. Box 1052 Ann Arbor, MI 48 106

Jessica Utts Division of Statistics University of California Davis, CA 956 16

Date post:	05-Dec-2015
Category:	Documents
Upload:	jonathanthorpe
View:	29 times
Download:	10 times

Journal of Scientific Exploration, Volume 5, Issue 1

Documents