Chapter 9 "Smoking, Cancer and the Final Frontier," from Roderick D. Buchanan, Playing with Fire: The Controversial Career of Hans J. Eysenck, (Oxford: Oxford University Press, 2010).

Abstract

Gives the detailed context for Eysenck's controversial collaboration with Ronald Grossarth-Maticek.

Full text

Chapter 9
Smoking, cancer, and the
final frontier
Time has a curious way of passing judgement on the actions of men and
women, a capricious capacity for turning out heroes and villains after the fact.
Time may well be relatively kind on Hans Eysenck and his race and IQ inter-
vention. Future science may well bear out the biogenetic basis of cognitive
differences, much more so than his pronouncements on smoking, cancer, and
personality. The increasingly dogmatic medical outlook on the dangers of
smoking has been woven with modern day tales of the evils of Big Tobacco and
the bought advocate, typecasting those involved to the right or wrong side of
history. However, a careful reconstruction of this final chapter in the Eysenck
story—sans hindsight where possible—should provide an antidote to any rush
to judgement.
Much of this story takes place during the latter part of Eysenck’s life and
involves one of his last collaborations. For over 30 years, Eysenck claimed that
medical scientists and the public health lobby had got it wrong. They had
greatly exaggerated the ill effects of tobacco. In their haste to condemn smok-
ing, they had ignored the possibility that genetic and personality factors were
largely to blame. Increasingly pilloried for this heterodoxy, Eysenck held out
for evidence that would back his position.
In the early 1980s, Yugoslav sociologist Ronald Grossarth-Maticek offered
Eysenck a lifeline. Grossarth-Maticek’s data strongly suggested that personal-
ity coping styles greatly affected the course of various physical diseases. In a
series of papers published in the late 1980s and early 1990s, Eysenck and
Grossarth-Maticek reported a variety of results that demonstrated a strong
association between particular personality types and cancer, coronary heart
disease, and other ailments. Although these were mostly write-ups of studies
set in train by Grossarth-Maticek more than a decade earlier, Eysenck’s input
helped fine-tune the presentation, theoretical explanations, and analyses. A
number of intervention studies were also carried out, each suggesting that
particular forms of psychotherapy that targeted cancer sufferers or unhealthy
personalities could have remarkable therapeutic or preventive effects.
09-Buchanan-Ch09.indd 36109-Buchanan-Ch09.indd 361 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
362
As in the race and IQ brouhaha, a raft of questions arise from Eysenck’s
actions. Why did he go out of his way to defend a suspect product like ciga-
rettes in the face of mounting evidence? In the past, critics have pointed vaguely
but darkly to his links with the tobacco industry, implying his views were influ-
enced by the money he received. Why too did he take up with the notorious
Grossarth-Maticek, a man short on support and allies within the scientific
community? This scepticism only intensified with Eysenck’s unabashed pro-
motion of Grossarth-Maticek’s research in the English-speaking world. A
diverse chorus of opinion suggested Grossarth-Maticek’s astonishing results
had little credibility. So, where did this leave Eysenck? He was Grossarth-
Maticek’s most ardent and prominent supporter, the co-author of many papers
reporting these results.
Eysenck’s sceptical stance on the connection between smoking and illness
appeared relatively unremarkable in the beginning. However, his continued
defence of this stance, the strategies he invoked, and the criticism he took, all
had a way of drawing him toward someone like Grossarth-Maticek and his
wondrous results. Their fateful embrace saw the Eysenck story go from the
controversial, touch on the scandalous, and, finally, descend into the tragic.
Smoking, personality, and disease
Eysenck wrote that his interest in the topic of smoking and health, and the
disease-prone personality, began when ‘I had shown that cigarette smoking
was significantly correlated with personality, specifically with extraversion,
and followed a lawful, predictable course.’1 In what could be seen as simple
extension of his biosocial perspective, Eysenck took seriously the ancient folk
wisdom linking temperament with physical health. Eysenck had become a
smoker as a student in London, a habit that was cemented during the war, and
one that escalated during his 1949 sojourn in the US. When he returned to
England, Eysenck began to notice the effect his pack-a-day smoking was having
on his sense of well-being and fitness, not to mention his hip pocket. Moreover,
as a good behaviourist, he felt ashamed to be a slave to this habit. So Eysenck
promptly—and effortlessly he said—gave up cigarettes. Thus he claimed he rid
himself of the weed before the link between smoking and disease had begun to
be widely advertised by medical authorities in the mid-1950s.2 Pity those who
1 See p. 297 of Eysenck, H.J. (1991c), Reply to criticisms of the Grossarth–Maticek studies,
Psychological Inquiry 2, 297–323.
2 See Eysenck, H.J. (1997b), Rebel with a cause [revised and expanded], Transaction Press,
New Brunswick, New Jersey, Chapter 5, and his introduction to Eysenck, H.J. (1965b),
Smoking, health and personality, Basic Books, New York.
09-Buchanan-Ch09.indd 36209-Buchanan-Ch09.indd 362 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, PERSONALITY, AND DISEASE
363
did not, he added, for alarmist health warnings would make it an increasingly
guilty pleasure.
Concern about the adverse health effects of smoking can be traced all the
way back to the importation of tobacco to Europe. These warnings became
commonplace in Western countries but were not taken seriously until the late
1940s, when it became clear that lung cancer was increasing dramatically—and
not just as a function of changing patterns of diagnosis.3 With acute infectious
diseases on the wane, chronic conditions like cancer and heart disease loomed
as a new mountain to climb for modern medicine. Cancer posed a special chal-
lenge, however, with its long latencies and potentially multiple aetiologies.
Research into causes had mainly focused on individual susceptibility, possibly
as some kind of imperfection at the cellular level mediated by host (i.e. genetic)
propensity. Nevertheless the steep rise in the incidence of lung cancer saw a
number of environmental triggers also fall under suspicion—smoking and
atmospheric pollution amongst them.
Teasing out causality from a multitude of candidate factors would not be
easy. While experimentation with human subjects was not practical, the use of
animals was also fraught. It was not clear whether humans and animals reacted
to carcinogens in the same way or whether cancers developed in the same
manner. Appropriate exposure methods were also a problem. At the time,
epidemiology was a young and fragile ancillary field, typically utilizing cross-
sectional studies of morbidity to shed light on the spread of infectious condi-
tions but not on their causes. All this would change by the early 1950s when the
results of several ambitious longitudinal studies were published—most nota-
bly those of Ernest Wynder and Evarts Graham in the US and A. Bradford Hill
and Richard Doll in the UK. Collectively, this data raised the possibility of a
strong link between smoking and lung cancer.4
Hill and Doll’s work was particularly influential, with their ground-breaking
1950 report setting a new standard in case control methods. Supported by the
Medical Research Council (MRC), Hill and Doll’s 1954 follow-up helped shift
3 See Burnham, J.C. (1989), American physicians and tobacco use: two Surgeons General,
1929 and 1964, Bulletin of the History of Medicine 63, 1. For an update on the fate of lung
cancer treatments see Timmermann, C. (2007), As depressing as it was predictable? Lung
cancer, clinical trials, and the Medical Research Council in postwar Britain, Bulletin of the
History of Medicine 81, 312–34.
4 For an excellent historical account of this debate, see Talley, C. et al. (2004), Lung cancer,
chronic disease epidemiology, and medicine, 1948–1964, Journal of the History of Medicine
and Allied Sciences 59, 329–74.
09-Buchanan-Ch09.indd 36309-Buchanan-Ch09.indd 363 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
364
scientific opinion and made the Ministry of Health take notice.5 It also spurred
those in the British tobacco industry into defensive action. Initially the indus-
try funded research into the causes of cancer with donations to the MRC.
Dissatisfied with the results, British tobacco formed the Tobacco Research
Council (TRC) in 1956, originally designated the Tobacco Manufacturer’s
Standing Committee. This industry-representative group was entrusted to
perform ‘research on smoking and health … and make information available
to scientific workers and the public.’6 However, the overriding and obvious
function of the TRC was to fight science with science and, at a public relations
and legal level, to combat warnings about the dangers of smoking.
By the late 1950s, biochemical analyses and animal experimentation began
to reinforce the smoking–cancer link repeatedly suggested by epidemiological
research. Most scientists were convinced cigarette consumption could at least
partly explain the upsurge in lung and bronchial cancers and the statistical
profile of heart disease. Even so, the epidemiological field remained a con-
tested space. The distinguished, pipe-smoking statistician R.A. Fisher was not
alone in doubting whether such studies could demonstrate causality. In his
1955 trip to Berkeley, Eysenck said he mused on Fisher’s suggestion that genetic
factors might underlie the relationship between smoking and disease. Since
Eysenck was sure personality had a basis in genetics, he reasoned that person-
ality might be an intervening variable in this link. Those with a certain person-
ality might be predisposed to smoke and might also be predisposed to get lung
cancer and other diseases. Eysenck was subsequently approached by the social
research group Mass Observation to do media work on smoking and they, in
turn, had been sounded out by the TRC to study smoking patterns. From
1958, Mass Observation did a series of studies directed by Eysenck that sug-
gested there were ‘genotypic differences’ between smokers and non-smokers.
Smokers scored higher on extraversion though not necessarily on neuroti-
cism.7 Eysenck claimed that later studies revealed that the smokers, especially
women, did tend to be more neurotic.8
5 Doll, R, and Hill, A.B. (1950), Smoking and carcinoma of the lung: preliminary report,
British Medical Journal 2, 739.
6 Review of past and current activities, Tobacco Research Council, January 1963, http://
tobaccodocuments.org/ctr/11297751-7780.html, p. 5.
7 Eysenck, H.J., Tarrant, M., and Woolf, M. (1960). Smoking and personality. British
Medical Journal 11, 1456–60; Review of past and current activities, Tobacco Research
Council, p. 16.
8 Eysenck (1997b), Rebel with a cause, Chapter 5.
09-Buchanan-Ch09.indd 36409-Buchanan-Ch09.indd 364 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, PERSONALITY, AND DISEASE
365
From this work, Eysenck went on to look at smoking-related diseases.
He was ‘lucky enough to meet’ Dr David Kissen, an oncologist who ran a chest
clinic in Edinburgh and a psychosomatic research programme at the University
of Glasgow.9 Kissen also did research financed by the TRC and appeared at
conferences and conventions that the British industry helped organize.
Collaborative work with Kissen demonstrated an association between cancer
and personality, with low scores on neuroticism associated with cancer, and
high scores with the absence of cancer.10 Kissen suggested this bore out the old
idea that cancer was associated with the suppression of emotion or a difficulty
in its expression. Kissen died in the mid-1960s of heart disease and Eysenck
cast about amongst medical researchers for new collaborators. No one was
willing to work with him, Eysenck said, or even let him test patients. It seemed
Kissen’s approach did not have a great deal of support at Glasgow, and his
psychosomatic research group went with him.
In Britain, the Royal College of Physicians had intervened in the debate in
1959. In 1962, they issued their first report pointing out the dangers of smok-
ing. Smoking was linked specifically to lung cancer, and it was also implicated
in other forms of cancers and coronary heart disease. The US Surgeon General
followed suit in 1964. Smoking was now officially regarded as a health risk—
even if the carcinogenic action of cigarette smoke was still to be thrashed out.
Given that the statistical data could not meet the highest standard of medical
proof (i.e. Koch’s postulates), a new standard of causality had been introduced
in tandem with the burgeoning field of chronic disease epidemiology.11 The
scientific debate was all but over, but the legal and PR war between the tobacco
companies and the public health lobby had only just begun. Hans Eysenck
came in at the tail end of this scientific debate and made little impact. However,
Eysenck would revisit the topic time and time again over the years, recapitulat-
ing the old doubts and adding a few of his own to become a significant player
in this new phase of the tobacco wars.
To this end, he ramped-up his public profile on this issue in 1965 with the
book Smoking, health and personality. Written in an accessible manner, pre-
sumably for the educated lay person, Smoking, health and personality was a
triumph of discursive even-handedness. The book was tailored to a public
9 Eysenck (1991c), Reply to criticisms of the Grossarth–Maticek studies, p. 297.
10 Kissen, D.M. and Eysenck, H.J. (1962), Personality in male lung cancer patients, Journal
of Psychosomatic Research 6, 123–7.
11 Harkness, J.M. (2006), The US Public Health Service and smoking in the 1950s: the
tale of two more statements, Journal of the History of Medicine and Allied Sciences 62,
171–212.
09-Buchanan-Ch09.indd 36509-Buchanan-Ch09.indd 365 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
366
debate that had already begun to polarize. With his distinctive blend of provo-
cation and equivocation, Eysenck presented his version of the history of the
dispute, his interest in it, and the cases for and against the link between smok-
ing and disease. He argued that the causal role of cigarettes in various diseases,
specifically lung cancer, had not been convincingly proven. Eysenck did not
present a great deal of his own data; he had done little relevant research. The
evidence he did present was tied to the suggestion that host factors (i.e. genet-
ics and personality) might play a mediating role in the smoking–disease link.12
Eysenck remained cautious, however, for available data did not appear to bear
this hypothesis out. While extraverts did tend to smoke more, they seemed
only slightly more likely to get cancer or heart disease. ‘It cannot by any stretch
of the imagination, be held responsible for the total effect,’ Eysenck concluded.13
Neither could neuroticism explain the link, since cancer victims appeared to
be less neurotic while smokers, if anything, were more so. Even so, Eysenck
could point to many holes, inconsistencies, and unexpected patterns in the
epidemiological data, just as Fisher had.14 Most of the research was retrospec-
tive, vulnerable to recall and reporting problems. Mortality statistics were
unreliable and demonstrated historical fluctuations. Finally, smokers and
non-smokers could not be considered to be matched groups; they might sys-
tematically differ in ways other than just tobacco use. Alternative or multifac-
torial causation of cancer was thus a possibility, deflecting the blame away
from smoking.
Overall, Smoking, health and personality did still give considerable credence
to the ill-effects of smoking, though in a most ambiguous way. For example,
Eysenck assigned his review of the experimental animal research to the section
outlining the case against the link between smoking and disease, with the sec-
tion detailing the case for the link dominated by a review of the statistical/data.
In doing so, he gave the impression that the experimental evidence was
overwhelmingly negative. This was especially evident in his lengthy summary
of Mayo Clinic epidemiologist Joseph Berkson’s critical review of the evidence.
But in an intriguing endnote to this section, Eysenck indicated that he thought
Berkson’s account was selective and incomplete. Eysenck went on to say that,
while other research did not ‘give completely convincing direct evidence of a
12 It was a hypothesis Eysenck had rehearsed two years before. See Eysenck, H.J. (1963),
Personality and cigarette smoking, Life Science 3, 777–92.
13 Eysenck (1965b), Smoking, health and personality, p. 118.
14 See Fisher, R.A. (1959), Smoking: the cancer controversy, Oliver and Boyd, Edinburgh and
Stolley, P.D. (1991a), When genius errs: R.A. Fisher and the lung cancer controversy,
American Journal of Epidemiology 133, 416–25.
09-Buchanan-Ch09.indd 36609-Buchanan-Ch09.indd 366 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, PERSONALITY, AND DISEASE
367
relationship, it cannot be denied, I think, that they do strongly support the
statistical evidence’.15 No doubt Eysenck might have defended his editorial
decisions based on the format of his book, but this important point was buried
in a footnote.
In his final summing up, given in mock legal style, Eysenck opined that the
case against smoking was ‘strong’. Nevertheless, he warned that the evidence
was only circumstantial rather than causally direct. ‘Clearly smoking is not
likely to promote anyone’s health… However, in science, as in law, the accused
is presumed innocent unless proven guilty…’16 With a hint of self-congratula-
tion, Eysenck gave smoking the benefit of the doubt.
Eysenck’s studied agnosticism in Smoking, health and personality gave him
plenty of rhetorical wiggle-room and laid logical booby-traps for opponents to
stumble into. Moreover, the message was layered to the point where, if neces-
sary, it could be sharpened to further provoke or toned down to be eminently
defensible. In self-consciously balancing the evidence pro and con, Eysenck
created the impression of reasonable doubt while apparently avoiding a fixed
position himself. Of course, Eysenck did take a position, at least implicitly, in
his selection and interpretation of a diverse range of evidence, but it was buried
in ostentatious displays of even-handed discussion. What was left to dispute
were his grounds for rejecting any judgement as premature or unwarranted,
and his authority and judgment in doing so.
While Smoking, health and personality did cause a stir, its rebel halo took on
a brighter hue in hindsight. When it first appeared it got many favourable
reviews and sold quite well. Moreover, Eysenck was not alone amongst scien-
tists in questioning the evidence. His views were echoed by a number of more
tendentious popular writers, notably C. Harcourt Kitchin in his 1966 book
You may smoke.17 Together these works helped promote the public perception
that warnings from medical authorities were alarmist and misleading. While
15 Eysenck, H.J. (1965b), Smoking, health and personality, p. 157, note 11.
16 Ibid., p. 151.
17 Apart from the dissenting scientists already mentioned, such as Fisher and Berkson,
see Brownlee, K.A. (1965), A review of ‘Smoking and health’, Journal of the American
Statistical Association 60, 722–39. Later publications questioning the link included Seltzer,
C.C. (1967), Constitution and heredity in relation to tobacco smoking, Annals of the
New York Academy of Sciences 142, 322–30; Seltzer, C.C. (1972), Critical appraisal of the
Royal College of Physician’s report on smoking and health, The Lancet 300, 243–8; and
Burch, P. (1974), Does smoking cause lung cancer? New Scientist, 21 February 1974,
458–63. For popular writers, see Kitchin, C.H. (1966), You may smoke, Award Books,
New York and the notorious industry-sponsored pieces by Stanley Frank: Frank, S.
(1968), To smoke or not to smoke, True Magazine, 15 January 1968. and (under the
09-Buchanan-Ch09.indd 36709-Buchanan-Ch09.indd 367 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
368
Eysenck disagreed with the certainty expressed by the British and American
medical communities, his cautiously dissenting message provoked little
response because the scientific debate was largely seen as settled while the anti-
smoking public health campaign had barely got off the ground.
Eysenck gave the impression that the controversy surrounding Smoking,
health and personality was enough to ensure that the TRC stopped funding his
work.18 This is an odd point, since it might be assumed they would welcome
any such stir. In any case, tobacco industry documents reveal that the TRC
continued to fund Eysenck’s research, such as that on the positive, stimulating
effects of nicotine, until the summer of 1970.19 By this stage other, more lucra-
tive funding avenues had opened up.
Eysenck published very little research or commentary in this area again until
the 1980s.20 Even so, he continued various programmes aimed at elucidating
the hidden variables in the smoking–disease link, especially genetic and per-
sonality factors. While Eysenck happily disclosed his funding source as the
tobacco industry, he did not care to detail his deep involvement with the pow-
erful American companies.
Big Tobacco, big money
When the law is not on your side, try the facts; when the facts are not on your side,
argue a point of law; when neither facts nor the law is on your side, try the plaintiff.
(Legal defence maxim)
The key defence strategy [for the tobacco industry] is to try the plaintiff (Crist, P.G.
et al. (1986), Re: Jones/Day Liability Summary (‘Corporate Activity Project’), http://
tobaccodocuments.org/landman/37575.html)
It is not easy to trace how Eysenck became involved with the American tobacco
industry. It is clear, however, that those within the American industry were
well aware of him and had been monitoring his writings on smoking and
pseudonym Charles Golden) Golden, C. (1968), Cigarette cancer link is bunk, National
Enquirer, 3 March 1968.
18 Eysenck (1997b), Rebel with a cause, Chapter 5.
19 Tobacco Research Council, Review of activities 670000—690000, http://www.tobacco-
documents.org/lor/00500025-0085.html.
20 The most notable piece Eysenck published in this period proposed a model of smoking
behaviour. See Eysenck, H.J. (1973c), Personality and the maintenance of the smoking
habit, in Smoking behaviour: motives and incentives (ed. W.L. Dunn), pp. 113–46. Winston/
Wiley, Washington.
09-Buchanan-Ch09.indd 36809-Buchanan-Ch09.indd 368 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

BIG TOBACCO, BIG MONEY
369
health since the early 1960s.21 Just as clear is the trail of money Eysenck had
begun to receive from the Americans, beginning in the late 1960s.
Like their British counterparts, the American-based tobacco giants battled to
turn the negative tide against smoking. However, the American companies
were engaged in an even tougher fight in their domestic market; they had to
endure intense scrutiny in Congress and the never-ending threat of litigation
from those who had used their products. It made the actions and attitudes of
US industry leaders more ruthlessly instrumental and, crucially, these attitudes
and actions were harnessed to immense resources. The Americans were quick
to seize on anyone or anything that might prove useful in shoring up their
legal-cum-commercial position.
In an attempt to influence public opinion and gain legal purchase, those
within or associated with the American tobacco industry began to monitor
scientific debates and develop a role in them. Beginning in the mid-1950s, they
sponsored outside research projects and set up some in-house research pro-
grammes. Yet when it came to science, they quickly realized that any rapacious
acquisition of evidence and experts had to be tempered with a consideration of
what made science trustworthy and persuasive. To be legally and politically
bullet-proof, scientific research had to be seen to be independent.
In January 1954, the US tobacco industry responded to adverse publicity
with a widely publicized ‘Frank statement to cigarette smokers’. At the time,
this statement was more or less in tune with mainstream scientific opinion,
namely, that it had not been scientifically established that smoking caused
diseases like cancer. After 1964, however, only the tobacco industry insisted
that the question was still unresolved. In order to keep the appearance of
controversy alive the American industry went out of its way to sponsor and
promote those few scientists still prepared to dispute the smoking–disease link.
One of the main ways of doing so was through the Council for Tobacco
Research (CTR), originally the Tobacco Industry Research Council. Set up in
1954 by five of the six American companies, the CTR and its Scientific Advisory
Board (SAB) was made up of non-industry scientists and was supposed to be
hands-off. Within certain parameters it usually was. Most researchers were
given a free rein once they received the go-ahead, but they had to pass through
the eye of a needle to do so. Industry representatives routinely passed over
any researcher with a negative attitude on the health issue and screened out
21 See Henry H. Ramm to James I. Erickson, 27 July 1965, http://www.tobaccodocuments.
org/rjr/500887047-7048.html.
09-Buchanan-Ch09.indd 36909-Buchanan-Ch09.indd 369 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
370
‘dangerous project proposals’.22 Moreover, they monitored the work subse-
quently done, halting some projects (e.g. central nervous system research), and
censoring and harassing the few who did uncover and attempt to publish dam-
aging results (e.g. Dr Frederic Homberger). Over the years they learnt to hide
research that had the potential for trouble, by funding them ‘off the books’.23
For many years (at least since the early 1960s) American tobacco industry
leaders privately acknowledged that they could not disprove the case against
smoking; an aggressive counter-attack was never on. Instead, industry repre-
sentatives threw up their hands over conflicting evidence. While reluctantly
fulfilling their duty to warn and to investigate, their main tactic was to find
ways to muddy the scientific waters that made the see-no-evil public stance
more credible. It was the scientific arm of a much bigger PR effort that
critics have since dubbed a ‘disinformation machine’.24 The CTR was ostensi-
bly intended to resolve the ‘open question’ on tobacco and health but was,
in the words of founding SAB chairman Dr Clarence Little, actually created
to build a research foundation to ‘arrest continuing or future attacks’.25
Thus some in-house and CTR research was conducted on a need-to-know
basis; however, most of it was carried out because it might serve to cloud or
complicate the causal link between smoking and disease. It was a tactic they
executed brilliantly, but it was a time bomb. It bet against the results of future
research and it assumed that industry confidentiality could be maintained
indefinitely.
The American companies’ relationship with Eysenck started in the late 1960s,
soon after Smoking, health and personality was published. Up to the mid-1960s,
the American industry had been content to fund domestic researchers. Early
victories in smoking health litigation had led to a sense of complacency.
Industry leaders were given a rude shock in the mid-1960s, however, when
the negative assessment of the Surgeon General was quickly followed by
congressional hearings on possible health warnings. Legal counsel for the
industry were in dire need of expert witnesses to testify that association was
not equal to causation and that you can’t extrapolate from animals to humans.
22 One of the most definitive reports documenting the strategies the industry used and the
precarious legal position that the industry eventually found itself in is Crist, P.G. et al.
(1986), Re: Jones/Day Liability Summary (‘Corporate Activity Project’), http://tobacco-
documents.org/landman/37575.html.
23 See Hilts, P.J. (1997), Smokescreen: the truth behind the tobacco industry cover-up, Addison-
Wesley, Reading, Massachusetts, Chapter 2.
24 See Ibid.
25 Crist et al. (1986), Re: Jones/Day Liability Summary (‘Corporate Activity Project’), p. 18.
09-Buchanan-Ch09.indd 37009-Buchanan-Ch09.indd 370 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

BIG TOBACCO, BIG MONEY
371
However, the industry had no relevant research to present to congress. Tobacco
research had become unpopular; regular journals no longer published it, and
government funding and CTR applications had dried up. Thus, CTR Special
Projects was born in 1966, narrower in scope than the regular CTR grant pro-
grammes, and usually directed toward specific legal, and political applica-
tions.26 Industry insiders actively solicited researchers and projects, with the
underlying aim being the development of expert witnesses. Since it was diffi-
cult to find scientists who did not think tobacco was harmful, CTR Special
Projects served as a valuable recruiting tool. New witnesses were always needed
to supplement the same old roll call that judges tired of and prosecuting advo-
cates could learn to dissect.
Eysenck became part of the CTR team in 1969. Noted American cancer
researcher and CTR member Dr Arthur Furst helped secure Eysenck’s initial
$10, 000 Special Projects grant for an investigation of tumour induction in
‘emotional’ versus ‘non-emotional’ rats. Eysenck had wanted to explore
personality theories in cancer, but was advised at the time that the SAB was not
very sympathetic to psychological theories of cancer. Nevertheless, Furst’s
opinion was that there was ‘still a long shot that this work may prove
useful’.27
Eysenck thus entered a truly nether world of science, full of highly partisan
interests, heavy secrecy, and unspoken obligations. But the Americans surely
held out one great attraction to him: they had boatloads of money, especially
for the right kind of research. While Eysenck might have recognized the poten-
tially corrupting effects of tobacco money, he must have felt immune in some
sense. For Eysenck, it was simply business as usual, merely a shift in funding
source. As he commented to Furst in 1970, ‘we have built up an excellent
research unit with unusual facilities for experimental work on nicotine, smok-
ing etc. [and] it would seem a pity if this had to be disbanded.’28
The relationship was maintained by both sides, mostly at a respectful dis-
tance, given the quid pro quo involved. Eysenck took many proposals to the
American tobacco industry during the 1970s and early 1980s. It appeared that
many, if not most, were accepted and supported. Much of this work was
funded as part of CTR Special Projects, as well as under the regular CTR
banner. The projects were typically related and often overlapping. One large
26 Ibid., pp. 142–3.
27 Arthur Furst to David R. Hardy, 24 December 1969, http://tobaccodocuments.org/tplp/
MNATPRIV00013508.html, p. 1.
28 Hans Eysenck to Arthur Furst, 20 May 1970, http://www.tobaccodocuments.org/ctr/
CTRSP-FILES008794-8794.html, p. 1.
09-Buchanan-Ch09.indd 37109-Buchanan-Ch09.indd 371 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
372
project that commenced in the early 1970s was based on data derived from the
twin registry Eysenck had set up. With the help of Oxford biometrician Lindon
Eaves, the idea was to investigate the ‘inheritance of the smoking habit’.
As well as funding for specific research, Eysenck received additional funds
from the American tobacco industry that indicated he had been earmarked for
an exceptional role. From the late 1970s on, Eysenck received generous pay-
ments from special account #4. These special accounts were a secret source of
funds operated by industry lawyers. They were used to fund projects that oth-
erwise did not gain approval or fit CTR guidelines, projects judged to be
primarily of legal (as opposed to scientific) value. And because these accounts
were controlled by the lawyers, the work they financed could be retained as a
privileged legal product. This avoided unwanted disclosure to the Federal
Trade Commission and made results ‘non-discoverable’. In other words, these
special accounts allowed those in the industry to hide research from public
view, where it was expedient, especially from plaintiff lawyers. Using this scien-
tific evidence in court, incidentally, meant placing favourable results in the
hands of an expert witness, for legal counsel could not present such evidence
themselves.
Eysenck was paid at an annual rate of $US 19, 000 in 1977 from special
account #4. Payments were progressively boosted up to $40, 000 in the years
following to at least 1983, for what was described as ‘consultative research’.
It is difficult to ascertain just what this amounted to but it appeared that
Eysenck gave assistance of a strategic nature, advising the industry as to what
research might be worth doing and building contacts with willing researchers.
Eysenck was enlisted to help set up research projects in Europe and the US,
some of which were ‘strictly litigation-oriented’. He helped devise studies done
by students trained by him, such as Kieron O’Connor.29 Noted American psy-
chologist Charles Spielberger also received special account #4 industry funding
in his attempt to replicate the twin-study and personality correlates reported
by Eysenck. The tobacco industry specifically commissioned Spielberger’s
study to head off any legal challenge that Eysenck’s findings might not apply in
29 ‘Hans J. Eysenck—Special Account #4 Project,’ 20 October 1993, http://tobaccodocu-
ments.org/pm/2025502365-2369.html; ‘CTR Special Projects,’ 2 February 1979, http://
tobaccodocuments.org/bliley-pm/20975.html. For some of the published results, see
Eysenck, H.J. and O’Connor, K. (1979), Smoking, arousal and personality, in
Electrophysiological effects of nicotine (ed. A. Redmond and C. Izard), pp. 147-–57, Elsevier,
Amsterdam.
09-Buchanan-Ch09.indd 37209-Buchanan-Ch09.indd 372 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

BIG TOBACCO, BIG MONEY
373
an American context.30 Special account #4 also helped finance Eysenck’s inci-
dental expenses. He was treated to numerous expenses-paid trips to speak at
conferences that were organized by the industry or on industry-relevant topics.
A considerable largesse was involved; subsidized Concorde air travel, for exam-
ple, was not out of the question.
Special account #4 also bankrolled Eysenck’s late 1970s survey on alternative
satisfactions sought by smokers who gave up, and later twin-study investiga-
tions of the relationship between heart disease, smoking, and stress. By the
early 1980s, Eysenck had also begun advising the industry on personality fac-
tors and psychological motivations behind smoking, the better to construct
and market their product. For instance, Eysenck suggested ways to position
various brands to appeal to introverts or extraverts, and how the characteris-
tics of these cigarettes could be matched to their target market.31
In return, Eysenck was to appear as a key witness in several court cases and
political hearings—including the 1983 ‘Waxman’ US congressional hearings.32
He consistently testified that the link between smoking and disease was still an
open question, a link possibly moderated by biogenetic factors. A number of
books and papers by Eysenck that utilized research supported by American
tobacco all helped promote this message.
Through most of this period Eysenck’s industry ‘handler’ was Ed Jacob, a
lawyer with a background in science. Jacob was a key legal strategist for
R.J. Reynolds, working at the New York law firm of Jacob, Medinger, and
Finnegan. Although Eysenck was supported by a range of industry sources, he
reported directly to Jacob. While Jacob helped steer CTR Special Projects, he
also played a key role in vetting proposals funded by the secret special accounts.
30 For an account of this work and Spielberger’s positive opinion of Eysenck’s research, see
Spielberger, C. (1986), Smoking, personality and health, in Hans Eysenck: consensus and
controversy (ed. S. Modgil and C. Modgil), pp. 305–15. Falmer Press, Philadelphia.
31 See for example, J.D. Weber, ‘Meeting with Dr. Hans Eysenck,’ 9 July 1982, http://tobac-
codocuments.org/rjr/505004727-4730.html. Eysenck had discussions with industry rep-
resentatives about brand image promotion for different personality types and how it
might relate to cigarette composition. For instance, he suggested that introverts needed a
stronger initial nicotine ‘hit’ that should decline with continued puffing. Conversely,
extraverts smoked to avoid boredom and thus needed to get an increasing ‘hit’ from each
successive puff.
32 See Hearings Before the Subcommittee on Health and the Environment of the Committee
on Energy and Commerce House of Representatives Part 2 of 3, 9 March 1983, http://
tobaccodocuments.org/lor/03637300-7630.html, and Eysenck’s, ‘Lung cancer, coronary
heart disease and smoking’, Statement Regarding S. 772 Submitted to the Committee
Labor and Human Resources, 6 May 1983.
09-Buchanan-Ch09.indd 37309-Buchanan-Ch09.indd 373 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
374
In this capacity Jacob seemed to play a significant role overseeing the research
Eysenck did. From notes of meetings it appeared Eysenck would present ideas
to Jacob and Jacob, in turn, would comment on their desirability from an
industry point of view. In these meetings, Eysenck demonstrates his keenness
to explore alternative factors (i.e. genetics and personality) that might explain
the smoking–cancer link. Here and in public, Eysenck maintained that nico-
tine was not a drug of dependence but one that met pre-existing needs exacer-
bated by environmental stresses.33 Although Eysenck generally disclosed his
financial sources for specific research, he did not detail this ‘special’ avenue of
tobacco industry funding and the level of consultative intimacy that went with it.
An intellectual cart and a legal horse
The first of Eysenck’s books that detailed CTR-sponsored research was
The causes and effects of smoking, published in 1980.34 The first half briefly
rehashed Smoking, health and personality. However, the detached tone of
Eysenck’s earlier book had given way to more tendentious argument. Eysenck
again pointed to the flaws in the epidemiological evidence. He noted the pro-
visos the Royal College of Physicians should have acknowledged in their 1977
report, despite the fact that this report gave serious attention to the constitu-
tional hypothesis. Eysenck focused almost exclusively on the evidence and
arguments that disputed the smoking–disease link, notably those of University
of Leeds’s Philip R.J. Burch, who also advanced a constitutional hypothesis, and
Harvard’s Carl C. Seltzer, who was also supported by CTR funds. Eysenck also
reviewed the evidence linking personality, smoking, and disease to develop a set
of hypotheses for the new research recounted in the latter part of the book.
The second part of The causes and effects of smoking presented twin data
analyses done with Lindon Eaves. This was serious scientific research, densely
packed with quantitative, biometric data. Although some attempt was made to
explain and summarize the methods and measures used, it was clearly aimed
at presenting sufficient data to convince fellow scientists, rather than a simpli-
fied review aimed at the educated lay-person. In so doing, Eysenck put more
on the line: not just his capacity to explain and critique the research of others
but also his reputation as a scientist capable of redirecting the cutting edge of
research. Eysenck and Eaves’s data seemed to suggest there was a significant
33 Edwin Jacob, ‘Notes of discussion with Dr. Eysenck—10/22/76,’ 11 November 1976,
http://tobaccodocuments.org/bliley_bw/521029803-9808.html.
34 Eysenck, H.J. (with Lindon Eaves) (1980c), The causes and effects of smoking, Maurice
Temple Smith, London.
09-Buchanan-Ch09.indd 37409-Buchanan-Ch09.indd 374 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

AN INTELLECTUAL CART AND A LEGAL HORSE
375
genetic factor in smoking uptake and maintenance, but left considerable room
for environmental factors as well. Moreover, Eysenck and Eaves had to admit
that previous assumptions about smoking and personality were not strongly
supported. There was little ‘evidence that neuroticism makes any overall con-
tribution to differences between smokers and non smokers’, while the role of
the extravert personality ‘does not emerge in our data.’35 Only the Eysenck
Personality Questionnaire’s P (psychoticism) and Lie scale scores seemed to
relate to smoking behaviour. Indeed, the addition of psychoticism as a dimen-
sion of personality appeared to ‘have led to the incorporation in P of some of
the elements of E [extraversion] which previously mediated the correlation
with smoking.’36 As George Hill noted, The causes and effects of smoking dem-
onstrated that personality and genetics might be implicated in smoking,
but ‘not strongly’. There was ‘nothing in this to skittle them down at the Royal
College…’37, Paul Kline agreed. The evidence in The causes and effects of
smoking left Eysenck’s hypothesis of a factor predisposing both smoking and
illness as viable, and that was all.38 While Eysenck had concluded that some
form of interaction of genes and environment was the most promising model
for future investigations, details that went beyond this sketch were still to be
filled in.
Other reviews of The causes and effects of smoking were generally more nega-
tive than those for his earlier work, the scientific and public health landscape
having changed considerably. The biochemical mechanisms underlying the
smoking–cancer link were still not entirely clear. Researchers had produced
skin carcinomas in mice and rats with isolated components of tobacco smoke—
even though they still had trouble agreeing on exposure methods and animal
models. Fearing the worst, the tobacco industry embarked on a concerted
attempt to discredit animal research, for it might be harder to argue against
than the statistical data. It was not until the 1990s that scientific opinion closed
around experimental evidence directly demonstrating human-type lung cancer
caused by whole tobacco smoke. However by 1980, most scientists with an
interest in the area believed the epidemiological evidence had clinched the
causal connection between smoking and disease long ago. In any case, the
public health risk was of such a magnitude that absolute proof was deemed
35 Ibid., p. 314.
36 Ibid.
37 Hill, G. (1980), Review of The causes and effects of smoking by Hans Eysenck (with Lindon
Eaves), The Times, 12 December 1980.
38 Kline, P. (1981), No smoking. Review of The causes and effects of smoking by Hans Eysenck
(with Lindon Eaves), London Review of Books, 19 February–4 March 1981.
09-Buchanan-Ch09.indd 37509-Buchanan-Ch09.indd 375 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
376
a luxury. Certainly medical authorities on both sides of the Atlantic saw the
evidence as sufficient to warrant action, and the public education campaigns
were gathering pace. For example, the British Medical Association and Ministry
of Health had condemned smoking since the mid-1960s, and the government
had progressively restricted tobacco advertising.39
Eysenck downplayed the animal studies in The causes and effects of smoking.
He also ignored many more recent epidemiological studies in the relatively
brief first section of the book. More revealingly, his point of view had shifted
from interested bystander to frank advocate, one more implicated in the
debate. With an aggressively anti-smoking message now an integral part of the
public health agenda, Eysenck now had the look of a quixotic figure tilting at
the windmills of orthodoxy. The net effect of The causes and effects of smoking
was to redouble the negative chorus from health spokespeople. For example,
the passionate Richard Peto gate-crashed the press conference for Eysenck’s
book, chiding him for ignoring contradictory evidence, including the impor-
tant follow-up review he and Sir Richard Doll had published in 1976.40 Eysenck
explained uncertainly that it was too late at the time of publication to include
this evidence and that he had included very recent studies favourable to his
point of view only because he had received pre-prints. According to tobacco
industry observers, those present found this explanation unconvincing; in fact,
it was greeted with howls of laughter. Yet Peto went on to plead with the media
to not report on Eysenck’s ‘heresy’, an invocation for suppression that the New
Scientist labelled ‘injudicious’ and ‘unworthy’.41 On a subsequent BBC radio
talk show, Peto continued to attack Eysenck, with Eysenck on the defensive:
‘I have no doubt smoking is not a healthy habit. All I am saying is … that I am
not completely convinced by the evidence … [of] causal factors to the extent
that is suggested.’42 Eysenck might be seen to be rhetorically retreating but his
39 No form of public health policy action directed at the tobacco companies came into play
in Britain until the mid-1980s. See British Medical Association, Public Affairs Division
(1986), Smoking out the barons: the campaign against the tobacco industry, Wiley,
Chichester.
40 See Doll, R. and Peto, R. (1976), Mortality in relation to smoking: 20 years’ observations
on doctors, British Medical Journal 2 (6051), 1525–36.
41 New Scientist (1980), Smoking out censorship [editorial], New Scientist 88 (18/25
December 1980), 756.
42 Quoted in Susan von Hoffman to Joan Mebane, 3 March 1981, http://tobaccodocuments.
org/ness/1855.html, p. 2.
09-Buchanan-Ch09.indd 37609-Buchanan-Ch09.indd 376 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

AN INTELLECTUAL CART AND A LEGAL HORSE
377
high profile on the airwaves still helped to promote the idea of disagreement
between experts—just the sort of thing Peto feared.43
Eysenck can’t have enjoyed being the butt of jokes. He was now seen as quite
out of step with the public health lobby—those he likened to a ‘mafia’ spread-
ing ‘propaganda’—and was now very visibly locked into a position that ques-
tioned the smoking–disease link.44 However, he lacked good counterevidence.
Support for his alternative genetics-and-personality hypothesis was hardly
overwhelming, undermined by merely suggestive or disappointing results.
By the early 1980s, Eysenck was emphasizing the notion of environmental
stress as well as personality in the maintenance of the smoking habit, also
pointing to other research linking stress with the aetiology of cancer.45 For
Eysenck, smoking was a function of boredom and emotional strain. With their
lower level of cortical arousal, extraverts smoked to spice up what for them
were non-stimulating environments. In contrast, neurotics were characterized
by higher levels of emotional lability, and they tended to smoke to cope with
what for them were stressful circumstances. Finally, those high in P had their
smoking reinforced by a need to express non-conforming social tendencies.
The model was complicated by evidence of the biphasic properties of nicotine:
while small amounts of nicotine raised cortical arousal, large amounts tended
to suppress autonomic nervous system activity. Thus (somewhat neurotic)
introverts might smoke heavily to calm themselves and reduce social anxiety.
Nevertheless, there was still much work to be done and Eysenck was nearing
the end of his professional career. Retirement loomed; the connections and
privileges of a professorship were about to disappear. Already marginalized by
the race and IQ controversy, access to research funds would only become more
difficult. Moreover, lengthy and large-scale longitudinal research must have
seemed impractical at his age.
Eysenck had apparently heard of the Yugoslav researcher with longitudinal
evidence linking personality and disease soon after the publication of
The causes and effects of smoking in 1980. At this stage, Grossarth-Maticek was
43 On another occasion, Eysenck came in for criticism from David Simpson, director of
anti-smoking pressure group ASH, that almost amounted to ridicule: ‘Frankly he
[Eysenck] ought to turn to geography next. By missing out similar research there he could
prove the earth is flat. It is honestly on that sort of level when we deal with this book.’BBC
Radio I, ‘Newsbeat’, 11 December 1980.
44 See, for example, Fletcher, C. (1981), Plea for the guilty, Times Higher Education
Supplement, 16 January 1981 and Jill Turner, ‘Mind if I Smoke?’ Books, 1 January 1981.
45 Eysenck, H.J. (1984c), Stress, personality and smoking behaviour, In Stress and anxiety,
Vol. 9 (ed. C.D. Spielberger et al.), pp. 37–49. Hemisphere, Washington.
09-Buchanan-Ch09.indd 37709-Buchanan-Ch09.indd 377 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
378
largely unknown in the English-speaking world. Within European scientific
circles, especially the medical and psychological fields in Germany and the
Netherlands, Grossarth-Maticek had an ambiguous reputation. A few thought
him a visionary, but many distrusted him. According to Eysenck, he found a
man at ‘the end of his tether’—a dedicated and charismatic researcher with
radical ideas and the courage to explore them.46 Empathizing with Grossarth-
Maticek’s status as a scientific heretic, Eysenck portrays himself as duty bound
to assist a man who through no fault of his own had run out of friends and
money.
The Heidelberg connection
Dr. Grossarth-Maticek is a gifted and creative researcher... Rarely have I been so
impressed by a young research investigator in this difficult field. (Dr George F.
Solomon, Professor of Psychiatry, UCSF and UCLA Schools of Medicine, 15 November
1976—Letter of reference for Ronald Grossarth-Maticek)
Today Ronald Grossarth-Maticek performs consultative work and research for
pharmaceutical companies and is affiliated with the European Centre for Peace
and Development, part of the University of Peace established by the UN in
Costa Rica. Living in a fine house overlooking Heidelberg Castle in Germany
on the steep embankments of the Neckar river, he has come a long way from
the town of Sobor in war-torn former Yugoslavia where he grew up. Precise
details of Grossarth-Maticek’s qualifications and work history are hard to pin
down, however, and still subject to dispute. He was born in 1940 and educated
at the University of Belgrade. Grossarth-Maticek received a D.Phil. in sociol-
ogy from the University of Heidelberg in 1973 with a thesis dealing with stu-
dent behaviour.47 Psychology was the second discipline in his Rigorosum, the
oral defence made to faculty. It was not clear whether he had extensive formal
training in psychotherapy or much experience in the construction of personal-
ity questionnaires. However, he soon began working in the field of psychoso-
matic epidemiology, a notoriously difficult new field.
The relationship between affective states and physical disorders had been
debated since antiquity. Integrating mind into medicine had particular appeal
in the US, however, where psychosomatic medicine developed as a distinct
entity mid-century that bordered on internal medicine, psychiatry, and
psychology. Inspired by psychoanalysis, key figures such as Helen Flanders
46 Eysenck (1991c), Reply to criticisms of the Grossarth–Maticek studies, p. 298.
47 These details come courtesy of the CV Grossarth-Maticek supplied, translated for me by
Hermann Vetter.
09-Buchanan-Ch09.indd 37809-Buchanan-Ch09.indd 378 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

THE HEIDELBERG CONNECTION
379
Dunbar and Franz Alexander investigated the bodily effects of emotional reac-
tions and the chain linking psychological stimulus and organic end-result.
Greater medical specialization in the 1960s and 1970s tended to marginalize
attention to mind–body interactions. Internal medicine opted for a strictly
material, reductionist approach, while psychiatrists pursued non-analytic psy-
chotherapies, drug regimes, and neurobiological research. Psychosomatic
research was increasingly performed by psychologists, who shifted the focus
from the exploration of unconscious conflict to psychometric studies of mala-
daptive behaviour patterns. This approach was extended to the psychological
concomitants of ailments like heart disease, hypertension, and stroke, as well
as cancer—still seen as essentially somatic conditions, notwithstanding their
complex aetiologies.48 And, following the example of chronic disease epidemi-
ology, one way of getting a handle on these psychological concomitants was
through longitudinal surveys.
After Grossarth-Maticek moved to Germany in the 1970s, he was employed
as a research assistant in the Institute of Social Medicine for a short time and
then worked as a private scholar.49 Most significantly, Grossarth-Maticek initi-
ated a very ambitious prospective longitudinal research programme in 1973.
In the spring of 1977, he turned up at the University of Heidelberg psychology
department to present a 100-page manuscript for the purpose of his
Habilitation—a German degree equivalent to a second doctoral title and a
prerequisite for a full professorship. When he was informally told by the review
panel of Professors Amelang, Weinert, and Wottawa that his Habilitation
application would be refused, he requested the manuscripts be returned, citing
a need to distribute them to other interested parties. One member of the panel,
Manfred Amelang, refused to give his up because he had made comments on
his copy.50
According to Amelang, the document was rejected largely because the claims
made were so extraordinary. They seemed ‘too good to be true’—a judgement
that has dogged Grossarth-Maticek ever since. Written in German and
prepared in his own institute, the manuscript detailed a prospective study
48 See Theodore M. Brown’s general account ‘The rise and fall of American psychosomatic
medicine,’ at http://www.human-nature.com/free-associations/riseandfall.html.
49 Even these simple details are difficult to definitively ascertain. According to Manfred
Amelang, the University of Heidelberg wrote to him on 3 December 2004 to say they had
no record of Grossarth-Maticek’s employment as a research assistant in this period or any
other. Grossarth-Maticek may have been employed privately by Professor Blohmke.
Manfred Amelang, personal communication, 15 January 2008.
50 Manfred Amelang, interview, 1 February 2003.
09-Buchanan-Ch09.indd 37909-Buchanan-Ch09.indd 379 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
380
undertaken in Crvenka, Yugoslavia from 1965. Using a sample of 1353 people,
predictions of future mortality were made with the aid of the 88-item Ronald
Grossarth-Maticek (RGM) questionnaire. The questionnaire measured factors
like reactions to stress and the inhibited expression of psychological needs.
Based on a series of analyses, Grossarth-Maticek was able to predict 38 cases of
fatal cancer. When follow-up data were obtained ten years later, 37 of 38 had
indeed died of cancer, a strike rate of 97.3%, ‘even 100% if suspected cancer
diagnosis is included.’51 The strike rate for coronary heart failure was also a
remarkably high 92.1%. Moreover, the RGM questionnaire used an unusual
system of differential weighting to achieve this predictive success, with indi-
vidual item weights varying from 1 to 100. No other psychological measures
were mentioned in the Habilitation document.
The review panel at Heidelberg was left in a quandary as to what to make of
Grossarth-Maticek. Apart from the stunning results, they were concerned
about an insufficient separation of the predictor and follow-up data. Knowledge
of predictions might have in one way or another influenced the gathering of
mortality data (or even vice-versa). The following year, the panel advised
Grossarth-Maticek to deposit the predictor scores of his ongoing Heidelberg
studies in order to head off the criticism of criterion contamination. Grossarth-
Maticek did so, registering names and a subject coding list, but not before the
first follow-up had been done. Some but not all predictor data were also depos-
ited. However, there was no complete, official registration of the personality
predictor scores.
In the late 1970s and early 1980s, Grossarth-Maticek struggled to get English-
language versions of his ideas and research in peer-reviewed journals.
Nevertheless, he appeared at various conferences outside Germany and started
to become better known in the Netherlands, the UK, and the US. Various aspects
of the Yugoslav study were published in the early 1980s; none detailed the typo-
logical approach that would characterize his later publications with Hans
Eysenck. One of the first English-language publications on the Yugoslav study
was in the relatively obscure journal Psychotherapy and Psychosomatics. The
focus was on two major psychological factors, rationality and anti-emotionality,
51 Grossarth-Maticek, R. (1977), Social scientific aspects in the aetiology of organic diseases:
perspective, method, and results of a prospective study, Habilitation thesis, University of
Heidelberg. [Translated by Manfred Amelang, University of Heidelberg, 1977. Original
German title: Veröffentlichung im Rahmen des interdisziplinären Forschungsprojektes
sozialwissenschaftlicher Onkologie.], p. 23.
09-Buchanan-Ch09.indd 38009-Buchanan-Ch09.indd 380 1/16/2010 9:45:33 AM1/16/2010 9:45:33 AM

RECONCILING THE DIFFERENCES AND DOUBTS
381
and life events causing chronic hopelessness and depression.52 Using the RGM
questionnaire (that had grown to 109 items), it was reported that high ration-
ality and anti-emotionality scores predicted coronary heart disease, stroke, and
cancer, while life events leading to hopelessness and depression increased the
risk of cancer.
Some time in the early 1980s Eysenck became interested in Grossarth-
Maticek’s work, having read the 1980 Psychotherapy and Psychosomatics piece.
According to Grossarth-Maticek, Eysenck was intrigued and excited by his
work. He had immediately phoned Grossarth-Maticek, and promptly turned
up in Heidelberg with the express view of joining forces. However, there is no
concrete evidence that Eysenck became involved with Grossarth-Maticek
before 1984.
The reasons behind Eysenck’s interest in this work were obvious enough,
given that Grossarth-Maticek’s work appeared to reinforce his own ideas in the
area of personality and disease. Grossarth-Maticek had demonstrated that
easily measured individual differences in personality were perhaps the primary
variable, combining and interacting with other risk factors (e.g. smoking,
drinking, and diet) to determine long-term disease propensities. These pro-
spective studies made more detailed and specific the kind of speculation and
arguments he had long engaged in, and then some. According to Grossarth-
Maticek, it also pleased Eysenck greatly to be able to work with and admire
someone in Germany, contrasting with the distant relationship he had experi-
enced with his former Nazi father.53
Reconciling the differences and doubts
When interviewed about their work together, Grossarth-Maticek was at pains
to stress that this was not a blind embrace; Eysenck came in with his eyes open
to the scepticism Grossarth-Maticek’s work had aroused. Their collaboration
therefore had a critical condition, Grossarth-Maticek recalled. Eysenck
demanded: ‘You must let me check your data, for if you deceive me I will never
forgive you.’54 To this end, Eysenck proposed ‘three steps to trust’. The first
was to check mortality data for some extreme cases, such as when a whole
family was recording as having died of cancer, by visiting homes of next of kin.
The second was to match Grossarth-Maticek’s mortality data with that of the
52 Grossarth-Maticek, R. (1980a), Psychosocial predictors of cancer and internal diseases:
an overview, Psychotherapy and Psychosomatics 33, 122–8.
53 Ronald Grossarth-Maticek, interview, 31 January-1 February 2003.
54 Ibid.
09-Buchanan-Ch09.indd 38109-Buchanan-Ch09.indd 381 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
382
City of Heidelberg records. The final check was to compare the subject name
and code assignment lists that had been deposited with the mayor of Heidelberg
in 1978 with the data files used in subsequent analyses.55
There were also some quite significant differences in the backgrounds and
intellectual world-views of Eysenck and Grossarth-Maticek that had to be rec-
onciled, minimized, or glossed over. Grossarth-Maticek’s ideas grew out of a
European psychodynamic tradition. While his views on the role of personality
in the genesis or promotion of disease were not dissimilar to Eysenck’s, they
were not wholly in accord as to what kind of personality characteristics medi-
ated which particular diseases. For example, Eysenck and Kissen’s work had
suggested a personality profile of cancer sufferers as low in neuroticism and
the expression of emotion. Grossarth-Maticek’s cancer type presented a very
different picture, with those expressing feeling of hopelessness and depression
being more prone to cancer, and more neurotic, if anything. However, the
uncertain and contradictory results Eysenck had obtained on this score prob-
ably made him receptive to a new theoretical understanding of relationship he
was still sure existed. As he said to others when questioned about his uncritical
support for Grossarth-Maticek: ‘I like the data, and I like the theory.’56
According to Grossarth-Maticek, Eysenck had very little say over the formu-
lation of personality type concepts. Indeed, Eysenck had struggled to banish
typological concepts in favour of continuous dimensions for most of his career.
However, the kind of synergistic thinking that dominates the Crvenka and
Heidelberg projects, wherein risk factors combine to greater effect than in
isolation, clearly predated Eysenck’s involvement. Nevertheless, Eysenck
emphasized quasi-experimental intervention as a way of teasing out causality;
he pushed the empirical aspects of the project at the cost of minimizing its
theoretical underpinnings; and, he fine-tuned and polished the presentation of
the work so that some of the more glaring methodological mistakes (and the more
outrageous claims said the critics) no longer appeared. Eysenck was even able to
adjust to and adjust Grossarth-Maticek’s approach to psychotherapy. Grossarth-
Maticek initially termed it ‘social psychotherapy’ and later changed this to
‘creative novation (behaviour) therapy’.57 For Eysenck, Grossarth-Maticek’s
55 Ibid.
56 Henk van der Ploeg, interview, 11 February 2003.
57 Grossarth-Maticek, R. (1980b), Social psychotherapy and course of the disease: first expe-
riences with cancer patients, Psychotherapy and Psychosomatics 33, 129–38.
09-Buchanan-Ch09.indd 38209-Buchanan-Ch09.indd 382 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

BIG TOBACCO HEDGES ITS BETS
383
approach was acceptable insofar as it was a reformulation of psychoanalytic
ideas in ‘meaningful learning terms’.58
It was a curious collaboration, by turns appearing one-sided, team-based, or
that of mentor–student. It must be remembered that Eysenck played no role in
the initiation of these studies, nor had much influence over the process of most
of the data gathering. During the peak of their collaboration in the late 1980s,
Eysenck would visit Heidelberg for a few days, several times a year. In Grossarth-
Maticek’s telling of the story, Grossarth-Maticek would do most of the talking,
expounding on his ideas and the data supporting them. Eysenck would listen
closely, asking lots of critical questions until he could finally satisfy himself
that things were sufficiently clear. It was not just a matter of taking and trans-
lating ready-made results, however. Eysenck would suggest analyses given the
data that existed and even suggested that certain variables be more systemati-
cally explored. There was ample opportunity to select, tease out, or redirect
attention—given a data set that was apparently sprawling and chaotic, but also
rich and ambitious. It included many personality and physiological measures
taken over a long period, accompanied by much personal information. Some
of the analyses were done by Eysenck’s statisticians in London and some by
statistical consultant Hermann Vetter in Heidelberg. However, from the mid-
1980s, Eysenck did virtually all the writing for publication in English and
presumably exerted a strong editorial control.
Eysenck first co-authored an article with Grossarth-Maticek on the Crvenka
data in 1985, accompanied by a piece rehashing his doubts about orthodox
theories linking smoking and disease. In the years to follow, a bewildering set
of studies and measures was progressively brought forth and published. Yet at
no stage were these reports greeted with widespread acceptance and acclaim.
Big Tobacco hedges its bets
Tobacco industry scientists had been well aware of Grossarth-Maticek’s work.
It would be no exaggeration to say that many were extremely sceptical. For
those surprised that the cigarette companies did not embrace him, one could
emphasize that there were risks for them, risks magnified by their pariah status.
Not only were they wary of the charge of buying results and advocates, they
also appeared to value only research that was not ‘attackable’. So even the
tobacco industry had stayed clear of Grossarth-Maticek for the most part,
despite the seemingly favourable message his work conveyed. R.J. Reynolds’
chief scientist, Frank Colby, had such a low opinion of this work that he had
58 Ronald Grossarth-Maticek, interview, 31 January-1 February 2003.
09-Buchanan-Ch09.indd 38309-Buchanan-Ch09.indd 383 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
384
made no attempt to meet Grossarth-Maticek. When he finally did in 1980,
he came away more sceptical than ever, ‘unsure whether to feel sorry for him
or completely distrust him... I definitely recommend against any involvement
between Dr. Grossarth-Maticek and us, or the industry in general.’59 Scientists
associated with Phillip Morris, Tom Osdene and William L. Dunn, apparently
shared similar opinions.60
At first, only the Verband der Deutschen Ziagarettenindustrie—a German
body roughly equivalent to the CTR—offered Grossarth-Maticek some assist-
ance. However, by the early 1980s, they were reviewing their funding arrange-
ments and only agreed to support a further study with the assurance that
Professor Jan Bastiaans was involved. Bastiaans was a somewhat controversial
Dutch psychiatrist, famously treating holocaust survivors with unorthodox
forms of psychotherapy and LSD. At the time he was probably Grossarth-
Maticek’s strongest supporter outside Germany. Internal Verband documents
from 1984 reveal that industry representatives were especially concerned about
the integrity of his data.61 They allude to unorthodox methodological prac-
tices, such as minimizing missing data by filling in the gaps in a subject’s record
by matching them with an otherwise similar subject.62 The results of this prac-
tice were spotted in 1993 by Bastiaans’ former research assistant Henk van der
Ploeg, who later added that this was ‘bad news for trustworthiness of the
data’.63
It appears that Eysenck’s prestige and authority, and his keenness to collabo-
rate with Grossarth-Maticek was enough to paper over the doubts and bring
R.J. Reynolds on board in 1985. Still, they had to be dragged kicking and
screaming to fund their collaborative efforts. Those within the tobacco industry
59 Frank Colby, Report on a trip to Germany, 1 May 1980, http://tobaccodocuments.org/
rjr/503794792-4795.html, p. 14.
60 See W.L. Dunn to T.S. Osdene, The Grossarth-Maticek paper on the psychosomatics of
cancer, 2 February 1982, http://tobaccodocuments.org/pm/1000039191-9193.html.
61 See Minutes of VDC Scientific Committee Meeting, 17 October 1984, http://tobacco-
documents.org/pm/2023539765-9768.html, p.4.
62 See for example, J.H. Robinson to A. Wallace Hayes, Articles authored by R. Grossarth-
Maticek,’ 12 December 1985, http://tobaccodocuments.org/rjr/504231829-1830.html.
63 For van der Ploeg’s critiques (authored in conjunction with Hermann Vetter and Wim
Chr. Kleijn) and Eysenck’s reply (with contributions from Grossarth-Maticek), see van
der Ploeg, H. and Vetter, H. (1993), Two for the price of one: the empirical basis of the
Grossarth-Maticek interviews, Psychological Inquiry 4, 65–6 [quote on p. 66]; van der
Ploeg, H. and Kleijn, W.C. (1993), Some further doubts about Grossarth-Maticek’s data
base, Psychological Inquiry 4, 68–9; and Eysenck, H.J. (1993b), Reply to van der Ploeg,
Vetter and Kleijn, Psychological Inquiry 4, 70–3.
09-Buchanan-Ch09.indd 38409-Buchanan-Ch09.indd 384 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

THE EYSENCK INFLUENCE
385
could not help notice the inconsistency of Eysenck’s position. When asked
to comment on the two 1985 articles on smoking, personality, and health,
J.H. Robinson and D.G. Gilbert at R.J. Reynolds commended Eysenck’s clarity
in pointing to the flaws in studies that linked smoking and cancer, and admit-
ted the attraction of a theory relating personality to cancer. Yet they were
‘disturbed by Dr. Eysenck’s total acceptance’ of Grossarth-Maticek’s data.
‘While the author [Eysenck] spends about 80% of his time stressing the pitfalls
of assuming causation based on correlational data, he seems to make this leap
in logic himself in several places.’64
Running almost parallel to Eysenck’s involvement with Grossarth-Maticek
was that of a re-analysis project. The tobacco companies agreed to support
Eysenck only if a ‘rival’ team was set up to check Grossarth-Maticek’s data set
and the conclusions drawn from it.65 Clearly, industry representatives were
covering themselves. Grossarth-Maticek was obliged to share data and furnish
details of his methods and the analyses. Led by Charles Spielberger, who had
been introduced to Grossarth-Maticek by Eysenck, the team also included the
more critical van der Ploeg and Boston University psychologist Bernard Fox.
While Grossarth-Maticek saw this move as an act of bad faith, he cooperated
with the re-analysis team (albeit rather reluctantly according to van der Ploeg.)
However, the critical reports of various members of the re-analysis team made
sure that this tobacco industry patronage was relatively short-lived.
The Eysenck influence
Once Hans Eysenck came on board, the direction of Grossarth-Maticek’s
research appeared to shift appreciably. A 1987 article in the European Journal
of Psychiatry introduced English-speaking audiences to their notion of psycho-
logical types.66 Even though it centred on the Crvenka data, Eysenck was,
oddly, the sole author. Four types were outlined: (1) understimulation;
(2) overarousral; (3) ambivalence; and, (4) personal autonomy. Gone was the
prediction of disease from (weighted) item scale scores that characterized
64 J.H. Robinson and D.G. Gilbert to A. Wallace Hayes, Interoffice memo, 10 July 1985,
http://tobaccodocuments.org/rjr/504226599-6600.html, p.2.
65 See Charles D. Spielberger and Henk van der Ploeg, Proposal for further review and
evaluation of the Grossarth-Maticek studies, 6 January 1986, http://tobaccodocuments.
org/rjr/515805043-5047.html.
66 Eysenck, H.J. (1987b), Personality as a predictor of cancer and cardiovascular disease, and
the application of behaviour therapy in prophylaxis, European Journal of Psychiatry 1,
29–41. Grossarth-Maticek had published his typological theories in German the previous
year.
09-Buchanan-Ch09.indd 38509-Buchanan-Ch09.indd 385 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
386
Grossarth-Maticek’s first write-ups of the Crvenka data, despite the unusually
successful results achieved. Gone also was any discussion of the concept of
rationality and anti-emotionality, and the importance of depression and hope-
lessness tended to be de-emphasized in favour of personal autonomy. However,
the type-approach managed to bring Grossarth-Maticek’s ideas in line with
contemporary psychosomatic work, especially the cognitive–behavioural
typologies emanating from the US. Grossarth-Maticek’s type 1 roughly
matched up with the type C personality identified by Lydia Temoshok as
cancer-prone. His type 2 corresponded with the type A personality that Meyer
Friedman, Ray Rosenman, and David Jenkins had earmarked as prone to heart
disease. Sure enough, follow-up mortality analyses revealed that those with
type 1 and 2 personalities had a far greater likelihood of dying of cancer or
coronary heart disease, respectively. The inventory used to assess type mem-
bership was not published till 1988, and it appeared that it functioned as the
basis for structured interviews. Items were very wordy and cumbersome, which
often necessitated explanations from interviewers. Groups of 11 or more items
formed subscales for each type. They were yes/no in format, with the highest
subscale score (augmented by interviewer judgement) typically used to assign
subjects to type.67
An expanded six-type approach was used to analyse the more extensive
Heidelberg data. In the six-type approach, types 1 and 2 were again conceptu-
alized as cancer- and heart disease-prone, type 3 was linked with psychopathic
behaviour, and type 4 was regarded as healthy and autonomous. Eysenck and
Grossarth-Maticek added two new types: type 5 was characterized by (the
return of) rationality and anti-emotionality, and type 6 by anti-social behav-
iour and drug addiction. Again strong findings in relation to predicted mortal-
ity were put forward, backing up the pivotal role personality had in the course
of physical disease and destructive behaviour. The inventory that assessed
these six types was not published until 1990, despite first being used in 1974
(or 1973).68 It contained 182 yes/no items that were far more direct and suc-
cinct than the four-type inventory, with more sophisticated rules for assigning
type membership. Eysenck and Grossarth-Maticek also claimed the need for a
more satisfactory psychometric device prompted the construction of the six
67 Grossarth-Maticek, R., Eysenck, H.J., and Vetter, H. (1988), Personality type, smoking
habit and their interactions as predictors of cancer and coronary heart disease, Personality
and Individual Differences 9, 479–95.
68 Grossarth-Maticek, R. and Eysenck, H.J. (1990), Personality, stress and disease: descrip-
tion and validation of a new inventory, Psychological Reports 66, 355–73.
09-Buchanan-Ch09.indd 38609-Buchanan-Ch09.indd 386 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

1991: A SCEPTICAL PEAK
387
type inventory in the early 1970s—even though it appeared Grossarth-Maticek
had only limited feedback on his earlier measures at that time.
However, not all were convinced by the data and the conclusions drawn.
Re-analysis group member Bernard Fox was one of the first to go public with
his doubts in 1988, stating that he could not believe in the validity of the
Crvenka data he had been shown.69 Others in the Spielberger-led group, such
as Henk van der Ploeg, also began making disbelieving noises, as did medical
epidemiologist and statistician P.N. Lee in his industry-commissioned analy-
ses of the data. This was enough for the tobacco companies; in 1988, they took
the decision to cease sponsoring Eysenck’s work with Grossarth-Maticek,
citing a need to shift research directions.70 While this was not the end of
Eysenck’s relationship with the American tobacco industry, it appeared to be
the end of the industry’s dealings with Grossarth-Maticek.
1991: a sceptical peak
Eysenck was unbowed by the flak his collaboration with Grossarth-Maticek
had begun to attract and completed another short book on smoking and health
in 1991.71 Again he dismissed the epidemiological evidence, suggesting that by
itself smoking was a relatively weak risk factor. Only when combined with the
more potent effects of unhealthy personalities and life stress did it become dan-
gerous. For Eysenck it followed that only those with high-risk personalities—
perhaps a quarter and no more than a third of the population—should be
pinpointed for interventions that changed their lifestyle and habits. Those with
healthy personalities could puff away with relative impunity. Eysenck’s empha-
sis on the combined effects of personality and stress allowed him to add a new
line of attack on the insidious anti-smoking ‘mafia’, one that he had first
suggested in 1989.72 He argued the public health campaigns were unlikely to
have a positive net effect. Even if they helped reduce smoking, they might
69 Fox, B.H. (1988),Psychogenic factors in cancer, especially its incidence, in Topics in health
psychology (ed. S. Maes et al.), pp. 37–55. Wiley, Chichester.
70 A. Wallace Hayes to Hans J. Eysenck, 28 January 1988, http://tobaccodocuments.org/
rjr/509882911-2911.html.
71 Eysenck, H.J. (1991g). Smoking, personality and stress: psychosocial factors in the prevention
of cancer and coronary heart disease. Springer-Verlag, New York.
72 Grossarth-Maticek, R. and Eysenck, H.J. (1989), Is media information that smoking
causes illness a self-fulfilling prophecy? Psychological Reports 65, 177–8. See also Eysenck,
H.J. (1986c), Consensus and controversy: two types of science, in Hans Eysenck: consensus
and controversy (ed. S. Modgil and C. Modgil), pp. 375–98, Falmer Press, Philadelphia.
09-Buchanan-Ch09.indd 38709-Buchanan-Ch09.indd 387 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
388
unnecessarily stress those who didn’t manage to quit with potentially fatal
consequences.
In contrast, Eysenck claimed the best way to lower cancer and heart disease
rates was by giving those at risk a brief course of psychotherapy that targeted
personality change and stress reduction. Here and in other articles he pro-
duced data that apparently demonstrated this point. Grossarth-Maticek’s
‘creative novation behaviour therapy’ was able to markedly improve the life
expectancy of those with cancer- or heart disease-prone personalities, and
even greatly improve the prospects of those already suffering from these dis-
eases.73 Even an abbreviated form of ‘bibliotherapy’, which amounted to writ-
ten material accompanied by a short session of explanation and coaching,
could produce excellent results. Other data from Eysenck and Grossarth-
Maticek suggested that those identified with unhealthy personalities could,
even by their own efforts, move their personality styles in the healthy direction.
When they did so, they significantly improved their long-term health
outlook.
These were remarkable results—given that the targets for change were rela-
tively enduring behavioural characteristics, outcomes of genetic predisposi-
tions. Eysenck and Grossarth-Maticek’s use of psychotherapy had taken up a
promising new research line in cancer care, largely initiated by Stanford’s
David Spiegel. Using a controlled intervention design, Spiegel’s 1989 Lancet
article suggested psychotherapy could have strong positive effects for those
suffering from metastatic breast cancer.74 Given the attention and hope this
work generated, it was hardly surprising that many other researchers jumped
aboard and set similar studies in train. What made Eysenck and Grossarth-
Maticek stand out, though, was how quickly they were able to do likewise.
Reviewers of Eysenck’s 1991 book opined that, if such work could be inde-
pendently replicated, ‘it would cause us to revise substantially all our current
notions of personality, disease and the relation of mind to body. The results
73 Therapy intervention studies had been outlined in a number of previous articles by
Grossarth-Maticek. However, in 1988 this topic was put front and centre with a number
of pieces by Eysenck et al. in various forums. See for example, Eysenck, H.J. (1988a),
Personality as a predictor of cancer and cardiovascular disease, and the application of
behaviour therapy in prophylaxis, British Journal of Clinical and Social Psychiatry 6, 4–12.
A series of articles in 1991 in Eysenck’s high-impact journal BRAT gave them extensive,
prominent coverage. For a start, see Grossarth-Maticek, R. and Eysenck, H.J. (1991),
Creative novation behaviour therapy as a prophylactic treatment for cancer and coronary
heart disease. I. Description of treatment, Behaviour Research and Therapy 29, 1–16.
74 Spiegel, D., Bloom, J.R., Kraemer, H.C., and Gottheil, E. (1989), Effect of psychosocial
treatment on survival of patients with metastatic breast cancer, The Lancet 334, 888–901.
09-Buchanan-Ch09.indd 38809-Buchanan-Ch09.indd 388 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

1991: A SCEPTICAL PEAK
389
he reports are that extraordinary.’75 It is impossible not to discern an under-
current of extreme scepticism here; however, only those closest to the work
tended to articulate the reasons for their doubts explicitly.
This scepticism reached a crescendo that year in a special issue of Psychological
Inquiry. Eysenck presented a target article outlining results of prospective
studies in Heidelberg (and Crvenka), the predictive power of the four- and six-
type approach, and the effectiveness of psychotherapeutic intervention.76 Over
a dozen senior researchers were invited to comment, including leading figures
Temoshok and Spiegel. Almost all were critical or extremely critical. Yet it is
extremely difficult to summarize their critiques succinctly. Like the proverbial
blind men feeling the elephant, they each focused on different parts of the
sprawling project that had been put in front of them.
Nevertheless, all these researchers seemed to concur on a number of points.
All agreed these were stunning results. As P.N. Lee commented, the effect sizes
were ‘absolutely mammoth... so outside my experience as an epidemiologist
that I find it very difficult indeed to accept them as real.’77 Estimates of relative
risk conferred by personality type membership for particular fatal disease were
all high to very high, depending on what data was being analysed and how. For
example, Bernard Fox’s analysis of the Crvenka data put the relative risk for
cancer-prone personalities at 23.7, one that surely dwarfed all known other
factors like smoking or poor diet.78
Various respondents pointed to insufficient data controls, including the
possibility that prior knowledge of predictions might have compromised the
independence and reliability of the mortality data. Lydia Temoshok went fur-
ther in suggesting a post hoc matching of personality prediction and mortality
criteria had occurred, adding that Eysenck’s influence must have been behind
the latter shift toward the kind a typological thinking that partly borrowed
from her work. Fox pointed to unlikely characteristics of the data, including
analyses that showed validity exceeding reliability and 113% success rates.
75 Baker, T.B. and Fiore, M.C. (1992), Elvis is alive, the mafia killed JFK, and smoking is
good for you, Contemporary Psychology 37, 1014–16. Even Eysenck struggled to incorpo-
rate these findings within his, by now, well-articulated biosocial theory of personality.
See Eysenck (1991c), Reply to criticisms of the Grossarth–Maticek studies and Eysenck
(1986c), Consensus and controversy: two types of science.
76 Eysenck, H.J. (1991b), Personality, stress and disease: an interactionist perspective,
Psychological Inquiry 2, 221–32.
77 Lee, P.N. (1991), Personality and disease: a call for replication, Psychological Inquiry 2,
251–3.
78 Fox, B.H. (1991), Quandaries created by the unlikely numbers in some of Grossarth-
Maticek’s studies, Psychological Inquiry 2, 242–7.
09-Buchanan-Ch09.indd 38909-Buchanan-Ch09.indd 389 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
390
P.N. Lee did likewise with other aspects of the data, while Amelang related his
personal dealings with Grossarth-Maticek dating back to his initial negative
appraisal of Grossarth-Maticek’s Habilitation document. More worrying still,
Grossarth-Maticek’s statistician, Hermann Vetter, outlined his suspicions that
some sort of fudging was involved in the figures he was being fed.79
Perhaps the most eye-popping critique came from van der Ploeg.80
He recounted a bizarre episode of subject record substitution, fake names, and
research assistant blaming. When sent a set of subject records for analysis, van
der Ploeg found they lacked any predictive validity. When Grossarth-Maticek
was informed, he claimed a mistake had occurred. Mortality and personality
data were matched with a set of (deliberately) fictitious subject names that
were nonetheless crossed out to respect Germany’s privacy laws. Attempts to
match the correct name lists with the correct data went back and forth. Some
subjects still appeared to match up with lists they shouldn’t, while others
seemed to have ‘died’ several times from different illnesses. The matter has
never been fully explained, and the conclusions drawn differ depending on
whom you ask. For Grossarth-Maticek, it was an excusable mix-up by an
(unnamed) assistant that was finally resolved with a trip to the former city
mayor’s house to check the subject lists deposited there. Eysenck apparently
concurred. For van der Ploeg and for Amelang, the trip settled nothing.81
Dr Rainer Frentzel-Beyme, an epidemiologist at the German Cancer Research
Centre, was virtually alone amongst those in Psychological Inquiry in defending
Grossarth-Maticek. A co-author on several Grossarth-Maticek papers,
Frentzel-Beyme argued that in epidemiological research ‘every insider knows
that no satisfactory, flawless and impeccable large studies’ exist. The critics
carp over errors and ‘faulty’ data because ‘the whole direction of creative think-
ing is terrifying to them.’82 He singled out Amelang’s criticisms as a prime
example of this envious and fearful attitude in German academia. The controls
79 See Amelang, M. (1991), Tales from Crvenka and Heidelberg: what about the empirical
basis? Psychological Inquiry 2, 233–6; Temoshok, L. (1991), Assessing the assessment of
psychosocial factors, Psychological Inquiry 2, 276–80; and Vetter, H. (1991), Some obser-
vations on Grossarth-Maticek’s data base, Psychological Inquiry 2, 286–7.
80 van der Ploeg, H. (1991), What a wonderful world it would be: a reanalysis of some of the
work of Grossarth-Maticek, Psychological Inquiry 2, 280–5.
81 Manfred Amelang, interview, 1 February 2003; Henk van der Ploeg, interview, 11 February
2003. As Amelang recently told me: ‘I have never understood what the aim of that
enterprise was and at what time the lists had been given to the Mayor.’ Manfred Amelang,
personal communication, 16 January 2008.
82 See p. 292 of Frentzel-Beyme, R. (1991), Levels of interest in an epidemiological approach
of identifying psychomental risk factors for cancer, Psychological Inquiry 2, 290–3.
09-Buchanan-Ch09.indd 39009-Buchanan-Ch09.indd 390 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

1991: A SCEPTICAL PEAK
391
for ensuring the separation of predictor and criterion data that Grossarth-
Maticek was forced to submit by those at Heidelberg were, for Frentzel-Beyme,
unprecedented and based on ‘impossible’ demands.
Eysenck attempted to respond to each critique with some input from
Grossarth-Maticek. Yet much was left unanswered. How might the interested
reader react, Eysenck asked, with ‘Perplexity? Incomprehension? Bewilderment?
Mystification?…’83 In the face of this fearsome assault Eysenck (uncharacter-
istically) distanced himself from Grossarth-Maticek, though only strategically:
‘There were many aspects of the work about which I could only say that
I would not have done it that way, and regretted that I had not been there when
decisions about methodology and statistical analysis had been made.’84 While
acknowledging that mistakes had occurred, Eysenck still argued that this
research should be taken very seriously. But Eysenck was left with a delicate
balancing act. In defending the overall value of the studies he pointed out that,
despite the quibbles, they demonstrated the overriding importance of person-
ality in determining who succumbed to particular physical diseases. Conversely,
in heading off the too-good-to-be-true charge, he was forced to argue that
these results were merely in line with other previous research. For once he
appeared to be talking down his data rather than talking it up.
The following year, British psychiatrists Tony Pelosi and Louis Appleby
launched a scathing broadside at the Heidelberg intervention studies in the
British Medical Journal. More saliently, they accused Eysenck and Grossarth-
Maticek of not understanding the significance of their own data. According to
the British pair, Eysenck and Grossarth-Maticek had actually outlined risk
factors that were ‘the highest ever identified in non-infectious disease epidemi-
ology’. They calculated that type I personalities were 121 times more likely to
die of cancer than healthy personalities. Furthermore, they had described
intervention effects which, ‘if correct, would make creative novation therapy a
vital part of the public health policy throughout the world.’85 However, Pelosi
and Appleby questioned the reliability of the four-type approach, the spotty
description of methods and measures used, and the immense amount of
time Grossarth-Maticek must have necessarily put in to fulfil the therapeutic
83 Eysenck (1991c), Reply to criticisms of the Grossarth–Maticek studies, p. 316.
84 Ibid., p. 298.
85 See p. 1296 of Pelosi, A.J. and Appleby, L. (1992), Psychological influences on cancer and
ischaemic heart disease, British Medical Journal 304, 1295–8. Taking the mortality rates of
the healthy autonomous types as a baseline, they also calculated that Eysenck and
Grossarth-Maticek’s type 2 subjects were 27 times more likely to die of heart disease.
09-Buchanan-Ch09.indd 39109-Buchanan-Ch09.indd 391 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
392
interventions described, a point raised previously by David Spiegel.86 Eysenck’s
reply defended some of the charges but also raised further problems.87
Concluding the exchange, Pelosi and Appleby poured scorn on Eysenck’s
claim that the studies were in line with other research in the area.
A denouement of sorts
Other questions from the critics lingered. What relationship did the published
inventories and scales have to those that were actually used in Crvenka and
Heidelberg in the 1960s and 1970s? Most of these inventories and scales have
never been made available in the language they were originally presented in
decades before. Why did the concept of rationality and anti-emotionality
disappear and reappear as it did? It was first linked with cancer but later, inex-
plicably, with rheumatoid arthritis. More intriguingly, how was Grossarth-
Maticek able to achieve such a breakthrough—given that he was not ‘burdened’
by a formal education in psychometrics, epidemiological statistics, and psy-
chotherapeutics, nor an intimate knowledge of most of the relevant psycho-
logical literature since he barely spoke English?
Attempts to investigate and replicate ran into obstacles of various kinds.
Amelang has made the most effort on this score. His 1992 finding that
Grossarth-Maticek’s type 1 and type 2 were almost psychometrically indistin-
guishable was seen as very damaging.88 How could measures of what was
essentially the same thing predict different outcomes with such power and
accuracy? Grossarth-Maticek countered by arguing that one needed to use
additional items and measures. Pre-warned of this replication work, Eysenck
and Grossarth-Maticek also published two papers at this time that highlighted
the importance of the administration method when using the type invento-
ries.89 The best (i.e. most reliable and predictively valid) results were achieved
when the inventories were presented in an atmosphere of rapport and trust,
86 Spiegel, D. (1991), Second thoughts on personality, stress, and disease, Psychological
Inquiry 2, 266–8.
87 Eysenck, H.J. (1992a), Psychosocial factors, cancer and ischaemic heart disease, British
Medical Journal 305, 457–9; Pelosi, A.J. and Appleby, L. (1993), Personality and fatal
diseases, British Medical Journal 306, 1666–7.
88 Amelang, M. and Schmidt-Rathjens, C. (1992), Psychometric properties of modified
Grossarth-Maticek and Eysenck inventories, Psychological Reports 71, 1251–63.
89 Grossarth-Maticek, R., Eysenck, H.J., and Barrett, P. (1993), Prediction of cancer and
coronary heart disease as a function of questionnaire administration, Psychological Reports
73, 943–59; Grossarth-Maticek, R., Eysenck, H.J., and Boyle, G.J. (1995), Method of test
administration as a factor in test validity: the use of a personality questionnaire in the
09-Buchanan-Ch09.indd 39209-Buchanan-Ch09.indd 392 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

THE IMPLICATIONS OF EYSENCK’S INVOLVEMENT
393
accompanied by explanations that ensured understanding. It was a point that
had taken a curiously long time to come to light, and it made these studies far
more difficult, if not impossible, to duplicate. For Amelang, it was the final
step on the part of Grossarth-Maticek and Eysenck to insulate their work from
criticism.
Nonetheless, Eysenck became increasingly enraged, exasperated, and even
dejected by what he saw as the one-sidedness of the critique. His reaction to
the volley of criticism in Psychological Inquiry was possibly the most defensive
of his career. While he was back on the front foot in responding to the Pelosi
and Appleby attacks, an air of apostasy becomes apparent in his last dealings
with the subject. Nearing the end of his life, Eysenck told Grossarth-Maticek
that the controls put in place to ensure the integrity of the data were not
working.90 Clearly not; some saw them as insufficient, while others saw them
as evidence of a less than benign form of secretiveness, or even as a combina-
tion both. So by the end, Eysenck had had enough. According to Grossarth-
Maticek, Eysenck felt it was time for a new phase, to move on to replicating or
extending these studies. Uncharacteristically perhaps, Eysenck began to with-
draw from the debate and ceased to engage with critics. For example, he refused
to listen to a 1996 conference paper on this work given by Amelang in Ghent,
even though they had once been firm friends and collaborators.91 His last con-
tribution on the subject was added to an edited volume with the understand-
ing that it would not be subject to peer review.92
The implications of Eysenck’s involvement
Much has been left unresolved following Eysenck’s death. Some commenta-
tors—even Eysenck himself—argued Grossarth-Maticek’s research was still
worthy, even if he had cut corners and made silly errors. Grossarth-Maticek’s
personal investment might have also exerted an unconscious, biasing effect
prediction of cancer and coronary heart disease, Behaviour Research and Therapy 33,
705–10.
90 Ronald Grossarth-Maticek, interview, 31 January–1 February 2003.
91 Manfred Amelang, interview, 1 February 2003.
92 See Eysenck, H.J. (2000), Personality as a risk factor in cancer and coronary heart disease,
in Stress and health: research and clinical applications (ed. D.T. Kenny and J.G. Carlson),
pp. 291–318, Harwood, Amsterdam. This information came from Dianna Kenny via
Michael Eysenck, personal communication, 20 November 2001. See also the posthumous
article Marusic, A., Gudjonsson, G., Eysenck, H.J., and Starc, R. (1999), Biological and
psychosocial risk factors in ischaemic heart disease: empirical findings and a biopsycho-
social model, Personality and Individual Differences 26, 285–304.
09-Buchanan-Ch09.indd 39309-Buchanan-Ch09.indd 393 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
394
over his own judgement and that of his assistants. Questions regarding the
reliability of the mortality data might be seen to fall into this category, along
with suggestions that student interviewers falsified or adjusted records through
laziness or an effort to please their director.
Those who were more critical suggested the matter went beyond sloppiness,
that there were grounds to suspect something more serious was involved.
More specifically, they questioned the independence of the mortality and per-
sonality data. Certainly the 1977 Habilitation document, with the idiosyncratic
item weights that guaranteed almost perfect prediction, struck those at the
University of Heidelberg as a case of retrospective fitting masquerading as
genuinely blind prediction. Other problems, such as the dodgy descriptions of
subjects, numbers that didn’t add up or were not consistent from report to
report, also fell into this category. Some of those closest to the research—such
as Hermann Vetter, Manfred Amelang, and Henk van der Ploeg—even har-
boured doubts about Grossarth-Maticek’s raw numbers.93
The Heidelberg study was extraordinary in terms of its breadth and ambi-
tion. It must have been a huge undertaking. Grossarth-Maticek had to track
large numbers of people for a long time, requiring immense resources and
the cooperation of all sorts of agencies and individuals. Over 20, 000 people
were said to have been tested in a prospective design—a bigger, richer data
base than any other comparable study in the field. Such studies are typically
the province of major institutions and big research teams. Grossarth-Maticek
did it all as an independent scholar, and on a shoe-string. It was never clear
where the money came from, especially for financing the expensive labour-
intensive phase of initial data gathering. Grossarth-Maticek said he had private
family sources (i.e. overseas relatives), but otherwise preferred to keep his
funding arrangements private.
It seemed Grossarth-Maticek had supporters in the medical fields and was
able to obtain, one way or another, genuine mortality data. However, the
names (and the reality) of the thousands of subjects tested by Grossarth-
Maticek have remained a mystery, a situation confounded by strict German
privacy laws. Likewise, the student assistants employed on the project were
hard to track down. To my knowledge, none have been questioned on record,
although private accounts of their working experience are said to exist.
In addition, the co-therapists who apparently helped Grossarth-Maticek perform
the thousands of hours of psychotherapy were never clearly credited. Eysenck
and Grossarth-Maticek were ambiguous on this point. On some occasions they
93 Manfred Amelang, interview, 1 February 2003.
09-Buchanan-Ch09.indd 39409-Buchanan-Ch09.indd 394 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

THE IMPLICATIONS OF EYSENCK’S INVOLVEMENT
395
used the terms ‘we’ when describing the performance of therapy, while on others
they appeared to indicate that Grossarth-Maticek alone performed the treat-
ment. Both the Crvenka and Heidelberg studies were also remarkable for their
foresight, apparently made possible by Grossarth-Maticek’s prescient incorpo-
ration of ideas and measures that anticipated the future direction of psycho-
logical research (e.g. high-risk personality types) and advances in knowledge of
drug, disease, and lifestyle interactions.
If one takes a dim view of this work, it poses many questions about the
involvement of Hans Eysenck. Clearly, the ageing professor came in with his
eyes open knowing there were problems with this research that went beyond
the methodological. The ‘three steps to trust’ he reputably imposed were testa-
ment to this. Eysenck thought he could straighten out the data as well as polish
up its presentation. Many of those who took a close interest in this case were
very curious about Eysenck’s role in the reconstruction of the inventories.
Although never fully explained, it appears he had a major role in translating
and cleaning them up for publication. Sybil Eysenck confirmed this, recalling
that the original German language versions of the questionnaire(s) were
‘a mess’.94 She also told me that her late husband thought he could make sure
mortality information was only gathered at the appropriate juncture—all of
which pointed to Eysenck’s involvement in an enterprise he knew was ques-
tionable, even if his intentions were good.
In any event, good intentions may not have mattered, such was the Yugoslav’s
control over the whole project. According to Amelang, Grossarth-Maticek saw
himself as the project’s spiritus rector, publishing whatever details he cared to at
a time of his choosing.95 No one besides Grossarth-Maticek could vouch for
the integrity of the central methodological feature of this research—the match-
ing of personality predictions to mortality. Eysenck was never in a position to
fulfil the role of guarantor he set himself up in. Eysenck’s ‘three steps to trust’
were ultimately irrelevant, if indeed they were taken seriously at all. For exam-
ple, a check of subject lists deposited with the former mayor was only provoked
by the alarm bells van der Ploeg set off. Otherwise, why would Eysenck partici-
pate in a ritual he had already performed? No one else involved in this work,
not even the re-analysis team led by Spielberger, could confirm whether Eysenck
actually performed the other checks. It is only on Grossarth-Maticek’s assertion
that these checks existed. And even if Eysenck did perform them, it would
have still left open the integrity of the predictor score data. Tobacco industry
94 Sybil Eysenck, interview, 25 July 2003.
95 Manfred Amelang, interview, 1 February 2003.
09-Buchanan-Ch09.indd 39509-Buchanan-Ch09.indd 395 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
396
documents indicate that Eysenck also wanted to use his own personality meas-
ures (the EPQ presumably) to compare with and corroborate Grossarth-
Maticek’s. This work was apparently not done either. There was no one else on
hand and on record to corroborate many of the details of his work. Eysenck
could only ever deal with the most easily disposed of accusations and invaria-
bly called upon Grossarth-Maticek as the final witness. Since it was Grossarth-
Maticek’s credibility at issue in the first place, this never convinced the
sceptics.
Despite the damning judgements, Eysenck continued to hold Grossarth-
Maticek in high esteem and talked of him with sincere reverence as he did of
his other heroes like Galton and Pavlov.96 Eysenck was so entranced by the
potential of the data he took Grossarth-Maticek at his word. Having met
Ronald Grossarth-Maticek, I can only say that he can be highly persuasive. In
person, he presents as a man under siege from jealous rivals who persecute him
because of his radical but successful ideas. In his own defence, he says he has
neither the cynical outlook nor the mathematical sophistication to produce
the effects he is accused of, and takes great delight showing off his neat folders
of raw data to anyone who is interested. When questioned further, Grossarth-
Maticek is able to furnish all sorts of explanations and anecdotes to cover the
quirky aspects of his work. Clearly committed to his science, he is a charming
and expansive theorizer who apparently has the Midas touch—almost every
research area he has gone into has turned up empirical gold.97
Nevertheless, the whole affair suggested Eysenck’s judgement left something
to be desired. Excusing him as merely naïve in taking Grossarth-Maticek’s data
at face value would seem inappropriate as well as inaccurate. At the very least,
Eysenck appeared guilty of not questioning when he should have, of not exer-
cising the kind of scientific rectitude one might expect from a man of his
stature.
Stop press!!
Given that the case had some potentially sensational ingredients, it is curious
why it never made news headlines. After all, important scientific claims were
involved, and several thousand subjects and massive amounts of data were
96 Jeffrey Gray, interview, 25 July 2003.
97 See, for example, an account of Grossarth-Maticek’s on the effects of vitamin supplemen-
tation and autonomy in Nias, D. (1997), Psychology and medicine, in The scientific study
of human nature: tribute to Hans J. Eysenck at eighty (ed. H. Nyborg), pp. 92–108, Elsevier,
Oxford.
09-Buchanan-Ch09.indd 39609-Buchanan-Ch09.indd 396 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

STOP PRESS!!
397
at issue. Plus there was the mystery surrounding the therapy collaborators and
the scores of assistants, not to mention the uncertain patronage. The superfi-
cial explanation for this inattention would be that the case has never been
conclusively adjudicated. And unlike some other well-publicized controver-
sies, no crusading journalist has ever taken up the cause. In the mid-1990s,
Tony Pelosi and British Medical Journal editor Richard Smith had hoped the
news media would take this matter further. Pelosi talked to one reporter, but
no story subsequently appeared. Perhaps the technicalities of this research
were too difficult to turn into a clear and concise story, Pelosi suggested.98
Comment within the more popular outlets was otherwise relatively light, con-
fined to a few cautious news reports and the likes of a Psychology Today article
and a recent Der Spiegel article.99
Why then has so little been resolved? One of the lessons drawn in recent
times was that the necessary checks and balances can be undermined by an
intimidating, powerful figure. Certainly, Grossarth-Maticek had almost sole
control over the generation and analysis of the data. The English language ver-
sions of this research were mainly funnelled through journals in which Eysenck
had a strong editorial role.100 Eysenck’s collaboration with Grossarth-Maticek
could be seen as a somewhat counterproductive attempt to bring him in from
the cold. His astounding claims cut across, undermined, or seriously dimin-
ished the more modest efforts of other more mainstream research programmes.
It forced attention on the work and it forced a judgement from those who
might simply have preferred to ignore it. However, Grossarth-Maticek’s out-
sider status put these fellow researchers in a difficult position. In making their
negative assessments, they were all too aware of being seen to be rounding on
an inspired but poorly resourced underdog. Thus sympathetic allies like Rainer
Frentzel-Beyme could argue that Grossarth-Maticek was a man unfairly
maligned by a narrow-minded scientific establishment. Consequently, most
commentators kept their distance, citing only rational-empirical, public
domain reasons for not trusting this work. Tellingly, it was those closest to
98 Anthony Pelosi, personal communication, 29 September 2003. Nevertheless, Simon
Wessely, then at the IoP, wrote a sceptical evaluation: Wessely, S. (1993), Is cancer all in
the mind? The Times, 22 June 1993.
99 See Fishman, J. (1988), The character of controversy, Psychology Today, December 1988
and Kurz, F. (2002), Akademisches Schattenreich, Der Spiegel, 37, 9 October 2002.
100 These Eysenck-controlled journals were Behaviour Research and Therapy and Personality
and Individual Differences.
09-Buchanan-Ch09.indd 39709-Buchanan-Ch09.indd 397 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
398
Grossarth-Maticek—van der Ploeg, Vetter, and Amelang—who produced the
most damning and frankly accusatory appraisals.101
The whole saga highlighted the ineffable but essential ingredients of
science—of cooperation, openness, and good faith that underwrite trust.
According to Amelang, ‘no scientifically satisfactory evaluation of the claims
made by Grossarth-Maticek and by Eysenck can be achieved through critical
readings of their published works.’ Only an ‘independent and methodologi-
cally incontestable’ corroboration could resolve the matter, he wrote. 102 But a
close and therefore convincing replication would require far greater levels of
cooperation than Grossarth-Maticek was apparently willing to display. And
while a successful replication might support the validity of this kind of
approach, it would not dispel all the doubts about Grossarth-Maticek’s con-
duct and research. (See the concluding chapter for an extended discussion of
this point.)
Sanctions and brand name protection
It seems that all the ink given to the scientific review process in the last couple
of decades hasn’t helped much. Various checks and safeguards have been put
in place, including guidelines for data retention and sharing. While increased
responsibility has been sheeted home to employing institutions, independent
watchdog bodies have also been set up in several countries, including the US.
Despite these developments, the uncertain status of Grossarth-Maticek’s work
demonstrated some glaring weaknesses in the scrutiny process. This work was
only subjected to the standard peer review process, and only in some publica-
tions at that. And anyway, peer review was not capable of dispelling the scep-
tics’ doubts because, as a recent study concluded, peer review provides no
absolute barrier to misrepresentation and malpractice.103
Moreover, the internationalism of science did not mesh easily with the
national and institutional structure of its policing mechanisms, leading to an
unfortunate form of buck passing. Grossarth-Maticek’s outsider position and
émigré status were barriers to his credibility but they did tend to head off
potentially damaging independent investigations. The recent war-torn history
101 Apart from van der Ploeg and Vetter’s critiques already mentioned, see also Vetter, H.
(1993), Further dubious configurations in Grossarth-Maticek’s psychosomatic data,
Psychological Inquiry 4, 66–7.
102 Amelang (1991), Tales from Crvenka and Heidelberg: what about the empirical basis?,
p. 235.
103 See Hagan, P. (2003), Review queries usefulness of peer review, The Scientist, 28 January
2003.
09-Buchanan-Ch09.indd 39809-Buchanan-Ch09.indd 398 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SANCTIONS AND BRAND NAME PROTECTION
399
in Grossarth-Maticek’s native Yugoslavia made any inquiries there very diffi-
cult. Grossarth-Maticek did not belong to any mainstream academies or pro-
fessional associations in Germany and England. While he had left the Institute
of Social Medicine at Heidelberg in the mid-1970s after failing to share data
with Professor Maria Blohmke, once he set himself up as private scholar, he
was Teflon-coated.104 For example, the German psychological hierarchy was
unable to pursue Amelang’s complaint about potential scientific misconduct
because it did not have the jurisdiction to investigate. Following the damaging
published critiques of the early 1990s, those at the IoP sought to distance the
Institute from Grossarth-Maticek. The IoP’s Dean, Stuart Checkley, wrote to
the British Medical Journal, 31 July 1993, stating that, to the best of his knowl-
edge, Maudsley ethics committees had never had the opportunity to vet this
research, nor had the Institute ever awarded a title to Grossarth-Maticek, nor
was this research part of the joint research strategy of the Maudsley and the
Institute.105 But that was all they could do. There were also some moves to
censure Eysenck, with talk of stripping him of his emeritus professorship, but
this came to nothing.106 In 1995, Tony Pelosi even presented an ethics com-
plaint to the BPS relating to Grossarth-Maticek’s use of subjects with high
blood pressure in a case control therapy trial. No action was taken, however,
for reasons the Society did not care to reveal.107
In lieu of any special investigation of Grossarth-Maticek’s work, the pub-
lished record was left to speak for itself. It seemed that sanctions functioned to
protect the brand name of employing institutions and the reputation of
professional bodies; whether they ensured the integrity of the science was an
altogether different question. Grossarth-Maticek continues to do research at
the margins of science investigating unusual drug remedies and promoting his
lifestyle and health strategies. He is still cited approvingly by cancer patient
groups and the self-help movement that goes with this. Amazingly, Vetter
104 A. Wallace Hayes, Meeting with Professor Blohmke, Dr. Kulessa, and Mr. Stelzer,
4 November c. 1983, http://tobaccodocuments.org/bliley_rjr/505743303-3315.html.
105 Checkley, S. (1993), Letter to the editor, British Medical Journal 307, 329.
106 Louis Appleby, personal communication, 22 September 2003; Anthony J. Pelosi,
personal communication, 29 September 2003.
107 The BPS Investigatory Committee deemed it ‘inappropriate’ to set up an investigatory
panel to look into the material Pelosi had sent them, and henceforth considered the mat-
ter closed. Pelosi disagreed, of course, but was left with little recourse. Anthony J. Pelosi,
personal communication, 29 September 2003. See also Anthony J. Pelosi, The responsi-
bilities of academic institutions and professional organisations after accusations of scien-
tific misconduct, The COPE Report 1998 at BMJ.com, and Richard Smith, Editorial
misconduct: time to act’, at http://www.pitt.edu/~super1/lecture/lec14631/index.htm.
09-Buchanan-Ch09.indd 39909-Buchanan-Ch09.indd 399 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
400
continues to work with the personable Yugoslav, even though the profound
reservations he expressed in print have been only ‘partially resolved’.108
The negative judgements have had a discernible effect; very few textbooks in
health psychology and epidemiology mention Grossarth-Maticek’s work
favourably, if they mention it at all. Exceptions to this are the work of
Spielberger in the US and Rocio Fernandez-Ballesteros in Spain, who have
sought to extend or adapt aspects of this work.109 In the meantime, the psy-
chologists, oncologists, epidemiologists, and psychiatrists involved have simply
moved on. Few were prepared to do the detective work required, or persist
with liasing with Grossarth-Maticek. They would not be paid or rewarded for
this. It would not, ultimately, be in their interests as scientists. Only Amelang
maintains an active scepticism, and only because he has been able to incorpo-
rate personal investigations and critical replications as part of an on-going
research programme. Amelang’s most recent work has offered only limited
support for the personality–disease link, hardly vindicating the particulars and
the strength of the causal connection Grossarth-Maticek claimed.110 Moreover,
his specific appraisals of the Grossarth-Maticek questionnaires hardly sup-
ported their incremental or predictive validity.111
More generally, the interaction between personality and psychological stress
has been shown to have a small but significant role in the development of heart
disease. Intervention aimed at changing coping and lifestyle patterns is now a
relatively well-established part of heart patient care. The picture is murkier
108 Hermann Vetter, interview, 31 January–1 February 2003. For an example of these pro-
found reservations, see Vetter (1993), Further dubious configurations in Grossarth-
Maticek’s psychosomatic data, p. 67.
109 See, for example, Fernandez-Ballesteros, R. et al. (1997), Assessing emotional expression:
Spanish adaptation of the rationality/emotional defensiveness scale, Personality and
Individual Differences 22, 719–29.
110 See Stürmer, T. et al. (2006), Personality, lifestyle, and risk of cardiovascular disease and
cancer: follow-up of a population based cohort, British Medical Journal 332, 1359. Only
certain personality traits (e.g. internal locus of control and time urgency, rather than
anger control, psychoticism, and depression) were weakly associated with a reduced risk
for cancer and heart disease.
111 The most recent study is Amelang, M. et al. (2004), Personality, cardiovascular disease,
and cancer: first results from the Heidelberg Cohort Study of the Elderly, Zeitschrift für
Gesundheitspsychologie 12, 102–15. See also Amelang, M. (1997), Using personality
variables to predict cancer and heart disease, European Journal of Personality 11, 319–42
and Amelang, M. et al. (1996). Personality, cancer and coronary heart disease: further
evidence on a controversial issue, British Journal of Health Psychology 1, 191–205.
09-Buchanan-Ch09.indd 40009-Buchanan-Ch09.indd 400 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING GUNS AND PUBLIC DISCLOSURE
401
when it comes to cancer, however.112 A cancer-prone personality has not been
consistently established, nor is it clear whether psychological intervention pro-
longs survival. Positive outcomes, like Spiegel’s 1989 study, have not been
consistently corroborated.113
The field of psychosomatic medicine has all but split up. Perhaps the most
successful offshoot is the new subfield of psychoneuroimmunology, which
explores the interrelationships among the mind, the central nervous system,
the immune system, and the outcome of immunologically mediated diseases.
It is research in these areas, and in the genetics of disease propensities, that is
now regarded as the most promising.114
Smoking guns and public disclosure
I was unwilling to accept the so often heard argument that smoking is addictive,
because neither then [i.e. the late 1950s] nor now does the term have any scientific
content. I like playing tennis and writing books on psychology; does that mean I am
addicted to tennis and book writing? (Eysenck, H.J. (1997b), Rebel with a cause,
p. 168)
Hans Eysenck claimed that nicotine was not addictive but tobacco money may
well have been, and controversy likewise. He received upwards of £ 800, 000
(in today’s terms) from Big Tobacco in the US and UK. It was probably
Eysenck’s biggest single funding source and the means for supporting for his
post-retirement research.115 Tobacco money enabled him to maintain the
intellectual lifestyle and the research profile he became accustomed to in his
twilight years. It fuelled the kind of media juggernaut his career had become,
but it also had its drawbacks (pun again intended).
Can we therefore lumber Eysenck with the dreaded title ‘bought advocate’?
Well, that depends on what is meant by the term. Despite trawling through a
massive number of confidential industry documents, I found nothing to sug-
gest tobacco industry representatives directly influenced Eysenck’s views.
Moreover, there are good reasons to believe they would not have even tried to
do so. Those in the industry who understood the science also understood the
112 Garssen, B. (2004), Psychological factors and cancer development: evidence after 30
years of research, Clinical Psychology Review 24, 315–38.
113 Stephen, J. et al. (2007), What is the state of the evidence on the mind–cancer survival
question, and where do we go from here? Supportive Care in Cancer 15, 923–30.
114 Genes too have been shown to play a key role in the development of many cancers,
including cancer of the lung, partially bearing out what was a standard hypothesis for
one branch of cancer research, not just Hans Eysenck.
115 Michael Eysenck, personal communication, 12 December 2001.
09-Buchanan-Ch09.indd 40109-Buchanan-Ch09.indd 401 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
402
importance of its demonstrable independence. Unlike the more intervention-
ist style of other players in the funding business—the pharmaceutical industry
springs to mind—the American tobacco industry kept the outside researchers
they supported at a distance and were ambivalent about the usefulness of
in-house research. For example, legal advisor Ed Jacob nearly resigned over the
R.J. Reynolds in-house programme. Tobacco insiders were also very sensitive
about accusations of censorship, doing their best to (paradoxically) cover up
the few instances where it was deemed necessary. Tobacco industry leaders
were able to maintain their canny hands-off stance because they were not
interested in clarifying the tobacco–disease link but in obscuring it. They made
sure they did not have to intervene by mounting an open-ended quest to look
elsewhere, and by handpicking the personnel and the projects to do just that.
Without a doubt, Eysenck allowed himself to be used to defend the legal and
commercial interests of a ruthless industry. He and many other prominent and
respected scientists were an essential part of an industry strategy of public
denial, despite private acknowledgement, of the hazards of smoking.116
Up until around the time of Eysenck’s death, the tobacco industry had never
lost a single one of the over 800 liability cases brought in the US. One must give
industry representatives credit for this, for they played a bad hand very well.
To do so, they needed people like Eysenck. They parlayed his credibility as a
scientist to reassure their customers and foil liability claims.
It was always unclear how much knowledge Eysenck had of industry thinking.
It is hard to believe that Eysenck was not aware of the reasons those involved
in the tobacco industry valued him, given his deep involvement with Jacob at
R.J. Reynolds. He may not have shared their cynically instrumental perspec-
tive, however. Eysenck and many others were used to create the impression
that scientists still disagreed about the smoking–disease link. For Eysenck,
it apparently was an open question. He gave sworn testimony to the US
Congress to this effect, emphasizing his impartiality into the bargain.117
He claimed he came to these conclusions despite the money he received from
116 See Glantz, S.A. et al. (1996). The cigarette papers. University of California Press,
Berkeley.
117 See Hearings Before the Subcommittee on Health and the Environment of the Committee
on Energy and Commerce House of Representatives Part 2 of 3, 9 March 1983, http://
tobaccodocuments.org/lor/03637300-7630.html, and Eysenck’s, Lung cancer, coronary
heart disease and smoking, statement regarding S. 772 submitted to the Committee on
Labor and Human Resources, 6 May 1983. See also Eysenck’s testimony in a Finnish
liability case. Hans J. Eysenck, ‘Transcript of Testimony As Checked by the Court Exhibit
11 Subsection 1,400,’ January 1990 (est.), http://tobaccodocuments.org/pm/2501072480-
2506.html.
09-Buchanan-Ch09.indd 40209-Buchanan-Ch09.indd 402 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING GUNS AND PUBLIC DISCLOSURE
403
the industry and, ironically, he might have genuinely disagreed with the
private opinions of industry insiders who supported him. Yet here was a dis-
tinguished scientist richly sponsored by a powerful industry because he was
seen as authoritative, simply because his work had legal, and therefore com-
mercial, value.
While any attempt to directly influence Eysenck’s views may be discounted,
a prima facie case can be made for the contention that tobacco money appeared
to affect what he researched, if not the outcome of this research. Jacob and
Eysenck appeared to work together on what questions Eysenck should research,
of what would be useful in a ‘litigation-oriented’ sense.118 In these discussions,
Eysenck directly and indirectly suggested ways that the case against smoking
could be attacked. A tacit form of complicity can’t be ruled out in the other,
more ‘above board’ projects Eysenck proposed. The CTR only tended to sup-
port projects with likely favourable outcomes, and advance notice of this
helped secure grants.119 In other scientists’ projects, tobacco industry lawyers
not only had a say in what was researched, they also had some say in how it was
researched—tinkering with experimental designs to yield the most useful
results.120 While no such shenanigans were apparent in Eysenck’s case, the
suggestion remains that he tailored his research proposals to appeal to the
tobacco industry. He also appeared quite willing to lend a hand to help them
sell their product—although it appeared industry types took him less seriously
on this front.
Eysenck always gave smoking the benefit of the doubt, even though it become
less and less clear why he thought it deserved it. Having sincere reservations
about the case against smoking was one thing—Eysenck was certainly not
118 Here is a sample quote: ‘Jacob reported that Eysenck’s work indicates that people begin
to smoke due to environmental factors (peer pressure) but continue to smoke due to
genetic factors. He thinks that a psychological study showing that behavioral patterns are
genetically founded would be of great use. He wants to explore with Eysenck the possibil-
ity of getting a group of people in the US who have trained under Eysenck to work up
material that might be useful in litigation. This would not be a CTR project but would be
strictly litigation-oriented. Jacob has no real idea of the money involved but thinks that
$25, 000 to $50, 000 would be involved for a pilot project and that the whole project
could involve $100, 000 to $200, 000 over a couple of years. Jacob also noted that Eysenck
will be looking for support for a twin registry project in about 1 1/2 years but that this
would be appropriate for consideration by CTR.’ A.H. ‘Request From FTC For Input
Regarding Carbon Monoxide Testing,’ no date, c.1980, http://tobaccodocuments.org/
ness/43461.html, p.3.
119 Crist et al. (1986), Re: Jones/Day Liability Summary (‘Corporate Activity Project’).
120 See Glantz et al. (1996), The cigarette papers, Chapter 8.
09-Buchanan-Ch09.indd 40309-Buchanan-Ch09.indd 403 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
404
alone in this—but feeling compelled to broadcast them from the rooftops for
decades was another. Eysenck’s career path on this issue was one of acquired
dependency and diminishing options. Following in Fisher’s footsteps, Eysenck’s
initial stance of disengaged scepticism seemed rational and relatively unexcep-
tional. Courted by the American industry in the 1970s, minimal intellectual
compromise led to a lot more research money at a time when his notoriety was
beginning to work against him. By the early 1980s, however, as the data linking
cigarettes and disease accumulated, Eysenck must have been aware that his
position was becoming increasingly untenable.
In downplaying the animal studies and preferring to argue the point in the
epidemiological realm, he only succeeded in painting himself into a corner.
As one of a minority of scientists highlighting the shortcomings of retrospec-
tive investigations, Eysenck prompted epidemiological researchers to adopt
more sophisticated and more convincing designs. It seemed that he (of all
people) would never accept the new standard of causality this research gener-
ated. But it also meant that Eysenck needed better evidence for his alternative
genetic-and-personality explanation. As P.N. Lee pointed out, personality
could be the key intervening variable explaining most of the association of
cigarettes and cancer only if the relationship between personality and cancer
was at least as strong—a point Eysenck later acknowledged explicitly.121
Grossarth-Maticek’s data, with its startling relative risk estimates, were a math-
ematical necessity, Eysenck’s get-out-of-jail card. Even some of his closest
colleagues saw Eysenck’s championing of Grossarth-Maticek as a notable
example of his critical double-standard—applying the most rigorous scepti-
cism to rival accounts, while protecting his own work and that of others sup-
porting it.122 The contradictions were more nuanced than this, however.
On the one hand, Eysenck continued to question whether statistical epidemi-
ology could demonstrate a causal connection between smoking and disease, in
the process aligning himself with an old-style medical framework dominated
by a focus on individual susceptibility. On the other hand, he argued that this
121 See the exchange between Eysenck, Stolley, and Vandenbroucke regarding Fisher’s con-
tribution to the tobacco and cancer debate. Eysenck, H.J. (1991a), Were we really wrong?
American Journal of Epidemiology 133, 416–25; Stolley (1991a), When genius errs: R.A.
Fisher and the lung cancer controversy; Stolley, P.D. (1991b). Author’s response to ‘How
much retropsychology?’. American Journal of Epidemiology 133, 428; and p. 434 of
Vandenbroucke, J.P. (1991), How much retropsychology? and reply to Eysenck,
American Journal of Epidemiology 133, 426–7.
122 Jack Rachman, interview, 24 June 2002.
09-Buchanan-Ch09.indd 40409-Buchanan-Ch09.indd 404 1/16/2010 9:45:34 AM1/16/2010 9:45:34 AM

SMOKING GUNS AND PUBLIC DISCLOSURE
405
kind of epidemiological research could reveal just such a connection between
personality and disease.
Grossarth-Maticek’s results gave Eysenck renewed licence to continue his
distinctive line in pro-smoking advocacy through the late 1980s and early
1990s. Eysenck’s increasingly certain pronouncements on the qualified and
specific risks of smoking and the non-addictiveness of nicotine were widely
interpreted as providing a licence to smoke. His input had little bearing on an
intellectual debate that had long since closed. As Baker and Fiore put it in their
review of Smoking, personality and stress, the (non-)addictiveness of nicotine
had ‘virtually no bearing on whether smoking causes coronary heart disease or
lung cancer. … Rather, he [Eysenck] seems more interested in justifying and
defending tobacco use. It is hard to discern a theoretical or scholarly intent
here.’123 Instead Eysenck made himself a champion of consumer choice who
went out of his way to undermine the public health message. For example, in a
1995 piece Eysenck made much of the semantic ambiguity of the ‘smoking
kills’ warning. Somewhat pedantically he argued that, as a relatively weak, syn-
ergistic factor, smoking per se did not kill anyone. He remained a thorn in the
side of public health advocates, who declared him unqualified and irresponsi-
ble but could not afford to ignore him.124
Eysenck carried on with the fight, despite and perhaps because of denuncia-
tions he endured. As with the race and IQ debate, pig-headedness should be
added to the explanatory mix. He was certainly not a man who liked to admit
he was wrong. In his 1991 exchange with Vandenbroucke and Stolley, his
refusal to concede any ground was truly breath-taking. He wrote: ‘at the time
of writing the evidence was clearly insufficient … we never engaged in proph-
ecy as to what future evidence might disclose... our statements were correct at
the time.’ [orig. italics]125 The public image drove the private scholar.126
It made it hard for Eysenck to change tack or leave the debate alone once he
had staked out a position. It made for a paranoid style, finding vindication in
frustrating a litany of opponents instead of embracing new approaches.127
123 Baker, T.B. and Fiore, M.C. (1992), Elvis is alive, the mafia killed JFK, and smoking is
good for you, p. 1016. For another example of this kind of promotion, see Eysenck, H.J.
(1986d), Nicotine’s not all nasty, The Best of Health, February 1986.
124 Fletcher, C. (1965), Eysenck v. anti-smokers, The Observer, 27 June 1965.
125 Eysenck (1991a), Were we really wrong?, p. 429.
126 Fletcher (1981), Plea for the guilty.
127 As with his flirtations with parapsychology and astrology, race and IQ, Eysenck denied
that he was simply playing the enfant terrible. Nonetheless, he did allow himself a
09-Buchanan-Ch09.indd 40509-Buchanan-Ch09.indd 405 1/16/2010 9:45:35 AM1/16/2010 9:45:35 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
406
Eysenck always defended himself by saying that it was all valid research, no
matter who paid for it. And he was by no means alone in taking such money.
The tobacco industry supported a great range of scientists; many were reputa-
ble and influential, some were recipients of Nobel prizes. It may also be unfair
to condemn Eysenck in this manner, given that many scientists in many fields
face perennial funding crises. With only finite means available in a competitive
tendering environment, they must have an eye on the main chance, the most
likely opportunities to get money. Moreover, the tobacco industry may have
used him, but Eysenck also used the industry to fund projects that went beyond
smoking and health. Along with support from the Pioneer Fund, he used
tobacco money to piggyback his extensive post-retirement research on the
biology of intelligence by adding in a study of the effects of nicotine on per-
formance. Without such ‘tainted’ funds this important research would prob-
ably not have been possible. As Eysenck’s successor as head of psychology
Jeffrey Gray noted, it was always extremely difficult to get money for basic
personality and individual differences research from more mainstream sources
such as the MRC or the Wellcome Trust.128 Those at the IoP felt these funding
bodies tended to favour more traditional experimentalism and practical, med-
ically oriented research. Thus Eysenck was forced to fall back on more unor-
thodox sources. Eysenck cheerfully disclosed much of his tobacco sponsorship,
though he was often vague about specific amounts. When it was revealed
toward the end of his life that he had received funds from accounts the indus-
try preferred to keep secret, Eysenck’s evasive response was that ‘we get a lot of
research money’, implying that he could never keep track of where it all came
from.129 Like many scientists, the ideas and the ambition did not stop when he
reached 65. He did not want to put his feet up and felt no obligation to do so.
The tobacco industry enabled Eysenck to maintain a vigorous post-retirement
career and lifestyle—a reversal of the usual effect the industry’s product has on
its customers.
As he did in the race and IQ controversy, Eysenck always maintained
that the duty of a scientist is to advance human knowledge no matter whose
contrary streak of critical rationalism, coupled with a sneaking sympathy for the under-
dog. See Eysenck, H.J. (1986c), Consensus and controversy: two types of science.
128 Jeffrey Gray, interview, 25 July 2003. Gray was himself a recipient of tobacco money,
having been introduced to Ed Jacob by Eysenck. When he took over in 1983, Gray said
he was shocked to learn just how little money the IoP psychology department had been
receiving from established sources.
129 Freeman, J. (1997), The pugnacious psychologist [obituary for Hans Eysenck], The
Guardian, 8 September 1997.
09-Buchanan-Ch09.indd 40609-Buchanan-Ch09.indd 406 1/16/2010 9:45:35 AM1/16/2010 9:45:35 AM

SMOKING GUNS AND PUBLIC DISCLOSURE
407
interests it may serve. ‘Scientists look at any factual question as a legitimate
topic for science’, he testified.130 Given the evidence he saw before him, this
was the truth, the most reasonable position—and the consequences were not
for him to dictate or imagine. He turned Claude Bernard’s dictum, ‘when in
doubt, abstain’ in his favour, arguing that anti-smoking measures were unjus-
tified. Smokers should relax and continue to smoke if they wanted to; it was
the health care do-gooders who needed to lay off. Eysenck’s position was a use/
abuse model carried to an extreme, with science conceptualized as occurring in
some sort of neutral, politically valueless space. For some, this is science at its
most idealized, as it should be. For others, it is science at its most cynical, as it
too often is. Certainly, it was a stance that offered Eysenck plenty of freedom
of choice—of topics and collaborators, patrons and supporters. But it con-
flicted sharply with the world-view of those who saw science and its conse-
quences as more intimately intertwined, as an enterprise necessarily harnessed
to social goals. Eysenck may have convinced himself that no one told him what
to think, that in his heart of hearts he was no one’s lackey. Yet once again, his
insensitivity to conflicts of interest and guilt-by-association made him a target.
As Jeffrey Gray agreed, it tended to confirm his critics’ suspicions that there
was ‘something sleazy about Hans.’131
Today the large US tobacco companies are still vehemently contesting the
compensation claims of some of their best customers. Insider revelations in
the mid-1990s saw their see-no-evil posture exposed as conscious deception.
Having been forced to admit what they had known all along—that nicotine is
addictive and cigarettes kill—they are fighting a rearguard action against indi-
vidual and class-based liability claims.
How Hans Eysenck might have responded to the recently forced admissions
from the major tobacco companies one can only speculate. He certainly would
have found it more difficult to use them as a funding source. Most scientists
and scientific institutions in Britain were forced to wean themselves off tobacco
money by the late 1990s. In the case of the IoP, the pressure came first from
below—from conscientious junior staff members, and psychiatric trainees,
rather than from senior management who tended to take a more benign
view.132 Sniffing the political wind, many of the established funding bodies
(including the Wellcome and the MRC) responded to this kind of activism by
130 Hans J. Eysenck, ‘Transcript of testimony as checked by the court exhibit 11 Subsection
1,400,’ c. January 1990, http://tobaccodocuments.org/pm/2501072480-2506.html, p.24.
131 Jeffrey Gray, interview, 25 July 2003.
132 Ibid.
09-Buchanan-Ch09.indd 40709-Buchanan-Ch09.indd 407 1/16/2010 9:45:35 AM1/16/2010 9:45:35 AM

SMOKING, CANCER, AND THE FINAL FRONTIER
408
refusing to sponsor anyone supported by the tobacco industry—relegating
tobacco money to the renegades and the outsiders.133
A sad postscript
Although Eysenck was an extraordinarily focused, rational man, the last few
years of his life were distracting and demanding. The results were fatal, if you
take Grossarth-Maticek’s word for it, a final confirmation that personality and
stress were intimately related to the development of serious diseases.134 When
Eysenck contacted Grossarth-Maticek with the dark news, Grossarth-Maticek
urged him to undergo the kind of autonomy training the Yugoslav was sure
would prolong his life. Ever the empiricist, Eysenck thought the number of
cases linking unresolved stress to cancer was too low to trust. He was dead
within months.
For Grossarth-Maticek, this was the loss of a key ally. His publications in
English tapered off, and so did the controversy. Still it was not plain sailing.
It transpired that Eysenck had apparently left a great deal of money (i.e. several
million Deutschmarks!) to Grossarth-Maticek as a parting gift to allow him to
continue his research. Grossarth-Maticek said he elected to forgo this money,
given that pressing for it might damage relations with the Eysenck family.
However, his account of the situation was sharply contradicted by several
other people I discussed it with, Sybil Eysenck amongst them.135
133 For one of the latest instalments in the history of tobacco industry cover-ups, see Rego,
B. (2009), The polonium brief: a hidden history of cancer, radiation, and the tobacco
industry, Isis 100, 453–84.
134 Ronald Grossarth-Maticek, interview, 31 January–1 February 2003.
135 Sybil Eysenck, interview, 25 July 2003.
09-Buchanan-Ch09.indd 40809-Buchanan-Ch09.indd 408 1/16/2010 9:45:35 AM1/16/2010 9:45:35 AM