E XT RE ME FE MININIT Y IN B O YS 58

Sex and Gender

VOLUME II
THE TRANSSEXUAL EXPERIMENT

Robert J. Stoller, M.D.

Professor of Psychiatry
DEPARTMENT OF PSYCHIATRY,
SCHOOL OF MEDICINE
UNIVERSITY OF CALIFORNIA AT LOS ANGELES

JASON ARONSON
New York
 CONTENTS

Acknowledgements page vii

Introduction
1
Part I
THE HYPOTHESIS

1. Bisexuality: The ‘Bedrock’ of Masculinity and Femininity 7

2. Extreme Femininity in boys: The Creation of Illusion 19
3. The Transsexual Boy: Mother’s Feminized Phallus 38
4. Parental Influences in Male Transsexualism: Data 56
5. The Bisexual Identity of Transsexuals 74
6. The Oedipal Situation in Male Transsexualism 94
7. The Psychopath Quality in Male Transsexuals 109

Part II
TESTS
8. The Male Transsexual as ‘Experiment’ 117
9. Tests 126
10. * The Pre-Natal Hormone Theory of Transsexualism 134
11.The Term‘Transvestism’ 142
12.Transsexualism and Homosexuality 159
13.Transsexualism and Transvestism 170
14.Identical Twins 182
15.Two Male Transsexuals in One Family 187
16.The Thirteenth Case 193
17.Shaping 203
18.Etiological Factors in Female Transsexualism: A First Approximation 223

Part III
PROBLEMS
19 Male Transsexualism: Uneasiness 247
20 Follow-Up 257
21 Problems in Treatment 272
22 Conclusions: Masculinity in Males 281
References 298
Index 313
Part II
TESTS
 10

THE PRE-NATAL HORMONE THEORY

OF TRANSSEXUALISM

When and how much does biology command behavior (rather than
subserve it) ? We know that animals can be ranged on a scale according to
the degree to which their behavior is flexible or drive-bound. The most
flexible animal is man, about whom we want to know: (how much) does
sex (biology) produce gender identity (psychology) ? That is a question
this book is investigating; I am a psychoanalyst struggling to create a
methodology for adequately studying the psychological side of that
equation.
There are experts who believe transsexualism is the result of biological
factors, probably pre-natal hormones that affect brain organization during a
critical period (e.g. 2, 3, 6, 9, 10). This chapter presents no evidence that
their idea is wrong (though that is my belief) but questions its underlying
logic.
First, let us remember that we are considering a theory: no one claims it
is proven that transsexualism can only occur if there is underlying brain
damage, even those who are convinced it is so. Their conviction seems to
come from four sources: (1) on seeing a transsexual, one is often
impressed how profoundly the gender identity is fixed, how strong is the
urge to change sex, how early in life it started, and how successfully this
person manages the role of the opposite sex; (2) animal work has
demonstrated in each species tested so far that adult behavior can be
modified in the direction of the opposite sex by manipulation of peri-natal
hormonal variables; (3) hormonal disorders in humans occur that result in
behavior moving in the directions predicted from the animal work; (4)
biological theories of behavior have appeal as more scientific, more open
to experimentation.
For some reason, respected analytic colleagues who have commented on
my ideas about male transsexualism have said

134
 THE PRE-NATAL HORMONE THEORY 135

that I believe the condition is due to biological factors (5, 11). They are
right if they think I do believe, along with everyone else, that biological
factors can contribute to normal gender identity development (12, 13). But
I do not claim that in regard to transsexualism; rather, I keep saying that
the psychological influences in the family overpower the same biological
anlagen present in males who will be masculine. On the other hand, those
believing in the hormone induction theory may more clearly see, though
through condemning eyes, that my hypothesis says the etiology is
psychological.
Three factors contribute to gender identity: biological, psychological,
and ‘biopsychic’ (e.g. imprinting, conditioning). In most humans, sex and
gender identity development are congruent. Maleness and masculinity and
femaleness and femininity go well together. But if the strength of any one
of these three factors increases unduly prenatally or in infancy-childhood,
gender aberration results. There are data (reviewed in 14 and Chapter 1)
showing that the norm is breached in certain ‘experiments’: on the
biological side by chromosomal disorders, pre-, peri-, and post-natal
endocrine disorders, and surgical and hormonal invasions by physicians;
on the psychological side by powerful, aberrant parental, sibling, and peer
influences; and in a biopsychic sector bridging the other two, by
imprinting, classical, visceral, and operant conditioning, and explosively
traumatic experiences. It is a researcher’s task, in studying such disorders,
to find which sector was unduly strengthened (or weakened). In the case of
transsexualism, I have found these data in the powerful family dynamics.
Nonetheless, ‘experiments’ exist in which a change in the strength of one
of the other sectors causes gender reversal. It tempts the experimenter,
therefore, to apply his solid data from X (e.g. behavior reversal in male
rats given perinatal estrogens) as an explanation for Y (e.g. extreme
femininity in boys). Such postulation can lead to the discovery that
transsexualism is a paranatal hormonal disorder but not without
confirmatory data.
I am approaching an idea that is simple enough and yet apparently still
too complex for acceptance—because it has two parts:* (1) Transsexualism
is the result of family influences,

* Occam’s razor does not necessarily demand a single explanation for similar
phenomena: does all abdominal pain have the same etiology ?
136 THE TRANSSEXUAL EXPERIMENT

regardless of the fact that gender identity reversal can also be produced
experimentally in animals by purely biological forces. (2) In rare cases, for
reasons still not tracked, gender identity reversal can also be produced in
humans by biological forces, such as congenital hypogonadism in males.
In other words, two very different causes can produce reversal. (Whether
the clinical pictures in these gender reversals due to the two different
causes are alike is unclear to me; I have only studied seven patients in
depth whose gender disorder was associated with a sex disorder. Two of
these seemed as absolute in reversing gender identity as a classical
transsexual. The rest, while strongly aberrant and requesting sex change,
had features not found in transsexuals, e.g. periods of masculine behavior,
or perversion, such as foot fetishism.)
I am saying that there are two different conditions that lead to gender
reversal. The first, more common, is transsexualism, in which there is no
biological diathesis (except the infant’s beauty and grace) but powerful
parental effects. The second, very rare, is an intersexuality, in which there
is a powerful biological defect but no parental effect. Two different
conditions.
For those espousing a primarily biological cause, the animal work (see
Chapter 1) gives the strongest impetus to the belief that transsexualism is
biologically induced. In each experiment, however one treats the variables,
one can induce behavior of the opposite sex in an animal: expected adult
sexual behavior can be permanently reversed by biological manipulation
of the fetus or newborn. If this can be done in all animals tested so far, it
can probably be done in humans. In fact, we know that massive amounts
of circulating opposite-sex hormones in the fetus can cause measurable—
though not profound—cross- sexed behavior later in life (8), (and there are
studies, contradicted by other studies, that homosexuality and hormones
may be related [reviewed in 4]). May there then not be some as yet
unmeasured hormonal influence at the root of transsexualism?* If so, the
hormone induction theory will have to

* Money has a biological explanation that encompasses all kinds and degrees of
gender reversal: ‘The etiology of transexualism though much speculated about,
remains essentially unknown. . . . Whereas there are no known laboratory tests
available today that show transexuals, as a group,
 THE PRE-NATAL HORMONE THEORY 137

explain these following obscurities (for which the family dynamics theory
has answers in data already collected). Why does transsexualism almost
invariably occur to only one child in a family? If a maternal defect sets
loose the hormones that afflict the fetus, the defect ought to occur on
occasion in more than one pregnancy; on the other hand, if the infant
inherits a genetic defect that sets off in the fetus the abnormal presence of
opposite-sexed hormones, then why also is this genetic defect only found
in one child of the family, and why is it not seen in previous generations?
In fact, when a family was finally studied in which there were two
transsexual siblings, the typical family

to be consistently physically different chromosomally, hormonally, or

morphologically, from a randomly sampled group, there also are no mental formulae
or tests that show them to be consistently different in terms of psychodynamic
history. Such lack of discrimination may be a function of the crudeness of today’s
tests. It is possible, for example, that tomorrow’s tests will yield measures of the
residual influence of fetal sex hormones on the fetal brain. It has already been
established ... in animal experimentation that sex hormones do in fetal life influence
the brain, especially those hypothalamic centers that neurohumorally regulate the
cyclic or noncyclic function of the pituitary gland, and also those adjacent centers
that co-regulate sexual behavior patterns. Not only do sex hormones produced by the
fetus itself influence its brain, so also do exogenous hormones injected into the
pregnant mother. Moreover, exogenous substances which are hormone antagonists
will also influence the fetus if injected into the pregnant mother. For instance, if a
pregnant animal is injected with androgen in order to masculinize the fetal females,
but is also simultaneously treated with barbiturates, then the injected androgen will
have no effect. This experimental finding raises a very cogent question as to the
possible, but as yet unknown influence, on the neural sexual centers of the fetal brain,
of barbiturates and other medications taken by a woman during pregnancy. It is
possible that a pregnancy drug effect could handicap a baby in the early postnatal
period of psychosexual differentiation, making him or her easily vulnerable to an
error of gender identity. A similar handicap could, in the rare instance of Klinefelter’s
(47, XXY) syndrome, be genetically induced, for it has been found that males with
this rare syndrome are also transexual more than would be expected by chance alone
(Money and Pollitt 1964).’ He suggests that so far undiscovered psychological factors
may also contribute, once the brain has been thus disordered; he believes in some
cases the biological contributes more and in others the psychological, but in all cases,
the disordered brain, damaged by a pre- or peri- natal biological forces, is a sine qua
non. Since he also considers as transsexualism conditions I would diagnose as
homosexuality, transvestism, or other perversions, he asks his biological theory to
cover more ground than I would for any etiological explanation in gender disorders
(9, pp. 251-2).
138 THE TRANSSEXUAL EXPERIMENT

dynamics were found in the infancy of both children (see Chapter 15). And
when two pairs of identical twins were studied, in which one of each pair
was a transsexual and the other not, only the twins reared by the parents in
the manner predicted by my ‘interpersonal’ theory were transsexual (see
Chapter 14).
I am not suggesting that the brain of the transsexual is neutral at birth,
that is, that no hormonal-CNS prenatal factors were present. I presume
they were but that, as with almost all males in all species not tampered
with in the laboratory, their brains have been organized by androgens. My
argument is that, despite such an effect, an environmental force—mother’s
presence—overpowers these biological effects.
Animal experiments that tamper with fetal brain physiology can produce
adult gender reversal, but the absence of reports of completely gender-
reversed (‘transsexual’), anatomically normal, free-ranging animals can
make us unsure that the above experiments tell us much about real-life
etiology. Outside of the laboratory, in only those rare cases, such as
freemartins, where the opposite-sex hormonal influence is so great that it
creates hermaphroditic genitals, does one see marked, consistent cross-sex
behavior changes in an animal. While anything is possible, it seems
unlikely that data will appear to support a theory of hormone-induced
permanent cross-sex behavior in transsexuals, behavior that is unknown in
the free-ranging anatomically-intact members of all other species but
present in man. One should extrapolate from the animal studies to man
only when non-experimental animals also are found with profound gender
reversal. So the idea that the animal work already done is evidence of the
etiology of transsexualism is shaky.
Well before being convinced by the hormonal experiments, we should
also insert another link for our chain of logic: an attempt to reverse gender
behavior in animals by environmental changes, e.g., conditioning or
aberrant rearing, while hormones are not the manipulated variables.* So far
as I know,
* There is a puzzling absence of experimental work attempting to prove or
disprove in animals that environmental effects playing upon the biologically intact
organism can cause gender reversal. So far as I know, the hundreds of researchers
studying the development of gender behavior in animals are trying to produce
abnormal gender behavior only by changing the animal’s biological status, not by
environmental (learned) effects. I suspect this one-sided experimentation reflects a
prejudice present these days that
 THE PRE-NATAL HORMONE THEORY 139

these have not been done, but I think animal psychologists could invent
such experiments. And it will surely be possible, in biologically normal
animals, to discover that changes in rearing can produce gender reversal.
There are already hints of this in the behaviorally damaged offspring of
Harlow’s monkeys.
While there is no information in this chapter not published before, I
wish, nonetheless, to comment again about the patients who raise the
question whether a ‘biological force’ * may contribute to marked
aberrations in masculinity and femininity. Patients who fall into this
category have long been to me an exception to the transsexual hypothesis;
their lives seem to tell us that there are extremely rare humans who
confirm the animal work and in whom powerful biological factors can
override the psychosocial forces that usually dominate in the development
of human gender identity. But there is no evidence a skeptic should accept
proving that gender behavior in most humans is primarily the result of
biological forces.
In 1964 (12) I began describing people born and assigned to the sex
unquestionably indicated by their anatomical appearance who nonetheless
insisted from early childhood that they should be members of the sex
opposite that to which they were assigned at birth. Then, later in life,
laboratory tests established that, except for genital appearance, these
people had a chromosomal and/or hormonal defect congruent with the sex
they wanted to be and opposite that in which they had been forced to live.
The first case was that of a child raised from birth as a girl because of
normal-appearing female genitals but with an otherwise normal male
biology (unknown to anyone). In the face of

gender behavior and its disorders are pretty much a biological matter. At any rate, it
seems not to have occurred to researchers to attempt to produce gender reversal in
biologically normal animals with environmental change alone.
* The term ‘biological force’ is vague; it was chosen exactly to be so. By no means
is this term used to imply there is some supraordinate quality, of mysterious origin
and function, that acts to guide behavior. In the past, areas unknown to science have
been prematurely considered understood when encompassed by fancy ‘biological’
terms, such as ‘life force’ or ‘libido’ or ‘orgone’. All I mean by ‘biological force’ is
that biological variables not yet clearly delineated are apparently motivating a piece
of behavior.
140 THE TRANSSEXUAL EXPERIMENT

rearing that should have produced a feminine-enough girl, the child

forever insisted on being treated as a boy and was finally discovered in
fact to be a male, biologically normal in all ways except for female-
appearing exterior genitals. In writing on this subject in 1968, Baker and I
reported on six more patients, males who suffered from conditions known
to be present from birth on (1). In each case, the infant was born, so far as
could be determined when we saw him as an adult, an apparently
anatomically normal male, and only years later was it discovered that his
testes were inadequate; this was not known in childhood. In each case,
while the evidence regarding their rearing was more fragmentary than it is
in the cases of our transsexual boys, in none was there a hint there had
been an excessively close symbiosis with mother or that father was absent
in the dynamic manner with which we are now familiar for our extremely
feminine boys. In addition, we noted an unexpected number of reports in
the literature in which anatomical males with hypogonadism were reported
as having cross-gender behavior from childhood (behavior ranging from a
strongly transsexual picture to less feminine activities, such as homosexual
preferences or infrequent cross-dressing without excitement). Money and
Pollitt had noted this, in regard to hypogonadal males, before us (7). Such
cases of gender disorder in hypogonadal males continue to be reported and
more so than in other chromosomal, hormonal, or anatomical disorders of
sex. So far, the outstanding feature is that one does not hear or see reported
family influences comparable to those of the transsexuals.*
So, as a test for the hypothesis, we can say that it seems these
constitutionally hypogonadal males have become feminine in the absence
of the typical family constellation and in the presence of a demonstrable
biological abnormality. This is the exception to the rule confirmed
throughout so much of the

* It is still unclear in the published reports, however, how many of these patients
had normal-appearing genitals at birth; if the genitals were abnormal, then cross-
gender behavior is not unusual and can be accounted for by parental uncertainty.
Then too, in none of the cases, including my own, have careful studies of infancy and
childhood been possible. In only one case (12) have I been able to talk with the
parents of such people in order to see more clearly whether parental attitudes might
have played a part. One is forced to make do with fragmentary data.
 THE PRE-NATAL HORMONE THEORY 141

development of human personality that the effects of environment (such as

parents) can override biological effects.
To repeat, in the hope that the complicated may nonetheless be
understood: while there are rare cases in which cross-gender behavior of
differing degree can be caused by the sex hormone defects alone, another,
larger, series of cases are produced by psychological factors alone.
In summary, there is no convincing evidence yet that transsexualism is
caused by a hormonal defect. Since there is at present no way to test this
hormone hypothesis (e.g. by direct experiment on human fetuses), we
cannot say that it does not occur. However, I believe the burden of proof is
now on those who claim this biological etiology. While they have shown
cross-sex behavior can be hormonally induced in animals, they must show
that the parental factors described above are either not present or
coincidental. They will have to show, for instance, that the most feminine
of males, those without fetishism or any history of masculine behavior, do
not typically have the excessively close symbiosis with mother or an
absent father. They must show that these mothers do not have a
predilection for attaching themselves excessively to this infant, and that
the infant’s beauty, which I think sparks the symbiosis, in fact plays no
such part. Although they will be able to show that there are cases in which
not every factor I have described is present, they will have to show that the
constellation of factors I describe is either not present or is coincidental.
The evidence for the ‘family dynamics’ theory is fragmentary so far, but,
unlike most psychodynamic theories, the proposition can be tested. So far,
there have been no data disproving the hypothesis.
142 THE TRANSSEXUAL EXPERIMENT

Chapter 10
1. Baker, H. and Stoller, R. J. (1968). ‘Can a Biological Force
Contribute to Gender Identity?’ Am. J. Psychiat. 124.
2. Benjamin, H. (1966). The Transsexual Phenomenon. New York:
Julian Press.
3. Diamond, M. (1968). ‘Genetic-Endocrine Interactions and Human
Sexuality’. In Perspectives in Reproduction and Sexual Behavior,
ed. M. Diamond. Bloomington: Indiana Univ. Press.
4. Green, R. (1974). Sexual Identity Conflict in Children and Adults.
New York: Basic Books.
5. Heiman, M. (1968). ‘Comment on Dr. Stoller’s Paper’. Int. J.
Psycho-Anal. 49.
6. Ionescu, B., Maximilian, C., and Bucur, A, (1971). ‘Two Cases of
Transsexualism with Gonadal Dysgenesis’. Brit. J. Psychiat. 119.
7. Money, J. and Pollitt, E. (1964). ‘Cytogenetic and Psycho-Sexual
Ambiguity: Klinefelter’s Syndrome and Transvestism Compared’.
Arch. Gen. Psychiat. 11.
8. Money, J. (1969). ‘Sex Reassignment as Related to Hermaphroditism
and Transsexualism’. In Transsexualism and Sex Reassignment,
eds. R. Green and J. Money. Baltimore: Johns Hopkins Press.
9. — (1970). ‘Sex Reassignment’. Int. J. Psychiat. 9.
10. Neumann, F. and Steinbeck, H. (1972). ‘Influence of Sexual
Hormones on the Differentiation of Neural Centers’. Arch. Sex.
Behav. 2.
11. Socarides, C. W. (1969). ‘The Desire for Sexual Transformation’.
Am. J. Psychiat. 125.
12. Stoller, R. J. (1964). ‘A Contribution to the Study of Gender
Identity’. Int. J. Psycho-Anal. 45.
13. — (1968). Sex and Gender. New York: Science House; London:
Hogarth Press.
14. — (1973). ‘The Impact of New Advances in Sex Research on
Psychoanalytic Theory’. Am. J. Psychiat. 130.