MSU

LIBRARIES
m

 

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

Y.

MAY i 0 $539;

 

‘fl

 

EHRLY CHILDHOOD ”OTHER-DAUGHTER INUOLUEHENT,
FEHININITY/HRSCULINITY, AND EROTIC

ORIENTATION IN WOMEN

Bu

Deborah Kay Barry

A DISSERIQIIDN
Submitted to
flichiyan State University
in partial Fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology

1987

ABSTRHCT

ERRLY CHILDHOOD "OTHER-DQUGHTER INUOLUENENI,
FENININITY/HASCULINITY, nNO EROTIC
ORIENTATION IN wONEN
Bu

Deborah Kay Berry

This study investigated a model proposing that
femininity, masculinity, and erotic orientation in women
develop in the early childhood process of differentiation
from mother, and are mediated by differentiation in object
representations. It was hypothesized that greater involvement
with mother in early childhood is associated with
identification with mother, less differentiated object
representations, femininity, and sexual interest in men.
Greater distance in the early mother--daughter relationship
was hypothesized to be associated with more differentiated
object representations, masculinity, and sexual interest in
women.

A sample of 77 adult women (heterosexual, bisexual and
lesbian) was obtained through friendship networks.
Participants completed the following research measures; the

Personal attribute Questionnaire, the Erotic Orientation

Scale, a maternal Involvement Scale, the Semantic
Differential (assessing identification with each parent),
written TAT stories scored for object differentiation, and a
personal history questionnaire.

The results do not provide support for the model,
however, some of the predicted relationships are supported.
Early childhood maternal involvement was correlated with
bipolar gender characteristics for heterosexual women.
Heterosexual women who reported greater mother-daughter
closeness were more feminine and less masculine on the n-F
scale. women’s identification with their mothers was also
significantly associated with their gender characteristics.
High identification with mother was related to high feminity
and low masculinity. Neither erotic orientation nor
differentiation of object representations were associated

with these variables.

EXDIOPOBOPU analyses of the data revealed a number of

unpredicted findings. masculinity in heterosexual women was
positively correlated with the variety and frequency of their
sexual interests in men. Furthermore, for the sample as a
whole, scores in the masculine direction on the n-F scale
were correlated with higher sexual interest in women.
Finally, information about sexual abuse showed that 22 of
the 77 participants had been victims of sexual abuse.
Post-hoc analyses indicated that a significengly lower
incidence of sexual abuse victimization was associated with
two variables: greater early childhood maternal involvement

and higher identification with mother.

ACKNOWLEDBENENTS

I would like to thank the members of my committee for
their guidance and support. Dr’s Burt Karon, Don Grummon, and
Jeanne Bullehorn each provided important direction at various
stages of the process. Their faith and encouragement have
been invaluable. In addition, each has contributed to my
growth and development as a professional in many important
ways.

Special appreciation goes to those persons who
contributed their time, effort, and expertise to this study.
flarty eldenbrandt and Terry Stein offered valuable insights
which improved the research design. Steve Kincaid and Nancy
Hill coded the prOJective data in a sophisticated and
conscientious manner. And I thank Sherrie Pearson and Dianne
Roycraft for the considerable time and effort they devoted to
data collection.

finally, I am most grateful to those friends who provided
emotional support and encouragement during the times of
intensive work on this prOJect. Dianne and Phil Roycraft were
especially nurturing. Their hospitality and generosity are

greatly appreciated.

iv

PREFACE

Personhood is the product of a complex interplay of
variables. Givens consist only of the fundamental biological
state, drives, and predispositions of the human being, and
the ecological milieu which the individual inhabits.
Thereafter culture, socialization, maturation, experience and
a multitude of additional known and unknown factors interact
in the ongoing process of personality. Certainly ranking
among the most complex facets of human existence is
psychosexual life, gender identity, sex role, sexual
orientation and attitudes about these.

One of the earliest conceptions of sexual orientation is
a Greek myth, recorded in Plato’s Symposium. Aristophanes
tells the legend to his companions during a discussion of
love. according to this tale, at the origin of human
existence there were three sexes, man-man, woman-woman, and
androgyn or man-woman. Each had two faces, four arms, four
legs, and the like. These first humans conspired to scale to
heaven and lay hands upon the gods. Zeus, in order to humble
their pride and diminish their strength split them each in
two. Thereafter, human beings have sought reunion through
love, each individual searching for his or her original

counterpart. In the man-woman relationship lies the

potential for procreation, in the union of man-man the
highest expression of manliness, and for the woman-woman, the
manifestation of womanliness.

This conception of the etiology and existence of
homosexuality is strikingly value-free in comparison to the
vast majority of those more recent in history and time. The
legacy of the Judeo-Christian tradition has been a wholesale
condemnation of homosexuality which permeates lew, morality,
and the collective thoughts and feelings of individuals in
society, social scientists not exempt. Homosexuality has been
regarded as the equivalent of perversion, abnormality, and
vie. Persecution operates in many modes, among them social
stigma and the sometimes subtle psychological reactions of
individuals.

The past two decades have witnessed a radical
transformation in the multifaceted relation of homosexuality
and society. Within the context of a wider reevaluation of
human sexuality and many of its cultural determinants,
traditional thoughts and feelings about homosexuality have
been challenged and in many cases reformed. Legal mandates
are challenged, psychiatric and psychological views
confronted, and social awareness fostered through the
saliency of those who voice their concerns. But the
collective image and social stigma of homosexuality remains
largely adverse. Internalizad effect-laden values, be they
religious, ethical, cultural or familial are fundamentally
relativistic, but they are integral components of personality

and identity, and highly resistant to change.

vi

The aim and purpose of this study strives to be not
value-free, but based on a value system whose foundations are
those of a definition of psychological health in which
adaptation is central. Humans are social animals of
tremendous complexity, plasticity, and creativity for
adaptation. Homosexuality is not viewed as perverse or

abnormal, but a variant form of human sexuality.

vii

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES. . . . . . . . . . . . . . . . . .
INTRODUCTION

Statement of Purpose . . . . . . . . . . . . . .

Gender and Sexual Orientation. . . . . . . . .
Definition of Terms. . . . . . . . . . .

Page

.xi

xiii

Theories of Gender and Sexual Orientation Development.

Psychoanalytic Theories . . . . . . . . . . .

Sociological Perspectives . . . . . . . . . .

Object Relations Theories

Nancy Chodorow’s Theory . . . . . . . .

Dorothy Dinnerstein's Theory. . . . . . .
Critique and Overview. . . . . . . . . . .

REUIEw OF THE RESEARCH LITERATURE

Empirical Studies of women’s Gender Development.
Empirical studies of women’s Sexual Orientation.
Lesbianism and Psychopathology. . . . . . . .
Sexual Orientation and Gender Characteristics
Summary. . . . . .
Family Background and Sexual Orientation.
Summary. . . . . . . . . . . . . . . . .
Purpose and Perspective. . . . . . . . . . . . .

NETHODOLOGY

Hypotheses and Rationale . . . . . . . . . . . .
measurement of Uariables . . . . . . . . . . .
Early Childhood Maternal Involvement . .
Identification with Mother/Identification with
Father . . . . . . . .
Level of Differentiation of Object Relations
Representations. . . . . . .
Psychological Femininity/nasculinity . . .
Erotic Orientation . . . . . . . . . . . .
Secondary Uariables and Rationale. . . . . . . .
Procedure. . . . . . . . . . . .

Data Collection . . . . . . . . . . . . . . .
Sample Characteristics . . . . . . . . . .
Family Background and Personal Information . . .
Preliminary Notes on the Data Analysis . . . . .

03%

.39
.HS
.%8
.53
.S%

RESULTS

Hypotheses I: Early Childhood maternal Involvement
and Femininity/flasculinity . . . . . . .

Hypotheses II: Early Childhood Maternal Involvement

Erotic Orientation . . . . . .
Hypotheses III: maternal Involvement and Level of
Differentiation of Object Representations. . . .
Hypotheses IU: Differentiation of Object
Representations and Femininity/nasculinity . . . .
Hypotheses U: Differentiation of Object Represen-
tations and Erotic Orientation . . . . . . . . . .
Hypotheses UI: Erotic Orientation and
Femininity/nasculinity . . . . . . . . .
Hypotheses UII: Identification with flother . .
Hypotheses UII: Identification with Father . . .

Sexual Orientation and Gender Attributes: Exploratory

Findings 0 I O I O O O D O I O O I O l O O 0
Semantic Differential Item Analyses. . . . . . . .

DISCUSSION . . . . . . . . . . . . . .
Identification with mother and Femininity/
masculinity. . . . . . . . . . . . . . . . . . . .
Identification with Mother and Agentic Character-
istics . . . . . . . . . . . . . . . . . . . . .
maternal Involvement . . . . . . . . . .
The Bipolar H- F Scale. . . . . . . . . .
The Proposed mediating Uariable: Differentiation of
Object Relations Representations . . . . . . .
Sexual Orientation and Gender Characteristics. . .
women’ s Erotic Interest, Sexual Orientation, and
Feminity/flasculinity . . . . . . . . . . . . . .
Bisexuality. . . . . . . . . . . . . . . .

Patterns of Differences for Heterosexual and Lesbian

woman a O O O I O O I C O O O 0 O O I O I O I O I I
methodological Considerations. . . . . . . . . . .

.83
.88

.88
.88
.88
.88
.91
.SS

.85
101

103
105
107
108
108

110
111

112
115

115
117

APPENDICES

A

Ell-ICU

1671M

The maternal Involvement Scale.
Photocopies of TAT Cards and Instructions
The Semantic Differential . . . . .

Instructions for Scoring Level of Differenti-
ation of Object Representation.

The Personal Attribute Questionnaire.

The Erotic Orientation Scale.
Questionnaire

Flier

Letter to Participants.

The TAT measure . . . . . . . . . . . . .

The Secondary Uariables: Results of the Data
Analyses. . . . . . . . . . . . . . . . . .

Education and Sexual Orientation.

Semantic Differential Item Analyses

REFERENCES

118
180
183

18%
137
188
138
135
138
138

1%H
155

158

158

LIST OF TABLES

Table Page

1

10

11

18

13

13

means, standard deviations and intercorrelations
of women's self ratings of their maternal involve-
ment for four life stages . . . . . . . . . . . . .67

Socio-economic status of respondents. . . . . . . .78

Socio-economic status of participant’s family of
origin. I I I I I I I I I I I I I I I I I I I I I I73

Intercorrelations between the FAQ n-F scale and
Hatinv scales 1-& for heterosexual and lesbian

groups. . . . . . . . . . . . . . . . . . . . . . .BS
Partial correlation coefficients between object

representations differentiation and flatinvl . . . .87
Correlations between the FAQ and EROS Scales. . . .SO

Partial correlations between identification with
mother and levels of Maternal involvement 1
through a I I I I I I I I I I I I I I I I I I I I I 92

women’s joint classifications of their three-way
identification with mother by their two way cate-
gorization on the FAQ Scales. . . . . . . . . . . .93

Frequencies of women’s 3-way classification of
identification with mother, by their combined n-F
and N classification. . . . . . . . . . . . . . . .9&

Group means and correlations between self labeled
sexual orientation and the FAQ Scales . . . . . . .95

Number and percentage of women in each of the
sexual orientation groups falling into each of
the four joint classifications of H and F . . . . .97

means and standard deviations on the PAQ scales
and age for the 'low-erotic’ and ’high-erotic’
groups, and all subjects. . . . . . . . . . . . . lOO

Correlations between SD items on which women rated
themselves and their identification with mother
scores. (Characteristics of self associated with
identificatition with mother) . . . . . . . . . . 101

Joint frequency of women’s self classification on
flativinv l by their TAT Group classification . . . 1&1

15

18

17

18

18

8O

81

Correlations between TAT Scale 1, Scale 3, and each
of the primary research variables. . . . . . . . . 1&3

Frequencies and percentages of women who had, and
had not been victims of sexual abuse by their early
childhood maternal involvement levels. . . . . . . 152

Frequency of women who were victims and non-victims
of sexual abuse classified by their levels of iden-
tification with mother . . . . . . . . . . . . . 15H

Correlations between Hatinvl and the semantic
differential items for Mother (Women’s ratings of
their mother’s characteristics which are associ-

ated with maternal involvement) . . . . . . . . . 156

Correlations between SD items on which women rated
their mothers and their identification with mother
scores. (Characteristics of mother which are
asssociated with identification with mother.) . . 157

Item correlations between identification with
father and women’s ratings of their own character-
istics. . . . . . . . . . . . . . . . . . . . . . 157

Highest correlations for SD item on which women
rated their fathers and their identification with
father scores . . . . . . . . . . . . . . . . . . 158

vii

LIST OF FIGURES

Figure Page

1 Number of women in each of the gynoeroticism and
androeroticism categories . . . . . . . . . . . .SS

xiii

INTRODUCTION

Statement of Purpose

This project explores personality and developmental
factors which are hypothesized to be associated with
gender-role personality attributes and sexual orientation in
women. The guiding theoretical framework is derived from two
theories of gender development proposed by Nancy Chodorow
(1878) and Dorothy Dinnerstein (1577). A review of the
research literature suggests that these theories may be
relevant to variation in psychological femininity/masculinity
and sexual orientation among women.

Nancy Chodorow and Dorothy Dinnerstein both rely on an
object relations developmental framework and assert that
feminine and masculine personality attributes are engendered
in the differential qualities of the early mother-child
attachment and separation-individuation process experienced
by girls and boys. while these authors limit their
discussions to the relative differences in the development of
females vs. males, this researcher hypothesized that their
theories might explain some factors relevant to variation in
women’s psychosexual and psychosocial functioning. The
hypotheses derived from these two theories predicted that
there are associations between five primary variables: Early

Childhood maternal Involvement; Level of Differentiation of

8

Object Relations Representations; Identification with Mother;
Psychological Femininity/flasculinity; and Sexual/Erotic

Orientation.

Gender and Sexual Orientation

The search for variables relevant to the development of
gender-role characteristics and sexual orientation spans the
range of biological (genetic, cortical, hormonal),
intrapsychic, environmental, and socio-cultural realms.
Until quite recently in history biology was viewed as the
primary determinant of gender and psychosexual functioning.
Recent research has debunked this assumption and indicates
that most aspects of gender and sexual functioning are
diverse, varied, and fundamentally independent of biological

factors. A clear exposition of the terminology used in this

area of study is in order.

Definition of Terms

Shively and De Cocco (1977) have delineated four
different components of sexual identity. They are: l)
biological sex, 2) gender identity, 3) social sex-role, and
5) sexual orientation. Traditional conceptions of these
factors have viewed them as typological-—one is either
male/masculine or female/feminine--or varying along a
unidimensional, biploar continuum. Conceptualizations which

view them as either two orthogonal dimensions or as

multidimensional are now recognized as more accurate
reflections and understandings of these aspects of human
personhood.

Even at the level of biological sex, the pertinent
factors are not as limited as is commonly assumed.
Biological sex consists of at least seven anatomical and
physiological components. Included among these are; sex
chromosomes, internal reproductive organs, external
genitalia, gonads, secondary sex characteristics, hormones,
and cortical differentiation. Although these biological
components are usually consistent as either female or male
within the organism, this is not always the case.

The term ’Gender Identity’ has unfortunately not been
used with much consistency of definition. Shively and
DeCocco define gender identity as an individual’s subjective,
psychological sense of being either female or male, feminine
or masculine. It is an experienced awareness and usually
conscious labeling of oneself based on one’s identification
with the constellation of personality traits which define
gender-role.

Social sex-role and gender-role refer to the multiple
manifest physical, behavioral, and personalty traits which
are culturally prescribed as feminine or masculine. Recently
the women’s movement has fostered an awareness of the
culturally/socially determined origin of gender-role traits,
and called for a reevaluation Of the desirability of our
traditional prescriptions and proscriptions for sex-roles.

The effect has been an increasing tendency to see these

H

traits as independent of biological sex, and a capacity to
assess their merits outside of a value system in which they
are seen as apprOpriate or psychologically fit depending upon
ones biological sex. This has led to a recognition of the
diversity and variability displayed by both sexes, as opposed
to a limiting dichotomous view.

Thus the terms sex-role and gender-role refer to the many
aspects of personality which are culturally defined as either
feminine or masculine. In the context of this study, the
terms ’psychological femininity’ and ’psychological
masculinity' are also used to refer to these aspects of
personality.

Scales which assess femininity and masculinity examine
interests, attitudes, and personality characteristics. Early
scales were typically devised on a unidimensional-biploar
continuum. The Bem Sex Role Inventory (Bem, 1977) and the
Personal Attribute Questionnaire (Spence 8 Helmreich, 1578)
both represent an innovative advance over prior measures,
both construe femininity and masculinity as separate,
orthogonal dimensions. Here an individual achieves
independent scores on each scale, and the conjoint
relationships between femininity and masculinity can be
explored.

Sexual orientation, too, was previously perceived as
bipolar and as an integral component and product of
biological sex, gender identity and sex-role. Any deviation
or inconsistency was presumed to implicate other factors of a

’unified’ whole. Hence homosexuality was thought to reflect

and accompany deviation of other factors including biological
aspects (e.g., hormones), gender-identity, and social
sex-role. Conversely, gender-role behavior deemed
appropriate for the other sex aroused suspicions of
homosexuality. More recent research and conceptualizations
view sexual-orientation as a complex phenomena with many
varying components. Among these are: overt physical sexual
behavior, arousal, fantasy, desire, affectional and emotional
intimacy. Each of these may vary in terms of their degree,
frequency, proportion, quality, and conscious or unconscious
manifestation. In this context homosexuality and
heterosexuality are not separate and distinct categories, but
a whole series of dimensions which may vary independently.
The scale which has been used most frequently to measure
sexual orientation is the Kinsey Scale, which focuses on
sexual behavior. This scale is bipolar and unidimensional,
ranging from exclusive heterosexuality to exclusive
homosexuality. In a recent study, Storms (1978), proposed a
new conceptualization of sexual orientation. Paralleling
changing conceptions of femininity and masculinity, Storms’
Erotic Orientation Scale (EROS) is based on a two dimensional
orthogonal model in which hetero-erotic and homo-erotic
orientation vary independently. The EROS does not assess
behavior, but arousal, attraction and fantasy. Use of the
scale with college populations confirms the independence of
homoerotic and heteroerotic orientations, and further shows

that arousal is highly correlated with behavior as measured

by the Kinsey Scale.

while our contemporary views now recognize the diversity
and variablility which characterize psychosexual life, we
have relatively little understanding of how these aspects of
social and sexual self-expression are interrelated, or of how
they develop. The theoretical perspective of this research
is that psychosexual life is composed of multiple factors

which are interrelated in complex ways, not yet understood.
Theories of Gender and Sexual Orientation Development

Theories which attempt to explain the origin and
development of gender-role and sexual orientation can be
broadly classified into five main areas: Biological; Social
Learning Theory; Cognitive Developmental; Psychodynamic; and
Sociological.

In western history, in the two centuries prior to Freud,
sex-role and sexual orientation were believed to be
biologically determined. Homosexuality and cross-sex
behaviors were viewed as aberrations caused by genetic and
biological ’taint'. The relationship between biology and
psychosexual development has been the focus of research by
John Money (1578), who has demonstrated that biology and
physiology are not paramount in the determination of gender
identity, femininity/masculinity, and sexual orientation. In
his studies of cases of missassigned gender (individuals who
had been misclassified as female or male at birth), honey
found that individuals develop psychologically according to

their sex of assignment, not their biological sex. Thus

broadly speaking it is learning and socio-psychological
factors which determine psychosexual development.

Social Learning Theory asserts that gender-role and
sexual orientation are learned behaviors, learned largely via
differential reinforcement during childhood and adolescence.
within the behaviorist model, sexual orientation is viewed as
a generalized drive which is channeled in a specific
direction through environmental and socialization pressures.

In cognitive developmental theories the child is seen as
an active participant in the socialization process. Through
the stages of cognitive maturation and development, the child
strives to master the culturally and socially prescribed
behaviors and attitudes which correspond to the behavior of
others toward the child, and the child’s awareness of
sex-appropriate traits, and sexual interests.

Psychodynamic theories posit various interpersonal and
intrapsychic factors as interactive determinants of
psychosexual development. Sociological theories extend their
analyses to include broader societal and cultural patterns,
with an emphasis upon the functional arrangements for the
sexes in society. These theoretical perspectives are
relevant to the two theories which are examined in this

research, and will be discussed in greater detail.

Psychoanalytic Theories

It was Freud who first seriously sought to explain the

psychological development of gender and sexual orientation.

His views, especially on female development, underwent
numerous revisions, and subsequent theorists have further
revised and expanded his concepts. All psychoanalytic views
of gender development and sexual orientation focus on
experiential determinants engendered in parent-child
relations, and the resolution of unconscious conflicts of
psychosexual development during the formative periods of
infancy and early childhood.

In Freud’s theory of personality development,
psychosexual development is central to all personality
organization and functioning. Freud saw sexual orientation
and gender-role as highly related and interdependent
phenomena. He held that ’normal' development led to
psychological femininity and heterosexuality for women, while
certain forms of ’pathological’ parent child relations and
intrapsychic processes are responsible for masculine
personality characteristics or homosexual orientation in
women. Yet Freud asserted that the human being is
constitutionally and psychologically bisexual. By this he
meant that all individuals are essentially capable of both
homosexual and heterosexual interests and behavior. In the
earliest years of life Freud saw the child as capable of many
and diffuse forms of erotic pleasure and personality
expression. Through the stages of psychosexual development
the child's personality and sexuality move toward greater
specificity.

During the first two stages, the oral and anal,

development is the same for both sexes. The development of

gender characteristics and sexual orientation diverges during
the third, the phallic stage, and is precipitated by sex
differences in anatomy. The young girl enters this stage
with mother as the primary object of both love and
identification. The oedipal or electra complex begins for
the girl with her discovery that she lacks a penis. This and
the discovery that her mother too is deprived of a penis lead
her to turn against her mother in anger for perceived
castration. This opens the way for oedipal attachment to
father who is desired because he possesses the ’prized
organ’. A hostile identification with mother results as she
is now a competitor.

when the girl realizes she cannot possess a penis, she
gives up the wish and substitutes the wish for a baby. A
successful resolution involves seeing father, and later other
men as objects of love, and mother as the object of
gender-role identification. Freud stated that this process
leads to the formation of three key personality traits for
women; passivity, masochism, and narcissism. The females
displaced wish for a penis-~to be penetrated, impregnated,
and give birth are seen as passive and masochistic.

Any departure from this posited ’normal’ development is
seen to predispose the girl to variant forms of sexual and
gender-role expression. In Freud’s formulation the
psychodynamics of female masculinity and female homosexuality
are essentially the same (involving either or both of two key
dynamics--masculine gender identification and or pre-oedipal

attachment to mother as love object) though they may result

10

from a number of different courses. Homosexuality and
masculinity may be the result of the girl's failure to accept
her lack of a penis, and hence a failure to identify with
mother. As such she remains fixated at the stage of
pre-oedipal attachment to mother as a love object, and forms
a masculine gender-identity and sex-role. Or later, an
inability to transfer her attachment to father as love object
to other men, or disappointment in heterosexual love may lead
to a regression to female object choice. Thus masculine
gender traits and homosexuality in women are viewed as a
failure to appropriately enter, or resolve the oedipal
conflict, concomitant fixation or regression to maternal love
object choice and/or masculine gender-role identification.

Criticisms of the psychoanalytic formulation of female
development have focused on what causes the young girl to
switch from maternal to paternal love object choice. Freud’s
theory of penis envy has received harsh criticism. The
notion of penis envy was based on Freud’s assumption that the
penis is anatomically superior to female genitalia. flany
psychoanalysts reject this assumption in favor of a view of
penis envy as symbolic of female envy of the male position of
power and privelege in society.

Freud’s theory of psychosexual development, which is the
cornerstone of his theory of personality, is flawed by his
patriarchal bias. His failure involved his inability to
perceive the more complex and powerful role of mother in
early childhood, in both interpersonal and‘intrapsychic

realms. Instead, Freud manages to place father in a pivotal

11

and powerful position in early child develOpment. Thus
patriarchy attempts to prevail in one of the few domains of
human experience which is in fact dominated and organized
around women, and mothers. Feminist critiques of
psychoanalytic theory view these constructs as manifestations
of universal unconscious and defensive dynamics. The
importance and power of mothers is denied precisely because
mothers are so powerful in the phenomenology of early life.

The three central constructs of the early psychoanalytic
theory of female gender and sexual-orientation development
are penis envy, the oedipal conflict, and identification with
mother. More recent approaches to early child development in
psychoanalytic and other camps have emphasized the dynamics
of the relationship with mother, and other interpersonal
factors as more probable determinants.

Charlotte wolff (1971) suggests a variant of the
psychoanalytic model based on her extensive interview
research with lesbian and heterosexual women. The important
dynamics which lead to the choice of a female or male sexual
orientation are as follows. The young girl develops a
normal, intense attachment to her mother in early childhood.
when she realizes that mother values males more highly than
females, she experiences insecurity about her own value as a
female, and strives to compensate for this, and to preserve
or attain mother via one of two routes. The first involves
an ingratiating and emulating position in which the girl
chooses femininity and to be like her mother, in this manner

she possesses mother by internalizing and identifying with

18

her. In becomming feminine the girl strives to attain that
special position of primary intimacy and value with father
and other men, that she cannot attain with her mother. Thus
the male object choice is a secondary and substitute one.
The girl’s anger at her depriving mother motivates her to
compete with mother in an attempt to outdo her. This
developmental pathway results in insecurity about ones real
worthiness to be loved, and loneliness. As a substitute
choice, men can never provide the woman with the maternal
love she continues to crave. Her lonliness is alleviated
through intimate relationships with other women. wolff notes
that throughout childhood young girls engage in much physical
intimacy and affection, and later emotional intimacy
characterizes female bonds throughout the lifespan.

The route chosen by the prelesbian girl is different.
She strives to be worthy of mother’s love, and therefore
emulates masculinity. In so doing she strives for an
independent personal value and worth. The lesbian woman may
be viewed as immature in her intimate sexual-relations, but
she is seen as mature in her striving to be independent from
male superiority and domination. The heterosexual woman
remains immature in her relationships with men while
developing mature capacities for relatedness with other
women.

wolff’s attribution of the pertinent variables is
suggestive of the manner in which these developmental
processes are now viewed by many. Current thinking about

psychosexual development involves an integration of

13

sociological and psychological perspectives. with what is
essentially a reformulation of Freud’s theory of psychosexual
development and the oedipal conflict, gender-role
characteristics as well as other aspects of personality are
importantly seen to develop in the context of attachment to
and separation-individuation from mother, and the various
intrapsychic, interpersonal, and familial socialization
forces at play. These current views take as their point of

departure sociological analyses of family structure.

Sociological Perspectives

Sociological perspectives address feminine and masculine
personality attributes through an analysis of the
differential functional roles women and men perform within
the structure of the family and society. Herein a variety of
theorists have proposed conceptualizations of adult gender
role functioning that focus on a key distinction, that which
Talcott Parsons describes as the Instrumental-Expressive role
dichotomy.

Talcott Parsons (1951) asserts that women, and
especially mothers, engage in expressive behaviors,
performing nurturing, supportive and integrative functions in
the home. Fathers, in contrast are seen to serve primarily
the instrumental functions in the family. Ifathers, whose
primary duties vis a via the family, take them outside the
home and into society, are the link with the outside adult

world. They teach children of society and social norms.

13

while mothers love is relatively unconditional, father’s love
and regard is contingent upon the child’s fulfillment of
expectations and demands geared toward fostering the childs
independence and maturity. As such fathers are seen to pry
children loose from dependency upon mother.

This dichotomy has been challenged with the assertion
that these differing functions are to a large degree shared,
and that mothers perform a multitude of instrumental tasks in
rearing children. Additionally, the family has changed
dramatically since 1551, and a large percentage of mothers
are now employed outside the home. However, the essence of
Parsons’ role dichotomy remains true, as mothers continue to
be the primary nurturers and caretakers of children. And
despite the changes in gender roles of recent decades,
empirical studies continue to demonstrate that children
perceive their parents as differing along these dimensions
(Baumrind, 1575; Spence & Helmreich, 1575).

A similar distinction has been proposed by David Bakan
(1565). Bakan’s concepts of agency and communion have come
into wide usage, and these terms are often considered
synonomous with masculinity and femininity. Bakan describes
the dichotomy as follows:

I have adopted the terms ’agency’ and ’communion’

to characterize two fundamental modalities in the

existence of living forms, agency for the existence

of an organism as an individual and communion for

the participation of the individual in some larger

organism of which the individual is a part. Agency

manifests itself in self-protection, self-

assertion, and self-expansion; communion manifests

itself in the sense of being at one with other
organisms. Agency manifests itself in isolation,

15

alienation, and aloneness; communion in contact,

openness, and union. Agency manifests itself in

the urge to master; communion in noncontractual

cooperation. Agency manifests itself in the

repression of thought, feeling, and impulse;

communion in the lack and removal of repression.

In their theories of gender development Nancy Chodorow,
an anthropologist, and Dorothy Dinnerstein, a cognitive
psychologist, assert that these two fundamental differences
in personality functioning described by sociologists arise

through early childhood object relations development.

Object Relations Theories

A variety of contemporary theories of psychosexual
development owe a tremendous debt to the works of Melanie
Klein (1538), Edith Jacobson (1565), and Hargaret Mahler
(1575). Each of these women have offered elaborate theories
of early personality development which focus on the
experiences of the early mother-child relationship. Their
theories are referred to as object relations theories. In
what amounts to a revolution in psychoanalytic thought,
Freud’s structural model of the psyche is in recent decades
being usurped by object relations views. Psychoanalytic
thinkers are increasingly adopting object relations
constructs as central to the understanding and explanation of
personality development and functioning.

Althea Horner (1551) defines object relations as

follows:

18

The term ’object relations’ refers to specific
intrapsychic structures, to an aspect of ego
organization, and not to external interpersonal
relationships. However, these intrapsychic
structures, the mental representation of self and
other (the object), do become manifest in the
interpersonal situation. That is, ”the inner world
of object relations determines in a fundamental way
the individual’s relations with people in the
external world. This world... is basically the
residue of the individual’s relations with people
upon whom he was dependent for the satisfaction of
primitive needs in infancy and during the early
stages of maturation” (Phillipson, 1555, p.7)
(Quoted from Horner, 1551, p.3)

Horner goes on to summarize development as follows:

For us to understand the concept of object
relations and its complex ramifications, it is
helpful to consider how these psychic structures
come into being, to think in terms of the early
mental processes by means of which the newborn
infant organizes its world into meaningful
patterns. One basic pattern is that of the
self--the self-representation--while another is
that of the object--the object-representation. The
object refers to the primary mothering person or
persons in the environment of the infant and the
very young child. The structural and dynamic
relationships between the self-representation and
the object-representations constitute what we refer
to as object relations.(Horner, 1551, p.&)

An increasing number of theorists have come to view
object relations development as the central organizer of
personality development. The development of personality
traits are seen as occuring within, and being mediated by the
matrix of object relations representations. Thus early
social-relational experiences are internalized, and comprise
the structural and dynamic patterning of psychological and
personality organization and functioning.

flany object relations theorists (including Horner,
Jacobson and hahler) have focused on the developmental

process in which the infant and toddler internalize and

17

organize personal and interpersonal experiences. Hargaret
hahler in her book The Psychological Birth of the Human
Infant (1550), provides a clear explication of this process
which she calls Separation-Individuation. Mahler’s theory is

a stage theory of earliest intrapsychic and interpersonal

personality development; it is derived from her extensive
empirical observations. Described briefly, she asserts that
the infant begins life in a stage she refers to as autism,
where experience is largely undifferentiated. As cognitive
maturation ensues, the infant attaches to the mother and
experiences self and mother as one, this second stage is
variously referred to as ’dual unity’, ’primary
identification’ or for Hahler, ’normal symbiosis’. In the
third stage the separation-individuation process begins.
This process is viewed as consisting of two distinct though
intertwined aspects. Separation refers to the achieving of a
sense of separateness, an independent sense of self and
identity. Individuation in this framework represents the
acquisition of individual characteristics, capacities and
components of personality and ego functioning. (The term
’differentiation’ refers essentially to both aspects of
separation-individuation.) In a hahlerian framework it is
this task, and not the Oedipal Conflict, which is most
critical in early personality formation.

Though hahler does not address gender differentiation in
the context of development in the early mother-child
relationship, Nancy Chodorow and Dorothy Dinnerstein do.

Nancy Chodorow (157%, 1575) and Dorothy Dinnerstein (1577)

18

have sought to explain the development of feminine and
masculine gender personality differences through the
differential quality of mother-child relation, identity
internalization, psychodynamics, and socialization variables
experienced by the child in the process of attachment to and
separation—individuation from mother. These theorists feel
that universal sex-role differences are engendered and

perpetuated in female-mothering.

Nancy Chodorow’s Theory

Chodorow asserts that the universal phenomenon of female
mothering impacts in differential ways on the formation of
female and male gender development. She suggests that
personality structure and deeply ingrained affect-laden
personality traits that are central to an individual’s
consistent sense of self are derived from early attachment
to, and differentiation from mother.

The infant begins life in a state of dependence, usually
upon a female mother. This state is one of primary
identification with mother. The infant does not
differentiate her/himself from mother, with whom she/he
experiences a sense of oneness. Attenuation of this primary
identification with mother occurs with maturation and the
childs increasing cognitive and emotional awareness of
her/his autonomous existence. The early experience of

attachment to and differentiation from mother is different

for girls and boys.

18

Chodorow sees the mother of the female child as
experiencing a double identification (Deutsch, 1555). She
identifies both with her own mother as her maternal self, and
with her daughter as herself. (Mothers say they identify
more strongly with their female children, Chodorow, 157%.)
This establishes a more potent mother-child identification
through which primary identification and infantile dependency
are facilitated.

The young boy too begins with a primary identification
with mother. The boy’s first identification--with a female
mother--must later be supplanted by masculine gender
identity. The sources of masculine identity for
internalization are father, and other salient males, but in
early life children have very little contact with men. Thus
the primary source for learning masculine identity is
masculinity socialization as oppositional to the experience
of mother, and father/masculinity as invoked by mother.
Masculinity develops, according to Chodorow, because ”Mother
treats son in a manner which emphasizes his differentness
(masculinity) in Opposition to herself" (Chodorow, 157%,
p.%5). Father in this context is an abstraction, masculinity
a role, a set of prescribed principles and behaviors} The
source of identity is not relational, but positional. ”For
the boy, the tie between affective relations and masculine
gender identification and role learning (between libidinal
and ego development) is relatively attenuated” (Chodorow,
1575, p.55). The boy’s identity is a positional one, a

fantasized identification with a role which requires a

80

rejection of the emotional, interpersonal experience of
mother. Since there is no personal relationship with father,
masculine identity is formed by negation, denying and
denigrating anything feminine in self and others.
(Dinnerstein sees this as the source of the universal
denigation of women). For Chodorow, the Oedipal conflict is
the process through which the boy later and finally denies

and represses his attachment to mother, and identifies with

father. Chodorow concludes:

To summarize, four components of the attainment of
masculine gender identity are important. First,
masculinity becomes and remains a problematic issue
for a boy. Second, it involves denial of
attachment or relationship, particularly of what
the boy takes to be dependence or need for another,
and differentiation of himself from another.

Third, it involves the repression and devaluation
of femininity on both psychological and cultural
levels. Finally, identification with his father
does not usually develop in the context of a
satisfactory affective relationship, but consists
in the attempt to internalize components of a not
immmediately apprehensible role. (157%, p.51)

Feminine identity development for the girl is continuous
and personal. It is engendered in the context of an
interdependent and emotional relationship with the person
with whom the young girl has experienced her earliest and
most potent relationship. Here, early development of gender
characteristics does not involve a rejection of emotional
closeness and interdependence. The development of
gender-role is a continuous process built upon the foundation
of early primary identification and dependency.

Femininity and female role activities are

immediately apprehensible in the world of her daily

life....5ecause her mother is around, and she has
had a genuine relationship to her as a person, a

81

girl’s gender and gender role identification are

mediated by and depend upon real affective

relations. (Chodorow, 1573, p.51)

Embedded in these differing identification processes are
factors which mediate and produce gender personality and
sex-role traits. The mother-daughter relationship itself is
internalized as feminine ego-structure, and therefore
intrapsychic object relations are interpersonal in nature.
mother is internalized both as an ’object’ in relation to
self and as a ’self’ object. The feminine object world is
characterized by identification with the other and emotional
interdependence. Feminine identity than involves a greater
interpersonal and empathic orientation, and less
differentiation between self and other. Masculine identity,
by contrast consists of an abstract positional self and a
denial of affective identification. The behaviors and traits
which are fostered in the young boy are a set of prescribed
principles for behavior, and amount essentially to role
learning. Consequently psychological masculinity involves a
more radical differentiation and separation of self and
other. while a boy is encouraged and pushed to
differentiate, a girl is encouraged to remain enmeshed.
Ego-boundaries are thus firmly established for the boy,
independence of self and other are clearly demarcated.
Ego-boundaries are highly permeable and more diffuse for the
girl.

Chodorow asserts that exclusive female mothering, and
the differential behaviors of mothers toward their daughters

and sons are the critical antecedents in this developmental

88

process. Mothers reward, foster and encourage in multiple
ways differing behaviors from their daughters and sons, with
concomitantly differing ramifications for the child’s
interpersonal and psychodynamic experience and development.
Mothers reward and foster the emulation and imitation of
themselves, i.e., of feminine and maternal traits in their
daughters. Mothers encourage and foster dependency in their
daughters, enabling them to remain emotional, dependent, and
vulnerable, until much later in childhood. In their
continued responsiveness to their daughters’ dependency
needs, they maintain the potency of the mother-daughter
union. In contrast, mothers push their sons to separate and
individuate. Mothers are less apt to gratify their sons
dependency needs, more apt to encourage the inhibition of
emotional dependency and vulnerability, and in general foster
self-reliance and independence in their sons.

The Oedipal stage, with its ’normative resolution’ in
heterosexual erotic-orientation, proceeds differently for the
sexes. The development of heterosexual object-choice for the
male is essentially continuous. His internalized love-object
choice remains one which embodies characteristics of
mother--females and femininity. The girl’s erotic
orientation development is more complex. Her primary
love-object is mother, she turns to father, according to
Chodorow, in an attempt to escape her unity with and
dependence upon mother. But the girl is never completely
free of her early libidinal attachment to mother, her oedipal

stage attachments are to both her mother and her father.

83

Though her erotic and genital cathexis may become
predominantly male-oriented, her affective interests remain
dual.

In contrast to the boy, whose erotic and affective
object is one--mother and then other females, the girls
internal object world is more diffuse and complex. Father
for children of both sexes is more distantly experienced both
because his work takes him out of the home, and because his
own development and socialization are characterized by denial
and constriction of emotional expression. Thus relationships
with men for both women and men are less directly emotional.
The young girl does not find her relationship with her father
capable of fulfilling her needs for mutual affirmation and
love, it does not replace her relationship with mother.
Mother and women remain the primary source for emotional
relationships for both sexes. Uomen, throughout life, have
closer personal ties and effectively richer involvements with
relatives, other women and their children.

The internalized object world which defines psychic
structure and concomitantly external interpersonal relations
are different for females and males, in structure, quality
and intensity. Masculine identity consists of an object
independent, highly differentiated, and effectively insulated
identity. Identity and ’self’ for the male is less
relationally defined. Masculine personality development
outfits the male for an affect-free mode of involvement in

public and work spheres where interpersonal ties are based on

8%

role enactments, on contractual and universalistically
defined role expectations.

For girls, feminine development is derived from
interdependent, emotional and relational identification with
mother. Chodorow proposes that an object-dependent and less
differentiated object world is defining of self and identity
for women. In psychoanalytic terms women have more flexible
ego-boundaries, and a less individuated identity. women’s
development prepares them for mothering, for serving the

expressive and communal functions in the family.

Dorothy Dinnerstein

The theory of gender development offered by Dorothy
Dinnerstein shares many parallels with the writings of Nancy
Chodorow. Dinnerstein in her book, The Mermaid end the
Minotaur: Sexual Arrangements and Human Malaise, sets forth
a broad descriptive framework in which to understand her
concepts. She too takes as her primary point of departure

the functional asymmetry of the social-familial roles for

women and men.

Until quite recent technological developments
opened new possibilities (so far largely unused),
most of the world’s women have been obliged to
invest major energy in the biological task of
perpetuating the species. This reproductive task
has tended to make them specialists in the exercise
of certain essential human capacities, capacities
crucial for empathic care of the very young and for
maintenance of the social-emotional arrangements
that sustain primary group life. (1577, p.50)

Dinnerstein asserts that many elements of adult

personality, including emotional proclivities and sex-typed

85

modes of interpersonal relating are derived from both the
actual and the desired-fantasized qualities of the
prerational experience of mother. She see’s the
characteristics of sexual relationships as redintegrating the
early mother-child bond. Though this author acknowledges the
importance of ongoing socialization and learning as major
determining factors in this area of psychological
development, her focus is on the psychodynamic and object
relations developmental phenomena which characterize the
earliest years of life.

Dinnerstein’s exposition begins with a formulation of
early infant experience drawn from Melanie Klein’s theory of
early object relations development (1532). During early
infancy mother is experienced as the ground of existence, in
Primary identification, there is no differentiation of self,
mother, world. Experience is an unbounded oneness wherein
everything is holistically experienced as mother. Affect is
unmitigated and as such pleasure and pain are holistically
experienced. All that is positive and pleasurable is
internalized as an intrapsychic constellation of the ’good
mother’, while frustration, pain, and displeasureble
experiences form the constellation of ’bed mother’. As the
child matures mother begins to be differentiated out and
perceived as a separate, autonomous, intentional entity, and
a creature whose intentions are not always geared toward
giving the infant pleasure. A source of frustration for the
infant, she/he wishes and desires to exclusively possess the

good-responsive mother, one whose intentions are not

88

independent and never counter to the immediate needs and
desires of the infant. Here, Dinnerstein asserts that this

wish for exclusive possession of an all-good mother undergoes
divergent transformations for the two sexes. Girls who are
encouraged to imitate maternal behaviors and emulate mother,
resolve their desire to possess the ’good mother’ by becoming
her: The female gratifies her wish for possession vie
intrOjection and identification. For the woman’s recreation
of herself as mother she needs affirmation through another
who seeks her nurturance and succor. Mothering is an
identity which is dependent upon relation with another. The
male wish for the ’good mother’ is gratified more directly in
his intimate relationships with women. Males thus do not
experience the restrictions and limitations attendant to
their gender identity development.

The potency of the experience of mother is derived from
its primacy during the child’s preverbal, prerational period.
Maturation and increasing autonomy are experienced with both
the inherent joy at newfound capacities for mastery,
competence, and self-reliance, and simultaneously pain of the
lost oneness with mother. By the time father becomes
important in the child’s object world, cognitive maturation
engenders a very different phenomenological matrix in which
he comes to be known. Father does not pose the threat of
regressive merging. Father is idealized both through
displacement of the ’good mother’ and because he does not
represent the source of the conflicts and ambivalence the

child is now experiencing. with a degree of separation

87

achieved, entering into a relationship with father provides
the child with a safe alternative to mother and fosters the
process of separation-individuation. In a mode of
experiencing which is now verbal, rational, and relatively

differentiated, the child approaches father to engage in a

more mature relationship.

Dinnerstein sees father as important for heteroerotic
socialization of the young girl. It is he who elicits the
shift in her erotic alliance from maternal to paternal
object-choice. This is achieved by virtue of the fact that
father offers his daughter the desired non-infantile
relationship, which is construed in seductive, erotic terms
by father. Father acts sedutively toward his daughters, the
offer to leave mother is for erotic love, and to become like
mother in relation to another.

Dinnerstein muses that it is perhaps the cultural taboo
against homoerotic experience that prevents father from
displaying sensual and affectional behavior toward his son.
Heteroerotic or gynoerotic experience is continuous for the
male. Father fosters his son’s individuation from maternal
identification with offers of friendship, comraderie, and
access to the privilege of power in both the home and public
world.

The girl’s oedipal experience is more complicated. She
experiences both homoerotic and heteroerotic pulls, and dual
rivalry with both parents.

A variety of aspects of adult gender functioning are

cast in these early years. Both sexes strive to actualize in

88

women the wished for good mother, and to deny in women traits
which threaten or compromise the fantasy of an all good

mother. women are universally denied their full independence
and individuality because they represent our early wishes for

a mother who exists solely to gratify our needs, and our

frustration with the real autonomous and uncontrollable
mother--the loss of passive, effortless wish-fulfillment.
Thus both sexes strive to actualize in women the wished for
good mother, and to deny in women traits which threaten or
compromise the fantasy of an all good mother. Hence women as
mothers form an identity based on gratifying the needs of
others. Dinnerstein proposes that both women and men
encourage women to emulate personality attributes which
reflect our early notions of the good mother as an
all-giving, all-providing, nurturing entity whose behaviors,
thoughts, and feelings are always geared toward the
fulfillment of the needs of others.

The experience of sexual intimacy redintegrates the early
experience of mother. Sexual intimacy comprises a highly
tactile, intensely pleasurable experience with varying
degrees of interpersonal merging. Dinnerstein believes that
the early mother-child union is the prototype of
erotic-sexual intimacy. She describes at length the pleasure
associated with intimacy with mother in infancy, pleasure for
which intensity is derived from the phenomenological nature
of consciousness at this time of life, e.g. effects are

unmitigated. In adult life, the experience of romantic love,

655 EEOIIQ BlOOSUFe redintegrate this early experience.

88

Critique and Overview

These two theories of gender development converge on a
number of points. Both Chodorow and Dinnerstein seek to
explain gender-role asymmetries, i.e. the tendency of women
to express feminine traits and the greater expression of
masculine traits in men. Both authors assert that
gender-role asymmetries exist and are perpetuated because: 1)
Females are the primary and often sole mothers/caretakers of
children; 5) women bring to their mothering their own
asymmetrical psychological organization and functioning
relevant to gender issues; and finally 3) the asymmetries
result from the psychodynamic and object-relational phenomena
that each author hypothesizes as operating within the child
vis a via the mother. In this regard both authors concur
that the mothering of the female child fosters an
identification with mother, and the internalization of modes
of socio-emotional functioning based on maternal and
mother-child relationship dynamics. In contrast mothering of
males fosters an identity which is non-maternal, more
separate, differentiated, independent, and less related.

Chodorow and Dinnerstein postulate two distinctly
different modes of development for the two sexes. The many
factors which they describe as leading to gender
differentiation can be viewed as dimensions which they

characterize as functioning in polarized ways across the

30

sexes. However both common sense and empirical research
indicate that many of the inter- and intra-personal factors
addressed in these theories are highly complex and varied
both within and across sax. For example maternal behaviors
and the feminine and masculine personality traits which the
theories seek to explain are not unitary phenomena, but are
complex, multifactored and variable within as well as across
biological sex. Thus while the authors offer
conceptualizations which are dichotomous and polarized, the
phenomena they describe are more accurately viewed as
multivariate factors which vary along continuous dimensions.

Both theories offer causal models to explain gender
development. Framed simply, both theories postulate:

1) That mothers respond differently to their female and

male children.

8) That the differences in maternal behavior produce
differences in intrapsychic structures and
personality organization in the child.

3) And that these differences in personality
organization and intrapsychic structures produce
differences in gender attributes.

Implicit in this first postulate is the assumption that
universally, the biological sex of the child is a major and
key determinant of mother’s behavior toward her child.
Empirical research to data provides some support for this
general assumption, there is evidence that many differences
in maternal behavior are linked to the biological sex of the

child. (For reviews of this literature see Donelson a

31

Gullahorn, 1577; Hyde a Rosenberg, 1556; and Unger, 1575.)
However many other factors impact on the early mother-child
relationship including the personality traits of mother and
infant temperament. These factors lead to substantial
differences in maternal behavior and subsequent mother-child
interactions. If the psychodynamic processes described by
Chodorow and Dinnerstein do contribute to psychosexual
development, then it is maternal behavior and interaction
with child, regardless of origin or precipitator which should
account for differences in psychological femininity/
masculinity and sexual orientation both within and across
biological sex.

Psychological femininity and masculinity have been shown
to be multidimensional for females and for males. Spence and
Helmreich have identified sets of gender-role attributes
which are independent and may coexist, and sets which are
bipolar and mutually exclusive (1575). while as groups women
and men tend to display the stareotypically expected
differences in femininity and masculinity, within sex
variation is greater than across sex differences. In their
theories Chodorow and Dinnerstein do not address this
variability. Again, if the developmental processes they
describe are correct, these should explain the variation in
feminine/masculine traits among women alone.

Neither of the theories address variation in sexual
orientation in women or in men. Their goal is to explain
heterosexuality, and traditional gender-role development.

Nancy Chodorow does not offer a clear exposition of the

38

interpersonal and intrapsychic factors which lead to sexual
orientation, her primary focus is upon gender
characteristics. Dinnerstein’s theory does offer a
formulation of the process which leads to sexual-affectional
interests in women or men, though again she describes only
opposite sex attraction. Her explanation of the pertinent
dynamics is essentially the same as the process delineated by
Charlotte wolff. Thus wolff and Dinnerstein concur that the
child’s desire for affectional closeness and intimacy with
mother is gratified either by internalizing, emulating and
identifying with mother; or striving for intimacy with other
women.

Both theories are further limited by their failure to
address varying forms of the family system. Chodorow and
Dinnerstein begin with the assumption that children are
raised in two parent nuclear families characterized by
traditional parental roles, i.e., father is the sole wage
earner while housewife-mother is the primary caretaker and is
not employed outside the home. Recent statistics indicate
that this form of family system constitutes only 15% of the
families in the United States today. The family has changed
dramatically in recent decades. Many children are raised in
single parent families, and dual wage-earner families. Both
women’s and men’s roles in the family and in society are
changing. women who are employed are engaged in a broader
range of activities outside the home and are less exclusively
involved in the primary care of children. Roles appear to be

changing for men too. Trends in the media suggest that

33

fathers are becoming more involved with their children in
more nurturing ways. And though the numbers are small, an
increasing number of fathers are assuming the primary
caretaking role.

While Chodorow and Dinnerstein address only the relative
development of females vs. males, this writer reasons that
their theories can be extended to the relative differences in
development among women alone. This research study examines
the relationships between the variables these theories posit
as important in development across the sexes, in just women.
The interrelationships between early childhood
mother-daughter involvement, degrees of identification with
mother, differentiation of object relations representations,
femininity/masculinity and sexual orientation are assessed.

The empirical research on women’s gender and sexual
orientation suggests that these constructs are worthy of

investigation.

3%

REUIEU OF THE LITERATURE

Empirical Studies of women’s Gender Role Development

Since the 1550’s many researchers have investigated the
structure of gender roles in our society and the innumerable
variables which determine, perpetuate and maintain gender
roles. Most of the empirical studies of gender development
examine the differing socialization experiences of girls and
boys which might account for the differential development of
gender-role personality traits across the sexes.

The many differences described by Chodorow and
Dinnerstein in mother’s behaviors toward daughters vs. sons
are broadly supported by numerous research studies. Both
mothers and fathers have warmer, closer, and more
affectionate relationships with their daughters from the time
that their daughters are about six months of age (Block,
1575, 1555; Goldberg a Lewis, 1555; Lewis, 1578; Moss, 1557).
Both mothers and fathers encourage their daughters to be more
dependent and less autonomous in their behaviors (Block,
1575, 155%; Collard, 155%, cited in Donelson & Gullahorn,
1577; Lewis, 1578; Maccoby & Jacklin, 157%). And mother’s
engage in more childrearing practices which teach and foster
empathy and prosocial behaviors with their daughters than

with their sons (Block, 1575, 155%; Yarrow 8 maxler, 1577).

35

Comparatively few studies have examined the relation-
ships between parenting and family variables and the
differential development of feminine and masculine
characteristics in women. However, investigations from a
range of different areas do illuminate some of the factors
which are associated with differences in femininity and
masculinity among women. These areas of research include
studies of parent-child interactions, studies of achievement
motivation and maternal employment, among others.

Converging evidence from a variety of different studies
indicates that high degrees of maternal involvement,
nurturance, and warmth are associated with a lack of agentic
characteristics in daughters (Baumrind, 1575; Hoffman, 1575).
Studies have found that high levels of maternal warmth and
protectiveness are associated with personality traits
including passivity, conformity, and low achievement
orientation (Crandall & Battle, 1570, cited in Donelson &
Gullahorn, 1577; Kagan 3 Moss, 1558, cited in Unger, 1575).
Moderate levels of maternal involvement and warmth are
associated with greater independence and dominance in
daughters. High achievement orientation in girls and women
is related to moderate, not high levels of maternal
involvement, affection, and protectiveness (Anderson, 1573;
Bardwick, 1571; Hatfield, Rau & Alpert, 1557).

Another area of research which has examined family
variables and gender-related characteristics in daughters is
studies of maternal employment. These studies were often

undertaken to assess whether maternal employment outside the

38

home had detrimental effects on childrens’ development.
Findings of numerous researchers indicate that maternal
employment has positive effects on daughters. Daughters of
working mothers are more androgynous (Seigel, et.al., 1553,

cited in Hyde & Rosenberg, 1550); and are higher achievers

and have higher career aspirations (Almquist 8 Angrist, 1570;
Astin, 1555, cited in Donelson 8 Gullahorn, 1577; Astin,
Suniewick & Dweck, 1571, cited in Donelson & Gullahorn, 1577;
Marecek, 1575, cited in Unger, 1575).

The relationship between the gender role-characteristics
of parents and their daughters is complex. Parents who are
strongly sex-typed in the traditional asymmetrical pattern
have daughters who are most likely to be highly female
sex-typed (Spence & Helmreich, 1575). Androgyny in either
mother or father is associated with androgyny in daughters
(Spence & Helmreich, 1575). Research findings suggest that
the degree to which fathers are masculine sex-typed is highly
correlated with the level of sex-typed femininity in
daughters while mother’s gender role attributes are less
strongly related to daughter’s (Heatherington, 1557; Mussen &
Rutherford, 1553; Fling 8 Manosevitz, 1578). However,
daughters who identify with a feminine sex-typed mother are
most likely to be feminine sex-typed. (Douven & Adelson,
1555).

Gender-role attributes are significantly associated with
socio-economic status. Low and working class SE5 is related
to higher frequencies of traditional asymmetrical sax-typing

in both parents and children, while middle and upperclass SE5

37

is associated with more androgynous and less stereotyped
gender-role traits and attitudes (Hall I Keith, 1555;

Nadelman, 157%; Spence & Helmreich, 1575).

Empirical Studies of women’s Sexual Orientation

Empirical studies of female sexual orientation have of
course focused on attempts to understand homosexuality. To
this end homosexual women have been the primary focus of
study, while heterosexual women are often included as the
comparison or control group.

Two major methodological problems beset studies of
homosexuality. The first concerns the definition and
identification of who is homosexual. As discussed earlier,
sexual preference is a complex phenomena involving behavior,
fantasy, and affection. Across studies investigators have
used differing criteria ranging from self identification to
differing classifications derived from the Kinsey Scale. The
second problem involves the difficulties in obtaining a
representative sample. From the outset the social stigma
which prevents many homosexuals from publicly acknowledging
their sexual preference leads to biased samples. It is
conceivable that those individuals who publicly acknowledge
thier homosexuality are different than those who do not.

Empirical research on female homosexuality spans a broad
range of content areas. Among these are; incidence,
personality traits and types, psychopathological factors, the

’coming out’ process, and aspects of lifestyles and

38

relationships. Research into the development of lesbianism
has examined factors such as early intrefamiliel
relationships, socialization, learning, and ’critical’

experiences.

Lesbianism and Psychopathology

Until recently much of the focus of the literature and
research on female homosexuality has been on the ’deviant’
and ’psychopathological’ nature of homosexuality. Thus most
of the early studies sought to establish a link between
psychopathology and lesbianism. This pervasive assumption
proved false; on numerous measures of psychological
adjustment lesbian women typically emerge as no more neurotic
or disturbed than heterosexual women (Adelman, 1577; Laner,
1577; Oberstone 8 Suknoeck, 1577; Ohlson and wilson, 157%;
Rosen, 157%).

In many studies lesbians display greater psychological
health and adjustment than their heterosexual pears
(Freedman, 1575; Hopkins, 1555; Siegelman, 1578; wilson &
Green, 1571). In studies which do find significantly greater
disturbance among lesbians, it is primarily alcoholism which
is more prevalent in homosexual women, and authors conclude
that no specifiable clinical feature accompanies same-sex
sexual preference in women (Armon, 1550; Saghir, Robins,
walbran a Gentry, 1570; Swanson, Loomin, Lukash, Cronin, &
Smith 1578; Thompson, McCendlass a Strickland 1571). The

only two studies which have found lesbians to be

38

significantly more neurotic than controls were done in
Britian using British populations (Kenyon, 1555; wolff,
1571). In the study reported by Kenyon, lesbian women
obtained mean scores on the Meudsley Personlity Inventory
Neuroticism Scale which were higher than those of a
heterosexual control group, but lower than those of a mixed
group of English University students upon whom the scale was
standardized.

The preponderance of evidence thus warrants the
conclusion that female homosexuality is not a clinical
entity, and as such lesbianism is best viewed as a
variant—~not a deviant form of sexual expression.

Two broad areas of research which compare lesbian and
heterosexual women are relevant to this project, studies of

personality correlates, and of family background.

Sexual Orientation and Gender Characteristics in Women

The methodological problems which beset research studies
of personality variables are many. Studies which examine
personality traits are potentially biased by socio-cultural
differences important to geographical region, cohort, and
generational effects. Studies often compare lesbians to
heterosexual women who are married and unemployed. These
groups of women have vastly different lifestyles, demands,
responsibilities, and sources of need gratification. In
repeated studies the effects of marital status, employment,

and education are shown to be related to self esteem,

HO

psychological well-being, and femininity/masculinity among
both women and men (See Donelson 8 Gullahorn, 1550; Gove,
1575). Among women, being single and being employed are
associated with androgyny. This is a potential source of
confound for studies which do not control for these
variables.

In studies which have compared lesbian and heterosexual
women on measures of femininity/masculinity, lesbian women
generally emerge as more androgynous than heterosexual women.
In a variety of studies reported in the literature lesbian
women endorse significantly more masculine traits than their
heterosexual peers, while there are often no differences in
the endorsement of feminine characteristics. Significant
differences in feminine traits where they do occur, and
nonsignificant trends show heterosexual women to be more
feminine sex-typed. Many other studies report personality
differences which are relevant to gender attributes.

Frequent findings indicate that lesbian women are
significantly more dominant, assertive, and self-sufficient.
The following review presents the research on personality
correlates of sexual orientation in women in chronological
order.

In the earliest study of gender characteristics in
lesbian women, Terman and Miles (1535) studied femininity and
masculinity using a multiple choice test of 510 items. Their
15 lesbian subjects achieved total masculinity scores
equivalent to those of male M.D.’s and Ph.D.‘s, excelled only

by male athletes. When compared to heterosexual women of

31

similar age and education, the mean femininity/masculinity
scores on a bipolar scale show a difference of 3.8 standard
deviations. These lesbian women showed a high degree of
femininity on word association tasks. All other scales were
within the average range for women.

In 1557, Grygier (cited in Kenyon, 1555b) used a broad
array of measures to study lesbians. Though lesbian women
scored higher than heterosexual women on masculinity scales,
significant correlations with the professional status and
intelligence of subjects explained these differences.

Using a measure devised for his study, Steven Giannell
(1555) found that homosexual women scored significantly
higher than heterosexual women on scales of Autonomy,
Aggression and Need for Nurturance, while scoring lower on
Need for Deference. Homosexual and heterosexual women were
equal on a Succorance scale.

Freedman (1555, 1571, 1575) administered the Eyesenck
Personality Inventory (EPI), and the Personal Orientation
Inventory (OPI) to 55 lesbian women and 57 heterosexual women
recruited from women’s organizations. Though the groups were
matched for educational background, the majority of lesbian
women were single and employed while the heterosexual group
were largely unemployed and married. The lesbian group
achieved significantly higher scores (p<.Ol) on OPI scales of
Self-Actualizing Ualue, Existen- tiality, Feeling Reactivity,
Spontaneity, Acceptance of Aggression, and Capacity for
Intimacy. Lesbian women scored significantly lower on the

EFT lie scale, indicating greater candor and self-disclosure.

%8

Hopkins (1555) using the Cattell 15PF Questionnaire
compared 8% lesbian and 8% heterosexual women matched for
age, intelligence, education and profession. The lesbian
women scored higher on scales of Alert Poise and Independence
(p<.08). Heterosexual women scored significantly lower on
scales of Dominance, Assertiveness, Progressiveness, and
Self-Sufficiency (all p’s <.O8). The lesbian group was also
found to be more tense, driven, reserved, detached, critical,
and cool (p’s <.O8).

Wilson and Green (1571) used the California Personality
Inventory, the Edwards Personal Preference Schedule, and the
EPI in their investigation of %5 heterosexual and %5 lesbian
women. The two groups were matched for age, and all were
college graduates. Lesbian women scored higher on scales of
Dominance, Capacity for Status, Good Impression, Intellectual
Efficiency, and Endurance. The heterosexual group scored
significantly higher on scales measuring Neuroticism and
Femininty (p<.O5).

In a 1571 study, Thompson, McCandless, and Strickland,
compared 5% lesbian and 5% heterosexual women on three scales
of the Adjective Check List (AOL), and a Semantic Differ-
ential. Only one significant difference distinguished the
two groups. Lesbian women achieved significantly higher
scores on the Self-Confidence scale of the ACL.

Siegelman (1578) conducted a study comparing the
responses of 5% lesbian and 133 heterosexual women on the
Cattell Neuroticism Scale Questionnaire and the Crowns and

Marlowe Social Desirability Scale. All participants were

53

either college students or in professsional occupations,
however the heterosexual group were significantly younger.
Lesbian women were less apt to give socially desirable
responses illustrating a more candid response style.
Significant differences (p’s <.OOl) indicate that lesbian
women are more goal directed, self accepting, tenderminded,
and dominant. Heterosexual women scored significantly higher
on scales of Neuroticism, Depression, and Submissiveness (p’s
<.Ol).

Saghir and Robins (1573) interviewed 101 women, %%
heterosexual and 57 lesbians. Only three factors
differentiated between the two groups: Lesbian women
identified themselves as masculine more often, had higher
alcohol abuse rates, and had dropped out of college more
often.

In a study published in 157%, Ohlson and Wilson
recruited 5% lesbian and 5% heterosexual women matched for
age and education. An item analysis of the MMPI yielded 35
interpretable items which significantly differentiated the
two groups. Authors classified the items into clusters, the
lesbian group were distinguised from the heterosexual group
by lower anxiety, fewer physical complaints, a masculine
orientation, social introversion, and endorsement of overt
homosexuality.

Anne Steinmenn, (157%) found that lesbian women as a
group had the highest scores on her devised measure of
Self-Achieving Orientation vs. Family Orientation of several

thousand women worldwide.

%%

Hassell and Smith (1575) administered the AOL to 8%
lesbian and 8% heterosexual women matched for age and
education. Les- bians scored higher on scales of Autonomy,
Exhibition, and Change. Heterosexual women scored higher on
scales of Abasement, Deference. Defensiveness, Order and
Endurance. (All p’s <.O5) There were no differences on
scales measuring Succorance and Self Confidence.

Ward (157%; cited in Spence & Helmreich, 1575) obtained
data from 5% lesbian women on the Spence & Helmreich Personal
Attribute Questionnaire and compared these to the unselected
college sample of women upon whom the measure was
standardized. The lesbian sample scored significantly more
masculine and less feminine on all three scales of femininity
and masculinity. When subscales are combined, lesbian women
display a significantly elevated frequency of androgyny, and
are also more likely to be classified in the undifferentiated
and masculine sextype groups.

McCauley and Ehrhardt (1577) examined the responses
obtained from 15 lesbian women on the WAIS, the Ben Sex-Role
Inventory (BSRI), and Embedded Figures Test. Lesbians showed
significantly higher Uerbal than Performance subscale scores
on the WAIS, typical of the pattern for women in contrast to
men. The lesbian group’s mean score on the embedded figures
test was closer to the male than to the female norms
established for this test. Finally, lesbian women were most
apt to be androgynous on the BSRI.

The BSRI was also administered to a sample of 75 lesbian

women age 15 to 58, in a study conducted by Uan Cleave

%5

(1577). Significantly more of these women endorsed traits
indicative of androgyny, rather than feminine or masculine
sex-typed.

In a study of gender attributes, Spence and Helmreich
(1575) compared 101 heterosexual and 100 lesbian college
women on the Bem Sex Role Inventory. Heterosexual women were
significantly more likely to be feminine sex-typed, while
lesbians were more frequently classified as androgynous,
masculine or undifferentiated.

Shavelson, Biaggio, Cross and Lehman (1550) compared the
responses of 85 lesbian and 85 heterosexual women on the
BSRI. Significantly more lesbians reported attributes
indicative of high masculinity and low femininity, and
heterosexual women were significantly more likely to be
categorized as undifferentiated.

Storms (1550) found significant differences between
lesbian and heterosexual women on one of the three subscales
of the Spence Helmreich Personal Attribute Questionnaire.
Lesbian women achieved higher mean scores on a bipolar scale
of masculine and feminine traits, indicative of greater

masculinity.

Summary

Across studies, lesbians clearly emerge as more
androgynous, ie., displaying traits which include both those
ecscribed as feminine and masculine, than their heterosexual
counterparts. Across measures, lesbian women endorse items

which indicate that they are more dominant, autonomous,

H8

assertive, inner-directed, self-accepting, and self-

actualizing than heterosexual women. They are less

self-abasing, deferentiel, submissive and defensive.
It is not yet possible to determine whether these

differences are attributable to aspects of early personality

development (core personality) or whether they result from
the lesbian’s struggle to integrate her sexual preference and
independent lifestyle in a society which deems it deviant.

While most studies do find group differences between
lesbian and heterosexual women, within group variation is far
greater than between group differences. Authors consistently
agree that the personality profiles of lesbian and
heterosexual women as groups are much more diverse than
unified, comprising a wide spectrum of variation.
Heterosexual and homosexual women and men are heterogeneous
groups. Research on both male and female homosexuality has
recently begun to examine the diversity displayed, and to
search for discernable typologies. Three studies of this
nature are pertinent here.

Clingman and Fowler (1575) asked 185 male and female
homosexuals to classify themselves as predominantly
masculine, feminine, or neither. While most of their
subjects did not subscribe to a gender-role label, the
feminine vs. masculine women scored significantly different
on three scales of the Adjective Check List. Feminine women
had significantly higher Succorance, Abasement and Daference
mean scores, and reported realizing their homosexual

orientation at an older age. The masculine identified women

%7

scored higher on Dominance and lower on Abasement than any of
the remaining groups.

Uance (1577) in an unpublished doctoral dissertation,
interviewed %3 lesbian women, with a focus on the process
through which they became lesbian. The distinction she
proposed was validated. ”Socialized” lesbians were
socialized into a homosexual orientation early in childhood,
showed more rigid masculine gender characteristics, and were
more proficient at masculine behaviors. ”Enculturated”
lesbians adopted a homosexual orientation in late childhood
or adulthood, were more flexible in their gender-role
functioning, and less adept at masculine skills. Uance notes
that despite these differences, both groups were effective in
both feminine and masculine behaviors.

Results of these two studies concur, indicating that the
more masculine lesbian traces the development of her
awareness of her sexual preference to early childhood while a
later awareness of development of same-sex sexual preference
is associated with greater femininity.

Saghir and Robins (1570) suggest, based on an analysis
of their data that there are two kinds of lesbians, those who
were conventially feminine during childhood and those who
displayed a masculine sex-role orientation. These authors
define the ”tomboy syndrome” as a ”persistent aggregate of
behaviors that are preferred by girls, and that are primarily
appropriate to boys.” They go on to state ”it is not a short
passing phase during development but a persisting pattern of

conduct starting in childhood and extending at least into

%8

adolescence”.(1573, p. 153) In their early studies, over two
thirds of their lesbian subjects reported masculine interests
and behaviors during childhood, vs. 15% of the heterosexual
controls. These women differed from the conventionally
feminine lesbians in their responses to a variety of
questionaire items. Significantly more tomboy lesbians
reported a primary identification with father (%7% vs 5%) and

a lack of closeness to mother.
Family Background and Sexual Orientation

Studies of family background are beset with the problems
of post-hoe memory data. Memory is constructive and
selective. The cultural prohibitions against lesbianism are
likely to lead to difficulties in the lesbian’s relation to
her family. As such current conflicts or tension may bias
her perceptions of past family relationships. This factor is
likely to account for some of the self-reported differences
in the family relationships of lesbian and heterosexual
women, and should be considered as a contributing factor. In
the following review the empirical studies of family
background are presented chronologically.

Bene (1555) compared responses on the Bane-Anthony
Family Relations Questionnaire obtained from 37 lesbian and
50 married heterosexual women. Though differences did not
reach statistical significance, lesbians reported having
received less affection from their mothers and having felt

less warmth toward them. Significant differences were found

%8

on feelings of hostility toward fathers; lesbian women
reported more hostility toward and fear of their fathers.

In a study utilizing a patient pOpuletion, Kaye, Berl,
Clare, Eleston, Geschwin, Kogan, Torda, and Wilbur (1557)
obtained questionnaire data from the psychoanalysts of 8%
lesbian and 8% heterosexual women. No significant
differences were found for the two groups on items assessing
the mother-daughter relationship, but analysts reported that
their lesbian patients more often felt fearful of their
fathers and felt exploited by them.

Gundlach and Reiss (1555) used a questionnaire in a
study of 885 lesbian and 83% heterosexual women matched for
age, education, and geographic distribution. Significant
findings indicated that lesbian women more frequently felt
neglected and ignored by their mothers, treated impersonally
and without love. Lesbian women more often reported that
their fathers were indifferent, less warm and less
affectionate. Heterosexual women reported less parental
discord, had spent significantly more time with their
mothers, and though not statistically significant,
heterosexual women felt more dependent upon their mothers.
Lesbian women spent less time with both parents.

Kenyon (1555b, 1555c) administered a family history
questionnaire to 183 lesbian and 18% heterosexual women. One
half of the lesbians and three-fourths of the heterosexuals
stated that they got along well with their mothers. ”Quite
poor” relationships with mother were reported by 81% of

lesbians compared to 0.5% of heterosexuals (p<.I0l). Among

50

the heterosexual group, 53% stated that their parents were
happily married, in contrast to %5% of the leasbian group,
and more parents of lesbian women were divorced (88.5 vs
%.52)

The developmental model of female sexual orientation
proposed by Charlotte Wolff described earlier, was derived
from her data based on interviews and questionnaires from 105
lesbian and 183 heterosexual women (Wolff, 1571). Pertinent
family history differences were tested for statistical
significance using a chi-square analysis of the questionnaire
responses. Mothers of lesbian women were more often
negligent, indifferent, and less loving (p<.001). Fathers of
lesbian subjects were reported to be less loving and more
negligent (p<.01). Significantly more mothers of the
prehomosexual girls would have preferred the subject to be a
male, and significantly more favored a son in the family.
Mothers were more often the dominant parent and marriages
were more frequently unhappy and disrupted.

Poole (1578) devised a questionnaire to assess the
family background of 50 heterosexual and 50 homosexual women.
Statistically significant differences were found with
lesbians reporting fewer recollections of mother as
affectionate and understanding, fewer mothers being happy in
their role, and fewer fathers who displayed affection.

In the Swanson at. al., (1578) study cited earlier, case
records of %0 heterosexual and %0 lesbian women in

out-patient psychotherapy were compared. No group

51

differences were found, family disorganization and feelings
of parental rejection were high for both groups (752).

Loney (1573) compared the responses of 11 homosexual and
18 heterosexual women on the Elias Family Adjustment Test.
Statistically significant differences found lesbian women to
score higher on scales measuring negative attitudes toward
both mother and father, struggle for independence,
parent-child friction, parental rejection, and interparental
friction.

Saghir and Robins (1573) found a variety of differences
between their 5% lesbian and %3 heterosexual women in data
they collected from interviews. Lesbians more frequently
described poorer relationships with their mothers; while 55%
of heterosexual women reported that their relationships with
their mothers were good, only 83% of lesbian women did.
Lesbians more often reported a closer relationship with their
fathers, and 77% of lesbians vs 15% of heterosexual women did
not identifiy with their mothers, but identified with either
father or no parental figure. Family disorganization and
marital discord were more prevalent in the lesbian family
background.

Siegelman (157%) used a measure which assessed
recollection of family relationships--the Parent Child
Relations Questionnaire Form 8, and a biographical
questionnaire. Significant differences were found with
homosexual women reporting both mothers and fathers to be

less loving, fathers more rejecting, and mothers more

58

demanding. Heterosexual women were closer to mother and
closer to father.

Beverely (1575) used an open interview format with 5
lesbian women. Four of these women described a distant or
seriously disturbed mother.

Crouch (1577) examined family relationships of 57
heterosexual and 55 lessbian women using the Elias Family
Adjustment Test (FAT). No differences in total FAT scores
indicate that there are no significant differences in overall
family adjustment. However significant differences were
found with lesbian women more often reporting rejection by
one or both parents during early childhood.

Hogan, Fox and Kirchner (1577) recruited 805 homosexual
women from lesbian organizations throughout the state of
Illinois to complete a 800 item questionnaire. Forty-five
percent of these women indicated that they felt rejected by
their mothers, and 352 reported feeling rejected by their
fathers.

Shavelson, at. al., (1550) compared the responses of 85
lesbian and 85 heterosexual women on the Schaefer Child
Report of Parental Behavior Inventory. They found no
significant differences on any of the parent-child or
parental-relationship dimensions. A regression analysis of
all the variables did not yield any significant combination
in predicting sexual orientation.

Bell, Weinberg and Hammersmith (1551) conducted one of
the most comprehensive large scale studies of sexual

orientation. Using structured interviews, 885 caucesion

53

lesbian and 101 caucesion heterosexual women were compared on
a broad range of factors. In this sample significantly more
lesbian women spontaneously described their relationships
with their mothers as predominantly negative (%5% vs. 81%),
characterizing their mothers as more hostile and rejecting.
Heterosexual women were more apt to report childhood
identification with mother, and feeling that they were
similar to their.mothers as they were growing up (55% vs
53%). The lesbian women report significantly higher
frequencies of parental discord, mother-dominant marriages,
and detached, hostile fathers. The investigators entered all
significantly differentiating variables into a path analysis.
Of the many factors examined, the lesbians’ greater tendency
to report having been masculine during childhood (58% vs 10%)
was the single strongest predictor of adolescent and adult
homosexuality. The path analysis indicates that the family
relationship variables explored explain only 852 of the
variance of childhood masculinity. The researchers conclude
that there are many different pathways to adult sexual
preference, and that the family relationship variables they
examined do contribute although the total effects are of

modest magnitude.

Summary

In accordance with the recently pervasive belief that
homosexuality was an aberrant form of erotic expression, and
that the determinants of same-sex sexual preference could be

found in pathological parent-child relationships, many of the

5%

studies reviewed above searched for differences in the family
history of lesbian vs. heterosexual women. The empirical
data indicates that disturbances in family relationships are
neither ubiquitous nor in any consistent form exclusively
characteristic of lesbian women.

Yet, in most of the studies, lesbian women do report
significantly more emotional dissatisfaction with their
mothers, and with their fathers, and significantly more
interparental discord. Across studies lesbian women are more
apt to perceive their mothers as less loving, affectionate,
understanding, and involved; and lesbians are more likely to
perceive fathers as distant and hostile. Where percentages
are reported, 50% to 70% of lesbian women characterize their
relationships with their mothers and with their fathers as
less close and less positive. In contrast heterosexual women
tend to report greater closeness and warmth in their
relationships with both parents. These differences may be
relevant components in a complex array of factors which are
influential in the development of same-sex vs. opposite-sex

erotic-responsiveness.

Purpose and Perspective

Empirical studies of the development of gender-role
attributes and sexual orientation in women suggest that the
theories offered by Chodorow and Dinnerstein to explain

gender development across the sexes may be relevant to the

55

development of psychological femininity and masculinity and
varying sexual orientation in women alone.

Thus, according to Chodorow traditional femininity
develops in the context of an early childhood relationship
with mother which is characterized by high levels of
closeness, nurturance and warmth, leading to extended
dependency and a potent and prolonged identification with
mother. Chodorow asserts that the greater closeness mothers
experience with their daughters, the prolonged dependency
allowed the infant girl, and the expression and emulation of
maternal behaviors which are allowed to the female child, all
lead to the formation of an internal object world in which
self and other representations are characterized by
identification and highly effective relatedness. Personality
organization is such that the self is defined in relational
terms, is object-dependent and relatively undifferentiated.
As Dinnerstein (and Charlotte Wolff) suggest, a daughter
resolves her feelings of maternal rejection or loss of
oneness with mother by internalizing and striving to become
the wished for ’good mother’. These factors, together with
further socialization processes produce an object-dependent
ego structure leading to personality functioning
characterized by a high degree of empathic merging, an
other-defined self, and the numerous communal personality
traits considered feminine. Hence, woman are nurturing,
compassionate, loyal, subjective, emotional, submissive,

dependent, compliant, etc...

58

Psychological masculinity in contrast develops when
mothers are less close, less nurturing and relate to their
infants in more differentiating ways. As such masculinity
develops via a rejection of identification with mother, hence
a rejection of affective relatedness. Masculinity involves
the differentiation of self from other, and forms of
relatedness which emphasize differentness, and affective
boundaries between self and other. The masculine object
world is characterized by greater differentiation of self,
which is not dependent upon or defined in relation to others.
Personality traits which emerge are ones which reflect an
insulated and independent identity, among these are
dominance, autonomy, aggression, assertiveness, objectivity,
and self determination. The longed for oneness with the good
mother is achieved in intimacy with women.

The empirical research which was reviewed here suggests
a broad base of support for the basic tenets of these
theories. Studies which examine the relationship between
family variables and femininity/masculinity in women indicate
that maternal distance, and demandingnass are associated with
the presence of agentic characteristics in daughters, while
identification with highly feminine mothers can lead to high
femininity. In either or both parents androgyny or highly
sex-typed characteristics are associated with similar
patterns of gender-role expression in women. As Diana
Baumrind and many other developmental psychologists have
noted, mother-daughter relationships which involve high

degrees of maternal warmth, closeness and involvement are

57

associated with a variety of undesirable characteristics in
daughters, reflective of the detrimental attributes of high
femininity, and low masculinity (e.g., dependency,
submissiveness, low self-esteem, low achievement motivation,
etc...).

In studies which compared lesbian to heterosexual women,
the majority of findings show lesbian women to be more
androgynous, i.e., endorsing traits which include both those
ascribed as feminine and masculine. In many of the studies
reported, lesbian women endorsed significantly more masculine
and agentic traits than their heterosexual peers; while there
are generally no significant differences in the endorsement
of feminine traits. Significant differences in feminine
attributes where they do occur, and nonsignificant trends
show heterosexual women to be more female sex-typed than
lesbian women. Across a wide range of measures lesbian women
consistently emerge as more dominant, assertive, autonomous,
and inner directed than heterosexual women. Thus the
personality attributes which distinguish lesbian from
heterosexual women are those which reflect the agentic,
masculine, and more interpersonally independent identity.

In investigations that compared the family backgrounds
of heterosexual and homosexual women, the significant
findings indicate that homosexual women are more apt to have
experienced mother as distant, uninvolved, and neglectful or
rejecting; while heterosexual women report having experienced
greater maternal nurturence, and greater dependency upon

their mothers. Prevalent patterns of differences include

58

lesbian women’s perceptions of their mothers as less loving,
affectionate, understanding and involved, more hostile, cold,
and rejecting; and a father who is perceived as distant,
hostile and exploitative. While these patterns were found to
significantly differentiate lesbian women from heterosexual
women, they are by no means pervasive nor ubiquitous in
characterizing the family backgrounds of lesbian women.
Across all studies reviewed, these differences are reported
by 50 to 70% of the lesbian women. According to the
theoretical conceptualization of this research, the maternal
distance experienced by the lesbian may be an important
component of the developmental processes which lead to both
masculine personality attributes and female sexual
preference. The lesbian woman thus does not experience the
dependency-identification mode of feminine socialization,
though she certainly experiences a wealth of other
socialization forces common for girls, perhaps explaining her
androgyny. The maternal behavior experienced by the
prehomosexual girl may be no more distant or rejecting than
that which typifies the mother-son relationship. It is
possible that the lesbian openly acknowledges her feelings of
maternal rejection because she lacks the full range of
masculine socialization pressures which prevent the boy from
acknowledging dependency needs and force him to construe
maternal distance as fostering of his own independence,
self-reliance and freedom.

The personality traits which distinguish lesbian from

heterosexual women, those such as autonomy, dominance,

58

aggression, etc..., reflect the more interpersonally
independent identity which characterizes masculine
attributes. It is conceivable than that the early
mother-child relationship pattern and separation-
individuation development of the prehomosexual girl more
closely approximates that which Chodorow and Dinnerstein
describe as typical of this developmental process for boys,
and that this route of differentiation from mother and
identity formation is related to female sexual preference for
both females and males. Thus the early childhood experience
of the prehomosexual girl would reflect the absence of
prolonged maternal nurturence, greater differentiation from
mother, and the development of an identity which is not
highly object-related but rather more differentiated and
effectively insulated.

How might this lead to female sexual preference?
According to Dinnerstein the child’s desire to possess the
’good mother’ is met either by internalizing, intrOjecting
and becoming the good mother through identification with
mother, the good mother, or one maintains the desire for
physical-emotional intimacy with females. Remember
Dinnerstein proposes that the infant’s global experience of
mother is a highly tactile-sensual-physical and emotional
knowing of a woman’s body, and that for all individuals the
erotic experiences of adulthood redintegrate this early
sensual contact. For the heterosexual male and the lesbian

women, it is a women which provides this experience. The

In

80

heterosexual woman according to Dinnerstein actualizes the
’good mother’ in herself in sexual intimacy.

If the developmental processes described by Chodorow and
Dinnerstein are correct, then the nature of the early

mother-child relationship should be associated with degrees

of identification with mother, differences in
object-relations representations, psychological
femininity/masculinity, and sexual orientation in women.
Specifically, high maternal involvement, warmth and
nurturence which accompany mother-child enmeshment should be
associated with strong identification with mother, relatively
undifferentiated and emotionally interrelated
object-relations representations, psychological femininity,
and sexual preference for males. An early childhood
relationship with mother which is characterized by less
warmth and nurturence, and greater distance (greater
mother-child differentiation) should produce more highly
differentiated and isolated object representations,
disidentification with mother, psychological masculinity, and
sexual preference for females. In fact the empirical
examination of these theories using only women subjects
serves to isolate the theoretically important variables in
the mother-child relationship from numerous other potentially
confounding factors, including possible biological sex-linked
determinants and the influence of other socialization forces
such as peers, other adults, the media, etc.., which differ
for girls and boys. It serves as a control for the numerous

socialization factors which differ for the two sexes.

81

As earlier discussions have noted, sexual orientation
and femininity/masculinity are neither unitary, single
dimensional nor bipolar. Women may display a range of
feminine and masculine personality attributes in varying
degrees, and may express sexual interest in either or both
women and men in varying degrees. This study was designed to
examine these variables in a manner which allows for

diversity and variation.

88

METHODOLOGY

Hypotheses and Rationale

I Early Childhood Maternal Involvement
and Femininity/Masculinity

1) Maternal involvement in early childhood is
positively associated with femininity in women. High early
childhood maternal involvement is associated with high
femininity.

8) Maternal involvement in early childhood is inversely
associated with masculinity in women. High maternal
involvement in early childhood is associated with low
masculinity.

These hypotheses were designed to test Chodorow’s theory
of gender development. They examine her assertion that
psychological femininity and masculinity result from the

degrees of involvement and differentiation in the early

mother-child relationship.

II Early Childhood Maternal Involvement
and Erotic Orientation

3) Maternal involvement in early childhood is
positively related to androeroticism (sexual interest in
men). High maternal involvement is associated with high
androeroticism.

%) Early childhood maternal involvement is inversely
associated with gynoeroticism (sexual interest in women).
High maternal involvement is related to low gynoeroticism.

The proposed relationship between gyno/andro-eroticism

and the early mother-child relationship are deduced from

83

Dinnerstein’s theory of erotic attraction. If these
variables are significantly related, they would lend support
to Dinnerstein’s assertion that the gender-related dynamics
of erotic-responsiveness are derived from differences in the

early mother-child relationship.

111 Early Childhood Maternal Involvement and Level
of Differentiation of Object Representations

5) Levels of early childhood maternal involvement are
inversely associated with levels of differentiation in object
relations representations. High maternal involvement in
early childhood is associated with undifferentiated and
interrelated object representations. Low maternal
involvement is associated with highly differentiated and
isolated object relations representations.

This hypothesis describes Chodorow’s assertion that the
degrees of differentiation in the early childhood
relationship with mother are internalized and become the

basis for the enduring structure of mental schemata of self

and other representations.

IU Level of Differentiation of Object Relations
Representations and Femininity/Masculinity

5) Differentiation in object representations is
positively associated with masculine personality attributes.
Highly differentiated object representations is associated
with high masculinity.

7) Differentiation of object representations is
inversely associated with femininity. Low levels of object
representation differentiation is related to higher
femininity.

These hypotheses examine the relationship between the

levels of differentiation of object relations representations

 

8%

and gender attributes. They test Chodorow’s theory that
feminine and masculine personality attributes are mediated by

differences in the mental schema for interpersonal relations.

U Level of Differentiation of Object Relations
Representation and Erotic Orientation

5) High differentiation in object-representations is
positively associated with gynoeroticism.

5) Low differentiation in object-representations are
positively related to androeroticism.

These hypotheses investigate whether the object
relational paradigms from Chodorow’s theory are associated

with erotic orientation.

UI Erotic Orientation and Femininity/Masculinity

10) Gynoeroticism is associated with masculine
personality attributes.

11) Androeroticism is positively related to feminine
personality attributes.

UII Identification with Mother, and
Identification with Father

18) High identification with mother is associated with
high early childhood maternal involvement.

13) High identification with mother is associated with
low levels of differentiation of object relations
representations.

1%) Identification with mother is positively correlated
with femininity.

15) Identification with mother is positively associated
with androeroticism.

15) Disidentification with father is positively
correlated with femininity and androeroticism. High
identification with father is associated with low early
childhood maternal involvement and levels of object relations
representation characterized by high differentiation.

85

Measurement of Uariables

Early Childhood Maternal Involvement

Early childhood maternal involvement was measured with a

six point (or level) rating scale on which subjects rate

their early childhood relationship with mother)" Each scale
point is described and defined for subjects. Each of the
scale levels describes differences in the degree of mother’s
involvement, warmth, closeness and nurturence; and
concomitantly, daughters feelings of involvement, closeness,
separateness and differentness. Thus each of six paragraphs
which define each of the six levels summarizes a range of
behaviors and feelings, in an effort to capture women’s
overall or general experiences of their mothers along these
dimensions. The points along the scale are arranged to
represent a continuum, ranging from high maternal involvement
at point 1, to low maternal involvement at point 5. The
scale is contained in Appendix A. This scale is abbreviated
Matinvl.

In addition to women’s ratings of their relationship

with their mother during early childhood (defined for

I A second measure of early childhood maternal involvement
was developed in a pilot study, and consists of a scoring
system for TAT stories. Results of data analysis revealed
that the two measures are only modestly associated, and
furthermore use of the TAT measure involves substantial
reduction of the sample size. As such only the Maternal
Involvement Scale is used to assess this variable.
Description and results of the data analysis with the TAT
Measure are provided in Appendix J.

88

participants as the years from birth to age six), respondents
were asked to rate their mother-daughter relationships using
the same scale for late childhood, for adolescence, and for
their present relationships with their mothers. (These
ratings are abbreviated Matinv8, Matinv3, and Matinv%,

reSpectively.)

The Maternal Involvement scale is a straight forward and
reasonable self-report measure. As such it has face
validity, though the reliability of the scale is unknown.
However, the pattern of means, standard deviations, and
intercorrelations of women’s ratings of their mother-
daughter relationships for the four life stages are
consistent with intuitive expectations and strongly indicate
that this is a valid and reliable measure. These data are
presented in Table 1. As the group means reveal,
mother-daughter involvement is highest during early
childhood, and lowest during adolescence, as would be
expected. The intercorrelations between mother-daughter
involvement across the four time periods show the strongest
associations for contiguous time periods with step-wise
decreases with each life phase. The correlations reveal
strong linear relationships across time periods, highlighting
considerable stability in the closeness vs. distance and
differentiation in mother-daughter relationships over the
life-span.

When these means and intercorrelations are computed

separately for lesbian and heterosexual women as groups the

”In.

87

values are nearly identical. These data further suggest that

the scale has stable psychometric properties.

Table 1 Means, standard deviations, and intercorrelations
of women’s self ratings of their maternal involvement for
four life stages

 

Mean 5.0. Intercorrelations
Early Childhood 8.5 1.3 1.0
Late Childhood 3.0 1.3 .57 1.0
Adolescence 3.5 1.5 .55 .55 1.0
Adulthood 3.% 1.3 .%0 .%5 .55

 

Identification with Mother/Identification with Father

Levels of identification and disidentification with both
mother and father are measured with a semantic differential
(50) designed for this study. (These variables are
abbreviated IDMO and IDFA.) Semantic Differential is the
term used to describe any instrument which asks subjects to
rate concepts (in this case ’yourself’, ’your mother’ and
’your father’) on a Likert format scale of bipolar
adjectives. Thirty six bipolar adjectives were chosen for
the 50. These adjectives reflect a broad range of
personality characteristics. A number of the items from
Charles Osgood’s recomendad scales are included. Women
complete three semantic differentials, one for ’yourself’, 1
one for ’your mother’, and one for ’your father’. The

semantic differential is contained in Appendix C.

88

The degree of identification vs. disidentification with
each parent was measured by taking the absolute differences
of subject’s scores for self and mother, and self and father,
on each item, then summing over the 35 items. This yielded a
single score for the degree of similarity between self and
mother, and a single score for self and father. Low scores
reflect greater similarity and greater identification, while
high scores are reflective of dissimilarity and
disidentification.

The data obtained consist of a distribution of scores for
all respondents which are used in correlational analyses.
For the purposes of group comparisons, the women were divided
into three groups reflecting high identification, middle, and
low or disidentification. The criteria for group
classification was simple percentage, i.e., the range of
scores were divided so that one third of the participants

fall into each of the three groups.

Level of Differentiation of Object Relations Representations

The Level of Differentiation of object representations
(L0) is measured with a scale (or scoring dimension) which is
scored for subject’s TAT stories. The dimension is scored on
one of five levels which are:

1) Merged and Undifferentiated.

8) Superficially Individuated and Interrelated
3) Superficially Individuated and Isolated

%) Fully Individuated and Interrelated

5) Fully Individuated and Isolated

88

This dimension is designed to measure the object
relations constructs proposed by Chodorow’s theory. She
hypothesizes that a prolonged, close relationship with mother
leads to a relatively undifferentiated ’self’ organization
which is object—related or object-dependent. A more distant
early childhood relationship with mother and earlier
differentiation is proposed to result in a highly
differentiated and less interrelated ’self’ representation.
These constructs are assessed by examining the levels of
individuation, and of interrelatedness of characters in
subject’s TAT stories. The term ’individuation’ in the
scoring dimension refers to the degree to which a character
is elaborated, developed, and differentiated--separate from
other characters. The ’interrelated’ and ’isolated’
components refer to the extent of interaction and the nature
of the character’s relatedness to other characters in the
story told. The assessment of ’self’ representation is based
on the character of the heroine or pivotal figure in
subject’s stories, while other characters in the story are
rated as representations of ’others’. Instructions for
scoring the LD are contained in Appendix 0.

Each of the participant’s two TAT stories were scored
two times on this scoring dimension, once for each story with
one scoring for the heroine of the story and the other for
the other characters. Photocopies of the two TAT cards
chosen for this study, and the instructions provided to

participants are contained in Appendix B.

70

Two individuals with Bachelors degree’s in psychology
scored the TAT stories on the level of differentiation
dimension. These coders had been trained to use this scoring
dimension for an earlier pilot study (see Berry, 1555).
Working independently, the coders had achieved a high degree
of interrater reliability with the pilot data (percent
agreement was 55). For the purposes of this study, two
refesher training sessions were held. Both coders scored the
first 80 TAT stories independently. Their percent agreement
was 552. For the remainder of the data set, each
respondent’s TAT stories were scored by only one of the

coders.

Psychological Femininity/Masculinity

Women’s femininity/masculinity was measured with the
Spence-Helmreich Personal Attribute Questionnaire (PAQ) (see
Appendix E). The FAQ contains three subscales of 5 items
each. Based on their research, the subscales were shown to
consist of an independent Feminine dimension (F), an
independent Masculine dimension (M), and a dimension which
reflects bipolar feminine and masculine traits (M-F).

The PAD is a self-report measure of gender linked
personality attributes. Each of the 8% items are measured
on a 5 point bipolar scale. Each item describes a
gender-role characteristic and was chosen for the instrument
based on significantly different mean ratings achieved by

women and men in the stareotypically expected directions.

71

Psychometric properties of the scales indicate that a
tripartite conceptualization of masculinity-feminity is
valid. Intercorrelations among the scales suggest that the
masculine and feminine scales are separate and distinct
dimensions. (They have low positive correlation for both

sexes). The M-F scale was shown to have internal validity
and sex linked bipolarity. The M-F scale was moderately
correlated with the M scale, and had a low but substantial
negative correlation with the F scale.

Items on the M and F scales were rated as socially
desirable for both sexes, but more characteristic of one sex.
The items on the F scale describe interpersonal awareness and
concern for others. The items on the M scale refer to
agentic or masculine personality characteristics including
independence, competitiveness, perseverence, etc... The items
of the M-F scale are all scored in the masculine direction.
This scale contains two masculine items assessing aggression
and dominance. The remaining six items describe emotional
vulnerability, expressiveness and dependency. These reflect
susceptibility to emotional hurt and need for emotional
support. High scores indicate aggression, dominance and lack
of emotional vulnerability and dependency. The M-F scale
items were each rated as socially desirable for one sex,
consistent with stereotyped expectations. Despite this
finding, the scale reflects undesirable characteristics at
both extremes.

The three scales are linear and continuous measures and

78

were used in correlational computations. To assess the
conjoint relationships between the scales, Spence and
Helmreich recommend a two way classification for each scale
based on the median scores. Thus participants are classified
as high or low on each of the three scales, and the conjoint
relationships between each pair of scales are expressed by
means of two by two tables. Such categorical analyses were

also computed.

Erotic Orientation

Erotic orientation is measured with a scale called the
Erotic Orientation Scale (EROS) developed by Michael Storms
(1550. 1551). Storms devised this scale based on a model of
erotic-responsiveness in which sexual attraction to women
(gynoeroticism) and sexual interest in men (androeroticism)
are separate dimensions. Subjects respond to eight
Guttman-format items on two distinct subscales, one measuring
the frequency and nature of erotic interests and fantasies of
women; the other measuring the frequency and nature of erotic
interests and fantasies of men. Scores on the two scales
were orthogonally combined to yield four categories:
’low-erotic’, ’high-erotic’, ’androerotic’, and ’gynoerotic’
groups. Storms tested the scale on a population of 155
college students. Subjects completed the EROS, the Kinsey
Scale, and a questionnaire item which asked subjects to label

themselves as homosexual, heterosexual or bisexual.

73

Statistical results confirmed the hypothesis that gynoerotic

and androerotic orientations vary independently and are

therefore best conceptualized as independent and orthogonal
dimensions. Findings of this study also confirmed that
erotic fantasy and interest are related to behavior as
measured by the Kinsey scale, and by subjects’ self
assignment of their sexual orientations. Storms found that
homosexual subjects report more heterosexual fantasy than
vice—verse, and that bisexual subjects report as much
heteroerotic fantasy as the heterosexual group and as much
homoerotic fantasy as the homosexual group.

Psychometric properties of the scale indicate that it
has high internal reliability and validity, and that it is
consistently reliable across sexual orientation groups.

The EROS is superior to the Kinsey Scale in a number of
ways. It discriminates between bisexual and asexual
individuals, while the Kinsey Scale places them in the same
range. The EROS makes finer discriminations of erotic
interest, and accurately assesses gynoeroticism and
androeroticism as separate dimensions.

The EROS provides two sets of scores which are linear,
and were used for correlational analyses. For the purpose of
group comparisons, a median split classfication of women’s
scores on the two subscales were combined as described above.
Group comparisons on the basis of respondents’ self
designated sexual orientation were also used. The EROS is

contained in Appendix F.

7%

Secondary Uariables and Rationale

A variety of additional information relevant to this
area of research was obtained from participants on the
questionnaire they completed. (See Appendix G for
questionnaire.) The items on the questionnaire requested
information about demographic factors, and a number of
secondary variables which previous studies have shown to be
related to aspects of women’s psychosexual development and
expression.

The demographic information obtained included age,
race/ethnicity, education, occupation, and the respondents’
classification of their present socio-economic status and the
socio-economic status of their families during their
childhoods. These factors are known to be related to gender
role attributes in women.

Respondents were also asked whether they were raised in
single or two parent families, whether their parents
separated or divorced during their childhoods. who their
primary caretaker was during early childhood and whether
there was any significant time period of separation from this
person. In addition to the rating scale which assesses the
subject’s early childhood relationship with mother, women
were asked to indicate whether any of a brief list of
adjectives characterizes their father’s behavior toward them
during childhood, and whether they would characterize

themselves as having been ”Daddy’s Princess” or ”Mother’s

Girl”.

75

A number of questionnaire items requested information
relevant to intimate relationships. Women were asked to
indicate their ’relationship status’, i.e., single, coupled
or married. They were asked whether they were currently in
an intimate relationship, and how long they have been in this
relationship. Two further questions asked women how many
’significant’ intimate relationships they have had and the
length of their longest intimate relationship. The
interrelationships among these items and the primary research
variables were examined to discern what if any differences
might characterize women with differing relationship
patterns.

Previous studies of both heterosexual and lesbian women
have found that feminist ideology is associated with
androgyny. Participants were asked to rate their views of
women as either traditional, liberal, or feminist. The
relationship between women’s gender ideology and the other
variables was examined.

Three items refer to areas of personal/developmental
history which were examined for relationships with the other
research variables. The first item which has been shown to
differentiate lesbian women from heterosexual women asked
subjects if they were ”tomboys” during childhood. The next
two items asked women how old they were when they first
experienced sexual sensations or feelings, and how old they
were when they first became aware of their sexual
orientation. Previous studies have found that lesbian women

who realize their lesbian sexual orientation later in life

78

are more feminine than lesbians who are aware of their
homosexuality at an earlier age (Clingman & Fowler, 1557;
Saghir 8 Robins, 1573; Uance, 1577).

The final questionnaire item asked whether respondents
had ever been the victim of sexual abuse, rape, or incest,
who the abuser was and how old they were at the time of the
abuse. A comment which preceded this question stressed that
participants did not have to answer the question if they
chose not to. Sexual abuse is a traumatic life event which
can affect many areas of women’s personality functioning.
The relationship between sexual abuse victimization and the
other research variables was examined.

Whereever demographic or family background variables were
significantly associated with the research variables, these
effects are controlled for with partial correlation
coefficients. (Results of exploratory analyses of the

secondary variables are provided in Appendix K.)

Procedure

Data Collection

The sample for this study was obtained through the
voluntary and anonymous participation of women who were
recruited through three sources: 1) Extended friendship
networks; 8) Places of work; 3) Women’s organizations.

A flier which briefly described the purpose and
procedure of the study and invited participation, was
distributed through friendship networks and was posted on
bulletin boards in a variety of work settings and in offices
of women’s organizations. For each of the public places
where fliers were posted permission was gained from the
person in authority. Each flier designated a contact person
in that particular setting from whom the research measures
could be obtained.(The flier is contained in Appendix H.)

Uolunteers were given a large sealed envelope which
contained a cover letter from the researcher (Appendix I),
the research instruments and instructions (Appendices A to
G), and a stamped envelope addressed to the researcher. The
cover letter contains information about the study,
instructions about the procedure, and orients volunteers to
the research instruments. The cover letter stresses that
participation is completely voluntary and anonymous, and 1

describes the confidentiality safeguards. Participants were

78

instructed to mail their completed research forms to the
researcher in the envelope provided. The researcher’s name,
address and phone number were included with the cover letter,
and women were informed that they may contact the researcher
if they had any questions or wanted further information about
the study.

The research forms included in the envelope contained
the Questionnaire, the Maternal Involvement Scale, the FAQ
Scale, the Eros, the Semantic Differential, and photocopies
of the two TAT cards, with instructions and paper for the

written stories.

Sample Characteristics

A total of 77 women completed and returned the research
measures for this study.

The sample ranges in age from 15 to 55 years old with a
mean age of 38 and a standard deviation of 5 years. One half
of these women are between 85 and 35 years of age. The
racial composition of the sample is almost exclusively white,
55 participants or 50% of the sample identified themselves as
caucesian. Information on educational background indicates
that this is a highly educated sample: 8% women have
completed graduate or professional training; 83 have
completed four years of college; 80 women have completed at
least one year of college and the remaining 10 respondents
have completed High School. Seventy-five percent or 55 of

the participants are employed, six women are not employed,

78

and 13 of the women are students.

Women were asked to indicate their present
socio-economic class, and additionally the standardized
Hollingshead and Redlich index of socio-economic status was
computed using the formula which combines education and

occupation. The Hollingshead and Redlich SE5 composition of

the sample is:

 

Table 8 Socio-economic status of respondents
SE5 Class Number of
Subjects
(highest) I 7
II 80
III 85
IU 18
U 5
Unclassified 5

 

The breakdown of Women’s self ratings of their present SES
and their ratings of their families socio-economic class

during their childhoods are as follows:

Table 3 Socio~economic status of participant’s
family of origin

 

Self Family

Upper Class O 1
Upper Middle Class 15 80
Middle Class 35 85
Working Class 83 85
Lower Class 1 8

 

Family Background and Personal Information

Thirty nine of the participants classified themselves as

heterosexual, 85 classified themselves as lesbian and 10

80

classified themselves as bisexual. Due to the small number
of bisexual women, this group is not included in forms of
data analysis which involve group comparisons of women based
on their sexual orientation. In response to the question on
relationship status, 55 of the women were in an intimate
relationship, 15 of the respondents view themselves as
married, and 85 view themselves as ’coupled’. Questions
about family background indicate that 55 of the women were
raised in intact two parent families, 17 of the respondents
are from divorced families. Seventy-two of the women
indicated that their biological mothers were their primary

caretakers during early childhood.

Preliminary Notes on the Data Analysis

The raw data for each subject consists of the
Questionnaire, the Maternal Involvement Scale, the PAQ, EROS,
the Semantic Differential, and two TAT stories. All 77
participants completed the questionnaire; the PAQ, EROS, and
50 were each completed by 75 respondents, however only 55
women returned written TAT stories.

The means and distributions of women’s scores on the FAQ
scales and the EROS scales are consistent with those reported
in the literature for the standardization samples. Women’s
scores on all three PAQ Scales and on the androeroticism
scale are essentially normally distributed. However women’s

scores on the gynoeroticsm scale are bimodally distributed.

81

Differences within Sexual Orientation Groups

An exploratory examination of the data revealed that the
patterns of relationships between the research variables for
the sample as a whole are often different from the patterns
of relationships within each of the two major sexual
orientation groups when examined independently. In a number
of cases this discrepancy is due to differences in the
distributions of scores across lesbian and heterosexual
women. However, in other instances it is evident that
findings which emerged from the data analysis of the sample
as a whole reflect the characteristics of one sexual
orientation group but not the other. Furthermore, the
relationships between the research variables often differ for
each of the sexual orientation groups. In these instances,
statistical analysis and results are presented for each of
the two major sexual orientation groups independently. In
the presentation of the findings in the Results chapter,

these differences are presented and later discussed.

The Measurement of Object Relations Representations

The data from respondents who returned TAT stories
provides four scores for each participant on the level of
differentiation of object representations dimension. For
each of their two TAT stories the dimension is scored once
for the pivotal figure in the stories and once for the other
characters. An examination of the interrcorrelations between

these four scores indicates that womens’ prOjections of self

88

and others in the two stories are very similar. For the
purposes of simplifying the analyses of the data, the four
scores for each participant are summed to provide a single
score. Hence each subject has a single score representing
the sum of their scores on the four ratings, and these values

are used in the data analysis.

Methodological Considerations

A number of features of the sample support its
representativeness. The overall sample means and standard
deviations on all of the standardized research measures are
consistent with those reported in previous studies with these
measures for college populations. Many of the potential
sources of confound in studies which compare lesbian and
heterosexual women are minimized in this investigation. The
lesbian and heterosexual women are matched for age, and
almost all are employed. The majority of women from both
groups are college educated; and while 15 of the heterosexual
women are married, the differences in women’s relationship
status (single vs. coupled) and the length of their current
relationship were examined as these related to the research
variables.

The generalizability of the findings are limited by
three constraints on the representativeness of the sample.
The women who participated in the study are highly educated.
While the effects of education are controlled for wherever it

was a relevant variable, the lower range of education is not

88

adequately represented in the sample. A second

unrepresentative feature involves group differences in the

socio-economic family background of lesbian and heterosexual
women. Forty percent of the lesbian women in this sample are
from upper or upper-middle class family backgrounds, in
contrast to only 1 z of the heterosexual group. Family SE5
is Significantly associated with many of the research
variables, and again the effects of family SES are controlled
for wherever possible. And finally, the many difficulties
which accompany defining and obtaining a representative
sample of lesbian and bisexual women further restrict the

generalizability of findings for these groups.

8%

RESULTS

Hypotheses I
Early Childhood Maternal Involvement and

Femininity/Masculinity

1) Maternal involvement in early childhood is
positively associated with femininity in women. High early
childhood maternal involvement is associated with high
femininity.

8) Maternal involvement in early childhood is inversely
associated with masculinity in women. High maternal
involvement in early childhood is associated with low
masculinity.

These hypotheses predicted relationships between the
level of early childhood maternal involvement and
respondents’ femininity/masculinity as measured by the three
PAQ Scales, F, M, and M-F. Pearson correlation coefficients
were computed for each of the three PAQ scales and the
Matinvl scale. For the sample as a whole, there was no
significant association between women’s ratings of their
earliest childhood relationships with their mothers and their
gender characteristics. When the relationship between early
childhood maternal involvement and femininity/masculinity is
assessed within sexual orientation groups, the findings
reveal one significant partial correlation between Matinvl
(early childhood maternal involvement) and the bipolar M-F

scale for heterosexual women in the predicted direction (see

Table %, partial r controls for education). Here greater

85

maternal nurturence and closeness is related to the
endorsement of feminine traits which are reflective of
emotional vulnerability and dependency, and the absence of
aggressiveness. Conversely, low maternal involvement in
early childhood is associated with bipolar masculine
attributes which reflect aggression, dominance and lack of
need for emotional support.

Table % reveals the correlation coefficients for Matinvl
through % and the M-F scale for each of the sexual

orientation groups.

Table % Intercorrelations between the FAQ M-F scale and
Matinv scales l-% for heterosexual and lesbian groups.

 

PAQ M-F Scale

 

Heterosexual Lesbian
Matinvl .33‘ .O7
Matinv8 .%O" .11
Matinv3 .83 .38‘
Matinv% .17 .08
* p<.05, *’ p<.01 (Correlations for heterosexual women are

partial correlation coeficients which control for education.)

As the table reveals, bipolar femininity/masculinity is
related to maternal involvement in the predicted direction
during early and late childhood for heterosexual women and
during adolescence for lesbian women. There are no

statistically significant relationships between the

88

independent Femininity and Masculinity scales and maternal

involvement at any of the four time periods assessed.

Hypotheses II
Early Childhood Maternal Involvement and

Erotic Orientation

3) Maternal involvement in early childhood is
positively related to androeroticism for women. High
maternal involvement is associated with high androeroticism.

%) Early childhood maternal involvement is inversely
associated with gynoeroticism in women. High maternal
involvement is related to low gynoeroticism.

The association between women’s early childhood
relationship with mother and their erotic orientation was
tested with the correlations between Matinvl and the EROS
gynoeroticism and androeroticism scales. Results show no
Significant relationships between early childhood
mother-daughter closeness and erotic orientation. The
hypotheses that distant and differentiating mother-daughter
relationships are associated with greater sexual interest in

women; and that close and enmeshed mother-daughter bonds

result in erotic interest in men are not supported.

Hypotheses III
Maternal Involvement and

Level of Differentiation of Object Representations

5) Levels of early childhood maternal involvement are
inversely associated with levels of differentiation in
prOjected object relations representations. High maternal

87

involvement in early childhood is associated with
undifferentiated and interrelated object representations.

Low maternal involvement is associated with highly
differentiated and isolated object relations representations.

Early childhood maternal involvement is significantly
correlated with the levels of differentiation of object
representations in TAT stories as predicted, for lesbian but
not for heterosexual women. Table 5 contains the partial
correlation coefficients (controlling for education and
family SE5) between maternal involvement and object

differentiation, for each of the sexual orientation groups.

Table 5 Partial correlation coefficients between object
representations differentiation and Matinvl.

 

Lesbian Heterosexual

 

Matinvl .%7* .OO
Matinv8 .%8* -.15
Matinv3 .55‘ .05
Matinv% .%%* -.85
“significant at p<.05 All partial correlations control for

education and family SE5.

Thus among lesbian women, high levels of early childhood
maternal involvement are associated with less differentiated
object representations, while low early childhood maternal
involvement is associated with highly differentiated object
relations representations.

(The direction of the correlation between education and

object representations differentiation differs for lesbian

88

and heterosexual groups, and is noteworthy because this group
difference also characterizes the relationship between
education and both gynoeroticism, and the bipolar M-F scale.
In each case, the heterosexual group shows a significant
positive relationship between education and each of the three
variables (object differentiation, masculinity, and
gynoeroticism); while education shows a significant negative
correlation with these variable for lesbian women. For

further discussion see appendix L.)

Hypotheses IU
Differentiation of Object Representations

and Femininity/Masculinity

5) Differentiation in object representations is
positively associated with masculine personality attributes.
Highly differentiated object representations is associated
with high masculinity.

7) Differentiation of object representations is
inversely associated with femininity. Low levels of object
representation differentiation is related to higher
femininity.

There is no evidence for the hypothesized association
between women’s gender characteristics and the differen-
tiation and interrelatedness of characters they prOject into
TAT stories of family relationships. The results of an
analysis of variance, and also correlation coefficients
indicate that there are no significant relationships between
the level of differentiation of object representations and

the three scales used to assess femininity/masculinity.

88

Femininity is not associated with prOjections of object-
dependent--and highly interrelated characters, and likewise
masculinity is not related to differentiated and more

isolated object representations.

Hypotheses U
Differentiation of Object Representations

and Erotic Orientation

5) High differentiation in object-representations is
positively associated with gynoeroticism.

5) Low differentiation in object-representations are
positively related to androeroticism.

The hypotheses that differentiation of object
representations would be associated with women’s erotic
interests were not supported. The correlations between the
EROS scales and object differentiation for the entire sample
of women and for each of the sexual orientation groups

assessed independently show no statistically significant

relationships.

Hypotheses UI

Erotic Orientation and Femininity/Masculinity

10) Gynoeroticism is associated with masculine
personality attributes.

11) Androeroticism is positively related to feminine
personality attributes.

The hypotheses which predicted relationships between the

FAQ and EROS scales were investigated with correlation

80

coefficients. These hypotheses predicted that gynoeroticism
would be positively associated with masculinity, while
androeroticism would be positively associated with
femininity. For the sample as a whole, the correlations
between the three PAQ scales and the two EROS scales reveal
only one significant relationship, between gynoeroticism and
the M-F scale in the predicted direction. These

intercorrelations are presented in Table 5.

Table 5 Correlations between the PAQ and EROS Scales

 

PAQ Scales
EROS F M M-F
Gynoeroticism -.1%@ .15 .83'
Androeroticism -.05 -.O5 .01

 

* p<.05 (G r is partial correlation controlling for
education.)

The significant but modest correlation between
gynoeroticism and the M-F scale supports the hypothesis that
women’s sexual interest in women is positively associated
with masculine personality characteristics.

The correlations between the PAQ scales and the EROS
scales for lesbian and heterosexual groups when computed
separately reveal three significant relationships between
gender attributes and erotic interests for heterosexual
women. Heterosexual women’s sexual interest in women is
inversely associated with their femininity as measured by the
F scale (r--.85, p<.05). As such more feminine heterosexual

women report less sexual interest in women. Significant

81

positive correlations between androeroticism and both the M
(r-.37) and M-F (r-.%7) scales indicate that heterosexual
women who score more masculine on these scales report a
greater range of sexual interests in men. (All correlations
reported here are partial corelations which control for
education, age and family SE5.)

When the correlations between the PAQ and EROS subscales
are computed independently for the lesbian group, there are

no statistically significant relationships.

Hypotheses UII

Identification with Mother

18). High identification with mother is be associated

with high early childhood maternal involvement.
Identification with mother is associated with early

childhood maternal involvement in the predicted direction,
however the cOrrelation does not reach statistical
significance. Both variables are significantly associated
with family SE5. When partial correlation coefficients are
computed to control for family 555 only Matinv3 (adolescence)
and Matinv% (adulthood) are significantly associated with
identification with mother. Table 7 reveals the partial
correlation coefficients for identification with mother and
Matinv1-%. (When these correlations are computed for lesbian
and heterosexual groups separately, there are no differences

in significance levels.)

88

Table 7 Partial correlations between identification with
mother and levels of Maternal involvement 1 through %.

 

Identification with Mother

 

Partial r _E
Matinvl .15 .08
Matinv8 .1% .13
Matinv3 .38 .003.
Matinv% .%0 .001'

 

13) High identification with mother is associated with
low levels of differentiation of object relations
representations.

The hypothesized relationship between identification with

mother and differentiation of object relations

representations was not supported.

1%) Identification with mother is positively correlated
with femininity.

The predicted relationship between identification with
mother and femininity was supported. Identification with
mother is correlated with the bipolar M-F scale, r-.8l,
p<.05. Women who perceive themselves and their mothers as
alike score in the feminine direction on this scale, while
those who perceive themselves and their mothers as disSImilar
score in the masculine direction. The correlations between
identification with mother and both the M and F scales are in
the predicted direction, but do not achieve statistical

significance.

83

Using a three way group classification based on women’s
identification with mother scores, and the median split
method of group classification on each of the PAQ scales,
Kendall’s Tau-C rank order correlation reveals significant
associations between identification with mother and all three
femininity/masculinity measures. The group of women who are
most highly identified with their mothers are less masculine
on both the M and M-F scales, and are more feminine on both
the M-F and F scales. Table 5 shows the jOInt classification
of women’s levels of identification with mother, and their

PAQ classifications.

Table 5 Women’s joint classifications of their three-way
identification with mother by their two way categorization on
the PAQ Scales

 

Identification with Mother

High Moderate Low

PAQ F

Low 5 11 1%

High 17 18 10 Tau c--.8%, p<.05
PAQ M

Low 17 11 10

High 5 11 15 Tau C-.85, p<.05
PAQ M-F

Low 17 13 5

High 5 10 15 Tau c-.31, p<.01

 

When the relationship between women’s identification
with mother grouping and their combined classifications on
the PAQ scales are examined the most dramatic differences are
found for the combined M and M-F scales. As the data in

Table 5 illustrates, the majority of those who are most

8%

highly identified with their mothers achieve combined low
scores on both measures of masculinity. In contrast, those
who are most disidentified with their mothers tend to achieve

high scores in the masculine direction on both scales.

Table 5 Frequencies of women’s 3-way classification of
identification with mother, by their combined M-F and M
claSSification.

 

Identification with Mother

High Moderate Low
LO M-F/LO M 15 8 7
HI M-F/LO M 8 3 3
L0 M-F/HI M 8 5 1
HI M-F/HI M 8 8 13

 

15) Identification with mother is positively assoc13ted
with androeroticism.

The hypotheSIs which predicted that identification with
mother would be associated with androeroticsm was not
supported. Women’s perceived similarity between themselves
and their mothers was not significantly associated with
either the androeroticism or the gynoeroticism scales.
Furthermore there is no relationship between identification

with mother and self-labeled sexual orientation.

85

Identification with Father
15) Disidentification with father is positively
correlated with femininity and androeroticism. High
identification with father is associated with low early
childhood maternal involvement and levels of object relations
representation characterized by high differentiation.
Identification with father was not significantly related

to any of the variables in the directions predicted by the

hypotheses.

Sexual Orientation and Gender Attributes:

Exploratory Findings

Exploratory analyses of the relationships between women‘s
self classification of their sexual orientation, their erotic
interests as measured by the EROS scales, and their gender
characteristics were carried out. These findings are

presented here.

Self Labeled Sexual Orientation and Gender Attributes

Correlations between women’s self classification of
their sexual orientation and their scores on the PAQ scales
are presented in Table 10. As the table reveals, when
women’s sexual orientation moves from heterosexual (1), to
bisexual (8), to lesbian (3), their scores on all of the PAQ

scales increase in a positive direction. Only the M-F scale
is significantly correlated with sexual orientation at a

p<.05.

88

Table 10 Group means and correlations between self lableled
sexual orientation and the PAQ Scales

 

PAQ Heterosexual Bisexual Lesbian r P
Scale
F 83.3 83.5 8%.5 .1% .11
M 80 80.5 81.5 .15 .05
M-F 18.5 13.7 1%.5 .88 .03

 

A one-way analysis of variance was computed to determine
if there are significant group differences for lesbian and
heterosexual women on each of the PAQ scales. The results
show no significant group differences, only the M-F scale
shows a trend of significance with an F value of 3.85,
(1,55,df) and p-.O7.

Using the median split method of classifying subjects’
conjOInt scores on the three pairings of PAQ scales, the most
striking difference shows that a higher percentage of
heterosexual women are classified as undifferentiated (i.e.,
achieving low scores on both F and M scales). Table 11
reveals the number of women who are classified into each of
the four quadrants using the F and M scales. A Chi-square
analysis of this data is significant at the p<.01 in a one
tailed test (Chi-square - 15.%5, df-5).

As the table reveals, lesbian and bisexual women are
much less likely to achieve combined low scores on the M and
F scales, though only the lesbian group shows a greater
frequency of androgynous women. Heterosexual women are least

likely to be androgynous, while both lesbian and heterosexual

87

groups show comparable frequencies of both feminine and

masculine sex-typed women.

Table 11 Number and percentage of women in each of the

sexual orientation groups falling into each of the four joint
classifications of M and F.

 

Heterosexual Lesbian Bisexual

Undifferentiated 1% 8 0

(L0 F/ LO M) (35%) (7%)

Feminine Sex-typed 5 5 5

(HI F/ LO M) (88%) (302) (55%)
Masculine Sex-typed 5 5 8

(L0 F/ HI M) (882) (302) (882)
Androgynous 7 8 1

(HI F/ HI M) (15%) (33%) (182)

 

Self-DeSIgnated Sexual Orientation and the EROS Scales

Self labeled sexual orientation was as expected, highly
associated with women’s scores on both gynoeroticism and
androeroticism scales. The correlation between gynoeroticism
and self-labeled sexual orientation is .75, androeroticism is
correlated -.55. Three trends are noteworthy. First, while
androeroticism is normally distributed, gynoeroticism has a
bimodal distribution. As this and the correlations suggest,
self designated sexual orientation is more strongly
predictive of womens scores on the gynoeroticism scale, while
sexual interest in men shows greater variance which is not
associated with women’s classification of themselves as
heterosexual, bisexual or lesbian. Secondly, the

androeroticism and gynoeroticism scales are inversely

88

correlated with each other with an r of -.31, p-.OO%. Yet
when a partial correlation coefficient is computed
controlling for sexual orientation, the gynoeroticism and
androeroticism scales are positively correlated (r-85,
p-.08). Thus within sexual orientation women’s sexual
interests in both men and women are positively associated.
The scales measure not only the direction of erotic interest,
but also the extent of erotic interest which covaries within
sexual orientation groups. And finally, lesbian women report
greater cross-sexual-preference sexual interests than
heterosexual women do. The lesbian group’s mean score on the
androeroticism scale is significantly higher than the

heterosexual group’s mean score on the gynoeroticism scale.

Erotic Orientation and Gender Attributes

A final investigation of the relationship between gender
characteristics and erotic orientation utilized a median
split classification of women on the gynoeroticism and
androeroticism scales. In this classification scheme women
are categorized as ’androerotic’, ’gynoerotic’, ’low-erotic’
and ’high-erotic’, based on their conjoint classification on
the two scales which measure their sexual interests in women
and in men. A four way classification of women on the EROS

scales is shown in Figure 1.

88

 

Gynoeroticism
Low High
High 85 Heterosexual 8 Heterosexual
Andro- 1 Bisexual 5 Bisexual
eroticism 0 Lesbian 5 Lesbian
Low 5 Heterosexual O Heterosexual
Andro- 1 Bisexual 3 Bisexual
erotICIsm 8 Lesbian 80 Lesbian

 

Figure 1 Number of women in each of the gynoeroticism and
androeroticism categories

As would be expected, there is substantial congruence
between women’s self labeled sexual orientation and their
classification on the combined gynoerotiCism and
androeroticism scales. The majority of heterosexual women
are in the category reflecting high sexual interest in men
and low sexual interest in women; while the majority of
lesbian women report high gynoeroticism and low andro-
eroticism. Five of the ten self-labeled bisexual women
report high sexual interest in both women and men. A
comparison of women’s scores on the PAQ scales based on this
classification scheme shows no differences between the
heterosexual and lesbian categories, but strong differences
characterize the ’low-erotic’ group (who report low sexual
interest in women and in men), and mild differences
characterize ’high-erotic’ group. The feature which most
distinguishes ’low-erotic’ women from the other groups is
their high femininity, reflected by their low mean score on

the M-F scale, as shown in Table 18.

100

Table 18 Means and standard deviations on the PAQ scales
and age for the ’low-erotic’ and ’high-erotic’ groups, and
all subjects

 

 

Asexual Bisexual All Subjects
Mean 5.0. Mean 5.0. Mean 5.0.
PAO M—F 5.8 8.5 15.7 5.% 13.5 %.5
FAQ M 15 3.5 88 %.5 80.5 %.3
PAD F 8% 3.5 88.8 %.0 83.5 %.8
Age 35 11 3O 5 38 8

 

While the small number of women who are categorized here
as ’low-erotic’ or as ’high-erotic’ precludes any generali-
zations from these findings, the data do suggest that
masculine personality characteristics are related to low vs.
high sexual interests in general. Further investigation of
the differences which distinguish the ’low-erotic’ women from
the other groups shows that a slightly greater proportion of
these women are in relationships of longer duration. Eight
of the twelve low-erotic women are in a relationship, with a
mean length of 11 years. The mean length of current
relationship for the 55 coupled women in the sample is 7.5
years. Thus the low-erotic women, who report low sexual
interests, are slightly but not significantly older, and on
the average they have been in intimate relationships slightly

longer.

101

Semantic Differential Item Analysis

Further investigation of the semantic differential
explored the correlations between the items on each of the
three semantic differentials and the distributions of women’s
scores reflecting their identification with their mothers,
identification with their fathers, and also their scores on
the maternal involvement scale. While these data are
provided in appendix M, the highest item correlations between
women’s ratings of themselves and their overall
identification with mother scores are presented here. Table
13 reveals the semantic differential items which were
significantly correlated with women’s identification with
mother scores. The adjective listed first represents the

direction associated with high identification with mother.

Table 13 Correlations between 50 items on which women
rated themselves and their identification with mother scores.
(Characteristics of self associated with identificatition
with mother)

 

Item Correlation
Conforming--Individualistic .30
Restrained--Expressive .85

Delicate--Rugged .85
Controlled--Spontaneous .83
Passive--Active .83
Providing--Needy .83
Submissive--Dominant .81
Cooperative--Competitive .81

 

Women’s ratings of their own characteristics which
correlated most highly with their identification with the;r

mothers suggest a dimension characterized by passive

108

conformity at one end and active indIVidualism at the other.
The direction of association is consistent with the findings
of many previous studies which have noted that women who have
highly involved relationships with their mothers, and
identify strongly with them show a pattern of feminine

characteristics which reflect the absence of agentic traits.

103

DISCUSSION

This study sought to empirically assess hypotheses which
propose that femininity, masculinity, and sexual orientation
in women develop in the early childhood process of attachment
to and differentiation from mother, and are mediated by
intrapsychic object relations representations. The
hypotheses were derived from two theories of gender
development, proposed by Nancy Chodorow and Dorothy
Dinnerstein. These theories pOSIt that aspects of the early
mother-child relationship lead to differences in gender
characteristics across the sexes. Extrapolating from the
theories, this investigation assessed whether the constructs
proposed by Chodorow and Dinnerstein are relevant to the
development of gender attributes and sexual orientation among
women alone. The hypotheses predicted relationships between
five primary variables; early childhood maternal involvement,
identification with mother, projected object relations
representations, femininity/masculinity, and sexual
orientation and erotic interests. The results of statistical
analyses show that few of the variables are significantly
related as predicted, and of these, even fewer of the
relationships are of strong magnitude. Thus overall there is

modest evidence for some tenets of the theories, while others

10%

are not supported. Of the significant findings, the
strongest support for both theories derives from the
significant relationship found between daughters’
identification with their mothers and their gender
characteristics. Here, as both theories propose, high
identification with mother (perceived similarity between self
and mother) is significantly associated with feminine
personality characteristics and inversely associated with

masculine personality characteristics.

Identification with Mother and Femininity/Masculinity

Women who perceive themselves and their mothers as alike
are more feminine and less masculine on all three scales of
femininity/masculinity. Chodorow and Dinnerstein both assert
that identity and personality organization that are based on
identification, similarity, mirroring, mutuality, and
emulation of mother are the central organizers of feminine
personality. They assert that where children disidentify
with their mothers, masculine attributes which are based on
separation and differentiation of self from others develop.
The findings of this study strongly support the relationship
between identification with mother and women’s psychological
femininity and masculinity. Identification with mother is
significantly associated with all of the two-way
classifications of the three Femininity and Masculinity
scales. Women who are most highly identified with mother are

high in feminine characteristics and low in masculine

105

characteristics, while women who are least identified with

their mothers are high in masculine characteristics and low

in feminine characteristics. When the combinations of
feminine and masculine characteristics are examined together,
the most dramatic differences emerge for the pairing of
independent masculine attributes and bipolar feminine/
masculine attributes (the M and M-F scales). Here the
majority of women who perceive themselves and their mothers
as most Similar endorse combined low scores on both scales,
demonstrating low masculine characteristics. In contrast,
the majority of women who were disidentified with their
mothers obtain high scores on both scales in the masculine
direction, reflecting high masculinity.

In the theoretical paradigm under investigation
’identification with mother’ is viewed as a process which
begins, importantly in earliest childhood. Chodorow and
Dinnerstein focus on early childhood development, proposing
that during this formative period of initial personality
organization, the extent of the child’s identification with
mother produces gender related differences in personality
organization. In this study, it is women’s adulthood
identification with mother that is assessed. Most
psychological theories view adulthood identification with
parents as the product of a longstanding, continuous and
stable developmental process. While in all likelihood
considerable stability in identification with a parent over
the child’s lifespan is probably the rule, change and

inconsistency also occur. While the degree of adulthood

108

identification with mother probably does result from an
ongoing childhood identification with her, this is an
untested assumption.

So comparatively, highly feminine women perceive
themselves and their mothers as alike, and highly masculine
women perceive themselves and their mothers as different.

Two alternative interpretations can be raised. First, it
could be argued that the participant’s femininity and
masculinity lead to differences in their perceptions of
similarity between self and others, and this is in fact not
inconsistent with the tenets of the theories examined.
However women’s femininity and masculinity are more strongly
associated with the extent to which they identify with their
mothers, and are less associated with the extent to which
they report being like their fathers. This difference
supports the claim that it is identification with mother, and
not father which is related to psychological femininity and
masculinity. Secondly, it may be that more feminine women
are similar to their mothers because mothers, in general, are
more feminine than daughters due to socio-cultural changes.

The positive relationship between maternal involvement
and identification with mother demonstrates that the levels
of closeness vs. differentiation daughters feel with their
mothers, and the extent to which they perceive their mothers
as involved, warm, and nurturing is associated with their
perceptions of similarity between self and mother. In a
similar manner the item analysis of the semantic differential

(Appendix M), reveals that overall identification mother is

107

most significantly related to a constellation of perceptions
of mother as nurturing vs. depriving, Open vs. closed, warm
vs. cold, etc... As expected, qualities such as mother’s
involvement, warmth, and nurturence throughout life are
assoc1ated With daughter’s identification with their mothers.
These findings are consistent with the well established role
of warmth vs. hostility in fostering childrens’

identification with, and emulation of, models and parents.

Identification with Mother and Agentic Characteristics

The semantic differential item analysis of women’s
ratings of themselves suggests that the degree of women’s
identification with their mother is associated with a
cohesive constellation of personality characteristics
compriSIng agentic vs. nonagentic traits. While the item
correlations between women’s ratings of themselves and their
overall identification with mother scores are modest--of all
the items on this scale--thOse with the strongest
correlations are conSistently reflective of a highly coheSIve
dimension characterized by passive-conformity vs. active
assertiveness. This finding is consistent with those of
numerous studies in the areas of child development,
achievement motivation and gender deveIOpment; high
identification and involvement with mother, especially
mothers who are themselves highly feminine sex-typed,
non-demanding and less differentiating is associated with a

range of personality characteristics in daughters reflecting

108

a lack of agentic characteristics. As many researchers have
concluded, daughters who do not individuate from their
mothers lack a range of desirable agentic personality
attributes.

Of the two variables which assessed aspects of the
mother-daughter relationship, the hypotheses of this study
focused primarily on maternal involvement in early childhood,
predicting that this variable would be associated with gender
characteristics (as well as other variables). The results
show that adulthood identification with mother--and not early
childhood maternal involvement--is more strongly related to
women’s femininity/masculinity. Thus it is similarity
between self and mother, and not how close or differentiated
daughter’s felt in early childhood, that is related to a

broader range of gender attributes.

Maternal Involvement

Women’s ratings of their early childhood relationship
with mother were significantly associated with gender
characteristics for heterosexual, but not for lesbian women.
Closeness vs. distance in the early childhood mother-daughter
relationship is associated with bipolar gender attributes
which reflect emotional expressiveness and need for others.
When the relationship between bipolar gender characteristics
and maternal involvement is assessed for other life stages,
heterosexual women show the strongest correlations for all

time periods except adolescence. Thus heterosexual women’s

108

early and late childhood relationships with their mothers are

significantly related to these gender-role characteristics,

while only adolescent maternal involvement is associated with
bipolar gender characteristics for lesbian women. Chodorow's
assertion that differentiation in the early childhood
relationship with mother produces differences in gender
characteristics is supported for heterosexual, but not for
lesbian women. Among lesbian women, psychological femininity
and masculinity are independent of their maternal involvement
in early and late childhood.

A note on causality in the relationship between maternal
involvement and adulthood personality functioning is in order
here. The parameters of women’s early childhood experiences
are not empirically examined, but only postdictively assessed
by self report. As such all significant findings are
associations, which only suggest support for causal
interpretations.

There is no relationship between early childhood
maternal involvement and the other two measures of gender

attributes, nor women’s erotic interest in either women or

men .

The Bipolar M-F Scale

Throughout all forms of data analysis, the scale which
measures bipolar feminine and masculine attributes shows the
strongest significant relationships with the other variables

examined in this study. Bipolar femininity/masculinity is

110

associated with sexual orientation, gynoeroticism,
identification with mother, and, for heterosexual women,
early childhood maternal involvement. It appears that the
gender characteristics which this scale measures, and not
those measured by the other scales, are most strongly related
to the mother-daughter and sexual orientation variables
explored. As discussed earlier, Chodorow and Dinnerstein do
construe psychological femininity/masculinity in bipolar
terms. The results of this study indicate that of the range
of gender characteristics, those which are bipolar are indeed
more likely to be related to the mother-child relationship

and related development for women.

The Proposed Mediating Uariable, Differentiation of

Object Relations Representation

This variable--differentiation of object relations
representations was hypothesized to mediate between the early
childhood mother-daughter relationship and both gender
characteristics and erotic orientation. The pattern of
statistical findings does not support a model of gender
differentiation mediated by structural differences in the
object relations representations women prOject into family
context. Thus as operationalized here, there is no evidence
for Chodorow’s assertion that differences in mental
representations of self and other underly feminine vs.

masculine personality attributes.

111

Maternal involvement in early childhood (and in fact at
all time periods) is strongly associated with differentiation

in object representations for lesbian, but not for
heterosexual women. A possible explanation for this sexual
orientation group difference implicates the greater strain in
family relationships which the lesbian woman experiences as a
result of her differing sexual orientation. For the lesbian,
maternal involvement, and degrees of closeness and warmth,
throughout life are likely to be more important to her real
and prOjected interrelatedness and differentiation in self

and family members.

Sexual Orientation and Gender Characteristics

The pattern of relationships between women’s sexual
orientation, and their gender characteristics are consistent
with the findings of previous studies. Prior studies have
also found lesbian women to be more masculine but not less
feminine. The lesbian women in this sample were more likely
to achieve combined high scores on the masculine and feminine
scales. Heterosexual women show a greater frequency of
combined low scores on feminine and masculine measures. Thus
the lesbian group show a higher frequency of androgynous
women, while a higher percentage of undifferentiated women
characterize the heterosexual group.

The strongest group differences emerge when examining
the combinations of feminine and masculine characteristics.

No group differences were found for each of the scales

118

independently, thus there are no strong differences between
lesbian and heterosexual women’s gender characteristics when
these are are investigated as single dimensions. Many recent
investigations of gender characteristics have noted similar
patterns, indicating that it is the combined influence of
feminine and masculine attributes which are most meaningfully
related to many areas of personality functioning, and group

differences.

Women’s Erotic Interests, Sexual Orientation

and Femininity/Masculinity

A number of interesting findings emerged from the data
provided by the EROS. These include further support for the
independence of gynoeroticism and androeroticism as separate
dimensions; differences in cross sexual-preference sexual
interests displayed by heterosexual vs. lesbian women; the
suggestive findings on gender attribute differences between
the low-erotic group and other groups; and the suggestion by
the date that the EROS scales may share a common dimension,
reflecting a general sexual appetite which may be related to
masculine personality characteristics in women.

The findings provide further evidence for the
independence of gynoerotic and androerotic orientations.
Though the gynoeroticism and androeroticism scales are mildly
inversely correlated across sexual orientation, they are
positively correlated within sexual orienation groups, and

Often enter into different patterns of relationships with

113

other variables. A number of variables studied here are
related to both measures in the same direction, indicating
that the degree of sexual interest in both women and men may
share some common dimensions.

Within each of the major sexual orientation groups,
women’s sexual interests in men and in women are modestly
positively correlated. The scales measure both the sex of
ones object of desire, and an additional factor, the
frequency and variety of sexual interests women have in both
women and men--which covaries. Group differences between
women who achieve high vs. low scores on both scales combined
suggest that women who report low sexual interests in both
women and men show signficantly lower scores on the bipolar
measure of feminine/masculine personality characteristics
(less masculine), and somewhat lower scores on the
independent measure of masculine characteristics.
Furthermore, among heterosexual women, both measures of
masculinity are positively and significantly related to the
degree of their sexual interest in men. Thus in women,
especially heterosexual women, masculine gender traits are
associated with increased sexual interests, appetite, and
arousal per se. While feminine characteristics reflecting
emotional vulnerability and interdependency are associated
with low sexual interests, masculine characteristics,
especially those reflecting the absence of emotional

dependency and greater aggressiveness are related to higher

levels of sexual interests.

11%

As Storms and others have reported, lesbian women show
greater sexual interests in men than heterosexual women show
in women. Most authors interpret this as a result of the
lesbian woman’s exposure to heterosexual socialization and
this undoubtedly is a major factor. SOCiety is largely
heterosexual, and in short, heterosexuality is widely
advertised. However the relationship between masculinity and
sexual interests suggests that another variable may also
contribute to this relationship. The slightly greater
tendency of lesbian women to be more masculine than
heterosexual women may increase their sexual interests in
general. Further support for this speculation derives from
the claim that lesbian women are more sexual than
heterosexual women. In studies of women’s sexual lives,
lesbian women have been found to have more active and more
varied sexual interests and behaviors. In this study
differences between the respective means on the gynoeroticism
and androeroticism scales shows that lesbian women report
more frequent and more varied feelings of sexual attraction,
desire, and fantasy. (Lesbian women report more frequent and
varied sexual interests in women than heterosexual women
report in men.)

Women’s masculinity was also significantly correlated
with their first awarenesss of sexual feelings, and with the
number of intimate relationships they have had. These
further suggest that masculine characteristics may lead to
earlier sexual awareness, and a larger number of intimate

relationships. (This latter variable refers not to number of

115

sexual partners, but to number of intimate relationships.)

Bisexuality

The ten bisexual women in this study obtained mean
scores which fall between the heterosexual and lesbian women
on the three femininity/masculinity measures. As Storms had
found in a previous study, bisexual women’s sexual interests
in women and in men are additive, i.e., bisexual women report

as much sexual interest in men as heterosexual women do; and

as much sexual interest in women as lesbian women do.

Patterns of Differences for Heterosexual and Lesbian Women

Many of the hypotheSIzed relationships between variables
were significant for only one of the sexual orientation
groups. In general the patterns of relationships between
variables often differ markedly, for lesbian and heterosexual
women. This unexpected trend highlights the importance of
women’s sexual orientation and may imply that many
differences in personality organization accompany differences
in sexual orientation for women.

In this study, differences in the directions of
interactions between gender attributes, erotic interests,
mother daughter relationships, and demographic factors
including especially education were found for lesbian vs

heterosexual groups.

Education was highly associated with many of the

118

research variables, and shows differing patterns of
relationships for heterosexual vs lesbian women. Among
heterosexual women, higher education is related to increases
in the differentiation of object representations, higher
gynoeroticism, and higher M-F scale scores (more masculine).
In contrast higher education is associated with decreases in
each of these variables for lesbian women. When the groups
are compared, lesbian and heterosexual women at higher levels
of education are more similar. If education is a causal
variable in these relationships, then it has a homogenizing
effect on differences which characterize lesbian and
heterosexual groups at lower levels of education. In all
probability, the better interpretation holds that women who
are more Similar along these dimensions are more achievement
oriented, and achieve higher levels of education. Some
evidence for this explanation can be found in previous
research with women’s scores on the M-F scale. Spence and
Helmreich found that for women, significant positive
correlations exist for the M—F scale and two achievement
motivation scales, and all three were related to career
aspirations of High School students. Other studies also
supported the role of the M-F scale in educational and
professsional achievement realms, especially for women
(Spence & Helmreich, 1575). Thus the similarities between
lesbian and heterosexual women at high levels of education
may not be the result of high education, but rather factors
which contributed to these women’s attainment of higher

education.

117

Methodological Considerations

Of the various research instruments two are insuffiCient
measures of the constructs they assess. Both the Maternal
Involvement Scale and the scale used to measure
differentiation Of object relations representations are
simple measures of complex phenomena. Greater
objectification and accuracy in assessing these variables
would necessitate more elaborate measures. And, methods
which would yield more precise discriminations of the
constructs are necessary for a clearer understanding of these
variables and the relationships they enter into. For
example, further illumination of the role of early childhood
maternal involvement in women’s gender and psychosexual
development would require a more elaborate assessment of the
many factors and complex patterning of mother-daughter

relationships.

APPENDIX A

118

APPENDIX A

THE MATERNAL INUOLUEMENT SCALE

Mother’s Relationship with You

Highly Maternal, Emotionally Close

1) Extremelinnvolved

Mother is very highly nurturing and maternal. She is
highly involved with her daughter and highly invested in her
role as mother. Mother encourages daughter to be ”Mommy’s
Girl” either by imitating maternal traits or by being close,
and sharing a great deal with mother. There is a strong
feeling of emotional closeness between mother and daughter.

 

8) Highly Involved

This mother is also highly nurturing and maternal, though
not to an extreme degree. She is involved in a very motherly
way, and encourages warmth and closeness. Daughter feels
quite close to mother, though not in an all-encompassing way.

 

Moderately Maternal, Medium Closeness

3) Moderately Involved

Mother is moderately nurturing and involved. She
empathizes with daughter, and responds to daughter’s needs
though she is also a person independent of her role as
mother. This mother encourages daughter’s independence and
separateness and to an equal degree she responds to closeness
and dependency.

 

%) Mildly Involved

This mother is nurturing and maternal at times, but not
frequently. She is not rejecting, but is not always maternal
and close and she encourages her daughter’s independence.
There is some sense of closeness but also a sense of
separateness.

 

Low Maternal, Emotionally Distant

5) Moderately Distant

Mother is not very nurturing or maternal. In either
active or passive ways she is distant. There is very little
feeling of closeness, and daughter feels distant and
separate. Daughter developes a strong sense of independence
quite early.

 

118

5) Extremely Distant

This mother is non-maternal and non-nurturing. She is
not warm or close, and there is a strong sense of distance
and of separateness. Mother is not responsive to daughter’s
needs, she is either actively or passively indifferent,
neglectful or even rejecting. Daughter is very independent.

APPENDIX B

180

APPENDIX B

PHOTOCOPIES OF TAT CARDS AND INSTRUCTIONS

Please allow for a minimum of 80 minutes to complete
this task. On the following pages you will find two
pictures. You are to write a story for each of the pictures
separately. Your story should have a beginning, a middle,
and an end. In other words, what is happening, what led up
to the events you describe, and what is the outcome? Also
include the thoughts and feelings of the characters. Please
write as you create your story, without formulating it
beforehand. Your stories may be as long as you like, but
spend at least 10 minutes, and not much more than 15 minutes
on each story. You will need only a pencil or pen, paper is
provided here.

Begin by looking at the first picture for about one
minute. Place it in front of you and begin a story.
Spontaneity and the natural unfolding of your story as you
create it are very important. Please write everything that
occurs to you down. If you wish to change your story, or
delete something, do not erase but draw a single line through
the words or portions you wish to delete. When you finish be
sure that you have included everything and add anything you
might have neglected—-for example the outcome of the story.

 

121

v5.” 1P7” .0

..
JMMM .

I:

Wu.

1..

 

122

 

APPENDIX C

 

123

Instructions This task is similar to the previous one. On
the next three pages you will find a series of words that
can be used to describe people. The words are arranged as
scales of opposites. Thus each word has an opposite at the
other end of a scale. You are asked to rate yourself, your
Mother, and your Father on these scales. For example:

 

 

POLITICIANS
Fair : : : : : : Unfair
Honest : : : : : : Dishonest

 

Think carefully about the person you are describing. If
you feel that the word or trait at one end of the scale
fully and strongly describes this person, place a check
mark in the space right next to that word. like this:

Fair X : : : : : : Unfair
or
Fair : : : : : : X Unfair
If you feel that the word closely describes this person, but
not completely, you should place your check mark as follows:
Fair : X : : : : : Unfair
or
Fair : : : : x X : Unfair
If one of the words is mildly or slightly characteristic
of this person place your check mark like this:
Fair : : X : : x : Unfair
or
Fair : : : : X : : Unfair
And finally, if you think that both words are neutral or

irrelevant in describing this person, place your check mark
in the middle space:

Fair : : : X : i : Unfair

Be sure to check every scale. do not omit any. and never put
more than one check mark on a single scale. As you proceed.
do not look back and forth through the scales. and do not

try to remember how you checked previous scales. Work at a
fairly high speed. and do not worry or puzzle over scales.
Your first impression and immediate feelings are usually best,
but please do not be careless. we want your true impressions.

You may find it helpful to think of the scales like this:

123a
Please rate your Mother

 

 

 

 

 

 

 

 

 

 

 

 

 

brave_____:____:__:___:____:___:___ Fearful
emotional_____:___:___:____:_____:___:____unemotional
possessive : :___:____:___:____:___nonpossessive
accessible__:__:_____:___:__:_____:___aloof
weak :_ __ ____°_____-___: strong
spontaneous___:__:____:__:___:___:____controlled
individualistic___:___:____:____:___:___:____conforming
needy____:____:____:____:__:_____: providing
compassionate :___:____:____:___:___: hard-hearted
bold :____:___:_____:_____:___:____meek
playful :______:____:___:_____:__: serious
hard____:____:__:___:___:____:_____soft
beautiful :___:_____:__:___:___:____ugly
1arge___:____:___:____:____:____:____small
nurturing :_____:___:_____:_____:_____:___depriving
compliant__:__:___:___:___:____:____stubborn
warm__:_______:_____:_____:___:___:____cold
rational__:___:___:____:___:___:____intuitive
accepting : : :____:___:______: rejecting
restrained____:__:____:____.____:___: expressive
independent_____:_____:___:____:_____:____:____dependent
open____:____:___:____:___:___:____closed
competitive___:___:_____:____:____:___:___cooperative
calm______:____:____:____:___:___:____agitated
harsh__:___:____:___:__:____: gentle
generous____:___:_____:__:___:___:_____stingy
rugged_:_____:___:___:_____:___:___delicate
good__:____:____:____:___:___:_____bad
peaceable___:___:___:____:___:___:__hostile
distant___:___:_____:__:___:__:___close
passive_____:__:___:___:___:____: active
dominant___:____:___:___:___:__:____submissive

1ntolerant___:__:___:___:____:___:____understanding
namra1___:____:___:___:_____:___: pretentious
loyal__:____:____:___:___:____:___disloyal
rigid : : : : : :____flexib1e

 

HPPEND I X D

 

 

18%

RPPENDIX D

INSTRUCTIONS FOR SCORING LEUEL 0F DIFFERENTIRIION

This dimension measures two components of the
differentiation of characters in the story. The first
component is the interrelatedness and refers to the extent to

which characters are merged, interrelated or isolated from
one another. The second component is called individuation

and refers to the complexity of the character’s personality.
The levels here are merged, superficially individuated and
fully individuated. The levels of these two components are:

Interrelatedness Individuation
merged Merged
Interrelated Superficial
Isolated Full

These are combined on the scoring dimension as follows:
Level of Differentiation

1) merged

E) Superficial Individuation and Interrelated
3) Superficial Individuation and Isolated

%) full Individuation and Interrelated

5) Full Individuation and Isolated

At level 1 the characters in the story are not
acknowledged or described individually in any way. There is
no mention of any independent thoughts or feelings of the
characters. They are depicted as a single unit, all
experiencing the same thing, as though they were one person.

For example: ” This family Just had dinner and they are
relaxing together”.

Each of the next four levels have two separate but
related components. These are the level of interrelatedness
and level of individuation. In the first category, Just
described, characters are not at all individuated or
separated and are therefore interrelated at the
extreme--merged. all stories not coded level 1 will at least

address and describe the characters as individuals to some
degree.

Interrelatedness
This factor refers to interpersonal relatedness, the

nature and degree to which characters respond to one another.
People respond to others in a variety of ways. They may

185

respond with sympathy, empathy and understanding, or an
emotional reaction such as anger. Is there a sense of
recognition, understanding, communication, reacting? Does
the character perceive and react to others, or is there a
sense of isolation and insulation from others. The quality
of the emotional tone is an important consideration.
Shallowness, defensiveness, and emotional distance may belie
isolation. anger and conflict do not necessarily entail
isolation. The key consideration is: Is the character
responding or reacting to others? If so code interrelated.
where the character does not respond to others, code
isolated.

Individuation

The individuation factor refers to the degree to which the
numerous possible traits and aspects of an individual’s
personality are developed, expressed or implied. This is
similar to the notion of ’character development’ as we think
of this in describing literature. Personalities are complex,
people have thoughts, feelings, motives, intentions,
attitudes, values, beliefs, etc.. A fully developed
character is one that strikes us as having complexity, depth,
integration, and reality. a superficially individuated
character is one that seems shallow, narrow, and sketchy.

The amount of description of a character does not necessarily
constitute full individuation. Rely on your intuitive sense
of the character, Just a few descriptors may give an
impression of a solid and unique individual.

An expanded description of the scoring categories is
presented below.

I) Merged: Family members are merged, a single unit. Are
different characters acknowledged? are their experiences
separate and different, or could they be essentially one
person?

a) §gperficial Individuation and Interrelated: There is
superficial individuation, individuals are acknowledged but
are two dimensional and are interrelated to one another.
Here the characters are lacking depth or complexity. Their
experiences are relatively shallow or simple. They are all
either sharing the same experience or interacting with each
other without a sense of an individual responding because of
her/his unique personality.

 

3) gyperficial Individuation and Isolated: The character is
isolated from others, acknowledged but superficially
individuated. Again characters lack depth or complexity.
There may be some reaction to each other, but it is without
interaction. They may be aware of the experiences of others
but are focused on their own experiences.

185

a) Full Individuation and Interrelated: Thoughts and
feelings are acknowledged in a full manner. One has the
impression that this is a real person responding to others in
a manner which indicates genuine relatedness.

S) Eyll Individuation and Isolated: There is a sense of a
fully individuated character, but characters are isolated
from others. In some manner there is an inability or
unwillingness to communicate, know and understand the
experiences of other characters.

 

I.

HPPEND I X E

think you are.

187

APPENDIX E

The Personal Attribute Questionnaire

The items below inquire about what kind of person you

Each item consists of a pair of

characteristics with the letters A-E in between.
example:

is you cannot be both at the same time.
scale between two extremes.

Not at all artistic

A..B..C..D..E

For

Uery Artistic

Each pair describes contradictory characteristics--that

you think describes where you fall on the scale.

example,
choose A.

D.

1)
8)
3)
H)
5)

8)
7)

B)
S)

10)
11)
18)
13)

15)
15)

16)
17)
18)
18)
80)
81)

88)

83)
8%)

If you are only medium you might choose

Not at all aggressive

Not at all independent

Not at all emotional
Uery submissive

Not at all excitable
in a maJor crisis
Uery passive

Not at all able to devote
self completely to others

Uery rough
Not at all helpful
to others

Not at all competitive

Uery home oriented
Not at all kind
Indifferent to
other’s approval

Feelings not easily hurt

Not at all aware of
feelings of others
Can make decisions
easily

Gives up easily
Never cries

Not at all self-confident

Feels very inferior

Not at all understanding

of others

Uery cold in relations

with others

Uery little need for
security

Goes to pieces under
pressure

D

D DDDDD D DD DDDD DD DD DDDD

b nnnnn b hh hhhb no no nnnn

b mmmbb w bb bmmm mm mm mmmm

.B..C.
.BCOCO
..B..C.

u ccbub u 66 6666 no no Dean

R hhhhh m hm hhmm mm mm mmmm

GD. GE
.D..E
.D..E

The letters form a
You are to circle a letter which

For

if you think you have no artistic ability you would
If you think you are pretty good, you might choose

C, and so forth.

Uery
Uery
Uery
Uery
Uery

aggressive
independent
emotional
dominant
excitable in a

maJor crisis

Uery

active

Able to devote self
completely to others
Uery gentle

Uery

helpful

to others

Uery
Uery
Uery

competitive
worldly
kind

Highly needful of
other’s approval
feelings easily hurt

Uery

aware of

feelings of others
Has difficulty making
decisions

Never gives up easily
Cries very easily

Uery

self-confident

Feels very superior

Uery

understanding

of others

Uery

warm in relation

with others

Uery strong need for
security

Stands up well under

pressure

188

Scoring for the PAQ

The following items are scored on each of the three

scales:

The Femininity scale; Item numbers 3, 7, B, S, 18, 15, 81,
22.

The Masculinity Scale; Item numbers a, 6, 10, 16, 17, 18, 20,
8%.

The Masculinity-Feminity Scale; Item numbers 1, s, S, 11, 13,
1%, 18, 33.

APPENDIX F

188

The Erotic Orientation Scale

Sexual Experience Questionnaire

The following questions ask about your sexual
experiences and feelings during the past 12 months. Please
read each question carefully and indicate whether you have
had the experiences or feeling being asked about-~never (0),
only once or twice (1-2), three to six times (3-6), seven to
twelve times (7-12), an average of once or twice a month

(monthly), or an average of once or twice a week (weekly), or
almost daily or more (daily), during the past 12 months.
Please circle your answers.

1) How often have you noticed that a man you’ve seen or met
for the first time is physically attractive to you?

Number of Times During the Past 12 Months
0 1-2 3-6 7-12 monthly weekly daily

2) How often have you noticed that a woman you’ve seen or met
for the first time is physically attractive to you?

Number of Times During the Past 12 Months
0 1-2 3-5 7-12 monthly weekly daily

3) How often have you had any sexual feelings (even the
slightest) while looking at a man?

Number of Times During the Past 12 Months
0 1-2 3-5 7-12 monthly weekly daily

H) How often have you had any sexual feelings (even the
slightest) while looking at a woman?

Number of Times During the Past 12 Months
0 1-2 3-5 7-12 monthly weekly daily

5) How often have you felt some sexual arousal from touching
or being touched by a man?

Number of Times During the Past 12 Months
0 1-2 3-5 7-12 monthly weekly daily

130

5) How often have you felt some sexual arousal from touching
or being touched by a woman?

Number of Times During the Past 12 Months
0 1-2 3-6 7-12 monthly weekly daily

7) How often have you thought about what it would be like to
have a sexual experience with a man?

Number of Times During the Past 12 Months
0 1-2 3-5 7-12 monthly weekly daily

8) How often have you thought about what it would be like to
have a sexual experience with a woman?

Number of Times During the Past 18 Months
0 1-2 3-5 7-12 monthly weekly daily

8) How often have you felt a desire to have a sexual
experience with a particular man you know?

Number of Times During the Past 12 Months
0 1-2 3-6 7-12 monthly weekly daily

18) How often have you felt a desire to have a sexual
experience with a particular woman you know?

Number of Times During the Past 12 Months
0 1-2 3-6 7-12 monthly weekly daily

11) How often have you daydreamed about having a sexual
experience with a man?

Number of Times During the Past 12 Months
0 1-2 3-6 7-12 monthly weekly daily

12) How often have you daydreamed about having a sexual
experience with a woman?

Number of Times During the Past 12 Months
0 1-2 3-6 7-12 monthly weekly daily

13) How often have you dreamed at

night about having a sexual
experience with a man?

Number of Times During the Past 12 Months
0 1-2 3-6 7-12 monthly weekly daily

1%) How often have you dreamed at night about having a sexual
experience with a woman?

Number of Times During the Past 12 Months
0 1-2 3-5 7-12 monthly weekly daily

131

15) How often have you masturbated while fantasizing a sexual
experience with a man?

Number of Times During the Past 12 Months
0 1-2 3-6 7-12 monthly weekly daily

18) How often have you masturbated while fantasizing a sexual
experience with a woman?

Number of Times During the Past 12 Months
0 1-2 3-5 7-12 monthly weekly daily

 

APPENDIX B

138

APPENDIX B

Questionnaire

1) what is your age?

8)

3)

d)

5)

E)

7)

S)

10)

what is your race or ethnicity?

Please indicate the highest level ofeducation that you
have completed:

Less than seven years of school
Junior High School

Partial High School

High School Graduate

1 year of College

a years of College

Graduate or Professional Training

Please indicate your current employment status:
Unemployed Employed Student

Briefly describe the nature of your occupation:

what is your present socio-economic class?

Lower Class
working Class
Middle Class
Upper-Middle Class
Upper Class

Please indicate your present relationship status:

Single
Coupled
Married

How would you describe your views of women?

Traditional
Liberal
Femininist

How would you describe your sexual orientation?

Heterosexual
Bisexual
Lesbian

11)

18)

13)

1%)

15)

133

If you are currently in an intimate relationship please
answer the following:

a) How long have you been in this relationship?

b) Do you live with your partner? yes no

c) Is your partner a man woman

Please estimate the total number of significant intimate
relationships that you have had (for example
relationships that lasted a year or more, or

in which you lived together):

0 1-2 3-& 5-7 B or more
How long did your longest significant relationship last?
How old were you when you first became aware of having
sexual sensations or feelings?

How old were you when you first became aware of your
sexual orientation?

Family Background Information

16)

17)

18)

18)

Please respond to any of the following that describe
your family background:

a) I was raised with both parents throughout childhood

b) My parents separated or divorced when I was years
old, and I continued to live with my

c) I was raised by one parent, my

d) I was raised by somone onther than parent (explain)

As a little girl were you ”Daddy’s Princess?”? Yes No
As a little girl were you ”Mother’s Girl”? Yes No
As a young girl were you a ”tomboy”? Yes No

Please check any of the following words that you feel
describe your father’s behavior toward you during your
childhood before you were age 10: (If you knew

your father)

warm Supportive Close Flirtatious Demanding
Distant Unemotional ReJecting Hostile Seductive

was your biological mother your primary caretaker (the
person who took care of you) from birth to age 5?

If not, who was your primary caretaker during

those years?

13%

20) was there any period of separation in your relationship

with your mother or primary caretaker before you were
5 years of age? Yes No

If so, how old were you at the time?
How long were you separated?

21) On the following page you will find six categories that
describe mother-daughter relationships. Please check
the one, and only one that best describes your
relationship with your mother when you were age 5
and under. If your primary caretaker was someone other
than your mother, consider that this was your ’mothering
figure’and check the category that describes your
relationship with that person. These categories focus
on differences in how nurturing and ’maternal’ your
mothering figure was, and on how close you felt to that
person during your preschool years.

22) Refering back to the categories on the previous page,
please fill in the number of the category that best
describes your relationship with your mother or
mothering figure when you were:

Age 6 to 11
Age 12 to 18
Your present relationship

You do not have to answer this question if you choose not to.

23) In recent years it has become clear that sexual abuse,
rape and incest occur more frequently than was previously

thought. were you ever a victim of sexual abuse, rape or
incest? Yes No

If so, who was the abuser?

Father Brother an acquaintance
Stepfather Sister a stranger
Grandfather Mother Other

Uncle Aunt

How old were you at the time?

I want to thank you for your participation in this
study. Please keep the letter that includes my name and
address, if you would like a summary of the finkings, send me
a post card with your address and a summary will be sent to
you this summer. Please be sure to return the remainder of
this packet in the envelope here.

Sincerely,

APPENDIX H

135

APPENDIX H

MOMEN NEOED FOR PARTICIPATION IN A RESEARCH STUDY ON WOMEN’S
PERSONALITY DEUELOPMENT

I am conducting a research study of women’s personality
development and I need women to participate. The study
involves completing 5 paper and pencil measures in the

privacy of your own home. About 1 to 1 and 1/2 hours of your
time will be needed. No identifying information is

requested, so your responses are completely anonymous.

The study is designed to increase our understanding of
similarities and differences in women’s experiences of
themselves and others. The measures to be completed are
interesting and varied. One invites you to use your
imagination, while others ask you to think about yourself or
others.

Your participation is completely voluntary, you may
withdraw at any time. If you are interested in
participating, you can obtain the packet of research measures
from
When you complete the measures, you can return them to me in
the stamped, addressed envelope that is included. Again, no
identifying information is requested so your responses are
anonymous.

If you are interested in more information about the
study, or have any questions, please feel free to contact me.
I hope you will consider participating, in an effort to
increase our knowledge of women’s personality development.

APPENDIX I

138

APPENDIX I

Letter to participants

Dear participant,

I want to thank you for your willingness to contribute
to this research study. The study focuses on a number of
aspects of women’s development and experience. These include
the differences women may have experienced in their childhood
relationships with their mothers and fathers, their current
experiences of themselves, of intimate relationships and of
sexuality. During the past two decades issues regarding
women’s development, personality similarities and differences
and sexuality have increasingly begun to be met with
openness, honesty, humanness, and acceptance. The
differences among women in their life experiences and forms
of self-expression in all realms are increasingly
acknowledged and respected. In this atmosphere we are better
able to seek an honest understanding of ourselves and of
others.

Your participation in this study is completely anonymous
and voluntary. If at any point in time you choose not to
participate, please return the unused research measures in
the envelope provided. In order to assure anonymity, no
identifying information is requested.

The contents of this packet should take one to one and a
half hours to complete. Instructions are provided for each
of the five tasks. None of the measures require great
concentration, in fact spontaniety is recommended for the
first four. Each of the five tasks will take about 10 to 15
minutes to complete. Please complete each individual measure
at one sitting, though you may complete the different
measures at different times. It is very important that you
complete each measure in the given order, and that you not
look ahead to other measures.

This study looks at aspects of personality and of
women’s relationships. The measures are not designed to
assess or look for ’mental disturbance’ of any kind. Two of
the measures inquire about sexuality and sexual orientation.
Becauses of biases in some past research, some participants
may be Justifiably concerned about the perspective of the
study. The research perspective has been explained to a
number of community prefessionals and women’s groups
including the Michigan State University Lesbian/Gay Council,
Terry S. Stein, M.D., and Marty Aldenbrandt, Ph.D. These
persons and organizations have endorsed the study with their
approval.

Again, no identifying information is requested of you,
Please be careful that you do not add any identifying
information on the research forms. Your completion and
returning of the research measures will indicate your

137

voluntary consent and participation in the study. The
returned measures will be treated with the strictest
confidence. No persons except the researcher and her
comittee members will have any claim to direct access to the
measures completed by participants.

1 would like to request your aid in this study, in order
to enhance our knowledge of women, our past and current
relationships, our experiences of ourselves and our
sexuality. Please redspond to the enclosed measures in an

openm, honest fashion. Your honesty is crucial if we are to
obtain a realistic understanding of these aspects of our life

experiences as women.

I hope to collect all of the study responses before
Your effort in returing them to me as soon as possible will
be greatly appreciated.

Thank you for your time and cooperation,
Sincerely,

If you would like a summary of the findings of this
research study, please write to:

APPENDIX J

138

APPENDIX J

The TAT Measure

Development of the TAT Measure

The TAT Measure was developed in a prior study (Berry,
1985) and was designed for this research proJect. The pilot
study developed this TAT Measure to assess the nature and
level of maternal involvement in early childhood, and
concomitant degrees of differentiation from mother. This
measure consists of an interpretive scoring system for
stories told to two TAT cards. The TAT measure classifies
respondents into one of four groups based on differences in
the structural and thematic content of their TAT stories.
The four groups are: 1) High, nurturing maternal involvement
and low differentiation from mother; 2) Moderate level of
maternal involvement and nurturence and moderate
differentiation from mother; 3) Low levels of maternal
involvement and nurturence and high differentiation from
mother; &) Unclassified.

The TAT measure was developed in the following manner.
A scoring system based on the concepts of Chodorow’s theory
was constructed for stories told to two TAT cards. SubJects
who were clients in psychotherapy provided TAT stories while

their therapists classified them into one of three groups

133

which were consistent with the first three groups (groups 1,
2 and 3) Just described. Thirteen of the 23 items of the
scoring system were found to be significantly associated, in
the predicted direction, with the therapists’ group
classification of subJects. Only these 13 items which
significantly differentiated between groups were retained for
the measure. Based on the findings, three scoring scales
were constructed using three different methods of scoring the
13 items. Each of these three scales measures the presence
of thematic and structural differences in the subJect’s TAT
stories which were predictably associated with one of the
three early childhood differentiation groups. For example,
Scale 3 awards points for the themes which are associated
with the psychodynamics of women from Group 3, those who were
highly differentiated from their mothers in early childhood.
Thus each scale consists of 13 items, scored differently, in
order to maximize group differences. Scale 1 maximizes the
scores of women from Group 1, women who were close to and not
differentiated from their mother’s in early childhood. Scale
2 maximizes the scores of women from the moderate
differentiation group, Group 2. And high scores on Scale 3
are associated with high differentiation in the early
mother-child relationship, representing Group 3. Using these
three scales, the maJority of subJects in the pilot study
were found to achieve high scores on only one of the scales.
SubJects who fail to achieve criterion (high) scores on any
scale are not classified. The interpretive scoring system is

FD

1&0

for use with two TAT cards, the Family Scene Card, and the
Mother-Daughter Card.
The data from the TAT measure consists of women’s scores
on three scales, and, based on these scores, their
classification into one of three mother-daughter

differentiation groups.

Comparison of the Two Measures of Differentiation

from Mother in Early Childhood

Two independent measures were originally intended to be
used to assess mother-daughter closeness vs. differentiation
in early childhood. The first measure consists of a six
level rating scale on which women classified their early
childhood relationship with mother (Maternal Involvement-1,
or Matinvl). This scale is designed as a continuum, varying
from low differentiation at level one, to high
differentiation at level six. The second method of assessing
early mother-child differentiation consists of the TAT
Measure Just described.

Comparing women’s self-classification of their early
childhood relationship with mother on the Matinvl scale, with
their group classifications derived from the interpretive TAT
Scales it is evident that the two measures are related though
not very strongly. Table 1% shows the relationship between
women’s six level maternal involvement-1 ratings and their

TAT Group classification. Kendall’s Tau C, a rank order

 

1&1

correlation coefficient yields a correlation of .16, p<.QS.

Table 13 Joint frequency of women’s self classification on
Matinv1 by their TAT Group classification.

 

Maternal Involvement-1 Scale

 

1 2 3 S 5 6 N
TAT Group
Group 1 2 1 5 Q Q Q S
Group 2 2 S 2 S S Q 22
Group 3 3 3 a s 3 2 13

 

If the Matinv1 scale is collapsed into three categories
the match between the two methods of classifying subJects is
poor, i.e., more subJects are mismatched than are matched on
the two classifications.

Further comparison of these two measures examined the
correlations between the hatinvi scale and TAT Scales 1 and
3, all of which were designed as linear measures. Pearsons r
correlation coefficients yield a small but significant
correlation between Scale 1 and Matinv1 (r--22, p<.05) in the
predicted direction; while the correlation between Scale 3
and Matinv1 is in the predicted direction but does not
achieve significance (r-.1S, p-.@7). Thus the two measures
of early mother-child differentiation are related, though the
magnitude of their association is small. while both methods
are measuring similar or related variables, they are clearly
also measuring different and distinct variables.

Given the modest association between the measures, it

became necessary to choose one of the methods for the data

1&8

analysis. As described earlier, only 58 of the 77
respondents in this study returned TAT Stories. And
furthermore, B of the women who provided TAT Stories were
unclassifiable on the TAT Measure. As such the use of the
TAT Measure involved substantial reduction in the sample size
available for data analysis. The Maternal Involvement scale
was thus used to assess mother-daughter closeness in early
childhood for the hypotheses of this study. A presentation
and discussion of further analysis of the TAT measure is

contained here.

The TAT Measure

The hypotheses of this study had predicted that the
construct, early mother-child differentiation, would be
related to women’s identification with each parent, their
gender characteristics, erotic interests, and obJect
relations representations. Table 15 reveals the correlation
coefficients between each of these variables and the TAT

measure .

 

1&3

Table 15 Correlations between TAT Group, Scale 1, Scale 3,
and each of the primary research variables

 

Scale 1 Scale 3
Uariable
Matinv1 -.22* .18
Matinv2 -.15 .16
Matinv3 -.08 .18
Matinv& .06 .DD
PAQ F .08 -.1S
PAQ M .06 -.1&
PAQ M-F -.08 .0&
Sexual-Drien .02 .00
Gynoeroticism -.1& .11
Androeroticism .02 -.25*
ID Mother -.05 ‘ .27'
ID Father .03 -.1Q

 

Overall, few of the correlations are significant, and
those which are show only modest relationships. The
statistically significant relationships are as follows: Scale
3, is significantly correlated with androeroticism. women
who are more highly differentiated from mother on the TAT
scale 3, show, as predicted, lower sexual interest in men.
The significant correlation between scale 3 and
identification with mother indicated that women who are
highly differentiated from mother on Scale 3 are less
identified with her, as predicted.

APPENDIX K

1&&

APPENDIX K

The Secondary Uariables

Hypotheses for the Secondary Uariables

A number of variables were assessed on the questionnaire
respondents completed. Hypotheses were offered for four of
these variables and the primary research variables as
follows:

17) Feminist ideology is positively associated with
androgyny.

18) Tomboyism is positively associated with gynoeroticism.
18) Having been ”Daddy’s Princess” and ”Mother’s Girl”in
childhood is positively associated with feminine personality
attributes and androeroticism.

20) Age of realization of lesbian sexual orientation is
positively associated with femininity and inversely
associated with masculinity.

Significance tests for these hypotheses utilized the Chi
Square for hypotheses 16, 17, and 18, and correlational
analysis for hypothesis 18.

No hypotheses were offered for the remaining secondary
variables, however the relationships between these and other

variables were explored.

1&5

Results of Data Analysis: The Secondary Uariables

Feminist Ideology

The participants rated their views of women as
1-Traditional, 2-Liberal, or 3-Feminist. The hypothesis that
feminist ideology would be positively associated with
androgyny was not supported. women’s views of women are not
significantly related to any of the three scales which
measure gender characteristics. An examination of the
associations between this variable and the other research
variables yields the following pattern of relationships:
Feminist ideology is significantly associated with sexual
orientation, lesbian women are more apt to classify
themselves as feminist, while the maJority of heterosexual
women classify themselves as liberal. Twenty of the 28
lesbian women are feminist, while only 8 of 38 heterosexual
women view themselves as feminist. Feminist ideology is not
related to age or to the participants’ self ratings of their
socio-economic status. It is, however, related to higher
education, high family SE5, single marital status, and the
Hollingshead and Redlich index of subJects SES (higher SE5).
There is no relationship between women’s views of women and
their relationships with their mothers at any of the four
time periods. Feminism is significantly associated with
women’s perceptions of their fathers’ behavior toward them as

demanding (Chi Square p-.03). Feminist women saw their

1&6

fathers as having been demanding more often than did liberal

or traditional women.

Tomboyism

The hypothesis that women who had been tomboys would
report higher levels of sexual interest in women as measured
by the gynoeroticism scale was not supported. A
nonsignificant trend (p<.07) does suggest that heterosexual
women who were tomboys have higher scores on the
gynoeroticism scale. Tomboyism is significantly related to
sexual orientation, however, the maJority of women in this
sample from both lesbian and heterosexual groups indicate
that they had been tomboys during their childhoods.
Twenty-four of the 28 lesbian women, 25 of the 3S
heterosexual women and three of the ten bisexual women
reported having been tomboys. women who stated that they had
been tomboys reported an earlier awareness of sexual
feelings, and for lesbian women tomboyism was positively
associated with an earlier awareness of their sexual
orientation. None of the femininity/masculinity scales are
significantly associated with tomboyism, although a
nonsignificant trend (p<.06) suggests that heterosexual women
who were tomboys are less feminine on the F scale. And
finally, tomboyism is significantly associated with higher

levels of education.(All significant p’s <.QS.)

1&7

Hypothesis 15

This hypothesis predicted that women who had been

”Daddy’s Princess” and ”Mother’s Girl” during their

childhoods would be more feminine, and more androerotic.
Among heterosexual women these were both significantly
associated with lower scores on the F and the M-F scales,
while lower scores on the M scale characterize only women who
had been ”Daddy’s Princess”. Heterosexual women who had been
mother’s girl or daddy’s princess have lower, not higher
scores on androeroticism. Thus for heterosexual women,
having been daddy’s princess in childhood is associated with
lower levels of sexual interests in men, and lower scores on
all three scales of gender attributes. Lesbian women show
fewer significant relationships with these variables, but
like their heterosexual peers lower scores on the M scale
characterize Daddy’s Princesses’s, and lower androeroticism
is associated with having been Mother’s Girl. Having been
’Mother’s Girl’ is significantly associated with low levels
of early childhood maternal involvement, and high family SES

for all subJects. (All p’s (.05)

Hypothesis 20

Age of realization of sexual orientation was not related
to gender characteristics for lesbian women, as had been

hypothesized. Only one of the femininity/masculinity scales,

1&8

the M-F scale was associated with this variable for
heterosexual women. Here, earlier awareness of sexual
orientation is related to higher scores, and thus greater
masculinity. Significant group differences characterize
women’s age of first realization of sexual orientation with
lesbian women reporting later awareness of their sexual
orientation. For the sample as a whole, the range of age of
realization of sexual orientation spans from 3 to &1 with a
mean age of 15 and a standard deyiation of 6.7 years. The
skew of the distribution shows that 20% of the women fall
before age 11, 762 before age 17 and a gradually decreasing
frequency of women realize their sexual orientation after age
17. The group differences for heterosexual and lesbian women
are as follows; the mean age of realization of sexual
orientation for heterosexual women is 12 with a standard
deviation of 3 years; the mean age for lesbian women is 18
with a standard deviation of 7 years.

Lesbian and heterosexual women report no differences in
the age at which they first became aware of having sexual
feelings per se. Both groups report a mean age of 11, the
standard deviation for heterosexual women is 3 years, while
it is 5 years for lesbian women. The distribution spans from
age 3 to age 18.

For lesbian women, significant correlations between age
of first awareness of sexual feelings and both the
androeroticism and gynoeroticism scales show that earlier
awareness of sexual feelings is related to higher sexual

interests in both women and men (p’s<.Dl). A similar

1&8

pattern of nonsignificant trends indicates that earlier first
awareness of sexual orientation is perhaps associated with
both higher gynoeroticism and androeroticism (p’s (.10). In
both instances, lesbian women who report earlier awareness of
sexual feelings, and of their sexual orientation have higher
sexual interests in both women and men. There are no
significant relationships between these variables for

heterosexual women.

Relationship Uariables

A variety of differences in relationship patterns are
evident for lesbian and heterosexual women. Not
surprisingly, lesbian women report that their longest
relationships are much shorter than those of heterosexual
women. Among the heterosexual women in this sample the mean
length of their longest intimate relationship is 11 years,
with a standard deviation of 16 years. For lesbian
respondents, the mean length of their longest relationship is
5 years, with a standard deviation of 5 years. (There are no
age differences in the two groups--means for age are 31 and
32, with standard deviations of 8 years for both groups.)
There are no sexual orientation group differences in current
relationship status, or the number of previous significant
intimate relationships women have had. The maJority of women
in the sample have had either one or two significant
relationships (31 women), or three or foUr significant

relationships (35 women).

150

women who report having had fewer significant
relationships tend to be in a relationship at present, and
are more likely to be traditional or liberal as opposed to
feminist in their views of women. These women also report a
later awareness of sexual feelings.

The relationship between women’s femininity/masculinity
and their relationship patterns shows that high scores on the
bipolar M-F scale (in the masculine direction) are associated
with having had more intimate relationships. Among
heterosexual women those who are currently in relationships
have higher scores on the F scale, as such femininity is
significantly related to current involvement in an intimate
relationship for heterosexual participants.

The number of previous significant relationships women
report is positively associated with both androeroticism and
gynoeroticism. Greater sexual interests in both women and
men characterize women who have had more previous

relationships.

Sexual Abuse

All of the women who participated in this study answered
the questionnaire item which asked if they had ever been the
victim of sexual abuse. Of the 77 women respondents, 22
indicated that they have been victims of sexual abuse. In
this sample, 30% or almost one third of women have been
sexually abused. Official statistics on incest, rape, and

sexual assaults often cite percentages of 15% to 25%, and

151

emphasize that these are probably gross underestimates.

The questionnaire item asked women how old they were at
the time of the incident and who the abuser was. A number of
the women who reported having been sexually abused wrote in
additional information about their victimization. Their
stories were often poignant and compelling.

From the information provided by respondents, eight
women reported having been sexually abused by more than one
person. Fourteen women had been victims of male family
members, and many of these women indicated that the
incestuous relationships lasted for many years. Three women
reported father-daughter incest, these women stated that the
incestuous relationships began when they were &, 7, and 12
years of age. Three women had been sexually abused by a
step-father, with the first incidents occuring when they were
s, 1% and 1% years old. Four women reported incestuous abuse
by a brother, and four women had been sexually victimized by
an uncle; all eight of these women were between four and ten
years of age at the time of the first incident of abuse.
Fourteen women were victims of sexual assault perpetrated by
non-family members. Eleven women had been sexually abused by
an acquaintance, three women had been victimized by a
stranger. The women’s ages at the time these incidents
occured range from four to 26.

Statistical analysis of these data showed that having
been the victim of sexual abuse, whether incestuous or
non-incestuous, was not related to any of the demographic

variables including family SES, age, education, or marital

158

status or relationship status. There was no relationship
between sexual orientation and having been sexually abused.
The relationship between sexual victimization and all of the
research variables investigated in this study showed only two
significant associations. Sexual abuse victimization was
significantly associated with maternal involvement for all
four time periods, and with identification with mother. The
association between sexual abuse and the Mother-Daughter
Differentiation Scale for early childhood yields a Chi Square
of 1%.3 and a p-.01&. women who had been sexually abused are
proportionately more apt to report more differentiated and
distant relationships with their mothers. Table 16 shows the
frequencies of women in each of the Mother-Daughter

involvement categories by their sexual abuse victim status.

Table 16 Frequencies and Percentages of women who had, and
had not been victims of sexual abuse by their early childhood
maternal involvement classifications.

 

Levels of Maternal Involvement

 

1 2 3 & 5 6
Uictim O & 7 6 3 2
Nonvictim 11 18 11 6 7 0

 

Examining womens ratings of their mother-daughter
relationships across the other age periods assessed reveals
the same significant linear pattern of association between
mother-daughter distance and sexual victimization. As the

table reveals, the relationship between victimization and

153

maternal involvement suggests that women who characterize
their relationships with their mothers as having been
moderately close (3 and &) or distant and highly
differentiated (5 and 6) are more apt to have been the

victims of sexual abuse. As the table demonstrates, as

mother-daugther closeness diminishes, the percentages of
women who have been sexually abused increases. The most
obvious explanation for this association would hold that
mothers who are less involved may supervise their daughters
less, thus as young girls these daughters are more apt to be
in unsupervised situations which put them at risk. Due to
the large number of women who cited incidents of sexual abuse
during childhood, this explanation may certainly be an
important contributor to this association. However, a
comparison of the mean scores on the maternal involvement
scale for women who had been abused before and after age 12
indicates that it does not explain the relationship.
Non-significant differences in the means suggest that women
who had been abused in childhood--when maternal supervision
may be an important factor--have slightly more involved and
closer relationships with their mothers than the women who
were victimized after age 12. (The means are 3.33 for the 10
women victimized before age 12, and 3.8 for the 12 women
victimized after age 12.)

Sexual abuse was also related to identification with
mother. women who were disidentified with mother were twice
as likely to have been victims of sexual abuse. Table 17

presents the frequencies of women in each of the

15&

identification with mother groups, by their victimization

status.

Table 17 Frequency of women who were victims and nonvictims
of sexual abuse classified by their levels of identification
with mother.

 

Identification with Mother

High Moderate Low
Uictims 5 5 10
Nonvictims 20 15 _ 15

 

These findings suggest that women who are highly
identified with their mothers, and who are less
differentiated from them are less likely to be victims of

sexual abuse.

APPENDIX 1.

155

APPENDIX L

Education and Sexual Orientation

The demographic variable--education, shows differing
patterns of relationships with the research variables for
heterosexual and lesbian women. Among heterosexual women,
education is significantly and positively correlated with
obJect differentiation (r-.50) and with gynoeroticism (r-.50)
indicating that higher education is associated with increases
in both gynoeroticism and levels of obJect differentiation.
In contrast, education is inversely associated with both
obJect differentiation (-.&3) and gynoeroticism (-.1S) for
lesbian women. The distributions of scores reveal that the
two groups have more similar scores at high levels of
education, while their scores on both measures are highly
discrepant at lower levels of education. when correlations
are computed for all women in the sample, education is
positively associated with both obJect relations

differentiation (r-.20), and gynoeroticism (r-.&3).

APPENDIX M

158

APPENDIX M

Semantic Differential Item Analysis
Exploratory analyses of the data provided by the
semantic differential are presented here.
The 5D items on which women rated their mothers which
correlated most highly with the maternal involvement scale
are consistent with the adJectives used in the scale. The

most highly correlated items are presented in Table 18.

Table 18 Correlations between Matinv1 and the semantic
differential items for Mother (women’s ratings of their
mother’s characteristics which are associated with maternal
involvement)

 

Item Correlation

Close--Distant .53
Accessible--Aloof .&7
Emotional--Unemotional .&3
Nurturing--Depriving .&2
Loyal--Disloyal .SO
Compassionate--Hard Hearted .36
Open--Closed .3&
Expressive--Restrained .32

 

Table 19 reveals the items on which women rated their

mothers which correlate most highly with their Identification

with mother scores.

157

Table 15 Correlations between SD items on which women rated
their mothers and their identification with mother scores.
(Characteristics of mother which are asssociated with
identification with mother.)

 

Item Correlation
Natura1--Pretentious .55
Open--Closed .53
Flexible~~Rigid .&8
Understanding--Intolerant .&&
Generous--Stingy .&&
Compassionate--Hard Hearted .&1
Nurturing--Depriving .&0
Close--Distant .&0
Accessible-~Aloof .38
warm--Cold .37

 

Table 20 presents the items on which women rated
themselves, which were most highly associated with their
identification with father scores. AdJectives on the left
were associated with high identification with father. As
table 21 reveals, perceived similarity between self and

father is related to communal characteristics in fathers.

Table 20 Item Correlations between identification with
father and womens ratings of their own characteristics

 

Item Correlation
Conforming--Individualistic .38
Playful--Serious .87
Delicate--Rugged .27
Submissive-~Dominant .25
Nonpossessive--Possessive .2&
Ugly-~Beautiful .23
Understanding--Intolerant .28

warm--Cold .22

 

158

Table 21 Highest correlations for SD item on which women
rated their fathers and their identification with father
scores

 

Item Correlation
Flexible--Rigid .55
Open--Closed .55
Understanding--Intolerant .&2
Close--Distant .&O
Peacable--Hostile .35
Accessible--Aloof .37
Nurturing--Depriving .35
Accepting--ReJecting .35

Generous--Stingy .32

 

158

References

Adelman, M. R. (1577). A comparison of professionally
employed lesbians and heterosexual women on the MMPI.
Archives of Sexual Behavior,_§, 153-201.

Almquist E. M., 8 Angrist, S. M. (1570). Career salience and
atypicality of occupational choice among college women.
Journal of Marriage and the Family, 32, 2&2-2&5.

 

Armon, U. (1560). Some personality variables in overt female
homosexuality. Journal of ProJective Technigues,
23, 852-305.

 

Bakan, D. (1566). The duality of existence: Isolation and
communion in western man. Boston: Beacon Press.

 

 

Baruch, G. K. (1572). Maternal influences upon college
women’s attitudes toward women and work. Developmental
Psychology, 6, 32-37.

 

Baruch, G. K. (1576). Girls who perceive themselves as
competent: Some antecedents and correlates. Psychology
of women Quarterly,_1, 38-&5.

 

 

Baumrind, D. (1575). New directions in socialization
research. Invited address prepared for the Society
for Research in Child Development, San Fransisco.

Bell, A. P., weinberg, M. 5., & Hammersmith, S. K. (1581).
§§xual Preference. Indiana: Indiana University Press.

Bem, S. L. (1577). The measurement of psychological
androgyny. Journal of Consulting_and Clinical Psychology,
12(2), 155-162.

Bene, E. (1565). On the genesis of female homosexuality.
British Journal of Psychiatry, 3, 815-821.

 

Berry, 0. K. (1585). A TAT research measure for assessing
adult correlates of early childhood separation-
individuationdynamics. ‘Unpublished masters theses,
Michigan State University. East Lansing.

 

 

Beverely, G. S. (1577). Some factors in the development of
homosexual obJect choice. Dissertation Abstracts
International, 32, (7-B), 3577.

189

Block, J. H. (1578). Another look at sex differentiation
in the socialization behaviors of mothers and fathers.
In J. Sherman & F. Denmark (Eds.), Psychology of women:
Future Directions of Research. New York: Psychological
Dimensions.

 

Block, J. H. (158%). Sex role identity and egg development.
San Fransisco: Jossey- Bass Publishers.

Chodorow, N. (157&). Family structure and feminine
personality. In M. Z. Rosaldo & L. Lamphere (Eds.),
woman, culture, and society. Stanford: University of
California Press.

Chodorow, N. (1578). The reproduction of Mothering_
Psychoanalysis and the sociology of gender. Berkeley:
University of California4 Press.

 

Clingman, J., & Fowler, M. G. (1576). Gender roles and human
sexuality. Journal of Personality Assessment, 39(3),
276-28&.

Crouch, A. M. (1578). A comparison of family relationships
and family constellations of adult female homosexuals
and adult female heterosexuals. Dissertation Abstracts
International, 38, (5-A) 586&.

Deutsch, H. (15&&). The gsychology of women (U01. 1).
New York: Grune and Stratton.

Dinnerstein, D. (1577). The mermaid and the minotaur: Sexual
arranggments and human malaise. New York: Harper
Colophon Books.

 

Donelson, E., & Gullahorn, J. E. (1577). women: A
psychological perspective. New York: wiley Press.

 

Douvan, E., 8 Adelson, J. (1566). The adolescent experience.
New York: wiley.

Freedman, M. (1568). Homosexuality among women and
psychological adJustment. Dissertation Abstracts,
International, 88, &25&B-&8558.

Freedman, M. (1571). Hgmoggxuglit and s cholo ical
functioning, Belmont, California: Brooks/Cole.

Freedman, M. (1575). Far from illness: Homosexuals may be
healthier than straights. Psychology Today, 8, 28-32.

 

Giannell, A. S. (1566). Giannell’ s criminosynthesis theory
applied to female homosexuality. Journal of Psychology,
65(2), 213- 222.

181

Goldberg, 5., 8 Lewis, M. (1565). Play behavior in the year-
old infant: Early sex differences. Child Development,
&0, 21-32.

Gove, w. R. (1576). The relationship between sex roles,
marital status, and mental illness. In H. 6. Kaplan
8 J. P. Sean (Eds.), Beyond sex role stereotypes.
Boston: Little Brown.

 

Gundlach, R. H., 8 Reiss, B. F. (1568). Self and sexual
identity in the female: A study of female homosexuals.

In B. F. Reiss (Ed.), New directions in mental health,
(U01. 1). New York: Grune and Stratton.

Hall, M., & Keith, R. A. (156&). Sex-role preference among
children of upper and lower social classs. Journal of
Social Psychology, 68, 101-110.

Hartley, R. E. (1560). Some implications of current changes
in sex-role patterns. Merrill Palmer Quarterly,
3, 153-16%.

 

Hassel J., a Smith, E. w. L. (1575). Female homosexuals’
concepts of self, men and women. Journal of Personality
Assessment, 35(2) 15&-155.

Heatherington, E. M. (1565). A developmental study of the
effects of sex of the dominant parent on sex-role
performance, identification and imitation in children.
Journal of Personalityand Social Psychology, 2, 188-15&.

Heatherington, E. M. (1567). The effects of familial
variables on sex-role typing, parent-child similarity,
and imitation in children. In J. P. Hill (Ed. ),
Minnesota Symposium on Child Development (1),
Minneapolis. University of Minnesota Press.

Heilbrun, A. B. (1568). Sex-role, instrumental-expressive
behavior, and psychopathology in females. Journal of
Abnormal Psychology, 73, 131- 136.

 

Hoffman, M. L. (1578). Toward a theory of empathic arousal
and development. In M. Lewis & L. A. Rosenblum (Eds.),
The development of affect. New York: Plenum Press.

 

Hogan, R. A., Fox, A. N., & Kirchner, A. H. (1577).
Attitudes, opinions, and sexual development of 205
homosexual women. Journal of Homosexuality, 352) 183-136.

 

Hopkins, J. H. (1565). The lesbian personality. British
Journal of Psychiatry, 115, 1&33-1&36.

 

Horner, A. J. (1575). ObJect relations and the developigg
Ego in therapy. New York: Jason Aronson.

 

 

188

Hyde, J. 5., & Rosenberg, B. 6., & Behrman, J. (1577).
Tomboyism. Psychology of women Quarterly,_§, 73-75.

 

Hyde, J. 5., & Rosenberg, B. G. (1580). Half the human
experience. Lexington Mass: D. C. Health.

 

 

Jacobson, E. (156&). The self and the obJect world.
New York: International Universities Press.

 

Kagan, J., & Moss, H. A. (1562). Birth to Maturity.
New York: wiley.

Kaye, H., Berl, 5., Carl, J., Eleston, M., Gershwin, B.,
Gershwin, P., Kogan, L., Torda, C. & wilbur, B. (1567).
Homosexuality in women. Archives of General Psychiatqy,
17, 626-63&.

 

Kenyon, F. E. (1568a). Studies in female homosexuality:
Psychological test results. Journal of Consulting and
Clinical Psychology, §§, 510-513.

 

Kenyon, F. E. (1568b). Studies in female homosexuality IU:
Social and psychiatric aspects. British Journgi
of Psychiatry, 11&, 1337-13&3.

Kenyon, F. E. (1568c). Sexual development, attitudes
and experience. British Journal of Psyghiatgy,
11&, 13%3-1350.

Klein, M. (1532) The psychoanalygis of children.
London: Hogarth Press.

 

Klein, M. (1557). Envy and gratitude. New York:
Dellacorte Press.

 

Laner, M. R. (1577). Permanent Partner priorities:
Gay and straight. Journal of Homosexuality, 3(1),
21-35. 4- .-

Lewis, M. (1572). State as an infant environment
interaction: An analysis of mother-infant interaction

as a function of sex. Merrill-Palmer Quarterly,
;§, 55-122.

Loney, J. (1573). Family dynamics in homosexual women.
Archives of Sexual Behavior,.§, 3&3-350.

Maccoby, E. E., & Jacklin, C. N. (157&). The psygholo of
sex differences. Stanford, CA: Stanford University ress.

 

Mahler, M. 5., Pine, F., & Bergman, A. (1575). The
psychological birth of the human infant: Symbiosis
and individuation. London: Hutchinson.

 

183

McCauley, E. A., & Ehrhardt, A. A. (1577). Role
expectations and definitions: A comparison of female
transexuals and lesbians. Journal of homosexuality,
_§(8), 137-1&7. -

 

Money, J., & Ehrhardt, A. (1572). Man and woman/Boy and
girl. Baltimore: The Johns Hopkins University Press.

 

Moss, E. A. (1567). Sex, age, and state as determinants of
mother-infant interaction. Merrill Palmer Quarterly
of Behavior and Development 33, 15-36.

Mussen, P. H. & Rutherford, E. (1563). Parent-child
relations and parental personality in relation to young
children’s sex-role preferences. Child Develogment,

33, 585-607.

Nadelman, L. (157&). Sex identity in American children:
Memory, knowledge and preference tests. Developmental

Psychology, 19, &13-&17.

 

Neutgarten, B. L., & Gutmann, D. L. (1558). Age-sex roles
and personality in middle age: A thematic apperception
study. Psychological Monographs, 18, (17, No. &70).

 

Ohlson E. L., & wilson, M. (157%). Differentiating female
homosexuals from female heterosexuals by use of the MMPI.
The Journal of Sex Researgfl. g9, 308-315.

Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1557). he

measurement of meaning. Urbana, IL: University of
Illinois Press.

Parsons, T. (1551). The social system. New York:
The Free Press.

 

Plato The works of Plato. New York: Tudor Publishing
Company.

Poole, K. A. (1578). The etiology of gender identity and
the lesbian. Journal of Social Psychology, 82, 51-57.

Rosen, D. H. (157&). Lesbianism: A studyyof female
homosexuality. Springfield, Ill: Charles C. Thomas Press.

 

Saghir, M., & Robins, E. (1573). Male and female
homosexuality: A comprehensive investigation. Baltimore:
williams and wilkins.

Saghir, M., Robins, E., walbran, 8., 8 Gentry, K. (1570).
Homosexuality IU: Psychiatric disorders and disability
in the female homosexual. American Journal of
Psychiatry, £22, 1&7-15&.

18&

Shavelson, E., Biaggio, M. K., Cross, H. H., & Lehman, R. E.
(1580). Lesbian womens’ perceptions of their parent-child
relationships. Journal of Homosexuality,_§ (3), 805-215.

Shively, M. G., 8 De Cecco, J. P. (1577). Components of
sexual identity. Journal of Homosexuality, 3(1), &1-&8.

Siegelman, M. (1572). AdJustment of homosexual and
heterosexual women. British Journal of Psychiatry, 120,
977-581. '7 7’

Siegelman, M. (157%). Parental background of homosexual
and heterosexual women. British Journal of ngchiatgy.
18&, 1&-81.

Spence, J. T., & Helmreich, R. L. (1578). Masculinityvand
femininigy: Their psycholggical dimensions, correlgtes

gnd antecedents. Austin, Texas:FThe University of Austin
Press.

 

\

Storms, M. D. (1580). Theories of sexual orientation.
Journal of Personaligy and Social Psychology, 33 (5),
783-752.

Swanson, D., Loomin, 5., Lukesh, R., Cronin, R., 8 Smith, J.
(1572). Clinical features of the female homosexual
patient: A comparison with the heterosexual patient.
Journal of Nervous and Mental Disease, £33, 115-12&.

Thompson, N., McCandless, R., & Strickland, B. (1571).
Personal adJustment of male and female homosexuals and
heterosexuals. Journal of Abnormal Psychology, 23,
237-2&O.

 

Unger, R. K. (1575). Female and male; Psychological
perspectives. New York: Harper & Row.

 

 

Uance, B. K. (1577). Female homosexuality: A social
psychological examination of attitudinal and etiological
characteristics of different groups. (Doctoral
dissertation, Oklahoma State University, 1577).
Dissertation Abstracts Internationai, 33, &518.

Uan Cleave, C. (1577). Self identification, self
identification discrepancy and environmental perspectives
of women with same-sex sexual preference. (Doctoral
dissertation, Ball State University, 1577). Dissertation

Abstracts International, 33, 5532A.

 

whiting, D. B., 8 Edwards, C. P. (1573). A cross-cultural
analysis of sex differences in the behavior of children
age 3 through 11. Journal of Social Psychology,
33, 101-110. - ‘—

 

185

wilson, M., 8 Greene, R. (1571). Personality characteristics
of female homosexuals. Egycholggical Reports, 33, &07-&12.

 

wolff, C. (1571). Love between women. New York: Harper 8 Row.

 

Yarrow, M. R., 8 waxler, C. 2. (1577). The emergence and
functions of prosocial behaviors in young children. In
R. C. Smart 8 M. 5. Smart (Eds.), Readings in Child
Development and Relationships. New‘York: MacMillan.

 

"IIIIIIIIIIII“

31