u 13:: mi [Uzmflflfllll in 11mm": 1! mm mm

85

"7 " “ill—u w Hw— - 'h'w‘..n_efi:g‘1
rd '3.
' LIBRA n Y ;
. Michigan State g
(#33 Um ch‘y .-5

wvv'w- W

1 [pg-r

This is to certify that the

thesis entitled

Perceptions of Sex-Linked Differences
in the Speech of Mothers and Fathers
Toward their Newborn Infants

presented by
Roberta Tekotte Hale

has been accepted towards fulfillment
of the requirements for

Ph.D. figmem Counselig Psychology

M» 4%;

Major professor

Date January 19, 1981

0-7639

l
i

l
1

w:

l " ’ ”,3 25¢ per day 98? item
‘3 Ill/“d“ . RETURNING LIBRARY MATERIALS:
55V?” Place in book return to remove
w“ A

charge from circulation records

 

 

‘-\ ”I
.37 Jwv

04 08 02

 

 

PERCEPTIONS OF SEX-LINKED DIFFERENCES
IN THE SPEECH OF MOTHERS AND FATHERS

TOWARD THEIR NEWBORN INFANTS

By

Roberta Tekotte Hale

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling and
Educational Psychology

1981

ABSTRACT

PERCEPTIONS 0F SEX-LINKED DIFFERENCES
IN THE SPEECH OF MOTHERS AND FATHERS
TOWARD THEIR NEWBORN INFANTS
By

Roberta Tekotte Hale

The purpose of this study was to determine whether
subjects perceived differences in the way mothers and
fathers talked to their infants and whether they could
identify the sex of the infant from the parent's speech.
Subjects were asked to identify the parent who was speak-
ing and the sex of the infant from reading excerpts of
parents' verbal interactions with their infants where
direct references about the true identity of the parents
and infants had been eliminated.

The primary concern in this study was whether women
and men, in their role as parents, use the English language
differently and whether the sex of the infant influenced
their speech in an identifiable way. A review of the lit-
erature suggested that differences exist in the way women
and men talk, but that differences were dependent upon sev-
eral factors, namely, age, socioeconomic background, occu-

pation, and the context or role one is assuming at a given

Roberta Tekotte Hale

time. There was also some evidence from the literature
that the sex of the infant influenced the behavior and
speech of parents. Experimental evidence of speech differ-
ences for women and men in their role as parents was non-
existent. There were also no experimental studies which
tested whether the sex of the infant could be identified
from what parents said to their infants. The potential
use of empirical evidence that speech differences exist
for women and men for subsequent research efforts into the
part that language plays in learning sex-role appropriate
behavior was the impetus for the study.

Fifteen married, Caucasian couples from the Lansing,
East Lansing, Michigan, area were administered a 40-item
instrument which asked them to identify the parent who was
speaking and the sex of the infant for each item or segment
of verbal interaction. The instrument had been developed
from tape recordings of verbal interactions between mothers,
fathers, and their infants. Subjects were also asked to
indicate their reasons for deciding which parent was speak-
ing and the sex of the infant. They were given a choice of
four categories which they could use as a basis for their
judgments. In addition, each subject was administered the
Bem Sex-Role Inventory which classified them as Masculine,
Feminine, or Androgynous.

A t-test for population means was used to test

the hypothesis that subjects could identify the parent who

Roberta Tekotte Hale

was speaking with greater than chance accuracy. Statis-
tical significance was reached for this hypothesis (p <
.0001). However, subjects were not able to identify the
sex of the infant with greater than chance accuracy

(p < .11).

Correlational analyses between the accuracy scores
of identifying the parent and the sex of the infant and
the Bem Masculinity and Femininity scores were performed.
The only statistically significant finding was that there
was a tendency for subjects who had high accuracy scores
for identifying the father to have a high femininity score
on the Bem. There was no ready explanation for this find—
ing.

Examination of the data also showed that more than
half of the time, subjects formed their judgments about
the parent who was speaking from the language that was
used in the item. When it came to how subjects formed
their judgments about the sex of the infant, however, there
were a greater variety of explanations.

Before any conclusions can be reached about
whether women and men, in their role as parents, talk dif—
ferently to their infants and whether the sex of the infant
impacts on parental speech, larger and more powerful

studies are needed.

ACKNOWLEDGMENTS

My appreciation and thanks to the members of my
dissertation committee. I thank Bill Hinds, the chairper-
son of my committee, for the helpful and gentle way he gave
feedback and suggestions. I thank Bill Farquhar, who has
been supportive throughout my years as a graduate student.
I thank John Schneider, who has played an important role
in my professional and personal development. It was
through my contacts with John as an undergraduate that I
was encouraged and inspired to go on for graduate work. I
am grateful to Verda Scheifley for her willingness to not
only join my committee under unusual circumstances and at
a late date, but who taught me statistics in a fun and
understandable way. Her constructive feedback, availabil—
ity, and hard work were unfailing and a continuous source
of support to me. It was also a joy to me that we could
share the experience of being pregnant at the same time.

I thank Tim Boyd for his patience and willingness
to spend many evenings and weekends in the Computer Center.

I am grateful to the couples who tape recorded
their conversations with their infants and the couples who
agreed to be subjects for this study. They shared will-
ingly and enthusiastically of their time and interest.

ii

I thank my friends, who were understanding and
supportive, especially Dianne Singleton, who pitched in and
helped at crucial points along the way.

I thank Norbert Enzer, whose interest and pre-
vious research in the way mothers and fathers talked to
their infants served as a foundation for this project.

A very special thanks to Cliff, my husband, who
shared the triumphs and the setbacks of this study with me.
He listened when I needed to be heard and encouraged me
when I needed a lift. His sensitivity and perceptiveness

were a great source of support to me.

iii

TABLE OF CONTENTS

LIST OF TABLES

Chapter

I.

II.

III.

IV.

INTRODUCTION

Theory

Need

Purpose .
Research Hypotheses
Overview .

REVIEW OF LITERATURE

Female/Male Speech Differences

Mother- Infant, Father-Infant Verbal.

Interaction

Impact of Sex of Infant on Parental.

Verbal Behavior
Summary

RESEARCH DESIGN AND METHODOLOGY

Sample .
Instrumentation
Reliability . .
Research Procedures
Research Design
Research Hypotheses

Procedures for the Analysis of Data

Summary of Procedures

ANALYSIS OF THE DATA

Descriptive Statistics . .
Results of Hypothesis Testing
Further Analyses of Interest
Summary

iv

Page

vi

t-J

OLDQDNIN

11
24

27
35

37

37
38

51
53
54
55
57

60

60
62
64
67

Page

V. SUMMARY, CONCLUSIONS, AND DISCUSSION . . . . 71

Summary . . . . . . . . . . . . . . . . . 71

Results . . . . . . . . . . . . . . . . . 75

Discussion . . . . . . . . . . . . . . . . 77

Limitations . . . . . . . . . . . . . . . 81

Recommendations for Further Research . . . 83

APPENDICES . . . . . . . . . . . . . . . . . . . . . 87
A. SUBJECT CONSENT FORMS FOR RESEARCH

PARTICIPATION . . . . . . . . . . . . . . 88

B. TELEPHONE SOLICITATION 0F SUBJECTS . . . . . 92

C. INSTRUMENTATION AND SCORING KEY . . . . . . 94

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . 108

LIST OF TABLES

Item Performance Analysis

Percentage of Consistent Responses Across
Two Administrations of Sixteen Items

Cell Means and Standard Deviations of
Accuracy Scores for Identifying the Sex
of the Parent and the Infant

Pearson Correlation Coefficients

Categories Selected as Basis for Judgments
About the Sex of the Parent and the Infant

vi

Page

45

49

61

66

68

CHAPTER I

INTRODUCTION

Although interest in linguistic sex differences
dates at least as far back as 1664 when women's and men's
forms of speech were cited in a published report about the
Carib Indians (Jespersen, 1922), it was not until the
women's movement in the sixties that the topic of differ-
ences in the way men and women use the English language as
well as other discrepancies between the sexes received much
attention. Since then there has been an emerging body of
literature on male/female differences in the use of the
English language. Linguists, lexicographers, psychologists,
sociologists, speech physiologists, anthropologists, novel—
ists, and speech communication experts have all contributed
to the analysis, findings, and commentaries on the sexual
differentiation of the English language.

Sex differences have been noted in phonology
(e.g., pronunciation, intonation patterns, pitch); in
semantics (e.g., sex as a dimension in pronominal referents
and in the elaboration and content of lexicons); in morph-
ology and grammar (e.g., sex differences in choice of

adverbs and in syntactic usage). Sex differences have also

been reported in the structuring of speech events (e.g.,
patterns of interruption and silences, development and
choice of conversational topics, and the allocation of
speaking terms).

However, the variability in the results of such
studies as well as recent evidence suggests that factors
such as age, social class, geographical location, ethnicity,
occupation, as well as sex must be considered in the final
analysis. In addition, authors Thorne and Henley (1975)
and Key (1972) point out the necessity of going beyond
linguistics to examining the context of the situation where
differences in the use of the English language have been
observed. "Speech," according to Thorne and Henley, ”is
intimately bound up with the social differentiation of the
sexes” (p. 10). The authors conclude that the effect of
different social settings and roles is crucial to under-
standing the linguistic behavior of men and women.

In the present study, perceived differences in the
speech of men and women within their roles as parents of a

newborn infant were explored.

Theory

Until recently, theories which existed about the
differential use of the English language by men and women
suggested biological and anatomical determinants. Jesper-

sen (1922), who was the first writer to include a separate

chapter on women's speech, claimed that women "instinc-
tively” shrink from ”coarse and gross expressions" (p. 246).
The anatomical hypothesis for male/female speech differences
was that the voice-producing mechanism of men and women was
different so that the sounds which were made were differ-
ent, allowing a relatively easy distinction. This theory
was challenged, however, by an experimental study done by
Sachs, Lieberman, and Ericksen (1973). Basing their study
on anatomical findings that prepubertal boys and girls of
the same height and weight have the same larynx and thus
the same fundamental frequencies, regardless of sex, these
investigators asked 83 judges to listen to tape recordings
and determine whether it was a girl's or a boy's voice they
were hearing. Eighty-one percent of the guesses made by
the judges were correct. Since there was no difference in
articulatory mechanism size, the differences observed were
thought to be the results of children learning culturally
determined ways of speaking, i.e., learning a particular
pitch and frequency, etc., that are viewed appropriate for
each sex.

Although some writers imply that sex differentia-
tion of speech has biological roots, the clear weight of
evidence suggests social and cultural components. The
phenomenon of taboo, for instance, has been offered as an
explanation for sexual differentiation of vocabulary. This

theory was advanced by Jesperson, particularly in reference

to the different forms of men's and women's speech found

in the Carib Indians. He believed that if taboos became
associated with particular activities or objects where,

say women, were forbidden to use the original name, new
words or phrases are likely to be used instead by the women,
thus resulting in a sexual differentiation of vocabulary.
This theory has been used to explain the finding that it is
more acceptable for men in our society to use swear words
than for women.

Peter Trudgill (1974), a sociolinguist, hypothe-
sized that the differences in the way men and women use the
English language are the result of social differences.

That is, different social attributes and different behav-
ior are expected from men and women. "Using a female
linguistic variety,” according to Trudgill, "is as much a
case of identifying oneself as a female, and of behaving
as a woman should, as is wearing a skirt" (p. 95).

The most comprehensive theory regarding the sexual
differentiation of language comes from Thorne and Henley
(1975). The authors cite three major themes in accounting
for the differences: (a) the social elaboration of gender;
(b) the structure of male dominance; and (c) the division
of labor by sex (the interests, activities, and positions
of women and men in society, including the socialization

of children, and forms of social bonding).

The social elaboration of gender is explained by
their belief that humans, compared with other species, are
only slightly sexually dimorphic. That is, they organize
such learned characteristics as posture, gesture, facial
expression, and speech to enhance sexual dimorphism.
Sexual differentiation of speech is conceptualized as part
of human gender display, as one way in which sex differ-
ences are socially marked, emphasized, and enacted.

With regard to language and male dominance, the
authors note that male dominance is strikingly apparent in
tfluacontentof words and that in the English culture, men
are more highly regarded than women. The authors cite
nonverbal gestures which function as gestures of dominance
and submission as well as asymmetry occurring at the verbal
level. The authors also note that there are constraints
against women and men using the speech style associated
with the opposite sex but that men who use female forms
seem to be more stigmatized than women who use male forms
of speech.

As far as the division of labor by sex is con-
cerned, Thorne and Henley (1975) believe that a strong
division of labor by sex still exists in our society,
with women largely responsible for housework and child
care and men defined primarily by their economic or occupa—
tional position. This division of labor, according to the

authors, is reflected in our language. For example,

conversational topics among men and women have been found
to parallel the division of labor. Women talk to other
women about family and interpersonal matters, while men
talk to male friends about cars, sports, work, motorcycles,
carpentry, and local politics. The authors suggest, how-
ever, that these patterns of conversation may be changing
with the growing employment of women outside the home.

The present study is based on the theoretical
assumption that sexual differentiation of language is
effected by the relationship of the participants, the
situation where the interaction takes place, and the par-
ticular role one is assuming at the moment. The importance
of noting the context of the situation in which speech
occurs as well as the effect of different social settings
and roles has been recognized by authors Key (1972),

Thorne and Henley (1975), and Trudgill (1974). Key (1972)
believes that what might be a male/female distinction among
one substratum of people might be more accurately described
along other lines or along with other lines of the many
dimensional varieties of language.

Trudgill (1974) noted that the same speaker may
use different linguistic varieties in different situations
and for different reasons. He theorized that the role
relationships and relative statuses of the speakers involved
in a discourse resulted in different linguistic varieties.

For example, speech between individuals of unequal rank

(due to social class, age, or some other factor) is likely
to be less relaxed and more formal than that between equals.

Thorne and Henley (1975) also concur that social
context must be considered, especially as it relates to
whether words, phrases, and sentence patterns are valued
or less valued, strong or weak. It is their belief that
given speech forms are not inherently strong or weak but
acquire those attributes only in a particular cultural or
social setting.

Overall, the theories reviewed suggest a social/
cultural basis for the sexual differentiation of language.
Additionally, the theories suggest that the context of the
situation effects the linguistic behavior of men and women
and that the impact of different settings and different
roles must be considered in understanding and interpreting

the linguistic behavior of men and women.

E929.
While there is growing evidence that differences
exist in the linguistic behavior of males and females, it
has also become clear that many factors must be taken into
account before conclusions can be drawn. An important
consideration is that of social context or the particular
situation and role in which discourse takes place. No
studies to date have investigated whether differences are
perceived in the speech of men and women in their roles as

parents.

Furthermore, although an increasing number of
studies have suggested that mothers and fathers treat their
infants differently on the basis of the sex of the infant,
no studies to date have investigated whether the sex of
the infant is identifiable from what mothers and fathers
say to their infants.

If it can be established that differences are per—
ceived in the speech of mothers and fathers toward their
infants and that the sex of the infant can be identified
by what is said to the infants, then researchers can begin
to explore the acquisition of speech styles associated
with each sex and the relationship between sex-typed lan—
guage and the learning of sex—related roles and behavior.
Although language is often taken for granted, it is dif-
ficult to imagine that language would not play a central
role in the socialization of children and their percep-
tions of what it means to be a male or a female in our
society.

The implication for learning how children learn
"appropriate" sex—role behavior is that we can then evalu-
ate whether our ideas of "appropriate” behavior have kept
pace with the demands of modern-day society. For example,
if 90% of the female population will be employed outside
of the home at least some of their adult lives, does it
make sense to perpetuate the belief that housewifery and

motherhood are the choice occupations for women? Also,

if men are going to be called upon to share increasingly

in child care and household responsibilities, are we pre-
paring them for those tasks by perpetuating the belief

that men are tough, strong, insensitive, and occupation
oriented? Sensitizing ourselves to all the complexities

of the process by which we develop sex roles can help us in
adequately preparing our children for the world in which

they will ultimately have to cope and survive.

Purpose

One purpose of this study was to determine whether
mothers and fathers can be identified accurately on the
basis of their speech to their infants. Additionally, it
was the purpose of this study to determine whether the sex
of the infant could be accurately identified from what
parents said to their infant. Specifically, this study
focused on whether naive subjects could accurately perceive
the parent who was speaking and the sex of the infant from
reading excerpts of transcripts where direct references to

sex had been eliminated.

Research Hypotheses

 

The following five research hypotheses were tested
in this study. These hypotheses are based on the results
of a preliminary test and from theories taken from the

literature.

10

H1: Female subjects will accurately identify
which parent was speaking a greater number
of times than male subjects.

H2: Male subjects will accurately identify
the sex of the infant a greater number of
times than female subjects.

H3: Subjects will accurately identify when the
mother is speaking a greater number of times
than they will identify when the father is
speaking.

H4: Subjects will identify the parent who is
speaking with greater than chance accuracy.

H5: Subjects will identify the sex of the infant
with greater than chance accuracy.
Overview

In Chapter II, the literature pertinent to male/
female speech differences and the impact of the sex of the
infant on parental speech will be reviewed. Chapter III
describes the subject sample, the instrumentation, the
research procedures, the hypotheses, the analyses, and
the experimental design. In Chapter IV, the analysis of
data for each hypothesis is presented along with the
results of supplementary analyses used for exploratory pur-
poses in this study. A summary of this investigation and
a discussion of the findings and limitations as well as
the implications for further research will be included in

Chapter V.

CHAPTER II

REVIEW OF LITERATURE

A growing body of evidence exists which suggests
that differences exist in the way men and women use the
English language. However, evidence is also emerging which
indicates that the specific social setting and role being
assumed highlight or blur male/female differences. In
this study sex differences in speech as they relate to
women and men in their roles as parents of newborns were
examined. The literature and theories on female/male
speech differences will be reviewed first. Next literature
pertaining to mother-infant and father-infant interaction
will be presented. Finally, the literature on parental
verbal behavior, especially as it relates to the sex of

the child, will be presented.

Female/Male Speech Differences
Historically, sex differences in language were
first reported in non—Western societies. Little was
written about sex differences in English until the Women's
movement in the late '603 and early '705. Since then,

there has been a rapid spread of interest in this area.

11

12

Students investigating female/male speech differ—
ences come from a variety of backgrounds which, in turn,
has led to diverse frameworks for research. A 1944 study
by Mary Haas exemplified the old ethnographic studies using
the classic linguistic method of informants. Other
researchers such as Labov (1972) and Swacker (1975)
obtained samples from their English-speaking informants
by the use of "elicitation.” Using oneself as an informant
in one's own language was used by Lakoff (1973). Jespersen
(1922) and Lakoff (1973) made use of anecdotal material as
well. "Observation” by tape recorders or by hand tran-
scription has been used frequently by such researchers as
Cherry (1975) and Gleason (1973). Other writers have done
structural analyses of the language such as Key (1972),
who wrote about the contexts for female-marked words.
Kramer (1974) has done a "content analysis" of dirty jokes
and cartoons. Some investigators have used the experiment
(Wood, 1966). "Tests" have been used by a number of writ-
ers such as Warshay (1972).

Sociolinguists Thorne and Henley (1975) have pro-
vided the most up-to-date and comprehensive information
on sex differences in language in their book, Language and
Sex: Difference and Dominance. An overview of the develop-
ment of the field as well as an extensive annotated bibli-
ography which covers both verbal and nonverbal sex

differences in language is presented. In addition, the

13

authors have included a selection of papers representing
examples of current sociolinguistic interests.

Author Ann Bodine (1975) has written a review of
cross-cultural literature on sex—based differentiation of
language, drawing heavily from ethnographic studies of non-
European societies, particularly Amerindian, Bengali,
Carib, Cham, Chukchee, Hebrew/Semitic, Indo-European,
Japanese, and Thai. She reviewed the attitudes of the
speakers and the linguists toward linguistic sex differen-
tiation. Bodine developed a classification system for how
sex-based language differences have been reported. The
axes of the classification scheme were: (a) type of lan-
guage differences and (b) the basis of language difference
(sex of speaker, receiver of the message, the speaker plus
the person being spoken to or spoken about). It was
Bodine's belief that the backgrounds of the ethnographers
reporting linguistic sex differences caused them to over-
state and exaggerate types of linguistic sex differentiation
which do not occur in their own languages and ignore or set
apart types which do occur in their own languages. She
strongly recommended, therefore, that a classification sys-
tem such as hers be used and that native speakers investi-
gate their own language rather than a foreign language for
sex differences.

Writer Marjorie Swacker (1975) pointed out the

importance of the sociolinguistic variable, speaker sex.

14

In reviewing the literature, Swacker noted the meagerness
of studies which consider the sex of the informants. Her
own research was designed to help fill the void. She used
34 informants: 17 men and 17 women who were between the
ages of 20 and 28 and were full-time students at California
State University. Each informant was asked to view three
pictures from the year 1500 and then describe what they
saw. They were given as much time as they needed for their
descriptions and were told to be as thorough as possible.
Their descriptions were tape recorded. The total time was
calculated for the descriptions of all three pictures.
Three areas of significant differences between male and
female speaking patterns were found. One area was "ver—
bosity." The female mean time for all three descriptions
was about 3.17 minutes, and for males the mean time was
about 13.0 minutes. The second area of difference in speak-
ing patterns was numerals. Men tended to use considerably
more numerals in their descriptive passages than did women.
Women were found to precede half of their numerals with
indicators of approximation. The third area of nonandroge-
nous verbal behavior was that of the "topic shift marker”
or pauses, interjections, conjunctions, or a combination of
two or more of these which were used to shift topics.

Women used significantly more conjunctions than did men.

Men used interjections to mark topic shifts.

15

Author Ruth Brend (1975) found definite prefer-
ences in the general usage and avoidance of some speech
patterns by men and women. For example, men rarely, if
ever, use the highest level of pitch that women use, and
men avoid final patterns which do not end at the lowest
level of pitch.

With regard to interruptions and silences in con-
versations, Zimmerman and West (1975) found that the power
and dominance enjoyed by men in other contexts are used
in their conversational interaction with women. The authors
examined transcripts of brief two-party conversations,
covertly recorded in public places. The pairs consisted
of 11 male/female, 10 female/female, and 10 male/male,
Caucasian, middle—class, from 20-35 years old. The rela—
tionships between the pairs ranged from close friendships
to casual acquaintanceships. The transcripts were examined
for overlaps (where the current speaker has reached a point
in his/her utterance which could be construed as a complete
sentence and a nonspeaker begins to talk while the current
speaker continues and the speaker might protest); and
interruptions (a nonspeaker begins to talk at a point which
could not be construed as the end of a sentence). There
were 7 interruptions and 22 overlaps, both symmetrically
distributed between speakers in the single-sex conversa-
tions. There were 48 interruptions and 9 overlaps, with

male speakers making 98% of the interruptions and 100% of

16

the overlaps in the male-female conversations. The women
who were interrupted did not protest in any of the cases.
The distribution of silence was nearly equal in same—sex
conversations, while in cross—sex conversations, women
showed more silence than males. The authors concluded that
male dominance pervades routine conversations since the
women's right to complete a speaking turn was routinely
encroached upon with no apparent consequences.

Sachs (1975) examined whether the sex of children
can be identified from their voices prior to puberty and
what some of the possible cues are for differentiating
female and male speakers. Citing three experimental
studies, Sachs concluded that prepubertal girls and boys
can be identified with greater than chance accuracy as to
sex from their voices. She also discovered that:

(a) Judges were able to guess the sex of the child from
hearing isolated vowels; (b) Judges could not accurately
determine sex from sentences played backwards, suggesting
that there is considerable information in normal sentences
that carries information about the sex of the speaker
beyond the phonetic aspects of the voices; (c) A factor
emerged that was correlated with the perceived masculinity
or femininity of the voice, along with two other factors,
Active-Passive and Fluent-Disfluent, when judges rated
spoken sentences on semantic differential scales. The

author concluded that an independent cue about the sex of

17

the speaker which does not involve how active the voice
sounds or how fluent it is may be Operating.

A study by Garcia-Zamor (1973) determined that
children were aware of male versus female language. The
author studied eight nursery school children, four males
and four females, all middle to upper middle class and
Sé-B years of age. The children were interviewed two
times. The first interview consisted of a questionnaire
designed to elicit overt attitudes toward sex roles. In
the second interview, children gave judgments about whether
a girl or boy doll prop uttered sentences containing terms
of endearment, hostility, etc. The results indicated that
there was more agreement among male subjects than among
female subjects about whether a sentence was said by a male
or a female. "Aggressive" and competitive expressions were
consistently associated with male dolls. The expression
"dum-dum" combined with the accidental breaking of something
was associated with female dolls. Bright colors describing
clothing were associated with males while light colors were
associated with females. Tag questions tended to be female
associated while swear words were seen as male terms. While
boys made more consistent judgments about the appropriate-
ness of language use on the basis of sex, girls had stronger
sex-biased notions, according to Garcia—Zamor. The author
concluded that boys learn a new language--male language—-

while girls continue in the language of the nursery.

18

In another study about children's perceptions of
differences between the way men and women use the English
language, Fillmer and Haswell (1977) examined the percep-
tions of 121 children from an inner-city school, all of
whom were from a low socioeconomic background. The results
of that study indicated that children perceived differences
both in language usage of men and women and between the
conception of occupational roles traditionally assigned to
each sex.

Writers Labov (1972) and Trudgill (1974) have
noted that women are more sensitive to prestige patterns
and often use the more advanced forms in their own casual
speech. Trudgill's explanation for why women use more
correct forms of English than men is that women in our
society are more status-conscious than men, and for that
reason they will be more sensitive to the social signifi—
cance of social-class-related linguistic variables. Men,
because of current concepts of masculinity in our society
such as "toughness," will use less "correct" speech than
that of women to identify themselves as masculine.

Among the earliest writers to include a separate
chapter on women's speech was Jespersen (1922). He dis-
cussed taboo words, female linguistic conservatism, adverbs,
word choice, as well as some general characteristics of
female speech. He believed that women's vocabulary was not

as extensive as men's and was concerned with immediate

19

surroundings and ornamental things whereas men's vocabu-
lary was more abstract and remote. Women, according to
Jespersen, used more intense words and more adverbs than
men. Jespersen also believed that women used incomplete
sentences while men did not and that women talked more
quickly than men did. Jespersen's claims, however, were
based on literature written by men.

In an experimental study, Gilley and Summers (1970)
found sex differences in the use of hostile verbs. Fifty
males and 50 females were asked to make up sentences from
a given pronoun and a given verb. The results indicated
that male subjects used hostile verbs at a greater average
frequency than did female subjects.

Gleser and Gottschalk (1959) tape recorded samples
of speech elicited from 90 subjects. They found that women
used significantly more words implying feeling, emotion,
or motivation (whether positive, negative, or neutral) and
made more references to self. Women also used more auxil—
iary words and negations. Men used significantly more
words implying time, space, quantity, and destructive
action. An interesting finding was that the more intelli-
gent women were more similar to men in less frequent use
of self-references and negation. Another important con—
sideration in interpreting the results of this study is
that it was done in the fifties prior to the women's move-

ment and the increasing numbers of women who were employed

20

outside the home. Both of these factors may influence the
content and topic of conversations.

Conversational interaction between men and women
was the focus of three other studies. Saskin and John
(1963) analyzed the talk of a couple during a 16-hour day.
The quantity of talk and the frequency of different message
types and functions were examined. They found that the
wife produced more affect—discharging messages while the
husband produced more directive and informational state-
ments. In private, the wife produced more expressive mes—
sages than when the couple was in the presence of others.
The husband claimed 29% of the talking time when he was
having golf lessons compared to 79% of the time when he
was talking to his wife. Hirschman (1973), in her study
of six dyadic conversations between men and women, found
that females used a much higher percentage of "fillers,”
e.g., uhm, you know, than males. No differences were found
in the proportion of "qualifiers," e.g., maybe, probably,
I think, I guess. Males used third—person references
while females more often used pronouns. The generaliza-
bility of these studies is brought into question because
of the small sample sizes and because no comparison data
are available for interactions in other contexts, i.e.,
analyzing the talk of men and women in the office or where

both men and women are employed outside the home.

21

The third study involving the analysis of con-
versational interaction between men and women was done by
Marion Wood (1966). In her experimental study she found
that men tended to use more words per utterance in a given
verbal task than women. She also found that the length of
the verbal output of males tended to increase under condi-
tions of ineffective communication and to level off under
conditions of successful communication. This was not true
for females.

With regard to the language style of men and women,
Warshay (1972) found that men tended to write with less
fluency than women. They referred to events in a verb
phrase, to be time oriented, to involve themselves more in
their references to the events, to locate the event in
their personal sphere of activity, and to refer less to
others. Females, on the other hand, referred to events in
a noun phrase, were less time oriented, tended to be less
involved in their event references, and referred more to
others.

Lakoff (1973) and Key (1972) have each written
comprehensive reviews of male/female distinctions in lan-
guage. Lakoff (1973) explored the concept of "woman's
language" in English. She provided examples of woman's
language with regard to lexicon (color terms, particles,
evaluative adjectives) and syntax (tag questions and

related aspects of intonation in answers to requests and

22

of requests and orders). The author also analyzed speech
about women, especially with regard to ladyzwoman, master:
mistress, widowzwidower, and Mr.:Mrs.:Miss. Lakoff hypothe-
sized that as children women are encouraged to be "little
ladies."
Little ladies don't scream as vociferously as little
boys, are chastised more severely for throwing tan-
trums or showing temper: ”high spirits" are expected
and therefore tolerated in little boys; docility and
resignation are the corresponding traits expected of
little girls. . . . Women are allowed to fuss and
complain, but only a man can bellow in rage (p. 51).
Lakoff described women's language as generally more polite,
that women make far more precise discriminations in naming
colors than men, that women tend to use "meaningless"
particles in their speech, and use less strong expressions.
In addition, Lakoff found a number of adjectives such as
adorable, charming, sweet, lovely, and divine which were
used almost entirely by women. Women are more apt to use
tag questions, according to Lakoff, more than men. A tag
question is midway between an outright statement and a
yes-no question, such as "The war in Vietnam is terrible
isn't it?" (p. 55). The author claimed that the use of dif-
ferent particles by men and women is a learned trait,
”merely mirroring nonlinguistic differences again, and again
pointing out an inequity that exists between the treatment
of men and women, and society's expectations of them and

the treatment of women” (p. 51). Lakoff concluded that

there is a discrepancy between the way men and women use

23

the English language and that the social discrepancy in
the positions of men and women is reflected in linguistic
disparities.

Mary Ritchie Key (1972) summarized the findings
of other researchers and made observations of her own
about phonological differences in women's and men's speech,
intonation patterns, male/female differences in syntax, the
semantic component, and pronominal and nominal referents.
In addition, Key commented on the importance of noting the
context of the situation in which men and women are using
language, i.e., the relationship of the participants, the
situation where the interaction takes place, and the whole
range of conditions involved, before determining whether
the difference is sex—linked. Key concluded that her
research showed there are differences in the linguistic
behavior of males and females as well as in references
made to males and females though these differences may not
occur when speakers are in roles other than the sex role.

Kramer (1974) in her discussion on male/female
speech differences noted the difference between what is
actually different and what is believed to be different
about men's and women's linguistic behavior. She reviewed
both previous research and folk linguistics taken from a
variety of sources including etiquette manuals, cartoons,
speech books, and novels. She demonstrated how the stereo—

type of a woman having particular characteristics of speech

24

emerged. Kramer noted that there was a conflict between
what women's speech is really like and what people think
it's like as well as what people think it should be like.
Kramer suggested that women learn to control their speech
to remain in the background. She concluded that if sex-
linked differences do exist, they are a way to maintain
the differentiation between the sex roles.

In another study, Kramer (1973) examined actual
and perceived differences in men's and women's speech. She
asked 17 white men and 17 white women from a college fresh-
man class to write paragraphs describing two photographs--
one of a building and one of people. In analyzing the
data, Kramer found no significant differences in adjective
or adverb usage by women or men as had been previously
theorized by Jespersen (1922) and others. However, when
she asked the subjects to write their perceptions of the
way men and women use language, they generally agreed that
there are differences. Kramer concluded that ”beliefs that
sex-related language differences exist may be as important
as the actual differences" (p. 85).

Mother-InfantiiFather-Infant
Verbal Interaction

 

 

Mother—infant verbal interaction has been the
focus of many investigators, most of whom have been inter-
ested in language acquisition and the language environment

of the child. Few researchers have examined father-infant

25

interaction in general, much less father—infant verbal
interaction.

Clarke-Stewart (1973) concluded that there was a
positive relationship between a child's language develop-
ment and the mother's verbal stimulation. In particular,
she found that mothers of girls were more highly verbal
than mothers of boys and that girls were later found to
be verbally more precocious than boys. However, research-
ers Cohen and Beckwith (1976) found contradictory results
from their study and concluded that maternal language
behavior to boys and girls was not different. Still other
results were found by Friedlander, Jacobs, Davis, and
Wetstone (1972), who found that although mothers were more
active than fathers in babies' language environment, fathers
played an important role but they were not specific as to
what the father's role was.

Maccoby and Jacklin (1974) in reviewing the lit—
erature available on mother-, father-infant verbal inter-
action concluded that recent evidence does not clearly
indicate that girls undergo more rapid verbal development
in the first few years of life. The studies they reviewed
showed inconsistent results as far as girls receiving more
verbal stimulation or reinforcement than boys from their
parents.

Trevarthan (1974) filmed five babies once a week

from birth to six months. His evidence showed that fathers

26

play a more boisterous kind of game with more jokes and
mimicry of prespeech grimaces and more poking of the body
than do mothers.

In one of the few studies on father-infant verbal
interaction, Rebelsky and Hanks (1971) showed that fathers
spend little time vocalizing to their infants (average
time was 37.7 seconds per day) and that the number of
interactions varies by time of day, age and sex of infant,
and the kind of activity occurring during the interaction.
No comparable data on the amount of time mothers and infants
interact were available, however, which was a weakness in the
design of the study.

Parke and O'Leary (1972) compared mother—infant and
father-infant interactions with their newborn infants.

They found that fathers were very active with the infants.
Fathers were twice as likely to hold the infant and equaled
the mother in the extent to which he looked at, touched,
and vocalized to the infant. These same investigators
reported in a later study (1975) that mothers vocalized
more when alone with their infants than they did when
fathers were present.

Directly related to the present study is an unpub-
lished study by Norbert Enzer (Enzer, Note 1). In personal
communication with Enzer, he reported that judges were able
to identify when the mother was speaking and when the

father was speaking with greater than chance accuracy.

27

Enzer asked mothers and fathers to tape record their ver-
bal interactions with their newborn infants in the hospital
within 24-48 hours after delivery. The recordings were
transcribed, eliminating references to the sex of the
speaker or the sex of the infant as well as any conversa-
tion which was not directly related to the infant. A team
of "expert" judges from the fields of psychiatry, psy-
chology, pediatrics, and nursing was then asked to read the
transcripts and make a guess as to when the father was
speaking, when the mother was speaking, and the sex of the
infant. Although statistical analyses were not done,

Enzer reported that the judges were able to accurately
guess the sex of the parent speaking and the sex of the
iinfant greater than 50% of the time. The judges, accord-
ing to Enzer, were unable to provide any reasons or clues
as to what they were responding to in the transcripts which
helped them make their judgments.

Impact of Sex of Infant on Parental
Verbal Behavior

 

Although the results are inconsistent, a growing
body of literature exists on the differential treatment of
sons and daughters by their parents. Maccoby and Jacklin
(1974), for example, who summarized existing studies on
parent—infant verbal interaction, concluded that the
results of such studies were highly variable but that the

majority of studies showed no difference in the amount and

28

kind of verbal interaction. However, these authors concede
that the evidence they reviewed may not be sufficiently
refined to reflect important interactional differences which
may be operating between the sex of the parent and the sex
of the infant.

Aiden Macfarlane (1977) tape recorded parents'
verbal interactions with their newborn infants immediately
after delivery. She then transcribed every word uttered.
Macfarlane found a preoccupation with gender and the attri-
bution of a number of sex stereotypes by both parents and
the medical staff. One mother upon learning she had just
delivered a baby girl said to her husband, "Oh, I'm sorry
darling," to which the doctor asked, "What are you sorry
about?" "He (the father) wanted a boy" (p. 61). And
later the mother said, "There's one——there-—oh look, she's
got hair. It's a girl--you're supposed to be all little"
(p. 64).

In a study by Rubin, Provenzano, and Luria (1974),
it was found that sex-typing and sex-role socialization
appears to begin at birth. Thirty primiparous parents,

15 with sons and 15 with daughters, were interviewed within
24 hours after delivery. The births were traditional in
that general anesthesia was used with deliveries and
fathers were not allowed in the delivery room. Fathers
were also not allowed to handle their babies during the

first 24 hours but would look at them through the display

29

windows in the hospital nursery. Each subject was asked,
"Describe your baby as you would to a close friend or
relative” (p. 515). The responses were tape recorded and
later coded. In addition, each subject was asked to com—
plete a short questionnaire which consisted of 18 pairs of
bipolar adjectives. Although there were no rating differ-
ences on the 18 scales as a function of sex of parent,
several significant effects emerged as a function of sex
of infant. Daughters were significantly more likely than
sons to be described as little, beautiful, pretty, and
cute, and as resembling their mothers. Fathers were more
extreme in their ratings and made more stereotypic rating
judgments of their newborns than did mothers.

Gould's story of a baby named "X" dramatically
portrayed people's reactions to not knowing the sex of an
infant from birth (Gould, 1972). The difficulties the
parents encountered as well as the confusion and anger
others experienced in not knowing the gender of the child
poignantly demonstrated the importance our society places
on gender.

An offshoot of this story was an experiment done
by Seavey and Katz (1975), who found that adults' sex-role
stereotypes influence their behavior toward boy and girl
infants. Adult subjects were asked to interact with a
3—month-old infant who they were told was a boy, a girl,

or were told nothing about the child's sex. Subjects were

30

randomly assigned to each of the three conditions. Adults
who were told the baby was a girl were more likely to
choose a doll for the baby to play with than those adults
who thought the baby was a boy. In the case where sub—
jects did not know the sex of the infant, each of them
spontaneously labeled the infant a boy or a girl. When
asked how they arrived at that judgment, the subjects
were more likely to have used the characteristics of the
baby. For instance, if the subjects thought the baby was
male, they referred to the baby's strength, and likewise
if they thought the baby was female, they cited the baby's
softness.
Jeanne Brooks-Gunn and Wendy Schempp Matthews
(1979) in reviewing the studies done on sex differences in
male and female infants concluded that
Whether parents perpetuate or amplify already exist-
ing sex differences or assume sex differences should
exist based on their own stereotypes, their inter-
active style with their infants has an impact on their
children's developing sex role. Because we unearthed
so few real differences between boy and girl infants,
we would be willing to assume that the development of
increasingly extensive sex differences depends a great
deal on the parent's differential treatment of boy and
girl babies (p. 79).
Howard Moss (1974) found that mothers and fathers
use stereotyped sex—role attitudes in relating to their
7-week-old infants. Parents were asked to work together

in administering a series of nine tests to their infants.

The tasks the parents were asked to administer included

31

trying to get the infant to visually follow objects, to
exhibit physical strength and coordination, to smile, and
to vocalize. The parents were instructed to continue with
the task until they were satisfied that their infant had
successfully completed the task or would be unable to suc-
ceed. There was no time limit. The authors found that
mothers and fathers spent significantly more time partici-
pating with female than male infants in trying to get their
infant to smile. Similarly, when the parents were asked
to get their child to vocalize, both mothers and fathers
exhibited greater participation time with the females. Of
particular relevance to the present study, the author tape
recorded the entire procedure and a frequency count was
made of the number of affectionate terms of address the
parents used in talking to their infants during the test
itself. The author found that both mothers and fathers
used significantly more affectionate terms when addressing
the girls than the boys. Moss concluded that differences
between the sexes are probably "amplified over time through
the stereotyped sex—role attitudes of parents” (p. 162).
Cherry and Lewis (1976) in their study of sex-
differentiated aspects of mother—child verbal interactions
found that mothers of girls talked more to their daughters
than mothers of boys, asked more questions, used repeti-

tion, and used longer utterances. There were no comparable

32

studies of sex-differentiated aspects of father—child
verbal interactions.

Rebelsky and Hanks (1971), as mentioned in the
previous section, tape recorded the verbal interactions of
fathers and their infants for a 24-hour period every 2 weeks
from the time the infant was 2 weeks of age to the time
he/she was 3 months old. They found that 7 out of the 10
fathers who were recorded vocalized less to their infants
during the latter half of the study and that this finding
was more marked among fathers of female infants.

Goldberg and Lewis (1969), in their study using
direct observation of mothers and children in a play situa-
tion, collected data on differential treatment of mothers
toward their sons and daughters. As part of their analy-
sis, the amount of verbal interactions were recorded. A
frequency count was made which showed that mothers spent
more time vocalizing to daughters than to sons. No com-
parable data on the verbal interaction of fathers and sons
and daughters were available .

Gleason (1973) found that the manners of address
were different for boy and girl children. Gleason did an
observational study of children's emerging control of dif—
ferent language styles. He based his findings on the
observations of natural conversations of families with
children ranging in age from infancy to 8 years old.

According to Gleason, boy babies might be addressed in a

33

"Hail-Baby-Well-Met style" especially by fathers while
being played with heartily. Girl babies, on the other
hand, were dealt with more gently, physically and verbally.

Rothbart and Maccoby (1966) in examining parents'
differential reactions to sons and daughters found that
mothers tended to be more permissive with boys while
fathers tended to be more permissive with girls with regard
to dependency and aggression. A child's voice that was
ambiguous with respect to sex was used as a stimulus. One
group of parents was told they were hearing a boy's voice,
another than they were hearing a girl's voice. The parents
were asked to imagine that he or she was at home reading
with his or her 4-year-old boy, Johnny (or girl, Susan),
playing with a puzzle in the next room. With the child is
a l-year-old baby. Parents were asked to write down what
they would say or do in response to each statement made by
the 4 year old.

Investigators Osofsky and O'Connell (1972) looked
at dependence-independence variables in mothers' and
fathers' interactions with their daughters. Both the
mothers and fathers were observed in structured interac-
tion sessions with their daughters. Mothers and fathers
were also observed separately in interaction with their
child. There were also separate interviews for the parents
about their child-rearing attitudes and behaviors. In gen-

eral, the investigators found that the father is more likely

34

to help the child physically or become detached during
dependence—producing situations while the mother is more
likely to encourage independence and explain, question,
and make comments to the child. An interesting addition
to this study would have been to include sons in the study
to determine whether a difference in behavior based on the
sex of the child would occur.

Parke and O'Leary (1975) reported the results of
two observational studies they conducted comparing mater-
nal and paternal behavior toward their newborn infant.
They explored the impact that the sex and ordinal position
of the infant had on parental interaction patterns. In
the first study, two sets of observations were made:

(a) mother-father-infant and (b) mother-infant alone. In
the second study, three sets of observations were made:
(a) mother—father-infant, (b) mother-infant alone, and

(c) father-infant alone. Both studies were conducted in
the hospital following delivery. In the first study
fathers were present during both labor and delivery, with
one exception; half of the couples had attended Lamaze
childbirth classes. Nineteen couples and their first—born
infants served as subjects. In the second study 51 white
and 31 black families participated. Unlike the first study,
where the families were from middle-class backgrounds, the
families in the second study were of lower socioeconomic

status. Also in contrast to the first study, none of the

35

fathers were present during delivery. The authors found
that fathers vocalized more to first-born boys than to
first-born girls, while fathers vocalized equally to later-
born infants irrespective of sex. They concluded that both
sex and ordinal position are important modifying variables

in early parent-infant interaction.

Summary

A review of the literature found that, in general,
there were differences in the way males and females use
the English language but that the differences were high-
lighted or blurred depending on the context of the situa-
tion and such factors as age, socioeconomic level, and
occupation. The most common difference in males' and
females' linguistic behavior and the one with the most
empirical support is that women are more "polite” than men.
They avoid strong expressions or swearing and will more
often choose the "correct" way of talking. Early training
and cultural influences have been the more accepted and
supported theories for observed differences in the way men
and women use the English language.

Few studies have looked at what mothers and fathers
say to their infants and the impact that the sex of the
infant has on the parent's speech. Moss (1974) reported
that parents related to their 7—week-old infants in stereo-

typed sex-role ways. A preoccupation with gender as well

36

as the attribution of a number of sex stereotypes by par-
ents and medical staff could be found in the transcripts

of parents following the delivery of their infant (Macfar-

lane, 1977).

CHAPTER III

RESEARCH DESIGN AND METHODOLOGY

This chapter covers the following areas: sample,
instrumentation, reliability, research procedures, design,
research hypotheses, and procedures for analysis of the

data.

Sample

Fifteen married, Caucasian, middle-class couples
who ranged in age from 21 to 35 volunteered to participate
as subjects in this study. Middle-class, for the purposes
of this study, was defined using Hollingshead's (1957) two-
factor index of social position, i.e., amount of education
of head of household and type of work. The couples were
contacted through birth announcements in the local news-
paper and by word-of—mouth. The couples were from the
Lansing, East Lansing area. All of the couples had at
least one child. The children ranged in age from 6 weeks
to 2 years. Each couple was paid $10 for their participa-
tion in the study. Feedback about the results of the study

was offered.

37

38

Instrumentation

 

The instrument used to assess responses about the
sex of the parents and the sex of the infants consisted of
40 items. The items were segments of actual conversations
between mothers, fathers, and their infants. Two separate
answer sheets accompanied each test instrument. Directions
included with the test instrument instructed each subject
to read each item and then fill in the appropriate space
on the designated answer sheet for the sex of the infant
and the parent who was speaking.

There were two stages in the development of this
instrument. During the first stage, two couples were
initially contacted to tape record 90 minutes of their
verbal interactions with their infants. One of the infants
was a female and the other infant was a male. The couples
were instructed to make their verbal interactions as normal
as possible without trying to equalize the amount of inter—
action each of them had with their infant. The tape
recordings were then transcribed in their entirety. Numeri-
cal codings were used to designate a change in speaker or
a lapse of time. Direct references about the sex of the
infant and which parent was speaking were eliminated
throughout each transcript. These references included
names; pronouns such as he, she, him, her; nouns such as

Mom, Dad, Mother, Father, girl, boy, lady, and man.

39

As a preliminary test to determine the suitability
of this instrument for further research, 10 employees of
the College of Human Medicine, Michigan State University,
were asked to participate as judges. The directions accom-
panying the transcripts asked them to read through each
transcript and to list five times, by code number, when
they thought the mother was speaking, five times when they
thought the father was speaking, and the sex of the infant
at the end of each transcript. The judgments about the
sex of the parents and the sex of the infants were done
independently by each of the participants.

The results from the preliminary testing of the
instrument showed that 63% of the judgments about the
parent who was speaking and 70% of the judgments about the
sex of the infant were accurate.

The instrument was modified and procedural changes
were made based on the findings of the preliminary test.
Thus, for the second stage in the development of the instru-
ment, four couples, instead of two, were asked to tape
record their verbal interactions with their infants. Two
of the infants were females and two of the infants were
males. Equal numbers of female and male infants were used
to avoid bias. The number of couples was increased to
broaden the sample of parent-infant verbal interactions.
The couples were recruited by word—of—mouth. Their instruc-

tions were similar to those given to the first two couples

40

who tape recorded their interactions. They were reminded
that they did not need to equalize the amount of conversa-
tion each of them had with their infant and to make the
interactions as normal as possible.

Again, each of the tape recordings was transcribed
in its entirety. Then, the recordings were divided into
60-second segments or 50 words, depending on whether a
change in speaker occurred prior to the 60—second elapse
of time. Five segments of the mother speaking and five
segments of the father speaking were randomly selected
from each of the four transcripts for a total of 40 seg-
ments or items. As was done previously, direct references
about the sex of the infant and the parent who was speak-
ing were eliminated throughout the segments. By randomiz—
ing the order of the 40 items, three forms of the
instrument were made. The directions were changed in that
determinations about the sex of the infant and which par-
ent was speaking had to be made for each of the 40 seg-
ments or items. Separate answer sheets for judgments
about the sex of the infant and the sex of the parent were
provided.

By modifying the instrument in the previously
described manner, it was believed that several variables
noted in the preliminary test results would be controlled.
One of the factors noted was the variability in the length

of the transcripts themselves. Most of the participants

41

who agreed to serve as subjects for the preliminary test
stated they thought the transcripts were too long and
tedious. The subjects also reported that the longer
transcript gave more information by its length alone.
Second, the subjects reported they became used to a "style
of dialog" between the parents, which aided them in select-
ing which parent was speaking and the sex of the infant.
Third, the length of the verbalizations by each parent was
used in making judgments rather than any other cues.
Finally, by asking the subjects to list five times when
the mother was speaking and five times when they thought
the father was speaking, there was no way of knowing
whether it was the coded verbalization itself which was
being used as a cue or the verbalizations leading up to
the one chosen or following the one chosen. There was
also no way of knowing why some verbalizations were picked
and not others.

Categorization of Reasons
for Judgments

 

 

In addition to making identifications about the
sex of the infant and the sex of the parent, the subjects
in the preliminary testing of the instrument were asked to
write an explanation about their reasons or how they went
about making their judgments. From their explanations and
comments, four different categories of explanations were

compiled. Category I referred to female/male differences

42

as they related to: (a) sex-role function; (b) personal
attributes, behaviors, anatomy, and dress; and (c) occupa-
tion. Category II referred to the language that was used;
i.e., either the words seemed distinctly masculine or
feminine or the structure of the sentence seemed masculine
or feminine. Category III related to the context or set-
ting in which the interaction took place. Category IV
referred to any explanations which were not included in

the other three categories. For the present study, a
description of each of the four categories was given to the
subjects. They were instructed to check one of the cate-
gories each time they made a judgment about the sex of the
infant and the parent who was speaking. If Category IV was
checked, the subject was also asked to write out his/her
own reason since it was to be checked when the reasons con-
tained in the other three categories seemed inappropriate

or unsuitable.

Bem Sex-Role Inventory

 

The Bem Sex—Role Inventory, although not included
at the time of the preliminary testing of the instrument,
was included in the present study. This instrument was
originally designed to implement empirical research on
psychological androgyny, a term that denotes the integra—
tion of masculinity and femininity within a single indi-

vidual. This inventory was one of the first instruments

43

to treat masculinity and femininity as two independent
dimensions whereas in the past, masculinity and femininity
were treated as two ends of a single dimension. Thus, an
individual had to be either masculine or feminine but not
both. This sex-role dichotomy has not allowed that many
individuals could be "androgynous"; that is, that they
might be both masculine and feminine, both assertive and
yielding, both instrumental and expressive, depending on
the situation and appropriateness of these behaviors. Like-
wise, strongly sex-typed individuals might be restricted in
the range of behaviors available to them as they go from
one situation to the next. It has been pointed out by Bem
(1974, p. 155) that highly sex-typed individuals are moti-
vated to keep their behavior consistent with an internal-
ized sex-role standard. Bem and Lenney (1976) later
demonstrated this notion in an experimental study where
they saw the avoidance of cross-sex behavior by sex-typed
individuals.

The Bem Sex-Role Inventory was selected for use in
this study to gain information about the subjects and to
determine whether a relationship existed between the scores
the subjects received on the instrument used to assess
whether they could identify the sex of the parent and the
sex of the infant and the Bem Masculinity and Femininity
scores. That is, is there a relationship between being

typed as Masculine, Feminine, or Androgynous and the number

44

of accurate identifications about the sex of the infant
or the sex of the parent? Does having an internal percep-
tion of one's own sex role influence how he/she perceives

others?

Reliability

 

Reliability refers to the accuracy and consistency
of the measuring instrument. There are many varieties of
test reliability. With regard to the instrument used in
the present study to assess whether subjects could accu-
rately identify the sex of the infant and which parent
was speaking, test-retest reliability was considered rele-
vant. In addition, reliability as measured by whether the
subjects responded in the same way, regardless of their
accuracy, was also considered relevant to this study.

Sixteen items were readministered to the subjects.
The 16 items were selected after the subjects had completed
the 40-item instrument. Table 3.1 shows the number of accu-
rate identifications about the sex of the infant and the
parent who was speaking for each of the 40 items. The num-
ber of correct responses was divided by the number of sub—
jects (n = 30) to arrive at a percentage of correct
responses.

Upon further examination, it was found that the
items could be grouped according to the following:

(a) whether greater than 50% of the responses about the sex

45

 

m H om m amm
SS mm Sm om «mm
om Sm Sw vH dHm
Sm HH Sm SH om
Sm HH SN w mmH
OS Hm Sm SH me
mm oH Sm SH SH
ow om Sm HH oH
ow NH SS mm mH
Sm om mS mm avH
om m Sv vH mH
om Sm Sm SH dNH
Sv vH SN w HH
Sm SH mS mm «OH
me mH Sm on m
Sm SH mS mm mm
mm mH mv mH S
mv mH Sm SH m
SH m mm mm um
om m Sm HH . v
ow vm om mH m
mm oH mm mH N
Sm om om m H
pcwmcH Ho xom pcamcH Ho xom Ho“ vacuum Ho xom paopmm Ho xom 90H .
now mowcoamom womaommom poohhoo how noncommom momcoamom uoonhoo Bow”

poonnoo mo &

Ho amassz Hence

poohhoo Ho &

Ho amnesz Hapoe

 

mHmSHdc¢ ooquHoHnom EopH

H.m mHDmB

46

.coHpmupchHEUm nsooow map cH povsHocH mums mEopH omosed

 

 

mS NN Sm HH c¢
Sm mN mm NH «mm
So 0N mm oH mm
ow vN SS NN «Sm
SN w 0v NH on
om m mm mN «mm
So ON mm ®H Nvm
mm mH mS NN mm
mm mH mm oH Nm
mm oH mS NN Hm
Sm SH Sm SH Nom
Sm SH mm NH NNN
OS HN on mH NN
mm mH mv mH SN
SN m ow SN mN
om mH SN w mN
SH m SS mN vN
pcwmzH Ho xom pcwmcH Ho wow now vacuum mo xom vacuum Ho xom noH .0
ROM womqoamom noncommom poohgoo HoH monsoawom momsoumom poonuou Sou”
poounoo Ho S mo nonesz Hapoe womanou Ho S Ho gonadz Hmpoe

 

Hemsanpcoov H.m mHnae

47

of the infant and the sex of the parent were accurate,
(b) whether greater than 50% of the responses about the
parent who was speaking were accurate but less than 50% of
the responses about the sex of the infant were accurate,
(c) whether less than 50% of the responses about the parent
who was speaking and the sex of the infant were accurate.
It was decided to include some items from each of the pre-
viously mentioned groups for readministration rather than
all 40 items. The concern in administering only 16 of the
items was that the subjects would feel fatigued and drop
out prior to the completion of the study. Table 3.1 indi-
cates which items were selected for readministration.

The items were readministered approximately one to
two weeks after the first administration. Test-retest
reliability was computed using the Pearson correlation coef-
ficient. Correlation coefficients were computed separately
for identifying which parent was speaking and the sex of the
infant. The coefficient for identifying the sex of the

infant was -.12. This correlation coefficient suggests that

there was little relationship between the two sets of scores.

The correlation coefficient for identifying which parent
was speaking was .48, which indicates a moderate degree of
positive relationship between the two sets of scores. That
is, in identifying which parent was speaking, there was
some tendency for those subjects who performed well on the

first administration to perform well on the second

48

administration and vice versa. In addition, the correla-

tion for identifying which parent was speaking was statis-
tically significant beyond the .01 level (p < .003). Thus
the two sets of scores for identifying the parent who was

speaking are truly correlated, even though the correlation
was not great.

Perhaps if all 40 items had been readministered, the
reliability coefficients would have been higher since relia-
bility is affected by the length of the test (Anastasi,
1968, p. 83). Another possible contributor to the low
reliability may have been the items which were selected for
readministration. Some of the items included those where
less than 50% of the subjects were able to accurately iden-
tify which parent was speaking and the sex of the infant
when the item was first administered. Since the computation
of the Pearson correlation coefficient is done on the num—
ber of accurate responses, those items where there were few
correct responses could have lowered the total correlation
coefficient.

Another index of reliability is to examine whether the
same subject responded to the same item on a different occa-
sion in the same way, regardless of whether the response was
correct. Table 3.2 shows the percentage of the subjects who
responded in the same way to the same item across the two
administrations. That is, if 20 of the 30 subjects identi-

fied which parent was speaking exactly the same way for the

49

Table 3.2
Percentage of Consistent Responses

Across Two Administrations of Sixteen Items

 

 

 

Item % Consistent Responses % Consistent Responses
No. for Sex of Parent for Sex of Parent
1 67 67
2 4O 53
3 73 4O
4 57 50
5 57 53
6 63 30
7 60 47
8 43 30
9 63 67
10 63 60
ll 43 60
12 47 60
13 43 57
14 63 63
15 73 60
16 3O 37
Average % of Average % of
consistent consistent
responses for responses for

sex of parent = 55 sex of parent = 52

 

50

two administrations of the item, the percentage of consis-
tent or same responses would be 67%. An overall average
of consistent responses was calculated by summing over the
percentages and dividing by the number of items (16). On
the average, 55% of the subjects identified the sex of the
parent in the same way across the two administrations of
the items while 52% of the subjects identified the sex of
the infant in the same way. Possibly, if all 40 items had
been readministered the consistency of the responses for
identifying which parent was speaking and the sex of the
infant may have been greater.

Reliability and Normative Data
for the Bem Sex-Role Inventory

 

 

The normative data for the Bem Sex-Role Inventory
consisted of 444 male and 279 female students at Stanford
and 117 male and 77 female paid volunteers at Foothill
Junior College in California. Internal consistency for the
Bem was estimated by computing coefficient alphas separately
for the Masculine, Feminine, and Social Desirability scores.
Masculinitycx= .86; Femininity<1= .80; Social Desirability
cx=.75for the Stanford sample. In the Foothill sample,
Masculinitycx= .86; Femininity<1= .82; and Social Desira—
bilitycx= .70. The reliability of the Androgyny differ-
ence score was .85 for the Stanford sample and .86 for the

Foothill sample.

51

A test-retest reliability was computed from the
Stanford sample with the following results: Masculinity
r = .90; Femininity r = .90; Androgyny r = .93; Social
Desirability r = .89. In the Stanford sample, 34% of the
males and 27% of the females were classified as androgynous,
36% of the males and 8% of the females were classified as
masculine, 35% of the females and 6% of the males were
classified as feminine. In the Foothill sample, 44% of
the males and 38% of the females were classified as androgy—
nous, while 22% of the males and 8% of the females were
classified as masculine and 40% of the females and 9% of
the males were classified as feminine.

In the present study, 33% of the male subjects
were classified as masculine and none of the female sub-
jects; 30% of the female subjects and 7% of the male sub-
jects were classified as feminine; 10% of the male subjects
and 7% of the female subjects were classified as androgy-
nous; and 13% of the female subjects and none of the male
subjects were classified as undifferentiated. Fewer of
the subjects in this sample were classified as androgynous

than in either of the normative samples.

Research Procedures
Subjects who participated in this study were
initially contacted through birth announcements in the

local paper or through word-of-mouth. Once the subjects

52

agreed to participate, a date and time were set. The sub-
jects were randomly assigned to one of the three forms of
the test instrument. Ten subjects were assigned to each
form.

Data collection was done in two lé—hour sessions
in the homes of the subjects. The subjects were contacted
one or two days prior to the beginning of data collection
to remind them of the appointment time.

At the time of the first session, each subject was
asked to read and sign the consent forms (Appendix A).
Subjects were told what was to happen during this session.
They were also reminded that all responses were to be com-
pleted independently. The subjects were then provided
with the test instrument and the answer sheets. They were
instructed to read the directions and ask any questions
they might have before they began.

At the end of the session, another appointment was
made one to two weeks later. As a reminder, subjects were
contacted one or two days prior to the second appointment.
Upon arrival, the subjects were briefed about what would
happen during that session.

Sixteen items from the original 40 items were
readministered at the time of the second session. The
directions and answer sheets were the same as was previously
given. In addition, each subject was given a description of

the categories which were to be used to indicate their

53

reasons about the judgments they made. They were instructed
to check one of the categories for each judgment about the
sex of the infant and one of the categories for each judg-
ment about the parent who was speaking. The Bem Sex-Role
Inventory was also administered. A description of these
instruments is included in the Instrumentation section of
this chapter. Upon completion of the instruments, subjects
were thanked for their participation and each couple was

paid $10.

Research Design

 

The design for this study was one way with two
levels. That is, there was one independent variable, the
sex of the subjects, with two levels, females and males.
Each subject was asked to make 40 judgments about which
parent was speaking and 40 judgments about whether the
infant was a girl or a boy. Thus, all of the subjects
received the same treatment condition.

The dependent variables were the number of correct
identifications (accuracy scores) for: (a) the sex of the
parent who was speaking and (b) the sex of the infant.

For some of the analyses, these scores were further broken
down into the number of correct responses for identifying:
(a) when the mother was speaking, (b) when the father was
speaking, (c) when the infant was a boy, and (d) when the

infant was a girl. The number of possible correct responses

54

for identifying the sex of the parent was 40. Twenty of
the 40 possible correct responses were when the mother

was speaking, and 20 were when the father was speaking.
Similarly, there were 40 possible correct responses for
identifying the sex of the infant. Twenty of the 40 pos-
sible correct responses were when the infant was a girl,
and 20 were when the infant was a boy. The effect of the
independent variable, the sex of the subject, was assessed
by comparing the average score on the dependent measures
for female and male subjects.

In addition to receiving an accuracy score for
identifying which parent was speaking and the sex of the
infant, each subject received a Masculinity and Femininity
score from the Bem Sex-Role Inventory. These scores were
used in a correlational analysis to determine whether a
relationship existed between the Masculinity and Femininity
scores and the accuracy scores for identifying which parent

was speaking and the sex of the infant.

Research Hypotheses

 

Five research hypotheses were generated to
empirically test whether subjects perceived the sex of
the infant and the parent who was speaking.

The null hypotheses may be stated as follows:

Null Hypothesis 1: No differences between

female and male subjects on the accuracy

scores of identifying which parent was speak-
ing will be found.

 

55

Null Hypothesis II: No differences between

female and male subjects on the accuracy

scores of identifying the sex of the infant

will be found.

Null Hypothesis III: No differences between

the number of accurate responses for identi—

fying when the mother is speaking and when

the father is speaking will be found.

Null Hypothesis IV: No differences from chance

(H0: p:=20) will be found in identifying which

parent is speaking, as measured by the accuracy

scores.

Null Hypothesis V: No differences from chance

(H0: p==20) will be found in identifying the

sex of the infant, as measured by the accuracy

scores.

Procedures for the Analysis of Data

The major focus of this study was to determine
whether subjects could accurately identify which parent
was speaking and the sex of the infant from reading
excerpts of verbal interactions between parents and
infants. The data used for the analyses included accuracy
scores (number of correct identifications) for judging
which parent was speaking and the sex of the infant.

To test the first two hypotheses a multivariate
analysis of variance test procedure was used. The eXpec—
tation that the dependent measures, i.e., the accuracy
scores for identifying which parent was speaking and the
sex of the infant, were interrelated suggested that a multi-

variate analysis be used. The alpha level for the multi-

variate approach was set at .05.

56

A t—test of dependent means was used to analyze
the data for Hypothesis III. Equality of population vari-
ance and normality of distribution were assumed. The maxi-
mum score for identifying when the mother was speaking was
20. Likewise, the maximum score for identifying when the
father was speaking was 20.

To test for Hypotheses IV and V, t-tests for popu-
lation means were used. Equality of population variance
and normality of distribution were assumed. There were 40
possible correct responses for identifying which parent was
speaking and 40 possible correct responses for identifying
the sex of the infant. Chance or 50% of the total possible
correct responses equaled 20 for identifying which parent
was speaking and the sex of the infant. Thus, symbolically,
the null hypothesis was HO:II= 20 for both Hypotheses IV
and V.

Additional analyses were also done that provided
descriptive data beyond the formal hypothesis testing. A
one-way analysis of variance procedure was used to test
whether there were significant differences between the
accuracy scores for the three different forms of the
instrument. Correlational analyses were used to explore
the relationship between the Masculinity and Femininity
scores on the Bem and the accuracy scores for identifying

which parent was speaking and the sex of the infant.

57

A frequency count was also made of the categories
chosen by the subjects which they used to explain why they
decided the parent was the mother or the father and whether

the infant was a girl or a boy.

Summary of Procedures

 

Fifteen married, Caucasian couples with one or
more children under the age of two volunteered to partici-
pate as subjects in this study. The subjects were from
the Lansing, East Lansing area and were contacted by birth
announcements in the local newspaper and through word-of-
mouth.

The instrument used in the study to assess whether
subjects could identify which parent was speaking and the
sex of the infant was developed from tape recordings of
mothers and fathers talking to their infants. A prelimi-
nary testing of the instrument resulted in several modifi-
cations and changes. Ultimately, the instrument consisted
of 40 items or segments of actual verbal interactions
mothers and fathers had with their infants. The subjects
were instructed to read each item or segment and on sepa-
rate answer sheets make a judgment about which parent was
speaking and the sex of the infant. All responses were
made independently.

Reliability of the instrument was determined by

readministering 16 of the 40 items. A criterion for

58

inclusion of an item in the second administration was
developed. Test-retest reliability was computed, and con-
sistency of the responses across the two administrations

was examined. There was little relationship between the two
sets of scores for identifying the sex of the infant. A
moderate degree of relationship was found between the two
sets of scores for identifying which parent was speaking.
Although this correlation was found to be statistically sig-
nificant, it would have been desirable if it had been
higher.

At the time of the second administration of the
items, the subjects completed the Bem Sex-Role Inventory.
The subjects were also given a description of categories
which they were to use to explain how they made each of
their judgments. Each couple was paid $10 for their par-
ticipation in the study.

The design for the study was one way with two
levels. The independent variable was the sex of the sub-
jects, females and males. The dependent variables were
the number of correct identifications (accuracy scores)
for the sex of the infant and the parent who was speaking.

Because the dependent measures were believed to be
highly interrelated, multivariate analysis of variance was
used to analyze some of the data. T—tests for dependent means
and population means were also used. Additional analyses

were done, correlating the scores from identifying which

59

parent was speaking and the sex of the infants and the Bem
Masculinity and Femininity raw scores. A frequency count
was also done of the categories chosen for each judgment
about the sex of the infant and sex of the parent. The

results of data analysis are presented in Chapter IV.

CHAPTER IV

ANALYSIS OF THE DATA

This chapter is divided into three sections for
the purpose of reporting the results of the investigation.
The first section presents the descriptive statistics, the
second section reports the results of the hypothesis test-
ing, and the third section presents further analyses of

interest.

Descriptive Statistics

The effect of the independent variable, the sex
of the subject, was tested using the data from the depen—
dent variables. Table 4.1 contains the cell means and
standard deviations for the accuracy scores identifying
the sex of the infant and the parent who was speaking.

A Pearson product-moment correlation coefficient
was computed to determine whether a relationship existed
between the accuracy scores for identifying which parent
was speaking and the sex of the infant. The correlation
coefficient was -.21, which was not statistically signifi-
cant (p < .44). Thus, there was little relationship between
the accuracy scores for identifying the parent who was
speaking and the sex of the infant.

60

Cell Means and Standard Deviations of
Accuracy Scores for Identifying the

Sex of the Parent and the Infant

61

Table 4.1

 

 

 

 

Subjects Parent Infant
i = 23.20 i = 20.53
Females
s.d. = 3.19 .d. = 3.44
x = 22.00 3? = 20.87
Males
s.d. = 2.54 .d. = 2.97
i = 22.60 SE = 20.70
Total
s.d. = 2.90 .d. = 3.16

 

62

Results of Hypothesis Testing

Since the dependent measures were believed to be
highly interrelated, multivariate analysis of variance was
used to analyze the data for two of the five hypotheses.
A t-test for dependent means was used to analyze the data
for Hypothesis III. T—tests for population means were used
to analyze the data for Hypotheses IV and V. To facilitate
reading, a restatement of each null hypothesis in nonsta-
tistical form will be presented along with the results of
the analysis. The results of Hypotheses I and II will be
presented together since they were analyzed using the
multivariate analysis of variance.

Null Hypothesis I: No differences between female

and male subjects on the accuracy scores of iden—
tifying which parent was speaking will be found.

 

Null Hypothesis II: No differences between female
and male subjects on the accuracy scores of iden-
tifying the sex of the infant will be found.

 

The multivariate test used to determine whether
male and female subjects differed in their accuracy scores
of identifying which parent was speaking and the sex of
the infant was not statistically significant F (2,27) =
.63, p < .54. Table 4.1 reports the means and standard
deviations. The null hypotheses were not rejected. That
is, male and female subjects have similar accuracy scores.

Null Hypothesis III: No differences between the

number of accurate responses in identifying when

the mother is speaking and when the father is
speaking will be found.

 

63

A t-test for dependent means was used to analyze
the data. The level of significance for this hypothesis
was p < .35, indicating that there were no differences at
the stipulated alpha level (p < .05). Thus, mothers were
not accurately identified a greater number of times than
fathers.

Null Hypothesis IV: No differences from chance

(H0211: 20) will be found in identifying which

parent was speaking, as measured by the accuracy
scores.

 

Hypothesis IV is the only effect that obtained
statistical significance at the stipulated alpha level
(p < .05). The mean score for identifying which parent
was speaking was 22.6, which was statistically significantly
greater than the chance level of 20 (t = 4.91, p < .0001).
A t—test for population means was used to analyze the data.
The null hypothesis was rejected in favor of the alternate
hypothesis that there was a significant difference from
chance (H1:II>20) in identifying which parent was speaking.

Null Hypothesis V: No differences from chance

(HO:I1= 20) will be found in identifying the

sex of the infant, as measured by the accuracy
scores.

 

A t-test for population means was used to analyze
the data. The mean score for identifying the sex of the
infant was 20.7, which was not significantly different from
the chance level of 20 (t = 1.21, p < .11). The null

hypothesis was not rejected, indicating that subjects could

64

not identify the sex of the infant with greater than chance

accuracy.

Further Analyses of Interest

The results of the multivariate analysis of vari—
ance were not significant in finding whether female and
male subjects differed in identifying the sex of the infant
and which parent was speaking. An additional analysis was
performed to test whether there was a sex-of—subject effect
when the difference between the accuracy scores of identi-
fying the parent who was speaking and the sex of the infant
were compared. There was a difference of 2.67 for female
subjects in identifying the parent who was speaking versus
identifying the sex of the infant. Male subjects had a
difference of 1.13 in accuracy scores for identifying the
sex of the infant and the parent who was speaking. An
analysis of variance test procedure failed to find any
significant difference between female and male subjects
(p < .38).

Analysis of the data for Hypothesis III revealed
that there was no statistically significant difference
between the number of accurate responses for identifying
the mother and the number of accurate responses for iden-
tifying the father. Analyses were also performed to test
whether there was a greater number of accurate responses

for identifying female infants versus male infants. A

65

t-test was done and was not found to be statistically
significant (p < .54).

An analysis of variance procedure was used to
test whether the accuracy scores were significantly dif-
ferent on the three forms of the test. The results showed
there were no significant differences between the accuracy
scores received by the subjects who had Form #1 versus the
subjects who received Form #2 or Form #3 (p < .75).

Correlational analyses were performed to explore
whether a relationship existed between the accuracy scores
for identifying which parent was speaking and the sex of
the infant and the Masculinity and Femininity scores
received from the Bem Sex-Role Inventory. Pearson product-
moment correlation coefficients are shown in Table 4.2.
As indicated in the table, the correlation coefficient for
the relationship between the Bem Femininity score and the
accuracy score for identifying which parent was speaking
was found to be significantly different from zero. More
specifically, the accuracy score for identifying when the
father was speaking and the Bem Femininity score were cor—
related. Thus, there was a tendency for those subjects
who had a high accuracy score for identifying when the
father was speaking to have a high Femininity score.

A frequency count was made to determine which of
the categories was used most by the subjects in giving

their reasons for selecting the sex of the infant and

66

Table 4.2

Pearson Correlation Coefficients

 

 

Accuracy Scores BEM Masculinity BEM Femininity
for Identifying: Score Score
Parent —.05 .30*
Mother .03 .04
Father .09 .33*
Infant .11 .05
Girl .19 -.05
Boy -.03 .09

 

*p < .05.

67

which parent was speaking. The number of responses for
each category was divided by the total number of responses
to arrive at a percentage. The percentages are shown in
Table 4.3.

As indicated in Table 4.3, Category II was chosen
most frequently for both the selection of the parent and
the infant. Category II related to male/female differences
as they related to language. That is, the subjects checked
Category 11 if they felt the basis of their judgment about
the sex of the parent or the infant had to do with the
language that was used in the item which they associated as
being characteristic of a particular sex. In identifying
the parent who was speaking, the subjects checked Cate-
gory II 58% of the time. In identifying the sex of the
infant, the subjects checked Category II 48% of the time.
The least—selected category for the sex of the parents was
Category IV, which was to be chosen when the other cate-
gories seemed inappropriate or unrepresentative. The
least-selected category for the sex of the infant was

Category III, context.

Summary
Five hypotheses were tested. Hypotheses I and II

used a multivariate analysis of variance testing procedure.
Results of the analysis indicated there was no significant

difference between male and female subjects in accurately

68

Table 4.3
Categories Selected as Basis for Judgments

About the Sex of the Parent and the Infant

 

 

Sex of Parent Sex of Infant
Category I 23% 23%
Category II 58% 48%
Category III 14% 8%
Category IV 5% 21%

 

Category I referred to female/male differences as they
related to sex-role function, physical appearance, or
occupation.

Category II referred to female/male differences in
the language that was used, either words or phrases.

Category III referred to the context of the situation
of the setting described in the item.

Category IV was an open-ended category which could be
checked when the other categories seemed inappropriate
or unsuitable.

69

perceiving the sex of the parent and the sex of the
infant.

Hypothesis III was tested using a t—test. The
conclusion drawn was that there was no significant differ—
ence in the number of correct responses for identifying
when the mother was speaking and when the father was
speaking.

Analysis of the data for Hypothesis IV, using a
t-test for population means, showed statistical signifi~
cance for the subjects being able to perceive which parent
was speaking with greater than chance accuracy (p < .0001).

A t-test for population means was used to test
Hypothesis V. Statistical significance was not reached.
The null hypothesis was not rejected. The sex of the
infant could not be identified with greater than chance
accuracy.

Further analyses were conducted to test whether
there was a sex-of—subject effect. An analysis of vari—
ance procedure used to test the data resulted in no sig-
nificant differences between female and male subjects.

To test whether there was a difference in the
accuracy scores for the three forms of the test, an analy-
sis of variance procedure was used. No statistical signifi-
cance was found (p < .75).

Correlational analyses were done to explore the

relationship between the accuracy scores and the Masculinity

70

and Femininity scores from the Bem Sex-Role Inventory. A
statistically significant correlation was found between the
accuracy score for identifying the father and the Bem
Femininity score. The correlation coefficient, however,
was low (.33), which suggested that the relationship was
not close.

Subjects chose Category II 58% of the time as con—
taining the reasons for their decisions about the sex of
the parent, while they chose the same category 48% of the
time when making judgments about the sex of the infant.
Category II referred to male/female differences with

regard to the language used in the items.

CHAPTER V

SUMMARY, CONCLUSIONS, AND DISCUSSION

In this chapter, the study is summarized, and
conclusions based on the data analysis are explored. A
discussion of the results as well as the limitations of
the study are included, along with suggestions for future

research in this area.

Summary

The primary purpose of this study was to determine
whether differences were perceived in the speech of mothers
and fathers toward their infants such that the sex of the
infant and which parent was speaking could be accurately
identified by naive subjects. The assumption that men and
women use the English language differently, depending upon
the situational context, the role that one is assuming at
a given time, age, socioeconomic background, and occupation,
served as the theoretical basis for this study. The addi-
tional knowledge that would be gained from empirical data
supporting the notion that men and women use the English
language differently in their roles as parents provided

the impetus for this study. Also, very little experimental

71

72

research had been reported on this theoretical assumption
to date.

A review of the literature was conducted in three
areas: Female/Male Speech Differences, Mother-Infant,
Father-Infant Verbal Interaction, and the Impact of Sex of
Infant on Parental Verbal Behavior. The summary of female/
male speech differences suggested that, in general, there
were differences in the way females and males use the
English language. The finding with the most empirical
support is that women are more "polite" than men. They
avoid strong expressions or swearing and will more often
choose the "correct" way of talking.

Moss (1974), one of the few researchers to comment
on what mothers and fathers say to their infants and the
impact that the child's sex has on parents' speech, noted
that parents were relating to their 7-week-old infants in
stereotyped sex—role ways. Macfarlane (1978) audio-tape
recorded parents immediately following delivery of their
infant and found a preoccupation with gender as well as
the attribution of a number of sex stereotypes by the par—
ents and the medical staff.

Fifteen married, Caucasian couples with one or
more children under the age of two were paid volunteer
subjects for this study. The subjects were from the

Lansing, East Lansing area and were contacted by birth

73

announcements in the local newspaper and through word-of-
mouth.

The instrument used to assess whether subjects
could accurately identify which parent was speaking and
the sex of the infant was developed from tape recordings
of mothers and fathers talking to their infants. A prelimi-
nary testing of the instrument showed that 63% of the judg-
ments about the parent who was speaking were accurate, while
70% of the judgments about the sex of the infant were accu—
rate.

Several modifications and procedural changes were
made as a result of the preliminary testing of the instru-
ment. The number of couples who agreed to tape record
their verbal interactions was expanded from two couples to
four couples. Instead of using entire transcripts as had
been done for the preliminary test, the transcripts were
divided into 60—second segments or 50 words. Five segments
of the mother speaking and five segments of the father
speaking were randomly selected from the four transcripts
for a total of 40 segments or items. By randomizing the
order of the 40 items, three forms of the instrument were
made. The directions were changed in that subjects had to
identify which parent was speaking and the sex of the infant
for each of the 40 segments or items. Separate answer
sheets for identifying which parent was speaking and the sex

of the infant were provided for each subject.

74

The reliability of the instrument was low for
identifying the sex of the infant and only moderate for
identifying which parent was speaking. Reliability was
tested by readministering 16 of the 40 items one to two
weeks after the administration of the 40-item instrument.
Test—retest reliability was computed using a Pearson
product-moment correlation coefficient. The correlation
coefficient for identifying which parent was speaking was
.48, while the correlation coefficient for identifying the
sex of the infant was -.12. Reliability was also tested by
examining whether the subjects responded in the same way to
the same item across the two administrations of the items.
On the average, 55% of the subjects identified which parent
was speaking in the same way across the two administrations,
while 52% of the subjects identified the sex of the infant
in the same way. Again, the reliability was low.

At the time of the second administration of the 16
items, the subjects were also asked to indicate their
reasons for deciding which parent was speaking and the sex
of the infant. The subjects were given a choice of four
categories which they could use as a basis for their judg-
ments. A description of each category used and an answer
sheet were provided. In addition, each subject was admin-
istered the Bem Sex-Role Inventory, which classified them

as Masculine, Feminine, or Androgynous.

75

The design was one way with two levels. The inde—
pendent variable was the sex of the subjects. The dependent
variables were the accuracy scores for identifying the sex
of the infant and the parent who was speaking. Five
hypotheses were generated to test whether subjects could
identify with greater than chance accuracy (H0: u==20)
which parent was speaking and the sex of the infant. Since
the dependent measures were believed to be highly inter-
related, a multivariate analysis of variance procedure was
used to analyze some of the data. T—tests for dependent
means and for population means were also used for some of
the analyses. Correlational analyses were included to
determine whether a relationship existed between the Bem
Masculinity and Femininity scores and the accuracy scores
for identifying which parent was speaking and the sex of

the infant.

Results

A multivariate analysis of variance procedure was
used to test whether there was a sex-of—subject effect.
Results of the analysis indicated that there were no sig-
nificant differences between females and males in either
correctly identifying which parent was speaking or in
identifying the sex of the infant.

A t-test for dependent means was used to test

whether mothers were accurately identified a greater number

76

of times than fathers. Analysis of the data showed there
was no significant difference between the number of cor-
rect responses for identifying either the mother or the
father.

The results of the data analysis for testing
whether subjects could identify which parent was speaking
with greater than chance accuracy (HO: u==20) showed sta-
tistical significance (p < .0001). That is, subjects could
correctly identify which parent was speaking with greater
than chance accuracy (H1: u > 20).

No statistical significance was found after testing
whether subjects could identify the sex of the infant with
greater than chance accuracy (HO: u==20).

Further analyses were performed to test whether
there was a sex-of—subject effect. Differences in accuracy
scores between identifying the sex of the infant and the
parent who was speaking were compared for female and male
subjects. The results of the analysis of variance proce-
dure showed no sex—of—subject effect (p < .38).

An analysis of variance procedure was used to test
whether there were significant differences between the
accuracy scores of the three forms of the instrument. No
difference was found (p < .75).

Correlational analyses were performed between the
Bem Masculinity and Femininity scores and the accuracy

scores identifying which parent was speaking and the sex

77

of the infant. A statistically significant correlation was
found between the Femininity scores on the Bem and the
accuracy score identifying which parent was speaking. More
specifically, the higher the Femininity score, the higher
the accuracy score for identifying when the father was
speaking. The correlation coefficient was low (.30), how—
ever, which suggests that the relationship was not close.
Subjects selected Category II most frequently as
the basis for being able to identify both the sex of the
infant and the parent who was speaking. Category 11
stated that the language used in the items seemed associ-
ated with a particular sex or that the phrasing or language

style seemed masculine or feminine.

Discussion

 

Only one of the five hypotheses tested achieved
statistical significance. That is, subjects could identify
which parent was speaking with greater than chance accuracy
(p < .0001). There was no effect for the independent vari-
able, sex of subject.

The findings from this study were consistent with
earlier research that there are differences in the speech
of women and men (Fillmer & Haswell, 1977; Garcia-Zamor,
1973; Gilley & Summers, 1970; Gleser & Gottschalk, 1959;
Key, 1972; Kramer, 1973; Lakoff, 1973; Thorne & Henley,

1975). The results of this study, however, are interesting

78

from the standpoint that the sex of the infant could not
be identified from what the parents said to their infants.
That is, the sex of the infant did not appear to influence
what parents said in any identifiable way. This is in
contrast to a limited amount of research which indicates
that parents speak and behave differently on the basis of
the sex of the infant (Macfarlane, 1977; Moss, 1974). Addi-
tionally, in the preliminary testing of the instrument, 70%
of the judgments about the sex of the infant were accurate,
which is also in contrast to the present findings. The
question of why the subjects were unable to accurately
identify the sex of the infant warrants further explanation.
One possibility is that the instrument itself was
not powerful enough to detect the differences. The items
for the instrument were taken from four couples' verbal
interactions with their infants. The parents themselves
chose what they would record. Thus, the number of couples
who recorded was low so that the representativeness of the
verbal interactions is called into question as well as the
method used to collect the samples of verbal interactions.
Increasing the number of couples who tape recorded would
have added to the generalizability and representativeness
of the sample of verbal interactions which, in turn, would
have increased 13M; power of the test. In addition, a set-up
similar to the one used by Rebelsky and Hanks (1971) would

have been desirable. In their set-up, the microphone was

79

operated by a voice key to eliminate silent periods, and
the recordings took place over a longer period of time.
With the recorder activated by a voice at specified inter-
vals rather than the couples choosing when they would
record, a wider, unbiased sampling could have occurred.

A number of methodological and procedural changes
occurred as a result of the preliminary testing of the
instrument which also may have resulted in the subjects'
inability to correctly identify the sex of the infant. One
of the largest changes from the preliminary testing was in
using 60—second segments or 50 words instead of entire
transcripts. Perhaps more cues or different cues were
needed to identify the sex of the infant than were needed
to identify which parent was speaking. By not using the
entire transcript, the cues may have been lost. It is also
possible that the percentage of time that the sex of the
infant does influence the speech of parents is very small
and that those segments containing cues about the sex of
the infant were not included in randomly choosing the seg-
ments.

Any conclusions or interpretations of the findings
of this study are limited by the questionable reliability
of the instrument. Perhaps if all 40 items, instead of
16 items, had been readministered, the reliability of the
instrument would have been raised. With an increase in the

reliability, there would have been stronger support for the

80

finding that subjects could identify which parent was speak-
ing. However, since the reliability was so low for identi-
fying the sex of the infant (—.12), it is doubtful whether
additional items would have raised the reliability coeffi-
cient substantially. The instrument would probably still

be unreliable with respect to identifying the sex of the
infant.

With regard to the correlational analyses, there
appeared to be a positive relationship between the Femi-
ninity score on the Bem Sex-Role Inventory and the accuracy
score for identifying when the father was speaking. There
was no ready explanation or theoretical underpinning for
this finding. The correlation coefficient was low, how-
ever, which suggests that although there was a tendency
for those subjects who had a high accuracy score in identi-
fying when the father was speaking to also score high on
the Femininity scale, the relationship was not close.

There was no readily available explanation or theoretical
underpinning for this finding.

Perhaps more meaningful information could have been
gained by the use of an instrument such as the Myers-Briggs
Type Indicator. The Myers—Briggs contends that individuals
differ in their basic preferences with regard to perception
and judgment and classifies individuals into dichotomous
type categories. One of the categories, Sensing or Intui-

tion, has to do with whether an individual perceives

81

directly in a factual, realistic way or indirectly through
associated ideas and imaginative implications. With regard
to this study, it would be interesting to determine whether
there is a difference in the accuracy scores between
intuitors and sensors. That is, are intuitors or sensors
able to identify the sex of the infant and the parent who

is speaking more accurately? If a difference was found
between sensors and intuitors, perhaps the process the sub-
jects went through in making their judgments could be better
understood.

In addition, if an indepth interview followed the
subjects' identification of the sex of the infant and which
parent was speaking, more information about the process of
decision making might become available. For instance,
instead of just knowing that the subjects thought the lan-
guage used was distinctly masculine or feminine, it would
be possible to find out the exact word or phrase that was
used. This method seems preferable to the checking of
categories which was used in this study or may have been

used as a supplement to checking the categories.

Limitations

 

When interpreting the results of this research,
several things must be kept in mind. The way in which the
research subjects were obtained, the size of the samples,

and the instrument used for assessing whether subjects

82

perceived differences in the way mothers and fathers talk
to their infants.

Couples from a city where there is a large mid—
western university volunteered to participate as subjects
in this research. Volunteers may differ from other groups
of individuals in the general population and thus affect
their response set for completing the instruments. Their
willingness to participate may indicate that they are more
sensitive and perceptive and aware of themselves as parents
than the general population. In meeting with the couples,
there was a sense that they wanted to perform well on the
instruments even though it was explained to them that they
would not receive specific feedback about their perform-
ance. As a group, they seemed highly motivated and
expressed a strong interest in knowing the results of the
experiment. These attitudes may not reflect the attitudes
of the general population.

A second limitation is with the couples who agreed
to audiotape record their verbal interactions with their
infants. Four couples volunteered to participate. Because
of the low number of participants, any conclusions or gen-
eralizations about the representativeness of their verbal
interactions to couples in the general population is quite
limited. In addition, their cooperativeness and willing-
ness to participate may set these volunteers apart from the

population as well as their small number.

83

A third limitation lies with the test instrument
itself. As mentioned previously, test—retest reliability
for identifying the sex of the infant was low and only
moderate for identifying the parent who was speaking. It
was not possible to predict an individual's score on the
second administration of the item with his/her performance
on the first administration. Thus, even though subjects
could identify which parent was speaking with greater than
chance accuracy, any conclusions are limited by the low

consistency and accuracy of the instrument.

Recommendations for Further Research

Further exploration and investigation would be
appropriate and useful based on the discussion of the
results found in this study. The notion that women and
men, in their roles as parents, use the English language
differently is intriguing and important to our understand-
ing about sex-related differences. If differences in the
speech of men and women can be firmly established, perhaps
the theoretical underpinnings for why differences exist can
be further developed and explored. From there, an explora-
tion could begin about the role that language plays in the
learning of sex roles and sex—appropriate behavior in our
society.

Recommendations and suggestions for further research

include replication of the study but with correction of the

84

shortcomings discussed in the preceding Discussion and
Limitations sections. More specifically, it would be
desirable to test whether the sex of the infants and the
parent who was speaking could be identified by a larger
sample. Furthermore, it would be desirable to conduct the
research using a nonvolunteer sample such as with all par—
ents who enter a hospital to have a baby. By increasing
the size of the sample and using a nonvolunteer group, the
results would be more generalizable to the population.
Improvements in the instrumentation are also highly
desirable. The number of couples who tape record their
verbal interactions should be increased, and as mentioned
previously, the method of collecting the sample of verbal
interactions should be changed. The recordings should take
place over a longer period of time and at specified inter-
vals, i.e., a 24-hour period of recording every week. A
microphone that is operated by a voice key to eliminate
silent periods would be useful. The tapes should be
transcribed by naive coders to eliminate experimenter bias.
Administer the Myers-Briggs Type Indicator instead
of the Bem Sex-Role Inventory to learn more about the ways
subjects perceive things. Do some subjects perceive
directly in a factual, realistic way, or do they perceive
indirectly via associated ideas and imaginative implica-
tions? Do subjects who perceive indirectly identify the

sex of the infant and the parent who is speaking more

85

accurately than subjects who perceive things in a factual,
realistic way? Is there something inherent in making judg-
ments about the sex of the infant and which parent is speak-
ing which makes it easier for those subjects typed as
intuitors versus those who are typed as sensors to attain

a higher level of accuracy?

Include an indepth interview after the subject has
made his/her judgments about the sex of the infant and which
parent is speaking. The subject can be asked directly about
each judgment and what cues or decision-making process he/
she used. This information would be especially useful if
differences were found in the speech of mothers and fathers.
It would then be possible to know exactly what differences
were perceived as well as that they exist.

Another interesting research project relating to
this one would contrast the speech of women and men in
their roles as parents with women and men in work situa-
tions. The instrument could be developed in a similar way
using tape recordings, but eliminate asking subjects to
identify the sex of the infant, only whether it was a woman
or a man speaking. The focus of the study would be on
whether women and men use the English language differently,
depending on the context of the situation and the particu-
lar role they are playing.

Another way of establishing whether differences

in the speech of mothers and fathers exist would be to

86

collect and analyze large samples of verbal interactions.
The recordings could be transcribed and coded according to
the vocabulary used, the language style, the frequency and
amount of verbal interaction, the number of silences and
interruptions, the number of references to the sex of the
infant, etc. By coding the interactions, it would be pos-
sible to establish how many, if any, of the verbal inter—
actions have to do with the sex of the infant and how many
have to do with unrelated topics.

In summary, the evidence from this study suggests
that accurate judgments can be made about the sex of the
parent from their speech, but the sex of the infant cannot
be accurately identified from their speech alone. However,
before any conclusions can be drawn, more powerful, rigorous

studies need to be designed and implemented.

APPENDICES

87

APPENDIX A

SUBJECT CONSENT FORMS FOR

RESEARCH PARTICIPATION

88

APPENDIX A

SUBJECT CONSENT FORMS FOR
RESEARCH PARTICIPATION

Consent Form for Mother-, Father-Infant
Speech Study

 

I understand that participation in this study will
involve audiotape recordings of conversations between
myself and my newborn infant. I understand that the audio—
tape recordings are for the purpose of research for this
study and that my confidentiality will be strictly main-
tained.

I understand that I am free to discontinue my par-
ticipation in the study at any time without penalty.

I have freely given my consent to participate in
this study being conducted by Roberta T. Hale under the
supervision of Dr. William Hinds. I have not been offered,
nor do I expect to receive, any remuneration for my par-
ticipation in this study.

 

Signature of Participant

 

Date

 

Signature of Witness

 

Date

89

90

Consent Form for Mother-, Father-Infant
Speech Study

 

 

I have freely consented to take part in a research
study being conducted by Roberta Hale under the super-
vision of William Hinds, Ed.D.

I understand the explanation that has been given
and what my participation will involve.

At the conclusion of my spouse's and my participa-
tion, I understand that together we will receive $10.00
in remuneration.

I understand that I am free to discontinue my par-
ticipation in the study but will not receive any part of
the $10.00 remuneration nor will my spouse.

I understand that the results of the study will be
treated in strict confidence and that I will remain anony-
mous.

 

Signature of Participant

 

Date

 

Signature ofFWitness

 

Date

91

General Information Sheet

 

Name of Parents: Mother: Father:
Age of Parents: Mother: Father:
Occupation of Parents: Mother: Father:
Education of Parents: Mother: Father:

Sex of Infant: Boy Girl

APPENDIX B

TELEPHONE SOLICITATION OF SUBJECTS

92

APPENDIX B

TELEPHONE SOLICITATION OF SUBJECTS

Telephone Conversation

 

Mr./Mrs. , my name is Roberta Hale.
I'm a doctoral student in Counseling Psychology at MSU. As
my research project, I am investigating what mothers and
fathers say to their children. I saw your name in the birth
announcements in the State Journal and am calling to see
whether you and your husband/wife would be interested in
participating. I'll be offering $10 for each couple who
participate in both parts of the project.

 

Your participation would involve you and your
husband/wife reading through some excerpts of actual con-
versations that mothers and fathers had with their newborn
and then independently making judgments as to the sex of
the parent and the sex of the infant. I will come to your
home while you complete this. In approximately 2-3 weeks
from that visit, I'll return to your home with a second set
of excerpts of parents' speech and again ask you to make
judgments as to the sex of the parent and the sex of the
infant. There will also be a questionnaire about why you
made the judgments you did, as well as a brief personality
inventory. I expect the first and second visits to each
take approximately 1% hours. I will pay you the $10 at
the conclusion of the second visit.

Do you have any questions you'd like to ask me?

93

APPENDIX C

INSTRUMENTATION AND SCORING KEY

94

APPENDIX C

INSTRUMENTATION AND SCORING KEY

Mother-, Father-Infant Speech Study

 

DIRECTIONS: The attached instrument is composed of 40
excerpts of actual conversations mothers and fathers had
with their infants. References to whether the mother or
father is speaking and the sex of the infant have been
eliminated. On the answer sheets, you will be asked to

make a judgment as to whether the mother or the father was
speaking and the sex of the infant for each of the 40
excerpts. ON ANSWER SHEET #1 PLEASE RECORD YOUR JUDGMENTS
AS TO WHETHER YOU THOUGHT THE MOTHER WAS SPEAKING OR WHETHER
YOU THOUGHT THE FATHER WAS SPEAKING BY BLACKENING SPACE #1
IF YOU THOUGHT THE FATHER WAS SPEAKING AND SPACE #2 IF YOU
THOUGHT THE MOTHER WAS SPEAKING. ON ANSWER SHEET #2, PLEASE
RECORD YOUR JUDGMENT AS TO WHETHER THE INFANT WAS MALE OR
FEMALE BY BLACKENING SPACE #1 IF YOU THOUGHT THE INFANT WAS
A GIRL AND SPACE #2 IF YOU THOUGHT THE INFANT WAS A BOY.

95

96

Mother—y_Father—Infant Speech Study

DIRECTIONS: Please read each excerpt and record your deci-
sions as to the sex of the speaker and the sex of the infant
on the answer sheet. PLEASE DO NOT WRITE ON THIS FORM.

1. You didn't want much of that a minute ago... says my
gives me lots....For such a little one, you come
up with the big ones, huh?... 's looking more like
now. You're a lucky looking like your

instead of your

2. Let's see if can go through the whole thing with-
out a cry. Spoke too soon. Record is still two in a
row. has to be able to go through more than two
diaper changes without a scream or a holler or a hoot.

really tightens up stomach you know?

3. '8 been a real good . I'll tell you, you've
been a real good ... Who are you sticking that
tongue out at? Are you sticking that tongue out at me?
Umm, tastes good... Don't cry... Hi, Sweetsl... Are you
going to cry? No. "I'm going to smile I believe,

." What's that little eye doing?

4. Say something. What's that coming?... Say something.

Can you say "Hi, "... You going to the bathroom?

...Yea, look at that face of yours, it's turning all

red, turning all red... Look at your over here.

What have you got to say to your ? Talk to your
Say "Hi, .”

 

5. What has this country done to our babies? What else
can we give you? Pizza, want a pizza? I think we can
get a nipple big enough to fit a mushroom or a piece of
pepperoni, huh? Still hungry? Round 2, so far the
bottle won the first round. didn't eat that much.

just attacks the damn thing, I mean the darn thing.

6. Well, wasn't following the lights yesterday.
sat and looked... Then ___ got bored with it...
Heavy, you're getting heavy. Oh, what's the matter?
Okay, here goes. Are you ready to go into the water?
It's okay. You've got a funny look on your face.

10.

ll.

12.

97

”Oh my . This is such hard work and all I wanted
was a little more food. You keep insisting on these
burps. Oh, what do I do? I'm just so hungry, so hungry.
I've only had 32 fluid ounces, you're starving me.”
Here we go, have a little more . What the heck.
Oh, take it slow, take it slow. We sure got rid of
those hiccups, didn't we? Yea, we got rid of the hic—
cups, for sure. We don't want to get rid of all the
milk, now, so we better go a little bit slower. Yea,
it's better if you take it slower.

 

Are you looking forward to Grandma and Grandpa coming?
Yea, they're coming soon. Yea, they're coming espe—
cially to see you. Aren't you lucky? You have a nice
Grandma and Grandpa. Yes, you do. Your Grandma's gonna
be surprised at how big you've gotten, too. Yea, you
smile likethat for them too, okay? Are you gonna smile
like that for them?

We can't stay here all day even though I think you'd
like to... How come you're all wet? On no, baby, don't
cry. No, my goodness, my goodness. I'd cry too if my
clothes were all wet. Oh, stinky baby. We'll get you
out of those wet clothes.

 

loves Pluto... That's what happens when you don't
burp... Are your pajamas all wet?... One arm in... Good
One arm in. Oh, . Two arms in. "Oh my,
I'm getting so hungry.”... Oh, goodness gracious. Try
once more. Oh, my goodness gracious. These hiccups,
oh my. What do you think , want to try to eat?

All right, let's try to eat.

You're the cutest, little starer, yes you are, the
cutest little starer. Are you having a nice, mid-
morning breakfast. It's around lunchtime, you know,
and you still haven't finished breakfast. Been munch-
ing all morning long. That'd better do or I don't know
what we'll do. Yea, I don't know what we'll do. Okay,
try a little burp now. How was that? That was pretty
good and you want more already, huh?

You are a doll face. Talk to , talk to .
What a big smile you've got. You've got a beautifu
smile. You are gorgeous. You are so sweet. You gonna
talk: to ? Yea, talk to . Can you say "Hi,

." You're such a pretty little . "Look at
those socks and feet and legs. Bathroom time.

 

 

13.

14.

15.

l6.

17.

18.

19.

20.

98

just doesn't want to. I know why you're grunting
You've got a surprise for us, don't you? looks
like a little munchkin, a chubby munchkin, huh? Putting
on chubbies with all the chow-chow? You want to say
something? What a good

A nice resounding burp, . How about one of those?
Whee...over the shoulder. Oh, way up high. Oh, a
sneeze. You're working on it. We'll try it this way.
We'll try this down here and this here and lean forward

just a little bit.

Oh, so tired. Eyes closed. Hold still. Gotta change
the diaper. A little smile. Oh a frown, little frown.
. A couple more minutes and we'll change you. I
don't know, you rest so often. Oh, baby face.

made my breakfast for me. Isn't that nice of your

 

?
Giving up, are you? "Hi , my name's Teddy."
There, Teddy Bear's on his side. Both of you going to
sleep?... Stay warm. You gonna stay warm? Nice and

warm? Give me a big smile. What else do you have to
say this morning? What else? Come on, how about smil-
ing for me? Oh, just a smile, come on.

Gonna talk to ? You gonna talk to ? Oh, all
those noises. What? See the pretty baby in the mirror?
See the pretty baby? Yea, you're smiling at the baby,
aren't you? You're such a big smiler. Yes you are.

”Oh hurry. I don't want to get cold." You gotta
have a bath today. There they are, look at those
kittens... "Oh hurry , hurry." Oh, not a crummy
tee shirt. This is a bad one. There. You're such a
good . Here we go. Want to burp?

Where are you going? You gonna be a deep sea diver,
huh? Sliding down into the water... What you got your
toes clenched for, huh? You like splashing around in
the water? It seems like it. You have such a rough
day.

Oh, stretch, stretch, stretch. Come back. Did you
burp? That's pretty puky. It'll be all right,

 

 

's so tired... Oh, you're so hungry. "Oh, hurry
.” Here we go, . Over here. Now, get your
tongue out. There, that's a . Good . Want

to go to Grandma's tomorrow?

21.

22.

23.

24.

25.

26.

27.

99

 

How are you doing big ? Are you doing pretty

good? Yea, you really are doing well today. How come?
How come? Are you gonna talk to today? Make some
noises for , can you? What kind of noises can you

make? Can you make some noises? Sure you can. Can
you do that?

You look so cute. Pretty good, huh?... You smelling
your toes? They smell pretty?... Look at this dry
skin... Nice and clean... Yea, for a little while.
For a little while then '11 realize that 's
hungry again, then start screaming.

 

You're busy all the time aren't you? Gotta find some-
thing to do. Yea. Doesn't take much to keep you happy

 

does it? says, "Just give me a hand, and I'm
happy or the nightgown." '8 going to be a boxer.
Look at fists. You gonna be a boxer, ? Is

 

that a good flavor?

Good kid. Ready to wet again, aren't you? "Yep, soon
as I can.” Made it, made it through another diaper
change. Want to say a few words to the audience? Go
ahead, say a few words. Don't be bashful, go ahead.
Tell them how glad you are to be here.

That's a cute smile on that face of yours. That's a
real cute smile. What are you doing? We gotta cover
you up. Look at that white tongue. Oh, still sleepy?
You still sleepy? What a good . You gonna smile
for ? You gonna smile for , huh? You ready
to eat some more?

 

 

What's going on in here?... Taking a swimming lesson?
Gee, look at the size of that belly? 's expanding
like crazy. You like the water?... It's like being by
poolside, right?... sure opens eyes wide,
you know?... doesn't know if should be
enjoying it or should be scared.

You're really different from yesterday. Smiles...
says, "I'm a whole new person today, since I

went to the bathroom.”... Hey , want to talk to

your ? You're so relaxed. It's nice to see you
so relaxed for a change. Is that a yawn or a smile?
That was definitely a smile.

 

28.

29.

30.

31.

32.

33.

34.

100

You're going to be just fine, aren't you . You're
an awful hungry, hungry, hungry little . The
problem is you're always hungry. You're going to eat us
out of house and home, maybe. Is that what you're try-
ing to do, eat us out of house and home? You sure do
look cute, yes you do.

A great big smile. One more, let's see it. Stick your
tongue out and smile. Can you say something to us? Are
you ticklish? Not ticklish, huh? Stick your tongue out
and smile. Can you say something to us?

Is that what we're going to do, get another chair for
you? What do you have to say for yourself? You're get-
ting so excited. What are you getting so excited about?
Oh, my goodness. What? You want to talk so badly.
What? What are you going to say? You just want to
talk.

Are you ready to be awake for awhile? Yes, you little
sleepyhead? Time for you to be awake. What? What do
you want to tell ? "Goo" to you too. What?
What's the problem? What's your problem? What you
little fussbudget? What's wrong? What's the problem?

You just want to be picked up again. You can't be

picked up all the time. 's got other things to do.

"Can I rest, just for a little bit? Oh, just so hungry.”
'5 ready for seconds, . "Skip the burps. I

want my second course." Here you go... Hurry

hurry... Was that a hiccup?... You're so expressive...

This is gonna be a quick one cuz Mom and Dad have

to go eat breakfast.

Oh wow, great smile! If I pull your ear like this,

will you smile? How about your cheeks? Who's that sit—
ting over on the couch? Is that your ? Can you
see ? Can you see through the covers?

You've heard that voice before. You can see

can't you?

 

You stink don't you?... Want to hug your Teddy Bear?
Teddy Bear's talking to you... Kiss him on the lips.
Oh, good kiss. Want to hug him? Oh, Teddy's hugging

Kiss him on the nose. We'll just sit here and
talk to you. Do you think Teddy Bear will crush you?
Do you think Teddy Bear will sit on you?

35.

36.

37.

38.

39.

40.

101

What's with the two hats, huh? Are you just gonna be
an ? Maybe that's it... You better like .
's with you a lot more than me. I think 's

 

starting to focus now. I mean, not just today over

yesterday, but it just seems like the last couple of
days would follow something a lot easier.

How can you get that worried look? , there's
nothing to worry about. Nothing to worry about at your
age. You're really drinking that down. You make me
feel like I didn't give you anything... 's going
to end up drinking that whole thing. , your

'5 good to you.

Oh goodness. What's the matter, honey? Oh, we let you
sleep today too long, huh? You gonna let us sleep
tonight? Oh, honey. Tell me about it. Let me see,

do you have wet pants? Do you? Let me check and see.
Oh, wet pants, wet pants. No wonder you're unhappy.
Okay, we'll go change you.

It's Easter. You know, little Easter bunnies and
Easter baskets and carrots and candy. Too much for you
these days, huh? You'll know soon and then you'll have
lots of fun. Right now you just enjoy baths and eating.
Sleeping, eating, changing your diapers and baths and
rocking in your chair, that's what you like huh?

You have fat cheeks. Wake up... Let's see, little
How do you expect to get your hand out? Tell

me, how do you expect to get your hand out?... You sure

are strong. Tell me what I can do for you?... You cute
little doll, you. You're cute. Hey sweets. You going
to talk with ?

Look how warm and cozy. Oh there. All better now.
With no burps though. Your burping mechanism is hay—
wire. Oh, there we go. That was a good one, honey.

Say, ” I'm all right." Such a hiccup, so loud.
Hi Tiger. Hi, little. Oh you looking at the trees?
"Oh, I'm so angry at . Where's my breakfast?

Where's my breakfast?"

102

Scoring Key for Identifying the
Parent Who Was Speaking

00

10.
11.
12.
13.
14.
15.
16.
17.
18.
19.

20.

COCDS‘ICDCfiI-b

l—’

{ONION

21. 2
22. 2
23. 2
24. 1
25. l
26. 1
27. 1
28. 1
29. 1
30. 2
31. 2
32. l
33. l
34. 1
35. l
36. 2
27. 7
38. 2
39. 1
40. 2

.5

10.
11.
12.
13.
14.
15.
16.
17.
18.
19.

20.

(DOOSICDU'I

NNl—‘NNH

l—'

NNNNN

103

Scoring Key for Identifying the

Sex of the Infant

21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.

40.

NNNNNNNNH

...:

104

Mother—, Father-Infant Speech Study

Part 11

Directions: Please read the descriptions of each of the
categories below and on the basis of your judgment as to
which parent was speaking, select the category which best
describes the reason for your choice and place an X in the
appropriate column on the attached answer sheet. Similarly,
based on your judgment as to the sex of the infant, select
the category which best describes the reason for your choice.
CHOOSE ONLY ONE CATEGORY FOR YOUR JUDGMENT AS TO WHICH PARENT
WAS SPEAKING AND ONE CATEGORY FOR THE SEX OF THE INFANT FOR
EACH OF THE EXCERPTS.

 

Categories

 

I. Male/Female Differences as they relate to:

A. Sex-role function--in particular, division of labor
by sex. Example: You might believe that one parent
is more likely to do the caretaking of the infant
more than the other parent.

 

B. Personal attributes, behaviors, anatomy, dress, etc.
Example: You might believe that one sex expresses
feelings more than the other sex.

 

C. Occupation
Example: You might associate certain occupations
with one sex more than the other sex.

 

11. Language

A. The words used seem distinctly masculine or femi-
nine. Example: You might believe that certain
adjectives such as pretty, sweetheart, toughie,
etc. are used by one sex more than the other or
that certain adjectives are used to describe one
sex more than the other.

B. The structure of the sentence(s)
Example: You might believe that one sex makes
statements that sound like questions, or that cer-
tain phrases sound more like those used by one sex
more than the other.

III.

IV.

105

Context of the situation--Something about the setting
or assumptions you made about the setting helped you
in making a judgment about the sex of the speaker or
the sex of the infant. Example: You might believe
that because a particular time of the day was alluded
to in the excerpt, that one sex would be more apt to
be home at that time than the other sex.

Other-—This category is to be used when your reason
does not fit into any of the above categories. If

you check this on your answer sheet, please state below
it what your reason is.

106

Answer Sheet for Mother-iiFather-Infant Speech Study
Part II

Example: If the reason you decided the mother was speaking
had to do with a task that was mentioned, you would put a
checkmark under Category I. If you thought the infant was

a girl based on the occupation that was mentioned, you would
check Category I again beside "infant.” PLEASE REMEMBER
THAT IF YOU CHECK CATEGORY IV--OTHER, YOU MUST NOTE BELOW
WHAT YOUR REASON IS.

 

(hwegmw'I
Male/Female Category 11 Category 111 Category IV

*
Differences Language Context Other

 

Hmmmt
Infinm
Paman
Infiun
Paman
hflhnt
Pamxm
Inflnn
Pamxm
Inflnn
IRnent
Influn
IRuent
Influm
Ifiuent
Inflnm
IRuent
Influn
Pamam
Infiun
Panxm
hflhnt

*Other:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10

 

 

11

 

 

 

 

 

107

 

(hmegmw'l
Male/Female
Differences

Cahxgny'll
Iangxge

Category III
Omflext

(kmegmerV
Other*

 

Pwmmt

 

Inflnn

 

13

Pamam

 

Infinm

 

l4

IRnent

 

Infinn

 

15

Pamxm

 

Infinfi

 

l6

Rmmmt

 

Influm

 

 

 

 

 

*Other:

BI BLIOGRAPHY

108

NOTES

1. Enzer, N. Personal communication, August 20, 1979.

109

BIBLIOGRAPHY

Anastasi, A. Psychological testing. Toronto, Ontario:
The Macmillan Company, 1968.

 

Belsky, J. The effects of context on mother-child inter-
action: A complex issue. The Quarterly Newsletter of
the Laboratory of Comparative Human Cognition, 1971,
1(3), 29-32.

 

 

Bem, S. The measurement of psychological androgyny.
Journal of Consulting and Clinical Psychology, 1974,
‘4g(2), 155—162.

 

Bem, S. On the utility of alternative procedures for
assessing psychological androgyny. Journal of Con-
sulting and Clinical Psychology, 1977, 35(2), 196-205.

 

 

Bem, S., & Lenney, E. Sex typing and the avoidance of
cross-sex behavior. Journal of Personality and Social
Psychology, 1976, 33(1), 48—54.

 

Bodine, A. Sex differentiation in language. In B. Thorne
& N. Henley (Eds.), Language and sex: Difference and
dominance. Rowley, Mass.: Newbury House Publishers,
1975.

 

Brend, R. Male-female intonation patterns in American
English. In B. Thorne & N. Henley (Eds.), Language
and sex: Difference and dominance. Rowley, Mass.:
Newbury House Publishers, 1975.

Broen, P. The verbal environment of the language-learning
child. American Speech and Hearing Association Mono-
graph, 1972, 11.

Brooks-Gunn, J., & Matthews, W. S. He & she. Englewood
Cliffs,IiuJ.: Prentice-Hall, Inc., 1971.

Cherry, L. Teacher-child verbal interaction: An approach
to the study of sex differences. In B. Thorne &
N. Henley (Eds.), Language and sex: Difference and
dominance. Rowley, Mass.: Newbury House Publishers,
1975.

 

Cherry, L., & Lewis, M. Mothers and two-year-olds: A study
of sex-differentiated aspects of verbal interactions.
Developmental Psychology, 1976, 12(4), 278-282.

110

111

Clarke-Stewart, K. A. Interaction between mothers and
their young children: Characteristics and consequences.
Monographs of the Society for Research in Child
Development, 1973, 33(6—7, Serial No. 153).

Cohen, S. E., & Beckwith, L. Maternal language in infancy.
Developmental Psychology, 1976, 13(4), 371-373.

Conklin, N. F. Toward a feminist analysis of linguistic
behavior. The University of Michigan Papers in

1

Women's Studies, 1974, 1(1), 51—73.

Downey, P. The invisible woman--Where is she? Graduate
Woman, March/April 1979, 8-13.

Emmerich, W. Variations in the parent role as a function
of the parent's sex and the child's sex and age.
Merrill Palmer Quarterly, 1962, 3, 3-11.

 

Farwell, C. B. The language spoken to children. Human
Development, 1975, 13(4), 288-309.

 

Fein, R. A. Consideration of men's experiences and the
birth of a first child. In W. Miller & L. Newman
(Eds.), The first child and family formation. North
Carolina: Carolina Population Center, 1978.

 

Fillmer, H. T., & Haswell, L. Sex-role stereotyping in
English usage. Sex Roles, 1977, 3(3), 257-263.

 

Friedlander, B. 2., Jacobs, A. C., Davis, B. B., & Wetstone,
H. D. Time-sampling analysis of infants' natural lan-

guage environments in the home. Child Development,
1972, 33, 730-740.

 

Garcia-Zamor, M. Child awareness of sex-role distinctions
in language use. Paper presented at Linguistic
Society of America, December 1973.

Gilley, H. M., & Summers, C. S. Sex differences in the use
of hostile verbs. Journal of Psychology, 1970, 13,
33-37.

 

Gleason, J. B. Code switching in children's language. In
Timothy E. Moore (Ed.), Cognitive development and the
acquisition of language: New York: Academic Press,
1973.

 

 

112

Gleser, G. C., Gottschalk, L. A., & Walkins, J. The rela—
tionship of sex and intelligence to choice of words:
A normative study of verbal behavior. Journal of
Clinical Psychology, 1959, 13, 182-191.

 

 

Gould, L. W. S: A fabulous child's story. Mg, December
1972, 74-76, 105-106.

Greenberg, S., & Formanek, R. Social class difference in
spontaneous verbal interaction. Child Study Journal,
1974, 3(3), 145—153.

 

Haas, M. Men's and women's speech in Koasati. In D. Hymes
(Ed.), Language in culture and society. New York:
Harper & Row, 1964.

 

Heider, F. Perceiving the other person. In R. Tagiuri &
L. Petrullo (Eds.), Person_perceppion and interper-
sonal behavior. Stanford, Calif.: Stanford University
Press, 1958.

 

 

Hirschman, L. Female-male differences in conversational
interaction. Paper presented at the Linguistic Society
of America, December 1973.

Hollingshead, A. B. Two factor index of socialgposition.
New Haven, Conn.: Department of Sociology, Yale Uni—
versity, 1957.

 

Holzman, M. The verbal environment provided by mothers
for their very young children. Merrill-Palmer Quar-
terly, 1974, 33(1), 31-42.

Jespersen, 0. Language: Its nature, development and origin.
London: George Allen & Unwin, Ltd., 1922.

Key, M. R. Linguistic behavior of male and female.
Linguistics, 1972, 33, 15-31.

 

Korner, A. The effect of the infant's state, level of
arousal, sex and ontogenetic stage on the caregiver.
In M. Lewis & L. Rosenblum (Eds.), The effgct of the
infant on its caregiver. New York: John Wiley &
Sons, Inc., 1974.

 

Kramer, C. Actual and perceived sex-linked differences
in adverb and adjective usage. Unpublished paper
presented at meetings of the Linguistic Society of
America, 1973.

113

Kramer, C. Women's speech: Separate but unequal? Quarterly
Journal of Speech, April 1974, 14-24.

 

 

Kramer, C. Folklinguistics. Psychology Today, June 1974,
8, 82-85.

 

Lakoff, R. Language and woman's place. Language in
Society, 1974, 3, 45-80.

 

Lamb, M. E. (Ed.). The role of the father in child develop-
ment. New York: John Wiley & Sons, Inc., 1976.

Lamb, M. E. Father-infant and mother-infant interaction in
the first year of life. Child Development, 1977, 33,
167-181.

 

Lewis, M. Parents and children: Sex—role development.
School Review, 1972, 33, 229-240.

 

Lewis, M., & Weinraub, M. Sex of parent x sex of child:
Socioemotional development. In R. Friedman,
R. Richart, & R. VanderWiele (Eds.), Sex differences
in behavior. New York: John Wiley & Sons, 1974.

 

Lynn, D. B. The father: His role in child development.
Monterey, Calif.: Brooks/Cole Publishing Co., 1974.

Maccoby, E. E. (Ed.). The development of sex differences.
Stanford: Stanford University Press, 1966.

Maccoby, E. E., & Jacklin, C. N. The psychology of sex
differences. Stanford: Stanford University Press,
1974.

 

Macfarlane, A. The psychology of childbirth. Cambridge,
Mass.: Harvard University Press, 1977.

Moss, H. A. Sex, age, and state as determinants of mother-
infant interaction. Merrill-Palmer Quarterly, 1967,
13, 19-36.

Moss, H. A. Early sex differences and mother-infant inter-
action. In R. Friedman, R. Richart, & R. VanderWiele
(Eds.), Sex differences in behavior. New York: John
Wiley & Sons, 1974.

O'Leary, S. E. Mother-father-infant interaction in the
first two days of life. Unpublished doctoral disser—
tation, University of Wisconsin, 1974.

 

114

Osofsky, J. D., & O'Connell, E. J. Parent—child interac-
tions: Daughter's effects upon mother's and father's
behaviors. Developmental Psychology, 1972, 7(2),
157—168. "

Parke, R. D., & O'Leary, S. Father-mother—infant interac-
tion in the newborn period: Some findings, some obser—
vations, and some unresolved issues. In K. Riegel &
J. Meacham (Eds.), The developing individual in a
changing world, Vol. 2. The Hague: Mouton, 1975.

Parke, R. D., O'Leary, S. E., & West, S. Mother-father-
newborn interaction: Effects of maternal medication,
labor, and sex of infant. From Proceedings, 80th
Annual Convention, American Psychological Association,
1972.

 

 

Pedersen, F. Relationships between paternal behavior and
mother-infant interaction. Paper presented at the
annual meeting of the American Psychological Associa—
tion, Chicago, 1975.

Rebelsky, F., & Hanks, C. Father's verbal interaction with
infants in the first three months of life. Child
Development, 1971, 33, 63—68.

 

Rothbart, M., & Maccoby, E. Parents' differential reac—
tions to sons and daughters. Journal of Personality
and Social Psychology, 1966, 3(3), 237-243.

 

 

Rubin, J., Provenzano, F., & Luria, Z. The eye of the
beholder: Parents' views on sex of newborns. American
Journal of Orthopsychiatry, 1974, 33(4), 512-519.

 

Sachs, J. Cues to the identification of sex in children's
speech. In B. Thorne & N. Henley (Eds.), Language
and sex: Difference and dominance. Rowley, Mass.:
Newbury House Publishers, Inc., 1975.

 

Sachs, J., Lieberman, P., & Erickson, D. Anatomical and
cultural determinants of male and female speech. In
R. W. Shuy & R. W. Fasold (Eds.), Language attitudes:
Current trends and prospects. Washington, D.C.:
Georgetown University Press, 1973.

 

 

Sander, L. W., Stechler, G., Burns, P., & Julia, H. Early
mother-infant interaction and 24-hour patterns of

activity and sleep. Journal of American Academy of
Child Psychiatry, 1970, 3, 103-123.

 

115

Saskin, W. F., & John, V. P. The study of spontaneous talk.
In R. Barker (Ed.), The stream of behavior. New York:
Appleton-Century-Crofts, 1963.

 

Sears, R., Rau, L., & Alpert, R. Sex-typing and gender-
role. In R. Sears, L. Rau, & Alpert, R. (Eds.),
Identification and child rearing. Stanford: Stanford
University Press, 1965.

 

Seavey, C. A., & Katz, P. A. The effect of gender labels
on adult responses to infants. Sex Roles, 1975, 1(2),
103-109.

 

Snow, C. E. The development of conversation between mothers
and babies. Journal of Child Langpage, 1977, 3(1),
1-22.

 

St. Clair, L. A. Some considerations for the study of sex
differences in social behavior. Paper presented at
the Speech Communication Association's Summer Confer-
ence XI, 1975.

Stockard, J., & Johnson, M. Sex Roles. Englewood Cliffs,
N.J.: Prentice—Hall, Inc., 1980.

 

Swacker, M. The sex of the speaker as a sociolinguistic
variable. In B. Thorne & N. Henley (Eds.), Language
and sex: Difference and dominance. Rowley, Mass.:
Newbury Publishers, Inc., 1975.

 

Trevarthan, C. Conversations with a 2-month-old. New
Scientist, May 2, 1974, 230-235.

 

Trudgill, P. Sociolinguistics: An introduction. Middlesex,
England: Penguin Books, Ltd., 1974.

Warshay, D. W. Sex differences in language style. In
C. Safilios-Rothschild (Ed.), Toward a sociology of
women. Lexington, Mass.: Xerox College Publishing,

Wood, M. M. The influence of sex and knowledge of commu-
nication effectiveness on spontaneous speech. Word,
1966, 33(1,2,3), 112—113.

Zalk, S. R., & Katz, P. A. Gender attitudes in children.
Sex Roles, 1978, 3(3), 349-357.

 

Zimmerman, D. H., & West, C. Sex-roles, interruptions, and
silences in conversation. In B. Thorne & N. Henley
(Eds.), Language and sex: Differeppe and dominance.
Rowley, Mass.: Newbury House, 1975.

MIMEI“