Citation
The role of education in the use of masculine generics

Material Information

Title:
The role of education in the use of masculine generics
Creator:
Harvey, Vickie L
Publication Date:
Language:
English
Physical Description:
vii, 117 leaves : forms ; 29 cm

Subjects

Subjects / Keywords:
English language -- Gender ( lcsh )
Genericalness (Linguistics) ( lcsh )
English language -- Gender ( fast )
Genericalness (Linguistics) ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 114-117).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Department of Communication.
Statement of Responsibility:
by Vickie L. Harvey.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
22879554 ( OCLC )
ocm22879554
Classification:
LD1190.L48 1990m .H37 ( lcc )

Downloads

This item has the following downloads:


Full Text
THE ROLE OF EDUCATION IN THE USE OF MASCULINE GENERICS
by
Vickie L. Harvey
B.A., University of Colorado, 1988
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Arts
Department of Communication
1990


This thesis for the Master of Arts
degree by
Vickie L. Harvey
has been approved for the
School of
Communication
by
Michael Monsour
Date


Harvey, Vickie L. ( M.A., Speech Communication)
The Role of Education in the Use of Masculine Generics
Thesis directed by Michael Monsour, Assistant Professor
The question under study concerns perception of generic
terms such as "man," man related terms with "man" as a
prefix or suffix and the pronoun "his" and their
relationship to one's educational level and gender. Two
independent variables, gender and education level were
correlated to five dependent variables of generic terms.
This study is designed to examine subjects' use and
understanding of generic language. Significant
statistical support was found for sentences that contained
both the suffix "man" and pronoun "his" together. These
results found that graduate students rated the terms more
masculine than undergraduates and both groups rated the
terms less generic overall. Interaction effects on the
"pre" variable were marginally supported. While the two
hypothesis were not completely supported, the results
provide minor support for the often cited claim that the
use of generic terms lead to increased thoughts of men.
The form and content of this abstract are approved. I
recommend its publication.
Signed
Michael Monsour
iii


CONTENTS
CHAPTER
1. INTRODUCTION ................................ 1
Rationale ..... .............................. 2
Problem.......................................11
2. BACKGROUND AND REVIEW OF LITERATURE .... 21
Overview......................................22
Summary.......................................38
A Correlation of Education &
Generic Language ............................ 39
Gender Differences .......................... 42
Purposed Research Questions & Hypotheses .46
3. METHODS........................................48
Measurement Instrument ....................... 49
Independent Variables ..................... 49
Dependent Variables ....................... 50
Subjects.......................................51
Data Collection Procedures ................... 51
Pilot Study.................................51
Survey Procedures ......................... 52
Analysis.......................................53
Coding......................................53
Statistical Analysis ...................... 53
4. RESULTS.................................... 59
Internal Measurement Reliability ............. 60
Pearson Correlation Coefficients .... 60
Reliability Coefficients .................. 61
iv


Hypotheses Testing ........................... 62
Hypothesis One..............................62
Hypothesis Two............................ 65
Analysis of Variance .................. 67
5. DISCUSSION.....................................75
Hypothesis One.................................78
Hypothesis Two.................................85
Interaction Effects .......................... 85
Gender as a Contribution to
Interaction Effects .......................... 88
College Education Contributions to
Interaction Effects .......................... 94
Limitations of Present Study ................ 100
Suggestions for Further Research ............ 101
Review of Conclusions ....................... 103
APPENDIX
A. Questionnaire and cover letters ............. 107
B. Survey........................................110
LIST OF REFERENCES.................................114


TABLES
Table
4.1 Pearson Correlation Coefficients .......... 61
4.2 Reliability Coefficients ................... 62
4.3 Education Results .......................... 64
4.4 Gender Results...............................66
4.5 ANOVA Interaction of Gender &
Educational Level"Pre" Variable ........... 70
4.6 ANOVA Interaction of Gender &
Educational Level"Suffix" Variable ... 70
4.7 ANOVA Interaction of Gender &
Educational Level"Both" Variable .... 71
4.8 ANOVA Interaction of Gender &
Educational Level"Man" Variable .... 71
4.9 ANOVA Interaction of Gender &
Educational Level"His" Variable .... 72
VI


FIGURES
Figure
1.1. Graph AThe ANOVA Male Trend...........67
1.2. Graph BThe ANOVA Female Trend .... 68
1.3. Graph CANOVA "His" Trends ........... 73
1.4. Graph DANOVA "Pre" Trends ........... 88
vii


CHAPTER ONE
INTRODUCTION
Of this remarkable protagonist Man must it not be
said that his capacity to adapt his universe and its
physical laws to his own needs and desires and
purposeshis science and technologyis one of the
qualities of his greatness?
Miller & Swift, 1988.
Are the term "man" and the masculine pronoun "his" being
used generically or is the passage only referring to males?
Past research has demonstrated that generic terms ("man,"
man related terms and "his" or "he") do not function
generically (Martyna, 1978; Hyde, 1984; Hamilton, 1988).
The terms bias the reader or listener toward a male
interpretation of the material rather than toward a male
and female interpretation. The literature reviewed and
research presented in this study look at masculine generics
from a theoretical, historical and analytical perspective.
Later in this chapter the central concern of this research
is presented which examines perception of the generic
terms: "man," man related terms and the pronoun "his" and
their relationship to one's gender and one's level of
college education. The upcoming section discusses the
Whorfian hypothesis in connection with masculine generics.
Linguistic relativity is introduced as a foundation from
which to expand.


Rationale
Whorf (1956) and Sapir (as cited in Whorf, 1956)
conducted intensive research concerning the relationship
between language, thought and social behavior. The
Whorfian Hypothesis, linguistic relativity, provides the
theoretical foundation that states that the "soul of a
people" and the "mind of a people" were not only reflected
by its language but, actually, formed by it (Mandelbaum,
1949, p.162). The language of the real world is
intertwined with one's thoughts and one's behavior.
Language serves as a reflection and reveals the way our
culture projects its image. An example of language
reflecting reality from the 1960's is the creation of the
term Black Power. It was born of anger and resistance by
a group of people who felt misrepresented by the terms
society used to refer to them such as "colored people".
The term Black Power expressed an emotional post-Malcomist
statement of black consciousness and racial pride rather
than a concrete program (Blair, 1977). The newly born
Black movement was endorsed by the term Black Power which
quickly caught on and helped the spread of the movement
beyond the South.
Sapir (1929) refers to language as the "symbolic guide
to culture." Language is probably the greatest force of
socialization that exists. It serves as a symbol of the
social solidarity of those that speak the language (Sapir,
2


1921) The Whorfian Hypothesis linked language to creation
and perpetuation of the reality of those who create
language (traditionally men). They suggest our reality is
built upon our perceptions and that language affects us in
direct and less conscious ways.
Two obvious characteristics of the structure of our
language are grammar and semantics. Semantics is the
foundation from which we devise experience (Sapir, 1929).
In English, semantics and usage use masculine generics to
generalize to females. In this regard, English differs
from other languages. In Spanish all nouns including those
denoting non-living things are either masculine or
feminine. They do not use a masculine or a feminine noun
to include the other gender. These supposedly generic
terms, he, his and him name males as active agents in our
society more than females. It is grammatically correct to
assume from the English language that agents or active
persons are men. The masculine pronouns used generically
may or may not be interpreted as gender neutral by their
audience, but still this usage communicates distinctive
cultural patterns associated with English users' perception
of the world. Language is a major medium for the transfer
and maintenance of any given culture. The way a
civilization expresses itself linguistically is the road
map to that culture (Sapir, 1929).
The implications of language do not stand apart from
3


experience but run parallel to direct experience (Sapir,
1929). Language is at the base of our sense of self. As
speakers of the language we identify with it and are
referents of the language. It is the language itself that
contains the sex bias, not necessarily the purpose of its
speakers. Our psychological experience of language is just
as important as the language itself. This perspective
invites a thorough presentation of the history of the usage
of nouns and pronouns.
Nouns and pronouns play essential roles in our language,
and it is through these words that we may unconsciously
shape our world. Nouns and pronouns may be either
masculine, feminine or neuter. Masculine nouns ("man",
"husband", "mailman", etc.) all refer to the male gender.
Feminine nouns ("woman", "wife", "daughter", etc.) all
refer to the female gender. Neuter nouns ("garden", "sky",
"sweater", etc.) all refer to objects without gender.
Masculine pronouns ("he", "him", or "his") refer to the
male gender. Feminine pronouns ("she", "her", or "hers")
refer to the female gender. Neuter pronouns ("it" or
"its") refer to objects without gender. A masculine noun
or pronoun will generally refer to any male regardless of
race, religion, age, geographical location, time or
occupation. The same applies for the feminine and neutral
nouns and pronouns (Evans & Evans, 1957).
Masculine nouns ("man", "manpower" or "fireman") used
4


generically reflect a language's loyalty and superiority
of the male gender by excluding women in this semantic
structure (Fishman, 1960). The effect of exclusion is the
invisibility of women, invisibility in the sense that women
are not included in the imagery of masculine terms. It has
been empirically shown that women are excluded by the "he"
while men are merely made less prominent by alternatives of
"he or she" or "they" (Martyna, 1978). McConnell-Binet,
Borker, & Furman (1980) suggest this exclusion in language
has been historically unrecognized and only recently is it
being brought up for examination. Following the thought of
the Whorfian Hypothesis, if women are invisible in our
language they are likely to be less visible as active
agents in our society.
If masculine pronouns and nouns are meant to refer only
to males, then why are the terms used generically? One
begins to find formal arguments from English grammarians
attempting to crystalize the male pronouns and nouns as
gender neutral in the 1700's. Bodine (1975) writes about
early grammarians, John Kirby, 1746, among others, who
refer to the masculine gender as the "worthier" gender and
this was the reason given that the masculine terms; "he" and
"man" should be used for women. Bodine continues with a
quote by Ward, 1765 who wrote An Essay on Grammar that
stated that pronouns must agree in gender; "he" must
represent a male; "she," a female; and "it," an object of
5


no sex. Yet another grammarian stated that "he or she"
should not be used because it draws attention to women and
makes them appear in a special league separate from people
(McCawley, 1974, cited in Bodine, 1975).
An act of Parliament in 1850 by grammarians (all male)
legalized the gender specific term "he" to be the
"official" pronoun for convenience' sake and to alleviate
the clumsiness of the phrase "he or she." Now the male
pronoun would include females. They also argued that
"they" was the plural pronoun and could not be used as the
singular form, but "they" continued to be the more widely
used. The striking thing about the male pronouns, "he, his
and him" was that it took 50 years for people to believe
that "he" was a generic in English and it took 20 more
years for people to believe it had always been this way
(Stanley, 1978). There was tremendous resistance by the
users of the language in adopting and believing the
masculine terms were generic. An attack on the language is
essentially an attack on the speakers of that language. It
took 50 years to implement this language change and it
still continues to be a source of debate.
One year after the passage of the Parliament Abbrevation
Bill on June 16,1851, an attempt was made in the House of
Commons to repeal the Act. John Stuart was concerned that
the bill inadvertently gave women rights which they should
not have. A Reform Bill concerning housing and voting
6


rights was written using the masculine pronoun "he" and
this would now give women the right to vote since women
were now included in the masculine terms. The attorney
general claimed that this wording would not allow women to
vote and if it did, so be it. (Baron, 1986).
Twenty-five years after the formalization of "he" into
a generic term, interpretation still remained open in the
legal arena. Lavina Goodell in 1875 was denied admission
to the bar because the statute regulating bar admission
referred to attorneys as "he." Her petition asserted that
the term had been legally changed to include females as
well as males. The court denied her request on the grounds
that "the language of the statute, of itself, confessedly
applied to males only" (Cox and Ray as cited in Hamilton,
Hunter, & Stuart-Smith, 1990). Penalties or duties
written in the masculine form applied to women but
privileges or benefits were upheld by the courts as only
inclusive of males. The ambiguity of the term continued to
be open to discussion and interpretation.
The Dictionary of Contemporary American Usage (Evans and
Evans, 1957) defines the masculine pronouns "he", "his" and
"him" in natural English as always referring to males.
Theoretically "he" can refer to either a male or a female
when the sex is unknown. In an example, "Each student must
turn in his test," "his" is not being generically used
because the sex of the student is known. There are boys
7


and girls in the classroom not unknown persons. The
Parliament Act of 1850 was used to shorten the language so
that all masculine pronouns ("he", "his", "him") would be
understood to include females. This Parliamentary Act or
all the grammar rules in the world can not change the fact
that, if we are told "somebody telephoned while you were
out," we will say "did they leave a message?" In natural
English, the words "they," "them," and "their" are all used
more often than "he," "his" or "him" when referring to an
unknown person who may possibly be a woman. "He" is also
used 10 to 20 times more often to refer only to males than
to both males and females (Hughes & Casey, 1986).
The ambiguity of these generic words lends itself to
women's invisibility in their meaning. Ambiguous is
defined by Oxford American Dictionary (1980) as having two
or more possible meanings: "doubtful, uncertain."
The word "man" and pronouns "he" and "his" have had to
do a double semantic duty. They have served both gender
specific and gender neutral roles. The generic uses of
"he," "his" and "man" have no set contexts to resolve their
ambiguity (Moulton, Robinson, & Elias, 1978). There is no
sentence structure or usage that identifies a generic term
as generic. It is only when a feminine term, "woman" is
used with a masculine term, "man" or "he" that the generic
term automatically become masculine in meaning. There is
no rule that states when a masculine term is used without
8


a feminine term it is automatically generic. Generic terms
frequently need further clarification to resolve their
intended meaning as either generic or gender specific. All
readers must read further when the term "man" is used to
see if the generic use or the gender specific use is
intended. However, men automatically know they are
included in these masculine words while women must search
for the intended meaning to find themselves either included
or excluded. A sentence referring to man and his destiny
may start out generically and as the sentence progresses
become gender specific, "part of man's happiness is
achieved through his work, environment, wife and children."
Suddenly the generic "man" which woman was originally
included in now causes woman to vanish from the meaning of
the sentence and the minds of its readers. Ambiguity
itself can be resolved by adding detailed information or by
reading or listening further for gender intention. While
this may be inconvenient, time consuming and biased toward
masculine language, the real problem concerns the nature of
true generic language.
Generic terms may fail as gender neutral terms regardless
of intentions and efforts of speakers to use them neutrally
and the terms may accidentally not be applied to neutral
situations even when speakers provide explicit information
that the terms are neutral (Moulton et al., 1978). The
problem arises because the terms are not truly generic. An
9


authentic generic term can be used in different contexts to
denote the same meaning. The term "person" can be used
generically in two ways: "she is the person I hired" or "he
is the person I hired." The noun "person" easily changes
gender with the feminine and masculine pronoun used. This
is how one may tell if a word is generic. The terms "he"
and "man" are not interchangeable with feminine nouns and
pronouns: "he is the prettiest girl here" or "she is our
new weatherman." If "he" and "man" are genuinely gender
neutral, then they ought to be applicable to any person
regardless of gender (Moulton et al., 1978). A term that
is used to represent a class or group that does not apply
to all of its members is a false generic.
In the upcoming section two problems are connected to
generic language. The first one concerns women's
invisibility in generic terms. The most obvious examples
are when the masculine form is used as a generic form, as
in "policeman," "mailman," "the average man," or "mankind."
Linqusitics call the masculine form the unmarked while the
feminine is the marked since an additional marker must be
used to denote a female. An unmarked word is the more
natural one and is closely associated with the norm, while
the marked word is viewed as the exception and needs more
information to differentiate it from the natural state
(Salter, 1979). The second problem concerns the
reinforcement of sex role stereotyping. Sex role
10


AB0M ASI 7465


stereotyping of professions does have certain linquistic
consequences. For example, when referring to the chairman,
that person has traditionally been a male but with more
women in higher social positions it is common for the chair
to be a female. The same can be said of secretaries or
elementary teachers being primarily females in the past but
present positions of these jobs have had an increase of
males employed in them.
Problem
Generic language does not accurately represent women,
which leads to two connected problems: 1) women are made
invisible by generic language which implies a male is
responsible for the job or situation and this affects their
self-esteem and identity; 2) generic language reinforces
sex role stereotyping.
Generic language makes women invisible by implying that
a male is responsible for the job or situation (Henley,
Gruber & Lerner, 1988). While ambiguity of the terms imply
a confusion of meaning, the male image is the first image
brought to mind by generic language (Hamilton, 1988;
Stanley, 1978) A job title that includes a masculine term
such as "newsman," "congressman" or "chairman" implies that
a male is employed in that position. The comparative
visibility of women and men in such job titles raises the
question of discrimination (Ward, 1975). Men are much more
visible in these terms than women regardless of the gender
11


of the jobholder. A situation that refers to the lack of
manpower available or the manhours it takes to complete a
task also reflects women's invisibility and implies man's
responsibility for the situation. The words "manpower" and
"manhours" deny women's efforts while highlighting men's
involvement. Women's roles are not accurately reflected in
generic language. A woman may be part of the "manpower" in
a project or work "manhours" to accomplish a feat and not
be acknowledged for it in masculine or generic language.
The implication would be that men participated in the
project rather than women. There is also empirical
evidence that supports the disregard for female's ability
to perform the job (Hyde, 1984). If women are to be full
participants in our society, our language needs to reflect
their ability and involvement. Not only is generic
language inaccurate, but it also leads the listener/reader
into perceiving the person as male rather than as male or
female (Martyna, 1978; Stericker, 1981).
This inaccuracy and misperception of women in our
communication lead to an increased awareness of masculine
generics as sexist language. It is empirically unknown for
certain what the extent of those effects are on women and
communication. Two effects of women's invisibility in
generic language that receive frequent publication are lack
of self-esteem and uncertain identity.
Women do not view themselves in generic language and when
12


they are referred to as men this has deep effects on their
self-esteem. This decrease in self-esteem causes a decrease
in communication due to lack of confidence (Martyna, 1978;
Henley, Gruber, & Lerner, 1985). Henley et al. (1985)
found boys had more positive change in self-esteem using
masculine generics and that girls had more positive change
in self-esteem using neutral pronouns. Empirical evidence
points out that pronouns have significant effect upon self-
esteem on both females and males. The masculine pronouns
and nouns can lead to lower self-esteem as compared with
unbiased alternatives for females ("he" or "she" or
"humans") (Henley et al., 1985). Inequality in
communication and in society is further perpetrated by
sexist language. Benjamin Spock (1970) agrees that by
using the male pronoun little harm is seemingly done but
when language is looked at in its entirety, its impact at
keeping women at a disadvantage is potentially enormous.
Equality needs to be represented in the language of the
people who seek equality. One's identity in communication
is also linked to one's culture's language.
Generic language affects identity in communication.
Identity is the sense of self one has. Identity combines
a person's social role and true self. Language has a
strong impact on the development of girls and women, and
this in turn influences their identity. Women's
invisibility in language and communication patterns
13


identify women as the less active agent in society. When
a woman is identified through a masculine identity
(masculine terms) rather than having her own feminine
reference or a unity of language of neutral terms, she
tends to view her identity as one of lesser status
(Irigaray, 1985). Masculine language is accepted as the
norm. Male is equated with being human while female is
equated only with women (Silveira, 1980; Enders-
Dragaesser, 1988). The masculine has become the norm in
our communication and with it women's identity as
peripheral and invisible. A woman may be referred to as
one of mankind, working as a repairman while attending
college part-time as a freshman. A woman's own verbal
interactions will deny her own participation in her life.
Women learn that their identity in communication is not as
visible or as vital as men's.
If visibility is a key factor to influence how a person
is represented and ranked in society, women's identity is
clearly underrated and subordinate to men's. Women's
identity in generic language is omitted and makes them
linguistic strangers in a world they are not a part of
simply because they are female (Enders-Dragaesser, 1988).
This identification connection is supported by memory
recall research done by Crawford and English (1984).
Memory recall has been investigated in connection with
generic terms. Crawford and English (1984) tested 50
14


females and 28 male college students and predicted male
students would remember material better that was learned
with "generic" masculine language and females would
remember better with material that specifically includes
them ("he" or "she," or "they"). These hypotheses were
supported. The male good learners were able to recall
better than the female good learners (good in comparison
to bad learners in the study) after having read an essay
using "he" pronouns. The females recalled more than the
males on the "he or she" and "they" pronoun essays. The
good learners were more likely to interpret the material
literally rather than interpret it in a generic form as did
the poor learners. If women can identify themselves in
language, they are more likely to remember the material
than if they cannot associate themselves with or in the
language. It was shown that sensitivity to gender
references affects performance in a memory task. The
readers of both genders literally interpreted the generic
terms.
One's identity is grounded in words, and the dependency
of people upon words works toward a semantic result binding
language and reality. The standard language allows that
women may be in the real world but does not allow the
generic pronoun or noun to reflect this reality. Whorf's
principle of linguistic relativity states that the
structure of language affects our perception of reality,
15


and reality reflects itself in our language. Language is
an objective reality by which people structure and organize
their external world through and because of our
communication (Fishman, 1960). The thought process of
communication directly relates words with an outward
representation (Whorf, 1956). Women's identity is not
communicated in generic language. Females may be less able
to project themselves visually into a sentence when "he" is
used to refer to the antecedent than are males, thus
causing women to have less identification and more
exclusion in generic language (MacKay & Fulkerson, 1979;
Silveira, 1980).
The fact that generic language does not accurately
represent women leads to the second stated problem.
Generic language reinforces sex role stereotyping. A
stereotype "denotes beliefs about classes of individuals,
groups or objects which are preconceived, i.e. resulting
not from fresh appraisals of each phenomenon but from
routinized habits of judgement and expectation" (A
Dictionary of the Social Sciences. 1964). Attitudes that
result in oversimplifications of experiences or judgements
are commonly referred to as stereotypes. The danger in
stereotypes is due to a premature generalization which may
distort attributes of the external world. Stereotyped
thinking dramatically interferes with the perception of
women in our society. Generic language stereotypes
16


individuals into roles without regard to their gender. Sex
role stereotyping reinforces traditional role behavior in
areas such as employment. Women are discouraged from
entering or thinking of entering jobs that have masculine
titles such as cameraman, flagman or weatherman.
As children we are taught to make sex-category
identifications; e.g., "policemen" are not women so girls
may not grow up to think they can be police officers due
to the job title. A mother who agreed to play "fire
department" with her young son provided that she could be
a firewoman was told by her child that she could not be a
firewoman because there was no such thing (Farguhar, Dunn,
& Burr, 1972). His books about fire trucks only talked
about firemen riding them. Every child knows that only men
are firemen. Children are educated by their learning
materials in and out of formal schooling.
Research indicates women may not be entering fields of
work based on sex-biased job titles. Bern and Bern (1973)
studied 60 women and 60 men and found job advertising
language greatly affected the women and men who would apply
for which jobs. Bern and Bern examined sex-biased ads and
recruiting brochures using "telephone framerfian" and
"telephone lineman" (the terms "he" and "craftsman" were
also included in the job description) and found the
percentage of men and women interested in applying for
either opposite-sex job varied greatly with each altered
17


ad. The sex-biased structure (telephone frameman and
lineman) produced no more than 5% female interest. When
the same job was advertised in an unbiased structure
(telephone frameworker and lineworker), 25% of the women
surveyed were interested in the job. When this same ad was
written in feminine terms (frame-woman and linewoman) the
percentage rose to 45% of women surveyed. Jobs written in
the feminine structure discouraged men from applying and
when the ads were written in a sex unbiased or masculine
sex biased structure men's interest rose from 30% to 75%.
These results clearly indicate that jobs with masculine
titles serve to aid and abet discrimination by discouraging
women from applying for "opposite-sex" jobs merely by their
masculine termed description.
Briere and Lanktree (1983) researched 72 female and 57
male subjects on psychology as a future career for women
and men using the APA "Ethical Standards of Psychologists"
text with three different versions: the uncorrected 1972
version using generic "man"; version 2 in which "he" was
replaced with "he" or "she"; and version 3 which used "she"
or "he" in place of "he." Subjects rated psychology from
all 3 conditions on its appeal as a future career for women
and men using a 7 point Likert scale. Version 1
(exclusively male nouns and pronouns) rated a career in
psychology as less attractive for women than did either
version 2 ("he" or "she") or version 3 ("she" or "he") The
18


only different result of version 2 and 3 were when the
feminine pronoun was shown first in version 3 ("she" or
"he") the males were less likely to suggest to a male
friend that he go to a psychologist. Generic masculine
nouns and pronouns were connected with decreases in
attractiveness for females but not for males. Even mild
sex-biased wording affects subjects' perceptions of
employment in psychology for women.
Benoit and Shell (1985) discovered sex-biased
communication can affect business students' career
decisions. Using ACT'S "Assessment of Career Development"
(ACD) instrument, questions referred to a person,
unidentified by sex, in a work situation. There were both
a neutral and a sex-biased version. The test was to
examine if sex-biased communication had an influence upon
career choice. The results showed that females picked 60
correct answers in the neutral condition and 37 correct in
the biased condition. Males picked 78% correct in the
neutral condition, but only 47% correct in the sex-biased
condition. Males were unwilling to place a female (Mrs.
Jarvis) in a nontraditional occupation (i.e. plumber).
Subjects were more likely to choose jobs that were
traditionally held by their own sex rather than
nontraditional jobs. The author encourages educators to
examine their teaching materials for sexist wording so all
students can consider careers without regard to one's sex.
19


Language can lend itself not only to employment problems
but as the research indicates also communication problems.
Language is an important tool to understanding both
historical problems and human communication problems
(Sapir, 1929).
In summary, this chapter has examined linguistic
relativity as a means of linking language to our
perceptions of reality. A historical view of masculine
pronouns and nouns has been analyzed in terms of their
original etymology. Due to changing times the language
itself now contains sex bias, not necessarily in the
purpose of the speaker. The ambiguity and lack of
authenticity of generic terms lends itself to two problems:
1) women's invisibility in generic language has negative
effects on self-esteem and identity; 2) generic language
reinforces sex role stereotyping.
20


CHAPTER TWO
BACKGROUND AND REVIEW OF LITERATURE
A goal of communication research is the identification
and description of variables which have an effect on the
communication process. Much research has been completed
on generic language (Kidd, 1971; Martyna,1978; Hamilton,
1988). Researchers in speech communication have examined
models from social psychology, linguistics and other
disciplines to isolate variables which explain differences
in listener perception and speaker intention. The rapidly
expanding investigation of generic language reveals its
importance in the communication process. Since the English
language is central to our communication process and
culture, this literature review examines studies that have
English as its dominate language. While some foreign
languages have neutral pronouns, the English language does
not and it is this language usage in examination. Other
studies have been omitted due to their investigation design
that was not in line with the focus of this study. This
includes those studies that did not find statistically
significant results or where the design or result was not
free from bias. It is beyond the scope of this review to
comment on all theories, findings, arguments and counter-
arguments concerning generic terms in American culture and


language. This review discusses those studies which are
characterized by empirical soundnessi.e., where
acceptance of social science methods, analysis and design
were employed.
Overview
The background and review of the literature surveys five
main areas of research: research conducted on 1) younger
children (kindergarten through high school), 2) college
students, 3) faculty members, 4) an argument for a direct
link between correlation of education and generic language,
and 5) gender differences. I have separated the first
three areas into divisions by educational level and type of
generic studied ("he," "man" or both). The reviews begin
with kindergarten and elementary school, progress to high
school, college and then faculty members. They include
written and oral methods of investigation. Studies
examined concern kindergarten and first grade students
response to the generic "he" (Fisk, 1985); grade school
children's response to the generic "man" (Harrison &
Passero, 1975) ; kindergarten through college level students
response to the generic "he" (Martyna, 1978; Hyde, 1984);
high school students response to biased job advertising
(Bern & Bern, 1973); college students response to the generic
"he" (Kidd, 1971; Moulton et al., 1978; Murdock & Forsyth,
1980; Hamilton, 1988; Hamilton, Hunter, & Stuart-Smith,
1990); college students response to the generic "man"
22


(Schneider & Hacker, 1973); college students response to
the generic "he" and "man" (Jacobson & Insko, 1985); and
faculty members response to the generic "he" and "man"
(Bate, 1978).
The studies are arranged by the earliest date to the
latest date within the division. The research theories
remain relatively constant throughout the 1970's and
1980's. The message concerning the generic pronoun "he"
and the term "man" remain consistent in that they are not
perceived as genuine generics. The concluding section then
progresses to further empirical support which provides
focused data pertinent to the conceptualization between
language and education that lends itself to the research
questions and purposed hypotheses.
The literature review which is organized by education
level of subjects in related studies starts by examining
pronoun use at the first formal level of education
(kindergarten) and advances through the grades, continuing
to faculty members. The review examines both the
correlation between education and generic language and when
appropriate the gender of subjects.
Harrison and Passero (1975) analyzed elementary
children's responses to the generic "man." Children were
shown drawings of stick people (gender identifiable) along
with an explanation of their behavior. The descriptions
were written with neutral terms (people, salesperson and
23


hand-made) or masculine generics (men, salesman, and man-
made) .
The responses to the generics ranged from 49% to 85%
towards a masculine interpretation. It was found that
"man" did not visually include woman 85% of the time in the
statement: "If a man could wish upon a star to make his
dreams come true, do you think he would?" Eighty-five
percent of eight-year-old children presented with these
generic examples circled only male in response.
Harrison and Passero explain their results in an
editorial style, "the extremely high response can be
attributed to our culture's bias toward males." Any
dominating element in the language reflects that culture's
highly valued elements. The authors continue their claim by
noting "that a culture that is male dominated will create
a male dominated language." The supposedly generic terms
"man," "men" and "he" are permitted to include females but
female terms may refer to people only if no possibility
exists that a male might be included in the reference. An
example of this is that a few years back (the early 1980's)
people starting referring to elementary teachers as "she"
since the vast majority of grade school teachers were
females. Male elementary teachers and administrators
strongly objected to this since some of the teachers were
male (less than 10%) But masculine terms are allowed to
include females when the ratio is higher for females than
24


for males. Mankind consists of 52% females and 48% males
but still the term is used to refer to humanity. The term
"man" is also used for many prefixes and suffixes of words.
The authors theorize that the language habits of the
dominant group even if they are less in number create the
language for the entire group (culture).
Martyna (1978) conducted research on sentence fragments
using 400 students of all ages from kindergarten through
college level. Sentences in a written format were
completed concerning male related roles (i.e., engineer),
female related roles (i.e., secretary) and neutral roles
(i.e., teenager) in connection with pronoun choice. Many
filler topics were used (sports and weather) so attention
would be focused away from generic pronouns. The subjects
were questioned afterwards for imagery responses. Both
sexes reported receiving imagery of males in the male-
related topics and imagery of females in the female-
related topics. A different trend of imagery appeared in
the neutral sentence subjects. Sixty percent of male
college students reported a male image when they chose
generic "he" in a neutral sentence (i.e., human being,
teenager, person). Only 10% of female college students
reported any imagery when they chose these same terms and
the imagery reported was 100% male. When females were
questioned afterwards, they replied they had selected "he"
because they'd been trained to or that it was an automatic
25


response. The males replied, "I pictured males probably
because I'm a male" and "I thought of myself."
The pattern in this study leans toward males picturing
male imagery in masculine pronouns and females choosing the
male pronoun based on a grammatical standard of
correctness. Martyna did not report the ratio of females
and males but did indicate a sex difference occurring with
females using generic "he" less often than males regardless
of antecedent.
Hyde (1984) conducted research on 310 subjects ranging
from first graders to college students. There were a total
of 140 males and 170 females. Each child was individually
given information about a character and asked to create a
story about this character. Three different pronouns were
used to describe the character by the experimenter:"he,"
"he" or "she" or "they." The subjects were also asked if
they knew the generic pronoun rule.
Information given about characters using the masculine
pronoun "he" resulted in 88% of the stories being about
males. Not a single first grade boy created a story with
a female character. Third grade boys created no females
characters with the pronoun "he" but females characters
increased to 30% when the alternative "he or she" was used
to originally describe the character.
In response to the standard grammar rule of the
prescriptive "he," knowledge of the rule increased with
26


age: 28% of first graders, 32% of third graders, 42% of
fifth graders, 84% of college students replied they knew
the rule but disagree with it '"yes, he refers to males
so one should use "he" or "she" to be clear that everyone
is included.'"
A recent investigation by Fisk (1985) found subjects did
not perceive the generic "he" as gender neutral. Subjects
were 36 first graders and 36 kindergarten students. Half
of the subjects were girls and half were boys. Children
were told a story using either "he," "he" or "she" or
"they." Each story's wording was the same with the
exception of the alternative pronouns. The child was asked
to retell the story, and his or her pronoun usage was
recorded. The child was also asked to pick either a boy
picture or a girl picture to represent the character in the
story. The results indicate the children did not respond to
the pronoun "he" as if it were a generic "he."
In the group that received the story with the pronoun
"he," 22 out of 24 subjects retold the story using the
pronoun "he" and picked the boy picture (the other two did
not respond). No sex differences occurred in this
condition. This research corresponds with previous
research in showing male-biased responses to the use of
"he" in what is intended to be a neutral presentation.
The findings for the "they" pronoun produced 22 out of
24 children retelling the story using the "they" pronoun.
27


Of the 22 subjects, seven chose the boy picture, nine chose
the girl picture and six said both. The sex differences
varied greatly in the "they" presentation with all six of
the kindergarten boys choosing boy pictures. The girls in
both levels chose a girl or both picture and only one first
grade boy chose a boy picture while the rest of the first
grade boys chose girl or both pictures.
The "he or she" presentation netted a variety of pronoun
usages in retelling the story; two "he," five "she," four
"he" or "she," eight "they," three mixed. The pictures
chosen were five boy, eleven girl, and six both. No sex
differences occurred in this group. Since only boy and
girl pictures were presented, the both responses were
probably underestimated. During the "he or she" and "they"
presentations, the children looked back and forth at the
available pictures prior to choosing. The subjects who
said both must have been certain of the answer because
there was no both picture to choose. Many of the children
were likely obeying the experimenter and chose one of the
pictures when they may have wanted to pick both. This
response is especially key because socialization has deep
impact after just a few years.
Fisk concluded that by the age of six for girls and the
age of seven for boys, children respond to pronouns in a
learned manner rather than just assigning their own sex to
the subject if the reference is vague (they) or mixed (he
28


or she) Boys until the age of seven could only respond to
the pronoun "he" with a masculine interpretation. It is
also suggested that girls do not go through the same
process as boys in responding to pronoun presentations.
The research on grade school children reveals a pattern:
the younger the child the less likely the child will know
and use generics neutrally. Children tend to use masculine
generics to include only males rather than males and
females. As elementary education continues children are
more apt to learn and use masculine generics neutrally.
This relationship between generic usage and education level
continues through college in the studies just reviewed.
Studies involving only high school and college students
perceptions of masculine generics are now examined.
Kidd (1971) conducted an important first empirical study
on sex bias in language use. Sixty-eight college students
were asked to openly respond to three questions pertaining
to pronouns. 1) Do people respond to generic pronouns as
though the antecedents are neutral? The first question
received an emphatic no in response to people viewing the
generic pronoun "he" as neutral. Subjects responded with
407 male image responses and 31 female image responses to
a visual image of the generic "he." 2) If forced to respond
in terms of gender, do people describe the antecedent as
"male" and "female" in equal proportion? The second
question forced the subject to respond with either a male
29


or female antecedent in a neutral generic sentence. "Male"
was chosen 526 times over "female" being chosen 53 times.
3) If some pronoun usages evoke heavy male responses and
some strong female responses, what consistencies can be
uncovered in the nature of these usages? The third
question concerned societal roles connected with pronouns.
The students responded by choosing masculine images for the
active roles and administratively superior positions.
Women were chosen for the "emotional child, the teacher and
the consumer roles." "Other" as an answer surfaced in the
abstract and unspecific roles. These results coincided
well with traditional stereotypes of the male and female
roles.
As previously discussed, Bern and Bern (1973) examined 120
senior high school students response to booklets containing
sex-biased job advertisements. They found job descriptions
written with masculine generics (telephone lineman or
frameman) produced no more than 5% female interest. The
same job advertisements were changed to include neutral
terms (telephone lineworker or frameworker) and female
interest rose to 25%. When the descriptions were rewritten
using feminine terms (telephone linewoman or framewoman)
women's interest rose even higher to 45%. Clearly these
results emphasize a trend of masculine generics excluding
females.
Schneider and Hacker (1973) tested 306 college students
30


for their interpretation of the generic term "man." No sex
breakdown was given. Students were asked to supply
pictures to represent chapters in an introductory sociology
textbook. Three separate schools complied with this
request. Two dependent variables were employed: one form
presented chapter titles containing the term "man," and the
other form without the term.
All pictures were coded by gender: "males only," "females
only," "both males and females," "no people present," "sex
indistinguishable" and no picture submitted. Overall, 64%
of the students who received the man-linked titles
submitted pictures containing only males. The chapter
titled "industrial man" received the highest number of male
only pictures with a percentage of 87%. Rarely in any of
the titles either man-linked or otherwise wore only women
shown. The results indicate the tendency to perceive males
rather than people or women in the supposedly generic term
"man".
Moulton et al. (1978) tested 226 male and 264 female
college students response to the generic "he." They were
given written instructions to create a character who would
fit in one of the themes given. The themes consisted of
neutral subjects, either a student or a person used in a
sentence with either the pronoun "his," "his" or "her" or
"they." Students were told not to write about themselves.
Results found that when the pronoun "his" was used, 65%
31


of the story characters were male; the "his or her"
instructions netted 44% male characters and the "they"
instructions netted 54% male characters. Students had to
name their characters as well and use a pronoun to describe
them so that it could be clear of which gender the
characters were.
These results continue the trend that the using the
pronoun "his" increases the possibility that people will
think of a male even in contexts that are intended to be
neutral. Generic terms are not necessarily gender neutral
in interpretation.
Research by Murdock and Forsyth (1980) examined reactions
in a naturalistic context by asking 47 men and 87 women
adult students (college level) to examine a written essay
using all plural pronouns, the generic pronoun(s) and
evaluate stereotypical phrasing. They also evaluated
sexism in sentences that either contained no sex bias, the
generic "man" or "he," or referred to women in an unfair,
stereocast manner. Their results showed that generic
phrasing ("businessman" and "he") and stereotyping
(mothering or manhandled) were perceived to be biased and
sexist. The neutral alternatives such as police officer,
the plural pronoun "they" or when man and woman were both
used were the only terms deemed nonsexist.
Jacobson and Insko (1985) surveyed 234 female and 164
male college students for pronoun preferences in
32


accountant or
traditionally masculine occupations (i.e.,
physician), stereotypical feminine positions (i.e.,
secretary or nurse) and titles considered to be generic
(congressman or spokesman). Even though man related terms
have historically been associated with men more than women
i
they are still considered generic. While men gave more
"he" responses than females overall, the biggest finding
concerned the choices made on the generic terms. The
generic terms had the fewest "he/she" choices even over the
masculine items. The researchers concluded that the ending
"man," while it is suppose to refer to both men and women
"apparently induces subjects to expect the individual to be
male" and thus they picked "he" as the correct pronoun for
these terms. The researchers did not report the sex
breakdown but claimed subject sex played a significant
predictor on both masculine and feminine responses. Males
chose "he" more often than females on masculine items and
"she" more often on feminine responses rather than "he or
she" on either items.
Current research by Hamilton (1988) studied masculine
generics in connection with linguistic relativity (that
language can shape thought), prototypicality (the most
common "he" is probably a man) and activation of multiple
meanings ("he" has both gender-specific and gender- neutral
denotations and both may be activated even when the gender-
neutral meaning is intended).
33


Sixty female and 60 male college students participated
in four experimental conditions: traditional vs. modern
instructions, (used to induce uses of "he" vs. "he" or
"she" or "they") crossed with plural vs. singular sentence
fragments. The varying instructions were to measure if
variation in the content of imagery or in name assignment
was caused by the instructions. Subjects were required to
fill in sentences with the pronoun "he," "he" or "she," or
"their." The imagery questions asked for a description of
any image the subject had for the person in the sentence.
This included the persons age, sex, activity, etc. The
subjects also had to think of a name for the person who
would fit in the sentence.
Significant results indicated that people's imagery
excluded females twice as often as it included females and
male names were given two and one half times more often
than female names. Male subjects displayed more male bias
than female subjects. No other interaction effects were
significant.
This study supports other findings that state generic
terms generate more male imagery in the mind of the user
than does using unbiased or true generics (Harrison &
Passero, 1975; Martyna, 1978 & Fisk, 1985). This study
indicates that a speaker may not, on some level, know what
he or she is saying. Someone who is told to hire the "best
man for the job" may be influenced to hire a man. These
34


findings extend themselves to the Whorfian Hypothesis
(Whorf, 1956) Language assists in shaping, allowing and
encouraging certain thoughts while discouraging others.
Masculine generics discourage people from perceiving
females in their imagery. The link between difficulty in
mental inclusion and the exclusion of women in further
thought and action could potentially be made. There is
empirical evidence for this exclusion in sex biased job
advertisements (Bern & Bern 1973) and on the effects of
masculine generics concerning a self-defense verdict in
court (Hamilton, Hunter & Stuart-Smith, 1990).
Hamilton, Hunter, and Stuart-Smith (1990) presented 72
college students with one of three written versions of jury
instructions examining a murder case. The subjects who
read the generic editions of "he" were less likely to
believe that the female defendant took action in self-
defense than those who read one of the alternative terms
conditions of "he" or "she" or "she."
The research conducted on high school students and
college students is consistent with some of the research
done on younger children. However, while younger
children's interpretation of generics increased with each
grade level, college students masculine interpretation
declined or maintained during their college years. High
school female students showed more interest in jobs written
in feminine terms (i.e., linewoman) rather than in generic
35


terms (i.e., lineman). College students that were forced
to choose either a male or female antecedent in response to
the generic pronoun "his" chose male 99% of the time.
College students response to creating a fictional character
produced 65% male characters in response to the generic
pronoun "his." Adult students perceived man related terms
(i.e., businessman) as biased and sexist and only neutral
terms (those without the term "man" in them) were deemed
nonsexist. The college studies examined reveal a positive
correlation between education and one's perception of
generic terms. As one's college education level increased,
masculine perception of generic terms increased. Does this
positive correlation continue beyond college? Research on
faculty members is now examined as the last level of
education in connection with generic language.
Bate (1978) interviewed 10 female and 10 male faculty
members concerning their level of comfort with gender
specific and gender neutral terms (i.e.,chairman and
chair), their pronoun preference in written and oral
contexts and their opinion of generic terms as sexist.
Eight female faculty and nine male faculty picked the
"he" or "she" usage in written form in response to a
sentence fragment "a person...." "They" was more widely
chosen for verbal use. Martyna (1978) found these same
results in her investigation. She also found females used
alternatives to the term "he" more often than males.
36


Bate presented cards with different alternatives on them
for faculty to chose which terms they were most comfortable
using. The term "chairman" received positive support with
six females and eight males choosing this term. The term
"chair" received support from six females and seven males
which made them almost equal in language usage (14 to 13
respectively). The term humanity surpassed mankind 20 to
11 (15 females and 16 males) Many of the faculty felt
they were in transition about the terms. A male faculty
member said he was eliminating the term "man" from his
speech out of respect for his female colleagues. Another
male faculty said he was open to arguments against such
language and would change his speech patterns if his social
setting changed. The faculty members were not as likely to
view the man related terms as sexist as much as limiting
opportunities and affecting self concepts of females.
This study introduces a new perspective on the perception
of generic language. In review, it ws found that the
younger the child the less likely he or she will be to know
generic rules. As education level increases, knowledge and
use of generics become formalized and regulated. This
trend increases until college where it maintains or
declines. While students perceive degrees of inclusion and
exclusion of females no other reactions are noted and it is
only at the post-education level (M.A.,Ph.d. and faculty
levels) that potential for bias enters. Faculty members
37


found that generic terms affect self concepts and limit
opportunities for females. While words appear objective
because they name and define, other ramifications of words
go unnoticed such as self-image and the role modeling
process of language and communication.
Summary
The studies reviewed concerning generic language
demonstrated that "he" and "man" have a masculine
interpretation more often than its intended neutral
interpretation. Studies completed on children drew similar
conclusions (Harrison & Passero, 1975; Martyna, 1978; Hyde,
1984; Fisk, 1985). Kindergarten through elementary school
aged children are more likely to interpret generic terms
with a male antecedent rather than a male and female
antecedent.
Studies done on college students drew similar conclusions
as those done on younger children. As education level
increased, the knowledge of the rule increased and while
84% of college students knew the rule 6% disagreed with its
use and meaning. It was suggested that a speaker may use a
generic term without realizing its bias towards males and
its exclusion of females.
Little research has examined faculty members as subjects.
Bate (1978) did conduct one such study and found that
faculty were more likely to choose "he or she" over the
38


generic "he" in written form and "they" in verbal
interactions. Faculty did not feel comfortable using "a
person ... he" because if failed to include females.
Faculty were not comfortable using "they" in written form
because it was not standard English. However, in
conversation "they" was selected as a modest way to avoid
choosing the gender of a subject.
Bate concluded that many speakers are willing to make
changes to their language habits if they have information
to support its change. Faculty members were also the only
group that perceived the affect of generic terms on
peoples' self-image. This would include either
interpersonal or professional situations warranting change.
The masculine generics controversy continues to increase
people's awareness of the advocacy for change.
Overall, general findings in the literature suggest that
the terms "he" and "man" are not producing neutral
responses throughout all levels of education and beyond.
Substantial numbers of subjects are responding toward a
masculine interpretation instead of a masculine and
feminine interpretation. Research that more directly
correlates education and generic language is examined next.
A Correlation of Education and Generic Language
Though education and generic language have been looked
at indirectly (Fisk, 1985; Harrigan & Lucie, 1988; Hyde,
1984; Hughes & Casey, 1986; Martyna, 1978; Nilsen, 1977),
39


one area that has not been directly examined is the
association between one's level of college education and
the propensity to view the terms, "man," man related terms
and the pronoun "his" as generic. There is logical reason
and empirical evidence to suggest that there should be a
connection.
The logical reason that links one's level of education
and one's understanding of generic language is that
children prior to grade school (under five years of age)
literally interpret generic language ("he" and "man")
towards the masculine meaning (Hyde,1984). The male
pronoun always has a male antecedent for children (he=male)
(Fisk, 1985). This male imagery starts at a young age and
continues through adulthood (there have been no studies
done on children less than 5 years of age) Children
literally interpret terms that refer to the male gender
(he, paperboy, man, fireman, husband, etc.) as having only
masculine meanings. Once children are taught masculine,
feminine and neuter nouns and pronouns then they are taught
the generic use and meaning. This is where confusion
begins. Hyde (1984) found only 28% of first graders knew
and used the grammatical rule of generic "he". Other
results in this study that link education and generic
language use were that 32% of third graders, 42% of fifth
graders and 84% of college students knew and used the
generic "he" but an additional 6% protested against its use
40


even though they knew the rule but disagreed with it for
example, "yes, he refers to males, so one should use "he or
she" to be clear that everyone is included". Martyna
(1978) found that males were more likely to choose male
pronouns due to male imagery while females chose male
pronouns because they had been formally trained (in the
academic environment). As education level increases one's
knowledge and use of generic terms increase due to what one
has been taught in school. If one is taught that generic
terms ("he" and "man") are not gender neutral but gender
specific, one would be less likely to view and use the
terms generically (due to influence of academic learning).
Therefore the more college education one has the less
likely they would be to view terms generically. There is
empirical evidence to support this hypothesis.
Empirical evidence supports the link between education,
generic language and influence of authority. Hughes and
Casey (1986) found this influence of authority (education
in this case) to be responsible for the predominance of
masculine pronoun use and that students who were taught to
avoid "he" or "she" or "he/she" usage were also taught "he"
was a universal or generic pronoun. Harrigan and Lucic
(1988) studied 5 groups: members of a local NOW, faculty
members, medical students, English graduate students and
psychology graduate students. Subjects completed a
questionnaire concerning gender bias in our language.
41


They found that the influence of other people encouraged
pronoun use. The groups most influenced by other people in
authority were the three student groups. It was found that
the student groups were more likely to use a new gender
neutral pronoun when a person in authority made the
request. Females were a bit more influenced by an
authority influence than males. Since students are the
most influenced group, what educators teach in our schools
regarding generic language is definitely worth research.
Teaching alternatives to masculine pronouns and nouns in
education is not only worth the effort but studies show
that what is learned in early education concerning language
rules stays with the student throughout his or her life
(Nilsen, 1977; Henley et al., 1988). Boys as well as girls
need to learn alternative terms so that the terms may be
genuinely inclusive. It is through education that this
language/ communication change can occur. Lastly, research
on gender differences is examined.
Gender Differences
Another area that has been relatively ignored is the
possible association between gender and the propensity to
view the terms "man," man related terms and the pronouns
"he" and "his" as gender neutral rather than gender
specific.
42


Sex differences reported in the studies are logical given
that males and females learn communication patterns
differently. Males learn feminine and masculine terms
(pronouns and nouns) and then use masculine terms as
generic terms. Males may add a female interpretation to
masculine terms to make them generic. Males may also just
continue to use masculine terms without consciously
including females but with the knowledge that masculine
terms are more commonly used as generics rather than
alternative terms (Nilsen, 1977). It is less important that
males know when a term is inclusive of females because
these same terms always include males in any given context.
Generics are only gender neutral in a defined context, out
of that context they retain their original masculine
interpretation. Males can project or image themselves in
generic terms without changing their original meanings.
Females, on the other hand, learn feminine, masculine and
generic terms as three separate units. Females have
learned that masculine terms apply only to males and when
these same terms become gender generic to include them
(females) they must relearn the contextual meanings so that
they are included. They also must learn when to include
themselves and when not to (when the term is used
generically and when it is not). Some empirical evidence
for these learning differences exist (Fisk, 1985; Henley,
1989; Nilsen, 1977). Fisk (1985) found that males and
43


females learn pronoun usage at different ages with girls
learning them at an earlier age than boys.
Empirical evidence concerning sex differences is divided.
Research examining sex differences includes early research
by Martyna (1978) found that 60% of male college students
were able to perceive themselves in generic language while
only 10% of females reported imagery. Men imaged either
themselves and/or males in generic nouns while women had
little imaging of any person at all. Men have learned they
are included in masculine and generic terms but women do
not feel included in either.
Current research by Hamilton (1988) found male subjects
tested using generic terms displayed male bias twice as
often as female subjects. Men were more likely to
visualize themselves in generics than were women to include
themselves in generics.
Benoit and Shell (1985) found males were more likely to
pick correct answers in a sex-biased condition than females
in a sex-biased condition (47 to 37 respectfully). Males
were also more apt to stereotype people into traditional
jobs than females were.
Fisk (1985) found that the younger the male child was
the more likely he would pick a picture of a boy child for
a story using "they" as the only pronoun. Female children
were able to pick female & male pictures but 5 years old
boys could only project themselves or another male self
44


into the story.
Research that has not reported a sex difference or noted
that one did not occur in their study includes research by
Harrison and Passero (1975) on elementary students. Kidd
(1971) tested college students and found more male images
to female images in generic "he" use (407 to 31
respectfully) but did not test for sex differences.
Jacobson and Insko (1985) found that invisibility of women
in man related terms and the term "man" exists but did not
give any breakdown of sex differences. Murdock and Forsyth
(1980) tested reactions to gender biased language
(masculine generics) and found them to be perceived by
subjects as biased and sexist but sex differences were not
reported. Among other studies covered in this paper that
did not report or test for sex differences are Scheider and
Hacker, 1973; Hughes and Casey, 1986; Harrigan and Lucic,
1988. Either the sex differences were not significant
enough or there has not been enough research on them.
Either way, research is lacking in definite support of or
in opposition to gender differences.
This literature review answers some questions concerning
generic language but others remain open. What is the
effect of college education on generic usage? Do male and
female college students differ on their interpretations of
generic language? Is there a preference toward a masculine
interpretation at different levels of college education and
45


if so, at what level? The next section introduces the
research questions and hypotheses proposed.
Proposed Research Questions and Hypotheses
The lack of research on education and generic language
supports the importance of the present research. Generic
language is learned communication that must be learned in
a specific environments Formal education is one such
environment. This leads to the first research question.
Research Question One: What is the connection between post
secondary education and the use and understanding of
generic language?
Hypothesis One: The more college education individuals
have the more likely they will view the terms "man", man
related terms and the pronoun "his" toward a more masculine
interpretation.
There is logical reason and empirical evidence to suggest
that some connection exists between gender and the
propensity to view generic terms as masculine rather than
gender neutral although neither reason nor evidence is
definitive. Gender differences are feasible and this
possibility lends itself to research question two and
hypothesis two.
Research Question Two: What is the connection between
46


one's gender and one's use and understanding of the generic
terms, "man", man related terms and the pronoun "his"?
Hypothesis Two: Males will view the generic terms "man,"
man related terms and the pronoun "his" toward a more
masculine interpretation than females at the same
educational level.
Hypothesis One states that the more college education
one has the more likely one will view the generic terms as
masculine. Hypothesis Two theorizes that males will view
the terms more masculine than females. A combined effect
of the independent variables (college education level and
gender) of the two hypotheses may occur.
47


CHAPTER THREE
METHODS
The questions examined in this study concern perception
of the generic terms: "man," related terms with "man1' as
the prefix or suffix and the pronoun "his" and their
relationship to one's gender and one's level of college
education. This chapter will examine the investigation
methodology including: 1) methods- the survey design 2)
measurement instrument- the independent and dependent
variables 3) subjects 4) data collection procedures- the
pilot study and written survey 5) analysis of methodology-
data coding and statistical procedures chosen.
The choice of research methodology was based on other
researchers traditional use of a written survey in this
area of study. The use of a written survey allowed for a
quicker response time and a greater number of responses
rather than conducting personal interviews. The use of a
survey method is based upon its effectiveness in gaining
focused information about a well defined social phenomena.
The data collected by way of survey enabled information to
be gathered with relative ease from a large number of
people in a defined population (university setting).
A number of limitations and drawbacks are recognized in


the use of a survey method. These limitations include the
following. First, there is no control over unknown
variables affecting the results. Second, the potential
force that social influence may have exerted over
respondents. Generic langauge is more socially accepted
as the norm and this factor may have caused respondents to
rate the terms toward a more neutral intrepretation than
they actually believe. Respondents may have chosen answers
they thought were socially acceptable rather than their own
opinion or perception. Third, the written survey asks
about issues that are lifted from a defined context. In
interpersonal conversations, one is able to assess issue
meanings from the conversation, persons involved and by
asking for further clarification. The survey does not
define or clarify words or issues for the respondent. The
words are not in any given context but in a general
context.
Measurement Instrument
Independent Variables
Variables were measured using a six page questionnaire
(see Appendix A). The independent variables were gender
and college education level. Gender was indicated by a
mark on the line next to the male or female response
options. College education level was indicated by a
written reply or a circle around the printed response
49


options of freshman, sophomore, junior, senior or graduate
student. Freshman, sophomore, junior and senior responses
were combined into the undergraduate category. Faculty
members were given a separate information sheet and asked
to indicate the year their degrees were earned.
Dependent Variables
The five dependent variables consisted of distinct uses
of the masculine generic terms "man" and or "his". The
five dependent variables were labeled "pre, suffix, both,
man and his". The variable "pre" consisted of five terms
with the prefix "man": manpower, mankind, manhood,
manhours, and manned each used in a sentence (see questions
2,11,13,14 and 21 in the survey). The variable "suffix"
consisted of five terms with the suffix "man": baseman,
caveman, policeman, layman, and chairman each used in a
sentence (see questions 4,6,8,9 and 24). The variable
"both" consisted of the five terms foreman, weatherman,
congressman, spokesman, and salesman each used with the
pronoun "his" in the same sentence (see questions
12,16,18,20 and 25). The variable "man" consisted of the
five terms average man, man, working man, dirty old man,
and good men each used in a sentence (see questions
1,7,17,19 and 22). The variable "his" consisted of the
five neutral terms student, customer, student, professor,
and citizen used with the pronoun "his" in the same
sentence (see questions 3, 5,10,15 and 23).
50


Sub~i ects
Respondents for the study were students and faculty
members from the University of Colorado at Denver and
Metropolitan State College. There were 68 undergraduates,
53 graduate students and 57 faculty members who
participated in this research. There were a total of 30
male undergraduates, 27 male graduate students and 29 male
faculty and 38 female undergraduates, 26 female graduate
students and 28 female faculty. The undergraduate and
graduate students were recruited from various communication
and education classes. Faculty members were chosen by the
researcher upon chance visits to faculty and department
offices. An attempt was made to get faculty members with
different educational backgrounds (LAS, business, EE,
etc.). This was an attempt to get faculty from fields
other than the humanities. There was no attempt made to
obtain a racially representative sample. The majority of
the subjects were Caucasians, but there were some
Hispanics, Blacks and Oriental subjects as well. Data was
collected over a three month period.
Data Collection Procedures
Pilot Study
A pilot study was given to a class of 45 undergraduate
and graduate students. The instructions indicated that this
was a probability study dealing with general knowledge.
51


Students were to read the question and indicate to what
degree was its likelihood or probability. Gender was not
mentioned as an operating factor. There were 100 items
with 25 generic usage sentences and 75 fillers provided in
an attempt to mask the generic terms. The researcher asked
the students to give their responses to the true nature of
this survey and found this type of survey failed to mask
the generic sentences. The students guessed the intent to
have something to do with gender and the "man" terms. The
survey was then redesigned using only the 25 generic terms.
In comparing the results of the pilot survey and the final
survey given, no differences were discovered in
respondents1 answers.
Survey Procedures
Respondents completed the final form of the written
survey containing 25 questions given to them by myself or
their professor. Verbal instructions given were only that
this study was for a graduate thesis and concerned gender.
The entire questionnaire packet contained six pages. The
consent form and instructions were given on the first page
of the survey. It was presented as a probability study
dealing with gender. Questions from the survey were
written in this manner: "How likely it it that the term
layman refers to the everyday female?" The questionnaire
used well known man-related terms in an attempt to gain
understanding of individual interpretations of generic
52


terms. It asked the respondents for their personal
opinion. It was also stated that there are no right or
wrong answers. A sample question was presented on page
one. The second page contained personal questions such as
age, sex, level of education, major and other factors to be
correlated with the responses to the questions. Pages
three through six contained the questions concerning
masculine generics. All questions were randomly arranged
in the survey. A 7-point Likert type scale was used to
measure probability or likelihood for each question.
Please see Appendix A for the survey.
Analysis
Coding
The data collected were coded numerically. Sex was coded
1 for male and 2 for female. Level of college education
was coded 1-freshman, 2-sophomore, 3-junior, 4-senior, 5-
graduate and 6-faculty. Dependent variables were coded on
a 7-point Likert scale with 1 being the most gender neutral
and 7 being the most masculine. Non-responses were coded
as 9's.
Statistical Analysis
The credibility of a study rests heavily on the
appropriateness of the statistical analysis used. The use
of the wrong statistical test can lead to false conclusions
and speculation that call the findings into question.
53


Every statistical test has certain conditions for its
appropriate use. A test that allows the researcher to draw
specific conclusions from the data is the more powerful and
therefore more desirable test. The term power used here is
based on the mathematical ability of a statistic to
correctly test a hypothesis. Three considerations that
directly influence one's choice of statistical test are: 1)
sampling methods used 2) nature of the research population
3) level of measurement for the variables. The most
important of these criteria for this study is the level of
interval measurement of variables.
Hypothesis One states that the more college education
one has the more likely one will be to view the generic
terms as masculine. Hypothesis Two states that males will
view the generic terms more masculine than will females.
Both hypotheses used independent groups t-tests employing
education for Hypothesis One and gender for Hypothesis Two.
An analysis of variance (ANOVA) was used to test for main
effects and any interaction effects. The statistical tests
chosen for this study include the following. Pearson
Correlation was used to measure the relationships between
the dependent variables. The reliability test employed
Alpha to test for stability among dependent variables.
The independent groups t-test was used to measure the
difference between group means for dependent variables.
The test was employed 10 times- one for each of the five
54


dependent variables times 2 for the two independent
variables of gender and education. The analysis of
variance (ANOVA) was used to test for main effects and any
interaction of the two independent variables. ANOVA was
employed five times.
The Pearson Correlation was employed to provide a
numerical indication of both the strength and the direction
of any relationship among the five dependent variables.
Correlation coefficients range along a continuum from -1.0
(perfect negative) at one extreme to a +1.0 (perfect
positive) at the other extreme with 0.0 (no correlation) at
the midpoint. The closer the numerical value of the
correlation coefficient is to either extreme the stronger
the relationship between the variables. The closer the
coefficient is to the midpoint the weaker the relationship.
A correlation of 0.0 suggests a lack of a relationship.
The plus or minus sign indicate the direction of the
relationship. A positive relationship is indicated by the
absence of a minus sign while a minus sign indicates a
reversal or negative relationship.
The Pearson R was the logical choice given that early on
in this inquiry it was observed that the five dependent
variables may have constituted only one dependent variable.
The Pearson R was needed to reveal the relationships among
the variables. It was speculated that the dependent
variables were distinct but the extent of their
55


independence was left open to statistical test.
Reliability of measurement is used by social scientists
to test for the degree of stability or generalizability of
empirical findings. The search for reliability raises the
question of how stable the variables can be when observed,
measured, repeated and inferred. The results are
interpreted similar to the Pearson Correlation Coefficients
with +1.0 indicating a strong positive relationship, -1.0
indicating a strong negative relationship and 0.0
indicating no relationship. The reliability coefficient
was attained in this study by comparing each of the
individual items with the total score of the five dependent
variables.
The coefficient interpretation is based on the
assumption that the survey setting was the same for each
instance. The results will be the same when the conditions
are constant. A high coefficient supports this assumption
while a low coefficient denies it. A reliability test was
necessary to determine if the relationships among variables
were due to true differences and how much was due to
inconsistencies in measurement.
The independent groups t-test was chosen to test the
hypotheses because it can be used on a relatively small
sample size, the groups do not have to have an equal number
of cases and it employs interval level data. The t-test is
an analysis of the amount of difference between group
56


means. The t-test used was a 1-tailed test because the
study's hypotheses were directional in nature. The
independent groups t-test was employed to help test the
significance of the differences of means for the groups
associated with each of the independent variables values.
Hence, a t-test was conducted on gender and college
education level groups and their relationship with the five
dependent variables.
The analysis of variance (ANOVA) is the technique that
separates the differences or variation of independent
variables into separate means units. The ANOVA measures
the differences between independent variables group means.
It compares within group variation of scores and the
variation between groups. These two variance measures
are then analyzed to test hypotheses.
The ANOVA was the appropriate choice to test for main
effects and especially for any interaction between the
independent variables. This study's data level had the two
elements necessary in a two-way ANOVA test: 1) the
dependent variables used interval data 2) there were two
or more independent variables being investigated.
In summary,f the data tested 178 subjects' response to
their perception of the generic terms "man", man related
terms and the pronoun "his" and their relationship to one's
gender and one's level of college education. The survey
instrument was pilot tested resulting in 25 unmasked
57


questions concerning generic terms for the final
questionnaire. The data were collected over a three month
period. Data were coded numerically and statistical tests
employed were: Pearson Correlation for the relationship
between measures, Alpha Reliability test for stability
among dependent measures, independent groups t-tests for
gender and college education level, and the ANOVA for main
and interaction effects. Next, the results of these
statistical tests are given and examined.
58


CHAPTER FOUR
RESULTS
This study was designed to examine subjects' use and
understanding of generic language. Hypotheses proposed
that those individuals with more college education would be
more likely to view generic terms toward a more masculine
interpretation than neutral interpretation and that males
over females would view these same terms more masculine
than females. Two independent variables, gender and
college education level were tested as predictors of five
dependent variables of generic terms. Statistical support
was found for the relationship between education on and
sentences that contained both the suffix "man" and the
pronoun "his" together. These results indicate that
graduate students rated the terms more masculine than did
undergraduates. Both groups rated the terms toward a more
masculine bias overall. Marginal support was found for a
consistently masculine bias by male graduate students over
all the other groups in one other dependent variable.
While the two hypotheses tested with five dependent
variables were not statistically supported overall, the
results provide minor support for the often cited claim
that the use of generic terms lead to a greater likelihood
of a masculine interpretation.
The results of the statistical procedures employed are


examined and statistically significant emergent patterns of
the variables are depicted at length. The discussion of
results begins with the report of the Pearson Correlation
Coefficients for the dependent measures (see table 4.1).
These are followed by an examination of the internal
reliability for the five dependent measurements (see table
4.2). T-test results on the separate effects of college
education level and gender are reported next (see tables
4.3 & 4.4). Lastly, the discussion examines analysis of
variance (ANOVA) results showing main and interaction
effects. Please see the complete ANOVA results on tables
4.5 4.9.
Internal Measurement Reliability
Pearson Correlation Coefficients
The dependent variables were correlated with one another
in an attempt to discover if a single construct was being
measured or if the items constituted five different
variables (see table 4.1). The variable "pre" had the
weakest relationship with the variable "both" at .5803.
The strongest variables relationship test occurred between
the variables "pre" and "suffix" at .8010. The other
variable tests measured between .5803 and .8010. A strong
to very strong relationship between each variable was
indicated by the data. However, since only one pair of
variables reached the .80 level, alternative tests for a
60


single construct must be considered independently (Betty,
1990). Only the "pre" and "suffix" variables are
sufficiently statistically correlated as one construct.
This suggests that different constructs exist here.
However, the remaining variables are highly correlated.
Table 4.1
Pearson Correlation Coefficients
Pre Suffix Both Man His
Pre 1.000 .8010 .5803 .7407 . 6306
Suffix 1.000 .6457 .7659 .6248
Both 1.000 .5966 .7215
Man 1.000 .6306
His 1.000
Reliability Coefficients
The consistency of single variables in relationship to
the total score of the five dependent variables were
compared and attained reliability coefficients ranging
from a low of .7713 to a high of .8947 (see table 4.2).
This indicates a strong to very strong reliability
relationship between individual scores and the total of
scores on all five measurements. There were 178 cases
measured.
61


Table 4.2
Reliability Coefficients
N of cases = 178
N of items =25
Variable Measured Reliability
Pre
7850
Suffix
7937
Both
8852
Man
7713
His
8947
Total of 5 items = .9071
Hypotheses Testing
The following hypotheses have been proposed:
Hypothesis One: The more college education individuals
have the more likely they will be to view the terms "man,"
man related terms and the pronoun "his" toward a masculine
interpretation.
Hypothesis Two: Males will view the generic terms "man,"
man related terms and the pronoun "his" toward a more
masculine interpretation than will females at the same
educational level.
Hypothesis One
Hypothesis One was statistically supported in one of the
five dependent variables tested. Using the t-test to
compare undergraduate students with graduate students the
62


"both dependent variable reached the .05 significance
level on a 1-tailed test. The variable "both" consisted of
five questions each combining the suffix "man" and the
pronoun "his" in each sentence. Faculty rated the terms
slightly more masculine than students on every variable
with the exception of the "his" variable. No findings were
statistically significant for t-tests completed on faculty
and undergraduate or graduate students. See table 4.3.
63


Table 4.3
Education Results
T-Test for Undergradates & Graduates for each Dependent
Variable
Pooled Var Est
Variable # cases mean sd se t value df 1 -tail orob
pre
undergrad 68 4.16 1.09 . 133 -.55 119 .290
graduate 53 4.27 1.10 .152
suffix
undergrad 68 4.44 1.21 .148 -.29 119 .385
graduate 53 4.51 1.22 .168
both
undergrad 68 4.85 1.35 .164 -1.63 119 .050
graduate 53 5.24 1.23 .170
man
undergrad 68 5.00 1.14 .139 -.34 119 .367
graduate 53 5.07 1.10 .152
his
undergrad 68 4.72 1.59 .193 .99 119 .156
graduate 53 4.45 1.28 .176
The mean of graduate scores on the 'both" variable was
5.24 in comparison to undergraduates' mean of 4.85. While
this may not be a practical difference in mean scores,
statistical difference was reached so these findings
warrant attention and further research. These results
64


indicate that graduate students were more likely to
perceive the generic terms as masculine than were
undergraduates. While other results were not statistically
significant at the .05 level, this pattern throughout the
education factor continued. The graduate students rated
the generic terms more masculine than did the
undergraduates, with one exception. Overall both the
graduate and undergraduate students rated the terms toward
a more masculine interpretation rather than gender neutral
(4.16 was the lowest rating with a rating of point 1 being
the most gender neutral and point 7 being the most
masculine).
Hypothesis Two
Hypothesis Two was not statistically supported in any of
the five dependent measures using the t-test. However, the
results do consistently illustrate males rating the terms
more masculine than females. Please see table 4.4 for
results.
65


Table 4.4
Gender Results
pooled variance
est.
variable number of cases mean sd se t value df 1-
tailed prob
pre
male
female
suffix
male
.132
female
both
male
.182
female
86
92
86
92
86
92
4.36 T.15 .124 1.19 176 .117
\
4.14 1.25 .131
4.59 1.23 .132 1.12 176
4.38 1.33 .139
5.09 1.45 .156 .91 176
4.90 1.31 .137
man
male 86
.240
female 92
his
male 86
.211
female 92 4.48 1.64 .171
Very little rating difference occurred between the
5.09 1.12 .121 .70 176
4.97 1.21 .127
4.67 1.46 .157 .80 176
66


genders. Males rated the terms more masculine by only .1
or .2 more than females. Even if these results had gained
statistical significance the differences would not have
been practical enough to warrant a claim of socially
important gender differences.
Analysis of Variance
ANOVA results registered one variable statistically
significant at the .07 level and another marginally
significant at the .14 level. See tables 4.5 and 4.9
respectively.
The .07 significance occurred on the "pre" variable which
was the first dependent measure. The findings related to
this variable concern the mean scores for male respondents
that began at 4.25 for undergraduates, peaked at 4.66 for
graduates and declined to their lowest rating of 4.18 for
male faculty. This indicated that faculty males rated the
terms the most gender neutral for the three male groups.
The trend for males is one of rising at the graduate level
and declining at the faculty level (see graph A).
Graph A- The ANOVA Male Trend
4.5
3.5
male-undergrad male-grad
67
male-faculty


These trend lines have nonparallel effects for males and
females. The same analysis conducted on females reveal an
opposing trend (see table 4.9). The findings of mean
scores for female respondents began at 4.09 for
undergraduates, declined to 3.88 for graduates and peaked
at 4.45 for female faculty. This indicated that female
graduate students rated the terms the most gender neutral
of the three female groups. The trend for females is one
of a decline, at the graduate level and rising at the
faculty level (see graph B) It was only at the faculty
level that the females rated the terms more masculine than
did males.
Graph B- The ANOVA Female Trend
Only on one variable did female undergraduate students
rate the terms more masculine than any other group. This
occurred on the "his" variable and achieved 14 level of
significance (see ANOVA table 4.9).
On this variable, the male mean score began with the
undergraduates lowest rating of 4.51, rising to 4.78 and
slightly declining to 4.72 for male faculty. This trend
differs from the previously discussed trend concerning
68


table 4.5 because here male undergraduates had the most
gender neutral rating and there the male faculty had the
most gender neutral perception.
The trend for female mean scores began at with their
highest rating of 4.88 for undergraduates, declined to 4.12
for graduates and rose to 4.26 for female faculty (see
table 4.9). This trend also differs from the previously
discussed female findings on table 4.9. This indicates
that female graduates rated the terms the most gender
neutral for the three female groups. The trend for females
is one of peaking at the undergraduate level, declining at
the graduate level and a slight rise at the faculty level
but one that is still lower than the undergraduate level
Findings here were that undergraduate females rated the
terms toward a more masculine interpretation than any other
female or male group regardless of college education level.
For complete ANOVA results see tables 4.5 4.9.
69


Table 4.5
ANOVA Interaction of Gender & Educational Level
Pre Variable
education level
undergrad grad fac
sex male 4.25 4.66 4.18
female 4.09 3.88 4.45
2-way interactions SS DF Mean Sq
Sig of F
sex education level 7.67 2 3.84
.072
Total 257. 177 1.45
Table 4.6
ANOVA Interaction of Gender & Educational level
Suffix Variable
education level
undergrad grad fac
sex male 4.44 4.75 4.59
female 4.44 4.26 4.41
2-way interactions
Sig of F
sex education level
.611
Total
SS DF Mean Sq
1.65 2
.825
291. 177 1.64
F
.67
F
494
70


Table 4.7
ANOVA Interaction of Gender & Educational Level
Both Variable
education level
undergrad grad fac
male 4.81 5.57 4.94
female 4.89 4.90 4.93
2-way interactions SS DF Mean Sq F
Sig of F
sex education level 4.72 2 2.36 1.24
.290
Total 336. 177
Table 4.8
ANOVA Interaction of Gender & Educational Level
Man Variable
education level
undergrad grad fac
male 5.01 5.27 5.02
female 5.01 4.88 5.01
2-way interactions SS DF Mean Sq F
Sig of F
sex education level 1.36 2 .680 .490
.614
Total
241. 177 1.36
71


Table 4.9
ANOVA Interaction of Gender & Educational Level
His Variable
education level
undergrad
sex male 4.51
female 4.88
2-way interactions
Sig of F
sex education level
.145
Total
grad
4.78
4.12
SS DF
9.38 2
427.
177
fac
4.72
4.26
Mean Sg
4.69
2.42
1.95
An interaction effect was statistically supported at the
.07 level on the "pre" variable (table 4.5) and marginally
at the .14 level on the "his" variable (table 4.9).
The "pre" variable results significant at the .07 level
are examined first. The trend concerning males depicted a
pattern of graduate students rating the terms toward a more
masculine interpretation with an decrease toward a more
gender neutral interpretation at the faculty level. The
pattern for females was opposite of the male pattern.
Female graduate students rated the terms more gender
neutral with an increase toward a masculine interpretation
at the faculty level. The interaction effect on the "pre"
72


variable showed male graduate students rated the terms more
masculine than any other group regardless of gender or
college education level.
The "his" variable results, although only marginally
significant at the .14 level, need to be examined because
this was the only variable that resulted in female
undergraduates rating the terms more masculine than any
other group. The finding concerning females was depicted
by female undergraduates rating the terms more masculine
followed by a shift toward a gender neutral perception at
the graduate level and back toward a masculine perception
at the faculty level. Male graduate students rated the
terms more masculine with a slight shift toward a gender
neutral perception at the faculty level. Only on the "his"
variable did female undergraduate students rate the terms
toward a more masculine interpretation than any other group
regardless of gender or college education level (see graph
C) .
The patterns revealed give some indication that the term
"man," man related terms and the pronoun "his" are being
Graph C ANOVA "His" Trends
+ Moles
O Females
3.5
3
2.5
2
t.5
0.5
0
undergrad1 gred 1 faculty
73


perceived toward a more masculine interpretation rather
than gender neutral among college students and faculty
members in this study. What do these finding signify?
Analysis and speculation concerning these results are
presented in the next section.
74


CHAPTER FIVE
DISCUSSION
This study was designed to examine subjects' use and
understanding of the generic terms: man, man related terms
and the pronoun "his". The study's main result indicates
that graduate students are more likely than undergraduate
students to perceive the generic terms as masculine. This
finding was statistically significant on one measurement at
the .05 level. An interaction effect indicated that male
graduate students interpreted the terms more masculine than
any other group regardless of gender or college education
level at the .07 level of significance. While neither was
significant at the .01 level, both results indicate a trend
of graduate students over undergraduate students rating the
terms toward a more masculine interpretation. It was shown
that generic language that has been taught to be gender
neutral throughout education is no longer being perceived
this way. What might this reversal in language habits
mean? What change has occurred for male and female college
students perception of generic language? Why are males
limiting their perception of generic terms primarily to
males while females are perceiving their own inclusion?
Is it so that females do not feel left out of a substantial
part of the English language and communicative process?


These questions are examined in the following discussion.
This chapter first examines the education element of
Hypothesis One. Next, Hypothesis Two which did not reach
statistical significance is briefly examined. Then
interaction effects are covered that include gender and
education factors as contributors. Lastly, suggestions for
further research are offered followed by a review of the
conclusions. Now, the study's two hypotheses and
interaction effects are discussed in terms of statistical
findings and their implications.
Hypothesis One: The more college education individuals
have the more likely they will be to view the terms, "man",
man related terms and the pronoun "his" as masculine.
The first hypothesis concerns college education level
and interpretation of generic terms. Statistical support
was observed on one test for this hypothesis. Three
alternative explanations for this finding are provided in
the following discussion. These explanations focus on the
issues associated with the learning of generic terms, the
application of generic terms and the role authority plays
in the learning and application of generic terms. The
discussion attempts to present explanations for the "both"
variable that received statistical support and the four
other measurements that did not.
76


Hypothesis Two: Males will view the generic terras, "man",
man related terms and the pronoun "his" toward a more
masculine interpretation then will females at the same
educational level.
t
The second hypothesis concerns gender and interpretation
of generic terms. The statistical test of main effects
indicated that differences between males and females were
not sufficient to warrant a claim of socially important
gender differences. On the t-test males consistently rated
the terms toward a more masculine perception but the
difference was a slight one.
Both hypotheses lent themselves to test for interaction
effects. Six interaction effects were concluded on the
ANOVA testing. One interaction effect tested was
significant at the .07 level. On the variable measures
that were not statistically significant, most of the
interaction trends continued.
This chapter's discussion of effects begins with
discussion of Hypothesis One. The explanations focus on
one's gender identification in language and gender
differences in learning and responding to pronoun
interpretation. Hypothesis Two is briefly examined but
since no major gender differences were supported little
analysis can be completed. Next, interaction effects are
analyzed by gender and by college education theories.
77


Gender identification and equality in language and learning
and application of generic terms are interwoven into these
explanations in an attempt to explain the six interaction
effects.
Hypothesis One
Hypothesis One received statistically significant support
using the independent groups t-test on the "both" variable
(.05). This t-test examined the college education factor
in connection to subjects' use and understanding of the
generic terms. Graduate students rated the combination of
the man related terms and the pronoun "his" toward a more
masculine interpretation then did undergraduate students.
Tests regarding the other variables, "pre", "suffix", "man"
and "his" did not receive support but with the exception of
the testing of the "his" variable continued the same trend
of graduate students rating the terms as more masculine.
What are the implications of these results and how are they
connected with one another?
When the generic terms foreman, weatherman, congressman,
spokesman and salesman were individually combined with the
pronoun "his" in the same sentence (the "both" variable),
the sentences were rated more masculine by graduate
students and reached statistical significance of .05. It
was only when these two generic terms were combined that
statistical significance occurred (i.e., the foreman gave
78


his verdict). The combination of terms had a stronger
influence on students to rate them as masculine than when
only one generic term was Used in a sentence. Graduate
students rated the "both" variable the most masculine at
5.24 (the highest masculine rating on any variable) in
comparison to undergraduate's rating of 4.85 (the third
most masculine rating on any variable). It was here that
the greatest difference between mean scores of the two
groups occurred. It must be noted that although
statistical significance was reached on this variable, one
must address the issue of the finding's practical
importance. This slight difference between mean scores is
not great enough to warrant an observable behavioral
difference between the groups. It is unlikely that the
graduate students over undergraduate students who completed
this survey would exhibit observable linguistic differences
concerning generic terms (the use of less generic
language).
The "pre", "suffix" and "man" variables demonstrated the
same trend. Graduate students rated the terms more
masculine than undergraduates. These variables along with
the "both" variable all contained the word "man" as a
prefix, suffix or used alone. It appears that the students
perceived the different uses of "man" in a similar manner
as the "both" variable but statistical significance was not
reached.
79


The trend changed on the "his" variable. Undergraduate
students rated the terms more masculine than did graduates.
There was a .3 difference between mean scores of the two
groups. Students must have looked at the pronoun "his"
somewhat differently from the man related terms.
Undergraduates were less likely to include females in the
pronoun usage. The discussion on the "his" variable
follows this first discussion.
The following discussion section offers analysis on all
of the findings (except those of the "his" variable). The
analysis offered attempts to directly link the role of
education to generic language learning and interpretation.
The first issue under analysis is associated with the
learning of generic terms. The academic environment is a
formal educational setting for learning. Children enter
elementary school with the basics of language learned at
home and in informal settings (i.e., child care). The
children's language skills include word choice and order
but no formal language rules. Generic language learning
follows general knowledge of language learning. Children
only know what they have been told, seen role modeled, or
learned intuitively on their own. Hyde (1984) found only
28% of first graders knew the grammatically correct generic
"he" rule. But by college 84% knew and used the rule.
It is in formal learning (school) that children are
taught generic terms' usage and semantics. Children learn
80


language rules and abide by them in order to succeed (i.e.,
gain approval, get good grades, etc.). Children adhere to
the rules in this structured environment that does not
encourage much deviation. A child in grade school may
learn generic rules, be forced to use them in a structure
that does not offer alternative terms and grow up knowing
they are prescriptive grammar and yet never truly
understanding their inclusiveness. Education plays the
primary role in generic language learning. It is through
primary education that one learns generic terms.
A second important issue is the application of these
terms. This application takes two forms, one involving
teacher application and influence and the other involving
the student him/herself. Children learn and obey the
expected application of generic terms and it is only
further along in their education that people start to
question this practice or that teachers offer alternative
terms to students. Even in college, the novelty of
creating one's own ideas and papers after 12 years of
restricted classroom instruction may result in
indiscriminate pronoun use and the variety of generic "man"
conjunctions. Perhaps only by graduate school do students
no longer perceive generic terms as gender neutral.
Graduate students were more likely to perceive generic
terms as masculine, so somewhere in their college education
they have been influenced (required) by professors to use
81


alternative terms rather than generic terms to refer to
neutral or feminine referents. It also follows that
teachers must have used alternative terms when teaching
(i.e., "he or she" or "men and women") as an example to be
followed. Since it is unlikely (but not impossible) that
elementary teachers or high school teachers used
alternative terms, students must have been taught these
terms in college. It is at the graduate level that a
change in perception of generic terms occurs. What might
account for this change?
At the undergraduate level, students still ask for and
receive much structured guidance concerning tests, papers
and research projects. They are expected to complete work
according to the professor. They are given a topic to
write on and possibly told not to use the masculine pronoun
"his" when referring to referents. The student responds to
this task possibly without giving too much thought to
pronoun meaning. Once again, the student is following the
rules imposed by the teacher. After four years of this
instruction, a change may occur in generic language
awareness.
For the graduate student a change has occurred in generic
language awareness. Graduate students were more likely to
rate the generic terms as masculine. Some reasons for this
change may involve the graduate student's new level of
responsibility for his or her own education. This would
82


include the writing of numerous research papers which
causes the student to become more familiar with the
structure of writing and with the subtleties of meaning.
Graduate students become aware that generic terms are not
inclusive and that using only them excludes females.
The third issue under analysis concerns the role
authority plays in the learning and application of generic
language. Education is a link between generic rules and
application and their interpretation. Harrigan & Lucic
(1988) found students over non-students were the most
influenced by authority figures. It could be guessed that
students would be the most influenced group since they are
in an atmosphere of learning from adults. Perhaps teachers
who require the use of non-sexist language and themselves
use language alternatives such as, "he or she" instead of
"he" influence their students to do the same. I can only
speculate that teachers at UCD and Metro State are more
likely to use these alternative terms in graduate courses
as a model for students. While some students may be
transfer students not all the students can be transfer
students, so some of their formal usage exposure must have
come from the Auraria campus.
What can explain the differences in student responses to
the "his" variable as opposed to the other man related
terms? The difference between man related terms and the
pronoun "his" concerns complementary terms. The term "he"
83


has the term "she" and "his" has "hers" as their
complements to indicate a female. Many of the man related
terms have no complementary feminine term or only recently
has a term been coined specifically for females but this
use has not gained the popularity of the masculine term.
The term "mankind" has served as a universal inclusion for
males and females. It has no parallel term to indicate
females so that "mankind" can stand alone for maleness as
and does "he". Although the term "womankind" does exist,
it is rarely used or accepted as a term to indicate
females. A main effect detected undergraduate students as
rating the pronoun "his" more masculine than graduate
students. This result could have occurred due to
undergraduates' introduction into the academic environment.
As already discussed, students are highly influenced by
authority figures (professors). Teachers in college are
encouraging independent thought and language
discrimination. They are asking the students to break free
from the restrictions imposed on them as high schoolers and
connect language to thought. Words are tools used to
convey meaning and have an impact that goes beyond the
printed word. This personalization of language may affect
generic terms and their implications. This teaching may
affect students' usage and response to generic terms. The
students are just beginning to make a place in the world
for themselves and are marking this distinction through
84


their language. They may be saying, let "he represent
males and "she" females, let each gender have its own
distinctive pronoun. When both genders are included, use
"he or she" or "they" but not just "he". Perhaps
undergraduate students are reversing the generic pronoun
rules they learned throughout school as they become better
as writers and more socially aware of the power of
language.
Hypothesis Two
Hypothesis Two proposed that males more than females at
the same educational level would rate the terms more
masculine. There were no statistically significant
results, although the trend on every variable showed males
rating the terms more masculine. This means males were
more apt to exclude females and females were more likely to
include females in generic terms. This trend differs from
most of the research on generic language which indicates
that females rate such terms more masculine then males
(Benoit & Shell, 1985; Fisk, 1975; Hamilton, 1988; and
Martyna, 1978). Perhaps there is a change occurring
concerning gender and generic language interpretation. It
may be that the Auraria campus has different linguistic
norms than other campuses or that this lack of gender
differences is a signal for educated people in our society.
85


Interaction Effects
Hypothesis One states that the more college education
one has the more likely one will be to view the generic
terms as masculine. Hypothesis Two states that males will
view the generic terms more masculine than will females.
Analysis was performed for five interaction effects
associated with the factors of education and gender and
their connection to generic terms. These analyses resulted
in several findings concerning male respondents who
exhibited a pattern of means similar to an inverted V.
Male undergraduates and male and female faculty rated the
terms more generic than the highest masculine rating of the
graduate students on five out of five measures. The second
interaction concerns female respondents who exhibited a
pattern of means similar to an upright V. The female trend
revealed undergraduate students and faculty rated the terms
more masculine while graduate students rated the terms more
generic on four out of five measures. Male graduate
students rated the terms more masculine than any other male
group on five out of five measures. Male undergraduate
students rated the terms the least masculine of any male
group on four out of five measures. Female graduate
students consistently rated the terms more generic than any
other group on four out of five measures. Male graduate
students rated the terms more masculine than any other
group on four of the five measures with female
86


undergraduate students rating the terms more masculine on
the fifth measure.
This discussion first examines the "pre" variable which
was the only variable that demonstrated statistical support
for an interaction effect concerning gender and college
education level. The following section examines gender and
college education as two interrelated explanatory variables
for the interaction findings. Gender is treated first,
since research has been done on females and males
concerning generic language, and only loosely connected
research has been done on education and generic language.
The gender explanations concern female exclusion in the
generic terms by males and females and female inclusion in
the generic terms by males and females. The college
education level explanations concern undergraduates,
graduates, and faculty exclusion and inclusion of females
in the generic terms. The interaction effects are
delineated within these appropriate discussions. Following
these discussions, suggestions for further research are
presented and then followed by a summary of conclusions.
The "pre" variable interaction effect was statistically
supported at .07 level. No other interaction effects were
statistically significant. So this section will only
discuss those findings associated with the "pre" variable.
The means for the males on this variable revealed
undergraduates to have the middle score (4.25), graduates
87


the highest or most masculine (4.66) and faculty the lowest
or most generic (4.18) See Graph D for the group means of
the "pre" variable. The means for females exhibited an
responses at the middle score (4.09) as did male
undergraduates. However, here the trend reverses for
females, as graduates had the lowest score (3.88) and
faculty the highest (4.45). This pattern of means is one
of an upright V for females. Male graduate students rated
the terms more masculine than any other group. Female
graduate students rated the terms the least masculine
(hence, more generic) of all groups. What are the
implications of the interaction effects when combining
gender and college education level?
What do we make of the differences between male and
inclusiveness of females in the generic terms when the
males were undergraduate students and faculty members but
more exclusiveness when they were graduate students? Why
opposite trend. Female undergraduate students made
Graph D ANOVA Pre Trends
3.5
3
2.5
2
1.5
0.5
0
undergred1 grad 1 faculty
female graduate students? Why did males indicate
88


do females show the opposite trends? Several explanations
are offered. They begin with analysis of those males and
females who rated the generic terms the least masculine and
hence most inclusive of females. These explanations are
interwoven with gender identification and gender equality
in language as well as the learning and application of
generic terms previously covered in the discussion of
Hypothesis One.
Gender as a Contribution to Interaction Effects
What is one to make of males who rated the generic terms
the least masculine and hence, the most inclusive of
females? This discussion includes the pattern of mean
scores for males, especially male undergraduates and
faculty who rated the terms less masculine than male
graduate students. On the "pre" variable male faculty
rated the terms less masculine than male undergraduates and
graduates. These males have used generic terms throughout
their lives, are comfortable with them and may not perceive
the need to change or alter them in any way. They included
female in the terms because it has always been done this
way and they may not be aware of the uproar concerning the
terms. Although male faculty must be aware of the uproar
considering their role as an acamedician. These males may
have a lack of awareness necessary to change their language
use. Or it may be the opposite case. These males may be
89


strident against changing the language they believe in, are
comfortable with, or think is equal as it is. They may
believe their own gender identification is secure in these
terms and should be also for females as well since they
also grew up using these terms. The language is as equal
for women as it is for men and just because the terms are
originally masculine in nature does not make them less
applicable for females.
What about those females who rated the generic terms the
least masculine and hence, most inclusive of females? This
discussion concerns the pattern of mean scores for females
especially female graduate students who consistently rated
the terms the least masculine. Gender identification may
play a key role. Women were more likely to include
themselves in the terms because otherwise they would not
have been identified in the terms at all. Women rated the
terms slightly less masculine so there would be at least a
chance of female inclusion. If generic language is not
gender inclusive females remain absent or invisible in
generic terms. Women's identity becomes less notable and
valued if they are not identified in generic language.
Generic terms are widely used and women's inclusiveness is
necessary for their representation. People identify
themselves in their language as they identify with their
own name. Women want to believe they are included in
generic terms. The more identification women have in
90


language, the larger potential they have in their society
(Silveira, 1980). One's identity is central to one's
being. Language is a primary source of identity in which
r
to identify ourself, others and the world in which we live.
Language is also a source where equality can be expressed.
For some females, equality in language is the tendency
for females to do what males naturally do in generic terms-
-include themselves in the generic terms in any context.
Generic terms operate only in a defined context, out of
that defined context the terms go back to their original
masculine intention. Females may be seeking the same
inclusion or identification that males have, which is one
of complete inclusion in generic terms. They are seeking
equality in language but it is in a language that has a
masculine foundation of semantics. The terms were
originally masculine in definition and interpretation.
Other analysis of female inclusion concerns the
difference in learning and responding to pronoun
interpretation. As already covered in the discussion on
Hypothesis One, there is a learning similarity between
generic nouns and pronoun complements. The term "man has
"woman" as its complement as does "he" have "she" as its
complement. Children learn generic pronoun and generic
noun interpretation in much the same manner. Girls learn
at an earlier age that generic pronouns may be inclusive of
both males and females. Females must formally learn
91


generic meanings in school, those of which may be misused
and corrected. Males are less likely to be corrected
because they are using the "correct" pronoun but it is not
known if they have fully grasped the generic inclusive
meaning. Girls and women have had to learn the formal
rules of generic terms and use them if they wanted to be
included in the textbook, lesson, discussion, language or
in whatever context the generic terms were being applied.
Females have been taught to refer to "a person" with the
masculine pronoun "he" and so they must reverse their
natural tendency to use themselves as reference of the
subject. Now in order to have their gender exist in
generic terms, females must believe in, perhaps insist upon
its inclusiveness (Henley, 1989).
What is one to make of males who rated the generic terms
the most masculine and hence, the most exclusive of female
interpretation? This discussion covers the pattern of mean
scores for males especially male graduate students who
rated the terms more masculine than any other group on the
"pre" variable. In terms of gender identification and
equality in generic terms, these males identified with
generic language to the point of the exclusion of females.
Masculine terms that are "officially" declared generic
terms continue to describe males, include males and are
being perceived toward a more masculine interpretation than
neutral one. Males have heard and used the terms to
92


Full Text

PAGE 1

THE ROLE OF EDUCATION IN THE USE OF MASCULINE GENERICS by Vickie L. Harvey B.A., University of Colorado, 1988 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Department of Communication 1990

PAGE 2

This thesis for the Master of Arts degree by Vickie L. Harvey has been approved for the School of Communication by I Michael Monsour 1 Jon A. Winterton I I ii 1-31-Cfa Date

PAGE 3

Harvey, Vickie L. (M.A., Speech Communication) The Role of Education in the Use of Masculine Generics. Thesis directed by Michael Monsour, Assistant Professor The question under studyconcerns perception of generic terns such as "man," man related terms with "man" as a prefix or suffix and the pronoun "his" and their relationship to one's educational level and gender. Two . independent variables, gender and education level were correlated. to five dependent variables of generic terms. This study is designed to examine subjects' use and understanding of generic language. . statistical support was found for sentences that contained both the suffix "man" and pronoun "his" together. These results_found that graduate students rated the terms more masculine than undergraduates and both groups rated the terms less generic overall. Interaction effects on the "pre" variable were marginally supported. While the two hypothesis were not completely supported, the. results provide minor support for the often cited claim that the use of generic terms lead to increased thoughts of men. The form and content of this abstract are approved. I re_coinmend its publication. Signed Michael Monsour

PAGE 4

CONTENTS CHAPTER 1. 2. INTRODUCTION Rationale Problem . . . . . BACKGROUND AND REVIEW OF LITERATURE Overview o Summary . A Correlation of Education & Generic Language . o Gender Differences o 1 2 11 21 0 22 38 0 0 39 0 42 Purposed Research Questions & Hypotheses . 46 3. 4o METHODS o o o o Measurement Instrument o Independent Variables Dependent Variables Subjects Data Collection Procedures Pilot study Survey Procedures o Analysis o o o o Coding o o o o o Statistical Analysis RESULTS o o o o o o Internal Measurement Reliability o Pearson Correlation Coefficients Reliability Coefficients o o o o iv 0 0 48 49 49 50 51 51 51 52 53 0 0 53 0 53 59 60 60 61

PAGE 5

Hypotheses Testing . . . . . . 62 Hypothesis One 0 0 62 Hypothesis Two 0 65 Analysis of Variance 67 5. DISCUSSION . 75 Hypothesis one 78 Hypothesis Two 85 Interaction Effects . . . o o o o 85 Gender as a Contribution to Interaction Effects . o o o 88 College Education Contributions to Interaction Effects . . . . o o o 94 Limitations of Present Study o o 0 100 suggestions for Further Research 0 101 Review of Conclusions 103 APPENDIX A. Questionnaire and cover letters . . o o 107 B. Survey . . . . . . . . . 110 LIST OF REFERENCES . . . o 114

PAGE 6

TABLES Table 4.1 Pearson Correlation coefficients 61 4.2 Reliability Coefficients . 62 4.3 Education Results ...... 64 4.4 Gender Results . 66 4.5 ANOVA Interaction of Gender & Educational Level--"Pre" Variable . 70 4.6 4.7 4.8 4.9 ANOVA Interaction of Gender & Educational Level--"Suffix" Variable ANOVA Interaction of Gender & Educational Level--"Both" Variable ANOVA Interaction of Gender & Educational Level--"Man" Variable ANOVA Interaction of Gender & Educational Level--"His" Variable vi 70 71 71 72

PAGE 7

FIGURES Figure 1.1. Graph A--The AN OVA Male Trend 1.2. Graph B--The AN OVA Female Trend 1.3. Graph C--ANOVA "His" Trends 1.4. Graph D--ANOVA "Pre" Trends vii . . . 67 68 73 88

PAGE 8

CHAPTER ONE INTRODUCTION Of this remarkable protagonist must it not be said that his capacity to adapt his universe and its physical laws to his own needs and desires and purposes--his science and technology--is one of the qualities of his greatness? Miller & swift, 1988. Are the term 11man11 and the masculine pronoun 11his11 being used generically or is the passage only referring to males? Past research has.demonstrated that generic terms (11man,11 man related terms and 11his11 or 11he11) do not function generically (Martyna, 1978; Hyde, 1984; Hamilton, 1988). The terms bias the reader or listener toward a male interpretation of the material rather than toward a male and female interpretation. The literature reviewed and research presented in this study look at masculine generics from a theoretical, historical and analytical perspective. Later in this chapter the central concern of this research is presented which examines perception of the generic terms: "man," man related terms and the pronoun "his" and their relationship to one 1 s gender and one 1 s level of college education. The upcoming section discusses the Whorfian hypothesis in connection with masculine generics. Linguistic relativity is introduced as a foundation from which to expand.

PAGE 9

Rationale Whorf (1956) and Sapir (as cited in Whorf, 1956) conducted intensive research concerning the relationship between language, thought and social behavior. The Whorfian Hypothesis, linguistic relativity, provides the theoretical foundation that states that the "soul of a people" and the "mind of a people" were not only reflected by its language but, actually, formed by it (Mandelbaum, 1949, p.l62). The language of the real world is intertwined with one's thoughts and one's behavior. Language serves as a reflection and reveals the way our culture projects its image. An example of language reflecting reality from the 1960's is the creation of the term Black Power. It was born of anger and resistance by a group of people who felt misrepresented by the terms society used to refer to them such as "colored people". The term Black Power expressed an emotional post-Malcomist statement of black consciousness and racial pride rather than a concrete program (Blair, 1977) The newly born Black movement was endorsed by the term Black Power which quickly caught on and helped the spread of the movement beyond the Sapir (1929) refers to language as the "symbolic guide to culture." Language is probably the greatest force of socialization that exists. It serves as a symbol of the social solidarity of those that speak the language (Sapir, 2

PAGE 10

1921). The Whorfian Hypothesis linked language to creation and perpetuation of the reality of those who create language (traditionally men). They suggest our reality is built upon our perceptions and that language affects us in direct and less conscious ways. Two obvious characteristics of the structure of our language are grammar and semantics. Semantics is the foundation from which we devise experience (Sapir, 1929). In English, semantics and usage use masculine generics to generalize to females. In this regard, English differs from other languages. In Spanish all nouns including those denoting non-living things are either masculine or feminine. They do not use a masculine or a feminine noun to include the other gender. These supposedly generic terms, he, his and him name males as active agents in our society more than females. It is grammatically correct to assume from the English language that agents or active persons are men. The masculine pronouns used generically may or may not be interpreted as gender neutral by their audience, but still this usage communicates distinctive cultural patterns associated with English users' perception of the world. Language is a major medium for the transfer and maintenance of any given culture. The way a civilization expresses itself linguistically is the road map to that culture (Sapir, 1929). The implications of language do not stand apart from 3

PAGE 11

experience but run parallel to direct experience (Sapir, 1929). Language is at the base of our sense of self. As speakers of the language we identify with it and are referents of the language. It is the language itself that contains the sex bias, not necessarily the purpose of its speakers. Our psychological experience of language is just as important as the language itself. This perspective invites a thorough presentation of the history of the usage of nouns and pronouns. Nouns and pronouns play essential roles in our language, and it is through these words that we may unconsciously shape our world. Nouns and pronouns may be either masculine, feminine or neuter. Masculine nouns ("man", "husband", "mailman", etc.) all refer to the male gender. Feminine nouns ("woman", "wife", "daughter", etc.) all refer to the female gender. Neuter nouns ("garden", "sky", "sweater", etc.) all refer to objects without gender. Masculine pronouns ("he", "him", or "his") refer to the male gender. Feminine pronouns ("she", "her", or "hers") refer to the female gender Neuter pronouns ("it" or "its") refer to objects without gender. A masculine noun or pronoun will generally refer to any male regardless of race, religion, age, geographical location, time or occupation. The same applies for the feminine and neutral nouns and pronouns (Evans & Evans, 1957). Masculine nouns ("man", "manpower" or "fireman") used 4

PAGE 12

generically reflect a language's loyalty and superiority of the male gender by excluding women in this semantic structure (Fishman, 1960). The effect of exclusion is the invisibility of women, invisibility in the sense that women are not included in the imagery of masculine terms. It has been empirically shown that women are excluded by the "he" while men are merely made less prominent by alternatives of "he or she" or "they" (Martyna, 1978). McConnell-Binet, Borker, & Furman (1980) suggest this exclusion in language has been historically unrecognized and only recently is it being brought up for examination. Following the thought of the Whorfian Hypothesis, if women are invisible in our language they are likely to be less visible as active agents in our society. If masculine pronouns and nouns are meant to refer only to males, then why are the terms used generically? One begins to find formal arguments from English grammarians attempting to crystalize the male pronouns and nouns as gender neutral in the 1700's. Bodine (1975) writes about early John Kirby, 1746, among others, who refer to the masculine gender as the "worthier" gender and this was the reason given that the masculine terms "he" and "man" should be used for women. Bodine continues with a quote by Ward, 1765 who wrote An Essay on Grammar that stated that pronouns must agree in gender; "he" must represent a male; "she," a female; and "it," an object of 5

PAGE 13

no sex. Yet another grammarian stated that "he or she" should not be used because it draws attention to women and makes them appear in a special league separate from people (McCawley, 1974, cited in Bodine, 1975). An act of Parliament in 1850 by grammarians (all male) legalized the gender specific term "he" to be the "official" pronoun for convenience' sake and to alleviate the ciumsiness of the phrase "he or she. 11 Now the male pronoun would include females. They also argued that "they" was the plural pronoun and could not be used as the singular form, but "they" continued to be the more widely used. The striking thing about the male pronouns, "he, his and him" was that it took 50 years for people to believe that "he" was a generic in English and it took 20 more years for people to believe it had always been this way (Stanley, 1978). There was tremendous resistance by the users of the language in adopting and believing the masculine terms were generic. An attack on the language is essentially an attack on the speakers of that language. It took 50 years to implement this language change and it still continues to be a source of debate. One year after the passage of the Parliament Abbrevation Bill on June 16,1851, an attempt was made in the House of Commons to repeal the Act. John Stuart was concerned that the bill inadvertently gave women rights which they should not have. A Reform Bill concerning housing and voting 6

PAGE 14

rights was written using the masculine pronoun "he" and this would now give women the right to vote since women were now included in the masculine terms. The attorney general claimed that this wording would not allow women to vote and if it did, so be it. (Baron, 1986). Twenty-five years after the formalization of "he" into a generic term, interpretation still remained open in the legal arena. Lavina Goodell in 1875 was denied admission to the bar because the statute regulating bar admission referred to attorneys as "he." Her petition asserted that the term had been legally changed to include females as well as males. The court denied her request on the grounds that "the language of the statute, of itself, confessedly applied to males only" (Cox and Ray as cited in Hamilton, Hunter, & stuart-Smith, 1990). Penalties or duties written in the masculine form applied to women but privileges or benefits were upheld by the courts as only inclusive of males. The ambiguity of the term continued to be open to discussion and interpretation. The Dictionary of Contemporary American Usage (Evans and Evans, 1957) defines the masculine pronouns "he", "his" and "him" in natural English as always referring to males. Theoretically "he" can refer to either a male or a female when the sex is unknown. In an example, "Each student must turn in his test," "his" is not being generically used because the sex of the student is known. There are boys 7

PAGE 15

and girls in the classroom not unknown persons. The Parliament Act of 1850 was used to shorten the language so that all masculine pronouns ("he", "his", "him") would be understood to include females. This Parliamentary Act or all the grammar rules in the world can not change the fact that, if we are told "somebody telephoned while you were out," we will say "did they leave a message?" In natural English, the words "they," "them," and "their" are all used more often than "he," "his" or "him" when referring to an unknown person who may possibly be a woman. "He" is also used 10 to 20 times more often to refer only to males than to both males and females (Hughes & Casey, 1986). The ambiguity of these generic words lends itself to women's invisibility in their meaning. Ambiguous is defined by Oxford American Dictionary (1980) as having two or more possible meanings: "doubtful, uncertain." The word "man" and pronouns "he" and "his" have had to do a double semantic duty. They have served both gender specific and gender neutral roles. The generic uses of "he," "his" and "man" have no set contexts to resolve their ambiguity (Moulton, Robinson, & Elias, 1978). There is no sentence structure or usage that identifies a generic term as generic. It is only when a feminine term, "woman" is used with amasculine term, "man" or "he" that the generic term automatically become masculine in meaning. There is no rule that states when a masculine term is used without 8

PAGE 16

a feminine term it is automatically generic. Generic terms frequently need further clarification to resolve their intended meaning as either generic or gender specific. All readers must read further when the term "man" is used to see if the generic use or the gender specific use is intended. However, men automatically know they are included in these masculine words while women must search for the-intended meaning to find either included or excluded. A sentence referring to man and his destiny may start out generically and as the sentence progresses become gender specific, "part of man's happiness is achieved through his work, environment, wife and children." Suddenly the generic "man" which woman was originally included in now causes woman to vanish from the meaning of the sentence and the minds of its readers. Ambiguity itself can be resolved by adding detailed information or by reading or listening further for gender intention. While this may be inconvenient, time consuming and biased toward masculine language, the real problem concerns the nature of true generic language. Generic terms may fail as gender neutral terms regardless of intentions and efforts of speakers to use them neutrally and the terms may accidentally not be applied to neutral situations even when speakers provide explicit information that the terms are neutral (Moulton et al., 1978). The problem arises because the terms are not truly generic. An 9

PAGE 17

authentic generic term can be used in different contexts to denote the same meaning. The term "person" can be used generically in two ways: "she is the person I hired" or "he is the person I hired." The noun "person" easily changes gender with the feminine and masculine pronoun used. This is how one may tell if a word is generic. The terms "he" and "man" are not interchangeable with feminine nouns and pronouns: "he is the prettiest girl here" or "she is our new weatherman." If "he" and "man" are genuinely gender neutral, then they ought to be applicable to any person regardless of gender (Moulton et al., 1978). A term that is used to represent a class or group that does not apply to all of its members is a false generic. In the upcoming section two problems are connected to generic language. The first one concerns women's invisibility in generic terms. The most obvious examples are when the masculine form is used as a generic form, as in "policeman," "mailman," "the average man," or "mankind." Linqusitics call the masculine form the unmarked while the feminine is the marked since an additional marker must be used to denote a female. An unmarked word is the more natural one and is closely associated with the norm, while the marked word is viewed as the exception and needs more information to differentiate it from the natural state (Salter, 1979). The second problem concerns the reinforcement of sex role stereotyping. Sex role 10

PAGE 18

'V'O ((R)(()) O\\Q
PAGE 19

stereotyping of professions does have certain linquistic consequences. For example, when referring to the chairman, that person has traditionally been a male but with more women in higher social positions it is common for the chair to be a female. The same can be of secretaries or elementary teachers being primarily females in the past but present positions of these jobs have had an increase of males employed in them. Problem Generic language does not accurately represent women, which leads to two connected problems: 1) women are made invisible by generic language which implies a male is respopsible for the job or situation and this affects their self-esteem and identity; 2) generic language reinforces sex role stereotyping. Generic language makes women invisible by implying that a male is responsible for the job or situation (Henley, Gruber & Lerner, 1988). While ambiguity of the terms imply a confusion of meaning, the male image is the first image brought to mind by generic language (Hamilton, 1988; Stanley, 1978). A job title that includes a masculine term such as "newsman," "congressman" or "chairman" implies that a male is employed in that position. The comparative visibility of women and men in such job titles raises the question of discrimination (Ward, 1975). Men are much more visible in these terms than women regardless of the gender 11

PAGE 20

of the jobholder. A situation that refers to the lack of manpower available or the manhours it takes to complete a task also reflects women's invisibility and implies man's responsibility for the situation. The words "manpower" and "manhours" deny women's efforts while highlighting men's involvement. Women's roles are not accurately reflected in generic language. A woman may be part of the "manpower" in a project or work "manhours" to accomplish a feat and not be acknowledged for it in masculine or generic language. The implication would be that men participated in the project rather than women. There is also empirical evidence that supports the disregard for female's ability to perform the job (Hyde, 1984). If women are to be full participants in our society, our language needs to reflect their ability and involvement. Not only is generic language inaccurate, but it also leads the listener/reader into perceiving the person as male rather than as male or female (Martyna, 1978; Stericker, 1981). This inaccuracy and misperception of women in our communication lead to an increased awareness of masculine generics as sexist language. It is empirically unknown for certain what the extent of those effects are on women and communication. Two effects of women's invisibility in generic language that receive frequent publication are lack of self-esteem and uncertain identity. Women do not view themselves in generic language and when 12

PAGE 21

they are referred to as men this has deep effects on their self-esteem. This decrease in self-esteem causes a decrease in communication due to lack of confidence (Martyna, 1978; Henley, Gruber, & Lerner, 1985). Henley et al. (1985) found boys had more positive change in self-esteem using masculine generics and that girls had more positive change in self-esteem using neutral pronouns. Empirical evidence points out that pronouns have significant effect upon selfesteem on both females and males. The masculine pronouns and nouns can lead to lower self-esteem as compared with unbiased alternatives for females ("he" or "she" or "humans") (Henley et al., 1985). Inequality in communication and in society is further perpetrated by sexist language. Benjamin Spock (1970) agrees that by using the male pronoun little harm is seemingly done but when language is looked at in its entirety, its impact at keeping women at a disadvantage is potentially enormous. Equality needs to be represented in the language of the people who seek equality. One's identity in communication is also linked to one's culture's language. Generic language affects identity in communication. Identity is the sense of self one has. Identity combines a person's social role and true self. Language has a strong impact on the development of girls and women, and this in turn influences their identity. Women's invisibility in language and communication patterns 13

PAGE 22

identify women as the less active agent in society. When a woman is identified through a masculine identity (masculine terms) rather than having her own feminine reference or a unity of language of neutral terms, she tends to view her identity as one of lesser status (Irigaray, 1985). Mascuiine is accepted as the norm. Male is equated with being human while female is equated only' with women (Silveira, 1980; EndersDragaesser, 1988). The masculine has become the norm in our communication and with it women's identity as peripheral and invisible. A woman may be referred to as one of mankind, working as a repairman while attending college part-time as a freshman. A woman's own verbal interactions will deny her own participation in her life. Women learn that their identity in communication is not as visible or as vital as men's. If visibility is a key factor to influence how a person is represented and ranked in society, women's identity is clearly underrated and subordinate to men's. Women's identity in generic language is omitted and makes them linguistic strangers in a world they are not a part of simply because they are female (Enders-Dragaesser; 1988). This identification connection is supported by memory recall research done by Crawford and English (1984). Memory recall has been investigated in connection with generic terms. Crawford and English ( 1984) tested 50 14

PAGE 23

females and 28 male college students and predicted male students would remember material better that was learned with "generic" masculine language and females would remember better with material that specifically includes them ("he" or "she," or "they"). These hypotheses were supported. The male good learners were able to recall better than the female good learners (good in comparison to bad learners in the study) after having read an essay using "he" pronouns. The females recalled more than the males on the "he or she" and "they" pronoun essays. The good learners were more likely to interpret the material literally rather than interpret it in a generic form as did the poor learners. If women can identify themselves in language, they are more likely to remember the material than if they cannot associate themselves with or in the language. It was shown that sensitivity to gender references affects performance in a memory task. The readers of both genders literally interpreted the generic terms. One's identity is grounded in words, and the dependency of people upon words works toward a semantic result binding language and reality. The standard language allows that women may be in the real world but does not allow the generic pronoun or noun to reflect this reality. Wharf's principle of linguistic relativity states that the structure of language affects our perception of reality, 15

PAGE 24

and reality reflects itself in our language. Language is an objective reality by which people structure and organize their external world through and because of our communication (Fishman, 1960). The thought process of communication directly relates words with an outward representation (Whorf, 1956). Women's identity is not communicated in generic language. Females may be less able to projectthemselves visually' into a sentence wheri "he" is used to refer to the antecedent than are males, thus causing women to have less identification and more exclusion in generic language (MacKay & Fulkerson, 1979; Silveira, 1980). The fact that represent women Generic language generic language does not accurately leads to the second stated problem. reinforces sex role stereotyping. A stereotype "denotes beliefs about classes of individuals, groups or objects which are preconceived, i.e. resulting not from fresh appraisals of each phenomenon but from routinized habits of judgement and expectation" (A Dictionary of the Social Sciences, 1964). Attitudes that result in oversimplifications of experiences or judgements are commonly referred to as stereotypes. The danger in stereotypes is due to a premature generalization which may distort attributes of the external world. Stereotyped thinking dramatically interferes with the perception of women in our society. Generic language stereotypes 16

PAGE 25

individuals into roles without regard to their gender. sex role stereotyping reinforces traditional role behavior in areas such as employment. Women are discouraged from entering or thinking of entering jobs that have masculine titles such as cameraman, flagman or weatherman. As children we are taught to make sex-category identifications; e.g., "policemen" are not women so girls may not grow up to think they can be police officers due to the job title. A mother who agreed to play "fire department" with her young son provided that she could be a firewoman was told by her child that she could not be a firewoman because there was no such thing (Farquhar, Dunn, & Burr, 1972). His books about fire trucks only talked about firemen riding them. Every child knows that only men are firemen. Children are educated by their learning materials in and out of formal schooling. Research indicates women may not be entering fields of work based on sex-biased job titles. Bem and Bem (1973) studied 60 women and 60 men and found job advertising language greatly affected the women and men who would apply for which jobs. Bern and Bern examined sex-biased ads and recruiting brochures using "telephone frame:n'tan" and "telephone lineman" (the terms "he" and "craftsman" were also included in the job description) and found the percentage of men and women interested in applying for either opposite-sex job varied greatly with each altered 17

PAGE 26

ad. The sex-biased structure (telephone frameman and lineman) produced no more than 5% female interest. When the same job was advertised in an unbiased structure (telephone frameworker and lineworker), 25% of the women surveyed were interested in the job. When this same ad was written in feminine terms (frame-woman and linewoman) the percentage rose to 45% of women surveyed. Jobs written in the feminine structure discouraged men from applying and when the ads were written in a sex unbiased or masculine sex biased structure men's interest rose from 30% to 75%. These results clearly indicate that jobs with masculine titles serve to aid and abet discrimination by discouraging women from applying for "opposite-sex" jobs merely by their masculine termed description. Briere and Lanktree (1983) researched 72 female and 57 male subjects on psychology as a future career for women and men using the APA "Ethical Standards of Psychologists" text with three different versions: the uncorrected 1972 version using generic "man"; version 2 in which "he" was replaced with "he" or "she"; and version 3 which used "she" or "he" in place of "he." Subjects rated psychology from all 3 conditions on its appeal as a future career for women and men using a 7 point Likert scale. Version 1 (exclusively male nouns and pronouns) rated a career in psychology as less attractive for women than did either version 2 ("he" or "she") or version 3 ("she" or "he"). The 18

PAGE 27

only different result of version 2 and 3 were when the feminine pronoun was shown first in version 3 ("she" or "he") the males were less likely to suggest to a male friend that he go to a psychologist. Generic masculine nouns and pronouns were connected with decreases in attractiveness for females but not for males. Even mild sex-biased wording affects subjects' perceptions of employment in psychology for women. Benoit and Shell (1985) discovered sex-biased communication can affect business students' career decisions. Using ACT's "Assessment of career Development" (ACD) instrument, questions referred to a person, unidentified by sex, in a work situation. There were both a neutral and a sex-biased version. The test was to examine if sex-biased communication had an influence upon career choice. The results showed that females picked 60 correct answers in the neutral condition and 37 correct in the biased condition. Males picked 78% correct in the neutral condition, but only 47% correct in the sex-biased condition. Males were unwilling to place a female (Mrs. Jarvis) in a nontraditional occupation (i.e. plumber). Subjects were more likely to choose jobs that were traditionally held by their own sex rather than nontraditional jobs. The author encourages educators to examine their teaching materials for sexist wording so all students can consider careers without regard to one's sex. 19

PAGE 28

Language can lend itself not only to employment problems but as the research indicates also communication problems. Language is an important tool to understanding both historical problems and human communication problems (Sapir, 1929). In summary, this chapter has examined linguistic relativity as a means of linking language to our perceptions of A historical view of masculine pronouns and nouns has been analyzed in terms of their original etymology. Due to changing times the language itself now contains sex bias, not necessarily in the purpose of the speaker. The ambiguity and lack of authenticity of generic terms lends itself to two problems: 1) women's invisibility in generic language has negative effects on self-esteem and identity; 2) generic language reinforces sex role stereotyping. 20

PAGE 29

CHAPTER TWO BACKGROUND AND REVIEW OF LITERATURE A goal of communication research is the identification and description of variables which have an effect on the communication process. Much research has been completed on generic language (Kidd, 1971; Martyna,1978; Hamilton, 1988). Researchers in speech communication have examined models from social psychology, linguistics and other disciplines to isolate variables which explain differences in listener perception and speaker intention. The rapidly expanding investigation of generic language reveals its importance in the communication process. Since the English language is central to our communication process and culture, this literature review examines studies that have English as its dominate language. While some foreign languages have neutral pronouns, the English language does not and it is this language usage in examination. Other studies have been omitted due to their investigation design that was not in line with the focus of this study. This includes those studies that did not find statistically significant results or where the design or result was not free from bias. It is beyond the scope of this review to comment on all theories, findings, arguments and counterarguments concerning generic terms in American culture and

PAGE 30

language. This review discusses those studies which are characterized by empirical soundness--i.e., where acceptance of social science methods, analysis and design were employed. Overview The background and review of the literature surveys five main areas of research: research conducted on 1) younger children (kindergarten through high school), 2) college students, 3) faculty members, 4) an argument for a direct link between correlation of education and generic language, and 5) gender differences. I have separated the first three areas into divisions by educational level and type of generic studied ("he," "man" or both). The reviews begin with kindergarten and elementary school, progress to high school, college and then faculty members. They include written and oral methods of investigation. studies examined concern kindergarten and first grade students response to the generic "he" (Fisk, 1985); grade schbol children's response to the generic "man" (Harrison & Passero, 1975); kindergarten through college level students response to the generic "he" (Martyna, 1978; Hyde, 1984); high school students response to biased job advertising (Bem & Bem, 1973); college students response to the generic "he" (Kidd, 1971; Moulton et al., 1978; Murdock & Forsyth, 1980; Hamilton, 1990); college 1988; Hamilton, Hunter, & Stuart-Smith, students response to the generic "man" 22

PAGE 31

(Schneider & Hacker, 1973); college students response to the generic "he" and "man" (Jacobson & Insko, 1985); and faculty members response to the generic "he" and "man" (Bate, 1978) The studies are arranged by the earliest date to the latest date within the division. The research theories remain relatively constant throughout the 1970's and 1980's. The.message concerning the generic pronoun "he" and the term "man" remain consistent in that they are not perceived as genuine generics. The concluding section then progresses to further empirical support which provides focused data pertinent to the conceptualization between language and education that lends itself to the research questions and purposed hypotheses. The literature review which is organized by education level of subjects in related studies starts by examining pronoun use at the first formal level of education (kindergarten) and advances through the grades, continuing to faculty members. The review examines both the correlation between education and generic language and when appropriate the gender of subjects. Harrison and Passero (1975) analyzed elementary children's responses to the generic "man." Children were shown drawings of stick people (gender identifiable) along with an explanation of their behavior. The descriptions were written with neutral terms (people, salesperson and 23

PAGE 32

hand-made) or masculine generics (men, salesman, and manmade). The responses to the generics ranged from 49% to 85% towards a masculine interpretation. It was found that "man" did not visually include woman 85% of the time in the statement: "If a man could wish upon a star to make his dreams come true, do you think he would?" Eighty-five percent of eight-year-old children presented with these generic examples circled only male in response. Harrison and Passero explain their results in an editorial style, "the extremely high response can be attributed to our culture s bias toward males. Any dominating element in the language that culture's highly valued elements. The authors continue their claim by noting "that a culture that is male dominated will create a male dominated language." The supposedly generic terms "man," "men" and "he" are permitted to include females but female terms may refer to people only if no possibility exists that a male might be included in the reference. An example of this is that a few years back (the early 1980's) people starting referring to elementary teachers as "she" since the vast majority of grade school teachers were females. Male elementary teachers and administrators strongly objected to this since some of the teachers were male (less than 10%). But masculine terms are allowed to include females when the ratio is higher for females than 24

PAGE 33

for males. Mankind consists of 52% females and 48% males but still the term is used to refer to humanity. The term "man" is also used for many prefixes and suffixes of words. The authors theorize that the language habits of the dominant group even if they are less in number create the language for the entire group (culture). Martyna (1978) conducted research on sentence fragments using 400 students of all ages from kindergarten through college level. Sentences in a written format were completed concerning male related roles (i.e., engineer), female related roles (i.e., secretary) and neutral roles (i.e., teenager) in connection with pronoun choice. Many filler topics were used (sports and weather) so attention would be focused away from generic pronouns. The.subjects were questioned afterwards for imagery responses. Both sexes reported receiving imagery of males in the malerelated topics and imagery of females in the femalerelated topics. A different trend of imagery appeared in the neutral sentence subjects. Sixty percent of male college students reported a male image when they chose generic "he" in a neutral sentence (i.e., human being, teenager, person). Only 10% of female college students reported any imagery when they chose these same terms and the imagery reported was 100% male. When females were questioned afterwards, they replied they had selected "he" because they'd been trained to or that it was an automatic 25

PAGE 34

response. The males replied, "I pictured males probably because I'm a male" and "I thought of myself." The pattern in this study leans toward males picturing male imagery in masculine pronouns and females choosing the male pronoun based on a grammatical standard of correctness. Martyna did not report the ratio of females and males but did indicate a sex difference occurring with females using generic "he" less often than males regardless of antecedent. Hyde (1984) conducted research on 310 subjects ranging from first graders to college students. There were a total of 140 males and 170 females. Each child was individually given information about a character and asked to create a story about this character. Three different pronouns were used to describe the character by the experimenter: "he, 11he" or "she" or "they." The subjects were also asked if they knew the generic pronoun rule. Information given about characters using the masculine pronoun "he" resulted in 88% of the stories being about males. Not a single first grade boy created a story with a female character. Third grade boys created no females characters with the pronoun "he" but females characters increased to 30% when the alternative "he or she" was used to originally describe the character. In response to the standard grammar rule of the prescriptive "he," knowledge of the rule increased with 26

PAGE 35

age: 28% of first graders, 32% of third graders, 42% of fifth graders, 84% of college students replied they knew the rule but disagree with it-"'yes, he refers to males so one should use "he" or "she" to be clear that everyone is included.'" A recent investigation by Fisk (1985) found subjects did not perceive the generic "he" as gender neutral. Subjects were 36 and 36 kindergarten students.-Half of the subjects were girls and half were boys. Children were told a story using either "he," "he" or "she" or "they. 11 Each story's wording was the same with the exception of the alternative pronouns. The child was asked to retell the story, and his or her pronoun usage was recorded. The child was also asked to pick either a boy picture or a girl picture to represent the character in the story. The results indicate the children did not respond to the pronoun "he" as if it were a generic "he." In the group that received the story with the pronoun "he," 22 out of 24 subjects retold the story using the pronoun "he" and picked the boy picture (the other two did not respond) No sex differences occurred in this condition. This research corresponds with previous research in showing male-biased responses to the use of "he" in what is intended to be a neutral presentation. The findings for the "they" pronoun produced 22 out of 24 children retelling the story using the "they" pronoun. 27

PAGE 36

Of the 22 subjects, seven chose the boy picture, nine chose the girl picture and six said both. The sex differences varied greatly in the "they" presentation with all six of the kindergarten boys choosing boy pictures. The girls in both levels chose a girl or both picture and only one first grade boy chose a boy picture while the rest of the first grade boys chose girl or both pictures. The "he or she" presentation netted a variety of pronoun usages in retelling the story; two "he," five "she," four "he" or "she," eight "they," three mixed. The pictures chosen were five boy, eleven girl, and six both. No sex differences occurred in this group. Since only boy and girl pictures were presented, the both responses were probably underestimated. During the "he or she" and "they" presentations, the children looked back and forth at the available pictures prior to choosing. The subjects who said both must have been certain of the answer because there was no both picture to choose. Many of the children were likely obeying the experimenter and chose one of the pictures when they may have wanted to pick both. This response is especially key because socialization has deep impact after just a few years. Fisk concluded that by the age of six for girls and the age of seven for boys, children respond to pronouns in a learned manner rather than just assigning their own sex to the subject if the reference is vague (they) or mixed (he 28

PAGE 37

or she). Boys until the age of seven could only respond to the pronoun "he" with a masculine interpretation. It is also suggested that girls do not go through the same process as boys in responding to pronoun presentations. The research on grade school children reveals a pattern: the younger the child the less likely the child will know and use generics neutrally. Children tend to use masculine generics to include only males rather than males and females. As elementary education continues children are more apt to learn and use masculine generics neutrally. This relationship between generic usage and education level continues through college in the studies just reviewed. Studies involving only high school and college students perceptions of masculine generics are now examined. Kidd (1971) conducted an important first empirical study on sex bias in language use. Sixty-eight college students were asked to openly respond to three questions pertaining to pronouns. 1) Do people respond to generic pronouns as though the antecedents are neutral? The first question received an emphatic no in response to people viewing the generic pronoun "he" as neutral. Subjects responded with 407 male image responses and 31 female image responses to a visual image of the generic "he." 2) If forced to respond in terms of gender, do people describe the antecedent as "male" and "female" in equal proportion? The second question forced the subject to respond with either a male 29

PAGE 38

or female antecedent in a neutral generic sentence. "Male" was chosen 526 times over "female" being chosen 53 times. 3) If some pronoun usages evoke heavy male responses and some strong female responses, what consistencies can be uncovered in the nature of these usages? The third question concerned societal roles connected with pronouns. The students responded by choosing masculine images for the active roles and administratively superior positions. Women were chosen for the "emotional child, the teacher and the consumer roles." "Other" as an answer surfaced in the abstract and unspecific roles. These results coincided well with traditional stereotypes of the male and female roles. As previously discussed, Bern and Bern (1973) examined 120 senior high school students response to booklets containing sex-biased job advertisements. They found job descriptions written with masculine generics (telephone lineman or frameman) produced no more than 5% female interest. The same job advertisements were changed to include neutral terms (telephone lineworker or frameworker) and female interest rose to 25%. When the descriptions were rewritten using feminine terms (telephone linewoman or frafuewoman) women's interest rose even higher to 45%. Clearly these results emphasize a trend of masculine generics excluding females. Schneider and Hacker (1973) tested 306 college students 30

PAGE 39

for their interpretation of the generic term "man." No sex breakdown was given. students were asked to supply pictures to represent chapters in an introductory sociology textbook. Three separate schools complied with this request. Two dependent variables were employed: one form presented chapter titles containing the term "man," and the other form without the term. All pictures were coded by gender: "males only," "females only," "both males and females," "no people present," "sex indistinguishable" and no picture submitted. overall, 64% of the students who received the man-linked titles submitted pictures containing only males. The chapter titled "industrial man" received the highest number of male only pictures with a percentage of 87%. Rarely in any of the titles either man-linked or otherwise wore only women shown. The results indicate the tendency to perceive males rather than people or women in the supposedly generic term "man". Moulton et al. (1978) tested 226 male and 264 female college students response to the generic "he.11 They were given written instructions to create a character who would fit in one of the themes given. The themes consisted of neutral subjects, either a student or a person used in a sentence with either the pronoun 11his," "his" or "her" or "they." students were told not to write about themselves. Results found that when the pronoun "his" was used, 65% 31

PAGE 40

of the story characters were male; the "his or her" instructions netted 44% male characters and the "they" instructions netted 54% male characters. students had to name their characters as well and use a pronoun to describe them so that it could be clear of which gender the characters were. These results continue the trend that the u'sing the .Pronoun "his" increases the possibility that people will think of a male even in contexts that are intended to be neutral. Generic terms are not necessarily gender neutral in interpretation. Research by Murdock and Forsyth (1980) examined reactions in a naturalistic context by asking 47 men and 87 women adult students (college level) to examine. a written essay using all plural pronouns, the generic pronoun ( s) and evaluate stereotypical phrasing. They also evaluated sexism in sentences that either contained no sex bias, the generic "man" or "he," or referred to women in an unfair, stereocast manner. Their results showed that generic phrasing ("businessman" and "he") and stereotyping (mothering or manhandled) were perceived to be biased and sexist. The neutral alternatives such as police officer, the plural pronoun "they" or when man and woman were both used were the only terms deemed nonsexist. Jacobson and Insko (1985) surveyed 234 female and 164 male college students for pronoun preferences in 32

PAGE 41

traditionally masculine occupations (i.e., accountant or physician} stereotypical feminine positions ( L e. secretary or nurse} and titles considered to be generic (congressman or spokesman}. Even though man related terms have historically been associated with men more than women they are still considered generic. While men gave more "he" responses than females overall, the biggest finding concerned the choices made on the generic terms. The generic terms had the fewest "he/she" choices even over the masculine items. The researchers concluded that the ending "man," while it is suppose to refer to both men and women "apparently induces subjects to expect the individual to be male" and thus they picked "he" as the correct pronoun for these terms. The researchers did not report the sex breakdown but claimed subject sex played a significant predictor on both masculine and feminine responses. Males chose "he" more often than females on masculine items and "she" more often on feminine responses rather than "he or she" on either items. Current research by Hamilton (1988} studied masculine generics in connection with linguistic relativity (that language can shape thought}, prototypicality (the most common "he" is probably a man} and activation of multiple meanings ("he" has both gender-specific and gender-neutral denotations and both may be activated even when the genderneutral meaning is intended}. 33

PAGE 42

Sixty female and 60 male college students participated in four experimental conditions: traditional vs. modern instructions, (used to induce uses of "he" vs. "he" or "she" or "they") crossed with plural vs. singular sentence fragments. The varying instructions were to measure if variation in the content of imagery or in name assignment was caused by the instructions. Subjects were required to fill in sentences with the pronoun "he," "he" or "she," or "their." The imagery questions asked for a description of any image the subject had for the person in the sentence. This included the person's age, sex, activity, etc. The subjects also had to think of a name for the person who would fit in the sentence. Significant results indicated that people's imagery excluded females twice as often as it included females and male names were given two and one half times more often than female names. Male subjects displayed more male bias than female subjects. No other interaction effects were significant. This study supports other findings that state generic terms generate more male imagery in the mind of the user than does using unbiased or true generics Passero, 1975; Martyna, 1978 & Fisk, 1985). (Harrison & This study indicates that a speaker may not, on some level, know what he or she is saying. Someone who is told to hire the "best man for the job" may be influenced to hire a man. These 34

PAGE 43

findings extend themselves to the Whorfian Hypothesis (Whorf, 1956). Language assists in shaping, allowing and encouraging certain thoughts while discouraging others. Masculine generics discourage people from perceiving females in their imagery. The link between difficulty in mental inclusion and the exclusion of women in further thought and action could potentially be made. There is empirical evidence for this exclusion in sex biased job advertisements (Bem & Bern 1973) and on the effects of masculine generics concerning a self-defense verdict in court (Hamilton, Hunter & Stuart-Smith, 1990). Hamilton, Hunter, and Stuart-smith (1990) presented 72 college students with one of three written versions of jury instructions examining a murder case. The subjects who read the generic editions of "he" were less likely to believe that the female defendant took action in selfdefense than those who read one of the alternative terms conditions of "he" or "she" or "she." The research conducted on high school students and college students is consistent with some of the research done-on younger children. However, while younger children's interpretation of generics increased with each grade level, college students masculine interpretation declined or maintained during their college years. High school female students showed more interest in jobs written in feminine terms (i.e., linewoman) rather than in generic 35

PAGE 44

terms ( L e. lineman) College students that were forced to choose either a male or female antecedent in response to the generic pronoun "his" chose male 99% of the time. College students response to creating a fictional character produced 65% male characters in response to the generic pronoun "his." Adult students perceived man related terms (i.e., businessman) as biased and sexist and only neutral terms (those without the term "man" in them) were deemed nonsexist. The college studies examined reveal a positive correlation between education and one's perception of generic terms. As one's college education level increased, masculine perception of generic terms increased. Does this positive correlation continue beyond college? Research on faculty members is now examined as the last level of education in connection with generic language. Bate (1978) interviewed 10 female and 10 male faculty members concerning their level of comfort with gender specific and gender neutral terms chair), their pronoun preference in (i.e. ,chairman written and and oral contexts and their opinion of generic terms as sexist. Eight female faculty and nine male faculty picked the "he" or "she" usage in written form in response to a sentence fragment "a person .. "They" was more widely chosen for verbal use. Martyna (1978) found these same results in her investigation. She also found females used alternatives to the term "he" more often than males. 36

PAGE 45

Bate presented cards with different alternatives on them for faculty to chose which terms they were most comfortable using. The term "chairman" received positive support with six females and eight males choosing this term. The term "chair" received support from six females and seven males which made them almost equal in language usage (14 to 13 respectively). The term humanity surpassed mankind 20 to 11 (15 females and 16 males). Many of the faculty felt they were in transition about the terms. A male faculty member said he was eliminating the term "man" from his speech out of respect for his female colleagues. Another male faculty said he was open to arguments against such language and would change his speech patterns if his social setting changed. The faculty members were not as likely to view the. man related terms as sexist as much as limiting opportunities and affecting self concepts of females. This study introduces a new perspective on the perception of generic language. In review, it ws found that the younger the child the less likely he or she will be to know generic rules. As education level increases, knowledge and use of generics become formalized and regulated. This trend increases until college where it maintains or declines. While students perceive degrees of inclusion and exclusion of females no other reactions are noted and it is only at the post-education level (M.A.,Ph.d. and faculty levels) that potential for bias enters. Faculty members 37

PAGE 46

found that generic terms affect self concepts and limit opportunities for females. While words appear objective because they name and define, other ramifications of words go unnoticed such as self-image and the role modeling process of language and communication. Summary The studies reviewed concerning generic language demonstrated that "he" and "man" have a masculine interpretation more often than its intended neutral interpretation. Studies completed on children drew similar conclusions (Harrison & Passero, 1975; Martyna, 1978; Hyde, 1984; Fisk, 1985). Kindergarten through elementary school aged children are more likely to interpret generic terms with a male antecedent rather than a male and female antecedent. studies done on college students drew similar conclusions as those done on younger children. As education level increased, the knowledge of the rule increased and while 84% of college students knew the rule 6% disagreed with its use and meaning. It was suggested that a speaker may use a generic term without realizing its bias towards males and its exclusion of females. Little research has examined faculty members as subjects. Bate (1978) did conduct one such study and found that faculty were more likely to choose "he or she" over the 38

PAGE 47

generic "he" in written form and "they" in verbal interactions. Faculty did not feel comfortable using "a person he" because if failed to include females. Faculty were not comfortable using "they" in written form because it was not standard English. However, in conversation "they" was selected as a modest way to avoid choosing the gender of a subject. Bate concluded that many speakers are willing to make changes to their language habits if they have information to support its change. Faculty members were also the only group that perceived the affect of generic terms on peoples' self-image. This would include either interpersonal or professional situations warranting change. The masculine generics controversy continues to increase people's awareness of the advocacy for change. Overall, general findings in the literature suggest that the terms "he" and "man" are not producing neutral responses throughout all levels of education and beyond. Substantial numbers of subjects are responding toward a masculine interpretation instead of a masculine and feminine interpretation. Research that more directly correlates education and generic language is examined next. A Correlation of Education and Generic Language Though education and generic language have been looked at indirectly (Fisk, 1985; Harrigan & Luc.ic, 1988; Hyde, 1984; Hughes & Casey, 1986; Martyna, 1978; Nilsen, 1977), 39

PAGE 48

one area that has not been directly examined is the association between one's level of college education and the propensity to view the terms, "man," man related terms and the pronoun "his" as generic. There is logical reason and empirical evidence to suggest that there should be a connection. The logical reason that links one's level of education and one's understanding of generic language is that children prior to grade school (under five years of age) literally interpret generic language ("he" and "man") towards the masculine meaning (Hyde, 1984). The male pronoun always has a male antecedent for children (he=male) (Fisk, 1985). This male imagery starts at a young age and continues through adulthood (there have been no studies done on children less than 5 years of age). Children literally interpret terms that refer to the male gender (he, paperboy, man, fireman, husband, etc.) as having only masculine meanings. Once children are taught masculine, feminine and neuter nouns and pronouns then they are taught the generic use and meaning. This is where confusion begins. Hyde (1984) found only 28% of first graders knew and used the grammatical rule of generic Other results in this study that link education and generic language use were that 32% of third graders, 42% of fifth graders and 84% of college students knew and used the generic "he" but an additional 6% protested against its use 40

PAGE 49

even though they knew the rule but disagreed with it -for example, "yes, he refers to males, so one should use "he or she" to be clear that everyone is included". Martyna (1978) found that males were more likely to choose male pronouns due to male imagery while females chose male pronouns because they had been formally trained (in the academic environment). As education level increases one's knowledge and use of generic terms increase due to what one has been taught in school. If one is taught that generic terms ("he" and "man") are not gender neutral but gender specific, one would be less likely to view and use the terms generically (due to influence of academic learning). Therefore the more college education one has the less likely they would be to view terms generically. There is empirical evidence to support this hypothesis. Empirical evidence supports the link between education, generic language and influence of authority. Hughes and Casey (1986) found this influence of authority (education in this case) to be responsible for the predominance of masculine pronoun use and that students who were taught to avoid "he" or "she" or "he/she" usage were also taught "he" was a universal or generic pronoun. Harrigan arid Lucie (1988) studied 5 groups: members of a local NOW, faculty members, medical students, English graduate students and psychology graduate students. Subjects completed a questionnaire concerning gender bias in our language. 41

PAGE 50

They found that the influence of other people encouraged pronoun use. The groups most influenced by other people in authority were the threestudent groups. It was found that the student groups were more likely to use a new gender neutral pronoun when a person in authority made the request. Females were a bit more influenced by an authority influence than males. Since students are the most influenced group, what educators teach in our schools regarding generic language is definitely worth research. Teaching alternatives to masculine pronouns and nouns in education is not only worth the effort but studies show that what is learned in early education concerning language rules stays with the student throughout his or her life (Nilsen, 1977; Henley et al., 1988). Boys as well as girls need to learn alternative terms so that the terms may be genuinely inclusive. It is through education that this language/ communication change can occur. Lastly, research on gender differences is examined. Gender Differences Another area that has been relatively ignored is the possible association between gender and the propensity to view the terms "man," man related terms and the pronouns "he" and "his" as gender neutral rather than gender specific. 42

PAGE 51

Sex differences reported in the studies are logical given that males and females learn communication patterns differently. Males learn feminine and masculine terms (pronouns and nouns) and then use masculine terms as generic terms. Males may add a female interpretation to masculine terms to make them generic. Males may also just continue to use masculine terms without consciously including females but with the knowledge that masculine terms are more commonly used as generics rather than alternative terms (Nilsen, 1977). It is less important that males know when a term is inclusive of females because these same terms always include males in any given context. Generics are only gender neutral in a defined context, out of that context they retain their original masculine interpretation. Males can project or image themselves in generic terms without changing their original meanings. Females, on the other hand, learn feminine, masculine and generic terms as three separate units. Females have learned that masculine terms apply only to males and when these same terms become gender generic to include them (females) they must relearn the contextual meanings so that they are included. They also must learn when to include themselves and when not to (when the term is used generically and when it is not). Some empirical evidence for these learning differences exist (Fisk, 1985; Henley, 1989; Nilsen, 1977). Fisk {1985) found that males and 43

PAGE 52

females learn pronoun usage at different ages with girls learning them at an earlier age than boys. Empirical evidence concerning sex differences is divided. Research examining sex differences includes early research by Martyna (1978) found that 60% of male college students were able to perceive themselves in generic language while only 10% of females reported imagery. Men imaged either themselves andjor males in generic nouns while women had little imaging of any person at all. Men have learned they are included in masculine and generic terms but women do not feel included in either. Current research by Hamilton (1988) found male subjects tested using generic terms displayed male bias twice as often as female subjects. Men were more likely to visualize themselves in generics than were women to include themselves in generics. Benoit and Shell (1985) found males were more likely to pick correct answers in a sex-biased condition than females in a sex-biased condition (47 to 37 respectfully). Males were also more apt to stereotype people into traditional jobs than females were. Fisk (1985) found that the younger the male child was the more likely he would pick a picture of a boy child for a story using "they" as the only pronoun. Female children were able to pick female & male pictures but 5 years old boys could only project themselves or another male self 44

PAGE 53

into the story. Research that has not reported a sex difference or noted that one did not occur in their study includes research by Harrison and Passero (1975) on elementary students. Kidd (1971) tested college students and found more male images to female images in generic "he" use (407 to 31 respectfully) but did not test for sex differences. Jacobson and Insko (1985) found that invisibility of women in man related terms and the term "man" exists but did not give any breakdown of sex differences. Murdock and Forsyth ( 1980) tested reactions to gender biased language (masculine generics) and found them to be perceived by subjects as biased and sexist but sex differences were not reported. Among other studies covered in this paper that did not report or test for sex differences are Scheider and Hacker, 1973; Hughes and Casey, 1986; Harrigan and Lucie, 1988. Either the sex differences were not significant enough or there has not been enough research on them. Either way, research is lacking in definite support of or in opposition to gender differences. This literature review answers some questions concerning generic language but others remain open. What is the effect of college education on generic usage? Do male and female college students differ on their interpretations of generic language? Is there a preference toward a masculine interpretation at different levels of college education and 45

PAGE 54

if so, at what level? The next section introduces the research questions and hypotheses proposed. Proposed Research Questions and Hypotheses The lack of research on education and generic language supports the importance of the present research. Generic language is learned communication that must be learned in a specific environment. Formal education is one such environment. This leads to the first research question. Research Question One: What is the connection between post secondary education and the use and understanding of generic language? Hypothesis One: The more college education individuals have the more likely they will view the terms "man", man related terms and the pronoun "his" toward a more masculine interpretation. There is logical reason and empirical evidence to suggest that some connection exists between gender and the propensity to view generic terms as masculine rather than gender neutral although neither reason nor evidence is definitive. Gender differences are feasible and this possibility lends itself to research question two and hypothesis two. Research Question Two: What is the connection between 46

PAGE 55

one's gender and one's use and understanding of the generic terms, "man", man related terms and the pronoun "his"? Hypothesis Two: Males will view the generic terms "man," man related terms and the pronoun "his" toward a more masculine interpretation than females at the same educational level. Hypothesis One states that the more college education one has the more likely one will view the generic terms as masculine. Hypothesis Two theorizes that males will view the terms more masculine than females. A combined effect of the independent variables (college education level and gender) of the two hypotheses may occur. 47

PAGE 56

CHAPTER THREE METHODS The questions examined in this study concern perception of the generic terms: "man," related terms with "man" as the prefix or suffix and the pronoun "his" and their relationship tO One IS gender and One IS leVel Of COllege education. This chapter will examine the investigation methodology including: 1} methods-the survey design 2} measurement instrument-the independent and dependent variables 3) subjects 4) data collection procedures-the pilot study and written survey 5) analysis of methodology-data coding and statistical procedures chosen. The choice of research methodology was based on other researchers traditional use of a written survey in this area of study. The use of a written survey allowed for a quicker response time and a greater number of responses rather than conducting personal interviews. The use of a survey method is based upon its effectiveness in gaining focused information about a well defined social phenomena. The data collected by way of survey enabled information to be gathered with relative ease from a large number of people in a defined population (university setting) A number of limitations and drawbacks are recognized in

PAGE 57

the use of a survey method. These limitations include the following. First, there is no control over unknown variables affecting the results. Second, the potential force that social influence may have exerted over respondents. Generic langauge is more socially accepted as the norm and this factor may have caused respondents to rate the terms toward a more neutral intrepretation than they actually believe. Respondents may have chosen answers they thought were socially acceptable rather than their own opinion or perception. Third, the written survey asks about issues that are lifted from a defined context. In interpersonal conversations, one is able to assess issue meanings from the conversation, persons involved and by asking for further clarification. The survey does not define or clarify words or issues for the respondent. The words are not in any given context but in a general context. Measurement Instrument Independent Variables Variables were measured using a six page questionnaire (see Appendix A). The independent variables were gender and college education level. Gender was indicated by a mark on the line next to the male or options. College education level was female response indicated by a written reply or a circle around the printed response 49

PAGE 58

options of freshman, sophomore, junior, senior or graduate student. Freshman, sophomore, junior and senior responses were combined into the undergraduate category. Faculty members were given a separate information sheet and asked to indicate the year their degrees were earned. Dependent Variables The five dependent variables consisted of distinct uses of the masculine generic terms "man" and or "his". The five dependent variables were labeled "pre, suffix, both, man and his". The variable "pre" consisted of five terms with the prefix "man": manpower, mankind, manhood, manhours, and manned each used in a sentence (see questions 2,11,13,14 and 21 in the survey). The variable "suffix" consisted of five terms with the suffix "man": baseman, caveman, policeman, layman, and chairman each used in a sentence (see questions 4, 6, 8, 9 and 2 4) The variable "both" consisted of the five terms foreman, weatherman, congressman, spokesman, and salesman each used with the pronoun "his" in the same sentence (see questions 12,16,18,20 and 25). The variable "man" consisted of the five terms average man, man, working man, dirty old man, and good men each used in a sentence (see questions 1, 7 17 19 and 2 2) The variable "his" consisted of the five neutral terms student, customer, student, professor, and citizen used with the pronoun "his" in the same sentence (see questions 3, 5,10,15 and 23). 50

PAGE 59

Subjects Respondents for the study were students and faculty members from the University of Colorado at Denver and Metropolitan State College. There were 68 undergraduates, 53 graduate students and 57 faculty members who participated in this research. There were a total of 30 male undergraduates, 27 male graduate students and 29 male .faculty and 38 female undergraduates, 26 female graduate students and 28 female faculty. The undergraduate and graduate students were recruited from various communication and education classes. Faculty members were chosen by the researcher upon chance visits to faculty and department offices. An attempt was made to get faculty members with different educational backgrounds (LAS, business, EE, etc.). This was an attempt to get faculty from fields other than the humanities. There was no attempt made to obtain a racially representative sample. The majority of the subjects were Caucasians, but there were some Hispanics, Blacks and_Oriental subjects as well. Data was collected over a three month period. Data Collection Procedures Pilot study A pilot study was given to a class of 45 undergraduate and graduate students. The instructions indicated that this was a probability study dealing with general knowledge. 51

PAGE 60

Students were to read the question and indicate to what degree was its likelihood or probability. Gender was not mentioned as an operating factor. There were 100 items with 25 generic usage sentences and 75 fillers provided in an attempt to mask the generic terms. The researcher asked the students to give their responses to the true nature of this survey and found this type of survey failed to mask the generic sentences. The students guessed the intent to have something to do with gender and the "man" terms. The survey was then redesigned using only the 25 generic terms. In comparing the results of the pilot survey and the final survey given, no differences respondents' answers. Survey Procedures were discovered in Respondents completed the final form of the written survey containing 25 questions given to them by myself or their professor. Verbal instructions given were only that this study was for a graduate thesis and concerned gender. The entire questionnaire packet contained six pages. The consent form and instructions were given on the first page of the survey. It was presented as a probability study dealing with gender. Questions from the survey were written in this manner: "How likely it it that the term layman refers to the everyday female?" The questionnaire used well known man-related terms in an attempt to gain understanding of individual interpretations of generic 52

PAGE 61

terms. It asked the respondents for their personal opinion. It was also stated that there are no right or wrong answers. A sample question was presented on page one. The second page contained personal questions such as age, sex, level of education, major and other factors to be correlated with the responses to the questions. Pages three through six contained the questions concerning masculine generics. All questions were randomly arranged in the survey. A 7-point Likert type scale was used to measure probability or likelihood for each question. Please see Appendix A for the survey. Analysis Coding The data collected were coded numerically. Sex was coded 1 for male and 2 for female. Level of college education was coded 1-freshman, 2-sophomore, 3-junior, 4-senior, 5-graduate and 6-faculty. Dependent variables were coded on a 7-point Likert scale with 1 being the most gender neutral and 7 being the most masculine. Non-responses were coded as 9's. Statistical Analysis The credibility of a study rests heavily on the appropriateness of the statistical analysis used. The use of the wrong statistical test can lead to false conclusions and speculation that call the findings into question. 53

PAGE 62

Every statistical test has certain conditions for its appropriate use. A test that allows the researcher to draw specific conclusions from the data is the more powerful and therefore more desirable test. The term power used here is based on the mathematical ability of a statistic to correctly test a hypothesis. Three considerations that directly influence one's choice of statistical test are: 1) sampling methods used 2) nature of the research population 3) level of measurement for the variables. The most important of these criteria for this study is the level of interval measurement of variables. Hypothesis One states that the more college education one has the more likely one will be to view the generic terms as masculine. Hypothesis Two states that males will view the generic terms more masculine than will females. Both hypotheses used independent groups t-tests employing education for Hypothesis One and gender for Hypothesis Two. An analysis of variance (ANOVA) was used to test for main effects and any interaction effects. The statistical tests chosen for this study include the following. Pearson Correlation was used to measure the relationships between the dependent variables. The reliability test employed Alpha to test for stability among dependent variables. The independent groups t-test was used to measure the difference between group means for dependent variables. The test was employed 10 times-one for each of the five 54

PAGE 63

dependent variables times 2 for the two independent variables of gender and education. The analysis of variance (ANOVA) was used to test for main effects and any interaction of the two independent variables. ANOVA was employed five times. The Pearson Correlation was employed to provide a numerical indication of both the strength and the direction of any relationship among the five dependent variables. Correlation coefficients range along a continuum from -1.0 (perfect negative) at one extreme to a +1.0 (perfect positive) at the other extreme with 0.0 (no correlation) at the midpoint. The closer the numerical value of the correlation coefficient is to either extreme the stronger the relationship between the variables. The closer the coefficient is to the midpoint the weaker the relationship. A correlation of 0.0 suggests a lack of a relationship. The plus or minus sign indicate the direction of the relationship. A positive relationship is indicated by the absence of a minus sign while a minus sign indicates a reversal or negative relationship. The Pearson R was the logical choice given that early on in this inquiry it was observed that the five dependent variables may have constituted only one dependent variable. The Pearson R was needed to reveal the relationships among the variables. It was speculated that the dependent variables were distinct but the extent of their 55

PAGE 64

independence was left open to statistical test. Reliability of measurement is used by social scientists to test for the degree of stability or generalizability of empirical findings. The search for reliability raises the question of how stable the variables can be when observed, measured, repeated and inferred. The results are interpreted similar to the Pearson Correlation Coefficients with +1.0 indicating a strong positive relationship, -1.0 indicating a strong negative relationship and o.o indicating no relationship. The reliability coefficient was attained in this study by comparing each of the individual items with the total score of the five dependent variables. The coefficient interpretation is based on the assumption that the survey setting was the same for each instance. The results will be the same when the conditions are constant. A high coefficient supports this assumption while a low coefficient denies it. A reliability test was necessary to determine if the relationships among variables were due to true differences and how much was due to inconsistencies in measurement. The independent groups t-test was chosen to test the hypotheses because it can be used on a relatively small sample size, the groups do not have to have an equal number of cases and it employs interval level data. The t-test is an analysis of the amount of difference between group 56

PAGE 65

means. The t-test used was a 1-tailed test because the study's hypotheses were directional in nature. The independent groups t-test was employed to help test the significance of the differences of means for the groups associated with each of the independent variables values. Hence, a t-test was conducted on gender and college education level groups and their relationship with the five dependent variables. The analysis of variance (ANOVA) is the technique that separates the differences or variation of independent variables into separate means units. The ANOVA measures the differences between independent variables group means. It compares within group variation of scores and the variation between groups. These two variance measures are then analyzed to test hypotheses. The ANOVA was the appropriate choice to test for main effects and especially for any interaction between the independent variables. This study's data level had the two elements necessary in a two-way ANOVA test: 1) the dependent variables used interval data 2) there were two or more independent variables being investigated. In summary,, the data tested 178 subjects' response to their perception of the generic terms "man", man related terms and the pronoun "his" and their relationship to one's gender and one's level of college education. The survey instrument was pilot tested resulting in 25 unmasked 57

PAGE 66

questions concerning generic terms for the final questionnaire. The data were collected over a three month period. Data were coded numerically and statistical tests employed were: Pearson Correlation for the relationship between measures, Alpha Reliability test for stability among dependent measures, independent groups t-tests for gender and college education level, and the ANOVA for main and interaction effects. Next, the results of these statistical tests are given and examined. 58

PAGE 67

CHAPTER FOUR RESULTS This study was designed to examine subjects' use and understanding of generic language. Hypotheses proposed that those individuals with more college education would be more likely to view generic terms toward a more masculine interpretation than neutral interpretation and that males over females would view these same terms more masculine than females. Two independent variables, gender and college education level were tested as predictors of five dependent variables of generic terms. Statistical support was found for the relationship between education on and sentences that contained both the suffix "man" and the pronoun "his" together. These results indicate that graduate students rated the terms more masculine than did undergraduates. Both groups rated the terms toward a more masculine bias overall. Marginal support was found for a consistently masculine bias by male graduate students over all the other groups in one other dependent variable. While the two hypotheses tested with five dependent variables were not statistically supported overall, the results provide minor support for the often cited claim that the use of generic terms lead to a greater likelihood of a masculine interpretation. The results of the statistical procedures employed are

PAGE 68

examined and statistically significant emergent patterns of the variables are depicted at length. The discussion of results begins with the report of the Pearson Correlation Coefficients for the dependent measures (see table 4.1). These are followed by an examination of the internal reliability for the five dependent measurements (see table 4. 2). T-test results on the separate effects of college education level and gender are reported next (see tables 4.3 & 4.4). Lastly, the discussion examines analysis of variance (ANOVA) results showing main and interaction effects. Please see the complete ANOVA results on tables 4.5 -4.9. Internal Measurement Reliability Pearson Correlation Coefficients The dependent variables were correlated with one another in an attempt to discover if a single construct was being measured or if the items constituted five different variables (see table 4 .1). The variable "pre" had the weakest relationship with the variable "both" at 5803. The strongest variables relationship test occurred between the variables "pre" and "suffix" at 8010. The other variable tests measured between .5803 and .8010. A strong to very strong relationship between each variable was indicated by the data. However, since only one pair of variables reached the .80 level, alternative tests for a 60

PAGE 69

single construct must be considered independently (Betty, 1990). Only the "pre" and "suffix" variables are sufficiently statistically correlated as one construct. This suggests that different constructs exist here. However, the remaining variables are highly correlated. Table 4.1 Pearson Correlation Coefficients Pre Suffix Both Man His Pre 1. 000 Suffix .8010 1.000 Reliability Coefficients Both .5803 .6457 1. 000 Man .7407 .7659 .5966 1.000 His .6306 .6248 .7215 .6306 1.000 The consistency of single variables in relationship to the total score of the five dependent variables were compared and attained reliability coefficients ranging from a low of .7713 to a high of .8947 (see table 4.2). This indicates a strong to very strong reliability relationship between individual scores and the total of scores on all five measurements. measured. 61 There were 17 8 cases

PAGE 70

Table 4.2 Reliability Coefficients N of cases = 178 N of items = 25 Variable Measured Reliability Pre .7850 Suffix Both Man His Total of 5 items = .7937 .8852 .7713 .8947 .9071 Hypotheses Testing The following hypotheses have been proposed: Hypothesis One: The more college education individuals have the more likely they will be to view the terms "man," man related terms and the pronoun "his" toward a masculine interpretation. Hypothesis Two: Males will view the generic terms "man," man related terms and the pronoun "his toward a more masculine interpretation than will females at the same educational level. Hypothesis one Hypothesis One was statistically supported in one of the five dependent variables tested. Using the t-test to compare undergraduate students with graduate students the 62

PAGE 71

"both" dependent variable reached the 05 significance level on a 1-tailed test. The variable "both" consisted of five questions each combining the suffix "man" and the pronoun "his" in each sentence. Faculty rated the terms slightly more masculine than students on every variable with the exception of the "his" variable. No findings were statistically significant for t-tests completed on faculty and undergraduate or graduate students. See table 4.3. 63

PAGE 72

Table 4.3 Education Results T-Test for Undergradates & Graduates for each Dependent Variable Pooled Var Est Variable # cases mean sd se t value df 1-tail prob pre undergrad graduate suffix undergrad graduate both undergrad graduate man undergrad graduate his undergrad graduate 68 53 68 53 68 53 68 53 68 53 4.16 4.27 1. 09 .133 1.10 .152 -.55 119 4.44 1.21 .148 -.29 4.51 1.22 .168 119 4.85 1.35 5.24 1.23 5.00 1.14 5. 07 1.10 4.72 1.59 4.45 1.28 .164 -1.63 119 .170 .139 -.34 119 .152 .193 .176 .99 119 .290 .385 .050 .367 .156 The mean of graduate scores on the "both" variable was 5.24 in comparison to undergraduates' mean of 4.85. While this may not be a practical difference in mean scores, statistical difference was reached so these findings warrant attention and further research. These results 64

PAGE 73

indicate that graduate students were more likely to perceive the generic terms as masculine than were undergraduates. While other results were not statistically significant at the .05 level, this pattern throughout the education factor continued. The graduate students rated the generic terms more masculine than did the undergraduates, with one exception. Overall both the graduate and undergraduate students rated the terms toward a more masculine interpretation rather than gender neutral (4.16 was the lowest rating with a rating of point 1 being the most gender neutral and point 7 being the most masculine) Hypothesis Two Hypothesis Two was not statistically supported in any of the five dependent measures using the t-test. However, the results do consistently illustrate males rating the terms more masculine than females. results. 65 Please see table 4. 4 for

PAGE 74

Table 4.4 Gender Results pooled variance est. variable number of cases mean sd se t value df 1-:tailed prob pre male 86 4.36 1.15 .124 1.19 176 .117 female 92 4.14 1.25 .131 suffix male 86 4.59 1.23 .132 1.12 176 .132 female 92 4.38 1.33 .139 both male 86 5.09 1.45 .156 .91 176 .182 female 92 4.90 1.31 .137 man male 86 5. 09 1. 12 121 .70 176 .240 female 92 4.97 1.21 .127 his male 86 4.67 1.46 .157 .80 176 .211 female 92 4.48 1.64 .171 Very little rating difference occurred between the 66

PAGE 75

genders. Males rated the terms more masculine by only .1 or .2 more than females. Even if these results had gained statistical significance the differences would not have been practical enough to warrant a claim of socially important gender differences. Analysis of Variance ANOVA results registered one variable statistically significant at the .07 level and another marginally significant at the .14 level. respectively. See tables 4 5 and 4 9 The 07 significance occurred on the "pre" variable which was the first dependent measure. The findings related to this variable concern the mean scores for male respondents that began at 4.25 for undergraduates, peaked at 4.66 for graduates and declined to their lowest rating of 4.18 for male faculty. This indicated that faculty males rated the terms the most gender neutral for the three male groups. The trend for males is one of rising at the graduate level and declining at the faculty level (see graph A) Graph A-The ANOVA Male Trend 4.5 4 3.5 mal .. -under"gr"ad male-gr"ad 67

PAGE 76

These trend lines have nonparallel effects for males and females. The same analysis conducted on females reveal an opposing trend (see table 4. 9). The findings of mean scores for female respondents began at 4.09 for undergraduates, declined to 3.88 for graduates and peaked at 4.45 for female faculty. This indicated that female graduate students rated the terms the most gender neutral of the three female groups. The trend for. females is one of a decline, at the graduate level and rising at the faculty level (see graph B). It was only at the faculty level that the females rated the terms more masculine than did males. Graph B-The ANOVA Female Trend 4.5 4 3.5 Only on one variable did female undergraduate students rate the terms more masculine than any other group. This occurred on the "his" variable and achieved .14 level of significance (see ANOVA table 4.9). On this variable, the male mean score began with the undergraduates lowest rating of 4.51, rising to 4.78 and slightly declining to 4.72 for male faculty. This trend differs from the previously discussed trend concerning 68

PAGE 77

table 4. 5 because here male undergraduates had the most gender neutral rating and there the male faculty had the most gender neutral perception. The trend for female mean scores began at with their highest rating of 4.88 for undergraduates, declined to 4.12 for graduates and rose to 4. 26 for female faculty (see table 4.9). This trend also differs from the previously discussed female findings on table 4. 9. This indicates that female graduates rated the terms the most gender neutral for the three female groups. The trend for females is one of peaking at the undergraduate level, declining at the graduate level and a slight rise at the faculty level but one that is still lower than the undergraduate level Findings here were that undergraduate females rated the terms toward a more masculine interpretation than any other female or male group regardless of college education level. For complete ANOVA results see tables 4.5 -4.9. 69

PAGE 78

Table 4.5 ANOVA Interaction of Gender & Educational Level Pre Variable sex male female education level undergrad 4.25 4.09 2-way interactions Sig of F sex education level .072 Total grad 4.66 3.88 ss 7.67 257. Table 4.6 DF 2 177 fac 4.18 4.45 Mean Sq 3.84 1.45 ANOVA Interaction of Gender & Educational level Suffix Variable sex 2-way Sig of sex .611 Total education level male female undergrad 4.44 4.44 interactions F education level ss grad 4.75 4.26 DF 1.65 2 291. 177 70 fac 4.59 4.41 Mean Sq .825 1. 64 F 2.67 F .494

PAGE 79

Table 4.7 ANOVA Interaction of Gender & Educational Level Both Variable education level sex male female undergrad 4.81 4.89 2-way interactions Sig of F sex education level .290 Total ss 4.72 336. Table 4.8 grad 5.57 4.90 2 DF 177 fac 4.94 4.93 Mean Sq 2.36 ANOVA Interaction of Gender & Educational Level Man Variable education level undergrad grad fac sex male 5.01 5.27 5.02 female 5.01 4.88 5.01 2-way interactions ss DF Mean Sq Sig of F F 1.24 F sex education level 1.36 2 .680 .490 .614 Total 241. 177 1. 36 71

PAGE 80

Table 4.9 ANOVA Interaction of Gender & Educational Level His Variable education level sex male female undergrad 4.51 4.88 2-way interactions Sig of F ss grad 4.78 4.12 DF fac 4.72 4.26 Mean Sq F sex education level .145 9.38 2 4.69 1.95 Total 427. 177 2.42 An interaction effect was statistically supported at the .07 level on the "pre" variable (table 4.5) and marginally at the .14 level on the "his" variable (table 4.9). The "pre" variable results significant at the .07 level are examined first. The trend concerning males depicted a pattern of graduate students rating the terms toward a more masculine interpretation with an decrease toward a more gender neutral interpretation at the faculty level. The pattern for females was opposite of the male pattern. Female graduate students rated the terms more gender neutral with an increase toward a masculine interpretation at the faculty level. The interaction effect on the "pre" 72

PAGE 81

variable showed male graduate students rated the terms more masculine than any other group regardless of gender or college education level. The "his" variable results, although only marginally significant at the .14 level, need to be examined because this was the only variable that resulted in female undergraduates rating the terms more masculine than any other group. The finding concerning females was depicted by female undergraduates rating the terms more masculine followed by a shift toward a gender neutral perception at the graduate level and back toward a masculine perception at the faculty level. Male graduate students rated the terms more masculine with a slight shift toward a gender neutral perception at the faculty level. Only on the "his" variable did female undergraduate students rate the terms toward a more masculine interpretation than any other group regardless of gender or college education level (see graph C) Graph C ANOVA "His" Trends 5 a.__..___. Neles 4.5 .-..........._ -o 0 4 .....----- Iemal.es 3.5 3 2.5 2 1.5 I 0.5 0 f_ec_u_lt\1;1 The patterns revealed give some indication that the term "man,11 man related terms and the pronoun "his" are being 73

PAGE 82

perceived toward a more masculine interpretation rather than gender neutral among college students and faculty members in this study. What do these finding signify? Analysis and speculation concerning these results are presented in the next section. 74

PAGE 83

CHAPTER FIVE DISCUSSION This study was designed to examine subjects use and understanding of the generic terms: man, man related terms and the pronouri "his". The study's main result indicates that graduate students are more likely than undergraduate students to perceive the generic terms as masculine. This finding was statistically significant on one measurement at the .05 level. An interaction effect indicated that male graduate students interpreted the terms more masculine than any other group regardless of gender or college education level at the .07 level of significance. While neither was significant at the .01 level, both results indicate a trend of graduate students over undergraduate students rating the terms toward a more masculine interpretation. It was shown that generic language that has been taught to be gender neutral throughout education is no longer being perceived this way. What might this reversal in language habits mean? What change has occurred for male and female college students perception of generic language? Why are males limiting their perception of generic terms primarily to males while females are perceiving their own inclusion? Is it so that females do not feel left out of a substantial part of the English language and communicative process?

PAGE 84

These questions are examined in the following discussion. This chapter first examines the education element of Hypothesis One. Next, Hypothesis Two which did not reach statistical significance is briefly examined. Then interaction effects are covered that include gender and education factors as contributors. Lastly, suggestions for further research are offered followed by a review of the conclusions. Now, the study's. two hypotheses and interaction effects are discussed in terms of statistical findings and their implications. Hypothesis One: The more college education individuals have the more likely they will be to view the terms, "man", man related terms and the pronoun "his" as masculine. The first hypothesis concerns college education level and interpretation of generic terms. Statistical support was observed on one test for this hypothesis. Three alternative explanations for this finding are provided in the following discussion. These explanations focus on the issues associated with the learning of generic terms, the application of generic terms and the role authority plays in the learning and application of generic terms. The discussion attempts to present explanations for the "both" variable that received statistical support and the four other measurements that did not. 76

PAGE 85

Hypothesis Two: Males will view the generic terms, 11man11, man related terms and the pronoun 11his11 toward a more masculine interpretation then will females at the same educational level. The second hypothesis concerns gender and interpretation of generic terms. The statistical test of main effects indicated that differences between males and females were not sufficient to warrant a claim of socially important gender differences. On the t-test males consistently rated the terms toward a more masculine perception but the difference was a slight one. Both hypotheses lent themselves to test for interaction effects. Six interaction effects were concluded on the ANOVA testing. One interaction effect tested was significant at the .07 level. On the variable measures that were not statistically significant, most of the interaction trends continued. This chapter's discussion of effects begins with discussion of Hypothesis One. The explanations focus on one's gender identification in language and gender differences in learning and responding to pronoun interpretation. Hypothesis Two is briefly examined but since no major gender differences were supported little analysis can be completed. Next, interaction effects are analyzed by gender and by college education theories. 77

PAGE 86

Gender identification and equality in language and learning and application of generic terms are interwoven into these explanations in an attempt to explain the six interaction effects. Hypothesis One Hypothesis One received statistically significant support using the independent groups t-test on the "both" variable (.05). This t-test examined the college education factor in connection to subjects use and understanding of the generic terms. Graduate students rated the combination of the man related terms and the pronoun "his" toward a more masculine interpretation then did undergraduate students. Tests regarding the other variables, "pre", "suffix", "man" and "his" did not receive support but with the exception of the testing of the "his" variable continued the same trend of graduate students rating the terms as more masculine. What are the implications of these results and how are they connected with one another? When the generic terms foreman, weatherman, congressman, spokesman and salesman were individually combined with the pronoun "his" in the same sentence (the "both" variable), the sentences were rated more masculine by graduate students and reached statistical significance of .05. It was only when these two generic terms were combined that statistical significance occurred (i.e., the foreman gave 78

PAGE 87

his verdict) The combination of terms had a stronger influence on students to rate them as masculine than when only one generic term was used in a sentence. Graduate students rated the "both" variable the most masculine at 5. 24 (the highest masculine rating on any variable) in comparison to undergraduate's rating of 4. 85 (the third most masculine rating on any variable). It was here that the. greatest difference between mean scores of the two groups occurred. It must be noted that although statistical significance was reached on this variable, one must address the issue of the finding's practical importance. This slight difference between mean scores is not great enough to warrant an observable behavioral difference between the groups. It is unlikely that the graduate students over undergraduate students who completed this survey would exhibit observable linguistic differences concerning generic terms (the use of less generic language). The "pre", "suffix" and "man" variables demonstrated the same trend. Graduate students rated the terms more masculine than undergraduates. These variables along with the "both" variable all contained the word "man" as a prefix, suffix or used alone. It appears that the students perceived the different uses of "man" in a similar manner as the "both" variable but statistical significance was not reached. 79

PAGE 88

The trend changed on the "his" variable. Undergraduate students rated the terms more masculine than did graduates. There was a .3 difference between mean scores of the two groups. Students must have looked at the pronoun "his" somewhat differently from the man related terms. Undergraduates were less likely to include females in the pronoun usage. The discussion on the "his" variable follows this first discussion. The following discussion section offers analysis on all of the findings (except those of the "his" variable}. The analysis offered attempts to directly link the role of education to generic language learning and interpretation. The first issue under analysis is associated with the learning of generic terms. The academic environment is a formal educational setting for learning. Children enter elementary school with the basics of language learned at home and in informal settings (i.e., child care}. The children's language skills include word choice and order but no formal language rules. Generic language learning follows general knowledge of language learning. Children only know what they have been told, seen role modeled, or learned intuitively on their own. Hyde (1984} found only 28% of first graders knew the grammatically correct generic "he" rule. But by college 84% knew and used the rule. It is in formal learning (school} that children are taught generic terms' usage and semantics. Children learn 80

PAGE 89

language rules and abide by them in order to succeed (i.e., gain approval, get good grades, etc.). Children adhere to the rules in this structured environment that does not encourage much deviation. A child in grade school may learn generic rules, be forced to use them in a structure that does not offer alternative terms and grow up knowing they are prescriptive grammar and yet never truly .... understanding their inclusiveness. Education plays the primary role in generic language learning. It is through primary education that one learns generic terms. A second important issue is the application of these terms. This application takes two forms, one involving teacher application and influence and the other involving the student him/herself . Children learn and obey the expected application of generic terms and it is only further along in their education that people start to question this practice or that teachers offer alternative terms to students. Even in college, the novelty of creating one 1 s own ideas and papers after 12 years of restricted classroom instruction may result in indiscriminate pronoun use and the variety of generic "man" conjunctions. Perhaps only by graduate school do students no longer perceive generic terms as gender neutral. Graduate students were likely to perceive generic terms as masculine, so somewhere in their college education they have been influenced (required) by professors to use 81

PAGE 90

alternative terms rather than generic terms to refer to neutral or feminine referents. It also follows that teachers must have used alternative terms when teaching (i.e., "he or she" or "men and women") as an example to be followed. Since it is unlikely (but not impossible) that elementary teachers or high school teachers used alternative terms, students must have been taught these terms in college. It is at the graduate level that a change in perception of generic terms occurs. What might account for this change? At the undergraduate level, students still ask for and receive much structured guidance concerning tests, papers and research projects. They are expected to complete work according to the professor. They are given a topic to write on and possibly told not to use the masculine pronoun "his" when referring to referents. The student responds to this task possibly without giving too much thought to pronoun meaning. Once again, the student is following the rules imposed by the teacher. After four years of this instruction, a change may occur in generic language awareness. For the graduate student a change has occurred in generic language awareness. Graduate students were more likely to rate the generic terms as masculine. Some reasons for this change may involve the graduate student's new level of responsibility for his or her own education. This would 82

PAGE 91

include the writing of numerous research papers which causes the student to become more familiar with the structure of writing and with the subtleties of meaning. Graduate students become aware that generic terms are not inclusive and that using only them excludes females. The third issue under analysis concerns the role authority plays in the learning and application of generic language. Education is a link between generic rules and application and their interpretation. Harrigan & Lucie (1988) found students over non-students were the most influenced by authority figures. It could be guessed that students would be the most influenced group since they are in an atmosphere of learning from adults. Perhaps teachers who require the use of non-sexist language and themselves use language alternatives such as, "he or she" instead of "he" influence their students to do the same. I can only speculate that teachers at UCD and Metro State are more likely to use these alternative terms in graduate courses as a model for students. While some students may be transfer students not all the students can be transfer students, so some of their formal usage exposure must have come from the Auraria campus. What can explain the differences in student responses to the "his" variable as opposed to the other man related terms? The difference between man related terms and the pronoun "his" concerns complementary terms. The term "he" 83

PAGE 92

has the term "she" and "his" has "hers" as their complements to indicate a female. Many of the man related terms have no complementary feminine term or only recently has a term been coined specifically for females but this use has not gained the popularity of the masculine term. The term "mankind" has served as a universal inclusion for males and females. It has no parallel term to indicate females so that "mankind" can stand alone for maleness as and does "he11 Although the term "womankind" does exist, it is rarely used or accepted as a term to indicate females. A main effect detected undergraduate students as rating the pronoun "his" more masculine than graduate students. This result could have occurred due to undergraduates' introduction into the academic environment. As already discussed, students are highly influenced by authority figures (professors). Teachers in college are. encouraging independent thought and language discrimination. They are asking the students to break free from the restrictions imposed on them as high schoolers and connect language to thought. Words are tools used to convey meaning and have an impact that goes beyond the printed word. This personalization of language may affect generic terms and their implications. This teaching may affect students' usage and response to generic terms. The students are just beginning to make a place in the world for themselves and are marking this distinction through 84

PAGE 93

their language. They may be saying, let "he" represent males and "she" females, let each gender have its own distinctive pronoun. When both genders are included, use "he or she" or "they" but not just "he". Perhaps undergraduate students are reversing the generic pronoun rules they learned throughout school as they become better as writers and more socially aware of the power of language. Hypothesis Two Hypothesis Two proposed that males more than females at the same educational level would rate the terms more masculine. There were no statistically significant results, although the trend on every variable showed males rating the terms more masculine. This means males were more apt to exclude females and females were more likely to include females in generic terms. This trend differs from most of the research on generic language which indicates that females rate such terms more masculine then males (Benoit & Shell, 1985; Fisk, 1975; Hamilton, 1988; and Martyna, 1978). Perhaps there is a change occurring concerning gender and generic language interpretation. It may be that the Auraria campus has different linguistic norms than other campuses or that this lack of gender differences is a signal for educated people in our society. 85

PAGE 94

Interaction Effects Hypothesis One states that the more college education one has the more likely one will be to view the generic terms as masculine. Hypothesis Two states that males will view the generic terms more masculine than will females. Analysis was performed for five interaction effects associated with the factors of education and gender and their connection to generic terms. These analyses resulted in several findings concerning male respondents who exhibited a pattern of means similar to an inverted V. Male undergraduates and male and female faculty rated the terms more generic than the highest masculine rating of the graduate students on five out of five measures. The second interaction concerns female respondents who exhibited a pattern of means similar to an upright V. The female trend revealed undergraduate students and faculty rated the terms more masculine while graduate students rated the terms more generic on four out of five measures. Male graduate students rated the terms more masculine than any other male group on five out of five measures. Male undergraduate students rated the terms the least masculine of any male group on four out of five measures. Female graduate students consistently rated the terms more generic than any other group on four out of five measures. Male graduate students rated the terms more masculine than any other group on four of the five measures with female 86

PAGE 95

undergraduate students rating the terms more masculine on the fifth measure. This discussion first examines the ''pre" variable which was the only variable that demonstrated statistical support for an interaction effect concerning gender and college education level. The following section examines gender and college education as two interrelated explanatory variables for the interaction findings. Gender is first, since research has been done on females and males concerning generic language, and only loosely connected research has been done on education and generic language. The gender explanations concern female exclusion in the generic terms by males and females and female inclusion in the generic terms by males and females. The college education level explanations concern undergraduates, graduates, and faculty exclusion and inclusion of females in the generic terms. The interaction effects are delineated within these appropriate discussions. Following these discussions, suggestions for further research are presented and then followed by a summary of conclusions. The "pre" variable interaction effect was statistically supported at .07 level. No other interaction effects were statistically significant. So this section will only discuss those findings associated with the "pre" variable. The means for the males on this variable revealed undergraduates to have the middle score (4.25), graduates 87

PAGE 96

the highest or most masculine (4.66) and faculty the lowest or most generic (4.18). See Graph D for the group means of the "pre" variable. The means for females exhibited an opposite trend. Female undergraduate students made responses at the middle score (4.09) as did male undergraduates. However, here the trend reverses for females, as graduates had the lowest score (3.88) and faculty the highest (4.45). This pattern of means is one of an upright V for females. Male graduate students rated the terms more masculine than any other group. Female graduate students rated the terms the least masculine (hence, more generic) of all groups. What are the implications of the interaction effects when combining gender and college education level? Graph D ANOVA Pre Trends 5 3.5 3 2.5 2 1.5 1 0.5 0 Femelet Mele1 0 undergred What do we make of the differences between male and female graduate s.tudents? Why did males indicate inclusiveness of females in the generic terms when the males were undergraduate students and faculty members but more exclusiveness when they were graduate students? Why 88

PAGE 97

do females show the opposite trends? Several explanations are offered. They begin with analysis of those males and females who rated the generic terms the least masculine and hence most inclusive of females. These explanations are interwoven with gender identification and gender equality in language as well as the learning and application of generic terms previously covered in the discussion of Hypothesis One. Gender as a Contribution to Interaction Effects What is one to make of males who rated the generic terms the least masculine and hence, the most inclusive of females? This discussion includes the pattern of mean scores for males, especially male undergraduates and faculty who rated the terms less masculine than male graduate students. On the "pre" variable male faculty rated the terms less masculine than male undergraduates and graduates. These males have used generic terms throughout their lives, are comfortable with them and may not perceive the need to change or alter them in any way. They included female in the terms because it has always been done this way and they may not be aware of the uproar concerning the terms. Although male faculty must be aware of the uproar considering their role as an acamedician. These males may have a lack of awareness necessary to change their language use. or it may be the opposite case. These males may be 89

PAGE 98

strident against changing the language they believe in, are comfortable with, or think is equal as it is. They may believe their own gender identification is secure in these terms and should be also for females as well since they also grew up using these terms. The language is as for women as it is for men and just because the terms are originally masculine in nature does not make them less applicable for females. What about those females who rated the generic terms the least masculine and hence, most inclusive of females? This discussion concerns the pattern of mean scores for females especially female graduate students who consistently rated the terms the least masculine. play a key role. Women were Gender identification may more likely to include themselves in the terms because otherwise they would not have been identified in the terms at all. Women rated the terms slightly less masculine so there would be at least a chance of female inclusion. If generic language is not gender inclusive females remain absent or invisible in generic terms. Women's identity becomes less notable and valued if they are not identified in generic language. Generic terms are widely used and women's inclusiveness is necessary for their representation. People identify themselves in their language as they identify with their own name. Women want to believe they are included in generic terms. The more identification women have in 90

PAGE 99

language, the larger potential they have in their society (Silveira, 1980). One's identity is central to one's being. Language is a primary source of identity in which to identify ourself, others and the world in which we live. Language is also a source where equality can be expressed. For some females, equality in language is the tendency for females to do what males naturally do in generic terms-include themselves in the generic terms in any context. Generic terms operate only in a defined context, out of that defined context the terms go back to their original masculine intention. Females may be seeking the same inclusion or identification that males have, which is one of complete inclusion in generic terms. They are seeking equality in language but it is in a language that has a masculine foundation of semantics. The terms were originally masculine in definition and interpretation. Other analysis of female inclusion concerns the difference in learning and responding to pronoun interpretation. As already covered in the discussion on Hypothesis One, there is a learning similarity between generic nouns and pronoun complements. The term "man" has "woman" as its complement as does "he" have "she" as its complement. Children learn generic pronoun and generic noun interpretation in much the same manner. Girls learn at an earlier age that generic pronouns may be inclusive of both males and females. Females must formally learn 91

PAGE 100

generic meanings in school, those of which may be misused and corrected. Males are less likely to be corrected because they are using the "correct" pronoun but it is not known if they have fully grasped the generic inclusive meaning. Girls and women have had to learn the formal rules of generic terms and use them if they wanted to be included in the textbook, lesson, discussion, language or in whatever context the generic terms were being applied. Females have been taught to refer to "a person" with the masculine pronoun "he" and so they must reverse their natural tendency to use themselves as reference of the subject. Now in order to have their gender exist in generic terms, females must believe in, perhaps insist upon its inclusiveness (Henley, 1989). What is one to make of males who rated the generic terms the most masculine and hence, the most exclusive of female interpretation? This discussion covers the pattern of mean scores for males especially male graduate students who rated the terms more masculine than any other group on the "pre" variable. In terms of gender identification and equality in generic terms, these males identified with generic language to the point of the exclusion of females. Masculine terms that are "officially" declared generic terms continue to describe males, include males and are being perceived toward a more masculine interpretation than neutral one. Males have heard and used the terms to 92

PAGE 101

describe themselves and other males and only peripherally females. But it is primarily males who are imaged in these terms. The terms may not be interpreted as intentionally inclusive or exclusive of females but are being perceived as more masculine. These males are not perceiving equality in these terms because the terms are more referenced for males than for males and females. The possibility exists that some males may have language sensitivity toward generic terms. These males may believe in using language that represents both genders equally and not using masculine terms generically. They may believe that generic terms reference only males and so language inclusive terms need to be used when including females. Several reasons are offered for their behavior. Female exclusion by males concerns the learning and application of pronoun interpretation. Boys learn at a later age than girls that generic pronouns may include boys and girls. Boys respond to generic pronouns by assigning their own gender to the subject. Males may use the "correct" term but may not be fully aware of its generic intention and may not be cognitively corrected. Boys and men may not fully understand the gender neutral meaning of generic terms because they never had to fully learn or relearn them. Males may continue to automatically use "he" to refer to "a person" since this is what they did when they were learning the terms as children. They may 93

PAGE 102

continue to think of "he" a person, like themselves, as having a masculine interpretation. What about female faculty members? Those who rated the terms the most masculine and hence, the most exclusive of females. This discussion covers the pattern of mean scores of females (undergraduate students and faculty) who rated the terms more masculine than female graduate students. Since these females have had the same educational process as females who perceived female inclusiveness, something other than gender and the education context must be a factor. While some females have accepted a generic interpretation, other females are primarily referencing a male image or at least a masculine interpretation. Even in the context of being used with a gender neutral term, (i.e., citizen, professor, or student) these females are still identifying the pronoun "he" as including only males more than males and females. These females do not identify or perceive equality in generic terms. Something has changed from when they learned and applied generic terms as grade schoolers through high schoolers. After years of education that taught females to reverse their natural tendency to use themselves as references of the subject, they are now referencing only males to male nouns and pronouns. Perhaps these females do not judge themselves to be included in masculine terms and believe this inequality in language has 94

PAGE 103

gone on long enough. In order for there to be full equality in our society, there need to be equality in our language. Terms like chairman, policeman, or 11he11 continue to bring to mind a male more than a female and this referencing needs to be changed to represent both genders without special attention to one over the other. Language needs to be neutralized so that all people have an equal chance in the status references be it in physical reality or cognitively. College Education Contributions to Interaction Effects Why would graduate and undergraduate students respond differently to the questionnaire in this study? Why would female graduate students respond differently than male graduate students? Why would faculty respond differently than students? This section offers several answers. It is divided into four analyses: 1) the minor differences between female and male undergraduate students 2) the differences between undergraduate and graduate students 3) the significant differences between female and male graduate students and 4) the minor differences between female and male faculty. What analysis can be induced concerning female and male undergraduate students who rated the terms less masculine and hence, more inclusive of females? Since the gender differences in ratings are so slight, it can be deducted that education plays a larger role than gender when one is 95

PAGE 104

an undergraduate in college. By definition, undergraduate students are either recent graduates from high school or those out of school for an extended length of time. If they are recent graduates, they are fresh from an educational -background that required they use generic terms. It is this learning that they have operationalized for years prior to entering college. Their role as a college student is one similar to their high school student role and this includes their language perceptions. While college professors may be requiring them to use alternative terms over generic terms, they have not yet conceptualized them. Those returning students either away from high school for a time or those having already earned some college credits are similar to incoming or continuing undergraduates in that they have not yet established distinctive gender perceptions in generic terms or have decided they are not open to alternative terms. What is one to make of the significant differences between undergraduate and male graduate students ratings? Male graduate students rated the terms toward a more masculine interpretation hence, more exclusive of females of females than did undergraduates or female graduates. The role of the professor as well as the student probably play dominant functions. Professors must present male and female students with examples that are readily understandable and identifiable. This includes the use of 96 '-

PAGE 105

nouns and pronouns that are masculine, feminine, neuter and neutral. The noun "husband" identifies male inclusiveness while the noun "wife" identifies female inclusiveness. The animal as "it" or the "student" or "citizen" clearly allows for both a male and female interpretation. Professors may also be using alternative terms rather than generic terms. This includes using such terms as "he or she" when referencing any given person. Or using "he" and "she" alternatively so that students are presented with gender specific pronouns to know that "he" means only males when its complementary term "she" is used in the next example. Faculty at UCD and Metro State may be advocates of alternative terms often enough so that the message is received by the students. Professors often write on students' papers that the student should not use generic terms so that it is clear that both genders are included. At the graduate level, the role the professor plays in the indoctrination of students by using alternative terms over generic terms increases from the undergraduate level. Faculty have higher expectations for graduate students as well as serving as role models through their own language usage. Faculty not only use alternative terms themselves but expect the same of graduate students since they have already had at least four years of college education. They are expected to know the basic rules and guidelines of the academic environment and this includes the subtle use of 97

PAGE 106

words as tools which convey distinctive gender meanings. The role of the student increases at the graduate level as well. The student becomes more invested in his or her educational process and must participate with his or her own theories and ideas. Topics are no longer randomly assigned but are created by the student in response to a research project with only a due date given. students must go beyond their undergraduate structure and knowledge and start contributing to the research field. It must also be noted that what is learned at the undergraduate level may not be applied until graduate school. But what about the significant differences between female and male graduate students? Why did male graduate students rate the terms more masculine than female graduate students? Since both have had the same number of years of education, education cannot exclusively account for such differences. Other factors must be active and gender must be one of these factors. An anecdote may help to clarify this interaction effect. It is an example of a female seeking inclusion in generic language. A woman who had diligently worked her way up the corporate ladder was offered the position as chair of the department. When asked if she'd prefer the title, chairwoman or chair, she responded with "it took me 15 years to get this position and I. want the honorary title that goes with it--chairman". The term chairman is not 98

PAGE 107

only traditionally the common term but a term of prestige and power. Female graduate students may be seeking inclusion in what they consider the dominant and more powerful masculine language. Since men are traditionally portrayed as the active agents in our. society, masculine terms are deemed more powerful than the newer and less accepted feminine terms. These females may think now that they have achieved graduate status they deserve inclusion in the masculine and therefore more powerful terms. What about the slight difference between female and male faculty? As with undergraduates, these two groups showed minor differences between one another. There are several possible explanations for this finding. All faculty operate in a similar institution of higher education with similar roles. They must prepare and present lectures, complete research, report to the dean of their school and interact with students and other faculty on a daily basis. They are locked into roles with individual attitudes and styles as the main sources of differences. Faculty are authority figures who exchange information and are frequently role modeled by students. This person must present information in such a way as to gain understanding and compliance of students. This is accomplished through language. They may use generic terms or alternative terms depending on their etymology. Traditional grammarians believe in prescriptive rules and do not favor changing or 99

PAGE 108

replacing generic terms. Descriptive grammarians believe in describing the language of the people so when society changes, language follows closely behind. Faculty members have their own theories and opinions on language and pass this on to their students. It may only be by using the generic "he" or "man" in classroom examples that professors pass on their generic interpretations. Basically all professors have the same duty to educate their students. Lines are not drawn along gender. Ideally, female professors have the same tasks as male professors. Both are cogs in the same system probably for similar reasons. Their roles are not so different as to warrant a significant difference of generic language awareness. It is also for the above reasons that faculty responses are likely to differ from students. Another reason is that faculty are no longer being educated but are educating others. Their focus is not on what they can learn but what they can teach. The role of the student is one of active listening, learning and applying in a semi-formal environment. It is an exchange of one person to many people. Students are in this environment to learn academic material but faculty are learning that and more. They are learning to teach, do research and support themselves within an academic environment. While the roles of student and faculty merge, differences do exist that may cause differences in generic perception between the two groups. 100

PAGE 109

Now that the "pre" variable results have been examined, other findings that did not reach statistical significance are covered. Some of these findings were similar to those for the 11pre11 variable and continued to surface as trends. Male undergraduates and faculty rated the terms more generic than graduate students on five out of five variables. Female undergraduate and faculty rated the terms more masculine than graduate students on four out of five variables. Male graduate students rated the terms more masculine than any other male group on five out of five variables. Male undergraduate students rated the terms less masculine of any male group on five out of five variables. Male graduate students rated the terms more masculine than any other group on four out of five variables with female undergraduate students rating the terms more masculine on the fifth variable. The interaction effects and associated findings of this study lend themselves to questions requiring further research. Some of these questions are presented in the next section. Limitations of Present Study The lack of statistical significance found may be due to the dependent measurement. The subjects may have been too aware of the general issue. The questions were not masked and subjects were asked directly to identify the gender of the antecedent. Even though subjects were asked for their 101

PAGE 110

own opinion, social influence may have significantly altered answers. Generic language is more socially accepted as the norm and this factor may have caused subjects to go with what is acceptable rather than what they truly thought. Another factor may be that since the survey was given in a university setting, the education variable was more pronounced. Along this same line of reasoning it must be noted that Auraria campus where this survey was administered is not a typical campus. Many students are older, attend classes part-time, work fulltime and have already had an extensive exposure to life outside the academic environment. The fact that this was a lab survey may have changed natural responses and language patterns. A naturalist approach to linguistic behavior observations would have been ideal for this type of study but would have required extensive time and resources. Suggestions for Further Research The fact that there have been significant findings in the research concerning generic language is comforting considering that subjects are saturated by sex-biased language daily. However, new research continues to be needed to further discover the consequences of generic terms. The issues need to be directly confronted. Sex bias is not a historical, accidental by-product of 102

PAGE 111

language; it is each cultures stamp on its language (Penelope, 1990). Questions that still need investigation that could further enhance this study and other pressing issues to be researched are listed below. 1. Is there a shift in higher education levels for the individual to make clearer delineations of gender in generic speech? Findings concerning the present study differ from past research in this field and suggest that normative changes may be occurring in our universities. These questions could be researched in our universities since there is a higher sensi ti vi ty to language among academicians. 2. What psychological effects do masculine generics have on children? Even though the present study found minor perception differences between males and females, it is likely that there are vast psychological differences between the sexes. Although generic language continues to be taught in elementary school, there is little research done on its effects. These questions could be researched by surveying children's perceptions and understandings of masculine generics. Do young girls and boys perceive female inclusion and how do they feel about this inclusion or exclusion? There is no recent research done on small children's perceptions of generic terms and is an important area linguistically and psychologically. 103

PAGE 112

3. Is it possible to use neutral terms for job titles (i.e., postal carrier, sales representative, department chair, member of congress, etc. ) and eliminate gender specific terms completely? Societal roles do not have to represent the gender that is primarily employed in that area, as do the titles, chairman or workman represent males in those roles. It is also only male identified titles that exist in the English language suggesting that males are the dominate agents. Education of such alternative usage may begin in one's college years or it may need to be instilled at an earlier age to truly be integrated into one's belief system and linguistically. It needs to be discovered at what age it become harder to change one's semantics and usage. A new word or meaning becomes widely used when it is more accepted than the old word or meaning. These questions could be researched in various field studies (i.e., businesses, corporations, government, universities, etc.) plus by studying how long it takes a new term to become part of a culture's dominant language. Review of Conclusions The present study found little empirical evidence to support the trend of research on generic language. It was only when the term "man" was combined with the pronoun "his" that the results were statistically significant. It was then that male graduate students were more likely than 104

PAGE 113

any other group to view the generic terms toward a masculine interpretation than a neutral interpretation. This result was shown on three of the five ANOVA variable tests. In this study males rated generics more masculine than females, which contradicts previous research suggesting the opposite. Males were less likely to include females in .... generic language while females were more likely to include themselves. The females in this study have been formally educated for 13 to 21 years to perceive themselves in generic terms. It has been taught to them in an academic environment and through personal and less formal socialization. Females have learned the generic rules, abided by them and now evidence shows they believe they are included in them. On the other hand males are viewing the generics as more masculine. Although it must be noted that neither males or females at any education level (with one variable exception for female graduate students) ever rated below 4.0 on a 7 point Likert scale with 7 as the masculine extreme. All participants were more likely to view the generic terms toward a masculine interpretation than toward a neutral one. This research and analysis of the role of education and gender and generic understanding was conducted and presented to enlarge our understanding of this problem and sensitize students, teachers and researchers to sexist 105

PAGE 114

language in its various manifestations. Since none of us has ever lived in a nonsexist society, moving toward equality in language is similar to trying to comprehend a dimension we have not physically experienced (Eichler, 1988). We can describe it in theoretical terms, but we cannot fully appreciate its nature and consequence until we are able to life ourselves out of our current parameters. 106

PAGE 115

APPENDIX QUESTIONNAIRE & COVER LETTERS Dear Faculty Member, I am completing my Masters Thesis and need your assistance in filling out this questionnaire for me. I realize that your time is valuable and scarce and would be grateful for your Please complete all of the following questions as carefully as possible. All answers to this questionnaire are strictly confidential. Federal and University regulations regarding research require that you indicate your consent before participating. Sign your name in the space provided and return this survey to your department in one week's time. once it is coded all the data will be detached from this cover sheet. Your name will not be connected to the data in any way. Instructions: This is a probability study dealing with gender. The questions are not difficult but need to be given thought. Please read the question and indicate to what degree is its likelihood or probability. Please respond with your personal opinion and not what you assume other people think or feel. There are no right or wrong answers. Sample Question: How likely is it that the statement, "the fisherman went fishing" does the term fisherman refer to a female? Extremely Likely Thank You Vickie Harvey Moderately University of Colorado-Denver Name Extremely Unlikely ----------------

PAGE 116

FACULTY FORM I am I am male ---years old I have a M.A. degree in I have a Ph.D. degree in I graduated with my M.A. I graduate with my Ph.D. I teach at in in female (what year?) (what year?) (school at Auraria campus) I have been teaching here for number of years I have been teaching years (total number of years) My religious affiliation is 108

PAGE 117

STUDENT FORM I am male female ---I am years old I have years of education I am a in college (freshman, sophomore, junior, senior, graduate) If you already have a degree, what is it and what year did your earn it The degree I am working toward is a Ph. D.) My major is My religious affiliation is 109 (B.A. I M.A. I

PAGE 118

Extremely likely SURVEY Moderately Extremely Unlikely 1. How likely is it that the old adage "the average man on the street" refers to a female? 2. How likely is it that the term manpower refers to female strength and ability? 3. How likely is it that in the statement, "the student paid his tution" the term his refers to a female? 4. How likely is it that the term policeman refers to a female? 5. How likely is it that in the statement, "the customer spent his money" the term his refers to a female? 6. How likely is it that the term layman refers to the everyday female? 7. How likely is it that the statement, "man must fulfill his destiny11, refers to females? 110

PAGE 119

Extremely Likely Moderately Extremely Unlikely a .. How likely is it that the term chairman refers to a female? 9. How likely is it that the term caveman refers to a female? 10. How likely is it that in the statement, "the student wrote his paper" the term his refers to a female? 11. How likely is it that in the statement, "mankind has increased its awareness" the term mankind refers to females? 12. How likely is it that in the statement, "the foreman gave his verdict" the statement refers to a female? 13. How likely is it that manhood means adulthood for females? 14. How likely is it that manhours are the amount of hours a female works? 15. How likely is it that in the statement, "the professor taught his class" the term his refers to a female? 111

PAGE 120

Extremely Likely Moderately Extremely Unlikely 16. How likely is it that in the statement, "the weatherman gave his forecast" the statement refers to a female? 17. How likely is it that the phrase, "the working man" refers to females? 18. How likely is it that in the statement, "the congressman cast his vote" the statement refers to a female? 19. How likely is it that the phrase, "dirty old man" refers to a female? 20. How likely is it that in the statement, the spokesman directed his comments to the media" the statement refers to a female? 21. How likely is it that in the statement, "the booth was manned" the statement refers to a female? 22. How likely is it that in the statement, "it is time for all good men to come to the aid of their country" does the term men refer to females? 23. How likely is it that in the statement, "the citizen cast his vote" the term his refers to a female? 112

PAGE 121

Extremely Likely Moderately Extremely Unlikely 24. How likely is it that in the statement, "the first baseman caught the ball" the term baseman refers to a female? 25. How likely is it that inthe statement, "the salesman made his quota" the statement refers to a female? 113

PAGE 122

LIST OF REFERENCES Baron,D. (1986). Grammar and gender. New Haven:Yale University Press. Bate,B. (1978 Winter). What does "she" mean? Journal of Communication, 139-150. Bem s., & Bem D. (1973). Does sex-biased job advertising aid and abet sex discrimination? Journal of Applied Social Psychology, 6-18. Benoit,s., & Shell,W. (1985). Can sex biased communication affect students career decisions? Delta Pi Epsilon Journal, 28(2), 123-129. Betty,s. (1990). Unpublished Conversation. University of Colorado at Denver. Blair,T.L. (1977). Retreat to the ghetto: The end of a dream? New York:Hill & Wang Inc. Bodine,A. (1975). Androcentrism in prescriptive grammar:singular 'they, sex-indefinite 'he' and 'he or she. Language in 129-146. Briere,J., & Lanktree,c. (1983). Sex-role related effects of sex bias in language. Sex Roles, 625-632. Crawford,M., D.(1984). Generic versus specific inclusion of women in language:Effects of recall. Journal of Psycholinguistic Research, 11(5), 373-381. Dictionary of Social Sciences. ( 19 64) Toronto, Candada: Free Press of Glencoe. Eichler,M. (1988). Nonsexist research methods;A practical guide. Boston,MA:Allen & Unwin. Enders-Dragaesser,U. (1988). Paradoxical reality. Women's Studies International Forum, 11(6), 583-590. Evans, B., & Evans,C. (1957). Dictionary of contemporary american usage. New York:Random House Inc.

PAGE 123

Farquhar,N., Dunn,s., & Burr,E. (1972). Sex stereotypes ni elementary and secondary education. In N. Farquhar, s. Dunn & E. Burr (Eds.), Equal Treatment of the sexes:Guidelines for Authors & Editors. (pp. 1-8). Los Angeles,CA:California State University. Fishman,J. (1960). A systematization of the whorfian hypothesis. Behavioral Science, 2, 323-330. Fisk,W. (1985). Responses of neutral pronoun presentations and the development of sex-biased responding. Developmental Psychology, 21(3), 481-485. Hamilton,M. (1988) Using masculine generics:Does generic he increase male bias in the user's imagery? Sex Roles, 19(11/12) 1 785-799. Hamilton,M., Hunter,B., & stuart-Smith,S. (1990). Jury instructions worded in masculine generics. In J. Chrisler & D. Howard (Eds.), New York:SUNY Pess. Harrigan,J., & Lucic,K. (1988). Attitudes about gender bias in language. Sex Roles, 19(3/4), 129-140. Harrison,L., & Passero,R. (1975). Sexism in the language of elementary school textbooks. Language & Society, 12, 22-25. Henley,N. (1989). Molehill or mountain? What we know and don't know about sex bias in language. In M. Crawford & M. Gentry (Eds.), Gender & Thought:Psychological Perspectives (pp.59-78). New York: Springer-Verlag Inc. Henley,N., Gruber,B., & Lerner,L. (1985). Studies on the detrimental effects of "generic" masculine usage. Paper presented at the Eastern Psychological Association, Boston,MA. Henley,N., Gruber,B., & Lerner,L. masculine usage on attitudes and submitted for publication. (1988). Effects of self-esteem. Paper Hughes,D., & Casey,P. (1986). Pronoun choice for genderunspecified agent words. Language and Speech, 29, 59-68. Hyde,J. (1984). Children and sexist language. Developmental Psychology, 20(4), 697-706. 115

PAGE 124

Irigaray,L. (1985, Winter). Any theory of the subject has always been appropriated by the masculine. Trivia. A Journal of Ideas, 38-46. Jacobson,M., & Insko W. Jr. (1985). Use of nonsexist pronouns as a function of one's feminist orientation. Sex Roles, 19(11/12), 785-799. Kidd,V. (1971). Study of the images produced through the use of the male pronoun as the generic. Moments in Contemporary Rhetoric & Communication, 1(2), 25-30. MacKay,D., & Fulkerson,D. (1979). On the comprehension and production of pronouns. Journal of Verbal Learning & Verbal Behavior, 18, 661-673. Mandelbaum,D. (1949). Selected writings of edward saoir in language, culture and personality. Berkeley & Los Angeles: University of California Press. Martyna,W. (1978, Winter). What does 'he' mean?:Use of the generic masculine. Journal of Communication, 131-138. McConnell-Ginet,s., Borker,R. & Furman,N. (1980) Women in Literature and Society. New York:Praeger. Miller,c., & swift,K. (1988). The handbook of nonsexist writing. New York:Harper & Row,Publishers. Moulton,J., Robinson, G., & Elias,c. (1978). Sex bias in language use. American Psychologist, Nov, 1032-1036. Murdock,N., & Forsyth,D. (1980). Is gender-biased language sexist? A Perceptual Approach. Paper presented at the meeting of the Southeastern Psychological Association, Washington,DC. Nilsen,A. (1977). Linguistic sexism as a social issue. In A. Nilsen, H. Lee Gershuny & J. Stanley (Eds.) Sexism and language. Illinois: National Council of Teachers of English. Oxford American Dictionarv. (1980). New York:Oxford University Press. Penelope,J. (1990). Speaking freely: unlearning the lies of the fatherstongues. Elmsford,NY:Pergamon Press. Rosch,E. (1974). Linguistic relativity. In A. Silverstein (Ed.), Human Communication:Theoretical Explanations. Hillsdale, NJ:Erlbaum. 116

PAGE 125

Salter, M.M. (1979). The effects of the generic use of 1 she 1 1 he 1 and 1 they 1 : Sex of speaker and sex of listener on speaker credibility. unpublished master 1 s thesis, University of Georgia. Sapir, E. (1921).Language: An introduction to the study of speech. New York:Harcourt, Brace & Co. Inc. Sapir,E. (1929). Status of linguistics as a science. Language, a, 207-214. Schneider,J., & Hacker,s. (1973). Sex role imagery and use of the generic "man" in introductory texts. American Sociologist, 12-18. Silveira,J. (1980). Generic masculine words and thinking. In c. Kramarae (Ed.), The Voices & Words of Women and Men. Oxford:Pergamon Press. Spock,B. (1970). Decent & indecent. New York:McCall Publishing Co. Stanley,J. (1978). The sexist tradition:Words and meanings. Iowa English Bulletin, May, 5-9. Stericker,A. (i981). Does this "he or she" business really make a difference? Sex Roles, 2(6), 637-641. Ward,J. (1975). Attacking the king1s english. Journalism Quarterly, 52(4), 699-705. Whorf,B.L. (1940). Science and linqusitics. Technology Review, 42(6), 229-248. Whorf ,B.L. (1956). Lancruaae. thought and reality. Cambridge: Technology Press of Massachusetts Institute of Technology. 117