P1: JQX

Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

Archives of Sexual Behavior, Vol. 33, No. 2, April 2004, pp. 87–96 (°
C 2004)

The Inexorable Rise of Gender and the Decline of Sex:

Social Change in Academic Titles, 1945–2001

David Haig, Ph.D.1

Received January 28, 2003; revision received May 8, 2003; accepted May 8, 2003

More than 30 million titles of “academic” articles, from the years 1945–2001, were surveyed for
occurrences of the words sex and gender. At the beginning of this period, uses of gender were much
rarer than uses of sex, and often used in the sense of a grammatical category. By the end of this
period, uses of gender outnumbered uses of sex in the social sciences, arts, and humanities. Within
the natural sciences, there was now more than 1 use of gender for every 2 uses of sex. The beginnings
of this change in usage can be traced to Money’s introduction of the concept of “gender role” in 1955
(J. Money, 1955). However, the major expansion in the use of gender followed its adoption by feminists
to distinguish the social and cultural aspects of differences between men and women (gender) from
biological differences (sex). Since then, the use of gender has tended to expand to encompass the
biological, and a sex/gender distinction is now only fitfully observed.

KEY WORDS: sex; gender; gender role; feminism.

INTRODUCTION that the word should refer only to grammatical categories

(Fletcher, 1991; Goodhart, 1992; Smyth, 1968) or to so-
In The Mill on the Floss, the novelist George Eliot cially but not biologically determined differences
(Mary Ann Evans) (1860) wrote “Public opinion, in these (Fishman, Wick, & Koenig, 1999; Kim & Nafziger, 2000;
cases, is always of the feminine gender—not the world, Lewine, 1994; Pearson, 1996; Walker & Cook, 1998;
but the world’s wife . . .” As this literary example shows, Wilson, 2000).
the use of gender as a synonym for sex has a long pedigree In an attempt to document these changes in usage, I
and is not a recent aberration as is sometimes claimed. The surveyed the titles of over 30 million academic articles,
Oxford English Dictionary quotes uses of gender for sex from the years 1945–2001, for occurrences of the words
from the fifteenth century, although in the first edition of sex and gender. This work extends a similar analysis by
the Dictionary in 1899 this usage was described as jocular. Haig (2000) for the period 1988–1999. The quantitative
From the 1950s, however, a trickle of nonjocular uses of analysis is followed by a discussion of the shifts in mean-
gender began to appear in the academic literature and, by ing of gender over this period.
the 1980s, this trickle had become a flood.
The most important factor was the adoption of gen-
der in the 1970s by feminist scholars as a way of distin- METHOD
guishing “socially constructed” aspects of male–female
differences (gender) from “biologically determined” as- The ISI Web of Science is formed from the amalga-
pects (sex). This distinction is now only fitfully respected, mation of three databases: the Science Citation Index—
and gender is often used as a simple synonym of sex. Expanded (SCI) contains titles from 1945 until the present;
The rise of gender has been accompanied by complaints the Social Sciences Citation Index (SSCI) contains titles
from 1956 until the present; and the Arts & Humanities
1 Department of Organismic and Evolutionary Biology, Harvard Citation Index (AHCI) contains titles from 1975 until the
University, Cambridge, Massachusetts 02138; e-mail: dhaig@oeb. present. The contents of the three databases have consid-
harvard.edu. erable overlap. Thus, an article may be indexed in more

87
0004-0002/04/0400-0087/0 °
C 2004 Plenum Publishing Corporation
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

88 Haig

than one database. The combined database contains over suing correspondence, Harlap [1979] contributed six of 33
30 million titles for the years 1945–2001. gender-containing titles in SCI for 1979). In both articles,
The titles of all English-language articles in the Web however, gender appeared in the title but not in the text,
of Science for the years 1945–2001 were searched for oc- where sex was used. Occasionally, tensions came to the
currences of “sex” and “gender.” Such searches retrieved surface. Ounsted and Taylor (1972) wrote in their edited
titles that contain hyphenated constructs such as “sex- volume, “As between the words ‘sex’ and ‘gender’ even,
specific” or “gender-significant” but did not retrieve titles while preferring the scope of the latter term, we have ac-
in which sex and gender appear as derived forms, such as cepted our authors’ preference for the former where they
“sexual” or “gendered.” In total, the searches found 59,262 wish it” (p. vi). Despite this ecumenical principle, “gen-
sex-containing titles and 29,941 gender-containing titles der” was used in the title of two chapters that used “sex”
(with some titles belonging to both categories). throughout the text, and the title of a third chapter con-
The titles of articles in non-English languages of- tained “gender” in the Table of Contents but “sex” at the
ten appear in the Web of Science as English translations. head of the chapter.
These articles were excluded from searches for “sex” and The journals indexed in the databases varied from
“gender,” but I was unable to exclude them from counts year to year. Therefore, changes in the number of titles
of the number of titles being searched. Therefore, when- containing a particular word will depend only partly on
ever I calculated proportions of titles containing “sex” changes in usage, but will also be influenced by what was
or “gender,” the numerator contained English language and was not included in the database for a particular year.
articles only, but the denominator contained articles in For example, several psychology journals that were cov-
all languages. Most articles in the database are in ered by both SCI and SSCI in 1977 were no longer covered
English, but I do not have a measure of how the propor- by SCI in 1978 (e.g., Child Development, Journal of Per-
tion of non-English articles has changed over time. This sonality and Social Psychology). As a result, 28 SCI titles
is a potentially confounding factor in the interpretation of contained gender in 1977 but only nine contained gender
Figs. 2–4. However, overall trends were little affected by in 1978. The latter figure would have been increased to 25
including or excluding non-English language articles from if titles, now included only in SSCI, had still been included
searches. in SCI. Thus, a corporate decision at the Institute for Sci-
Both sex and gender have uses for which the other entific Information accounted for most of the seemingly
would rarely, if ever, be substituted. Since the late 1970s, anomalous increase in the sex-to-gender ratio of SCI titles
gender in its grammatical sense has contributed a small mi- in 1978 (Fig. 1), although this factor does not explain the
nority of all gender-containing titles. Of somewhat greater rebound to 33 SCI titles containing gender in 1979.
significance are biological uses of sex for which gender is As another example of changes in coverage, the num-
not used (e.g., sex, in the sense of genetic recombination; ber of articles included in SSCI increased by 13% between
sex chromosomes; sex hormones). But probably the most 1994 and 1995. This increase appears to be due to the in-
important uses of “sex” for which “gender” is not a syn- clusion of additional journals not previously covered by
onym relate to copulation and other sexual activities (e.g., SSCI. It is possible that the substantial increase in the pro-
sex, in the sense of sexual intercourse; anal sex; safe sex; portion of titles containing gender that occurred in 1995
sex worker; sex slave). Such uses contribute a relatively (Fig. 3), and the subsequent plateau in this measure, re-
small proportion of sex-containing titles in SCI, but a much flected a change in the composition of SSCI rather than
greater proportion in SSCI and AHCI. (Analysis of a small any change of usage in the academic community; how-
sample suggests that about half of all sex-containing titles ever, a 25% increase in the number of articles covered
in SSCI and AHCI for the year 2001 belong in this cate- by SSCI between 1975 and 1976 does not appear to have
gory. I suspect that the advent of AIDS has increased the affected the relative occurrence of sex and gender. An
frequency of titles in this category, especially in SSCI, but ideal analysis would separate effects of changes in usage
I did not undertake a formal analysis.) from changes in coverage, but I doubt that such an analy-
My analysis focused on usage in titles, but fashions sis would change the gross trends detected by the present
in titles may not entirely reflect the content of articles. Ar- much simpler, and more easily replicable, analysis.
ticles may use gender in the text without it appearing in The databases did not contain book titles, except in
the title, or vice versa. In some cases, titles appeared to re- book reviews, nor the texts of articles and books. More-
flect editorial rather than authorial choices. For example, over, it is probable that use of gender in the titles of articles
articles by Rothman and Liess (1976) and Harlap (1979) in indexed journals, at first, lagged behind conversational
contained the first nongrammatical uses of gender in titles use. My quantitative analysis is restricted to indexed titles.
from the New England Journal of Medicine (with the en- The narrative that follows the quantitative analysis makes
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

The Rise of Gender and Decline of Sex 89

use of other published sources that came to light in my (from 1975 for AHCI). The ratio is expressed on a logarith-
readings. mic scale because this is unbiased with respect to whether
sex or gender appears in the denominator (i.e., 1:2 and 2:1
ratios are represented as equidistant from 1:1). There was
RESULTS substantial noise in the signal for the early years of this
series because of the small number of gender-containing
Prior to the late 1960s, nongrammatical uses of gen- titles.
der were exceedingly rare. For the years 1945–1959, 1,685 Some general observations can be made. The sex-to-
(.14%) SCI titles out of 1,162,909 contained sex but only gender ratio has always been lower in SSCI than in SCI,
five (.0004%) contained gender. Of these, three used gen- but this became more pronounced after 1973 when the
der in a grammatical sense and two were sexological arti- SSCI initiated a sustained decline in the sex-to-gender ra-
cles, both by Money (Money, 1955; Money, Hampson, & tio, which then leveled off in the 1990s (by which time
Hampson, 1957). gender-containing titles outnumbered sex-containing ti-
For the years 1960–1966, 2,094 (.17%) out of tles). A similar decline in the sex-to-gender ratio for SCI
1,253,631 titles in SCI contained sex and eight (.0006%) titles did not start until about 1980 and is still continu-
contained gender, of which three were grammatical uses ing. The ratio for AHCI followed closely that of SSCI,
and five were sexological (including three articles by but with a slightly stronger preference for gender over
Money and coauthors). For these same years, 819 (.24%) sex. The first year for which gender-containing titles ex-
out of 353,069 titles in SSCI contained sex and 12 (.004%) ceeded sex-containing titles was 1987 for AHCI and 1990
contained gender (including four articles by Money and for SSCI. Sex-containing titles have always outnumbered
coauthors). Four gender-containing titles appeared in both gender-containing titles in SCI.
SCI and SSCI. In 1993, the United States Food and Drug Admin-
Figure 1 presents changes in the ratio of sex-contai- istration (FDA) issued a Guideline requiring studies of
ning and gender-containing titles for the years 1966–2001 “gender differences” in all new drug applications (Kessler,

Fig. 1. The ratio of titles containing sex to titles containing gender for all articles in the Science Citation Index (SCI), Social
Sciences Citation Index (SSCI), and Arts & Humanities Citation Index (AHCI).
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

90 Haig

Fig. 2. Proportion of titles in the Science Citation Index containing the word sex and proportion containing the word
gender.

1993). The decline in the sex-to-gender ratio in SCI began expense of sex. The FDA Guideline on the evaluation of
years before this Guideline and was not markedly affected gender differences was possibly responsible for the extra
by it, although there was a small acceleration in the decline large jump in the frequency of gender in 1993.
in 1993. If the ratio of titles in SCI containing “sex differ- For SSCI (Fig. 3), there was a dramatic increase in
ences” to titles containing “gender differences” is consid- the proportion of titles containing sex and/or gender from
ered, this subsidiary ratio had been declining more rapidly 3.4 per 1,000 in 1966 to 16.3 per 1,000 in 2001. Up until
than the overall sex-to-gender ratio since about 1985 (data 1980, both gender and sex increased in tandem. During
not shown). Titles containing “gender differences” first the 1980s, gender began a rapid rise in frequency at the
outnumbered titles containing “sex differences” in 1994 expense of sex. From 1990, the frequency of sex has been
(i.e., in the year following the Guideline) and have done roughly constant (as has the frequency of gender from
so in every year since (except 1995). 1995). Thus, there is a hint that the relative interest in
Figures 2–4 present changes in the proportion of arti- sex-related subjects has reached a plateau in the social
cles containing sex and gender (expressed as occurrences sciences.
per thousand titles) for each of the three databases for The AHCI database contains data from 1975 until
the same years as covered in Fig. 1. Note that the vertical present. Figure 4 shows a dramatic increase over this
scales have been adjusted to reflect the fact that the propor- period in the proportion of titles containing sex and/or
tion of titles containing sex and/or gender was far higher in gender, from .6 per 1,000 in 1975 to 7.1 per 1,000 in 2001,
SSCI than SCI, with AHCI intermediate. For SCI (Fig. 2), with a slight lag relative to the corresponding increase in
there was a small increase in the proportion of titles con- SSCI (on the other hand, the fall in the sex-to-gender ratio
taining sex and/or gender over this period, from 1.8 per in AHCI was slightly ahead of the decline in SSCI). The
1,000 in 1966 to 2.7 per 1,000 in 2001. From about 1980, rapid rise in the frequency of gender began in about 1982,
gender began a steady increase in frequency, partly at the with a slower rise of sex from the late 1980s. Unlike SCI
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

The Rise of Gender and Decline of Sex 91

Fig. 3. Proportion of titles in the Social Sciences Citation Index containing the word sex and proportion containing the
word gender.

and SSCI, the rapid rise of gender was not associated with The juxtaposition of role and status in the above def-
a decline in the frequency of sex. inition suggests that Money was influenced by Parson’s
concept of sex roles. Money received his PhD in 1952 from
the Department of Social Relations at Harvard University
DISCUSSION and listed Parsons among his teachers (Money, 1986, p.
5). For Parsons (1949), a status was “any patterned defini-
Sexological Origins tion of who and what a person is” whereas a role was “the
dynamic aspect of status, the behavior counterpart of the
The first title in SCI to use gender in a nongrammat- ideal or expected position defined by a status” (p. 43). Uses
ical sense was Hermaphroditism, gender and precocity of sex role from the 1940s can be found in Parsons (1940,
in hyperadrenocorticism: Psychologic findings (Money, 1942), Cottrell (1942), and Mead (1949, p. 73). One of
1955). This article introduced the concept of a gender the many ironies to emerge from my analysis is that dis-
role: “The term gender role is used to signify all those cussion of sex roles is now a staple of sociobiology (e.g.,
things that a person says or does to disclose himself or Vincent, 1994) without awareness of the term’s origin in
herself as having the status of boy or man, girl or woman, sociology.
respectively. It includes, but is not restricted to, sexuality Money (1996) later wrote that he imported the term
in the sense of eroticism.” This was one of a series of pa- gender into sexological science “to make it possible to
pers by Money and his collaborators that appeared in the write about people who came into one’s office as either
Bulletin of the Johns Hopkins Hospital during that year. male or female, but of whom it could not be said that
Other papers in the series employed the concept of gen- their sex role in the specific genital sense was either male
der role (Money, Hampson, & Hampson, 1955a, 1955b), or female insofar as they had a history of birth defect of
without gender appearing in their titles. the sex organs.” He then continued grandiloquently, “The
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

92 Haig

Fig. 4. Proportion of titles in the Arts & Humanities Citation Index containing the word sex and proportion containing the
word gender.

majority of people who contributed to this new meaning of At this stage, it is worth discussing the causal con-
gender were hermaphrodites or intersexes. To them social notations that had built up around gender. Money (1955)
science and social history overall owe a debt of gratitude. concluded that “Gender role and outlook as boy or man,
It is impossible to write about the political history of the girl or woman, was found to be in agreement with sex of
second half of the twentieth century without reference to rearing, except in three cases, and not to be automati-
the concept of gender. This is particularly true with respect cally or instinctively determined by chromosomes, go-
to the women’s movement in politics” (p. xii). For similar nads or hormones.” Similarly, Money et al. (1957) ob-
reminiscences, and claims of priority, see Money (1973, served that “the sex of assignment and rearing is consis-
1985, 1995, p. 17ff.). tently and conspicuously a more reliable prognosticator of
Excluding grammatical uses, most, if not all, of the a hermaphrodite’s gender role and orientation than is the
gender-containing titles in SCI and SSCI from the 1960s chromosomal sex, the gonadal sex, the hormonal sex, the
appear to have derived the term from Money. These papers accessory internal reproductive morphology, or the am-
were mostly published in psychological journals and, at biguous morphology of the external genitalia.” They em-
first, were concerned with individuals who did not con- phasized that “our findings indicate that neither a purely
form to sexual stereotypes (hermaphrodites, transsexuals, hereditary nor a purely environmental doctrine of the ori-
transvestites, homosexuals, feminine boys, and masculine gins of gender role and orientation—of psychologic sex—
girls). However, in the late 1960s and early 1970s, gen- is adequate.”
der began to appear in the titles of articles that addressed Money and his co-workers offered two revealing
the behaviors and choices of individuals who conformed analogies for the acquisition of a gender role: the first
to gender stereotypes, with an emphasis on the extent to was the child’s acquisition of a natural language (Money,
which the stereotypes were mutable or immutable, bio- 1955; Money et al., 1957); the second was the imprint-
logical or social. ing of a duckling on Konrad Lorenz when he imitated the
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

The Rise of Gender and Decline of Sex 93

quacking of a mother duck (Money et al., 1957). In both Feminist Adoption

these examples, the individual was seen as biologically
primed to acquire a language or mother figure, but The origins of the use of gender among feminist
which language was acquired, or what individual was scholars has been variously dated to the late 1960s
identified as mother, was determined by the environment. (Nicholson, 1994) or the mid-1970s (Unger & Crawford,
Consistent with these analogies, Money et al. (1957) 1993). My own analysis suggests that its widespread adop-
believed that gender role was acquired very early in a tion in feminist circles was delayed until the late 1970s or
child’s development and once acquired was resistant to early 1980s. The first gender-containing title in the Web of
change: “Though the sex of rearing could transcend ex- Science that had an explicitly feminist context was Some
ternal genital morphology in psychologic importance, ab- evolutionary aspects of human gender (Tobach, 1971),
sence or correction of ambiguous genital appearance was in an issue of the American Journal of Orthopsychiatry
psychologically beneficial. Reassignment of the sex of devoted to The Women’s Movement: Social and Psycho-
rearing after the early months of life was, without doubt, logical Perspectives. In this article, Tobach differentiated
psychologically injurious.” Although Money explicitly “biological sex” from “societally assigned gender” and
adopted an interactionist position as regards nature ver- warned against using “concepts from evolutionary biol-
sus nurture, his work was implicitly read as lying at the ogy to justify either retaining old traditions or changing
nurture-end of the spectrum. Because a person’s sex could them.” Her article cited neither Money nor Stoller.
differ from their gender role, gender became associ- Other early feminist uses of gender occurred in books
ated with a blurring of the male/female dichotomy, and the (not indexed in the Web of Science). Holter (1970) used
claim that upbringing trumped anatomy provided a power- sex and gender as interchangeable synonyms, seemingly
ful argument against the essential nature of sex for variety, whereas Millett (1970, p. 29) makes only pass-
differences. ing reference to a sex/gender distinction, which she illus-
An early person to employ the terminology of gender trates with a quote from Stoller (1968). Likewise, Bernard
was the psychoanalyst Stoller (1964a, 1964b). For Stoller (1971, p. 16) derived her definitions of sex and gender
(1965), sex was biological but gender was social. The lat- directly from Stoller (1968). Oakley (1972) defined sex
ter connoted “behavior learned from a tremendous pool as biological and gender as psychological and cultural
of cues present in every culture and from a massive, intri- (pp. 16, 158). After a discussion of the work of Money
cate, though usually subtle, system of rewards and punish- and Stoller, she posed the rhetorical question “Does biol-
ments in which every person lives from birth on” (p. 197). ogy play any role at all in determining the development of
Although he did not deny some role for biology, Stoller gender identity in normal individuals?” and answered:
(1968) wrote that “those aspects of sexuality that are called
The consensus of opinion seems to be that its role is
gender are primarily culturally determined” (p. xiii) and a minimal one, in that the biological predisposition to
that “gender is a term that has psychological or cultural a male or female gender identity (if such a condition
rather than biological connotations” (p. 9). Other psycho- exists) may be decisively and ineradicably overridden by
analysts adopted a similar distinction between biological cultural learning. Those who have worked in the field of
sex and social gender (e.g., Gershman, 1967; Ovesey & hermaphroditic disorders and problems of gender identity
seem very impressed by the power of culture to ignore
Person, 1973). biology altogether, (p. 170).
Stoller (1964b) and Greenson (1964) together intro-
duced the term gender identity at the 23rd International Differences between successive editions of Mascu-
Psycho-Analytical Congress in Stockholm (July–August line/Feminine or Human? (Chafetz, 1974, 1978) are par-
1963). The latter defined this to be “one’s sense of being a ticularly illuminating. In the first edition, Chafetz (1974)
member of a particular sex; it is expressed clinically in the contrasted innate gender with learned sex roles. This edi-
awareness of being a man or male in distinction to being tion contained no citations to Money. However, by the sec-
a woman or female.” For Stoller (1968), “gender identity ond edition (Chafetz, 1978), the terms had been swapped—
starts with the knowledge and awareness, whether con- innate sex was contrasted with learned gender roles—and
scious or unconscious, that one belongs to one sex and references were added to Money and Ehrhardt (1972). Al-
not the other . . . gender role is the overt behavior one dis- lowing for the time lags associated with publication, this
plays in society, the role which he plays, especially with suggests an absence of feminist consensus on the meaning
other people” (pp. 9–10). For Money and Ehrhardt (1972), of gender in the early 1970s with an emerging consensus
“gender role is the public expression of gender identity, by the late 1970s (see Gould & Kern-Daniels, 1977). This
and gender identity is the private expression of gender timing is supported by Unger (1979) who was able to
role” (p. 4). write in the American Psychologist, “The term gender is
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

94 Haig

introduced for those characteristics and traits sociocultur- avoidable despite her “dissatisfaction,” and in these places
ally considered appropriate to males and females” (my she “used it with reservation” (pp. 8–10).
emphasis). From these small beginnings, use of gender became
The only use of gender that I can find in Women, widely adopted by feminists during the 1980s. It is this
Culture, and Society (Rosaldo & Lamphere, 1974) is in adoption that I believe is principally responsible for the ex-
the psychoanalytic chapter by Chodorow. Significantly, plosive growth in gender-containing titles that is observed
Ortner did not use gender in her influential chapter—Is in SSCI and AHCI during that decade (see Figs. 3 and
Female to Male as Nature is to Culture? (an ammended 4). Feminists were able to embrace the concept of gender
version of Ortner, 1972)—but 7 years later she was an ed- as their own contribution to discourse as the term’s ear-
itor of Sexual Meanings: The Cultural Construction of lier association with sexological science shifted into the
Gender and Sexuality (Ortner & Whitehead, 1981). In background.
Gender and Sex in Society, Duberman (1975) defined sex Feminist usage converged on a contrast between so-
as “an ascribed social status referring to the biological cially constructed gender and biologically determined sex.
differences between people” whereas gender role referred However, it proved difficult to maintain such a distinction.
to “the socially learned patterns of behavior that differ- One problem with the simple dichotomization of biologi-
entiate men from women in a given society” (p. 26). In cal sex and social gender was that no term remained to refer
Toward an Anthropology of Women, Rubin (1975) dis- to situations in which causation was unknown, disputed,
cussed the sex/gender system, which she defined as “the or involved an interaction between biology and culture.
set of arrangements by which a society transforms bi- Thus, the choice of term for this middle ground became a
ological sexuality into products of human activity, and simple matter of preference, blurring the conceptual dis-
in which these transformed sexual needs are satisfied” tinction between terms. Moreover, among feminists, the
(p. 159). domain of gender had a tendency to expand to subsume
Trends in feminist use of gender were assessed by the category of sex, because the way that people talk about
scanning the contents of early issues of Feminist Stud- “male” and “female” bodies was also seen as socially
ies (first issue in 1972) and Signs: Journal of Women in constructed (discussed by Nicholson, 1994). Kessler and
Culture and Society (first issue in 1975). The first gender- McKenna (1978) provided an uncompromising example
containing titles in Feminist Studies did not appear until of this position. They saw the element of social construc-
Volume 5 (Davidoff, 1979) and Volume 6 (Vance, 1980). tion as primary in all aspects of maleness and femaleness:
These authors derived their uses of gender from Oakley even to invoke two categories was a social construct. To
(1972) and Rubin (1975), respectively. Yudkin (1978) had emphasize their contention, they wrote of gender chromo-
earlier used gender in a philosophical discussion of trans- somes and gender hormones. In a retrospective, McKenna
sexualism, but without the term appearing in the title. She and Kessler (2000) returned to this theme: “Retaining
constructed a trichotomy between biological sex, psycho- a separation between sex and gender, even if it is pro-
logical gender, and social sex role. Her use of gender de- posed that both are socially constructed, raises the ques-
rived from Money and Stoller. The first issue of Signs de- tion of why biology is so important that it merits a special
fined the journal’s scope as including both sex and gender category.”
(Stimpson, Burstyn, Stanton, & Whisler, 1975), but use of Given the expansion in the domain of gender, and a
gender was sparse in early issues (and predominantly by certain indeterminacy in its meaning, it is hardly surprising
male authors). The first gender-containing title in Signs that some authors who were unfamiliar with the subtleties
did not appear until the sixth volume (Baker, 1980), in a of feminist debate interpreted gender as a simple synonym
review of the biological literature on sex differences that for sex and adopted it as such in their own writings. This
contained numerous references to Money and coworkers. is unambiguously demonstrated when gender is used in
Gender-containing titles first exceeded sex-containing ti- relation to the physiology of nonhuman animals, without
tles in Volume 11 of Signs (1986–1987). any implication of a determining role of culture in the
Gender did not achieve uncontested acceptance by causation of observed differences. Such titles first appear
all feminists. In Transsexual Empire, Raymond (1979) in the 1970s (e.g., Hahn, Norton, & Fishman, 1977) and
treated gender as a technical or therapeutic term asso- are now common in SCI.
ciated with the work of Money and Stoller. She found the The appearance of gender in a title from the natural
term to have “certain problems for a feminist critic” as it sciences now communicates little if anything about cau-
gives “the impression that there is a fixed set of psychoso- sation or the ideology of the author. Among the reasons
cial conditions that determines gender identity and role.” that working scientists have given me for choosing gen-
Nevertheless, there were times that she found the word un- der rather than sex in biological contexts are desires to
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

The Rise of Gender and Decline of Sex 95

signal sympathy with feminist goals, to use a more aca- REFERENCES

demic term, or to avoid the connotation of copulation.
Baker, S. W. (1980). Biological influences on human sex and gender.
Signs: Journal of Women in Culture and Society, 6, 80–96.
Conclusion Bernard, J. (1971). Women and the public interest. Chicago: Aldine.
Chafetz, J. S. (1974). Masculine/feminine or human? An overview of the
sociology of sex roles. Itasca, IL: F. E. Peacock.
This article addressed the history of terminology. Chafetz, J. S. (1978). Masculine/feminine or human? An overview of the
During the first half of the twentieth century, gender sociology of the gender roles (2nd ed.). Itasca, IL: F. E. Peacock.
Cottrell, L. S. (1942). The adjustment of the individual to his age and
appears to have been used predominantly in its grammat- sex roles. American Sociological Review, 7, 617–620.
ical sense, but its existing (albeit rare) use as a synonym Davidoff, L. (1979). Class and gender in Victorian England: The diaries
of sex was readily available for anyone who wished to of Arthur J. Munby and Hannah Culwick. Feminist Studies, 5(1),
87–141.
emphasize a dichotomy between different sources of sex- Duberman, L. (1975). Gender and sex in society. New York: Praeger.
associated differences or to establish a separate domain Eliot, G. (1860). The mill on the floss. Edinburgh: Blackwood.
for territory that had previously been considered part of Fishman, J. R., Wick, J. G., & Koenig, B. A. (1999). The use of “sex” and
“gender” to define and characterize meaningful differences between
the realm of sex. The expansion of the use of gender in the men and women. In Agenda for research on women’s health for the
second half of the century appears to have derived from 21st century. (A report of the Task Force on the NIH Women’s
Money’s concept of a gender role, introduced in the 1950s Health Research Agenda for the 21st century), (Vol. 2, pp. 15–19).
Rockville, MD: Office of Research on Women’s Health.
to refer to the self-identification of individuals whose gen- Fletcher, D. E. (1991). The eternal battle of sex vs gender. JAMA, 266,
ital sex was ambiguous. Significantly, in Money’s usage, 2833.
an individual’s gender role could differ from various bi- Gershman, H. (1967). The evolution of gender identity. Bulletin of the
New York Academy of Medicine, 43, 1000–1018.
ological definitions of an individual’s sex. From this be- Goodhart, C. B. (1992). Sex and gender. Nature, 359, 182.
ginning, there was a slow but gradual increase in the use Gould, M., & Kern-Daniels, R. (1977). Toward a sociological theory of
of gender through the 1960s by writers, especially in the gender and sex. American Sociologist, 12, 182–189.
Greenson, R. R. (1964). On homosexuality and gender identity. Interna-
social sciences and among psychoanalysts, who wished to tional Journal of Psychoanalysis, 45, 217–219.
emphasize the environmental, social, or psychologic de- Hahn, E. F., Norton, B. I., & Fishman, J. (1977). Influence of gender and
terminants of psychologic/behavioral differences between castration on liver and CNS N-demethylation of morphine in rats.
Life Sciences, 20, 95–99.
men and women. Some of these writers would have con- Haig, D. (2000). Of sex and gender. Nature Genetics, 25, 373.
sidered themselves feminists or at least sympathetic to the Harlap, S. (1979). Gender of infants conceived on different days of the
goals of the women’s movement. Debates about nature menstrual cycle. New England Journal of Medicine, 300, 1445–
1448.
versus nurture, the biological versus the social, and the Holter, H. (1970). Sex roles and social structure. Oslo, Norway: Univer-
autonomy of the social from the natural sciences, were of sitetsforlaget.
course much older than their association with a termino- Kessler, D. A. (1993). Guideline for the study and evaluation of gender
differences in the clinical evaluation of drugs. Federal Register,
logical sex vs. gender distinction. 58(139), 39406–39416.
Prior to the early 1980s, the rise in the use of gender in Kessler, S. J., & McKenna, W. (1978). Gender: An ethnomethodological
academic titles was not associated with an appreciable de- approach. New York: Wiley.
Kim, J. S., & Nafziger, A. N. (2000). Is it sex or is it gender? Clinical
cline in the use of sex. The major increase in the use of gen- Pharmacology & Therapeutics, 68, 1–3.
der, and the associated decline of sex, occured in the 1980s Lewine, R. R. J. (1994). Sex: An imperfect marker of gender. Schizophre-
after the adoption of gender as a technical term in feminist nia Bulletin, 20, 777–779.
McKenna, W., & Kessler, S. J. (2000). Retrospective response. Feminism
discourse. The available evidence strongly suggests that and Psychology, 10, 66–72.
this usage was derived by descent with modification from Mead, M. (1949). Male and female: A study of the sexes in a changing
Money. As the sex-to-gender ratio has declined, gender world. New York: William Morrow.
Millett, K. (1970). Sexual politics. New York: Doubleday.
has come to be adopted as a simple synonym, perhaps a Money, J. (1955). Hermaphroditism, gender and precocity in hypera-
euphemism, for sex by many writers who are unfamiliar drenocorticism: Psychologic findings. Bulletin of the Johns Hopkins
with the term’s recent history. Hospital, 96, 253–264.
Money, J. (1973). Gender role, gender identity, core gender identity:
Usage and definition of terms. Journal of the American Academy
of Psychoanalysis, 1, 397–402.
ACKNOWLEDGMENTS Money, J. (1985). Gender history, theory and usage of the term in sexol-
ogy and its relationship to nature nurture. Journal of Sex and Marital
Therapy, 11, 71–79.
This paper has benefited from the comments of Janet Money, J. (1986). Venuses/penuses. Buffalo, NY: Prometheus.
Saltzman Chafetz, Sarah Hrdy, Stephanie Kenan, Joanne Money, J. (1995). Gendermaps. New York: Continuum.
Money, J. (1996). Preface. In J. Money & A. A. Ehrhardt, Man and
Meyerowitz, Lucas Mix, Judith Ryan, Kenneth J. Zucker, woman, boy and girl. Gender identity from conception to maturity.
and the anonymous referees. Northvale, NJ: Jason Aronson. (Original work published 1972)
P1: JQX
Archives of Sexual Behavior pp1108-aseb-480174 January 24, 2004 11:20 Style file version July 26, 1999

96 Haig

Money, J., & Ehrhardt, A. A. (1972). Man and woman, boy and girl. The Rosaldo, M. Z., & Lamphere, L., (Eds.). (1974). Women, culture, and
differentiation and dimorphism of gender identity from conception society. Stanford, CA: Stanford University Press.
to maturity. Baltimore: Johns Hopkins University Press. Rothman, K. J., & Liess, J. (1976). Gender of offspring after oral
Money, J., Hampson, J. G., & Hampson, J. L. (1955a). Hermaphroditism: contraceptive use. New England Journal of Medicine, 295, 859–
Recommendations concerning assignment of sex, change of sex, 861.
and psychologic management. Bulletin of the Johns Hopkins Hos- Rubin, G. (1975). The traffic in women: Notes on the “political economy”
pital, 97, 284–300. of sex. In R. R. Reiter (Ed.), Toward an anthropology of women (pp.
Money, J., Hampson, J. G., & Hampson, J. L. (1955b). An exami- 157–210). New York: Monthly Review Press.
nation of some basic sexual concepts: The evidence of human Smyth, D. H. (1968). Gender and sex. British Medical Journal, 2, 368.
hermaphroditism. Bulletin of the Johns Hopkins Hospital, 97, 301– Stimpson, C. R., Burstyn, J. N., Stanton, D. C., & Whisler, S. M. (1975).
319. Editorial. Signs: Journal of Women in Culture and Society, 1, v–viii.
Money, J., Hampson, J. G., & Hampson, J. L. (1957). Imprinting and the Stoller, R. J. (1964a). Gender-role change in intersexed patients. JAMA,
establishment of gender role. Archives of Neurology and Psychiatry, 188, 684–685.
77, 333–336. Stoller, R. J. (1964b). A contribution to the study of gender identity.
Nicholson, L. (1994). Interpreting gender. Signs: Journal of Women in International Journal of Psychoanalysis, 45, 220–226.
Culture and Society, 20, 79–105. Stoller, R. J. (1965). Passing and the continuum of gender identity. In
Oakley, A. (1972). Sex, gender, and society. New York: Harper Colophon. J. Marmor (Ed.), Sexual inversion (pp. 190–210). New York: Basic
Ortner, S. B. (1972). Is female to male as nature is to culture? Feminist Books.
Studies, 1(2), 5–31. Stoller, R. J. (1968). Sex and gender: The development of masculinity
Ortner, S. B., & Whitehead, H. (Eds.). (1981). Sexual meanings: The and femininity. London: Hogarth.
cultural construction of gender and sexuality. Cambridge, UK: Tobach, E. (1971). Some evolutionary aspects of human gender. Ameri-
Cambridge University Press. can Journal of Orthopsychiatry, 41, 710–715.
Ounsted, C., & Taylor, D. C. (Eds.). (1972). Gender differences: Their Unger, R. K. (1979). Toward a redefinition of sex and gender. American
ontogeny and significance. Edinburgh: Churchill Livingstone. Psychologist, 34, 1085–1094.
Ovesey, L., & Person, E. (1973). Gender identity and sexual psy- Unger, R. K., & Crawford, M. (1993). Sex and gender—The troubled
chopathology in men: A psychodynamic analysis of homosexu- relationship between terms and concepts. Psychological Science, 4,
ality, transsexualism, and tranvestitism. Journal of the American 122–124.
Academy of Psychoanalysis, 1, 53–72. Vance, C. S. (1980). Gender systems, ideology, and sex research: An
Parsons, T. (1940). An analytical approach to the theory of social strati- anthropological analysis. Feminist Studies, 6, 129–143.
fication. American Journal of Sociology, 45, 841–862. Vincent, A. C. J. (1994). Seahorses exhibit conventional sex roles in
Parsons, T. (1942). Age and sex in the social structure of the United mating competition, despite male pregnancy. Behaviour, 128, 135–
States. American Sociological Review, 7, 604–616. 151.
Parsons, T. (1949). Essays in sociological theory pure and applied. Glen- Walker, P. L., & Cook, D. C. (1998). Gender and sex: Vive la difference.
coe, IL: Free Press. American Journal of Physical Anthropology, 106, 255–259.
Pearson, G. A. (1996). Of sex and gender. Science, 274, 328–329. Wilson, D. R. (2000). Gender vs sex. JAMA, 284, 2997–2998.
Raymond, J. G. (1979). The transsexual empire: The making of the she- Yudkin, M. (1978). Transsexualism and women: A critical perspective.
male. Boston: Beacon. Feminist Studies, 4(3), 97–106.