Gender, Nature, and Nurture - Richard A. Lippa 2014
The Search Commences
Masculinity and Femininity: Gender Within Gender
.. I felt I had with an impious and secret finger traced a first wrinkle upon [my mother's] soul and made the first white hair shew upon her head. This thought redoubled my sobs, and then I saw that Mamma, who had never allowed herself to go to any length of tenderness with me, was suddenly overcome by my tears and had to struggle to keep back her own. Then, as she saw that I had noticed this, she said to me, with a smile: "Why, my little buttercup, my little canary-boy, he's going to make Mamma as silly as himself if this goes on ..."
—Remembrance of Things Past
Marcel Proust (1934)
One of the most revered novelists of 20th century, Marcel Proust possessed great literary, artistic, and musical sensibilities. He was introspective, emotionally sensitive, physically delicate, foppish, and averse to the rough-and-tumble. Witty, verbal, and drawn to the mannered life of aristocratic salons, he was inordinately attached to his mother and sexually attracted to men. In short, it seems reasonable to describe Proust as feminine.
Proust provides a concrete example of what common sense tells us; some men are more masculine and some more feminine than others. But what do the words masculine and feminine mean? Proust's traits suggest some possibilities. Femininity (the opposite of masculinity?) consists of emotional sensitivity; artistic sensibility; a focus on manners; a tendency to timidity and nonaggressiveness; a nurturant, attached orientation to others; and sexual attraction to men. Admittedly, all of these feminine characteristics are stereotypic. They reflect an essentialist view of femininity, that there are core qualities to femininity a Platonic essence if you will, which exist despite cultural and historical variations.
To research psychologists, the concepts of masculinity and femininity have referred to individual differences (i.e., variations) in people's gender-related traits and behaviors, variations that exist within each sex. Masculinity and femininity refer to those aspects of gender that vary among men and among women. Chapter 1 considered the question: How much do men and women differ? We turn now to the second key question related to gender: How do men vary in their masculinity, and how do women vary in their femininity?
Research on masculinity and femininity has a long, complex, and controversial history. This may be due in part to the questions addressed. Do masculinity and femininity really exist, and if so, how are they best defined and measured? What causes people's masculinity and femininity to vary: biological factors, parental rearing, or social and cultural learning? Are masculinity and femininity essential traits of the individual, that is, are they fixed traits that exist inside of people? Or are they social constructions, arbitrary concepts foisted upon us by sexist societies? A central question for us is: What molds and determines a person's degree of masculinity and femininity: nature or nurture?
Because the roles of men and women have been the subject of passionate debate in recent years, it is no wonder that masculinity and femininity research has become embroiled in the debate. On one hand, if masculinity and femininity are real traits—perhaps even genetically determined to a significant extent—then gender would seem to be partly wired into us. One the other hand, if masculinity and femininity are social constructions—learned patterns of behavior that are culturally and historically variable—then existing gender roles may be malleable and subject to liberating alternatives.
What in fact does science tell us about masculinity and femininity? To understand research on masculinity and femininity it helps to begin at the beginning, in Palo Alto, California, in the 1920s.
The Search Commences
In 1936 Lewis Terman and Catharine Cox Miles began the modern study of masculinity and femininity with the publication of a classic book, Sex and Personality. In their book Terman and Miles presented both a method for measuring masculinity-femininity and a decade's worth of research investigating masculinity-femininity.
The Analogy Between Masculinity—Femininity and Intelligence
Terman—a Stanford University psychologist—was famous for developing the Stanford-Binet intelligence test, a revised version of which remains to this day a respected and much used test. Miles—who had worked with Terman as a graduate student—was well known for her doctoral dissertation estimating the IQs of eminent historical figures based on biographical information (Cox, 1926), In the late 1920s, after working for at time as a clinical psychologist in Cincinnati, Miles returned to Stanford to assist Terman with his burgeoning research on masculinity-femininity, which Terman described as "about the most interesting thing I have ever tackled" (Lewin, 1984a, p. 161).
During the 1920s, Terman started a classic study of gifted children, in which he identified 856 boys and 672 girls with high IQs in order to trace their social and intellectual development over time. Terman observed that, despite their shared high intelligence, the gifted boys displayed quite different patterns of interests from the gifted girls. Terman reasoned that such sex differences might serve as a means to measure variations in psychological masculinity and femininity within each sex.
Term an proposed that, like intelligence, masculinity-femininity (M-F) was a trait that could he measured through an appropriately designed test. Just as IQ tests provided an objective means to assess intelligence, Term an hoped that his M-F test might "enable the clinician or other investigator to obtain a more exact and meaningful, as well as a more objective, rating of those aspects of personality in which the sexes tend to differ" (Terman & Miles, 1936, p. 6). What was the way to determine whether an item measured a person's masculinity-femininity? (Think of an item here as a question on a self-report questionnaire: for example, "True or False: I like to watch football games.") Terman and Miles proposed that a given question could serve as a measure of M-F if large groups of men and women (or boys and girls) responded to the statement differently, on average. If many more men than women, for example, responded "true" to the statement, "I like to watch football games," then Terman and Miles considered this item to measure M-F, with a true response indicating masculinity and a false response indicating femininity. In contrast, if about equal numbers of men and women answered true to a question (e.g., "I like to go to movies"), then the researchers considered that question to be unrelated to M-F.
It is a well-demonstrated statistical principle in psychological testing that no single test item can provide a reliable measure of the thing we are trying to measure. To obtain a reliable (i.e., a stable and repeatable) test score, researchers must use many test items. To obtain a reliable measure of M-F, Terman and Miles created a 456-item questionnaire, huge by modern standards, which they called the Attitude Interest Analysis Survey. The reason for their bland and uninformative title is that Terman and Miles did not want people who completed their test to realize that it was actually measuring their M-F.
The Attitude interest Analysis Survey was quite varied, including subscales that measured general knowledge, emotions, occupational interests, reading preferences, personality traits, word and picture associations, and attitudes. (See Table 2.1 for some actual items; see if you can guess which responses are masculine and which are feminine.) Some of Terman and Miles' M-F subscales proved to be more reliable than others, in particular, the subtests on knowledge, emotions, occupational preferences, and interests.
Terman and Miles acknowledged that their M-F test was not based on a theory of masculinity or femininity. They also conceded that their test might be culturally limited, based as it was on sex differences "in the present historical period of the Occidental culture of our own country" (p. 6). Their goal, as they saw it, was to accurately and reliably assess individuals' levels of M-F and to investigate whether these levels were related to other interesting physical and psychological characteristics, such as people's educational accomplishments, intelligence, personality traits, body types, and sexual orientation.
Terman and Miles remained open-minded about why men and women varied on M-F: "[The] M-F test rests upon no assumption with reference to the causes operative in determining an individual's score. These may be either physiological and biochemical, or psychological and cultural; or they may be the combined result of both types of influence" (p. 6). Thus Terman and Miles acknowledged the possibility that individual differences in M-F might be a function of both nature and nurture.
The Bipolar Assumption
Terman and Miles' test made an important assumption, that masculinity and femininity are opposites. This necessarily follows from the way their test was constructed and scored. If you answered a question the way women tend to, you necessarily were not answering the question the way men tend to, and vice versa. Raw scores on Terman and Miles' test ranged from negative scores (feminine) to positive scores (masculine). The scoring system, therefore, assumed a single dimension, ranging from feminine to masculine. The more masculine you are, the less feminine your are; and vice versa. Stated a bit more formally, the Terman and Miles proposed a unidimensional (i.e., single dimension) bipolar (either-or) approach to masculinity and femininity. To modify the words of Rudyard Kipling, "Masculine is masculine, and feminine is feminine, and never the
twain shall meet." Note that the hyphenated term masculinity-femininity embodies the bipolar assumption in its very structure.
Terman and Miles's conception of M-F provided the conceptual framework for many subsequent researchers. One noteworthy example was Edward Strong—a colleague of Terman's at Stanford University—who developed one of the first occupational interest tests, the Strong Vocational Interest Blank, which, in updated forms, is still used today (Campbell, 1971; Strong, 1936, 1943). People taking this test are asked to rate how much they like or dislike various occupations and hobbies (e.g., farming, sewing) and how interested they are in taking various school subjects (e.g., geometry, English). Based on his research, Strong came to believe that M-F constituted a major dimension underlying occupational preferences. Accordingly, he developed a M-F scale for his test.
What determined if an occupational preference item was placed on Strong's M-F scale? Like Terman and Miles, Strong selected items for his M-F scale that showed large and statistically significant (i.e., not due to chance) sex differences. If many more men than women expressed an interest in being a farmer and a race car driver, for example, then these items would be placed on the M-F scale, keyed in the masculine direction. Conversely, if many more women than men expressed an interest in being an elementary school teacher and librarian, then these items would be placed on the M-F scale, keyed in the feminine direction. When Strong gave his M-F scale and the Terman and Miles M-F test to the same group of people, he found only a weak correlation between people's scores on the two tests. This early piece of evidence hinted that various M-F scales were not always measuring the same thing.
The 1940s and 1950s witnessed the development of well-known omnibus (i.e., broad, multitrait) personality inventories, including the Guilford-Zimmerman Temperament Inventory (Guilford & Zimmerman, 1956), the California Psychological. Inventory (Gough, 1957), and the Minnesota Multiphasic Personality Inventory (MMPI; Hathaway & McKinley. 1951). Many of these inventories took the Terman and Miles approach to M-F; that is, they assumed that sex differences in response could be used to select and validate items intended to assess M-F.
Because various personality inventories included somewhat different questions, their portraits of M-F varied. The Guilford-Zimmerman scale of masculinity (which, by the bipolar assumption, is the opposite of femininity) assessed inhibited emotional expression, male-typical vocational interests, and a cluster of so-called masculine emotional traits (not being easily disgusted, fearlessness, and a lack of sympathy). The California Psychological Inventory (CPI) was developed to embody folk concepts of personality, that is, dimensions of personality that make sense to lay people. The CPI M-F scale, which was labeled the Fe (femininity) scale, assessed sensitivity, the ability to perceive the nuances of social interaction, acquiescence, compassion, niceness, female-typical work and interests, and lack of interest in politics and social issues. According to this conceptualization, the feminine individual is portrayed as nice, but rather passive, unengaged, and dependent, whereas the masculine individual is somewhat disagreeable, but active, engaged, and independent.
The MMPI is perhaps the best know clinical personality inventory in use. Since its inception in the 1930s and 1940s, the MMPI has been used to diagnose mental illness. Indeed, many of the scales of the MMPI are labeled by the kind of mental illness they are meant to measure and predict (e.g., depression, paranoia, hypochondriasis). As a result, the developers of the MMPI approached the measurement of M-F from the vantage point of psychopathology. In particular, they were interested in masculinity-femininity as a means of diagnosing gender identity disturbances and sexual inversion (i.e., the kind of homosexuality shown by men who act like women or by women who act like men).
Earlier, Terman and Miles had also been interested in the relationship between M-F and homosexuality. Indeed, several chapters of their 1936 book were devoted to this topic. In one study, they collected data from 134 gay men (many of whom were prison inmates) and found that gay men's scores on their M-F test were much more feminine than heterosexual men's, influenced by these findings, the developers of the original MMPI—Starke Hathaway and J. C. McKinley—made a rather unusual decision in developing their Mf (masculinity-femininity) scale (Hathaway, 1956). Rather than initially choosing a set of items that distinguished men from women, they chose instead items that distinguished gay men from heterosexual men. The groups they used to test their first Mf items were quite small: 13 gay men and 54 heterosexual men (all of whom were soldiers). It is not surprising that a number of the items on the original MMPI Mf scale directly addressed sexual orientation, same-sex attraction, and so-called unusual sexual behavior (e.g., "I am very strongly attracted to members of my own sex" and "I have never indulged in any unusual sex practices.")
Once Hathaway and McKinley identified their initial set of Mf items, they used Terman and Miles' strategy to further validate the items. That is, they demonstrated that their M-F items distinguished men from women, and they gathered data to show that their scale distinguished feminine men identified by Terman and Miles' test from so-called normal men. in other words, the MMPI Mf scale was, in part, validated against Terman and Miles' earlier test.
In addition to including items that asked explicitly about same-sex attraction, the original MMPI Mf scale contained items that assessed narcissism and hypersensitivity, stereotypic feminine and masculine interests, heterosexual discomfort and passivity, and imtrospectiveness and social reticence (Greene, 1991). Research has shown that the MMPI Mf scale distinguishes gay men from heterosexual men fairly well (Haslam. 1997). This is not terribly surprising, however, given that a number of items in the original Mf scale asked directly about same-sex attraction. The MMPI was revised and renormed (i.e., administered and calibrated against large contemporary samples of men and women) in the 1980s (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer. 1989). The revised Mf scale omits items that directly ask about same-sex attraction, but the other content remains much the same.
What Is Masculinity—Femininity Related to?
Most of the bipolar M-F tests developed in the 1930s through 1950s showed acceptable levels of reliability, that is, they measured something consistently. But did they also show validity, that is, did they predict real-life behaviors and criteria in a way that made both theoretical and practical sense?
In their early research, Terman and Miles found that school children's M-F scores did not correlate much with their teachers' ratings of how masculine or feminine they were. Similarly, college students' M-F scores did not correlate much with their self-ratings of how masculine or feminine they believed themselves to be. These results were puzzling, for they seemed to raise questions about the validity of the M-F scale, Terman and Miles speculated that these results were due in part to the unreliability of lay people's ratings of their own and other people's masculinity femininity. They proposed (perhaps self-servingly) that their carefully developed M-F test was considerably more reliable than lay judgments and therefore a sounder measurer of people's real M-F.
Terman and Miles investigated additional factors that were linked to M-F. They found, for example, that M-F was somewhat age-related, with individuals—particularly males—showing their highest levels of masculinity in their late teens and early 20s. Not surprisingly, M-F was related to people's interests and academic pursuits. Masculine men tended to be more interested in science and mechanical things and feminine men in cultural pursuits and the arts. Among high school and college-aged women, masculinity was found to be associated with broad interests, high levels of education, and intellectuality. In other words, for women, masculinity was associated with intellectual and educational accomplishment, and if we wanted to place a value judgment on these findings, we might conclude that in this regard, masculinity is good for women.
Later research extended and replicated these early results, indicating that feminine boys and masculine girls tend to show higher levels of creativity, scholastic achievement, and giftedness than more sex-typed children do (Lippa, 1998a; Lubinski & Humphreys, 1990; Maccoby, 1966). (Sex-typed children are those whose traits and behaviors are stereotypic for their sex.) Thus, in terms of creativity and intellectual achievement, femininity can be considered good for boys and masculinity good for girls.
As noted previously, Terman and Miles observed a significant relationship between M-F and sexual orientation. Many subsequent researchers have replicated this finding (Lippa, 2000, 2002; Pillard, 1991): Gay men tend to be more feminine than heterosexual men on M-F scales, and lesbian woman tend to be more masculine than hetereosexual women. Is this good or bad? in Terman and Miles' time, the psychological establishment, as well as society at large, tended to view homosexuality as a kind of mental illness. Thus Terman and Miles' findings were taken as evidence that femininity was bad for men and masculinity was bad for women, for they upped one's odds for sexual deviance. (It is important to note that since the early 1970s, both the American Psychological Association and the American Psychiatric Association have declared that homosexuality is not a mental illness.)
Research on M-F and sexual orientation points to an unstated, if implicit, value judgment that permeated early research on masculinity-femininity, that it is good for people to score in gender-appropriate ways. If you are a man, it's good to be masculine; and if you are a woman, it's good to be feminine. This assumption reflected psychological dogma common throughout the middle part of the 20th century. Developmental psychologists of that period earnestly studied gender socialization and sex typing, the ways in which children learn supposedly appropriate gender roles and behaviors from their parents and from society (Huston, 1983).
But you may recall one set of findings that challenged this assumption, namely, the data that linked boys' femininity and girls' masculinity to creativity and scholastic achievement. In the 1950s and 1960s, other evidence raised additional questions about whether extreme masculinity is necessarily ideal for males or extreme femininity ideal for females. For example, some studies showed that femininity in women was often associated with anxiety, depression, low self-esteem, and meekness, and masculinity in boys and men was associated with aggressiveness and acting out. Eleanor Maccoby (1966), a respected Stanford University developmental psychologist, hypothesized that highly masculine boys might be overly impulsive, whereas highly feminine girls might by overcontrolled, meek, and unassertive. In other words, masculinity in boys and femininity in girls may not be so desirable after all.