The Animal = Male Hypothesis: children's and adults' beliefs about the sex of non-sex-specific stuffed animals

Over the past 30 years, dozens of studies have demonstrated that masculine generic terms such as he and man create bias in the perceptions of adults (e.g., Hamilton, 1988; Martyna, 1978; Moulton, Robinson, & Elms, 1978) and children (e.g., Hyde, 1984; Switzer, 1990). For instance, many studies show that masculine terms lead to more male than female mental imagery (e.g., Hamilton, 1988; Hamilton & Henley, 1982; Martyna, 1978; Switzer, 1990), others reveal that perceivers do not think a woman fits in a sentence worded in the masculine generic (e.g., MacKay & Fulkerson, 1979; Martyna, 1978; Silveira, 1980), and yet others have highlighted concretely harmful effects caused by masculine generics. The harmful effects are as diverse as lowered job interest among women (Bem & Bem, 1973), biased beliefs about who will succeed in occupations (e.g., Hamilton, 1989; Hyde, 1984), and biased decisions about a woman's guilt in a mock murder trial (Hamilton, Hunter, & Stuart-Smith, 1994). As Hyde (1984) stated, "although his may be gender-neutral in a grammatical sense, it is not gender neutral in a psychological sense." (p. 698).

Neutral generics (e.g., person, people, singular or plural they) and inclusive generics (e.g., men and women, he or she), according to some of these same studies and other research, are less male-biased than are masculine generics. But they are not perfect. Silveira (1980) asserted a People = Male Hypothesis, claiming that even gender-neutral terms such as he or she and person are male-biased, in that a male is thought of as a more typical person than is a female. Put another way, there is a tendency to assume a person is male unless there is specific information to the contrary.

Hamilton (1991) performed three studies to test Silveira's hypothesis. In Study 1, college students heard one of two unbiased pronoun versions of a science fiction story that described the lives of people of the future. The neutral version contained phrases such as "When children began their education at age three, they..." and "People had to learn..." The inclusive version had "When a child began his or her education at age three, he or she..." and "Men and women had to learn..." Male participants, but not female participants, reported more male than female characters in their mental imagery in response to both forms of the story. In Study 2, when college students were asked to describe the most typical person they could imagine and to give the person a first name, they were more likely to describe a male than a female and to provide a male first name than a female first name. In Study 3, Hamilton discovered that female and male participants were more likely to refer to a male protagonist as a person or an i ndividual than to refer to a male with the sex-specific term man. In contrast, in referring to a female protagonist, participants more often used the sex-specific term woman. In other words, a man is a person; a woman is a woman.

Merritt and Kok (1995) replicated and extended Hamilton's work (Hamilton, 1991) by having college students read a scripted dialogue in which the target character was given the gender-neutral name Chris. There were three versions of the script: a feminine version (interpersonal script), a neutral version (education script), and a masculine version (business script). After reading the script, students were asked whether they believed Chris was female or male. The researchers found a strong people = male bias, in that more participants saw Chris as male than as female regardless of which script they had read.

Other researchers, in contrasting masculine to neutral generic terms, have provided indirect evidence for the People = Male Hypothesis. Neutral terms, such as child and adult (Wise & Rafferty, 1982), persons, people, and men and women (Sniezak & Jazwinski, 1986), and the singular they (Hyde, 1984), have been shown to be less male-biased than their equivalent masculine generic terms, but were not found to be truly neutral.

Some studies of masculine generic terms have shown sex-related differences in the degree of male bias, such that men display more male-biased responses to such terms as he and mankind than women do (e.g., Gastil, 1990; Hamilton, 1988; Martyna, 1978; Moulton et al., 1978), and boys display more male-biased responses than girls do (Hyde, 1984; Switzer, 1990). Silveira (1980) similarly predicted that males would have more people = male bias in response to gender-neutral terms than would females. Of the four people = male studies performed by Hamilton (1991) and Merritt and Kok (1995), only one supported this prediction: Male participants in Hamilton's science fiction story experiment showed the people = male bias (Hamilton, 1991); female participants did not. In contrast, in the study on referring to a male protagonist as a person but a female protagonist as a woman, Hamilton discovered that both sexes were subject to the bias with no sex-related difference in its strength. In Hamilton's typical person study, th ere were too few male participants to analyze results separately by participant sex. And finally, Merritt and Kok (1995) found that the people male bias in the tendency to believe Chris was male was equally strong for women and men. Of the studies that indirectly tested the People = Male Hypothesis (Hyde, 1984; Sniezak & Jazwinksi, 1986; Wise & Rafferty, 1982), only one showed the predicted sex-related difference. Sniezak and Jazwinski found that men drew more pictures of males and gave more male names than women did in response to the gender-neutral term people. Wise and Rafferty found no sex-related differences in response to the neutral pronoun, and Hyde did not test for sex-related differences on this variable. Thus Silveira's prediction of a sex-related difference in the people = male bias has received only marginal support to date.

One question that Silveira did not address is whether the people = male bias might extend beyond people to animals, a form of the bias we call the animal = male bias. There is certainly anecdotal evidence that living things, from mosquitoes to bears to flowers, are conceived of as male when their sex is unspecified. Mykol Hamilton has an extensive collection of comic strips, newspaper articles, and excerpts from textbooks and novels that portray sex-unspecified animals, and even inanimate objects, as male. Some of the animals and objects that are called he, given masculine names, or otherwise referred to as male in these sources are lobsters, mice, squirrels, goldfish, birds, polar bears, lizards, frogs, dragonflies, fireflies, horses (even one in which the drawing clearly depicts a mare), dinosaurs, Martians, robots, stuffed animals, grass, flowers, medicines, bridges, the sun, the moon,5 sunbeams, and moonbeams. Sadly, yet rather amusingly, the collection also contains many examples of a category we call th e impossible male, that is, animals referred to as male when the more visible members of the species--the workers or biters--are all female (for example, ants, bees, and mosquitoes). Finally, the collection contains a subsection of exceptions to the rule, or sources in which sex-unspecified living things or inanimate objects are referred to as female. This subsection is extremely small! An excerpt from a novel illustrates that perhaps it is not too much to posit a World = Male Hypothesis. In The Case of the Missing Bront (Barnard, 1983), the narrator states that all the world and his wife were present at a particular social event.

In addition to the anecdotal evidence, two studies lend empirical support to the Animal = Male Hypothesis. DeLoache, Cassidy, and Carpenter (1987) asked mothers to read picture books to their children, and they found that the mothers referred to animals of indeterminate sex as male 95% of the time. Incidental findings of a study on children's moral orientation (Beal & Garrod, 1997) indicate that even when children read about female animal characters, they tend to remember them as male unless illustrations clearly mark the characters' femaleness.