WHY ARE THERE SO FEW FEMALE COMPUTER SCIENTISTS? (C) 1991 Ellen Spertus Abstract Women pursue education and careers in computer science far less frequently than men do. In 1990, only 13% of PhDs in computer science went to women, and only 7.8% of computer science professors were female. Additionally, the percentage of female computer science students appears to be increasing at only a slow rate or even decreasing. Apart from ethical con- cerns at women's lack of participation in computer science, the demographics of the country are such that the United States will not have enough engineers and scientists unless underrep- resented groups increase their participation. This report ex- amines the influences against a woman's pursuing a career in a technical field, particularly computer science. Such factors include the different ways in which boys and girls are raised, the stereotypes of female engineers, subtle biases that females face, problems resulting from working in predominantly male environments, and sexual biases in language. Finally, I discuss effective and ineffective ways to encourage women. A theme of the report is that women's underrepresentation is not primarily due to direct discrimination but to subconscious behavior that tends to perpetuate the status quo. This report is MIT Artificial Intelligence Laboratory Technical Report 1315. Introduction Only a small percentage of computer scientists and computer professionals are female. In the most recent years for which statistics are available, women received a third of the bach- elor's degrees in computer science, 27% of master's degrees, and 13% of PhDs. Not only do women make up just 7.8% of computer science and computer engineering faculties, only 2.7% of tenured professors are female [Frenkel 1990, page 38] [Gries et al 1991]. Even worse, these numbers seem to be im- proving only very slowly or even dropping [Leveson 1989, page 3]. Many girls are still steered away from math and science or choose not to pursue careers in these fields. Additionally, many women who go into computer science are conscious of being treated differently from men or feel that they face addi- tional barriers. This report describes some of these barriers. As someone who has loved math and computers ever since I was a little girl, I have always found it strange that so few females share these interests. At the computer camp I attended, the boy-girl ratio was six to one. It was similar at high school math club and the math summer program I attended. At MIT, only about 20% of the computer science undergraduates are female, while other departments, such as biology, are at least 50% female. I researched this report to explore why females so consistently stay away from computer science, why people of both sexes seem to expect less from women, and why a woman is considered unfeminine if she is an engineer. Because math and computer programming came easily to me and to many other women who have had the opportunities, women clearly are not inherently unable to do well in them. Instead, girls and women are choosing, consciously or subconsciously, not to go into or stay in computer science. While one cannot rule out the possibility of some innate neurological or psychological differences that would make women less (or more) likely to excel in computer science, I found that the cultural biases against i women's pursuing such careers are so large that, even if inherent differences exist, they would not explain the entire gap. In this paper, I describe the biases that women face in pursuing careers in computer science and how they deal with them. Organization My report examines the following topics, each occupying a chapter: 1. Societal pressure against women's being successful, par- ticularly in engineering. 2. Ways in which the male-dominated environments discour- age women. 3. Inequalities in language, their causes and effects. 4. Negative consequences of some attempts to help women. 5. Conclusions and recommendations. At the end of the report are appendices which contain addi- tional information, such as the methods I used for data collec- tion and implications of recent research on sex-based intellec- tual differences. Finally, there is an annotated bibliography. The report does not directly address the problems of racial minority group members, such as blacks and hispanics, who are also underrepresented in computer science. Readers should not interpret my lack of material on this subject as implying that no bias exists against these groups. On the contrary, I have been told by a black female computer scientist that the color barrier is greater than the sex barrier. In many parts of this report, such as Section 1.2.1 on subconscious bias, the experience of people of color is analogous to that of women. Readers are welcome to contact the author with questions or comments. To do so, send electronic mail to ellens@ai.mit.edu or use the following address: Room 630 545 Technology Square Cambridge, MA 02139 ii Acknowledgments Because many contributors wished to remain anonymous, I can- not thank them by name. I am pleased, however, to have this opportunity to acknowledge Prof. Sherry Turkle, who encour- aged me and supervised the early stages of this project, and my friend Nate Osgood for his support and thoughtful com- ments. Additionally, I am grateful to Dr. Vicki Almstrum, Danielle Bernstein, Joost Bonsen, David Chaiken, Prof. Judy Goldsmith, Prof. Eric Grimson, Dana Henry, Magdalena Leuca, Prof. Barbara Liskov, Prof. Nancy Leveson, Dr. Fanya Mon- talvo, Philip Spertus, Prof. Lynn Stein, Becky Thomas, Chris- tine Tsien, Dr. Kim Wallen, Janet Wixson, Liz Wolf, and Mary Ellen Zurko for their comments on earlier versions of this pa- per. I am also grateful to Dr. Anita Borg for managing an electronic mailing list of female computer scientists and to the women on the list who have helped and encouraged me. Finally, I want to express my appreciation of the many students, staff, and professors at MIT, particularly in Electrical Engineering and Computer Science, who have been wonderfully support- ive of this work, especially EECS department head Prof. Paul Penfield, CS head Prof. Fernando Corbato, MIT ombudsper- son Prof. Mary Rowe, and EECS administrator Marilyn Pierce. Part of this work was done while I was receiving financial sup- port through a NSF graduate fellowship. iii iv Contents 1 Societal Factors 1 1.1 Stereotyping : : : : : : : : : : : : : : : : : : : : : 2 1.1.1 Background : : : : : : : : : : : : : : : : : 2 1.1.2 Bias in Children's Toys and Computer Games : : : : : : : : : : : : : : : : : : : : 3 1.1.3 Stereotypes of Boys and Girls : : : : : : : 6 1.1.4 The Effects of Stereotypes on Teachers and Advisors : : : : : : : : : : : : : : : : 9 1.2 Ways that Males and Females are Treated Dif- ferently : : : : : : : : : : : : : : : : : : : : : : : 11 1.2.1 Subtle Bias : : : : : : : : : : : : : : : : : 11 1.2.2 Different Expectations for Men and Women 15 1.2.3 Different Standards for Men and Women 17 1.2.4 Career-Related Success Unfeminine : : : : 19 1.2.5 Implications of Gender Double Standards 21 1.2.6 Specific Stereotypes Against Female En- gineers : : : : : : : : : : : : : : : : : : : : 22 1.3 Summary : : : : : : : : : : : : : : : : : : : : : : 24 2 The Masculine Environment 27 2.1 Sexist or Sexual Humor : : : : : : : : : : : : : : 27 2.2 Sexual Displays and Discussions : : : : : : : : : 29 2.2.1 Different Reactions : : : : : : : : : : : : : 29 2.2.2 Attempts at Changing Behavior : : : : : 31 2.3 Behavior Due to Sex-Correlated Differences : : : 32 2.3.1 Socializing with Co-workers : : : : : : : : 32 2.3.2 Different Communication Styles : : : : : : 34 2.4 Finding a Mentor : : : : : : : : : : : : : : : : : : 37 2.5 Behavior Specific to Technical Events : : : : : : 38 2.5.1 Trade Shows : : : : : : : : : : : : : : : : 38 2.5.2 Technical Conferences : : : : : : : : : : : 39 2.6 Different Priorities : : : : : : : : : : : : : : : : : 40 v 2.6.1 Family Life : : : : : : : : : : : : : : : : : 40 2.6.2 The Hacker Culture : : : : : : : : : : : : 42 2.7 Sexual Harassment : : : : : : : : : : : : : : : : : 44 2.8 Summary : : : : : : : : : : : : : : : : : : : : : : 45 3 Gender in Language 47 3.1 Referring to Unknown People : : : : : : : : : : : 48 3.2 Masculine Terms as Default : : : : : : : : : : : : 51 3.2.1 The Inequality of Masculine and Femi- nine Terms : : : : : : : : : : : : : : : : : 51 3.2.2 Intentional Use of Masculine Terms : : : : 52 3.2.3 Unintentional Use of Masculine Terms : : 54 3.3 Gender-Neutral English : : : : : : : : : : : : : : 54 3.3.1 Background : : : : : : : : : : : : : : : : : 54 3.3.2 Examples of Usage in Transition : : : : : 56 3.3.3 Reversed-Expectation Writing : : : : : : : 57 3.3.4 Reactions to Nontraditional Language : : 58 3.3.5 Summary : : : : : : : : : : : : : : : : : : 60 3.4 Summary : : : : : : : : : : : : : : : : : : : : : : 60 4 Problems with Solutions 61 4.1 The Perception of Lowered Qualifications : : : : 61 4.1.1 The Need for Affirmative Action : : : : : 62 4.1.2 Distrust of Qualifications : : : : : : : : : 63 4.1.3 Low Self-Confidence : : : : : : : : : : : : 64 4.1.4 Uncritical Faith in Test Scores : : : : : : 65 4.1.5 Conclusion : : : : : : : : : : : : : : : : : 65 4.2 Informal Special Treatment Harmful : : : : : : : 66 4.3 Special Awards for Women : : : : : : : : : : : : 68 4.3.1 Separate Categories for Males and Females 68 4.3.2 Fellowships for Women : : : : : : : : : : : 69 4.3.3 "Heck, We Want More Girls" : : : : : : : 71 4.4 Bad Consequences of Raising Consciousness : : : 73 4.4.1 Female Hypersensitivity : : : : : : : : : : 73 4.4.2 Male Hypersensitivity : : : : : : : : : : : 74 4.5 Summary : : : : : : : : : : : : : : : : : : : : : : 75 5 Recommendations and Conclusions 77 5.1 Recommendations : : : : : : : : : : : : : : : : : 78 5.1.1 Programs and Policies to Encourage Women 78 5.1.2 Ways for Women to Build Self Image : : : 80 5.1.3 Ways for Women to React to Biased Be- havior : : : : : : : : : : : : : : : : : : : : 82 5.1.4 Ways for Individuals to Encourage Women 88 vi 5.1.5 Discussion : : : : : : : : : : : : : : : : : : 91 5.2 Conclusions : : : : : : : : : : : : : : : : : : : : : 91 A About This Paper 93 A.1 Data Collection Methods : : : : : : : : : : : : : 93 A.2 The History of the Document : : : : : : : : : : : 95 A.3 MIT : : : : : : : : : : : : : : : : : : : : : : : : : 96 A.4 How to Obtain Additional Copies : : : : : : : : : 97 B Sex-Based Intellectual Differences 99 C Advantages for Women 103 vii viii Chapter 1 SOCIETAL FACTORS Women beware. You are on the brink of destruc- tion: You have hitherto been engaged in crushing your waists; now you are attempting to cultivate your mind: You have been merely dancing all night in the foul air of the ball-room; now you are be- ginning to spend your mornings in study. You have been incessantly stimulating your emotions with con- certs and operas, with French plays, and French novels; now you are exerting your understanding to learn Greek, and solve propositions in Euclid. Beware!! Science pronounces that the woman who studies is lost --- R. R. Coleman, M.D., 1889.[1] In this chapter, I will describe various ways in which women are steered away from professional success, particularly in tra- ditionally male fields such as engineering.[2] Girls and women are subject to societal pressure to stay away from such subjects or to accept themselves as "unfeminine". Academically successful _____________________ 1. Coleman, R. "Woman's Relations to the Higher Education and Pro- fessions, as Viewed from Physiological and Other Standpoints." Trans- actions, Medical Association of Alabama (1889), page 238. Quoted in [Ehrenreich et al 1978, page 128]. 2. Readers outside of the field may be confused by my use of the term "engineer" to include computer scientists. I do this because the field of computer science does not fit neatly into either engineering or science. Despite the "science" in the name, university computer science departments are often attached to electrical engineering departments or are part of the school of engineering. Additionally, computer programmers and designers tend to think of themselves more as engineers than as scientists, although some individuals and organizations consider computer science as a branch of the mathematical sciences. 1 women are generally seen as being less attractive and less happy than less highly-achieving women.[3] Additionally, there are still people who believe, consciously or subconsciously, that women are incapable of being top scientists and who take women less seriously in general. 1.1 Stereotyping 1.1.1 Background Currently, most people in positions of power and respect are male. Men are rhetorically asked in [Sandler 1986, page 3]: Imagine that your lawyer, your doctor, your priest, rabbi, or minister, your Senator and Representa- tive, your mayor, the president of your institution, most of its trustees, almost all of the deans and most of your colleagues were all women. How would you feel? Not only are these positions held by men, but the media propa- gate stereotypes of women. A recent study found that "women are often still depicted on television as half-clad and half-witted, and needing to be rescued by quick-thinking, fully clothed men" [Adelson 1990]. Whether or not people realize it, many of their expectations of men and women are based on what they have observed and by what messages their culture presents. As one Usenet [4] reader wrote: This group [comp.society] has been discussing various stereotypes for a long time now. The prob- lem is that we haven't acknowledged the impor- tance of stereotypes to human cognition. We could easily call stereotypes heuristics. A heuristic is a device that allows a processor to use some sort of knowledge to reduce a search. In very rough terms, stereotypes allow the wetware in our heads to do less search when looking for evaluations of people and behaviors to events. _____________________ 3. Of course, the stereotype also exists of male engineers being less at- tractive than other males. See, for example, [Wolpert et al 1988, page 2], [Turkle 1984, Chapter 6], and [Holland 1990, pages 164-165]. 4. Usenet, a large electronic bulletin board system, is described in the appendix on methods. 2 The human brain is the best known implemen- tation of these heuristics. In fact the human brain is so reliant upon heuristics, that when presented in- formation that contradicts such a heuristic we feel confusion and discomfort. So where is all this going? For better or worse, humans will continue to make stereotypes based on input from their environment. Since the average person rarely meets programmers, and the only pro- grammers that people see are male geek actors on the TV, we will still have the geek image. I have to admit, most programmers I know are not ex- jocks. Most programmers I know are male, have interests in math and the insides of computers and will allow their personal appearance to slip during those wonderful, coffee-filled, marathon program- ming sessions. See, I have the disease too. (disease = stereotype) [5] I believe that the author of the message is unusual not in hav- ing biases but in being aware of them. It has been empirically shown, as will be described in the following sections, that many people expect less of females without realizing it. These stereo- types are not as harmless as the author implies however, as they sometimes impair people from seeing past the stereotypes. Ad- ditionally, people tend to live up or down to the expectations that are communicated to them. 1.1.2 Bias in Children's Toys and Computer Games The social biases that influence females begin in childhood, where boys and girls are often treated differently on the basis of sexual stereotypes. From the earliest ages, girls are given different types of toys than boys. For example, one study of children from one to six years of age found: Boys had more vehicles, toy animals, military toys, educational-art materials, sports equipment, and spatial-temporal objects. On the other hand, girls had more dolls, doll houses, and domestic objects [Rheingold et al 1975]. _____________________ 5. Katz, Douglas J. Comp.society, May 14, 1990. 3 The difference in toys cannot be explained purely by the chil- dren's preferences --- the expectations of parents and other gift givers play a major role. Numerous studies, cited in [Pomerleau et al 1990, page 360] have found: When interacting with an infant who was intro- duced as a girl, adults used feminine toys (for in- stance, a doll) and talked more to `her'. When the infant was presented as a boy, they used masculine toys (e.g., a hammer) and encouraged more motor activity. These stereotypes are perpetuated by toy companies, which market toys in a stereotyped manner. A 1969 Life ad contained: Because girls dream about being a ballerina, Mattel makes Dancerina ... a pink confection in a silken blouse and ruffled tutu ... Barbie, a young fash- ion model, and her friends do the `in' things girls should do --- talk about new places to visit, new clothes to wear and new friends to meet.... Because boys were born to build and learn, Mattel makes Tog'l [a set of blocks for creative play].... Because boys are curious about things big and small, Mattel makes SuperEyes, a telescope that boys can have in one ingenious set of optically engineered lenses and scopes [Komisar 1972, page 305]. While such an ad would not appear today, it indicates the en- vironment in which today's young scientists were raised. As recently as 1985, a study found: [T]he content of toy catalogues and the pictures of children on the packages of toys are still strongly stereotyped. In catalogues and in stores, special sections are reserved for sex-stereotyped toys. Girls' sections contain dolls and accessories, doll houses, arts and crafts kits, toy beauty sets, and housekeep- ing and cooking toys. Building sets, sports-related toys, transportation toys, workbenches and tools are featured in the boys sections ([Schwartz et al 1985] in [Pomerleau et al 1990, page 365]). In addition to marketing toys in a stereotyped manner, such factors influence the design of toys. It was reported that: 4 Jaron Lanier, head of VPL, gave a talk at UIST in 1989 about his experience productizing the data glove for Nintendo games....[H]e said that [the] toy manufacturer was very strictly divided from the very highest levels into the girl's toy division and the boy's toy division. He strongly resisted the glove from becoming either a girls' toy or a boys' toy, but he lost. He said that they immediately categorized it as a boys' toy and put all kinds of black Darth- Vader-ish, sports-car-like paraphernalia all over it to make it appeal to boys. He also said, had it been categorized as a girl's toy, it probably would have been pink and frilly. Not only are there differences in varieties of "old-fashioned" toys given to children, but these biases are carrying over into the realm of computerized toys and games. These games are both based on traditionally male interests, such as war and sports, and are marketed toward boys. [Kiesler et al 1985, pages 456-457] reports: On one rack [in a computer store], covers in comic- book style depicted such games as Olympic De- cathlon (4 male athletes on cover), Cannonball Blitz (3 men in battle), Swashbuckler (9 male pirates), Thief (1 male detective), Alien Typhoon (1 male space explorer) and Money Munchers (1 man in a suit). In all, 28 men and 4 women were illustrated on the covers. The women were on the covers of Monopoly (2 men and 2 women playing the game), Palace in Thunderland (1 very fat queen), and Wiz- ard and the Princess (1 wizard standing, 1 princess in supplicating position on floor). Girls' lesser usage of computer games could be a factor in their being less positively disposed toward computers and in their lack of interest in computer courses [Lockheed 1985, page 118], particularly as students who have played computer games are more likely to do well in their first college computing course [Kiesler et al 1985, page 457]. It should be noted that nobody with whom I have spoken proposes that a conspiracy exists among manufacturers and advertisers to keep females in their place. Rather, companies aim their products at the largest segment of the population that is predisposed to use them. Additionally, females are more 5 willing to buy products advertised for males than vice versa [Courtney et al 1983]. 1.1.3 Stereotypes of Boys and Girls Anecdotal evidence suggests that when an infant is dressed in blue, passers-by say how smart he looks; if the same baby is dressed in pink, people say how pretty she is. Boys' clothing is often decorated with cars and trains; girls' clothing rarely is. More rigorously, numerous studies of sex stereotyping of infants are reviewed in [Stern et al 1989], including: Parents in one study, for example, were asked to rate and describe their newborns shortly after birth when the primary source of information about the baby was his or her gender (Rubin et al., 1974). Al- though the infants did not differ on any objective measures, girls were rated as littler, softer, finer featured, and more inattentive than boys. Other studies have revealed that parents treat male and female infants differently.... Fagot (1978) observed that parents of toddlers reacted differently to boys' and girls' behaviors. Parents responded more pos- itively to girls than boys when the toddlers played with dolls, and more critically to girls than boys when the toddlers engaged in large motor activity [Stern et al 1989, page 502]. Expecting different behavior from boys and girls can be a self- fulfilling prophecy: If one sort of behavior is expected and en- couraged, the child will be more likely to continue it. Children also have been shown to have formed sexual stereo- types as early as at two years old [Weinraub et al 1983, page 33]. For example, Preschool children also have a good grasp of adult- validated sex-stereotyped beliefs about children's behavior. When asked in an interview-like situa- tion which of two paper dolls --- `Michael' or `Lisa' --- would like to do certain activities in nursery school, end up in certain future roles, and have cer- tain character traits, children 2- to 3-years old showed an impressive depth of knowledge (Kuhn, Nash, & Brucken, 1978). Children believe that girls like to play with dolls, help mother, cook dinner, 6 clean house, talk a lot, never hit, and say `I need some help'; they also believe that boys like to play with cars, help father, build things, and say `I can hit you' [Weinraub et al 1983, page 34]. The careers that children imagine for males and females are influenced by sex stereotypes. By the age of three years, most children "know that girls will grow up to clean the house, be a nurse, or be a teacher, and boys will grow up to `be boss'" [Weinraub et al 1983, page 38]. These stereotypes affect the careers that children picture for themselves: Even preschool children express future aspirations along sex-stereotyped lines. Both preschool and el- ementary school girls choose a parenting role sig- nificantly more often than boys (Looft, 1971; Von- dracek & Kirchner, 1974). In addition, the range of occupational choice is more restricted for girls, with nurse and teacher being the most popular an- swers (Vondracek and Kirchner, 1974; Beuf, 1974). Boys' choices include more action oriented occu- pations (police officer, sports superstar) and more prestigious careers (doctor, public servant, pilot). Taking the question one step further, Beuf (1974) asked children 3 to 6 years of age what they would do if they were of the other sex. Approximately 70 percent of the children replied with a job con- sidered appropriate for the imagined sex. More interestingly, boys frequently imagined themselves as nurses and girls imagined themselves as doctors when asked, `What if you were a girl (boy)?' Sev- eral girls confided that they really would prefer to be doctors rather than nurses when they grew up, but couldn't because they were girls [Weinraub et al 1983, page 44]. Thus, from an early age, girls and boys learn to think of most careers as being appropriate for either men or women but not both. This will influence not just their career choice but how they view males and females aspiring to "inappropriate" roles. Unfortunately, these stereotypes are so pervasive that it is difficult for unprejudiced parents to prevent their children from accepting the stereotypes: oA female computer scientist told me: 7 We ... have a rather non-traditional house- hold, and I'm surprised at how traditional my two daughters seem to be turning out. Both my husband and I work full-time, but when we are home, [John] does almost all the cooking (I make a meal maybe once every three weeks), he cleans up after himself while cook- ing so I don't do much of the cleaning in the kitchen, I do the laundry (sometimes), and we let everything else go until a friend comes to clean our house and dig us out from under the laundry I never can seem to get to.... [Once,] I asked my 5-year-old who did most of the work around the house, me or her daddy. She said "you". Now, this kid is totally guile- less --- she has not learned yet how to say one thing to one person and another to another, so I'm sure she wasn't just telling me this because I was the one who asked the question. So I said, "What kind of work do I do around the house? In the living room, in the kitchen?" She said, "You clean the kitchen." I couldn't believe it! I might have believed her if she said I occasionally picked up in the living room --- but cleaned the kitchen? Her dad's domain? Where did that come from???? o Another parent reported: When our daughter was very young --- about 3 years old --- we audiotaped an interview about what she would be when she grew up. After mentioning a number of possibilities my wife said, `What about a doctor?' Jessica replied, `Yeah, I could be a doctor.' Our son who was 5 at the time interrupted saying, `I think you mean a nurse.` `Yeah, a nurse,' Jessica said. My wife said, `She could be a doctor if she wanted,' and our son replied, `I don't think so...I've never seen any, at least not in Iowa.' Stereotypes also exist specific to the computer world. One paper reports: We have even found that some young children be- lieve computer games and computers are for boys. 8 In one nursery school, Pratto (1982) asked girls and boys aged 3 to 5 to name the toys they played with. Both girls and boys reported that boys played with Atari; it was never mentioned as a game for girls. We returned to that school and asked 42 children whether they thought computers were for girls, and then we asked whether computers were for boys. Most children answered this question. Although the majority thought computers were for both genders, the boys were not as sure of this as were the girls (71% of the girls and 57% of the boys). Of the mi- nority, more children thought computers were for boys only (14% of the boys and 11% of the girls) than thought computers were for girls only (7% of the boys and 4% of the girls) [Kiesler et al 1985, page 456]. The point of this section can be illustrated by the following incident: A group of parents arranged a tour of a hospital for a group of twenty children: ten boys and ten girls. At the end of the tour, hospital officials presented each child with a cap: doctors' caps for the boys, nurses' caps for the girls. The parents, outraged at this sexism, went to see the hospital administration. They were promised that in the future, this would be corrected. The next year, a similar tour was ar- ranged, and at the end, the parents came by to pick up their children. What did they find, but the ex- act same thing --- all the boys had on doctors' hats, all the girls had on nurses' hats! Steaming, they stormed up to the director's office and demanded an explanation. The director gently told them, `But it was totally different this year: We offered them all whichever hat they wanted'" [Hofstadter 1986, page 156]. 1.1.4 The Effects of Stereotypes on Teachers and Advisors Additionally, stereotypes influence people who advise students, such as their parents, guidance counselors, and teachers. For example, [Stewart et al 1989] showed that, when given artifi- cial case studies, high school teachers were more likely to ad- 9 vise male students than otherwise-identical female students to take courses that would prepare them for post-secondary insti- tutions. Another study showed that high school girls "said that they had chosen business and commercial courses in order to prepare themselves for clerical jobs because they believed these were the jobs open to women" [Stewart et al 1989, page 261]. In response to a survey of female scientists, [M]any women felt they had been given inadequate advice on careers and choices of subject --- careers advisers seemed to be fixated on nursing and teach- ing, and some were completely floored by requests for information about nuclear physicists or process engineers [Ferry et al 1982, pages 27-28]. Interviews with high school guidance counselors yielded simi- larly prejudiced advice to girls: A counselor in her early 30's: `Well, if they bring me their registration card with (an AP [advanced placement] science course) listed I'll check to see if that's really what they meant... but I would never encourage it. I mean, it's usually their last year and there are so many fun things going on. I think they'll be busy enough and they can get into the serious work in college.' A counselor in her 20's: `I just hate to see a girl get in over her head. I always try to place students at a level where I know they'll be successful. I mean, wouldn't it be frightful to spoil a beautiful record by doing poorly in a course your senior year.' A male director of guidance, mid-forties: `Sure, I'm for the AP Program in general, but not for encour- aging girls in science necessarily. Have you looked at the Bureau of Labor Statistics? It's a contracting market. There are men with Ph.D.'s in physics all over the place who can't get jobs. Why should we encourage girls? Why, if they're successful, they'd be taking jobs away from men who need them. No, it wouldn't be fair to the girls' [Casserly 1979, page 12]. Unfortunately, as the interviewer goes on to report, "these com- ments were chosen not because they were unique but because 10 they represented all too commonly the attitudes of the coun- selors in many schools". Additionally, even when girls are in science classes, teachers sometimes treat them differently, as shown by the following remarks from an interview of junior high school girls: So this teacher came down from the high school to give a demonstration in physics and said, `Now this is going to make a pretty big noise, so any of you girls who don't like loud noises better cover your ears.' He said, `Now this is going to be dirty so we'd better have a boy do it.' And he (a high school science teacher perform- ing a demonstration to a sixth-grade class) said, `Now this will help you boys who fix your own bi- cycles, so pay attention!' (See also [Marriott 1991] and [Hall 1982].) The girls then go on to describe the difficulties they had in getting their parents to buy them tinker toys and chemistry sets, which are routinely bought for their brothers [Casserly 1979, page 9]. 1.2 Ways that Males and Females are Treated Differently In addition to the people who consciously believe women less capable, there are those who acknowledge that women can suc- ceed at engineering but consider female engineers to be "some- how suspect" [Turkle 1984, page 200]. I will examine several aspects of this problem: First, based on their preconceptions of women, people often exhibit subtle forms of subconscious bias that cause them to treat women differently from men. Second, men and women are often held to different standards. Strange as it sounds, behaviors --- such as succeeding --- are sometimes considered attractive in men but not in women. Third, there is something about our culture's view of male-dominated fields such as engineering that causes female aspirants to be consid- ered unattractive. 1.2.1 Subtle Bias In [Sandler 1986, Sandler 1988, Hall 1982], there are summaries of several studies of subtle, subconscious bias --- that is, people 11 observably acting in a biased manner but unaware of their doing so. I was apprised of the importance of subtle bias by the number of respondents who objected to my call for "egregious examples", writing that they thought the subtle behavior to be more damaging. [Hall 1982, Sandler 1988] report the following biases, of which both men and women are guilty: o Women are interrupted more than men. o Faculty members make eye contact with male students more often than with female students. o Faculty members are more likely to know and use the names of their male students than of female students. [6] o Women are often asked fewer or easier questions than males. As Sandler writes, "Singly, these behaviors probably have little effect. But when they occur again and again, they give a pow- erful message to women: they are not as worthwhile as men nor are they expected to participate fully in class, in college, or in life at large" [Sandler 1988, page 149]. Unfortunately, the mes- sage appears to have sunk in. Studies have shown that, when engineering students are asked to predict the academic per- formance relative to that of male and female colleagues, "both sexes anticipated that men would outperform women. This was paradoxical, since the average female student had both a higher grade point average and higher class rank from high school than the average male" ([Ott 1975] in [Zappert et al 1984, page 4]). Another study found that, when male and female college stu- dents were asked to predict their midterm test score before tak- ing it, men had higher expectations for themselves than women did for themselves, even though the two groups actually per- formed the same [Erkut 1983, page 229]. Studies have found that women are more likely than men to attribute success to luck instead of skill [Deaux et al 1974] and to attribute failure to lack of skill [Ernest 1976, page 599]. Women's lack of con- fidence, and one consequence, is illustrated by an incident at Columbia, reported by Professor Joan Birman: I learned last year, to my astonishment, that for about four years running the honors calculus course _____________________ 6. The experience of women I have talked with is that if females are in an extreme minority, they stand out so much that the teachers are likely to know their names. 12 had been all male, in spite of the fact that admis- sion was based on an open competitive examination. This fall, one of the senior mathematics majors and myself made an intensive effort to encourage women to try the exam! The typical answer was, `I know I won't pass it,' --- to which we replied over and over, `Well , if you try it, at worst you will confirm what you already know, and only an hour of time will have been lost.' After three days of such ad- vising, the big day came, the exam was given, and this year the class has five men and five women! [Ernest 1976, page 604]. Not surprisingly, girls at single-sex schools study physical science and math more than in comparable coed schools, "even though girls' schools frequently have less adequate laboratory provision than mixed schools" [Kelly 1982, page 497] Even more ominously, [Sandler 1986, page 6] reports: In one study, first done in 1968 and then replicated in 1983, college students were asked to rate identi- cal articles according to specific criteria. The au- thors' names attached to the articles were clearly male or female, but were reversed for each group of raters: what one group thought had been writ- ten by a male, the second group thought had been written by a female, and vice versa. Articles sup- posedly written by women were consistently ranked lower than when the very same articles were thought to have been written by a male [Goldberg 1968, Paludi et al 1985, Paludi et al 1983]. In a similar study, department chairs were asked to make hypo- thetical hiring decisions and to assign faculty rank on the basis of vita. For vitae with male names, chairs recommended the rank of associate professor; however, the identical vita with a female name mer- ited only the rank of assistant professor [Fidell 1975]. Anti-female bias is strongest in traditionally male fields [Top 1991, pages 96-97]. When discussing the results of such studies with fellow stu- dents, I found that the males have tended to be more surprised than the females, because many females recalled specific in- stances of biased behavior, several categories of which are rep- resented below. 13 In some cases, a woman was viewed as less serious than a man in a similar position: o A female computer science graduate student had the fol- lowing experience: I was working at a fairly small company whose communal coffee was awful. A group of 6 of us (4 men, 2 women) bought our own coffee maker and had decent coffee which we paid a few cents for to defray the costs of coffee and cream. Anyway, I usually bought the coffee and my male coworker usually bought the cream. A new member to the group (male) approached me and told me we were out of cream. I told him that M2 usually bought the cream. Later that same day, M1 again comes up and tells me we are out of cream. I once again tell him that M2 gets the cream. To this he says, `But how can I bother M2 with something as insignificant as buying cream?' Needless to say, I told this fellow exactly what I thought about that... o One female graduate student in mechanical engineering sent the following two stories: 1. When I first started the [graduate] program, the head of the department (male) was assigning desks to graduate students... ([T]here were two entering females to the particular program at that time.) As he ran out of desks, he said, `Well, just put the girls together on a desk.' 2. That same professor put me (a newly graduated math major) into the slower (undergraduate) statis- tics class, and put a guy who had had a vague intro- duction to stats 8 years earlier in the faster (gradu- ate) class. Both events took place in the past two years, a time pe- riod which she has found "very frustrating". o A female computer consultant wrote: Most of our users ask for and accept help from whichever consultant is available, but some in- sist on talking with one of the male consul- tants (only 2 of the 10 consultants are male). 14 One user persisted at this, even after Alan ex- plained to her that he didn't know the pack- age she was working with, and that she would be better off asking me because I specialize in that particular software. Another user did ask me her question, but when she didn't like the answer I gave her (I explained what she could not do and why), she insisted on taking her problem to David (a higher authority?), who proceeded to tell her exactly what I had just explained. What is interesting about the latter incident was that she did not seem to be after a second opinion, because she could have gotten that from a number of people (all female). She apparently specifically wanted a male opinion. oWhen a female computer science undergraduate visited one of the graduate schools to which she had been ad- mitted, she and a male prospective student met with a male graduate student to discuss the school. Whenever the woman asked a question, the graduate student di- rected his answer to the male prospective instead of to her, i.e. by making eye contact and gestures toward the male prospective. This treatment surprised the woman, as she had not encountered such behavior at her under- graduate institution. After the meeting, she delicately pointed out the behavior to the graduate student, and he apologized profusely and sincerely, clearly unaware of the bias while it was occurring. When they met later in the day, his behavior was markedly better. The same woman, however, in a later meeting with two other graduate stu- dents, one male and one female, found herself addressing most of her questions to the male until she recognized her behavior and corrected it. 1.2.2 Different Expectations for Men and Women The following examples show how people sometimes expect women to be less interested or competent in technical areas than they actually are: oAccording to a survey of female scientists: Women in mainly male environments are al- ways being taken for secretaries or junior lab- 15 oratory staff: queries may be addressed to a male technician rather than his female boss. An engineer offering to help a telephone caller was told `No dear, this is a technical enquiry. Can I speak to someone who can help me?' [Ferry et al 1982, page 28] o A female computer science professor told me this story: I was visiting a university and arrived before my (male) host. I approached the departmen- tal receptionist to try to make certain arrange- ments. In one case, I suggested that my host might have made some provision --- `or,' I said cordially, `maybe not.' `Oh, probably not,' replied the receptionist. `After all you know those pro- fessors...' Boy was her face red when she real- ized what she'd said. [7] o A female undergraduate at a women's college wrote: The summer after my first year at [X] I took Linear Algebra at [a coed college] nearby. Out of probably twenty people in the class, I was one of two women. I found that the mood of the class was stifling. It was obvious that the men of the class expected me to sit quietly in my chair and contribute nothing and ask no ques- tions. It was also made obvious to me that, in general, they felt they were far superior to me. Because I had had no contact with them out- side of the classroom, I must assume they were basing their decision solely on the fact that I am female. In addition, I found the material rela- tively easy and was getting an A in the class, _____________________ 7. It is fascinating to read about how female professionals and secretaries interact with each other. One professor reports that she used to sneak to the typewriter and type her own letters, rather than ask a secretary to do the work for her. Correspondingly, secretaries generally do not like to work for women. "They experience women's authority as `unnatural', whereas men's authority is taken for granted" [Pringle 1989, page 58]. Additionally, as hinted by a comment in [MIT 1983, page 21], female secretaries feel demeaned when female professionals complain about being mistaken for secretaries. On the other hand, most of the secretaries I have worked with have treated me the same as they treated my male colleagues, and one has even told me outright how happy she was to see women as computer professionals. 16 so they could not be basing it on my academic performance. One day as we were going over a difficult problem set we had had for homework, the professor asked if anyone was able to do a particular problem which I had been able to solve. When I raised my hand, [a student made] the comment `What?!?! How could you have solved that problem!?!?' He in no way hid his hostility or his feelings that if he, a far superior man, could not solve the problem, I could not have. I was completely shocked that he could make such a comment. No one else seemed to be. It is no wonder that women tend not to contribute in a male-dominated classroom. oA female computer scientist sent me a copy of the cover of a prestigious computer periodical that showed a family (parents and a boy) looking at a computer. A bubble next to each shows what they are thinking. The mother is imagining her son using the computer to learn math and the father using it to figure taxes. The son and the father both imagine using the computer to play space war games. [8] These diverse examples illustrate how women are sometimes treated as less capable or interested in technology than men, instead of being treated as individuals. Of course, there ex- ist professors and administrators who treat their male and fe- male students equally as well or even devote extra effort to encouraging women. However, negative events are still com- mon enough to be of substantial concern. Moreover, the above behaviors are the symptom of a more fundamental problem: lower expectations for females. Many of the above events are too blatant to have the insidious effect of subtle discrimination (which probably accompany them). Even if the perpetrators could be coerced into not so openly displaying sexism, it would not eliminate the fundamental biases which would be displayed less directly. 1.2.3 Different Standards for Men and Women As Sandler writes, the same behavior is viewed differently in women than in men: _____________________ 8. Cover, IEEE Computer, March 1977. 17 He is `assertive'; she is `aggressive' or `hostile'. He `lost his cool', implying it was an aberration; she's `emotional' or `menopausal'. Thus, her behavior is devalued, even when it is the same as his [Sandler 1988, page 151]. This claim can be illustrated by a recent lawsuit by a woman who "repeatedly heard that she had not been given a partner- ship at [a] huge accounting firm because she was too macho, universally disliked and in need of `a course at charm school'" [Lewin 1990]. This is despite having brought in more business than any of the other 88 candidates for partnership. "Com- ments from the lawsuit [say] that she should wear makeup and jewelry and learn to walk, talk and dress `more femininely'" [Lewin 1990]. A survey of female scientists found: Most think their male colleagues are more forceful and aggressive than they themselves want to be; some have resigned themselves to low status rather than changing their personalities, while others have decided to fight with men's weapons --- and are of- ten labeled unfeminine as a result [Ferry et al 1982, page 30]. One study found the same behavior judged more harshly in female professors than in males: [According to] Susan Kay's classroom studies... male students were far more likely to give lower ratings to those female faculty perceived to be hard graders... This finding is consistent with a series of experi- ments at the University of Dayton that indicated that college students of both sexes judged female authority figures who engaged in punitive behav- ior more harshly than they judged punitive males... ([Martin 1984, pages 484-485] in [Koblitz 1990]). See also [Kierstead et al 1988] and [Bennett 1982]. A "halo" effect seems to exist where people tend to interpret behavior according to their preconceptions. The same action is often interpreted differently, depending on whether it is per- formed by a woman or a man, as the following stories illustrate: o When a high-ranking female engineer was at the airport to make a business trip, she saw a male acquaintance who worked for the same company. He asked where she 18 was going, and she answered San Francisco. He then said something like, `Oh, going to do some shopping?' She told me how angry his remark made her, as she works extremely hard at the company, putting in long hours and taking frequent business trips, with too little free time for her to go on out-of-state shopping trips even if she were inclined to do so. oA female computer science graduate student told me that it is common to see different reactions to men and women dropping a class. According to her, when a woman drops a class, people remark that the class must have been too difficult for her; when a man quits, people say he must not have found it interesting. These examples are troubling because they show one way in which stereotypes are perpetuated. In each case, someone in- terpreted the actions of a woman based on their prejudices, reinforcing their own stereotypes. Additionally, women at American universities are often the victims of other cultures' stereotypes. Foreign nationals out- number Americans as students in doctoral engineering pro- grams [Widnall 1988, page 1740], and there are many foreign- born professors. In one survey, female graduate students at MIT "reported that foreign-educated faculty --- many from cul- tures where women are not held in high esteem --- pose prob- lems for women in graduate programs, both in class and in research" ([MIT 1987 ] in [Baum 1990 , page 49]). 1.2.4 Career-Related Success Unfeminine Not only are some strong traits considered unfeminine, but "femininity and individual achievement continue to be viewed as two desirable but mutually exclusive ends," a shocking po- sition argued in [Horner 1970, page 46], based on empirical re- search and interviews. In one of Horner's studies, females were given the sentence "After first-term finals, Anne finds herself at the top of her medical school class." Males were given a similar sentence with a male name. Subjects were asked to write a story about the student. While only 8 of the 88 male subjects exhibited fear of success through negative stories, 59 of the 90 females did. Horner divides the negative stories into three categories and includes sample stories, of which I include a subset: 19 1. Fear of Rejection --- "Anne doesn't want to be number one in her class. She feels she shouldn't rank so high because of social reasons. She drops down to ninth in the class and then marries the boy who graduates number one." 2. Concern about Normalcy or Femininity --- "Anne is com- pletely ecstatic but at the same time feels guilty. She wishes that she could stop studying so hard, but parental and personal pressures drive her. She will finally have a nervous breakdown and marry a successful young doc- tor." 3. Denial --- "Anne is really happy she's on top, though Tom is higher than she --- though that's as it should be.... Anne doesn't mind Tom winning" [Horner 1970, pages 60-62]. Additionally, when questioned about Anne, "[m]ore than 70% ... described Anne as having an unattractive face, figure, or manners" [Horner 1970, page 63]. Females thus consider suc- cess to lessen their femininity, a sacrifice many are not willing to make [Horner 1970, pages 69-72]. See also [Mednick et al 1975]. This attitude can also be illustrated by the following inci- dent, reported in [Franklin et al 1981, page 20]: One woman earned high grades in a traditionally male field. Her professor announced to a mostly male class that this represented an unusual achieve- ment `for a woman' and was an indication, first, that the woman student was probably not really feminine, and, second, that the males in the class were not truly masculine, since they allowed a woman to beat them. Instead, the proper area for a woman's success is seen as her ability to attract high-status men. In a study for the Na- tional Institute of Education, researchers Holland and Eisen- hart found: Men's prestige and correlated attractiveness come from the attention they receive from women and from success at sports, in school politics, and in other arenas. Women's prestige and correlated at- tractiveness come only from the attention they re- ceive from men [Holland 1990, page 104, emphasis in original]. 20 This attitude is exemplified by the way one college woman at- tempted to insult another: "[You] may be able to do calcu- lus, but I'm dating a football player" [Holland 1990, page 104, brackets in original]. The study found that female friends often did not even know each other's majors [Holland 1990, page 14], although they spent large amounts of time together discussing other matters, primarily boyfriends [Holland 1990, page 14]. 1.2.5 Implications of Gender Double Standards The double standards discussed above should be a significant concern. Aggression, competitiveness, and even some brashness are necessary for a graduate student, for example, who must compete with other students for equipment, funding, and at- tention from professors. One doesn't get far by politely waiting to be noticed or for other people to stop using the computer. In her paper on being a female graduate student at MIT, Candace Sidner addresses some stumbling blocks women face: Receiving an advanced degree, in fact, any de- gree, from MIT is rather like being admitted to a fraternity. One has a certain set of rituals to go through, and both the process and one's perfor- mance define one's position in the fraternity in the years that follow.... It surprises no woman to say that women are so- cialized differently than men in our cultures. What is surprising is the effect of that socialization when women take roles traditionally held only by men. The most significant role change centers [on] devel- oping confidence and competence. Part of the pro- cess of hurdle jumping is not just the getting over, it is the form which one presents in doing it. For the MIT fraternity ritual, the form is confidence; a woman student must use what I call strutting be- havior, that is, she must look and act like she knows what she is doing. While developing confidence from accompany- ing competence, is difficult for all initiates, for women there is a subtle, but remarkable difference; women in the everyday world are not supposed to appear very confident and competent.... As a result, women must not only build and show confidence and com- petence, just as their male counterparts do, but un- 21 like the men, they must decide first to unlearn their normal behavior patterns.... The strutting behavior appears slowly; there are stops and starts, forward and backward progress. A woman student begins to act from a little bit of con- fidence in her competence, and tests out this con- fidence among her peers and superiors. Two more difficulties follow. First, a woman feels less femi- nine, because in fact she is less feminine according to the prevailing behavior patterns. In her personal life, her feelings may be communicated to her part- ner(s) who may find her less attractive. This threat- ens her personal status. Eventually a woman can learn to find personal friends who value her confi- dent image, but the time in between is frightening [Sidner 1980, pages 2-3]. Empirically, a comparative study of male and female Stan- ford graduate students in technical areas [Zappert et al 1984] found that women were less self-confident and assertive than their male peers: [W]omen less frequently than men reported that they felt free to disagree with their advisors...and that their ideas were respected by their advisors [Zappert et al 1984, page 9].... [W]omen more often reported having trouble say- ing "no" and in giving criticism. Women also more often reported having difficulty sticking up for them- selves and tended to let annoyances pile up [Zappert et al 1984, page 12]. 1.2.6 Specific Stereotypes Against Female En- gineers As if the culture-wide inhibitions to success were not enough, there are additional barriers in engineering. Nowadays, high school girls from middle- and upper-middle-class families are expected to go to college and to do reasonably well, but going to a technical institute or majoring in a technical field is still considered unfeminine, as these anecdotes indicate: o When a female student at an engineering institute went home for vacation, her mother leafed through the book 22 of photographs of the freshman class and exclaimed in surprise, "Why, some of these girls are pretty!" oA male computer professional wrote: Back in 1983, I was a freshman here at [X] and one of my friends was a genius who hap- pened to be a pretty blonde girl.... She was also a freshman and spent one of her first days here searching for her advisor's office. While hunting around [Y] Hall, a man in his early 30's came up to her and asked if she needed help. She said that she was looking for her advisor's office. The man responded with a puzzled, `What major are you?' When she answered, `I'm in Electrical Engineering.' The man smiled at her and oozed, `Oh, you're far too pretty to be an EECS major.' [She] immediately left and told us in the dorms about this slimy guy. The next day we went to our first lecture for [the introductory computer class]. [She] gasped as one of the lecturers entered the hall. He was the same slimy guy she had encountered the day before.... I'll never forget the quote, "Oh, you're far too pretty to be an EECS major." [9] oA male computer professional wrote: I used to teach undergraduate computer science classes, and saw a number of cases in which very promising and talented women abandoned computer science, much to my disappointment since they were some of my better students.... At least where I was teaching, the discourage- ment from that field was given more by other women, particularly in the sororities, rather than from within the field itself. _____________________ 9. In a later note, the writer added: "[The teacher] was fired two years after this incident. According to my advisor at that time, his attitude toward female students was one of the reasons. (He was not tenure track. He was a lecturer only.)" This story was later confirmed by a former professor from the university. 23 It is worth repeating, however, that the stereotype of male engineers is almost as bad. Jokes and television portray male engineers as unattractive, unpopular, awkward, and either un- successful with or uninterested in women.[10] However, I believe that in our culture, females are more susceptible to such stereo- types. This is in large part because, as described earlier in this chapter, femininity is considered to be at odds with success, while masculinity is not. While the stereotype that female engineers are inherently unattractive seems to be without rational basis, scientific fields may well be in conflict with some values traditionally thought of as feminine and currently held by a majority of females. The situation seems not to have changed since the following was written: I think [women's lack of achievement] comes from the general orientation of girls, young women, and even older women, toward `others' (in David Reis- man's sense of being `other directed'). Women are constantly urged to consider `Am I doing the right thing?' and `What shall I be or do that will please my husband, children, and parents?' Occupational success never comes out as the positive answer to these questions. Pleasing others and doing the `right thing' always means holding back, and retreating from a position of strong ambition and career com- mitment [Epstein 1974, page 15]. I would add that being other-directed might steer women away from objective sciences into the humanities and the more people- oriented social sciences. Thus, the values that are encouraged in women would not only make them less career-oriented but more likely to avoid the sciences. 1.3 Summary In our society, males and females are regarded very differently. Assertiveness, confidence, and high achievement are considered consistent with masculinity but not with femininity. In addi- tion to the stigma associated with success in general, techni- cal fields are considered particularly unattractive for females. These factors can influence a girl not to pursue an interest in _____________________ 10. See [Turkle 1984] for an interesting discussion of the male engineering student's self-image. 24 math or engineering, and they can sabotage a woman's career because either she acts feminine, e.g. demure, and is not taken seriously, or she acts masculine and is met with disapproba- tion. Of course, as mentioned in [Sidner 1980, page 3] and [Horner 1970, page 70], confident females eventually find male and female friends who like and respect them. The problem is thus not insurmountable. Still, it is an additional barrier that females face, and the playing field will not be level while these stereotypes exist. 25 26 Chapter 2 THE MASCULINE ENVIRONMENT I do not see that the sex of the candidate is an ar- gument against her admission... After all, we are a university, not a bathing establishment --- David Hilbert, arguing unsuccessfully for the appointment of Emmy Noether to the faculty at G"ottingen. [1] Currently, the majority of professionals in computer science departments and workplaces are male. As a consequence, these places often have masculine environments in which women feel uncomfortable or unwelcome. The behaviors described in this chapter are generally not meant to be harmful to women, which makes the participants often hostile toward criticism. Such behaviors include the display of nude pictures, discussing sex, telling dirty jokes, and expressing negative stereotypes of women in an attempt at humor. Additionally, other activities are morally faultless, such as coworkers playing basketball together, but they may tend to make a woman feel less part of the group if she does not enjoy the same activities. 2.1 Sexist or Sexual Humor Often, men make sexist or sexual remarks in attempts at hu- mor. As the following examples show, this happens in class- rooms, computer magazines, and at conferences. In most of _____________________ 1. Weyl, Hermann. "Emmy Noether", Scripta Mathematica, VIII, 3 (July, 1935), page 207. 27 these cases, the speaker probably means no harm. However, the behavior makes some women feel uncomfortable. In order to highlight the effect such statements have on women, their reactions have been included where available. o A female graduate student had the following experience: [A professor] in the introductory part of a guest lecture on robotics to the graduate core AI class: (approximate quote) `Pretty soon we'll have robots that are sophisticated enough to wander around in shopping malls and pick up girls.' I didn't listen to the rest of the lecture, so I don't know what else he had to say. o A female computer science professor wrote: When I was in graduate school, the professor in automata theory introduced the topic of de- composition by saying: `Machines are a lot like women --- many forms for the same function (wink wink).' As the only woman in the class, you can imagine that I felt terrific. And all of a sudden the guys sitting next to me sort of tensed up --- instead of a fellow student, his re- mark had made them see me as something else, something kinda dirty. o The narrator of the industry gossip column in the trade journal Infoworld is an adult male with a young girlfriend Pammy, shallow and uninterested in computers, whose silly statements and actions pad the columns, such as her return to beauty school. ("[A]t 21 she's older and can handle the pressure now.") [2] o A female computer science professor wrote: [A]t a conference in France, a male speaker (French), who was speaking about the impor- tance of testing, showed an overhead slide of a naked woman with a caption of the sort --- `Would you buy this product without testing it first?' There were only 2 or 3 women in _____________________ 2. Cringely, Robert X. "Notes from the Field". Infoworld, September 17, 1990, page 108. 28 the audience (of about 150), but I had fleet- ing feelings of having accidentally walked into a stag party and wondering if he had either not expected any women to be there or had discounted the importance of directing his re- marks to the women in the audience. What these examples all share is that the male speaker or writer was attempting to make a cute statement but that females (and some males) had negative reactions. 2.2 Sexual Displays and Discussions 2.2.1 Different Reactions At the workplace, many women feel uncomfortable with the "locker room atmosphere", which includes pictures of nude or partially nude women on posters or computer screens and the telling of dirty jokes. Unlike the sexist remarks described above, however, there is disagreement among women on how inappro- priate these actions are, with a significant number of women not personally offended by the behavior (although some of these women oppose it on the grounds that it upsets other women). This point is illustrated by the different reactions in the follow- ing examples: oWhen a graduating engineer was touring a company that wanted to hire her, they took her through the laboratory, which had a pin-up on the wall. The other people in the lab (men) and the men showing her around seemed oblivious to the poster and to her discomfort. She felt uncomfortable with the idea of working in a laboratory with a picture like that up and ended up refusing the offer, partly for this reason. oWhen a graduating computer scientist was taken out to lunch by engineers of a small computer company, one of the topics the employees (all male) discussed was a series of lingerie shows in the region. The student did not feel uncomfortable about the subject matter, and it did not affect her decision about the company, but she thought it was something which might make other female intervie- wees uncomfortable. 29 o When a computer science undergraduate had recently be- gun working in a research group at her university, some male graduate students entered the office and began play- ing an "adult" computer game, "Leisure Suit Larry," crowd- ing around the screen, discussing the game loudly. The undergraduate left the office because the situation made her too uncomfortable for her to work there. When she recounted the story a year later, she said that she would not react the same way now and would either be able to keep working or would say, "Hey, get out of here, guys." The change was due to her feeling more confident about her position in the group and knowing the individuals personally. One female computer science student explained one reason that some women are offended by sexual humor while others do not understand what the fuss is about: I have noticed that how offended I am by [gender- related humor] depends very strongly about how comfortable it is to be female and in the present environment. When I first entered grad school in the CS [X] group at [Y], there were some women graduate stu- dents, but only a couple. A secretary deliberately placed me, when I arrived, sharing a desk with a male graduate student who was at that time des- perately trying to find a woman (she was trying to be nice) --- a professor had a `funny' newspa- per article about [a sexual topic] posted outside his door. Don't get me wrong, I found nearly no-one among the faculty and graduate students who was anti-women or took me or my work or my concerns any less seriously than any other first-year student. Still, the graduate students were 90 percent men, and they talked all the time about how hard it was to meet `available' women, and as a first year stu- dent trying to establish myself within their commu- nity, I found the `locker room' atmosphere oppres- sive and daunting. If someone had sent around [a sexist joke through email] that year, I think I would have hit the roof. In a world where I was struggling to find my place, it would have just helped to un- dermine it. 30 Today the graduate student population in CS [X] has quite a few more women, and is much more comfortable. Instead of the "guys" in school here, it's the `people' in school here.... In my current environment I might have easily [passed along the joke] to my [male] office-mate. 2.2.2 Attempts at Changing Behavior Some computer science graduate students and staff at Carnegie Mellon were sufficiently disturbed by the display of nude pic- tures as backgrounds on computer terminals that they got to- gether and tried to change the situation by publicly appealing to the community. [CMU 1989 ] is a fascinating report describ- ing their appeal and the friendly and hostile reactions. Their appeal included the following passage: When a woman sees such a display on your work- station, is she likely to believe that you take her seriously as a fully contributing member of the de- partment? Rather, she may feel that you could be a source of sexual harassment, and feel hostile to- wards you, or nervous about working with you. If so, that is a loss for you, for her, and for all of us. Among the visitors to the department, some of whom are prospective students, staff, or faculty, there are surely some who will view us as unpro- fessional if they see these displays, and this hurts us all, too. Conversely, an environment more hos- pitable to women --- specifically, one in which rela- tions between women and men are less strained --- is of clear benefit to men as well. For some people, displays of naked women on work- stations, or elsewhere in offices, remind them of the forces in our culture that view women as sexual playthings, not as men's peers. For others, such reactions do not occur. People who are offended will interpret such displays as derogatory, even if that is not your intent. We therefore ask you to re- frain from using them out of respect for those who are offended, even if you believe the offended people are just overly sensitive [CMU 1989 , page 2]. The appeal closed by making clear that they were not advo- cating banning such displays but were requesting that people 31 voluntarily remove them out of sensitivity to others. Responses about the appropriateness of the displays and of the appeal were mixed and are categorized in the report. Negative reac- tions included the position that the writers were advocating censorship "like the Nazis or the Ayatollah Khomeini," that people should not be asked to change their behavior merely because of what others might think, and that a public appeal was inappropriate but instead should have been made by in- dividuals to individuals. Of those agreeing, the majority of responses said that the request was reasonable and not an at- tempt at censorship, that it prevented people from uninten- tionally giving offense, and that it was effective at raising con- sciousness. In response to the criticism that individuals should complain personally, several women wrote that "[w]omen ask- ing for changes in behavior individually are exposed to ridicule and abuse" [CMU 1989 , page 4]. This point was echoed by a woman quoted in a paper about the "Garden", a laboratory in the MIT Media Lab: [W]hen comments are made about the offensive na- ture of the music or movies, they are often ignored, or belittled, or are chortled at. Ironically, once you are labeled a feminist in the Garden, your comments are taken less seriously, because you are considered radical and your judgment less fair [Tidwell 1990, page 14]. Both the Carnegie-Mellon and Garden papers conclude that the attempts at changing people's behavior were somewhat, but not highly, successful. 2.3 Behavior Due to Sex-Correlated Differences An additional category exists of behavior that is not directly based on sex but which nevertheless discourages women. While attempts at changing sexist behavior are partially effective, there seems little that can be done about this category. 2.3.1 Socializing with Co-workers Through no real fault on any side, a woman sometimes feels out of place being one of the few women in a semi-social gath- ering with a group of men, even in the absence of any behavior 32 directly related to the sex of the participants. One reason is that, in our culture, men are often interested in activities or topics that women tend not to relate to. A female graduate student complained about her experi- ences as a teaching assistant (TA) for a course in a particularly male-dominated area of computer science. She wrote: Perhaps because the percentage of males is often high, men tend to dominate non-academic discus- sions with topics of interest to them, such as sports and cars, topics which women are often uninformed about or uninterested in. The resulting inability to participate in discussions can make it difficult for women to bond socially, and often leads women to feel outright alienated... This is exactly what hap- pened to me at each T.A.-faculty meeting. Another female computer science graduate student described a similar experience: My first summer at [a certain computer company], I worked in a group that was otherwise all male. While I got along okay with them and never had any behavior to complain about, I didn't socialize with the group. For example, every day after lunch, they would go outside to `shoot some hoops' [play basketball], an activity that I just did not relate to. For my next summer, I joined a group that had other female programmers and a female man- ager. I was much happier in that group. We would have barbecues, celebrate people's birthdays, and socialize in other ways that I related to better than `shooting hoops'. My third summer, I chose to re- turn to this group and not to the first one or to find a new group. However, when she casually discussed her social dissatisfac- tion in an exit interview with the department head, he pointed out that, coming from a different country, he did not relate to American sports either. This raises the important point, which holds for all examples in this section, that dissatisfac- tion with certain activities is not strictly divided by sex. There are individual women who enjoy sports and are better at them than some men. Sex-based differences are a tendency, not a fixed rule. Sex-correlated preferences in our society, however, 33 are strong enough that these phenomena tend to work against women (or whoever is underrepresented in a group). Additionally, female group members do not always feel com- fortable joining male group members who go out drinking to- gether. Not only might they not enjoy drinking, but some men are inclined to making lewd remarks after a few drinks. Thus, there are often times when women feel unable to take part in activities to which, as group members, they are invited. Another problem is that some men do not feel comfortable socializing in a professional manner with a woman, as this anec- dote illustrates: I was ... the first full-time woman faculty member in my department. There really was difficulty among my male colleagues in associating with a woman as a colleague. I think they literally did not know how to talk to me, and as a consequence often just did not talk to me. They would ignore me. They would not invite me to have lunch with them, which was a very ordinary experience there ... they would walk past my office and ask the next person and never ask me. [Years later] I asked one of my col- leagues why this was so. And he said, `You know what would happen if I asked you to lunch ... Peo- ple would talk' ([Clark et al 1986, pages 36-37] in [Sandler 1986, pages 7-8]). 2.3.2 Different Communication Styles The language that women use often differs from that of men in subtle ways. As discussed in [Hall 1982, page 9], specific constructions appear more frequently in women's speech than in men's: o hesitation and false starts (`I think...I was wondering...') o high pitch o `tag' questions (`This is really important, don't you think') o a questioning intonation in making a statement (`The sec- ond chapter does most to clarify the theme?') o excessive use of qualifiers (`Don't you think that maybe sometimes...') 34 oother speech forms that are excessively polite and defer- ential (`This is probably not important, but...') As Hall concludes: If, for example, a woman student begins her com- ments hesitantly and uses many qualifiers, she may be immediately perceived by her teacher and by her classmates as unfocused and unsure of what she wants to say. Her `overly polite' style may seem to `invite' interruptions by, or inattention from, both teacher and other students. Indeed, even the most insightful points made in this manner --- especially by a woman --- may be taken less seriously than the identical points made by a man or delivered in a more `masculine' assertive style [Hall 1982, pages 9-10]. See also [Lakoff 1975]. A large part of Jenifer Tidwell's report on the Garden [Tidwell 1990] describes how men and women react differently to the same treatment, due to their expectations of communication styles: The persons I interviewed did not believe that women were treated worse than the men were, nor vice versa, when I asked them directly. (`Everyone is treated equally badly,' said one male informant.) Yet some informants, both male and female, com- mented that women may not be able to deal with the Garden's harshness as well as the men do --- not because of any inherent weakness, but simply be- cause they have not been brought up with the same expectations of `toughness' (in one man's terms) that men have. Furthermore, it seems more ac- ceptable for anyone there --- male or female --- to try to solve all one's problems by oneself than to habitually ask for help. All of the women that I interviewed commented on this expectation of in- dependence (whereas almost none of the men did); they did not like it [Tidwell 1990, page 9]. Tidwell writes that the attitude is that "[what could be con- strued as] harassment may just be social incompetence" [Tidwell 1990, page 15, brackets in original]. Anecdotal evidence suggests that women tend to be more sensitive than men to general obnoxiousness [Widnall 1988, page 35 1744]. When a female computer science student told me that a male TA had been inexcusably rude to her in front of the class, I looked up the student evaluations of the TA and found this comment, by another student: [He] knows his stuff cold, but he's too rude. Hon- estly, once you humble yourself and tolerate this, he is an excellent source of help. I owe a lot to him. A student unused or unwilling to being treated rudely would not be able to interact with such individuals. When I sent electronic mail to a group of women asking for criticism on the first version of this report, one woman replied that my request was unlikely to draw many responses. Instead of asking for "criticism", she told me, I should have asked for "feedback to help me improve the report", something women would feel more comfortable supplying. Steven Levy's Hackers: Heroes of the Computer Revolution extensively describes the hacker culture, including the hackers' apathy and even antipathy toward women. Maybe it would have been different if there had been more women around TMRC [Tech Model Railroad Club] and the ninth floor --- the few that did hang around paired off with hackers.... There were not too many of these women, since outsiders, male or female, were often put off by the group: the hackers talked strangely, they had bizarre hours, they ate weird food, and they spent all their time thinking about computers [Levy 1984, page 72]. Levy goes on to describe the poor hygiene of one of the most admired hackers, a young man who did not bathe [Levy 1984, page 73]. Male computer environments that exclude women have oc- curred as early as in preschool: Even in preschool, males dominate the school com- puters. In one preschool, the boys literally took over the computer, creating a computer club and refus- ing to let the girls either join the computer club or have access to the computer. When the teachers intervened and set up a time schedule for sharing computer access, the girls spent as much time on the computer as the boys [Kiesler et al 1985, page 454]. 36 While it is not clear that one can justify the stereotype of the engineer and computer hacker as socially backwards, women are deterred by the environment.[3] (See [Markoff 1989], for example.) Additionally, whether the stereotype is accurate is to some extent irrelevant: If females believe that to study or work with computers requires hanging around socially incom- petent nerds, the stereotype, true or false, may influence their decision. 2.4 Finding a Mentor Having a mentor or sponsor can be of vital importance to a graduate student or junior professor: The sponsor may serve many functions for the protege. First, the sponsor introduces and initiates the protege in the customs, demands, and expectations of aca- demic life. Second, the sponsor shares his or her wisdom and knowledge with the protege, and pro- vides encouragement and comments on his or her work. Third, the sponsor can provide career assis- tance for the protege by making recommendations to his or her colleagues at other institutions, or sim- ply by sharing a bit of the deflected glow from his or her own shining reputation. Perhaps most im- portant, the sponsor helps to form with the protege the sense of him or herself as a member of the pro- fession, encouraging and fostering a self-image as a legitimate member of the community of scholars [Simeone 1987, page 101]. Despite the importance of having a mentor, there are few for- mal policies to ensure that every graduate student or junior faculty member receives mentoring. Although every graduate student, for example, has a thesis supervisor, the supervisor typically devotes different amounts of energy to different stu- dents. It is reported that "women are more likely than men to be excluded from this sort of relationship with senior fac- ulty" [Simeone 1987, page 102] [Hall 1982, page 10]. There are several possible reasons for this exclusion. First, as discussed above, some men feel uncomfortable dealing with women as _____________________ 3. It is worth noting, however, that some men are disgusted by the hacker culture, such as MIT professor Joseph Weizenbaum [Levy 1984]. 37 professionals. Second, some "faculty men may see women as be- ing different from themselves, less intellectually able, less com- mitted and dedicated, or simply inappropriate for academic ca- reers" [Simeone 1987, page 103]. Third, when men and women work closely together, there is the risk of their being suspected of having an affair [Simeone 1987, pages 82-83]. Additionally, many people like to help people who are "like" them, i.e. of the same sex or race. (Indeed, that was a motivation for my writ- ing this report.) As long as most of the people in positions of power are men, and as long as differences in sex are considered to be of great importance, junior men will benefit. 2.5 Behavior Specific to Technical Events Computer trade shows and technical conferences are often even more male-dominated than the workplace and university. Some of the specific problems of these events are discussed in this section. 2.5.1 Trade Shows Men far outnumber women at industry and academic confer- ences. At trade shows, particularly, companies have "histori- cally employed attractive women to draw attention to product exhibits or booths" [LaPlante 1989]. While this practice has been declining, an industry journal article about Comdex 1989 reports: A number of companies still insisted on hiring scant- ily clad female models to attract attention to their booths. And a party thrown by Fujitsu went far be- yond questionable taste, seriously offending Comdex attendees of both sexes [LaPlante 1989]. Additionally, when women do attend, they are suspected of not being legitimate attendees: [PFS Inc. president Mary] Rich remembers attend- ing the National Computer Conference (NCC), a now-defunct computer convention, where the male attendees outnumbered the female ones by a ratio of 300 to 1. Convention officials as well as hotel staffs were extremely suspicious of single women, Rich said. Women were often suspected of being 38 prostitutes trying to solicit show attendees. Rich said she once tried to go to the hotel room of a col- league for a drink only to be kicked out by security when trying to get in an elevator. Rich, who co-chaired the 1986 NCC with another woman, said that as recently as three years ago one of the primary concerns was how women were being treated at the show. `We still had prob- lems with security not believing [the credentials of] women trying to get onto the show floor,' she said [LaPlante 1989]. 2.5.2 Technical Conferences Women are treated better at technical conferences than at trade shows. While I have never seen any reports of a female speaker or attendee being mistaken for a prostitute, people sometimes assume a woman to be the wife of an attendee and not an attendee herself. Additionally, female computer scientists com- plain of being propositioned by male attendees. While this probably happens at trade shows too, its happening at confer- ences leads to particularly touchy situations because the women do not want to offend potentially important men in their field and feel obliged to find delicate ways to reject indelicate offers. Because of this, many women act cold to men they meet at conferences, which has the side effect of discouraging friend- ship with colleagues who could be useful contacts. A female computer scientist, with experience in several fields, described another problem with being one of the few attendees who needed to bring along their child: One difference between biology and computer sci- ence that I have noticed is that it is not unusual for childcare to be provided at professional confer- ences in biology, while I have never seen it offered at AAAI [American Association for Artificial In- telligence], IJCAI [International Joint Conference on Artificial Intelligence], Cog Sci, or ACM [Asso- ciation for Computing Machinery] conferences. I have a baby that I won't be able to leave overnight for about 2 years (because of breastfeeding), but it is difficult to make daytime childcare arrangements for an out of town conference oneself. Even with an older child, it can be difficult to make the night- 39 time arrangements at home if the child is left behind (traveling spouse, single parent, etc.), so bringing the child along might be preferable if childcare were available. [4] Additionally, another female computer scientist told me that "some conferences organize `spouse events' which tend to run to fashion shows, shopping, etc. My husband finds this strangely unappealing!" Presumably, conference hosts do not mean to discriminate when they fail to provide childcare or provide activities of inter- est to wives of attendees --- I assume that, if the issue crosses their minds, they assume that the number of exceptional at- tendees would be too small to justify the expense of providing for them. This is an example of how, entirely in the absence of any bad intentions and purely due to the ratio, conditions can be such that a class of attendees cannot conveniently take part in a professional event. 2.6 Different Priorities 2.6.1 Family Life Having a support spouse, usually a wife, is a boon for any- one but especially for pre-tenure professors, for whom it is not unheard of to work hundred-hour weeks. Since women rarely have a spouse willing to tend house for them, while male pro- fessors and graduate students sometimes do, the women are often at a relative disadvantage. Additionally, wives are of- ten more willing to relocate for their husbands than vice versa [Ferry et al 1982, page 29]. As one computer scientist said: There was an article in Chronicles of Higher Edu- cation about 3 years ago by a male professor who wrote about how he and his wife (also a professor) needed to have a third party --- a wife... The gist of his argument is that faculty workload is based on antiquated notions of unquestioning, full-time sup- port from a spouse, and that universities need to re- vise their expectations of professors [Frenkel 1990, page 41]. _____________________ 4. I have been told that childcare was arranged for this year's AAAI. 40 Women in academia have the additional problem that the years in which they must work to get tenure are a significant por- tion of their child-bearing years. In [Frenkel 1990, page 41], one woman gives this as a reason for choosing industry over academia. Additionally, some women "prefer to take a less de- manding job than their qualifications fit them for, because they feel that the time and attention they can afford on top of their responsibilities at home is limited" [Ferry et al 1982, page 30]. This might be one reason that female computer science PhDs are less likely than their male peers to enter academia, more often choosing industry instead. Not only are some women unwilling to sacrifice family for work, a choice that men rarely have to make, but when women do decide to put their career first, it is still assumed that they do not take their career seriously. According to a survey of (English) female engineers and scientists: The most trying moments in almost every woman's life seem to have been spent in interviews. The women in our survey have nearly all taken the trou- ble to equip themselves with a qualification that might suggest they had ambitions beyond boiling nappies [diapers] and making their husbands' tea. Yet time and again they have found themselves be- ing pulled apart on the subject of whether or not they are likely to leave soon in order to marry or have children [Ferry et al 1982, page 28]. Similarly, a study of the hiring of scientists and technical staff at the National Health Service found that employers often assume that all women will leave to have babies and that wastage due to pregnancy is greater than for any other reason. The pervasiveness of [this myth] was shown by the way in which they influenced practices at selection (for instance, only women were asked questions about marital status and dependent chil- dren). They also influenced notions of who can be a manager [Homans 1987 , page 90]. So even if a woman chooses not to have children or not to take time out to raise them, employers will often assume oth- erwise and treat her accordingly. Furthermore, pregnancy does not always cause women to miss much work. An army study found that "even when pregnancy leave is included, [enlisted 41 women] take less time off than men, who lose it to sports and auto injuries and drug, alcohol and discipline problems" [McNeil 1991]. Power can be another factor in why women choose not to be professors. The following was written by a woman who had been a professor but had switched to an industrial research position: A year ago I would probably have agreed with the popular conclusion that academia is difficult for women because of the time demands, coincidence of tenure with child-rearing, etc. After a little more than a year in industry, I've discovered another reason that academia can be difficult for women. I now believe that to be highly successful in academic research, one has to be very interested (invested?) in hav- ing power. Power over grants, over students, over committees, etc. I often heard professors referred to as empire-builders, something that I see very lit- tle of where I am now. Often the most successful researchers in my current environment are the ones who actively avoid politics and power-struggles and just `do their work'. As a woman I don't think that I am especially comfortable or adept at the power- games that I witnessed in the university. Of course, there are also men that dislike power and women who revel in it. Nonetheless, the tendency of women to be less comfortable with power than men may hold in our current society. 2.6.2 The Hacker Culture Some hacker subcultures have the property that the hackers spend nearly all of their waking hours, and miss sleep, to use the computers. In his discussion of the absence of female hack- ers, Levy writes: "There were women programmers and some of them were good, but none seemed to take hacking as a holy calling the way Greenblatt, Gosper, and the others did" [Levy 1984, page 72]. Another hacker classic, Tracy Kidder's Soul of a New Ma- chine describes the intensity of the designers of a new computer: Going to work for the Eclipse Group could be a rough way to start out in your profession.... [Y]ou 42 don't have any time to meet women, to help your wife buy furniture for your apartment, or to explore the unfamiliar countryside. You work.... You're working at a place that looks like something psy- chologists build for testing the fortitude of small animals, and your boss won't even say hello to you. New and old hands told the same story. Chuck Holland: `I can hardly say I do anything else now. It takes about three days to get Eagle out of my mind, so if you have a three-day weekend, you're just sorry to see Monday come.' Microkid Betty Shanahan, the group's lone female engineer: `You can end up staying all night. You can forget to go home and eat dinner. My husband complained that the last three times he's had to do the laundry.' Jon Blau: `I've had difficulty forming sentences lately. In the middle of a story my mind'll go blank. Pieces of your life get dribbled away. I'm growing up, having all those experiences, and I don't want to shut them out for the sake of Data General or this big project' [Kidder 1982, pages 60-61]. I have been told that, after this book came out, MIT students lined up to interview with Data General, so the described work environment does appeal to some computer science students. In our society, however, women are often less willing or able to devote all of their life to a job, particularly because working full force is often difficult without a support spouse to take care of other parts of one's life. When reading how intensely and single-mindedly the hack- ers work with computers, it is hard not to question the people who love the computer to the exclusion of all else, an opin- ion expressed by the writer of the following letter, written in response to [Markoff 1989], but equally applicable to the envi- ronment described in Hackers and The Soul of a New Machine: To the Editor: You regretfully wonder why women have not done as well at computers as men. You define the issue inside out. The problem is not women's expe- rience with computers but men's. If a `passionate romance' with the machine is the key to excelling, we should pity the men who do rather than the women who don't. 43 I am not a feminist crank at either end of that interesting spectrum, but no girl or woman I know is so alienated from her fellows that she `spurns the real world to master a universe locked inside a com- puter.' Machines do seem better suited for use than for passionate romances. So why regret women's atti- tude toward them? Why not worry about yet an- other generation of men who are sealing themselves off from human contact [Harrigan 1989]? It is important to remember that women who do not throw themselves into the computer world might not be inferior to men but that sacrificing everything to computers might not be something that a psychologically healthy human being does. Perhaps men and women alike would be better off if some jobs and hacker cultures did not require giving up the rest of their lives. 2.7 Sexual Harassment Because I take it for granted that readers consider sexual ha- rassment to be offensive and harmful, I am writing little on the subject. It remains, however a serious problem. A recent survey of Harvard students and faculty revealed: Thirty-two percent of the tenured female professors, 49 percent of those without tenure, 41 percent of the female graduate students, and 34 percent of the undergraduate women reported having been sexu- ally harassed by a person in authority at least once during their time at Harvard. Fifteen percent of the graduate students and twelve percent of the undergraduates reporting harassment consequently changed their academic plans because of it. Most did not report the incidents [Simeone 1987, page 115-116]. Other studies, such as [Baker 1990] find even higher percent- ages of women sexually harassed. See also [Gross 1991], an ac- count of the hostile environment for women at Stanford Medical School, which was recently called to attention by the resigna- tion of a female neurosurgeon. 44 2.8 Summary Because computer workplaces are often overwhelmingly male, women find themselves in what sometimes feels like a locker room environment, having to put up with behavior they might find offensive, such as sexist or sexual humor and female pin- ups. Additionally, some men use sarcasm or insults to commu- nicate more than women do, causing women to interpret the environment as hostile, even when no offense is meant. Also, because males often have different interests than females do, such as in sports, women may not feel as though they fit in. These factors can cause a woman to feel out of place in a com- puter workplace and make it difficult for her to picture herself as a computer professional. 45 46 Chapter 3 GENDER IN LANGUAGE If a woman is swept off a ship into the water, the cry is `Man overboard!' If she is killed by a hit-and- run driver, the charge is `manslaughter.' If she is injured on the job, the coverage is `workmen's com- pensation.' But if she arrives at a threshold marked `Men Only,' she knows the admonition is not in- tended to bar animals or plants or inanimate ob- jects. It is meant for her --- Alma Graham [1] The Sapir-Whorf hypothesis of linguistics states that the limits of human thought are determined by the nature and the structure of the language in which thought occurs. One corol- lary, on which this chapter is based, is that the English lan- guage's use of gender forces people to think in terms of male and female, with its gender-specific third-person singular pronouns and its different titles, in some cases, for males and females.[2] While it is not necessarily bad to be immediately aware of the sex of someone being discussed, the connotation of male and female terms differ so greatly that the distinction not only im- plies difference but inequality. Biases in language are important because they show both the biases people hold and how they are communicated. _____________________ 1. Graham, Alma. "How to Make Trouble: The Making of a Nonsexist Dictionary." Ms., December 1973, p. 16. 2. Of course, many languages are even worse than English [Hofstadter 1986, chapter 7]. 47 3.1 Referring to Unknown People The English language has two sets of pronouns for the third- person singular: he/him/his and she/her/hers. It is practically impossible to talk grammatically about individuals without im- plicitly mentioning their sex. While this can seem benign, it has several significant implications, which will be examined in this chapter. When A mentions a person unknown to B ("my biology teacher"), B must find out whether the third person is male or female in order to know how to phrase questions ("How is he/she?"). If A does not provide the information, B will usually guess the default pronoun from what is known --- e.g. "he" for a professor, and "she" for a secretary. Most often, this is done subconsciously, showing the speaker's preconceptions. It seems reasonable to expect that when children hear their parents using their best guess for people discussed, they will subconsciously conclude that a man's being a scientist, for ex- ample, is normal, while a woman's being a scientist is unusual. It is common for female scientists to be incorrectly ad- dressed. They tend to consider these incidents humorous, and they trade horror stories: o One [female] computer scientist with an ambiguous first name wrote: [M]y favorite case in this regard occurred about two years ago, when I received a letter addressed to Mr. [Name], saying `Dear Mr. [Name]: I attended your presentation at the [A] Conference on [B]. Please send me copies of your related technical reports. Sincerely, Dr. X' Usually, my only response to such incorrect usage is to sign my return letter as Dr. [Name], but that one was too much for me to let slide unremarked. So I sent a reply as follows: Dear Dr. X: While I am accustomed to re- ceiving letters addressed to Mr. [Name] from colleagues who have never met nor seen me, I found your letter quite puzzling. You began by saying you attended my presentation at the [A] Conference on [B]. Surely you must have noticed that I am female! If this was a sec- retarial error, you might alert your secretary that female computer scientists exist and that 48 in cases of doubt, "Dr." is a genderless form that is unlikely to offend (and may in fact be correct). Sincerely, Dr. [Name]. I neither expected nor received a reply, but I felt better, and maybe I saved some other fe- male computer scientist from one of those let- ters. oIronically, when a science magazine ran a survey of its female readers, the female journalists who received the responses found that about half the women had addressed the replies, "Dear Sir" [Ferry et al 1982, page 27]. oA female computer science professor told me: When I was applying for jobs, two of the com- puter science departments requested letters of recommendation for Mr. [Jane Linda Smith]. One almost immediately sent a followup let- ter, explaining that they were terribly sorry, that they indeed knew that [Jane Linda Smith] was female, and that they would certainly read the recommendation letters more carefully than they'd proofread their own request. The other department was the one in which I was then a graduate student. oWhen I received an award from the MIT EECS depart- ment for the first version of this report, the award letter was addressed to "Mr. Ellen Spertus". These examples show that some people expect scientists and science writers to be men to such a great degree that they blatantly misaddress female scientists. Perhaps the most unusual story in this category is that of a woman who was elected to the National Academy of Sciences in the late seventies: The scroll that [geneticist] Vivian [Davidson] re- ceived from the Academy had her name engraved on it, and then went on to announce that Vivian Davidson was being honored for `his' accomplish- ments, and that `he' was now entitled, and `he' could, and `he' should. She was so amazed at all the `he's that she sent a letter to the Academy inquir- ing whether the source of the problem might be that 49 the engraver was British and had taken Vivian for a man's name, or was it perhaps that the printing process was lagging behind the process of election of women to the Academy. The letter she received back from the Academy secretary (a man) was an angry one informing her that she was the first per- son ever to complain, the scroll was an honor, its plate had been struck in 1868 by Abraham Lincoln, and it had a historic value the Academy was not about to tamper with. At the next Academy meeting in Washington, Vivian raised the matter of the scroll's wording with some of the other women scientists. Each one said she had never noticed the use of `he' instead of `she' on the scroll. `That's probably true,' Vivian said sadly. `They're so grateful to be allowed into the club, they wouldn't dream of making waves. In all likelihood they haven't noticed' [Gornick 1990, page 115]. Also of interest are the experiences of a male professor whose first name is also a woman's name: I think having that name has made me more sympathetic to problems women face, as in e-mail it is almost always assumed that I am a women. That means I get a rash of mash notes from the Math/CS students when they first learn e-mail and pick my name out of the user list.... It also means that I get a great response every time I post a com- puter problem to a usenet group. Since I haven't done the control condition where I have a male's name, I'll never know for sure, but the tone of so many responses are so solicitous that I have to won- der (even followups to see if everything worked out okay). The worst case was in a scientific exchange through the mail where the other scientist was being exceedlingly patronizing. It was only when he came to [X] for a convention and tacked on his room and phone number and invited me for a drink that it dawned on me that he thought he was correspond- ing with a woman. For me I can laugh about these occurrences (as well as the female roommates I rou- tinely get assigned at meetings), but it would cer- tainly be different if it never went away. 50 3.2 Masculine Terms as Default 3.2.1 The Inequality of Masculine and Femi- nine Terms Because male and female pronouns have different connotations, an individual is immediately categorized into a set and assumed to have certain characteristics as soon as their sex is known. The English language forces us to divide people by sex, and, be- cause people have different assumptions depending on whether someone is male or female, preconceptions are applied to them. Were it not for these differences, asking someone's sex would have no greater import than asking how to properly pronounce a person's name, and suggesting that females need female role models would make no more sense than arguing that green-eyed people need green-eyed role models. While there is no reason that specifying the sex of an individual is necessarily bad, it is in a culture where people associate so many characteristics with sex. When masculine and feminine versions of the same word exist, the connotations often greatly differ: oA teacher writes of trying to describe one of her brightest students: I found myself saying `She's really a prince.' Appalled as I was at my own pro-masculine description, I just couldn't say that she was a princess because princess connotes someone who is fussy and spoiled and accustomed to liv- ing in the lap of luxury [Miller et al 1980, page 58]. oA female computer science student told me: When I was at [X], I derived a good deal of satisfaction from watching my male friends de- scribe my performance in [electrical engineering courses] --- `She's a .....goddess?' Additionally, "woman" and "man" are not symmetrical: oAs observed in [Hofstadter 1986, page 155]: If I write, `In the nineteenth century, the kings of nonsense were Edward Lear and Lewis Car- roll', people will with no trouble get the mes- sage that those two men were the best of all 51 nonsense writers at that time. But now con- sider what happens if I write, `The queen of twentieth-century nonsense is Gertrude Stein'. The implication is unequivocal: Gertrude Stein is, among female writers of nonsense, the best. It leaves completely open her ranking relative to males. She might be way down the list! Now isn't this preposterous? Why is our language so asymmetric? This is hardly chivalry --- it is ut- ter condescension. o One female computer science professor told me: I got a recommendation from a college professor stating that I was `one of the top female stu- dents' in his class ( 100 students, 10 women, the women were not all at the top of the class, and I was).... I'm sure he meant it as a compli- ment. o Annie Edson Taylor is described as "the first person to go over Niagara Falls in a barrel", while Neil Armstrong is "the first man to walk on the moon". Because masculine and feminine terms have such different con- notations, the distinctions make females separate and unequal. 3.2.2 Intentional Use of Masculine Terms Some men consider "being a man" to be high praise even to a woman, and expect her to be flattered at being called one. Here are some examples of women's reactions to being called a man, two from biographies (not of computer scientists) and one in response to my call for data: o Henry Hazlitt said to Ayn [Rand] one day: `I just talked with Lu Mises a few days ago. He called you "the most courageous man in America."' `Did he say man?' asked Ayn. `Yes,' he replied. Ayn was delighted. [3] o [A] story --- which as far as I know, is all it was, --- once went the rounds of Israel to the effect that Ben- Gurion described me [Golda Meir] as the `only man' in his cabinet. What amused me about it was that obviously _____________________ 3. Branden, Barbara. The Passion of Ayn Rand. New York: Doubleday & Company, Inc., 1986, page 189. 52 he (or whoever invented the story) thought that this was the greatest possible compliment that could be paid to a woman. I very much doubt that any man would have been flattered if I had said about him that he was the only woman in the government! [4] o[This incident] was ... entertaining. I and another un- dergraduate were head teaching fellows for a computer course and we were working our butts off. He had one final that he was pretty worried about, so I greeted him afterwards with a six pack. His remark, `[Jane], you're such a guy.' --- A female graduate student. In all cases, the original speaker thought he was giving the woman a high compliment. What stands out from these quotes is the different reactions: glee in one case, indignation and amusement in the others. Accepting such compliments is psy- chologically dangerous because it entails a woman's looking down upon her own sex. For example, Ayn Rand, the only female above to be proud of being "a man", is known for her misogyny, an unhealthy trait for a woman. The same phe- nomenon occurs among some women in male-dominated ca- reers. One female computer science graduate student used the term "male-identified" to describe women who scorn other women and who aspire to be "one of the guys". Some examples of such statements are: oA male computer professional wrote: One way [bias is expressed] is the attitude of women themselves. A female programmer here found out that the company next door, run by women, installs PC systems in offices. Her com- ment was, `That's pretty good for a couple of women.' Her words, not mine. oA female computer science student, when one of two fe- male students in a high school advanced placement physics class, said: "I judge how hard a class is by how few fe- males are in it." While the above comments are shocking, they should be no less shocking than a man's expecting a woman to be flattered by being called a man. In the context of a society which praises _____________________ 4. Meir, Golda. My Life, New York: Dell Publishing Co., 1976, page 108. 53 women for being "like a man," it is no wonder some women accept the role.[5] [Persing 1978] contains many examples of masculine terms in everyday use. 3.2.3 Unintentional Use of Masculine Terms Other terms exist besides "being a man" and "one of the guys". A male high school teacher who recognized the sexism in "di- viding the men from the boys," told the class that he intention- ally replaced the terms with "wimps" and "studs," a usage he did not consider to have gender implications. Indeed, "stud" is now a popular way to compliment someone from work well done. What's ironic is that "stud" is an extremely male word. Its primary definition is a male animal used for breeding. It is absurd that some consider it a gender-free way to express admiration.[6] When a female professor expressed skepticism that anyone could think "stud" a neutral term, it was called to her attention that a few months earlier, she had told a roomful of female students that they would have to "gird their loins" and get to work. This also derives from male terms. 3.3 Gender-Neutral English One area currently under debate is gender-neutral English. Above, I discussed the issue of how to discuss a specific un- known person, but it is another question how to discuss a generic person, i.e. with "he" or "he or she", etc. Related to this is whether to use terms like "chairman" and "man-hours" when one does not wish to exclude females. 3.3.1 Background The most convincing argument that terms like "he" and "man" are not truly neutral comes not from abstract arguments but from empirical research: _____________________ 5. I have been able to find nothing in the literature about male identifi- cation of women in science, although the phenomenon is widespread. Most women seem to outgrow it and feel ashamed of their former misogyny, but the subject merits deeper study. 6. There is another whole issue as to what the corresponding female term would be and why none are flattering --- i.e. one based on sexual prowess. Brood sow? Whore? Nymphomaniac? This is too far off the subject to explore here. 54 In 1972, two sociologists at Drake University, Joseph Schneider and Sally Hacker, decided to test the hypothesis that man is generally understood to embrace woman. Some three hundred college students were asked to select from magazines and newspapers a variety of pictures that would appro- priately illustrate the different chapters of a sociol- ogy textbook being prepared for publication. Half the students were assigned chapter headings like "Social Man", "Industrial Man", and "Political Man". The other half was given different but correspond- ing headings like "Society", "Industrial Life", and "Political Behavior". Analysis of the pictures se- lected revealed that in the minds of students of both sexes use of the word man evoked, to a sta- tistically significant degree, images of males only --- filtering out recognition of women's participation in these major areas of life --- whereas the corre- sponding headings without man evoked images of both males and females. In some instances the dif- ferences reached magnitudes of 30 to 40 per cent. The authors concluded, `This is rather convincing evidence that when you use the word man generi- cally, people do tend to think male, and tend not to think female ([Miller et al 1980, pages 19-20]). Additionally, "a number of studies have shown that young peo- ple are influenced in their job preferences and their willing- ness to apply for advertised jobs by gender bias in the wording of the advertisements" ([Bem et al 1973] in [Frank et al 1983, page 90]). Several sentences can be found that demonstrate that "man" is often unintentionally used to exclude women: David Moser once .... observed that in books you will find many sentences in this vein: `Man has tra- ditionally been a hunter, and he has kept his females close to the hearth, where they could tend his chil- dren.'.... So much for the sexual neutrality of the generic `man'. I began to look for such anomalies, and soon ran across the following gem in a book on sexuality: `It is unknown in what way Man used to make love, when he was a primitive savage millions of years ago' [Hofstadter 1986, page 145]. 55 Consider also the 10thcommandment.[7] 3.3.2 Examples of Usage in Transition In an approximately one month period of observation, I was able to find many examples of people waffling on the issues of gender nonspecific language at and around the MIT Artificial Intelligence (AI) Lab. I consider ambivalence more revealing than conforming to the old ways, because it shows that many individuals are trying to grapple with the issues but are unable to do so in a consistent manner. These examples are presented not because the behavior was egregious but to show the conflict within individuals: o A conference was held in one of the most male-dom- inated areas of computer science. Nevertheless, a few women played prominent technical roles. Relevant inci- dents were: 1. A male attendee asked a female speaker: "How many [pause] engineer-years did this take?" His pause was not sarcastic; he evidently decided mid-sentence to use a neutral term instead of "man-years". One guesses he would have used the latter term if ask- ing the question of a male. 2. A female was introduced as a "chairperson" but, like the other (male) chairpeople, wore a ribbon that said "chairman". o During a talk, a professor showed a slide which said "he/she" for the generic computer architect, but said "he" when speaking. o An announcement was sent to members of the AI Lab, containing "I am looking for a few brave men (or women) willing to help..." o In a recent issue of a journal, a book reviewer referred to the prototypical researcher as "he", parenthetically adding "typically he is a he".[8] _____________________ 7. It is not entirely fair to put all the blame on the word "man". It is also easy to find sentences such as "People won't give up power. They'll give up anything else first --- money, home, wife, children --- but not power" [Miller et al 1980, pages 33-34]. 8. Mackworth, Alan K. "Book Reviews and Response". Artificial Intel- ligence, 38 (1989) 239-251. 56 As mentioned above, the trait all these examples share is am- bivalence. Many people use neither the old way nor the new way but some mixture. On such melanges, Douglas Hofstadter writes: This is not progress, in my opinion. In fact, in some ways, it is retrograde motion, and damages the cause of nonsexist language. The problem is that these people are simultaneously showing that they recognize that "he" is not truly generic and yet continuing to use it as if it were. They are thereby, at one and the same time, increasing other people's recognition of the sham of considering "he" as a generic, and yet reinforcing the old convention of using it anyway. It's a bad bind [Hofstadter 1986, page 150]. 3.3.3 Reversed-Expectation Writing Several writers, in order to argue for non-sexist writing, have written essays with other biases than the traditional male/female ones, and the results are (intentionally) shocking. In this sec- tion, I describe three such forays. oDouglas Hofstadter has written an essay ostensibly argu- ing for traditional usages but from an imaginary stand- point with different terms for whites and blacks analogous to those for men and women in our culture. For instance, "white" is used for "whites and blacks" (as "men" is used for "men and women"), and blacks have different hon- orifics and pronouns. Here is an excerpt of his (long) essay: Most of the clamor, as you certainly know by now, revolves around the age-old usage of the noun "white" and words built from it, such as chairwhite, mailwhite,... The negrists claim that using the word "white", either on its own or as a component, to talk about all the mem- bers of the human species is somehow degrad- ing to blacks and reinforces racism. Therefore the libbers propose that we substitute "person" everywhere where "white" now occurs. Sensi- tive speakers of our secretary tongue of course find this preposterous. There is great beauty 57 to a phrase such as "All whites are created equal." Our forebosses who framed the Decla- ration of Independence well understood the po- etry of our language. Think how ugly it would be to say "All persons are created equal", or "All whites and blacks are created equal".... [Hofstadter 1986, page 159] o Bobbye Sorrels Persing, in [Persing 1978], has written a powerful essay of an office scene with the male and fe- male roles reversed. Not only are male workers called "boy" and "sir chairwoman" (corresponding to "girl" and "madame chairman"), they are treated and talked about as men stereotypically treat women who work for them [Persing 1978, pages 1-5]. o A recent example exists in the computer world. In MacTech Quarterly (now MacTech Journal), "she" is used instead of "he" as the generic pronoun. An editorial justified the policy and announced that it would be used henceforth by the magazine [MTQ 1989A ]. Further examples of reversing expectations appear in Section 5.1.3. 3.3.4 Reactions to Nontraditional Language It is telling to look at the reaction to MacTech Quarterly's pol- icy, printed in the issue following the editorial announcement [MTQ 1989B ]. A follow-up article wrote that responses poured in, "impassioned on both sides". Some readers canceled their subscriptions, while others pledged to buy as much as possible from the parent organization. One of the most interesting pos- itive letters was from a female novice programmer who wrote: [A]s a woman who is a bit intimidated by her love of math and computers, I deeply appreciate being able to open the MacTech Quarterly and have the articles addressed to me, personally, a woman. Negative letters condemned the (male) editor for abusing his editorial position. One particularly angry writer repeatedly called into question the editor's manhood: Ms. Hines, 58 You seem to be suffering from severe gen