Unitary Models and Sex Difference Research
The most common conception of the psychology of gender is that women and men as groups have, different traits: different temperam&nls,,rehar. ac. t£rs_. outlooks and opinions, abilities, even whole structures of personality, t here is no accepted term for this concept; I will call it (rexuaTcHaracter’J The analogy is with David Riesman’s term ‘social character’; ‘sexual’ is more apt than ‘gender’ since in most usages the idea is specifically linked with sex.
Often it is assumed that there is just one set of traits that characterizes men in general and thus defines masculinity. Likewise there is one set of traits for women, which defines femininity. This unitary model of sexual character is a familiar part of sexual ideology. It can be quite explicit: /ru^Til^^womTn^^just like_a rnajx’. More often it is implicit. lokes againsT’Tvomen drivers’, or the ‘Mere Male’ column in the women’s rnagazii^7^MT7^a7work by calling into play shared assumptions of this kind: that women are hopeless with cars, that men are hopeless around the house.
More sophisticated, but logically similar, ideas appear in academic writing. In Talcott Parson’s classic work ‘instrumental’ versus ‘expressive’ traits are supposed to mark the hwo sexual characters that correspond to the male and female roles. A unitary model of sexual character underlies Nancy Chodorow’s feminist reworking of the same themes. Here the focus is on how women’s sexual character prepares them for mothering and men’s does not. Notions of unitary sexual character have also emerged in cultural feminism in the last ten years. The antipornography campaigns, for instance, have often presented a lust for domination as the core of male sexuality, and as being more or less undifferentiated among men.
Freud’s writings implied rather different conceptions of femininity and masculinity, but most of Freud’s followers returned to convention. Prominent in the psychoanalytic. shift towards conservatism was the Austrian/American (Theodor ReikyHis long essay ‘The Emotional Differences of the Sexes’ is a classic statement of the unitary model. It takes for granted that women and men have sharply different emotional characteristics, and postulates that these are based on their different-junctions in biological reproduction, Reik draws from this, mainly by speculation, complacent explanations of an extraordinary range of matters, from the double standard to cooking, cattiness, premenstrual tension, and why women are concerned about furniture.
In more recent psychoanalytic literature such conceptions of gender remain active. The American Robert May, for instance, makes the difference between ‘the male and the female fantasy patterns’ the central theme of his book Sex and Fantasy. He pursues a general distinction between ‘pride’ and ‘caring’ — compare Parson’s ‘instrumental’ and ‘expressive’ — through ancient myths and modern personality tests and clinical case histories.
It is clear that unitary conceptions of sexual character have a wide appeal and can give comfort to people of very different political persuasions. This is partly because having a unitary conception of feminine or masculine character does not in itself settle what the content of the two opposed characters might be. Speculation, assertion and inference from biology are the order of the day.
Yet there is a large research literature closely related to this problem: the psychological ‘sex difference’ research mentioned in chapter 2, which as Viola Klein and Rosalind Rosenberg show has been flowing since the turn of the century.
The facts at issue in this research are what might be called the block differences between women and men — for instance, differences bet^££a-the flVPrage reaction times nr tartflf ^ndtiwity of women and men; or between average, scores on tests o£ ур^Ья! ability, агщс-ty, or extraversion. There are well-established conventions of method. Comparable samples of women and men are needed, together with a reliable measure of some sort. A test is always made for the statistical significance of whatever difference turns up between average scores, or between the proportions of women and men who meet a given criterion.
‘Significance testing’ itself presses the research towards a focus
on block differences, since what is tested is not its size or psychological importance, but simply the probability that there exists some difference which is not the result of chance. The kind of cbTrehmun that passes from the journal articles into the textbooks and popular-psychology best sellers is that [women have higher verbal ability’ or (men are more aggressive’^ If a statistically significant block difference does not emerge, tHeresearcher is likely to be disappointed and the research may not get published, since it seems to have nothing to say. I have written this kind of data — dredging paper, about adolescents in Sydney, and, as usual in this genre, paid a lot of attention to items where sex differences did appear and much less attention to the larger number of items where they did not.
When block differences do appear they are conventionally explained by appeal to some underlying traits which distinguish women from men — in other words by a unitary conception of sexual character. Often sex role notions are brought in to provide a common-sense explanation of how sexual character is formed. Thus women are said to be less achievement-oriented than men because they have Ізееп socialized to dependency, and so on. Utten sex differencg^areliardly ~distinguished from sex roles at all; the two blur together irv-i—ainnrlr
A vast compendium of this kind of research was made in the mid-1970s by Eleanor Maccoby and Carol Jacklin. The Psychology of Sex Differences reported that block differences between women and men (generally white, affluent, North American college students) did appear fairly consistently in studies of some traits: verbal ability, visual/spatial ability, mathematical ability, aggress — ivenp^TA finding of no block difference fairly consistently appeared
^studies of rather more traits: sociability, suggestibility, selfesteem. types nf learning, cognitive styles, achievement motivation, sensi^L-morfetiTy. The authors entered an ‘open finding’ — i. e., no consistent pattern — on another range of traits: tactile sensitivity, timidity, activity level, competitiveness, dominance, compliance, пщіш^н-егг The research has flowed on since Maccoby and Jacklin wrote. Its results have been mixed, but if it has had any trend it has been to close some of the gaps between women and men that they left open^Robert Plomin and Terry Foch. for instance, tincf sex differences accounting for only 1 per cent of the variance in a study of children’s verbal and Quantitative abilities. Olive Johnson and Carolyn Harley report that being right — or left-handed is a
better predictor of cognitive ■abilities J_han is sex^Of course these are only two examples from a mass ofstudies, ana there are others which suggest stronger sex differences. Nevertheless one can be very confident of a negative point. Recent research has not shown that Maccoby and Jacklin systematically underestimated sex differences.
A striking conclusion emerges. The logic of the genre focuses on ‘difference’ and its explanation. In fact the main finding, from about eighty years of research, is a massive psychological similarity between women and men in the populations studied by psychologists. Clear-cut block differences are few, and confined to restricted topics. Small differences-on-average, in the context of a very large overlapping of the distributions of men and women, are usual even with traits where differences appear fairly consistently. If it O^erenot fbrtbe culturaPbias of both writers and readers, we might long ago have been talking about this as ‘sex similarity’ resean
In so far as these scales and measures can be trusted, the notion of distinct unitary sexual characters for women and men has been decisively refuted. With it, much of the common-sense understanding of sex and gender, together with most functionalist sex role theorizing, should collapse.
This has not happened. There is too much invested in the notion of sexual character for a simple factual refutation to destroy it. Yet there has been serious pressure to modify it. The simplest modification is to abandon the idea of categorical differences and interpret the results of data-dredging sex difference research through the formula of variation about a norm.
Overlapping distributions now matter less. Differences between means suffice to establish difference between a male and a female norm. Role theory provides a gloss. Variation about each norm reflects role distance, vagueness or conflict in the expectations, or even deviance. Those studies which fail to find mean differences are interpreted as evidence of overlapping expectations or the convergence of traditional sex roles.
As argued in chapter 3, role theory is infinitely plastic. There is no great interest in its exercises in retrieval. But there is something of more note here: the shift to a focus on variation. Concern with variation in the traits for which feminine and masculine sexual character are explanations is only a step from concern with variation in femininity or masculinity themselves.
This path leads to non-unitary conceptions of sexual character.
The need to take this path is now clear in the research literature; all interesting conceptions of sexual character from here on are non-unitary. Both femininity and masculinity vary, and understanding their variety is central to the psychology of gender. How is that variety to be understood? There are several possible approaches.