Everyone takes surveys.
Whoever makes a statement about human behavior has engaged in a survey of some sort. Consider four statements:
1)A Catholic Cardinal (Bernard Law) announces that the Catholic laity are
glad that the Pope
continues to remind them about the evil of birth control.
2) Richard Lewontin in an attack in the New York Review of Books -- "Sex, Lies, and Social science" -- on the recent survey of American sexual behavior by The National Opinion Research Center (The Social Organization of Sexuality)1 says that the scholars are "deaf and dumb" because they believe that 45% of the men between eighty and eighty four still have a sex partner.
3)Journalists report that "Generation X" is less concerned about politics than its predecessors.
4) A priest announces that no one has ever interviewed him or one of his friends in a survey and therefore he thinks that surveys are a fraud.
Each of these statements are the result of surveys that the one making the statement has taken -- unless they have wet their fingers and held them up to the wind. Cardinal Law, Professor Lewontin, the journalists and the priest, have made observations, analyzed their observations, and generalized from them. The Cardinal has talked to some Catholic laity; the Professor has (presumably) observed the behavior of men in their early eighties; the journalists have talked to young Americans; the priest, by his own admission, has interviewed his friends.
You cannot generalize about human behavior unless you observe it one way or another, analyze your observations, and then generate a statement about them. The difference between the survey takers and the rest of generalizing humankind is that the former (usually) observe large numbers of people in a representative sample that reflects the total population and are precise about their methods of data collection and analysis. Precisely because the professional survey taker is honest about his methods, he becomes an easy target for loud mouth critics who appeal to "what everyone knows" and "common sense."
Thus one could say at a dinner party that most Americans are faithful to
their spouses and a loud
mouth, semi-inebriated jerk can respond that the conclusion cannot be
true because look at all the
academics who sleep around at professional meetings and besides surveys
cannot be trusted. Sic
The methods and techniques of the survey on which one has based one's comment are out in the open and are easy targets for criticism whereas the methods of the loud mouthed jerk are not revealed. Bad research has driven out good research.
The person who engages in survey research knows his limitations, at least if he is wise and experienced. The critic who responds with his generalizations is serenely unaware of his own limitations. The survey taker knows that even with the most carefully designed sample, the most cautiously crafted questionnaire, the most responsible field staff, and the most ingenious analytic techniques, the best he can claim for his work is a rough approximation of human behavior. He cheerfully admits that social science is not a science in the same sense as, for example, is Professor Lewontin's studies of insects.
Lewontin's attack on The Social Organization of Sexuality seems designed
to make that point, but
social scientists conceded it long ago. "There are some things in the
world," he writes, "that we will
never know and many that we will never know exactly."
Surely, but, put off by perhaps by pedantic and sometimes turgid style of The Social Organization of Sexuality, he seems to assume that the authors of the book are claiming the same degree of certainty about human copulation that he can have about insect copulation. Only a survey researcher who is a charlatan or a fool would make such a claim. If all knowledge is problematic, survey knowledge is especially problematic because of the complexity of human behavior. The survey scholar is attempting merely to be as precise as possible in his approximations by being clear about his methods and assumptions. In assuming that the survey scholar is attempting more, Lewontin is simply talking past them.
Surveys are limited tools but they are better than holding the saliva- tipped finger in the wind or generalizing to all humankind on the basis of what one's academic colleagues might or might not do at professional meetings.
However, as Harrison White, a man with doctorates in both Astrophysics and Sociology once remarked, "Science is humankind pursuing truth with no holds barred." By that standard, social science is indeed science though in a difference sense than astrophysics -- or the study of insects (both of which deal with relatively elementary problems compared to those encountered in the study of human beings).
Indeed the only conclusion the survey researcher can advance with some degree of certainty (as constrained by statistical confidence estimates) is that if he did everything right, the responses with which he is working reflect the answers the whole population would give if the questionnaire were administered to everyone in the population. It has often seemed to me, however, that this is not unimportant information.
Lewontin's principle argument against the book is that the authors "pretend that one knows nothing about the experiences of being human, forcing the investigator to pretend that people usually know and tell the truth about important issues when we all know from our own lives how impossible that is. (Italics added).
Again Professor Lewontin has made his own survey. He has spoken to or observed enough people to generalize about what "we all know," playing the role of the loud mouthed jerk at the dinner party. He is in effect saying that his survey of what people know indicates that people don't know and don't tell the truth about the important issues in their lives. If telling the truth is something important, then how can he believe his own survey about whether they tell the truth? Obviously Professor Lewontin surveyed only himself. He knows that he would not tell the truth to a National Opinion Research Center interviewer (indeed he says he would decline to be interviewed which is of course his privilege). Yet how can he be so certain that other people are incapable of giving a rough approximation of the truth in their interviews? He confronts the authors with his own personal opinion and nothing more. Moreover he does not say whether he thinks that respondents will exaggerate or underestimate their sexual behavior.
The issue remains as to whether there is any means by which responses to a survey of sexual behavior can be verified. Election surveys can be validated by the actual outcome of an election (which apparently is not an "important issue" in life). Respondents seem to underestimate their income, overestimate their participation in voting, perhaps overestimate their church attendance (though that remains to be seen). Is it possible to verify the responses in a survey of sex? Or might people be unwilling or unable to tell the truth?
Lewontin says that there is no such means available and that in fact one of the best measures, the number of life-time sexual partners, fails because men report many more partners then women. I respond that there are at least three ways of obtaining partial validation and that together they create a converging probability towards roughly accurate portraits of sexual behavior. 1) If one varies the method of interviewing from very private to very public and response patterns remain the same, the chances that respondents are telling the truth increases. 2) If one varies the cultures in which surveys have been taken, and response patterns remain the same, the chances that the respondents are telling the truth increases. 3) If one varies the gender of respondents and many response patterns remain the same, the chances that the respondents are telling the truth increases.
To the first criterion: in the various sex surveys around the world methods of data collection have varied from highly public -- the telephone interview on which my work Faithful Attraction2 was based -- to the extremely private -- self administered questionnaires mailed to the survey center (as in the Irish study). Response patterns vary only slightly.
To the second criterion: as the authors of The Social Organization of Sexuality report in the book, there is very little variation across cultural lines. Presumably cultural attitudes towards sex are very different in France and Denmark than they are in England and the United States (and I would add Ireland). Yet response patterns in all these countries are virtually the same.
To the third criterion: While there is indeed a difference between men and women in reports of the total number of life time sexual partners, there are similarities in their responses to other questions. For example, as I reported in this journal before the publication of The Social Organization of Sexuality, the infidelity rate among both men and women is the same -- 15% -- when one compares working women with working men who don't have sex with prostitutes. Husbands and wives report approximately the same rates of at least weekly sex. They also report approximately the same frequency of oral and anal sex. Given the different physiological and psychological reactions to sex in the two genders, these similarities provide impressive quasi-verification. As to the difference between men and women in reports of life-time sexual partners, there is some evidence in the Irish research that much of the difference may be attributed to exaggeration by men. Thus in the United States a quarter of the unmarried men and two fifths of the unmarried women report that they had no sexual partner in the last year. In Ireland a little less than half of both genders report no sexual partner. Perhaps the Irish men see no need to exaggerate their sexual accomplishments.3 Similar arguments can be made about other parts of the book. Thus the increase in a forced first sexual encounter among younger women can hardly be the result of fantasy among such women. Why would they be more likely to indulge in such ugly fantasies than older women?
My argument is that these criteria create not certainty (which is not attainable in research on human sexuality) but a converging probability that the authors of The Social Organization of Sexuality have presented a picture of sexual behavior in America which is roughly accurate as far as it goes. That is a major accomplishment whose importance ought not to be minimized, much less ridiculed as the New York Review of Books ridicules it. The study is less than perfect and necessarily less than completely accurate and adequate. It is still the best effort so far to draw a broad portrait of sexual activity in this country.
No more and no less.
A survey is an imperfect way of studying humankind. Nonetheless it serves useful purpose and should be neither worshiped nor ridiculed merely because the "up-front" candor of the survey scholar about methods makes him vulnerable to ridicule.The survey is an imperfect way of knowing. So are all the others.
The Social Organization of Sexuality has its flaws. The presentation is often plodding and dull, even when one grants that it is de rigeur for social science writing to be heavy. The questionnaire was innocent of creativity and flair: as one of the reviewers in the American Journal of Sociology observed there are no questions about hugging, kissing, and love play, actions which the species seems to find enjoyable, especially, if one is to believe reports, women of the species. The most serious weakness of the study, it seems to me, is that it pays little attention to affection and none to love. Perhaps this omission is hard-headed sociology. Yet, if the most frequent sex and apparently the best sex is that between married partners who are faithful to one another, is there not a hint that affection might be an important aspect of sex? Even love? While the link between sex and love may be understood by poets and novelists, social science has yet to explore it, perhaps because it is considered too "soft." Does love tend towards sex? Does sex tend towards love? Is sexual love one of the binding energies which holds the social structure together?
It is perhaps unfair to my NORC colleagues to suggest that they should
have explored those
questions. But the questions do seem appropriate for the next study.
Despite these criticisms (which are not original with me), it is still the most cautious and careful study of American sexuality currently available to us. To argue, as Lewontin does, that it is no better than the Kinsey reports, is to lose all sense of balance and perspective.
Why the ferocity of the attack, especially since most other reviews have been favorable?The New York Review of Books serves a function for certain members of the intellectual elite that Hong Kong martial arts films serve for some kinds of American males -- it provides raw meat for those who love to see others torn to shreds. Perhaps the media reports about the "conservative" nature of American sexual behavior (the data on the sexual abuse of women, especially of younger women seem not to have attracted much media attention) may have offended some of the Review's editors. So too may have the findings about the incidence of homosexual and the dangers of AIDS contaging to the general population. Many men like Richard Lewontin don't realize that they routinely engage in their own informal surveys (or self-examination about what it means to be human) to sustain their attacks on the more formal surveys. There is always the jerk at the dinner table syndrome. Finally one should never exclude the impact of invidia academica on book-reviewing.
Even Professor Lewontin offers some validation4 of the generally
"conservative" portrait of American
sexuality that the book provides: "If I can believe even half of what I
have read in The Social
Organization of Sexuality, my own sex life is conventional to the point
of being old fashioned."5
So it would seem is the sex life of most everyone else.
1 Even though I had published a book and several articles on the subject
I had no part in the
enterprise from questionnaire design to report writing. One of the
authors explained to a group of
students that if I were part of it I would have a book written and a
press conference within six weeks
-- a comment the students promptly brought to me. That may be true
(though the book would have
been better written), but it would not have happened if I had given my
work not to violate an
embargo. Thus I do not have a vested interest in defending The Social
2 The monogamy and fidelity rates reported in Faithful Attraction were replicated in The Social Organization of Sexuality
3 In 1994 the General Social Survey questions were replicated in Ireland. In most indicators there was little difference between Ireland and the United States. Professor Lewontin who has a couple of snide anti-Catholic remarks in his review would be well advised to withhold them when commenting on Ireland. The Irish are, incidentally, more likely to report same gender sex than Americans.
4 Despite his assertion that he would not talk to an The National Opinion Research Center interviewer
5 One could be cynical about Professor Lewontin's claim to conventionality by citing his own words:" the investigator (is forced) to pretend that people usually know and tell the truth about important issues when we all know from our own lives how impossible that is."