Nonviolent, Sexually Explicit Material and Sexual Violence

Part: 
One
Chapter: 
3

Submitted by: Father Bruce Ritter

  1. Background
    1. Problem of Definitions
    2. Evidence and Standard of Proof
  2. The Evidence
    1. Changes in Rape Rates
    2. Correlational Evidence
      1. Danish and Other Cross-Cultural Data
      2. Sex-Magazine Circulation
      3. Sex Offenders and Pornography
      4. Conclusions from Correlational Evidence
    3. Experimental and Clinical Evidence
      1. Arousal
      2. Effects on Attitudes Toward Rape- "Disinhibition"
      3. Overall Evidence for "Causation"
    4. Evidence Against Causation
  3. Conclusion

 

  1. Background

     

    The alleged relationship of sexually explicit material and sexual violence has long been a subject of acrimonious but compelling debate. The "Effects Panel" of the 1970 Commission, often accused of denying such a link, instead stated a relatively moderate view of what was then an almost entirely new area of inquiry: "On the basis of the available data . . . it is not possible to conclude that erotic material is a significant cause of sex crime."[1] Recognizing the impossibility of ever proving "conclusively" the existence of such a casual connection, the 1970 Commission nevertheless determined that the evidence did not, at the time, suggest a "substantial basis" for such a proposition.[2]

    The findings of our predecessors, though beleaguered in this area by extensive professional criticism,[3] are entitled to significant deference, especially because the 1970 Commission took pains to explain the basis of its conclusions. Rape, however, is among the most violent and damaging of crimes: not only inflicting deep injury on its victims, but also standing as a powerful obstacle to the fight for sexual equality in a democratic society. It is, further, an evil which has increased at shocking rates over the last fifteen years. We thus have the grave, and undeniably unpleasant, duty to examine again the possibility that consumption of sexually explicit materials and some rapes are causally linked-and to report, on the basis of the evidence available now, whether a "substantial basis" exists for believing in such a link.

    We have with little trouble concluded that circulation of materials which themselves portray graphic sexual violence is a probable "cause" of rape-at least in the sense of being one factor among many (and not necessarily the most important) which increases the likelihood of rape. With regard to sexually explicit materials which do not include depictions of violence our task is more difficult because so many of our witnesses, so many professionals, and so many of our fellow citizens disagree vehemently on the issue. Tempting as it is simply to wash our hands of the question by noting the existence of the dispute and refusing to "take sides" in it, we cannot avoid sifting through the evidence and attempting to come to our own conclusions on the matter. Even if we cannot ultimately agree on the purport of each piece of evidence, or the meaning of all the data collectively, our views should be fully, and publicly explained.

     

    1. Problem of Definitions

      One serious obstacle to such explanations, unfortunately, arises immediately in the guise of defining the material under examination. For purposes of general discussion about the possible "harms" of sexually explicit material we have found it useful to divide that material into three somewhat imprecise, but nonetheless useful categories: that which is (1) violent; (2) "degrading" but not violent; and (3) neither violent nor "degrading". Unhappily our scheme was not anticipated in advance by researchers and, though a useful blueprint for future scientific inquiry, has not formed the basis for research conducted in the past. The only distinction adhered to with some consistency in the past research has been that between those materials which depict violence and those which do not. Obviously that distinction is a crude one given the wide range of nonviolent "pornographic" materials, yet it may in some sense correspond with popular perception: thus public opinion seems strongly opposed to free circulation of materials "that depict sexual violence," but sharply divided over the fate of materials that "show adults having sexual relations," with no further explanation of whether the materials in question are "degrading" or not.[4]

      For purposes of examining the evidence regarding sexually explicit materials and sexual violence, then, it seems useful to begin, at least, without clearcut distinctions based on the "degrading" character of particular items. Rather, the case for linking nonviolent materials and rape should be examined on its own terms-that is, on the basis of definitions contained in the relevant research-with attention, ultimately, to those pieces of evidence which bear on the question of distinctions among various categories of nonviolent materials. Until we sort through the evidence on this issue we cannot, after all, be certain that boundaries useful for distinguishing among materials on observable attitudinal effects are equally valuable with regard to behavioral impacts.

       

    2. Evidence and Standard of Proof

      The assumption that consumption of sexually explicit material "causes" sexual violence is one that some 73 percent of Americans would accept as true,[5] but it is unclear what evidence they would point to as crucial to their judgment. From our standpoint some forms of evidence are clearly more persuasive than others, but no one is useless and nondispositive. Evidence from the social sciences-correlational, clinical and experimental-seems by a wide margin the most important tool of analysis in this area, in part, paradoxically, because its limitations are most apparent. The results of individual experiments or studies can be rigorously challenged on terms universally accepted by social scientists, and can be examined as carefully for what they do not "prove" as for what they do. Anecdotal evidence, even that presented by skilled professionals, has an unfortunate tendency to touch on a wide range of questions without furnishing the basis for answering any single one of them.

      Particularly on an issue as bitterly fought and important as this one, therefore, reliance primarily on data from the social sciences seems appropriate and quite possibly imperative. That does not mean, however, that we are bound by the standards of "proof" which govern the work of social scientists. Our task after all, is to recommend policy based on existing knowledge in an area that will always be plagued by uncertainty. Because of limitations on the capacity of social science to measure events outside the laboratory, and because of clear ethical boundaries on what research can be conducted in this area even in the laboratory,[6] it seems wholly unlikely that the extremely high standards for "scientific proof" can ever be satisfied one way or the other on this issue.

      The standard more appropriate for our purposes is suggested by the phrase used by the 1970 Commission: is there a "substantial basis" for believing that nonviolent but sexually explicit material is causally linked to sexual violence? If so, what evidence suggests the opposite conclusion-that no such link exists? Finally, which evidence on balance is more persuasive? (This standard was used by us as "the totality of the evidence" in our discussions.) Because rape is so widespread and so dangerous an evil, government action against constitutionally unprotected material might be appropriate if a "substantial basis" for believing in a causal link between such material and sexual violence exists, and might seem imperative if the evidence allows a stronger assessment. Just as government action against cigarette advertising could not await final, irrebuttable "scientific proof" of the causal link between cigarette smoking (let alone cigarette advertising!) and lung cancer, so the government may not be able to await scientific consensus on the pornography/rape connection-even if such consensus were imaginable.

  2. The Evidence

     

    Because direct experimental research on the alleged causal relationship between sexually explicit materials and sexual violence is impossible, or at least unthinkable, we are unhappily left to examine evidence of an indirect nature. That evidence, when it comes from the work of social scientists, tends to take one of two forms: correlational studies and laboratory experiments. The former is a useful launching point for an overview of the issue, because it measures statistical relationships between actual violence and actual consumption of sexual materials. Were no significant relationship found to exist between those two phenomena even on a statistical level, any causal connections between that be extremely difficult to demonstrate through work in the "artificial" setting of a laboratory. Such a setting is useful, however, for exploring possible causal relationships between statistically correlated events; and that is the sense in which experimental evidence is relied on here. Before either correlational or experimental evidence is examined, however, it is crucial to consider first whether sexual violence is a problem which might ever be affected by social change, and whether, in fact, as an aggregate phenomenon it has increased during the period in which sexually explicit materials have been widely available.

     

    1. Changes in Rape Rates

      That first question is easily answered. Rape rates do seem to be related to social change, for they have increased alarmingly during the past 25 years. From 1960 to 1970 the rate of reported forcible rape rose by 95 percent, but that increase seems to have been no more than part of an explosion of violent crime generally, which rose fully 126 percent during the 1960's.' Since the report of the 1970 Commission, however, the rate of reported rape has risen almost twice as fast as violent crime generally;[8] from 1970 to 1983 the rape rate virtually doubled, while the rate of reported homicides, for example, remained constant.[9] In 1970 one out of every 20 violent crimes was a forcible rape; by 1983 the proportion had become one out of 16.[10]

      Was this extraordinary rise in rape a "real" occurrence, or merely a product of increased reporting of rape? The possibility that increased sensitivity to rape-fueled by movements for women's equality-led to increases in the willingness of individuals to report rapes is not one that can lightly be dismissed,[11] for rape is highly underreported crime.[12] Nevertheless at least three pieces of evidence suggest that the increase of reported rape is not tied to increased willingness-to-report. The National Crime Survey, to begin with, which attempts to gauge actual (as opposed to reported) crime figures through a scientific public survey, showed no significant change in the percent of rapes reported to police from the period 1973-1977 to that of 1978-1982.[13] Yet between those two periods the average number of estimated actual rapes increased substantially.[14]

      Second, the 1978 survey by Professor Diana Russell found an increase in the "true rape rate" throughout most of this century;[15] thus historically no serious misrepresentation of trends in this area is found in police data. Finally, correlational data from recent studies of state-by-state rape rates and measurements of the status of women indicate only a small, although significant, relationship between the two.[16]

      Rape appears, therefore, to be a phenomenon subject to fluctuation, and during the period that sexually explicit materials have come into general circulation it has been a phenomenon on the rapid increase. That last fact, however, in no sense "proves" or even substantially "suggests" a relationship between the two events; only detailed correlational analysis can begin to do that.

       

    2. Correlational Evidence

      Our predecessors on the 1970 Commission had no sophisticated "correlational" data before them. Indeed, the only "correlational" data which they considered was of the sort discussed above-general trends in the sex-crime rates measured for time periods in which sexual materials were becoming more available. Unfortunately, for reasons discussed below, that sort of evidence is far too crude to be of significant value, and points, in any case, in no particular direction. Far superior correlational data has in the meantime come to the fore, and it shows that a statistical relationship does appear to exist between consumption of certain types of sexual materials and rape rates. Both types of data invite the most careful attention.

       

      1. Danish and Other Cross-Cultural Data

        The 1970 Commission was impressed, as was the Williams Committee later, by studies on Denmark conducted by Berl Kutchinsky in which he found that relaxation of Danish pornography laws coincided with a decrease in reported sex crimes. Since that time Kutchinsky's work has been repeatedly criticized, and he himself has been forced to concede that, at least with regard to rape, liberalization of pornography laws was followed ultimately by increases in reports of rape to police.[7] Further, Kutchinsky's approach fails to be even minimally persuasive for two crucial reasons. First, he does not account in any meaningful way for other social forces which might have affected Danish sex crime rates independently of pornography consumption. He fails to note, for example, that sex crime rates in Denmark might have been artificially high during the 20 years after the German occupation of World War II, a conflict described by one historian of Scandinavia as "shattering physically as well as emotionally."[18] A drop in sex crimes during the late 1960's and after would thus be the result simply of recovery from social disintegration wrought by war. Second, and substantially related, Kutchinsky fails to consider the case of Norway-a country with a similar culture and a similar war experience-which has maintained far stricter laws against pornography,[19] and has apparently enjoyed even greater success in combatting sex crimes.[20] In the end Kutchinsky's analysis seems shallow and almost completely without value for analysis of the American experience and American policy.

        A more appealing cross-cultural approach, but one with only marginally greater usefulness for our purposes, is that taken by Dr. John Court (1984). His research has examined the temporal changes in rape rates in a wide variety of countries in periods of greater or lesser legal control of pornography. His conclusion, presented with considerable cogency, is simply that greater legal control of pornography appears to hold down rape rates as well. Yet for all its resourcefulness Court's work fails, like that of Kutchinsky, to place the changes studied in careful historical and cultural perspective: thus Singapore, South Africa, Australia and Hawaii are all compared with little contextual information. An additional, related limitation on the helpfulness of his findings arises from his inability to show, like Kutchinsky, whether actual consumption patterns fit neatly into the patterns of changing legal regulation of sexually explicit materials. Our experience of American enforcement of obscenity laws indicates that such laws are often honored as much in the breach as in the observance.

         

      2. Sex-Magazine Circulation

        Interesting as the work of Kutchinsky and Court is, we have had the benefit of receiving a body of correlational evidence of far greater power. The research of Baron and Strauss (1984, 1985) supplemented by others, has shown a strong statistical relationship between state-by-state circulation rates for the most widely read "men's magazines" and state-by-state reported-rape rates. That relationship persists even when every other factor theoretically associated with rape is controlled for: indeed, they found that the Sex Magazine Circulation Index has a consistently stronger statistical relationship with rape rates than any other factor tested." Further, in the model developed by Baron and Strauss other variables theoretically expected to be related to rape rates in fact met expectations: those factors (e.g., percent urban, percent poor) together with the Sex Magazine Circulation Index explain 83 percent of state-to-state variation in rape rates.[22] Two independent studies, by Scott (1985) and Jaffee and Strauss (1986) have not only replicated the Baron and Strauss results for different years, but have cast doubt on potential "third factors" which would make the sex-magazine/rape association spurious. Baron and Strauss offered two such factors as possibilities: (1) a cultural pattern emphasizing "compulsive masculinity"; and (2) the degree of "sexual openness" within states. The first of those suggestions was undercut by Scott's finding that circulation of men's "outdoor magazines" is not associated with state-by-state rape rates. In addition, Baron and Strauss found that controlling for the "index of legitimate violence" and the general violent-crime rate-both seemingly plausible measures of a culture of "compulsive masculinity"-in no way lessened the sex-magazine/rape correlation. Nor did controlling for measures of the status of women-a plausible inverse measure of the degree of "compulsive masculinity" within a given state. Finally, the recent work of Check (1984) and Zillman and Bryant (1984, 1985) indicates that under experimental conditions, massive exposure to mainstream pornography may cause male viewers to become more callous and domineering in their attitudes toward women. Thus pornography may itself be a causal factor in creating a culture of "compulsive masculinity," and even if a correlation could be shown between such a culture and the incidence of rape, the association of the latter with sex-magazine circulation would still not be proved spurious.

        As for the other "third factor" suggested-the degree of "sexual openness"-the recent study of Jaffee and Strauss (in press) measured the impact of the Sexual Liberalism Index on the Baron and Strauss formulae. While finding that sexual openness and tolerance is correlated, to a small but significant degree, with increases in reported rape rates, Jaffee and Strauss discovered the inclusion of the new index had no effect at all on the sex-magazine/rape association. While continuing to hold out hope-against all the evidence mentioned in the previous paragraph-that a relationship between "hypermasculine gender roles" and rape rates would render the sex-magazine correlation spurious, they felt compelled to conclude that their research "suggests that there may be more to the pornography-rape linkage than originally expected. That is, the type of material found in mass circulation sex-magazines may, as claimed by critics of such material, encourage or legitimate rape."[23]

         

      3. Sex Offenders and Pornography

        Somewhat less suggestive and useful, but nonetheless important, is correlational evidence exploring links between the use of sexually explicit material by sex offenders and their behavior. Dr. Gene Abel's (1985) study, in particular, is directly pertinent to the issues raised by Baron and Strauss: in treatment of 247 outpatient sex offenders (paraphiliacs), well over half admitted to use of adult men's magazines or similar material, and 56 percent of rapists stated that such materials "increased their deviant sexual interests." Comparison of those offenders who use "erotica" and those who do not produced only one statistically significant difference of direct relevance: users of "erotica" maintained their paraphilia far longer than nonusers. Between those whose deviant arousal was increased by "erotica" and those whose deviant arousal was not increased two statistically significant differences emerged: (1) the aroused-by-erotica subjects maintained their paraphilia longer; and (2) they had less "ability to control their behavior." On the whole, Dr. Abel concluded that "erotica ... does not appear to affect significantly the behavior of sex offenders."[24]

        Careful review of Dr. Abel's results and of his oral testimony, however, tends significantly to undercut that assertion. To begin with, the mean number of sex crimes committed by users of erotica was 29 percent higher than the mean for nonusers. Dr. Abel lists the difference as "not significant" but does not supply a "p value"; we thus cannot gauge what the actual probability is that the difference is explained only by chance.[25] The finding of no significance is particularly puzzling because, according to Dr. Abel's other findings, users of "erotica" commit the same number of sex crimes per month (actually 21 percent more, but once again the difference is listed as "not significant") and maintain their paraphilia for more total months. Mathematically this would seem to compel the conclusion (already suggested by the statistics on "mean number of sex crimes") that by the end of their paraphilia, the group using "erotica" will have committed more total sex crimes than nonusers. That indeed seemed to be the gist of his oral testimony, where he explained the "price" paid by sex offenders who use "erotica" to reduce their desire to commit sex crimes:

         

        ...when you use the deviant fantasy in order to ejaculate, instead of attacking a kid or raping someone, it does transiently stop you from carrying out that behavior. In many cases, that is the case, but it's a transient phenomena. And in so using that tactic, the price you pay is maintenance of your arousal. That is your arousal stays strong and will get a little stronger. So over time you are more likely to maintain your arousal over a longer period of time, that means you can commit more acts.[26]

         

        In view of these internal tensions, Dr. Abel's results are extremely difficult to use in their present form.[27] They seem clearly to indicate, and Dr. Abel said as much, that use of "erotica" by sex offenders (outside a treatment setting) is not "helpful."[28] On the other hand they do not seem to rule out, Dr. Abel's protests to the contrary notwithstanding, the possibility of some important statistical relationship between use of sexually explicit materials and commission of sex crimes by this population.

        The possibility of such a relationship is clearly enhanced by several other relevant studies. Thus Dr. William Marshall (1985) found in an outpatient study that a far higher percentage of sex offenders currently use "hard-core" pornography than do a group of demographically similar "normals." Professor Diana Russell found high correlation in her study of 930 randomly selected adult women: a surprisingly high number of women victimized by wife rape and stranger rape who said pornography had played a substantial role in the event. A similar survey of 200 prostitutes by Silbert and Pines (1982) found that 24 percent of the large number who had been raped "mentioned allusions to pornographic material on the part of the rapist"-this without any questioning or prompting by the interviewer. Law enforcement witnesses we have heard have also consistently stated that pornographic materials are routinely found on the person of, or in the residence of arrested rapists. While all of this is, like Dr. Abel's evidence, "merely" correlational data, it suggests reason for further inquiry and research on the use of sexually explicit nonviolent materials by sex offenders.

         

      4. Conclusions from Correlational Evidence

        An overview of "correlational" evidence available to us ultimately leads to only one firm conclusion. A highly significant, and not obviously spurious statistical relationship exists in the United States between state "adult magazine" circulation rates and sexual violence. That relationship may be explained by a causal connection or it may not; only careful attention to other forms of evidence can indicate which explanation is more plausible. Because "adult" magazines contain relatively little violence,[29] their connection (if one exists) to rape rates makes an excellent "test case" for considering the possible effects of the broader class of nonviolent but sexually explicit materials.

        No clear statistical relationships exist, on the other hand, between cross-cultural measures of rape and sexually explicit materials, although such measures if anything tend slightly to support some relationship between the two. Nor is there undisputed evidence regarding the correlation of "erotica" use by sex offenders and commission of sex crimes; it is at least strongly arguable, however, that such a relationship exists. Other sources of information may prove more informative in evaluating these ambiguities.

       

    3. Experimental and Clinical Evidence

      A "casual" connection between circulation of adult material and sexual violence may only be inferred if one or more plausible explanations exist for how such "causation" could exist. Experimental evidence is particularly important in testing the likelihood of such causal links; as noted above, however, ethical and practical constraints insure that such evidence will always be open to charges of artificiality and obliqueness.[30] Simply put, actual rapes cannot be staged in the laboratory, nor can known rapists be subjected to testing which might provoke future violence. Retrospective "clinical" evidence, although it does generally relate to "real" rapes by "real" offenders, has the even more crippling handicap of relying on faulty, and self-serving, memory. Yet experimental and clinical evidence remain in this area the most effective tools for testing the "validity" of correlational data.

      Searching the evidence for suggestions of a "cause-and-effect" pornography/rape connection inevitably leads down two different paths. The first observes the capacity of pornography to effect arousal in the viewer, and examines whether such arousal can be causally linked to sexual violence. The second, somewhat more indirect approach examines the effects of pornography consumption on viewer's attitudes, then considers whether such changes in attitudes could plausibly affect the incidence of rape.

       

      1. Arousal

        One of the few undisputed properties of sexually explicit materials is their capacity to cause sexual arousal in many, if not most viewers.[31] One strand of experimental research has attempted to determine whether this arousal, alone or in combination with other factors, increases or decreases aggressive behavior in laboratory settings.

         

        1. "Normals"

          With regard to "normal" subjects (usually college-age male volunteers), the results have been mixed, or at least highly complex. Thus highly arousing erotic materials, when combined with prior or subsequent anger, seem clearly to provoke heightened aggression by males against males.[32] But in a recent review of the research Professor Donnerstein made the following, more limited, statement about the effects of exposure to nonviolent pornography on male aggression toward women:

          ... The question of whether or not nonaggressive pornography has an influence on aggression against women is not simple to answer. For one thing, there is not that much experimental research on the topic. Also, studies investigating this issue have differed in many ways.... These studies indicate that under certain conditions exposure to pornography can increase subsequent aggression against women. What seems to be required, however, is a lowering of aggressive inhibitions. This change in aggressive predisposition can come about in a number of ways. First, a higher level of anger, or frustration, than that exhibited in a laboratory setting could influence the effects of pornography on aggression against women. There is no question that such levels are present in the real world. Second, as mentioned earlier, drugs, alcohol, and other aggression disinhibitors very likely increase aggressive response to pornography. The main mediating factor, however appears to be the type of material viewed prior to an aggressive opportunity.[33]

          While experimental findings are neither conclusive nor absolutely consistent, the bulk of research to date supports the conclusion: that where highly arousing nonviolent pornography is viewed in a context of anger or provocation, aggressive behavior against women increases. Outside the context of provocation, in Professor Donnerstein's view, nonviolent material which is "either mildly arousing or leads to a positive affective reaction" does not appear to increase subsequent aggressive behavior, while that which depicts "unequal power relationships with women" or "women as sexual objects" may provoke such behavior. As part of his belief that the issue warrants "much more investigation" he notes that the effects of nonaggressive pornography may not occur with only a single exposure,[34] which would explain varying results in experiments based on single exposure. Growing habituation to standard "pornography" over the years among likely experimental subjects may substantially affect the results of research.[35]

           

        2. Sex Offenders

          Along slightly different lines, a certain amount of experimental and clinical evidence suggests that rapists are aroused by nonviolent, sexually explicit materials, and that some consciously use such materials to prepare for and execute sexual violence. Thus rapists are normally as strongly aroused to consensual nonviolent pornography as nonrapists; they are, moreover, at least as aroused to images of mutually consenting sex as they are to those of rape.[36] Does this arousal to mutually-consenting imagery cause some of them to commit sex crimes which they might otherwise avoid? Evidence from at least Dr. William Marshall suggests that the answer may be yes: 33 percent of rapists interviewed for his study "had at least occasionally been incited to commit an offense by exposure to one or the other type of pornography specified in this study."[37] Of that group 75 percent reported that they had at least occasionally used 'consenting' pornography to elicit rape fantasies which in turn led to the commission of a rape (or an attempt at commiting a rape)."[38] A large number of other rapists in his sample used "consenting pornography" to "evoke rape fantasies" and consequent arousal. Indeed, fully 52 percent of the rapists in his sample (as compared to none of the "normals") used pornography "always" or "usually" during masturbation.[39]

          Dr. Abel, while stating the belief that direct incitement to rape can be traced to sexually explicit depictions only in "exceedingly rare" cases, also found that a very high proportion of rapists use consenting "erotica" to elicit and maintain deviant arousal. Recent research has shown a high correlation between sexually deviant fantasies and deviant behavior,[40] and many treatment programs for rapists have been predicated on altering their deviant behavior through changing their fantasies and arousal patterns.[41] Dr. Abel and his colleagues at one point called for recognition of "fantasy as the pivotal process leading to deviant behavior."[42] To the extent that nonviolent, "consensual" pornography contributes to provoke or maintain deviant fantasy and arousal in rapists, it may be considered a "cause" of their deviant behavior.

           

        3. General Population

          Turning back to the general population-that is, both sex offenders and "normals"-it is important to note two significant theories concerning sexually aggressive behavior which are predicated on the biological forces of simple arousal. The first, called the "general emotional arousal theory," is described in one study as predicting that "by arousing either the sexual or aggressive drives in an individual, the overall general level of arousal would be increased, thereby making both sexual and aggressive responses more probable"[43] The second theory, which is more subtle and more flattering to the human will, adds an additional cognitive layer to the general-arousal theory:

           

          While evolutionary forces may have provided a biological basis for a link between sex and aggression, it is our contention that learning variables may accentuate or attenuate this relationship. We hypothesize that in human beings the biological link plays a relatively minor role and that to a large extent the relationship between sexual arousal and aggression is mediated by learned inhibitory and disinhibitory cues.[44]

           

          Both theories associate arousal with aggression; the second merely adds the additional mediating factor of "learned inhibitory and disinhibitory cues." If this association is ultimately found valid, then a "casual" connection between circulation of highly arousing sexually explicit materials and the incidence of rape would be both clear and easy to explain: more sexual arousal in society (as a consequence of pornography) inevitably produces more sexual and more aggressive behavior, both of helpful and harmful varieties. If viewing sexually explicit materials cause Americans to have more sex, then some of that incremental sexual behavior will be of a sexually aggressive nature. The "rate" of rape as a percentage of all sexual intercourse will not change,[45] but the absolute number of rapes, and the number of people victimized by rape, will increase.[46]

          The ability of sexually explicit materials to arouse those who view them may, therefore, be in itself a "cause" of sexually aggressive behavior-perhaps simply for rapists, or perhaps in a more general way. This evidence does not distinguish sexual material as being more culpable than, say, alcohol as a causal factor in rape-but it does suggest that the more highly arousing the material is, the greater will be its ultimate effect. Thus highly explicit sexual material will likely have more of an impact than material which is less sexually arousing. The evidence does not indicate, moreover, that "learned" cultural mores and social attitudes have no effect on preventing rape; rather, those factors may play a significant role in mediating the negative biological forces that push men toward rape.

         

      2. Effects on Attitudes Toward Rape - "Disinhibition"

        If arousal to rape is mediated by learned attitudes, however, a change in those attitudes may in itself change the likelihood of rape occurring-may become a "cause" of sexual violence.[47]

        Thus it is crucial to consider what the available experimental evidence shows about the effects of viewing nonviolent sexually explicit materials on attitudes toward women and toward rape. Although Professor Neil Malamuth and others have examined in some depth that question with regard to sexually violent materials, only very recently has substantial evidence emerged about materials which are similar to much of what is contained in the "adult magazines" examined by Baron and Strauss.

        Despite some surface tension in the results, that evidence strongly suggests that such materials, when viewed in substantial quantities over extended periods of time, tend to increase callousness toward women and acceptance of "rape myths". Thus six hours of viewing "commonly available (nonviolent)pornography" over a six-week period caused men in several experiments to become more accepting of "gender dominance"[48] and "sex callousness"-to trivialize rape, and to discount the trauma suffered by its victims.[49] The careful and extensive study by Professor James Check found repeated exposure to the "most prevalent" form of nonviolent pornography currently available-that depicting the women subjects in a "dehumanized fashion"-had even stronger effects on subjects' "reported likelihood of rape" and "reported likelihood of forced sex acts," than sexually violent materials.[50] Both types of material had particularly profound effects, it is important to note, on those subjects with higher tendencies toward psychoticism.[51] Exposure to "nonviolent erotica"-described as being the type of depiction used in sex education and therapy materials-was found to have at best an ambivalent effect: likelihood-to-rape scores increased among those viewers to a level where they were not significantly different from either those in the "no exposure" or the "dehumanizing pornography" groups.[52]

        Only one study currently extant seems to cast doubt on the tendency of viewing nonviolent pornography to increase "rape myth acceptance:" In a recent doctoral dissertation Daniel Linz found that exposure of university psychology students to either two or five full-length X-rated nonviolent films over, respectively, a three- or ten-day period did not affect their attitudes toward a rapist or his victim in a simulated rape trial shown two days after exposure was completed.[53] Such attitudes were dramatically affected, by contrast, in a comparison group observing four extremely violent R-rated films with far less sexual content. Unfortunately, Linz' study is not directly comparable with previous ones in this area. First, Linz limited the time frame of exposure to less than two weeks.[54] Second, his study did not measure the subjects' scores on "likelihood-to-rape" or "likelihood-of-forced-sex-acts" scales similar to those used by Professor Check but rather studied subjects' reactions to a simulated rape trial. Reaction to the plight of a specific rape victim in a simulation is not as direct-and so at least arguably not as useful-a measure as answers to questions about what the subject himself desires to do. Because his study did not include, as did Check's, comparisons based on his subjects' prior viewing habits, Linz' results must be treated with extreme caution. It is possible that the strong reaction to R-rated violent films was simply a function of low prior exposure to those films-the films may have their effects because of "shock value."[55] (College-age participants in studies of this nature are known, by contrast, to have previously seen large quantities of "commercialized erotica" and so would not likely have been as jarred by seeing more of it.)[56] The study did not measure the effects of X-rated violent films, which would have served to indicate the role of sexual explicitness in mediating the effects of viewing violence.

        Despite its methodological limitations, the Linz dissertation does contribute one highly important finding to the data on nonviolent material. In a follow-up study of the participants in his experiment Linz conducted careful "debriefing" of all subjects with regard to the specific material each had seen, then measured their attitudes toward rape after a six-month period. For those subjects who had seen, then been "debriefed" regarding R-rated violent and R-rated nonviolent materials, a dramatic reduction in "rape myth acceptance" occurred-with virtually no difference between those two groups in their final scores. "Debriefing" was thus seen as a success for both groups. Subjects who had seen X-rated nonviolent materials, by contrast, showed only the most minimal decline in "rape myth acceptance" after "debriefing" the lapse of six months-so that at the point of follow-up measurement they showed substantially higher toleration of rape than either of the R-rated groups.[57] The significance of this finding, not recognized by Linz himself, is its tendency to show long-term effects of "X-rated" material even in the face of positive efforts to "educate" viewers. In the "real world", as opposed to the laboratory, viewers of sexually explicit materials normally receive messages-" inhibitory cues"-contradicting those in the materials they watch. The Linz study provides tentative evidence that for sexual materials with a high degree of explicitness, such real-life "debriefing" may be unsuccessful.

        The overall results of work on "long-term" exposure to standard, nonviolent pornography was confirmed and summarized in a statement by Professor Donnerstein in 1983:

         

        Let me end up talking in the last couple of minutes, about the long term research. Researchers like myself and Neil Malamuth at UCLA are looking at massive long term exposure to this material. Some interesting things occur. If you expose male subjects to six weeks' worth of standard hard-core pornography which does not contain overtly physical violence in it, you find changes in attitudes toward women. They become more calloused towards women. You find a trivialization towards rape which means after six weeks of exposure, male subjects are less likely to convict for a rape, less likely to give a harsh sentence to a rapist if in fact convicted.[58]

         

        Professor Donnerstein went on to say:

        In our own research we are looking at the same thing. Let me point out one thing. We use in our research very normal people. I keep stressing that because it is very, very important. What we are doing is exposing hundreds and hundreds of males and now females to a six-week diet of sexually violent films, R-rated or X-rated or explicit X-rated films. We preselect these people on a number of tests to make sure they are not hostile, anxious or psychotic.

        Let me point out the National Institute of Mental Health and the National Science Foundation and our own subjects committee will not allow us to take hostile males and expose them to this type of material because of the risk to the community. They obviously know something some of us do not.[59]

        Although Professor Donnerstein himself has recently emphasized most the harmful effects of violent depictions, the research strongly seems to support the proposition that longer-term, substantial exposure to "standard" nonviolent, sexually explicit materials acts as a "disinhibiting cue" for rape.

         

      3. Overall Evidence for "Causation"

        No experiment has, for the reasons suggested by Professor Donnerstein, tested the effects of nonviolent, sexually explicit material on the aggressive behavior of known sex offenders or, indeed, those with even a tendency toward psychoticism. Experiments with "normal" subjects, however, have suggested two separate, but quite possibly interdependent means by which such material could heighten the probability of sexual violence. The simple capacity of nonviolent material to produce strong arousal in both offenders and the general population may in and of itself produce higher levels of sexual violence. Of equal importance, "standard" commercial pornography may over time and with significant exposure work to undermine "learned" inhibitions against sexual violence. While "adult men's magazines" have not been the normal focus of experimental investigation, the material they contain is sufficiently arousing, and sufficiently tied to views of women only as "sexual objects;" as to make the reasonable inference that these findings are applicable to them as a class. Thus the Badgley Committee in Canada found that in a group of "adult" magazines essentially the same as those studied by Baron and Strauss, photographic depictions of sexual bondage were three times as frequent as oral-genital contact, five times as frequent as vaginal penetration with penis or finger, and ten percent more frequent even than any form of kissing.[60] While further research is clearly indicated to determine the effects of this extremely common material, at present it may fairly be seen as falling within the range of materials as to which current experimental and clinical evidence is highly relevant.[61]

       

    4. Evidence Against Causation

      Studies of both arousal and attitudinal effects of viewing nonviolent materials thus provide several suggestive "causal" links between such viewing and sexual violence. What is the evidence against such a connection? If substantial enough, such data might preclude forming any opinion about the plausibility of the causal link suggested by the correlational data, in combination with indirect experimental and clinical data.

      Unfortunately evidence which contraindicates the existence of a cause-and-effect relationship between nonviolent materials and sexual violence is slim. Short-term exposure of normal subjects to "mild erotica" has been shown to have negligible (and in some cases positive) effects on aggressive responses toward women in the laboratory.[62] As discussed above, results of short-term exposure to highly arousing material have been to the contrary, with enhancement of aggression occurring in cases with "prior anger."[63] Long-term exposure, however, which seems the condition most likely to resemble actual behavior, seems clearly to disinhibit subjects regarding sexual violence. And of course, the reaction of paraphiliacs even to brief exposure to "mild erotica" is far from clearly negligible; if anything, the studies point toward some use of such material by sex offenders to initiate and maintain the deviant fantasies which help push them toward more offending behavior.[64]

      Nor is there substantial evidence showing beneficial effects of "standard" nonviolent pornography. It is crucial to note that when asked whether exposure to pornographic materials can ever reduce commission of sex crimes by paraphiliacs over the long term, Dr. Abel responded with a flat denia1.[65] The Fraser Committee found, on a more general level, "there is no research documenting the beneficial effects of pornography," a proposition that is somewhat misleading but generally true. In sex therapy and sex education settings, research by Dr. Abel's and others suggest that such material may be useful, and the work of Professor Check, discussed above, indicates that materials which are overtly educational or therapeutic may be substantially "harmless" even when viewed outside a controlled environment. Studies for the 1970 Commission found that some sexual materials helped ease sexual tension and promote "liberal" attitudes toward sexuality-a result that may be seen as "beneficial" according to one's basic assumptions regarding sexual morality. Yet with regard to strongly arousing, nonviolent materials, both Dr. Abel's judgment concerning sex offenders and the Fraser Committee's findings about the general population seem well founded.

  3. Conclusion

     

    Ultimately the empirical evidence suggests the following conclusions: viewing nonviolent, sexually explicit material similar to widely circulated "adult magazines" is statistically related to a higher probability of rape. (Thus, for example, Wyoming has a "sex-magazine circulation rate" 45 percent higher than Montana's, with a rape rate 57 percent higher. Baron & Strauss (1985).) That relationship is not only highly significant, and constant from year to year, but it is not "spurious" when other potential "third factors" are considered. Evidence from both experimental and clinical studies demonstrates at least two possible ways in which that correlation might be explained by "causation": (1) through the simple arousal properties of such materials, and (2) through their disinhibiting qualities, their capacity to change attitudes regarding sexual aggression. The evidence is nonetheless far from conclusive, and points toward the need for substantially more, and better-focused research. At this point, little or no evidence exists which shows any beneficial effects of such materials.

    It is useful to consider the weight of this data against that which supports our previous finding that sexually violent material is causally related to sexual violence. For that conclusion we had no correlational evidence demonstrating a "real-world" statistical relationship between the material and the behavior. By contrast, the experimental evidence was somewhat stronger-showing, for example, "negative effects" from short-term as well as long-term exposure. Sexually violent material is no more arousing to viewers (even to known rapists) than is "standard" nonviolent material (Abel, Barlow, Blanchard & Guild (1977). In the one study which directly attempted to compare the effects on attitudes of sexually violent material with effects from "dehumanizing" material and "erotica," the results showed no significant difference in the most crucial areas.[67] Only a well-founded intuition that direct depictions of sexual violence are more likely to produce such violence allows us to conclude that they are more "harmful" than nonviolent materials; the evidence from social science is at best ambivalent on the issue.[68] Our task is not an easy one, because with widely different backgrounds and substantially different ideas about what constitutes "proof" of a given fact, we are highly unlikely to reach consensus on highly disputed questions. With regard to the relationship between sexually explicit materials and sexual violence we will each carry away different levels of skepticism about the state of currently available evidence. And we will know, too, that our stated conclusions may be swept away by new research. Yet that does not relieve us of the obligation to state, not as scientists proclaiming "fact" but as policymakers confronting risk and probability, that wide circulation and consumption of materials similar to "adult men's magazines" must be a matter of concern among those seeking to combat sexual violence. There is at least a substantial basis, if not a preponderance of the evidence, to believe that such materials are a part (if only a small part) of the explanation for that cruel plague.

Acknowledgement. I am deeply grateful to Dr. Edna Einsiedel, The Commission's staff social scientist, for her review of, and comments on, the preliminary versions of this statement. The foregoing represents, however, only my own views and not necessarily hers.

NOTE: All references in the text and notes are to studies cited in the Report on Social Services of the Commission, except where a full citation is given.

Notes

 

In their joint statement Commissioners Becker and Levine attempt to discount the importance of this correlational evidence by pointing to a letter from one of the researchers involved, Murray Strauss, which states (1) the correlational research does not "demonstrate" that pornography causes rape;" and (2) "the scientific evidence clearly indicates that the problem lies in the prevalence of violence in the media, not on sex in the media." Id., p. 13. Strauss' first statement is uncontestable: no correlation can, by itself, "demonstrate" causation. Strauss' concern about "misinterpretation" of his research seems somewhat bizarre in view of his published statement that his "findings suggest that the combination of a society that is characterized by a struggle to secure equal rights for women, by a high readership of sex magazines that depict women in ways that may legitimate violence, and by a context in which there is a high level of nonsexual violence, constitutes a mix of societal characteristics that precipitates rape." Baron & Strauss (1984), at 207. He then intimates that research suggests "social policies directed toward eliminating or mitigating the conditions that make rape more likely to occur." Id. It is Strauss, not the Commission, who has made suggestions of causal linkage based on correlational data alone. See also text to note 23.

With regard to his second observation, that violence in the media seems to be "the problem" rather than sex, the research is very far from "clearly" indicating any such thing. Thus it has been found that with regard to same-sex interactions, nonviolent but highly arousing erotic material facilitates aggression substantially more than "violent" material. Donnerstein (1983b). And when, angered males are shown a nonviolent, "erotic" film, then allowed a short delay before testing, their aggressive behavior toward women has been shown to increase dramatically, to levels far higher than for similarly treated subjects shown violent or neutral films. Donnerstein & Hallam (1978). The "delay" factor seems crucial, as measurements of aggression toward women taken immediately after film exposure tend to suggest that "erotic" material does not increase aggression. Donnerstein (19836); Donnerstein & Berkowitz (1981). This "delayed reaction" effect is similar to that found by Zillman & Bryant (1982, 1984, 1985), in which "massive exposure" to nonviolent, degrading pornography over six weeks produced dramatic increases in subjects' acceptance of "rape myths" and "sex callousness." (By contrast Linz (1985) did not find such effects after a substantially shorter exposure period.) Obviously this experimental data is still at a primitive stage, but it hardly warrants the interpretation Strauss gives it.

  1. 1970 Commission Report, at 287. See, Fraser Report, p. 99; Williams Report, p. 6186.
  2. 1970 Commission Report, pp. 286-87.
  3. For a review of many of those criticisms see Donnerstein & Malamuth (1984).
  4. 1985 Newsweek Poll. Forty-seven percent of respondents would ban magazines showing adults having sexual relations, but only 21 percent favored such a ban for magazines depicting "nudity". Because many current popular magazines are clearly "degrading" in their portrayals, the difference in views seems more related to sexual explicitness than to the positive or negative portrayal of the person depicted.
  5. Id.
  6. See, e.g., Linz (1985) (excluding subjects from experiment if "psychoticism" or "hostility" score exceeded 1.0 on Symptom-checklist 90); Check (1985).
  7. Sourcebook of Criminal Justice Statistics, (1984), p. 380, (hereinafter Sourcebook).
  8. Id. The high point of both general violent crime rates and reported forcible rapes came in 1980, the former having risen 60 percent and the latter 95 percent from 1970 levels. From 1980 to 1983 the rate of all violent crime fell 9 percent, while reported forcible rape rates dropped by 7.5 percent. Id.
  9. Id.
  10. Id.
  11. Rapid social change associated with "women's liberation" may also be viewed, of course, as making rape itself more likely-through setting up more possibilities of "acquaintance rape". See Geis & Geis, Rape in Stockholm: Is Permissiveness Relevant? 17 Criminology, (1979), p. 311. Women raped by "friends" may be less willing to involve criminal sanctions against their attackers. Thus it is at least arguable that "women's liberation" may in some respects have had a dampening effect on rape reporting rates.
  12. National Crime Survey figures indicate that no better than half of all rapes are reported. Sourcebook, supra note 6, pp. 274-275.
  13. Between 1973 and 1977 an average of 46.2 percent of all rapes went unreported according to the Survey; between 1978 and 1982 the average percentage of unreported rapes stood p. 48.2. Id.
  14. Between 1973 and 1977 the average estimated number of actual rapes per year was 152,877; between 1978 and 1982 the average stood at 173,353, an increase of 13 percent. Id.
  15. D. Russell, Sexual Exploitation, (1984), pp. 52-57. Professor Russell's survey was conducted in 1978, and so is of little value for determining recent trends in rape reporting. It does attest, however, to the fact that, historically, upward trends in police reports of rape have been consistent with actual incidence of the crime.
  16. Baron and Strauss (1984), for example, found that every change of one standard deviation in the Status of Women Index in a given state is associated with a change in the rape rate of only 0.43 rapes per 100,000 population. By contrast, such a change in the homicide rate would result in a swing of 1.70 rapes, and a one-standard-deviation change in the Sex Magazine Circulation Index would cause a swing of 6.99 rapes (the highest of any variable studied). Id., p. 200.
  17. 17. Kutchinsky (1984), pp. 24-25. Kutchinsky attempts to limit the damage of this concession by noting that the increase in rape reports did not substantially begin until 1977, several years after liberalization. He is not, however, able to rule out the possibility that Danish consumption of pornography took some time after legalization to reach substantial proportions.
  18. F.D. Scott, Scandinavia, (1975), p. 247.
  19. See, General Civil Penal Code of 22 May, 1902, Para. 211, as amended by Law of 24, May, 1985 (received in translated form from Jan Farberg, Norwegian Information Service).
  20. According to the Public Information Office of Interpol the rate of reported sexual offenses in Denmark dropped 14.2 percent from 1970 to 1981. In West Germany, another country with liberal obscenity laws used by Kutchinsky in support of his argument, the rate dropped 19.8 percent during that span. In Norway, however, the drop was 33.7 percent in reported sex offenses from 1970 to 1981. These figures are not necessarily computed in the same manner from country to country and should thus be considered only with extreme caution. Nevertheless they do suggest the grave problems in Kutchinsky's selective use of sex-crime figures from one or two locations unembarrassed by historical or cross-cultural analysis.
  21. See note 16, supra.
  22. Scott (1985a). In another study Scott (1985b) found that no significant statistical relationship existed between rape rates in the states and the number of "adult theaters" per 100,000 residents in each state. That finding, however, is of almost no value on several grounds: (1) the study did not use multiple regression analysis to examine possible interdependence of the variables; (2) the number of "adult theatres" is an almost completely meaningless figure in view of the fact that each such theatre will sell a different quantity of sexually explicit materials, and no account is taken of that variation; and (3) "adult theatres" are so restricted by zoning, obscenity laws, and the need for urban or semi-urban locations that they cannot be assumed to measure exposure to sexually explicit materials among males who can, if necessary, purchase such materials through the mail.
  23. Jaffee & Strauss (in press) p. 10. Rodney Stark, in Demonstrating Sociology (1985), has claimed to disprove the Baron and Strauss correlation, at least with respect to Playboy's circulation rates. Id., pp. 29-31. Because Stark's discussion of the issue is openly informal, and because the Baron and Strauss results have been replicated formally by others, Stark's view is not persuasive. See, Koss (1986) (in large sample of college students there existed a statistically significant relationship between prior consumption of pornography and self-reported sexual aggression).
  24. Abel (1985), p. 5.
  25. Dr. Abel has been asked to furnish the exact "p value" for this and other comparisons in his written testimony. For our purposes the appropriate level of "significance" in a matter such as this might be substantially different from that typically used in the social sciences. There a statistical difference between two groups is normally not described as "significant" unless there exists 95 percent probability that it did not occur by chance. The probability level appropriate for our use-which, after all, is only to determine whether a "substantial basis" far a finding exists-might be as low as 70 percent.
  26. Houston Tr. 100. Earlier Dr. Abel has said the use of erotica by sex offenders "maintains their arousal over time, and therefore greater opportunities to commit further crimes occur." Id., p. 88.
  27. Because of his limitation of his study to the role of "hard-core pornography" (not including the typical "adult magazines" referred to by Dr. Abel in his study) Dr. Marshall's results are in no sense directly comparable to those of Dr. Abel. He does, however, find a pattern of pornography being used so integrally in preparation for and commission of sex offenses as to make his evidence highly pertinent.
  28. Id., pp. 97, 100.
  29. Malamuth & Spinner (1979) (sexually violent content in Playboy and Penthouse from 1973 and 1977, amounted to less than 10 percent of total cartoon and pictorial content).
  30. Thus Gross (1983) has criticized the research of Zillman and Bryant (1982) because he suspects the subjects "were giving the researchers what they thought they wanted." Id. at III. This, despite the elaborate efforts of the researchers to deceive the subjects into believing that they were most interested in aesthetic qualities of materials viewed, rather than their efforts on attitudes. Unfortunately Gross' criticism may be applicable to virtually any experiment in this area, or indeed in other areas of inquiry. And he is unable to suggest any way to surmount the artificiality inherent in laboratory experiments.
  31. See e.g., Donnerstein (1980); 1970 Commission Report pp. 198-241.
  32. Donnerstein (1984); Donnerstein (1983b); Sapolsky (1984).
  33. Donnerstein (1984), p. 62.
  34. . Id. Compare Check (1985) with Linz (1984). For further discussion of varying research results see, supra note 22.
  35. See, Saplosky, (1984), p. 92; Wolchik, Braver & Jensen (1985).
  36. Barbaree, Marshall & Lanthier, (1978); Abel, Recker & Skinner, (1980).
  37. Marshall Statement p. 23.
  38. Id.
  39. Id.
  40. Marshall (1984); Abel, Roulean and Cunningham-Rathmer (1985).
  41. Abel, Blanchard & Jackson (1974); Marshall (1973); Marquis (1970); Davidson (1968).
  42. Abel, Blanchard & Jackson (1974), p. 474.
  43. Abel, Becker & Skinner (1980), p. 183. See e.g., Saplosky (1984).
  44. Malamuth, Feshback & Jaffe (1977); Donnerstein, Donnerstein & Evans (1975).
  45. Rape statistics, of course, measure only the number of such acts, and the "rate" of such acts for a constant population group. They do not, and cannot, measure rape as a percentage of all sexual behavior.
  46. Some general support for this hypothesis may be found in the fact that as rape dramatically increased in incidence in post-war America, so did sexual activity among the young-the age group most prone to sexual violence. Thus only about one-half of males 21 years of younger had engaged in sexual intercourse at the time of the first Kinsey study. A. Kinsey, et al., Sexual Behavior in the Human Male 316, while currently over 90 percent of boys appear to have begun such activity by age 17. R. Coles & Stokes, Sex and the American Teenager 73 (1985) (The Coles & Stokes sure is somewhat ambiguous on this point; in another table the percent of 18 year olds "having had intercourse" is listed at 46 percent. Id., p. 73. In any case the trend toward earlier and greater sexual involvement is clear, for in Kinsey's survey only some 31 percent of all 18 year-old males had experienced sexual intercourse. Kinsey, supra, p. 316.
  47. See generally, Malamuth (1984).
  48. Zillman & Bryant (1985b).
  49. Check (1985), Zillman & Bryant (1982, 1984); Donnerstein (1984).
  50. Check (1985), p. 49.
  51. Id., p. 53. Indeed, subjects with "low P" scores were not significantly affected by any of the sexually explicit materials, a finding which may call into question flat conclusions about the effects of pornography independent of the specific vulnerability of individual subjects, and which supports the role of a well-developed moral sense in mediating the effects of exposure.
  52. Id. pp. 49, 53. It is notable that on the three measures of sexual violence in which no-exposure and "violent pornography" scores were significantly different, the "erotica" scores were slightly closer to those of the latter. Professor Check thus seems to have overstated the importance of his findings that "erotica" and "no exposure" scores were not "statistically significant".
  53. Linz (1985).
  54. Zillman and Bryant (1982, 1984, 1985), by contrast, used a six-week exposure model. Check (1985) used a time frame similar to Linz, but tested for prior consumption of pornography-finding that only those viewers with high previous consumption were affected by exposure to new materials. Thus the negative findings of Linz may well have to do with low prior exposure to pornography among his subjects-precluding, in the short time used, development of the effects of long-term exposure. See, infra text to note 57.
  55. See, Zillman, Bryant & Carveth (1981) (viewing bestiality increased aggression due to "annoyance summation"). The shock value explanation for the Linz data is strengthened by the fact that later "debriefing" treatments over a six-month period seemed completely to reverse the effects of viewing these materials. Linz, p. 96.
  56. Wolchik, Beaver & Jensen (1983).
  57. Linz (1985), pp. 96-98.
  58. Public hrgs. on Ordinances to Add Pornography as Discrimination Against Women, Minneapolis City Council, Sess. I, p. 31 (Dec. 12, 1983).
  59. Id. at 32.
  60. Badgley Report p. 1223. Of course graphic depictions of genitalia of nude models in such magazines-often with pubic hair shaved-serves as well to reduce those shown to the status of "sexual objects". This general description of magazines evaluated by Baron and Strauss and others should not be taken as specific to any one of them. Individual differences in format, and style and content may be crucial.
  61. Thus Abel (1985) focused on such material in his study of sex offenders. As discussed above, supra text to notes 21-23, Dr. Abel's findings are ambivalent but troubling.
  62. See Donnerstein (1984, 1980A).
  63. Donnerstein & Hallam (1978).
  64. For a discussion of the evidence on sex offenders presented by Dr. Abel and Dr. Marshall, see, supra text to notes 21-24,34-40.
  65. Houston Tr., p. 100.
  66. Fraser Report p. 98.
  67. Check (1985). Indeed, Check found that on many measures sexually violent materials produced less "negative effects" than "dehumanizing pornography"-although not by "significant" margins. "Erotica," of course, was also found not to be "significantly" different in its effects than "no exposure:" See, supra note 52.
  68. It is useful, as well, to compare the strength of our conclusions in this area with those of the Advisory Committee to the Surgeon General in an area which was at the time similarly contentious and difficult-the health risks of cigarette smoking. The evidence relied on for the Committee's conclusion was overwhelmingly correlational-showing higher death and illness rates among smokers than in non-smokers. The Committee recognized fully that correlational evidence did not show causality and looked to animal experiments, clinical data, and "population studies" (i.e., retrospective studies of smokers vs. control groups). Surg. Gen'l of the Pub. Health Serv, U. S. Dept. of H.E.W., (1964), pp. 26-27. With regard to lung cancer, those additional forms of evidence were sufficiently supportive of the correlational data to allow the Committee to conclude that "cigarette smoking is causally related to lung cancer in men"; with regard to women the data allowed the lesser conclusion that the data "point in the same direction." Id. p. 31. As for heart diseases, the Committee found that there existed a strong correlation between coronary disease and smoking, but found that the current explanations for causation from experimental and other evidence "do not account well for the observed association". Id. p. 327. Instead of throwing up its hands in the face of difficult and conflicting evidence the Committee said simply: "It is . . . more prudent to assume that the established association between cigarette smoking and coronary disease has causative meaning than to suspend judgement until no uncertainty remains:" Id.

     

    It would be presumptuous to compare the quantity of evidence before us with that reviewed by the Surgeon General's Committee; research on "pornography" is still in its infancy. But our responsibility to be as prudent as possible is the same, and the correlational evidence before us combined with at least a substantial strain of experimental and clinical data make it prudent to advise the public of the risks of the materials for which statistical data do exist.