Archives of Sexual Behavior, VoL 7, No. 2, 1978
Stimuli Eliciting Sexual Arousal in Males Who Offend Adult Women: An Experimental Study Ale[ Koki~sk#, Ph.D., x Jaroslav Madlafousek, Ph.D., 1 and Vladimfra Novotmi, M.D.1
The sexually arousing effects o f short film scenes showing a naked actress's seductive behavior were phalloplethysmographically measured in 14 sexual deviates. These were males who had offended adult women, predominantly exhibitionists. Controls were 14 normal men. Deviates responded positively to the scenes and differentiated strong and weak seduction scenes similarly to normals. Consequently, the question arises o f why deviates avoid their victim's erotic cooperation and why they do not offend their regular sexual partners. Post hoc analysis o f five scenes which elicited a strikingly higher response in deviates than in normals suggested that these scenes contained reduced seductive behavior but unrestrained presentation o f the genitals. This finding further encourages the laboratory study o f stimulus conditions for abnormal sexual arousal which occurs during the sexual offense. KEY WORDS: sex offender; exhibitionism ; sexual assault; penile response.
INTRODUCTION Clinicians acquainted with a large number of case histories of male sex offenders against adult women have noted an interesting anamnestic report that if the victim's reaction to the offense is that of true erotic cooperation (e.g., "seduction"), the offender may lose interest in continuation and culmination of the act. Researchers point out that exhibitionists avoid exposure to females who might cooperate sexually (Gebhard et al., 1965), that exhibffionists are afraid of f u r t h e r contact with the victim (Mohr et al., 1964), and that these individuals seek unknown women as victims (Gebhard et al., 1965; Macdonald, 1973; Mohr 1Psychiatric Research Institute, Prague 8, Czechoslovakia. 79
0004-0002/78/0300-0079$05.00.©1978 P l e n u m
Publishing Corporation
80
Kolfi~sk~r,Madlafousek, and Novotnd
et aL, 1964). In our experience, some other sex offenders who seek to reach in-
timate bodily contact during the offense also avoid the victim's erotic cooperation (offensive touchers and aggressors). The offender's desire in a state of sexual excitement for female erotic uncooperativeness sharply contrasts with the desire of a sexually excited normal man and, therefore, can be qualified as a symptom of sexual deviation, i.e., of a disturbance in the program for sexual behavior. For this reason we call sexual offenders who show this symptom "deviates who offend adult women." We propose that the above symptom reflects an essential disturbance common to different kinds of sex deviates. The idea of this commonness is supported by relatively frequent cases of deviates who alternately practice different kinds of offensive acts, e.g., indecent exposure and offensive touching. An important clinical and research question is whether deviates who offend adult women are sexually arousable by normal female erotic behavior. Are they avoiding the victim's erotic cooperation because they are aversive toward all or some sexual behavioral stimuli? Although some deviates have a regular sexual partner, Garrett and Wright (1975) found that the relationship between the offender and his wife is atypical, even before the wife has learned about her husband's offenses. The existence of quasinormal sexual partnership does not provide good evidence of the deviate's capacity to become sexually aroused by normal female sexual behavior in thebroadest sense, i.e. including behavior exhibited in the introductory (preparatory) phases of sexual interaction (seduction). Therefore, and also because the deviate's verbal reports need not be reliable, the phallometric (phm) method has been employed for the assessment of the deviate's response to systematically applied sexual stimuli. In the first study (Kol~sk~ and Madlafousek, 1972), comparison was made o f the arousing effect of Film scenes showing a motionless, neutrally posing actress with that of scenes showing a "courting" actress and with the arousing effect of the actress's engagement in nonerotic activity. The difference between deviates and normal men was found with respect to the effect of the actress's nonerotic activity. Whereas normals were sexually inhibited by it, deviates were not. On the other hand, the actress's courting behavior elicited a higher phm response than the motionless scenes in both normal and deviant subjects. The latter finding thus made less acceptable the most simple explanation that the deviate seeks to perform intimate acts (exposure, touching) against an erotically unprepared and uncooperative woman because female courtship behavior under all circumstances inhibits him sexually. To further elucidate the problem of sexually arousing stimuli in deviates, the present study was designed to compare the phm responses of normals and deviates to a number of film scenes showing another actress, this time nude, exhibiting behavior with different kinds of "seductive" cues.
Sexual Arousal in Male Offenders
81 METHOD
Subjects The deviant sample was composed of 14 subjects. Half were new patients referred to the authors, the other half were volunteers from those previously examined by the authors. The inclusion of each subject in the deviant sample depended on whether the diagnosis of sexual deviation was substantiated by the interview (Kolfi~sk3~, 1968), predominantly on the evidence that normal (inoffensive) sexual behavior was either absent or insufficient for providing the subject with full sexual satisfaction. The preferred sexual object of all was an adult woman. Ten deviates were exhibitionists with no known history of other offenses, and four had a history of offensive touches ("sexual aggression") without rape. Eight deviates reported having also experienced normal sexual intercourse. The mean age in the deviant sample was 27.9 years, the range from 20 to 51. The control sample of nondeviant men consisted of 14 paid volunteering soldiers and employees of the hospital in whom verbal exploration provided no evidence of the presence of sexual deviation. The fifteenth subject was exctuded because of his report of partial erectile failure. The remaining 14 subjects were sexually successful and active males. Their mean age was 21.7 years, the range from 19 to 30 years. All subjects gave informed consent for the examination, after its nature had been fully explained to them.
Procedure and Measurement The phalloplethysmograph of Freund et al. (1965) was used for the purpose of measuring penile volume changes. Black and white silent film scenes were projected as stimuli. The duration of each was 10 sec. The dimensions of the pictures were 120 by 197 cm, and they were viewed by the subjects from a distance of 6 m. For other details, see the earlier report (Kol~sk2) and Madlafousek, 1972). Because the acceleration rate of penile volume increase usually culminated within 10 sec after the termination of the stimulus, the arousing effect of each stimulus was expressed as a difference between the penile volume at the tenth second after its end and the volume at its onset. Care was taken to dissipate sexual arousal (to restore the prestimulus level of penile volume)before presenting the next stimulus. The stimuli were presented only after the penile volume became stabilized. The first stimulus included in the design was preceded by two preparatory stimulus scenes.
82
Kohffsk:#,Madlafousek,and Novotni Stimuli and Design
The stimuli were film scenes showing the same young woman either naked during the entire scene (14 scenes) or reaching full nudity in its course. It was intended that all 18 scenes represented different kinds of possible seductive behavior, such as chastity or provocative display of the body. The choice of a particular behavior also depended on the actress. The strict limitation, however, was that behavior should not be coital or postcoital, but precoital. The actress was either standing or sitting, never lying down. The front side of her body could be more or less visible in all scenes except for one, in which the standing actress presented her swinging movements from behind. The visibility of particular body zones usually varied during the course of the scenes. (One purpose of seduction is to prevent habituation from constant viewing of the same body zone.) For further description of scenes, see the Results section. The 18 "seductive" scenes were presented to each subject in the same ran. domized order. Before the second half of the scenes was presented, each subject saw several other ("nonseductive") scenes with the same actress. These were "0" (the motionless face), "T" (the motionless torso from the frontal view), ' T ' (the whole actress dressed in a swimming suit, standing and being nonemotional), and " F " (lying on her back, naked, simulating high sexual arousal, being prepared for coitus). The purpose of applying these scenes was to test the effect of the supposed weaker (O,T,I) and stronger (F) stimuli as well as to separate the two halves of "seductive" scenes for statistical control of long-term changes in responsiveness. During this nonseductive interlude, several other scenes, which are not reported here, were presented for piloting purposes. Because the empirical data supported the assumption that deviates do not differ from normals as far as general sexual arousability is concerned, a direct comparison of the arousing efficiency of the scenes was carried out. The statistical analysis was made with logarithmically transformed data, because of their tendency to nonnormal distribution. Measurements in the phm experiments are characterized by considerable variability. This must be taken into account when interpreting nonsignificant results.
RESULTS
As expected, normal subjects responded more strongly to the coital scene (F), less to the "seductive" scenes (Sed. 1, Sed. 2), and least to the supposed weak scenes (O,T,I). This was also true for the deviates. The results (see Fig. 1) seem to substantiate the assumption that there is no difference in the general sexual arousability of normals and deviates. The observed differences between normals and deviates in responses to scenes O, T, I, and F were not significant (t test).
Sexual Arousal in Male Offenders
83
1.8 1.7 u-x-..=
1.6
NORHALS
e.-e--e DEVIATES
1.5 1.4 1.3 1.2 1.1
1.0 SED.1
0
T
I
F
SED.2
Fig. 1. Phm responses of normals and deviates. Ordinate, group mean responses (log) to weak (O,T,I) and strong coital (F) scenes and means from mean responses (log) to the first (SED. 1) and second (SED. 2) halves of "seductive" scenes.
The effect of "seductive" scenes was also similar in both groups as reveaied by a two-way analysis of variance: the main effect of subject group was not significant ( F = 0.136, df = 1, 26). Further results of this analysis o f variance were as follows: No change in responsiveness could be demonstrated from comparison of the effect of the first and second half o f "seductive" scenes ( F = 0.066, df = 1, 16). The interaction of subject group × "seductive" scenes halves was also not significant (F = 0.048, df = 1,442). There was a significant main effect of particular "seductive" scenes, irrespective of halves 1 and 2 ( F = 1.905, d f = 16,442, p < 0.025). This indicates that various kinds o f behavior in the "seductive" scenes had a differential arousing effect, as intended, and that the phm method is capable of reflecting it. An important finding here is that deviates and normals were differentially affected by particular "seductive" scenes, as evidenced by the significant interaction of subject group × particular "seductive" scenes (F = 1.873,!df = 17, 442, p < 0.025). Figure 2 shows that with respect to five scenes deviates strikingly differed from normals. The five scenes mentioned elicited a stronger phm response in deviates than in normals. They tend to belong to those less arousing within the normals' rank order. They will be referred to as "deviant" scenes. With the least arousing scene for normals the difference between both groups reached the 5% level of
84
Kohi~sk}', Madlafousek, and Novotmi ®
120 115 ® 0
110
105
0 0
0 0
O0 Q
•
100 O
0
95 I
|
!
!
1.4
1.5
1,6
1.7
Fig. 2. Comparison of phm responses of normals and deviates to 18 "seductive" scenes. Abscissa, group mean responses of normals (log); ordinate, mean responses of deviates in terms of percentage of the normals' mean responses.
significance, according to the Dunnett's test (1964) (t = 3.014), which takes into account the number of all possible comparisons and thus estimates the t value conservatively. Using the classical method for estimation of t value, the 5% level of significance for a two-tailed test is reached with three scenes (those with the highest difference between both groups). Preliminary observations of scenes, made by the authors, gave the impression that the most conspicuous difference between the "deviant" and "nondeviant" scenes is that for the former scenes behavior that is labeled as "seductive pseudoretreat" is absent, whereas it is present in "nondeviant" scenes. In an attempt to make the identification of this behavior objective, we have operationally defined "seductive pseudoretreat" as avoidance of the frontal body position with respect to the camera and a tendency to squeeze the thighs together. Three observers, unacquainted with the results, were asked to classify all 18 "seductive" scenes according to (1) whether the naked actress assumed, at least for a moment, a frontal position with respect to the camera ("frontal" scenes), and (2) whether the actress squeezed her thighs together. The observers concordantly classified all five "deviant" scenes as frontal and only two out of 13 "nondeviant" scenes as frontal. As far as squeezing the thighs is concerned, all ob-
Sexual Arousal in Male Offenders
85
servers did not reach agreement with one "deviant" scene as well as with two "nondeviant" scenes. However, all observers agreed that the remaining four "deviant" scenes did not contain squeezing of thighs, whereas this was the case with only two of the remaining 11 "nondeviant" scenes. Although this post hoc analysis is only a preliminary result, it throws doubts on the use of "deviant" scenes from the original design for comparing the phm responses of normals and deviates to seductive scenes. Therefore, this comparison was carried out only with the remaining 13 scenes, the seductive character of which did not appear doubtful (Fig. 3). Deviates were affected by the 13 scenes similarly to normals. Spearman rank order correlation is 0.88. No significant difference between the two groups of subjects was found with any scene. (Responses to the "deviant" scenes, represented by crosses, are also shown in Fig. 3 for comparison.)
X
1,7o" X
X
@ @
1.60.
o X
@ ® ®
®
@ ®
®
150.
@
/.4,0. @
@
r
T
Fig. 3. Agreement o f normals and deviates in discrimination o f 13 "nondeviant" scenes. Abscissa, group mean responses of normals (log); ordinate, group mean responses of deviates (log). For comparison, "deviant" scenes are added and represented by crosses.
86
Kol~sk#, Madlafousek, and Novotn~i DISCUSSION
As no systematic analysis of human female precoital erotic behavior ("seduction") has so far been carried out, the creation of stimulus scenes in this study had to be based largely on the intuition of both the actress and the authors. The scenes were intended to contain a variety of seductive cues with the assumption that the resulting effects of particular scenes would quantitatively differ. This proved to be the case. Those scenes which elicited highest reactions in normal subjects can be conceived as empirically verified seductive stimuli. The main finding is that with the majority of scenes deviates responded in a similar way as normals did. This is particularly true for verified seductive stimuli. Also, the deviates' discrimination between weak and strong seductive scenes was similar to that of normal subjects. These results fit the previous study (Kol~ffsk~ and Madlafousek, 1972), which also demonstrated the deviates' capacity to become sexually aroused by a (dressed) female in precoital erotic behavior. Thus the present study provides further evidence that the sex offenders studied share with normal men sexual arousability to normal female erotic behavior. We suggest that this is true for all deviates who offend adult women, although the sample of deviates was composed predominantly of exhibitionists and of a small group of offensive touchers. Separating these two subgroups, both were found to differentiate weak and strong seductive scenes in agreement with normal subjects, and in both subgroups the phm response to the "deviant" scenes was higher than for normal subjects. The question arises then why normal female erotic behavior, although sexually arousing for deviates, inhibits their sexual arousal during the offense and why deviates do not usually have the temptation to perform an offensive act against their regular sex partner. (Exhibitionists do not expose themselves against their sex partner. We have repeatedly seen highly dangerous deviates, who had killed unknown women, who were free of any aggression in interaction with their girlfriends.) At present it is difficult to give a satisfactory answer. Animal studies (Hinde, 1959, 1970) as well as indirect evidence concerning man (Kol~i~sk!) and Madlafousek, 1977) indicate that sexual behavior must not be conceived of as a product of a unitary sexual drive, but rather from the point of view of partial motivational states. The pathological dissociation of these states is, in our opinion, a primary condition which leads to intimate acts (exposure, touches) against an unfamiliar and erotically uncooperative woman. Abel et al. (1975) as well as Marshall (1973) have demonstrated the pos• sibility of eliciting in the phm laboratory abnormal sexual arousal in sex offenders against adult women. The deviates' abnormal sexual arousal also seemed to occur in this study, namely as a by-product of some scenes which did not belong in the verified seductive Stimuli. These scenes supposedly lacked some component(s) of normal female erotic behavior. In the earlier study (Kolfi~sk?? and
87
Sexual Arousal in Male Offenders
Madlafousek, 1972) we found that female nonerotic activity, although inhibiting sexual arousal o f normals, does not inhibit that o f deviates. Future phm studies will be aimed at exploration o f stimulus conditions for abnormal sexual arousal in deviates who offend adult women.
ACKNOWLEDGMENT The authors wish to acknowledge the collaboration of others from the Psychiatric Research Institute in Prague. The preparation o f film stimuli was made with the technical assistance of P. Gemergc and K. Knob. Phallometric measurements were made by K. Knob. Computations were carried out b y ing. Drfibek under the supervision of ing. Roth.
REFERENCES Abel, G. G., Blanchard, E. B., Barlow, D. H., and Mavissakedian, M. (1975). Identifying specific erotic cues in sexual deviations by audiotaped descriptions. J. Appl. Behav. Anal. 8: 247-260. Dunnett, C. W. (1964). New tables for multiple comparisons with a control. Biometrics 20: 482-491. Freund, K., Sedlacek, F., and Knob, K. (1965). A simple transducer for mechanical plethysmography of the male genital. Z Exp. Anal Behav. 8:169-170. Garrett, T. B., and Wright, R. (1975). Wives of rapists and incest offenders, or. Sex Res. 11: 149-157. Gebhaxd, P. H., Gagnon, J. H., Pomeroy, W. B., and Christenson, C. V. (1965). Sex Of. fenders: An Analysis of Types, Harper and Row, New York. Hinde, R. A. (1959). Unitary drives.Anita. Behav. 7: 130-141. Hinde, R. A. (1970). Animal Behavior: A Synthesis o f Ethology and Comparative Psychology, McGraw-Hill, New York. Kolinsky, A. (1968). Zamg/ov~n~ verb~lnf explorace p~i zji~:Eov~n~sexu~lnl"deviace u mu~e (Centering verbal exploration during the estimation of male sexued deviation). Cesk. Psyehiat. 64: 167-171. Kol~sk5,, A., and Madlafousek, J. (1972). Female behavior and sexual aroused in heterosexual male deviant offenders: An experimental study. J. Nerv. Ment. Dis. 155: 110-118. Kol~skg, A., and Madlafousek, J. (1977). Variability of stimulus effect in the course of phedlometric testing, Arch. Sex. Behav. 6: 135-141. Macdonald, J. M. (1973). Indecent Exposure, Thomas, Springfield, Ill. Marshall, W. L. (1973). The modification of sexual fantasies: A combined treatment approach to the reduction of deviant sexual behavior. Behav. Res. Ther. 11: 557-564. Mohr, J. W., Turner, R. E., and Jerry, M. B. (1964), Pedophilia and Exhibitionism, University of Toronto Press, Toronto.
Stimuli Eliciting Sexual Arousal in Males Who Offend Adult Women: An Experimental Study Ale[ Koki~sk#, Ph.D., x Jaroslav Madlafousek, Ph.D., 1 and Vladimfra Novotmi, M.D.1
The sexually arousing effects o f short film scenes showing a naked actress's seductive behavior were phalloplethysmographically measured in 14 sexual deviates. These were males who had offended adult women, predominantly exhibitionists. Controls were 14 normal men. Deviates responded positively to the scenes and differentiated strong and weak seduction scenes similarly to normals. Consequently, the question arises o f why deviates avoid their victim's erotic cooperation and why they do not offend their regular sexual partners. Post hoc analysis o f five scenes which elicited a strikingly higher response in deviates than in normals suggested that these scenes contained reduced seductive behavior but unrestrained presentation o f the genitals. This finding further encourages the laboratory study o f stimulus conditions for abnormal sexual arousal which occurs during the sexual offense. KEY WORDS: sex offender; exhibitionism ; sexual assault; penile response.
INTRODUCTION Clinicians acquainted with a large number of case histories of male sex offenders against adult women have noted an interesting anamnestic report that if the victim's reaction to the offense is that of true erotic cooperation (e.g., "seduction"), the offender may lose interest in continuation and culmination of the act. Researchers point out that exhibitionists avoid exposure to females who might cooperate sexually (Gebhard et al., 1965), that exhibffionists are afraid of f u r t h e r contact with the victim (Mohr et al., 1964), and that these individuals seek unknown women as victims (Gebhard et al., 1965; Macdonald, 1973; Mohr 1Psychiatric Research Institute, Prague 8, Czechoslovakia. 79
0004-0002/78/0300-0079$05.00.©1978 P l e n u m
Publishing Corporation
80
Kolfi~sk~r,Madlafousek, and Novotnd
et aL, 1964). In our experience, some other sex offenders who seek to reach in-
timate bodily contact during the offense also avoid the victim's erotic cooperation (offensive touchers and aggressors). The offender's desire in a state of sexual excitement for female erotic uncooperativeness sharply contrasts with the desire of a sexually excited normal man and, therefore, can be qualified as a symptom of sexual deviation, i.e., of a disturbance in the program for sexual behavior. For this reason we call sexual offenders who show this symptom "deviates who offend adult women." We propose that the above symptom reflects an essential disturbance common to different kinds of sex deviates. The idea of this commonness is supported by relatively frequent cases of deviates who alternately practice different kinds of offensive acts, e.g., indecent exposure and offensive touching. An important clinical and research question is whether deviates who offend adult women are sexually arousable by normal female erotic behavior. Are they avoiding the victim's erotic cooperation because they are aversive toward all or some sexual behavioral stimuli? Although some deviates have a regular sexual partner, Garrett and Wright (1975) found that the relationship between the offender and his wife is atypical, even before the wife has learned about her husband's offenses. The existence of quasinormal sexual partnership does not provide good evidence of the deviate's capacity to become sexually aroused by normal female sexual behavior in thebroadest sense, i.e. including behavior exhibited in the introductory (preparatory) phases of sexual interaction (seduction). Therefore, and also because the deviate's verbal reports need not be reliable, the phallometric (phm) method has been employed for the assessment of the deviate's response to systematically applied sexual stimuli. In the first study (Kol~sk~ and Madlafousek, 1972), comparison was made o f the arousing effect of Film scenes showing a motionless, neutrally posing actress with that of scenes showing a "courting" actress and with the arousing effect of the actress's engagement in nonerotic activity. The difference between deviates and normal men was found with respect to the effect of the actress's nonerotic activity. Whereas normals were sexually inhibited by it, deviates were not. On the other hand, the actress's courting behavior elicited a higher phm response than the motionless scenes in both normal and deviant subjects. The latter finding thus made less acceptable the most simple explanation that the deviate seeks to perform intimate acts (exposure, touching) against an erotically unprepared and uncooperative woman because female courtship behavior under all circumstances inhibits him sexually. To further elucidate the problem of sexually arousing stimuli in deviates, the present study was designed to compare the phm responses of normals and deviates to a number of film scenes showing another actress, this time nude, exhibiting behavior with different kinds of "seductive" cues.
Sexual Arousal in Male Offenders
81 METHOD
Subjects The deviant sample was composed of 14 subjects. Half were new patients referred to the authors, the other half were volunteers from those previously examined by the authors. The inclusion of each subject in the deviant sample depended on whether the diagnosis of sexual deviation was substantiated by the interview (Kolfi~sk3~, 1968), predominantly on the evidence that normal (inoffensive) sexual behavior was either absent or insufficient for providing the subject with full sexual satisfaction. The preferred sexual object of all was an adult woman. Ten deviates were exhibitionists with no known history of other offenses, and four had a history of offensive touches ("sexual aggression") without rape. Eight deviates reported having also experienced normal sexual intercourse. The mean age in the deviant sample was 27.9 years, the range from 20 to 51. The control sample of nondeviant men consisted of 14 paid volunteering soldiers and employees of the hospital in whom verbal exploration provided no evidence of the presence of sexual deviation. The fifteenth subject was exctuded because of his report of partial erectile failure. The remaining 14 subjects were sexually successful and active males. Their mean age was 21.7 years, the range from 19 to 30 years. All subjects gave informed consent for the examination, after its nature had been fully explained to them.
Procedure and Measurement The phalloplethysmograph of Freund et al. (1965) was used for the purpose of measuring penile volume changes. Black and white silent film scenes were projected as stimuli. The duration of each was 10 sec. The dimensions of the pictures were 120 by 197 cm, and they were viewed by the subjects from a distance of 6 m. For other details, see the earlier report (Kol~sk2) and Madlafousek, 1972). Because the acceleration rate of penile volume increase usually culminated within 10 sec after the termination of the stimulus, the arousing effect of each stimulus was expressed as a difference between the penile volume at the tenth second after its end and the volume at its onset. Care was taken to dissipate sexual arousal (to restore the prestimulus level of penile volume)before presenting the next stimulus. The stimuli were presented only after the penile volume became stabilized. The first stimulus included in the design was preceded by two preparatory stimulus scenes.
82
Kohffsk:#,Madlafousek,and Novotni Stimuli and Design
The stimuli were film scenes showing the same young woman either naked during the entire scene (14 scenes) or reaching full nudity in its course. It was intended that all 18 scenes represented different kinds of possible seductive behavior, such as chastity or provocative display of the body. The choice of a particular behavior also depended on the actress. The strict limitation, however, was that behavior should not be coital or postcoital, but precoital. The actress was either standing or sitting, never lying down. The front side of her body could be more or less visible in all scenes except for one, in which the standing actress presented her swinging movements from behind. The visibility of particular body zones usually varied during the course of the scenes. (One purpose of seduction is to prevent habituation from constant viewing of the same body zone.) For further description of scenes, see the Results section. The 18 "seductive" scenes were presented to each subject in the same ran. domized order. Before the second half of the scenes was presented, each subject saw several other ("nonseductive") scenes with the same actress. These were "0" (the motionless face), "T" (the motionless torso from the frontal view), ' T ' (the whole actress dressed in a swimming suit, standing and being nonemotional), and " F " (lying on her back, naked, simulating high sexual arousal, being prepared for coitus). The purpose of applying these scenes was to test the effect of the supposed weaker (O,T,I) and stronger (F) stimuli as well as to separate the two halves of "seductive" scenes for statistical control of long-term changes in responsiveness. During this nonseductive interlude, several other scenes, which are not reported here, were presented for piloting purposes. Because the empirical data supported the assumption that deviates do not differ from normals as far as general sexual arousability is concerned, a direct comparison of the arousing efficiency of the scenes was carried out. The statistical analysis was made with logarithmically transformed data, because of their tendency to nonnormal distribution. Measurements in the phm experiments are characterized by considerable variability. This must be taken into account when interpreting nonsignificant results.
RESULTS
As expected, normal subjects responded more strongly to the coital scene (F), less to the "seductive" scenes (Sed. 1, Sed. 2), and least to the supposed weak scenes (O,T,I). This was also true for the deviates. The results (see Fig. 1) seem to substantiate the assumption that there is no difference in the general sexual arousability of normals and deviates. The observed differences between normals and deviates in responses to scenes O, T, I, and F were not significant (t test).
Sexual Arousal in Male Offenders
83
1.8 1.7 u-x-..=
1.6
NORHALS
e.-e--e DEVIATES
1.5 1.4 1.3 1.2 1.1
1.0 SED.1
0
T
I
F
SED.2
Fig. 1. Phm responses of normals and deviates. Ordinate, group mean responses (log) to weak (O,T,I) and strong coital (F) scenes and means from mean responses (log) to the first (SED. 1) and second (SED. 2) halves of "seductive" scenes.
The effect of "seductive" scenes was also similar in both groups as reveaied by a two-way analysis of variance: the main effect of subject group was not significant ( F = 0.136, df = 1, 26). Further results of this analysis o f variance were as follows: No change in responsiveness could be demonstrated from comparison of the effect of the first and second half o f "seductive" scenes ( F = 0.066, df = 1, 16). The interaction of subject group × "seductive" scenes halves was also not significant (F = 0.048, df = 1,442). There was a significant main effect of particular "seductive" scenes, irrespective of halves 1 and 2 ( F = 1.905, d f = 16,442, p < 0.025). This indicates that various kinds o f behavior in the "seductive" scenes had a differential arousing effect, as intended, and that the phm method is capable of reflecting it. An important finding here is that deviates and normals were differentially affected by particular "seductive" scenes, as evidenced by the significant interaction of subject group × particular "seductive" scenes (F = 1.873,!df = 17, 442, p < 0.025). Figure 2 shows that with respect to five scenes deviates strikingly differed from normals. The five scenes mentioned elicited a stronger phm response in deviates than in normals. They tend to belong to those less arousing within the normals' rank order. They will be referred to as "deviant" scenes. With the least arousing scene for normals the difference between both groups reached the 5% level of
84
Kohi~sk}', Madlafousek, and Novotmi ®
120 115 ® 0
110
105
0 0
0 0
O0 Q
•
100 O
0
95 I
|
!
!
1.4
1.5
1,6
1.7
Fig. 2. Comparison of phm responses of normals and deviates to 18 "seductive" scenes. Abscissa, group mean responses of normals (log); ordinate, mean responses of deviates in terms of percentage of the normals' mean responses.
significance, according to the Dunnett's test (1964) (t = 3.014), which takes into account the number of all possible comparisons and thus estimates the t value conservatively. Using the classical method for estimation of t value, the 5% level of significance for a two-tailed test is reached with three scenes (those with the highest difference between both groups). Preliminary observations of scenes, made by the authors, gave the impression that the most conspicuous difference between the "deviant" and "nondeviant" scenes is that for the former scenes behavior that is labeled as "seductive pseudoretreat" is absent, whereas it is present in "nondeviant" scenes. In an attempt to make the identification of this behavior objective, we have operationally defined "seductive pseudoretreat" as avoidance of the frontal body position with respect to the camera and a tendency to squeeze the thighs together. Three observers, unacquainted with the results, were asked to classify all 18 "seductive" scenes according to (1) whether the naked actress assumed, at least for a moment, a frontal position with respect to the camera ("frontal" scenes), and (2) whether the actress squeezed her thighs together. The observers concordantly classified all five "deviant" scenes as frontal and only two out of 13 "nondeviant" scenes as frontal. As far as squeezing the thighs is concerned, all ob-
Sexual Arousal in Male Offenders
85
servers did not reach agreement with one "deviant" scene as well as with two "nondeviant" scenes. However, all observers agreed that the remaining four "deviant" scenes did not contain squeezing of thighs, whereas this was the case with only two of the remaining 11 "nondeviant" scenes. Although this post hoc analysis is only a preliminary result, it throws doubts on the use of "deviant" scenes from the original design for comparing the phm responses of normals and deviates to seductive scenes. Therefore, this comparison was carried out only with the remaining 13 scenes, the seductive character of which did not appear doubtful (Fig. 3). Deviates were affected by the 13 scenes similarly to normals. Spearman rank order correlation is 0.88. No significant difference between the two groups of subjects was found with any scene. (Responses to the "deviant" scenes, represented by crosses, are also shown in Fig. 3 for comparison.)
X
1,7o" X
X
@ @
1.60.
o X
@ ® ®
®
@ ®
®
150.
@
/.4,0. @
@
r
T
Fig. 3. Agreement o f normals and deviates in discrimination o f 13 "nondeviant" scenes. Abscissa, group mean responses of normals (log); ordinate, group mean responses of deviates (log). For comparison, "deviant" scenes are added and represented by crosses.
86
Kol~sk#, Madlafousek, and Novotn~i DISCUSSION
As no systematic analysis of human female precoital erotic behavior ("seduction") has so far been carried out, the creation of stimulus scenes in this study had to be based largely on the intuition of both the actress and the authors. The scenes were intended to contain a variety of seductive cues with the assumption that the resulting effects of particular scenes would quantitatively differ. This proved to be the case. Those scenes which elicited highest reactions in normal subjects can be conceived as empirically verified seductive stimuli. The main finding is that with the majority of scenes deviates responded in a similar way as normals did. This is particularly true for verified seductive stimuli. Also, the deviates' discrimination between weak and strong seductive scenes was similar to that of normal subjects. These results fit the previous study (Kol~ffsk~ and Madlafousek, 1972), which also demonstrated the deviates' capacity to become sexually aroused by a (dressed) female in precoital erotic behavior. Thus the present study provides further evidence that the sex offenders studied share with normal men sexual arousability to normal female erotic behavior. We suggest that this is true for all deviates who offend adult women, although the sample of deviates was composed predominantly of exhibitionists and of a small group of offensive touchers. Separating these two subgroups, both were found to differentiate weak and strong seductive scenes in agreement with normal subjects, and in both subgroups the phm response to the "deviant" scenes was higher than for normal subjects. The question arises then why normal female erotic behavior, although sexually arousing for deviates, inhibits their sexual arousal during the offense and why deviates do not usually have the temptation to perform an offensive act against their regular sex partner. (Exhibitionists do not expose themselves against their sex partner. We have repeatedly seen highly dangerous deviates, who had killed unknown women, who were free of any aggression in interaction with their girlfriends.) At present it is difficult to give a satisfactory answer. Animal studies (Hinde, 1959, 1970) as well as indirect evidence concerning man (Kol~i~sk!) and Madlafousek, 1977) indicate that sexual behavior must not be conceived of as a product of a unitary sexual drive, but rather from the point of view of partial motivational states. The pathological dissociation of these states is, in our opinion, a primary condition which leads to intimate acts (exposure, touches) against an unfamiliar and erotically uncooperative woman. Abel et al. (1975) as well as Marshall (1973) have demonstrated the pos• sibility of eliciting in the phm laboratory abnormal sexual arousal in sex offenders against adult women. The deviates' abnormal sexual arousal also seemed to occur in this study, namely as a by-product of some scenes which did not belong in the verified seductive Stimuli. These scenes supposedly lacked some component(s) of normal female erotic behavior. In the earlier study (Kolfi~sk?? and
87
Sexual Arousal in Male Offenders
Madlafousek, 1972) we found that female nonerotic activity, although inhibiting sexual arousal o f normals, does not inhibit that o f deviates. Future phm studies will be aimed at exploration o f stimulus conditions for abnormal sexual arousal in deviates who offend adult women.
ACKNOWLEDGMENT The authors wish to acknowledge the collaboration of others from the Psychiatric Research Institute in Prague. The preparation o f film stimuli was made with the technical assistance of P. Gemergc and K. Knob. Phallometric measurements were made by K. Knob. Computations were carried out b y ing. Drfibek under the supervision of ing. Roth.
REFERENCES Abel, G. G., Blanchard, E. B., Barlow, D. H., and Mavissakedian, M. (1975). Identifying specific erotic cues in sexual deviations by audiotaped descriptions. J. Appl. Behav. Anal. 8: 247-260. Dunnett, C. W. (1964). New tables for multiple comparisons with a control. Biometrics 20: 482-491. Freund, K., Sedlacek, F., and Knob, K. (1965). A simple transducer for mechanical plethysmography of the male genital. Z Exp. Anal Behav. 8:169-170. Garrett, T. B., and Wright, R. (1975). Wives of rapists and incest offenders, or. Sex Res. 11: 149-157. Gebhaxd, P. H., Gagnon, J. H., Pomeroy, W. B., and Christenson, C. V. (1965). Sex Of. fenders: An Analysis of Types, Harper and Row, New York. Hinde, R. A. (1959). Unitary drives.Anita. Behav. 7: 130-141. Hinde, R. A. (1970). Animal Behavior: A Synthesis o f Ethology and Comparative Psychology, McGraw-Hill, New York. Kolinsky, A. (1968). Zamg/ov~n~ verb~lnf explorace p~i zji~:Eov~n~sexu~lnl"deviace u mu~e (Centering verbal exploration during the estimation of male sexued deviation). Cesk. Psyehiat. 64: 167-171. Kol~sk5,, A., and Madlafousek, J. (1972). Female behavior and sexual aroused in heterosexual male deviant offenders: An experimental study. J. Nerv. Ment. Dis. 155: 110-118. Kol~skg, A., and Madlafousek, J. (1977). Variability of stimulus effect in the course of phedlometric testing, Arch. Sex. Behav. 6: 135-141. Macdonald, J. M. (1973). Indecent Exposure, Thomas, Springfield, Ill. Marshall, W. L. (1973). The modification of sexual fantasies: A combined treatment approach to the reduction of deviant sexual behavior. Behav. Res. Ther. 11: 557-564. Mohr, J. W., Turner, R. E., and Jerry, M. B. (1964), Pedophilia and Exhibitionism, University of Toronto Press, Toronto.
