No title

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

7:41

Style file version Nov. 19th, 1999

Archives of Sexual Behavior, Vol. 30, No. 3, 2001

Defining Premature Ejaculation for Experimental and Clinical Investigations David L. Rowland, Ph.D.,1,2 Stewart E. Cooper, Ph.D.,1 and Michelle Schneider, B.Sc.1

Researchers investigating premature ejaculation (PE) have employed widely diverse definitions of it, thereby limiting progress in the field. This study summarizes available research on PE, notes patterns that emerge from these studies, compares patterns across several types of studies, and suggests a common model for defining PE groups to guide future research. We surveyed two bibliographic databases, identifying 45 studies employing a definition or description of a PE group. From these, we extracted a range of information, including descriptions of the participants, recruitment procedures, if PE subtypes were identified, operational criteria for PE classification, relationship and partner information, and additional inclusion/exclusion criteria. Over 50% of studies reported no criteria, or relied on simple self-identification by participants to establish the PE group. Quantifiable behavioral criteria were used in 49% of the studies, with ejaculatory latency reported most frequently. This measure was also used as a criterion more frequently in studies focusing on assessment of sexual response, whereas the number of penile thrusts was used more frequently in studies prior to 1989. Partner information was often included but seldom used as part of the assessment procedure. Progress on research and treatment of PE will continue to be limited by the absence of commonly accepted criteria for PE group membership and by a lack of identification of relevant PE subtypes and etiologies. This paper suggests a flowchart, based on data and a rational analysis of 40 years of research, for characterizing PE in ways that could assist the development of the field. KEY WORDS: premature ejaculation; sexual dysfunction; ejaculatory latency and control.

1 Department

of Psychology, Valparaiso University, Valparaiso, Indiana 46383. whom correspondence should be addressed at Department of Psychology, Valparaiso University, Valparaiso, Indiana 46383; e-mail: [email protected].

2 To

235 C 2001 Plenum Publishing Corporation 0004-0002/01/0600-0235$19.50/0 °

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

236

7:41

Style file version Nov. 19th, 1999

Rowland, Cooper, and Schneider

INTRODUCTION Important to the clinical and experimental investigation of any sexual dysfunction are the criteria used in defining the disorder. For clinical investigations, a correct diagnosis for individuals seeking help increases the likelihood of appropriate and effective treatment. For experimental investigations, preselected criteria are typically used to establish dysfunctional study groups and to distinguish them from sexually functional comparison groups. To categorize a set of responses as sexually dysfunctional, consensus regarding the relevant criteria needs to be established. In 1993, an NIH Consensus Panel (NIH Consensus Development Panel on Impotence, 1993) provided such criteria for defining erectile dysfunction (ED), and an analogous conference was recently convened to focus on female sexual dysfunction (Basson et al., 2000). However, with respect to premature ejaculation (PE), a problem affecting 20%–30% of men (Laumann et al., 1994; Spector and Carey, 1990), criteria for establishing PE group membership have been decidedly inconsistent. Early clinical investigations into PE—those by Masters and Johnson (1970) and Kaplan (1974)—focused on simple objective criteria such as the number of penile thrusts and latency to ejaculation. Subsequent reports have further emphasized subjective elements such as ejaculating prior to the desired time (Levine, 1992), a lack of control over the ejaculatory response (Grenier and Byers, 1995; Rowland, Cooper et al., 1997), and partner dissatisfaction (Shilon and Paz, 1984). Since much research on PE is aimed at its understanding, diagnosis, and treatment, a number of investigations have referred to the DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, American Psychiatric Association [APA], 1994, p. 509) description for guidance. DSM-IV defines PE by three criteria. First, it involves “persistent or recurrent ejaculation with minimal sexual stimulation before, upon, or shortly after penetration and before the person wishes it.” Second, “the disturbance must cause marked distress or interpersonal difficulty.” And third, “the premature ejaculation is not due exclusively to the direct effects of a substance.” As with other diagnostic categories, DSM-IV describes associated features such as relationship avoidance and typical etiological factors such as emergence in a new relationship or cessation of alcohol use prior to sex. As with several other sexual problems, DSM-IV includes three subtype specifiers: lifelong versus acquired, generalized versus situational, and psychological versus combined etiology. However, despite its multiple parameters and subtype specifiers, DSM-IV makes no attempt to operationalize any component of the description. Although this approach permits flexibility on the part of the clinician, it may also result in an overly subjective diagnostic process, substantial variation in criteria and, at least from an experimental perspective, inconsistency regarding which individuals fall within the dysfunctional classification and which do not. Such consequences are counter to the central purpose of the DSM to

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

Defining Premature Ejaculation

February 19, 2001

7:41

Style file version Nov. 19th, 1999

237

produce a classification system with improved reliability and validity (Nathan and Langenbucher, 1999). Problems in defining PE undoubtedly arise from several sources. First, responses that lead to ejaculation (e.g., number of thrusts, latency, etc.) are continuous measures (Grenier and Byers, personal communication, January 10, 2000), yet for purposes of experimental investigation, these continuous variables are dichotomized to form functional and dysfunctional groups (e.g., <2 min latency = PE). Not surprisingly, there is little agreement as to where to draw the line separating functional and dysfunctional subjects on such measures. Second, as implied by the DSM-IV description, multiple parameters characterize PE, yet it is not clear which of these parameters should be considered necessary components and/or what cluster of parameters might be considered sufficient for a PE classification. Third, men with PE constitute a heterogeneous group. For example, some men show lifelong PE while others acquire the condition later in life or exhibit it with a specific partner. Still others develop concurrent erection problems. Such “PE subtypes” likely reflect multiple etiologies of PE and, as such, are relevant to both clinical and research investigations. Clinicians, for example, may utilize different therapy techniques depending upon these specifiers, and researchers investigating the disorder need to account for potential nonrandom variance arising from such subtypes. And fourth, unlike ED, PE has not been traced to a clear identifiable pathological or psychological origin. In fact, the dysfunctional status of PE emanates more from the sexual dissatisfaction of the patient and his partner than from an actual dysfunction that precludes coitus (except for anteportal ejaculation). Thus, patient and partner satisfaction—highly subjective assessments—may play a relatively greater role in a PE diagnosis than in other sexual dysfunction diagnoses. Such problems have led us to conduct a review of the literature to describe parameters currently used in clinical and research investigations for the description and/or definition of PE. We carried out this review: (a) To describe which and how many parameters are being used to define PE in research investigations, including the criterion level for continuouslyscaled measures. (b) To determine if research studies routinely specify subtypes within the PE sample and if these subtypes are consistent with those specified in the clinical literature, for example, lifelong versus acquired or global versus situational. (c) To compare definitions and descriptions of PE across specific subgroups of studies. These subgroups included recent studies versus those predating 1989; pharmacological versus nonpharmacological studies; studies using sexual arousal, erection, or ejaculation as the response measure versus those investigating some other parameter (e.g., psychometric profile, autonomic activation, etc.); and studies published in medically

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

238

7:41

Style file version Nov. 19th, 1999

Rowland, Cooper, and Schneider

focused journals versus those appearing in journals with a broader clinical and experimental emphasis. Specifically, our goal was to determine whether procedures for defining PE group membership might be related to the recency of the study, the type and goals of the study, or the primary audience to which the study was directed. (d) To develop a working model that might assist in defining and qualifying PE for future investigations. The use of concordant, or at least overlapping, definitions among researchers and clinicians would help ensure that the experimental investigation of PE has enhanced clinical validity and relevance.

METHOD We surveyed the PE literature extending back to 1963, checking MEDLINE and PsychINFO bibliographic databases as well as sifting through our own archives. This search located 61 studies dealing with PE, 45 of which were appropriate for inclusion in that a definition or description of PE formed a necessary step in the investigational procedure. Reports of individual case studies were not included. From each study, the following information was extracted: (a) The number and ages of participants, both overall and by PE group. (b) Reference to a conceptual definition of PE, including its derivation if stated (e.g., from Masters and Johnson, 1970; Kaplan, 1974; or APA, 1994.) (c) The recruitment procedure, including whether the PE men were selfselected or self-identified. (d) Quantifiable behavioral and subjectively assessed criteria for PE group membership, such as measures of “ejaculation latency,” “number of thrusts to ejaculation,” and “ejaculatory control.” (e) The marital or partner status of the participants. (f) Whether partner assessment or participation was included. (g) Categorization of subtypes of PE or, alternatively, specification of type of PE participant. (h) Additional inclusion information or exclusion criteria for establishing PE and/or comparison groups. We also compared this information across four subgroups of studies: those published after 1988, those testing the effect of a pharmacological agent, those examining sexual response (arousal, erection, or ejaculation) as the response variable, and those published in medically focused journals. For this last categorization, we distinguished between journals directed mainly toward a medical audience (defined

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

7:41

Style file version Nov. 19th, 1999

Defining Premature Ejaculation

239

by the clinical medical discipline or the sponsoring professional organization) and those directed toward a more general audience that includes nonbiomedical experimenters and researchers. Our rationale for establishing each of these filters was as follows. With increasing emphasis on methodological rigor, we expected recent studies (arbitrarily defined as occurring within the past decade) to include more quantifiable criteria for establishing PE groups than earlier studies. We also expected pharmacological and sexual response studies to place greater emphasis on group definition criteria resulting from the need for precise measurement of response variables. Finally, we wondered whether researchers and their audiences operated according to different sets of conventions depending on the journal of publication. Specifically, medical journals and their audiences may emphasize clinical efficacy and outcomes over adherence to or reporting of rigorous experimental methodologies. Table I provides a list of the studies examined in the analysis, along with selected information extracted from those studies, including the subgroup categorizations for each study. Author, journal, year, sample size, type of study, etc., are specified. Table II shows the percentage of studies comprising the post-1988, medical, drug, or erectile/ejaculatory response subgroups plus provides additional descriptive information about the 45 studies in the sample. RESULTS Conceptualization, Self-Identification, and Operational Criteria for PE Group Membership All 45 studies classified the sample or a subsample of men as having PE and therefore applied some criterion for establishing PE group membership. As indicated in Table I, 57.8% of the studies included a basic conceptual definition of PE that guided group classification. Table III summarizes the reported criteria for defining PE group membership for all the studies together and for the various subgroups. Briefly, 77.7% of studies indicated that patients self-identified having a PE problem, with 42.2% specifying that patients were also seeking professional treatment at the outset of the study. Quantifiable behavioral (e.g., latency, number of thrusts) or subjectively assessed (e.g., perceived lack of control) criteria for inclusion in the PE group were reported in 48.9% of the studies, with substantial variation occurring in the use of these criteria. Forty percent of the studies relied on a single criterion, while 8.9% used two criteria. Ejaculatory latency was the most commonly used measure (reported in 40% of studies) for establishing PE group membership, with the latency cutoff ranging from 1 to 5 min. In 15.5% of studies using a latency criterion, the defined latency

240 J. Clin. Psychiat. Int. Urol. Nephrol. Acta Eur. Fertil. J. Urol. Eur. Urol. J. Sex Marit. Ther. Andrologia J. Psychosom. Res. J. Sex Marit. Ther. J. Int. Med. Res. J. Urol. Andrologia Urology J. Sex Marit. Ther. J. Int. Med. Res. J. Urol. J. Clin. Psychopharmacol. J. Urol. J. Urol. Int. J. Impot. Res. Arch. Sex Behav. J. Sex Marit. Ther. J. Clin. Psychopharmacol.

Althof et al. Balbay et al. Beretta et al. Berkovitch et al.

Cavallini Cohen Colpi et al. Cooper et al. Cooper et al. Eaton

Ertekin et al. Fanciullacci et al. Fein Godpodinoff Goodman et al. Haensel et al. Haensel et al.

Page

28 23 18 28 19 1

126–130 208–211 583–586 331–336 276–288 432–434

14 36 29 16 18 11

X

DSM-IV

Kaplan DSM-III-R

X X X

DSM-IV DSM-IV

DSM-III-R

DSM-III-R

Kaplan Kaplan and M&J

X

X X

X X X X X

X X X

X X

X

X X X

X X X

X

X

X

X X X X

X X X

X X

X X X

X

X

X

X

X X X X

Ejaculation latency Ejaculation latency Penile blood flow Erectile response Ejaculation latency Ejaculation latency

Ejaculation latency “Responded or not” Ejaculation latency Patient scale of Improvement of EL Ejaculation latency Hormone level Latency perineal recordings Ejaculation latency Psychometric measures Clinician rating of improvement Sympathetic skin potentials Somatosensory potential “Cured or not” Bulbocavernosal response B: better NB: not better Ejaculation latency/control Ejaculation latency/control

7:41

14 36 29 40 18 11

18 43 8 45 16 14 18

101 12 85 12 43 13

15 16 15 11

February 19, 2001

1996 156 1631–1632 1998 159 425–427 1992 4 31–36 1980 9 477–493 1996 22 22–26 1996 16 379–382

22 68 8 45 16 22 33

101 12 137 12 65 13

15 16 15 11

Major dependent variable pp062-295246

1995 153 76–79 1988 20 326–330 1990 35 301–303 1989 15 130–134 1980 8 53–59 1996 156 1310–1315 1998

1995 1997 1986 1984 1993 1973

1995 56 402–407 1998 30 81–83 1986 17 43–45 1995 154 1360–1361

Year Vol.

Archives of Sexual Behavior [asb]

Kara et al. Kim et al. Kim et al. Kockott et al. La Pera et al. Lee et al.

Journal

Author

Mainly Conceptual Conceptual Sexual medical definition definition focus Total (n) PE (n) (X: yes) (X: yes) reference (X: yes)

Table I. Reference and Descriptive Information on the Studies in this Sample

P1: LMF/GDW/GEC P2: GCR Style file version Nov. 19th, 1999

J. Sex Marit. Ther. J. Abnorm. Psychol. 1984 J. Sex Marit. Ther. Arch. Sex Behav. Arch. Sex Behav. Sex Marit. Ther. Urology J. Clin. Psychopharmacol. Br. J. Urol. Am. J. Psychiat. Yonsei Med. J. J. Urol. Yonsei Med. J.

Spiess et al.

Strassberg et al. Strassberg et al.

Strassberg et al.

Trudel et al. Vignoli Waldinger et al.

Xin et al. Xin et al. Xin et al.

Note. M & J refers to Masters and Johnson.

1997 38 1996 156 1995 36

91–95 979–981 397–401

79 592–595 151 1377–1379

163–167 81–82 274–281

251–257

89–101 327–336

186 186 21

27 14

25 20

32

34 26

186 120 21

27 6

25 10

15

23 13

10

20

X X X

X X

X X

X X X

X

X X

X X X

X

X

M&J

DSM-III-R

Kaplan, DSM-IV

DSM-IV DSM-IV DSM-IV

X

X X

X

X

X

X X

X

X

X

X X

X X X X

Ejaculation latency Ejaculation latency/ No. of thrusts Ejaculation latency Penile sensitivity Dorsal nerve potentials

Ejaculation latency Bulbocavernosal response Ejaculation latency

Ejaculation latency

Psychometric measures Ejaculation latency/control Affective/erectile response Penile sensitivity Erectile response/ ejaculation latency Erectile response/ ejaculation latency Erectile response/ ejaculation latency Ejaculation latency Subjective/erectile response

Delay of ejaculation Ejaculation latency Ejaculation latency Ejaculation latency

7:41

1997 1994

2 11 18

19

25 16

24

20

19 26 31 17 14

X

X X

February 19, 2001

Waldinger et al. Waldinger et al.

1987 1978 1998

1990

1999 1987

242–245

198–200

38 39 53 50 22

31 37 46 48

pp062-295246

93

19

194–203 189–197 161–166

783–787

31 37 46 48

Archives of Sexual Behavior [asb]

1993

1978 43 2000 1996 105 1993 19 1997 34

Segraves et al.

Munjack et al. Rowland et al. Rowland et al. Rowland et al. Rowland et al.

1996 77 881–882 1998 159 1935–1938 1998 10 181–184 1995 15 341–346

Br. J. Urol. J. Urol. Int. J. Impot. Res. J. Clin. Psychopharmacol. Psychol. Rep. J. Psychosom. Res. J. Abnorm. Psychol. J. Sex Marit. Ther. J. Sex Res.

Ludovico et al. McMahon McMahon Mendels et al.

P1: LMF/GDW/GEC P2: GCR Style file version Nov. 19th, 1999

241

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

242

7:41

Style file version Nov. 19th, 1999

Rowland, Cooper, and Schneider Table II. Description of Studies Included in the Sample (n = 45) Characteristic of study (%) Post-1988 Medical Pharmacological Focus on sexual response Study designs (%) Between subject Within subject Mixed factorial Sample size Overall (n) PE group (n)

73.3 57.7 53.3 73.3 42.2 46.7 8.8 40.5 (Range: 8–186) 31.9 (Range: 8–186)

had to meet the criterion 50% of the time; in 4.4% of the investigations the latency had to meet the criterion 90% of the time. Another 22.2% of the studies did not include latency as a criterion, but provided individual or mean ejaculatory latencies as descriptive information. A much smaller percentage of the studies (8.8%) included the number of penile thrusts as a defining criterion, with this number ranging from 8 to 15 thrusts. An additional 6.6% of the studies provided descriptive information regarding the number of thrusts for the PE group but did not use this measure as a criterion variable. Ejaculatory control was reported as a criterion in 11.1% of the studies, with another 11.1% providing descriptive information on this parameter without using it as a criterion variable.

Comparison of Subgroups of Studies Using z tests to compare percentages across groups, we explored hypotheses about the inclusion of the above behavioral or subjective measures in relation to the four filters mentioned previously: the recency of the research, the journal audience (mainly medical vs. broader), and the type of study (pharmacological vs. nonpharmacological, or focus on sexual response vs. focus on another parameter). Although the number of comparisons was limited, an adjusted alpha of .01 was used to control for Type I error. Table III summarizes significant findings. The “number of penile thrusts” was used as a criterion more frequently in pre-1989 studies (z = 2.29; p = .01). “Ejaculatory latency” was used as a criterion for PE classification more frequently in studies focusing on sexual response (z = 2.67; p = .007). Although other patterns were discernible (e.g., the greater frequency of including descriptive data on “ejaculatory latency” in nonmedical journal studies and nonpharmacological studies), none reached the .01 criterion level for significance.

40 38 11 11 9 7

Latency criterion Latency description Control criterion Control description Thrust criterion Thrust description

42 25 8 0 25a 0

39 42 12 16 3 9

Post-1989

Year of publication Pre-1989 34 27 8 12 8 4

Med 47 53 16 11 11 11

Nonmed

Journal

50 25 13 8 4 4

Pharm

8 33 0 8 17 8 39 15 12 6 6

Other 52a

Response Sexual

February 19, 2001

a Compared

29 52 10 14 14 10

Nonpharm

Study

pp062-295246

Note. Med = medical, Pharm = pharmacological, Nonmed = Nonmedical; Nonpharm = Nonpharmacological. with opposing category, p ≤ .01.

All studies

Measure

Archives of Sexual Behavior [asb]

Table III. Percent of Studies Using Behavioral Criteria and Information for Defining/Describing PE Group Membership Both Overall and by Type of Study

P1: LMF/GDW/GEC P2: GCR 7:41

Defining Premature Ejaculation

Style file version Nov. 19th, 1999

243

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

244

7:41

Style file version Nov. 19th, 1999

Rowland, Cooper, and Schneider

Inclusion of Partner Information Information about and from the participant’s sexual partner is often used clinically to provide added verification of the problem as well as to assist in understanding the role of relationship dynamics in the disorder. Table IV indicates the extent to which partner information and/or assessment were included. A number of studies (44.4%) required that participants currently be in a sexual relationship, with some adding further stipulation that it be stable. Some studies (15.6%) simply reported the marital/partner status of the participants. Only a small number (4.4%) included an assessment of the problem by the sexual partner as an added criterion for inclusion in the PE group. None reported on sexual dysfunctions of the partner that might contribute to or underlie the PE (e.g., hypoactive sexual desire disorder, anorgasmia, or sexual aversion disorder). Descriptions and Subtypes of PE Men Clinical strategies for the treatment of PE often take into account etiological factors surrounding the dysfunction. Such factors may include the age of the Table IV. Percent of Studies Specifying Etiological Factors Relevant to PE Subjects Percent Age Partner and relationship information Specified as heterosexual in studya Married, cohabiting, partner criterion Marital status reported (not criterion) Specified consent of partner in study Included partner assessment of PE Included partner satisfaction Duration of PE was: Used as a criterion for inclusionb Reported in results PE sample and subtype specifications Sample specified as lifelong Sample specified as primitivec Primary vs. secondary PE vs. PE + ED True vs. partner-relatedc Type I vs. Type IIc E1 vs. E2 c

97.7

Description or comment Mean age of all subjects = 38.2 Mean age of PE subjects = 38.5

46.7 44.4 15.6 15.6 4.4 6.7 11.1 20.0

Duration requirement = 0.5–5 years Mean duration = 9.8 years (± 4.8)

17.7 2.2 24.4 11.2 4.4 2.2 2.2

64% used in the analysis of data 100% used in the analysis of data Not used in the analysis of data 100% used in the analysis of data Not used in the analysis of data

a If not specified, married subjects were assumed to be heterosexual. b This measure is arguably not relevant for studies with only subjects having lifelong or primary PE. c Primitive defined as “psychogenic PE unresponsive to psychotherapy.” True defined as PE “with

different partners, within 15 pelvic thrusts. Type I defined as “normal PE,” and Type II as PE with “unstable erection due to severe fluctuation in [penile] blood flow.” E1 defined as “those who ejaculated prematurely even during masturbation” and E2 as “those who had a problem only or mainly during intercourse.”

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

7:41

Style file version Nov. 19th, 1999

Defining Premature Ejaculation

245

patient, the duration of the disorder, and whether the dysfunction is primary or secondary, global or situational, due to somatic factors, or concurrent with other patient or partner sexual dysfunctions. Table IV presents information regarding the participant’s age, the duration of the dysfunction, and subtype specifications or categorizations reported in studies. 64% of the studies either specified the type of PE patient or classified patients into various subtypes. The most prevalent specification for the sample (17.7%) was “primary PE” or a phrase suggesting primary PE (e.g., “PE since adolescence,” or “most of their lives”) whereas the most prevalent categorization reported in the studies was primary versus secondary PE (24.4%). The remaining studies utilized a variety of less conventional categorizations (see Table IV). By providing information about subtypes, characteristics of the study sample are better understood. However, the extent to which such subtypes might influence results is clearest when this variable is taken into account in the analysis of data. Of the 19 studies that classified PE men according to subtypes, 11 (57.9% of the studies using subcategorizations, 24.4% of all studies) included the PE subtype as a variable in the analysis; another two studies conducted preliminary analyses to justify combining PE subtypes into a single group for analysis. Additional Inclusion Information and Exclusion Criteria Occasionally, patient descriptions were elaborated with further information, for example, when ED (15.5%) or psychological symptoms (6.7%) were present. In most studies, exclusion criteria were reported (86.7%) although these varied widely. Table V lists exclusion criteria and the frequency (in percent) with which such criteria occurred. The most common exclusion criteria were general health-related problems (73.3%), general or organic erectile dysfunction (48.8%), psychiatric problems (48.8%), prescription medications (42.2%), and alcohol or substance abuse (37.8%). The same exclusions were typically applied to sexually functional men when included as a comparison group. Other exclusion criteria relevant to the specific goals of the studies were also listed at times. For example, participants were excluded when the sexual relationship was unstable or the patient was already in treatment for the sexual problem. DISCUSSION This study demonstrates the wide variation used in defining PE group membership in experimental and clinical investigations, a problem recently underscored in a review by Grenier and Byers (1995). Nearly half of the studies either provided no information about how the PE group was formed or relied solely on self-identification and/or clinic referral as the criterion for PE group membership. The lack of quantifiable criteria for PE group membership in these studies may be attributable to several factors. Consistent with the mostly subjective DSM-IV

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

246

pp062-295246

February 19, 2001

7:41

Style file version Nov. 19th, 1999

Rowland, Cooper, and Schneider Table V. Patient Exclusion Criteria

Sexual problems (%) Erectile dysfunctiona Organic erectile dysfunctiona Inhibited orgasm Low sexual desire/frequency Paraphilia Psychiatric or psychological disorders (%) Prescription medications (%) Any medication Medications affecting sexual function only Psychoactive medication only Prolactin stimulating medication only Substance use or alcohol abuse (%) Medical and physical conditionsb (%) Excluded if not in good physical health, disease free Excluded only if cardiovascular, endocrine, urological, or neurological problems

40.0 8.9 8.9 6.7 2.2 46.7 26.7 13.3 6.7 2.2 37.8 53.3 20.0

a Erectile

and organic erectile dysfunctions were mutually exclusive groups; an additional 11% of studies did not exclude men with ED but included them as a separate subgroup in the analysis. Other sexual disorder exclusions were in addition to erectile dysfunction or each other. b The groups under “Medical and physical conditions” were mutually exclusive.

description, clinical researchers may feel capable of identifying PE symptoms without relying on precise, quantifiable criteria. Or, when referred through a medical clinic, researchers may assume that patients have already been thoroughly assessed and therefore need no further work-up to establish the presence and nature of the dysfunction. In other instances, researchers may have used additional quantifiable criteria but omitted them because of the expectations of the publishing journal. Whatever the case, verification of the sexual disorder is critical, given that a self-identified dysfunctional status by the patient is at times discordant with a clinical assessment of the problem (Grenier and Byers, personal communication, January 10, 2000). Men purporting to have PE may not fit the classical definition of the disorder, or they may suffer from another dysfunction (Wincze and Carey, 1991) or condition that either is mislabeled (by the patient) or is primary (Kuhr et al., 1995; Cohen, 1997). Even where quantifiable criteria were incorporated into the investigation, little consensus occurred regarding the key factors that define the dysfunction, or what demarcations might be applied to continuous measures, to differentiate dysfunctional men from functional counterparts. Consequently, men who fell into a PE classification in one study would have failed to do so in another. Furthermore, less than half the studies described participants according to the commonly accepted subtypes of dysfunctional men specified by DSM-IV. An equally low percentage of studies attempted to control for variance arising from different subtypes either by limiting the sample to a single subtype or by including and analyzing subtype classifications. The lack of information and analysis on subtypes increases Type II

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

Defining Premature Ejaculation

February 19, 2001

7:41

Style file version Nov. 19th, 1999

247

error by failing to account for nonrandom sources of variance. Equally important, this situation limits a study’s ability to differentiate effects among men with varying PE etiologies and to target diagnostic/ treatment procedures where they have the greatest impact. Studies could benefit from inclusion of other information relevant to the dysfunction (Rowland, Cooper et al., 1997). For example, men with concurrent PE and ED have substantially different etiologies and may need different interventions than men with PE alone (Rowland et al., 1996). To give an example, by identifying a group of PE men with concurrent ED, we were able to specify that clomipramine, although highly effective in lengthening ejaculatory latency in men with PE, is counterproductive in PE men with ED as this drug further exacerbates the erectile problem (Haensel et al., 1996). Further, few studies incorporate diagnoses of other non-ED sexual dysfunctions in the patient or his partner (e.g., anorgasmia). Such information may be critical to understanding, diagnosing, and treating PE (Rowland, Cooper et al., 1997) and may recast the sexual problem in a more systemic (i.e., relationship) framework, countering the bias in the sexual dysfunction field toward an individual etiology. In fact, relationship therapy, in conjunction with or in lieu of PE treatment, may be the more effective strategy for a man purporting to have PE when the partner is found to have a comorbid disorder such as sexual aversion disorder, anorgasmia, or hypoactive sexual desire disorder (Thexton, 1992; Wincze and Carey, 1991). An analysis across subtypes of studies using the four filters of recency, type (pharmacological, sexual response focus), and primary journal audience yielded two significant patterns. First, the use of “penile thrusts” as a criterion for PE classification has decreased in recent years, not surprising in view of reports questioning the reliability of this measure (Rowland, Kallan et al., 1997). Second, “ejaculatory latency” has been used more often as a criterion variable in studies focusing on sexual response compared with studies investigating other parameters (autonomic, psychometric, hormonal) associated with PE. In the former class of studies, greater precision in the measurement of behavioral criteria is undoubtedly critical to detecting sexual/ejaculatory response changes resulting from the experimental manipulation of variables. Nevertheless, the fact that so few differences emerged from comparisons across subtypes of studies, together with the finding that over half of all studies failed to use any quantifiable behavioral criteria for PE classification, is indicative of the widespread lack of consensus regarding standard procedures in defining PE for research investigations. The predominant exclusion criteria used in this sample of studies are readily justified and should be considered for all research on PE. Good physical and mental health and the absence of other sexual problems (unless included as separate subgroups) ensure that it is safe for the patient to engage in sexual intercourse, that PE is not caused or complicated by other factors, and that psychological impairment does not affect self-report data obtained from the patient. Likewise,

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

248

pp062-295246

February 19, 2001

7:41

Style file version Nov. 19th, 1999

Rowland, Cooper, and Schneider

recreational and prescription drug use and abuse have the potential to obscure effects or interact negatively with other treatment procedures. One caveat in examining and interpreting our data is that not all studies may represent independent observations. Specifically, some research groups have carried out multiple studies on PE and, to the extent that a particular group adheres to a specific protocol in defining PE membership, certain criteria may be overrepresented. Interestingly, however, even studies emanating from a single research group often showed variation in the criteria used to establish PE groups. A second caveat concerns the lack of clarity surrounding the variables of latency, control, and thrusts. Some studies did not indicate whether these variables were used as criteria for PE group membership or simply as descriptions of the sample. For the purposes of this study, we classified the variables as descriptive unless their use as a criterion was specified.

A Model for Future Investigations of PE Patterns from previous research on PE suggest that a rational procedure can and should be applied to the establishment of PE groups. To this end, we propose a flowchart model that leads to consistent decisions about PE group membership (Fig. 1). We began with DSM-IV, as this classification system has documented reliability and validity and is one of the most widely applied psychodiagnostic systems (Nathan and Langenbucher, 1999). DSM-IV incorporates four concepts into its PE description. (a) Ejaculation with minimal physical stimulation, typically operationalized by the number of penile thrusts preceding ejaculation. (b) Ejaculation occurring before, upon, or shortly after penetration, operationalized by ejaculatory latency in minutes and seconds. (c) A persistent or recurrent phenomenon, operationalized by the proportion of times the rapid ejaculation occurs. (d) Ejaculation occurring before the person wishes it, operationalized by a measure of efficacy such as ejaculatory control. Furthermore, the DSM-IV description implies several assumptions. First, that the man identifies himself as having a problem and is sufficiently dissatisfied to seek (or consider seeking) help. Second, that the problem is not merely a response to a concurrent sexual dysfunction of the partner. And third, that the response is not the result of a medical condition. Assuming the prospective participant meets the DSM-IV criteria listed above, the PE description is enhanced by information regarding etiology, including the specifiers delineated by DSM-IV: primary or secondary, global or situational, and psychological or combined etiology. Fig. 1 illustrates a process that generates the information necessary for making decisions

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

7:41

Style file version Nov. 19th, 1999

Defining Premature Ejaculation

249

Fig. 1. Flowchart model producing information relevant to defining PE group membership.

about PE group membership—information that can typically be obtained as part of the patient’s sexual history. Step 1. Recruitment. From a research perspective, PE groups will largely be composed of men who have identified themselves as having the dysfunction and experiencing it as problematic. Thus, Step 1 is self-identification by the participant himself, typically through response to a recruitment notice or referral by a clinic.

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

250

pp062-295246

February 19, 2001

7:41

Style file version Nov. 19th, 1999

Rowland, Cooper, and Schneider

Step 2. Sexual History to confirm PE. Once a candidate has been identified, clinical verification of the PE is necessary. Step 2 consists of obtaining a complete sexual history and includes five components: 2.1, information about each phase of the sexual response cycle (e.g., desire, arousal, orgasm/ejaculation); 2.2, quantifiable behavioral and subjective measures relevant to the ejaculatory phase; 2.3, confirmation of dissatisfaction and consequent duress; 2.4, a medical history; and 2.5, an assessment of partner functioning. The rationale for each step and the operational definition for each variable/construct are as follows. 2.1 The sexual response cycle. Information about other components of the sexual response cycle eliminates them as contributors to the ejaculatory disorder, or alternatively, elucidates their mitigating role in the disorder. For example, an overall low frequency of sexual intercourse (and more specifically, ejaculation/orgasm) has been implicated in the etiology of rapid ejaculation (Spiess et al., 1984; Grenier and Byers, personal communication, January 10, 2000). 2.2 Positive symptoms of PE. Because distinction of PE from functional men requires group classification derived from continuous behavioral measures (e.g., ejaculation latency, number of thrusts), there will probably never be complete agreement regarding optimal cutoff criteria. Nevertheless, we posit that quantifiable criteria are needed to operationalize the constructs contained in the DSM-IV, so differentiated (PE vs. functional) groups can be established and studied. Behavioral criteria, therefore, might include an ejaculatory latency of less than 2 min in at least 75% of attempts at sexual intercourse over the past 6 months. A 2–3 min latency represents an approximate median of the studies included in this sample and substantially less than the 8–16 min typically reported for intercourse by sexually functional men during the final intromission (Rowland et al., 1993; Rowland et al., 2000; Strassberg et al., 1987). Ideally, latencies should be timed with a stopwatch rather than simply estimated. The 75% frequency within the past 6 months operationalizes the DSM-IV concepts of “persistent and recurrent.” Latencies that do not meet these criteria may not always exclude the prospective participant from a PE classification but should elicit further investigation of the problem to determine the true PE status of the man. Furthermore, latencies under 2–3 min do not ultimately define a PE classification since sexually functional men may choose to ejaculate after minimal stimulation because of the circumstances (e.g., concern about interruption from children) or their own or partner’s preference. Therefore, additional information regarding ejaculatory control over rapid ejaculation is important—research indicates that a score of 1–3 on a 7-point scale measuring ejaculatory control (1 (no control), 7 (complete control)) is generally consistent with a PE classification (Rowland et al., 2000; Strassberg et al., 1999). The criterion of “minimal stimulation” is also noted as part of DSM-IV. However, “number of penile thrusts” (the common operationalization of

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

Defining Premature Ejaculation

February 19, 2001

7:41

Style file version Nov. 19th, 1999

251

minimal stimulation) is generally colinear with latency and has lower reliability (e.g., the reliability coefficient for “number of thrusts” was found to be under .60 in a study on a related topic [Rowland, Kallan et al., 1997]). Therefore, its use might be limited to situations where the latency measure is suspect (e.g., as might occur following penetration in the absence of thrusting). Data supporting a particular “thrust” criterion are scant, but 10 or under represents a number generally consistent with several empirically based descriptions of PE men (Haensel et al., 1996; Rowland et al., 1997; Kim et al., 1992; Segraves et al., 1993). 2.3 The disorder as a source of duress. Confirmation that the sexual disorder produces distress can be achieved by asking the participant to respond to a Likert-type item measuring “satisfaction with sexual response.” To be consistent with DSM-IV, we consider a score falling within the lower half to indicate sufficient dissatisfaction for a PE classification. Failure to meet the stated criteria on latency, control (Step 2), and satisfaction may alter the classification pathway to 2b (Fig. 1). 2.4 Medical history. Data from measures of latency and control require further interpretation within the context of the patient’s medical history. Specifically, a history devoid of major medical events leads further to a PE classification, whereas a history indicating probable pathophysiological etiology leads to 2a (Fig. 1) with subsequent elimination from the PE research group, unless the goal of the study includes the investigation of iatrogenic or pathophysiological PE. 2.5 Partner considerations. Whether and to what extent a partner dysfunction or the partner’s behavior plays a role in the PE needs to be understood. Ideally such information should be obtained directly from the partner. Realistically, this is not always possible, so the prospective participant may need to provide information about the partner’s sexual response. Evidence of a possible partner dysfunction such as sexual aversion or anorgasmia may lead to reclassification into the non-PE group, particularly if other criteria (2.2 and 2.3) are not met. For example, a man who ejaculates rapidly because of a sexual aversion by his partner yet who reports moderate to good control over ejaculation should generally not qualify for the PE classification because he chooses to ejaculate rapidly. Although assessment of the participant’s PE by the partner is helpful and should be included whenever possible, this information is often difficult to obtain. Step 3. Preliminary research decision. As indicated in Fig. 1, information obtained through the sexual history guides the researcher toward an empirically based decision regarding PE group membership. Step 4. Exclusion criteria. As suggested by the sample of studies in our analysis, researchers should consider the following exclusion criteria: generally poor health, psychological/psychiatric symptoms, medications affecting sexual response, and substance use or abuse.

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

February 19, 2001

252

7:41

Style file version Nov. 19th, 1999

Rowland, Cooper, and Schneider

Step 5. Final research decision. Once exclusion criteria are applied, the researcher has the necessary information to establish final membership for the PE group. Step 6. Elaborating the PE classification with additional information relevant to etiology. For men included in the PE classification, either limitation to a specific subtype or delineation of sample subtypes based on DSM-IV specifiers should be undertaken. Notation of concurrent sexual problems, particularly ED, and inclusion of such men as a separate study sample will enable more effective treatment procedures for a group of men who have largely been ignored in research studies on PE. Step 7. Treatment implications. The goal of this study was to identify criteria and a process for defining PE group membership. However, once the various types of participants have been categorized, certain treatment options are likely to follow. Broad recommendations related to the relative balance of somatic to psychogenic and relational etiologies can be made, but specific interventions will be sensitive to the patient, his partner, and their views toward use of medication and the benefits of psychotherapy (Rowland et al., 1997). A classification process like the one delineated in Fig. 1 will ultimately need to be evaluated for its clinical and predictive validity. Furthermore, a consensus conference related to premature ejaculation, similar to those called to develop criteria for erectile dysfunction and female sexual dysfunction, needs to be convened to determine agreed-upon conventions for diagnosing this dysfunction. However, we contend that, in the meantime, adhering to this or a similar process will produce significant positive outcomes for the ongoing study of PE. The establishment of PE group membership using the cluster of criteria suggested in this model conforms closely to the DSM-IV description. Furthermore, with its multiple criteria, this procedure is likely to result in fewer “false positives” or “misses” than single criterion strategies are, an ever present danger when attempting to generate dichotomous outcomes (PE vs. non-PE groups) based in part on continuous measures. Finally, consistency in the definition of PE groups will enhance group definition reliability, boost the confidence in conclusions drawn from individual studies, and allow greater generalization across those studies. Inclusion of information on PE subtypes and experimental/statistical control of variance arising from these subtypes will further assist in understanding salient factors in the etiology and treatment of PE.

REFERENCES American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders, 4th Ed., American Psychiatric Association, Washington, DC. Basson, R., Berman, J., Burnett, A., Derogatis, L., Ferguson, D., Fourcroy, J., Goldstein, I., Graziottin, A., Heiman, J., Laan, E., Leiblum, S., Padma-Nathan, H., Rosen, R., Segraves, R. T., Shabsigh, R., Sipski, M., Wagner, G., and Whipple, B. (2000). Female Urology—Report of international

P1: LMF/GDW/GEC

P2: GCR

Archives of Sexual Behavior [asb]

pp062-295246

Defining Premature Ejaculation

February 19, 2001

7:41

Style file version Nov. 19th, 1999

253

consensus development conference on female sexual dysfunction: Definitions and classifications. J. Urol. 163: 888. Cohen, P. G. (1997). The association of premature ejaculation and hypogonadotropic hypogonadism. J. Sex Marit. Ther. 23: 208–211. Grenier, G., and Byers, S. (1995). Rapid ejaculation: A review of conceptual, etiological, and treatment issues. Arch. Sex. Behav. 24: 447–472. Haensel, S. M., Rowland, D. L., Kallan, K., and Slob, A. K. (1996). Clomipramine and sexual function in men with premature ejaculation and controls. J. Urol. 56: 1310–1315. Kaplan, H. S. (1974). The New Sex Therapy, Brunner/Mazel, New York. Kim, Y. C., Choi, H. K., and Lho, Y. S. (1992). Penile hemodynamics in premature ejaculation. Int. J. Impot. Res. 4: 31–36. Kuhr, C. S., Heiman, J., Cardenas, D., Bradley, W., and Berger, R. E. (1995). Premature emission after spinal cord injury. J. Urol. 153: 429–431. Laumann, E., Gagnon, J., Michael, R., and Michaels, S. (1994). The Social Organization of Sexuality: Sexual Practices in the United States, University of Chicago Press, Chicago. Levine, S. (1992). Helping men to control ejaculation. In Sexual Life: A Clinician’s Guide, Plenum, New York, pp. 90–106. Masters, W., and Johnson, V. (1970). Human Sexual Inadequacy, Little Brown, Boston. Nathan, P. E., and Langenbucher, J. W. (1999). Psychopathology: Description and classification. Ann. Rev. Psychol. 50: 79–107. NIH Consensus Development Panel on Impotence. (1993). JAMA 270: 83–90. Rowland, D. L., Cooper, S. E., and Slob, A. K. (1996). Genital and psychoaffective response to erotic stimulation in sexually functional and dysfunctional men. J. Abnorm. Psychol. 105: 194–203. Rowland, D. L., Cooper, S. E., Slob, A. K., and Houtsmuller, E. J. (1997). The study of ejaculatory response in men in the psychophysiological laboratory. J. Sex Res. 34: 161–166. Rowland, D. L., Haensel, S. M., Blom, J., and Slob, A. K. (1993). Penile sensitivity in men with premature ejaculation and erectile dysfunction. J. Sex Marit. Ther. 19: 189–197. Rowland, D. L., Kallan, K., and Slob, A. K. (1997). Yohimbine, erectile capacity, and sexual response in men. Arch. Sex. Behav. 26: 49–62. Rowland, D. L., Strassberg, D. S., de Gouveia Brazao, C. A., and Slob, A. K. (2000). Ejaculatory latency and control in men with premature ejaculation: An analysis across sexual activities using multiple sources of information. J. Psychosom. Res. 48: 69–77. Segraves, R. T., Saran, A., Segraves, K., and Maguire, E. (1993). Clomipramine versus placebo in the treatment of premature ejaculation: A pilot study. J. Sex Marit. Ther. 19: 198–200. Shilon, M., and Paz, G. F. (1984). The use of phenoxybenzamine treatment in premature ejaculation. Fertil. Steril. 42: 659–661. Spector, I. P., and Carey, M. P. (1990). Incidence and prevalence of the sexual dysfunctions: A critical review of the empirical literature. Arch. Sex. Behav. 19: 389–403. Spiess, W. F., Greer, J. H., and O’Donohue, W. T. (1984). Premature ejaculation: Investigation of factors in ejaculatory latency. J. Abnorm. Psychol. 93: 242–245. Strassberg, D. S., de Gouveia Brazao, C., Rowland, D. L., Tan, P., and Slob, A. K. (1999). Clomipramine in the treatment of rapid (premature) ejaculation. J. Sex Marit. Ther. 25: 89–101. Strassberg, D. S., Kelly, M. P., Carroll, C., and Kircher, J. C. (1987). The psychophysiological nature of premature ejaculation. Arch. Sex. Behav. 16: 327–336. Thexton, R. (1992). Ejaculatory disturbances. In Lincoln, R. (ed.), Psychosocial Medicine: A Study of Underlying Themes, Chapman & Hall, Boston, pp. 36–48. Wincze, J. P., and Carey, M. P. (1991). Sexual Dysfunction: A Guide for Assessment and Treatment, The Guilford Press, New York.

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Style file version Nov. 19th, 1999

Archives of Sexual Behavior, Vol. 30, No. 3, 2001

Measures of Transgender Behavior Richard F. Docter, Ph.D.,1,2 and James S. Fleming, Ph.D.1

Using factor analysis, we sought to identify the components of transgenderism. Subjects were 455 transvestites and 61 male-to-female transsexuals, all biological males. A 70-item questionnaire was used, along with other structured questions concerning preferred and usual sex partners. Five factors were identified and interpreted: Transgender Identity, Role, Sexual Arousal, Androallure, and Pleasure. These factors represent the most salient dimensions of transgenderism. All fivefactor Means for transvestites and transsexuals differ. An examination of overlap of group distributions for each factor showed such overlap to range from only 6% for Identity to 46% for both Androallure and Pleasure. Factor intercorrelations for the obliquely rotated factors ranged from −.37 to .27. While transvestites and transsexuals have different lifestyles, their transgender cognition and behavior seem constructed upon different combinations of the same variables. A secondorder analysis of these five factors yielded two factors: Sexual Arousal loaded highest on the first factor (.91), and for the second the highest loading variable was Androallure (.57). Each of these highlights the primary importance of sexual arousal in transgender cognition and behavior. Studying possible age effects, we found that the younger versus older transvestite groups had significantly different scale Means for Androallure and Pleasure; there were no age differences between older and younger transsexuals on any of the five scales. Six percent of transvestites reported a male as their usual sex partner; 25% of the transsexuals reported a female as their usual sex partner. For each group, one-third indicated their usual sex practice was without any partner, while only 5% said they preferred this practice. We propose that the five variables identified offer a comprehensive approach to the description of individual differences in transgender experience and expression. KEY WORDS: autogynephilia; cross-dressing; gender dysphoria; gender identity; secondary transsexual; transgender; transsexual; transvestism.

1 Department

of Psychology, California State University Northridge, Northridge, California. whom correspondence should be addressed at Department of Psychology, California State University Northridge, Northridge, California 91330-8255.

2 To

255 C 2001 Plenum Publishing Corporation 0004-0002/01/0600-0255$19.50/0 °

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

256

10:53

Style file version Nov. 19th, 1999

Docter and Fleming

INTRODUCTION The major diagnostic features of both transvestism and transsexualism have been agreed upon for at least a half-century. Transsexuals typically report a longstanding, intense experience of gender discordance—that is, of being cast into the wrong body relative to one’s gender identity. The major feature of transvestism is periodic cross-dressing for the purpose of sexual arousal, in a heterosexual male (American Psychiatric Association [APA], 1994). We have no argument with these broad diagnostic descriptions which reflect the typological tradition of the medical model; no doubt they serve useful purposes. However, our research has not been concerned with diagnostic classification or typological issues; we are interested in the variables which may be the elements of transgender cognition and behavior. To us, it is imperative to put aside diagnostic classifications and concentrate on the component parts of transgenderism, for in both transsexuals and transvestites, there are abundant examples of individuals whose gender identity, gender-role behavior, and sexual history do not fit neatly into the framework of diagnostic categories. For example, there are scores of well-documented cases of men who have functioned for decades as unremarkable transvestites, often renouncing feelings of intense gender discordance, who ultimately come to live permanently as women, with or without sex reassignment surgery. Such individuals have been labelled secondary transsexuals, since an extensive period of alternative behavior (e.g., transvestism) preceeded a fulltime gender change (Person and Ovesey, 1974). Additionally, among transvestites, there are marked individual differences in gender identity, gender-role behavior, and in the significance of sexual reinforcers across the lifespan; these differences have not been adequately documented or measured, and much of what is known is based on personal histories and case reports. Hence, the development of assessment tools is a vital research priority. The three parameters noted above are important examples, but they are not the only variables calling for additional study. Sexual orientation has also long been considered a variable of importance in both transsexuals and transvestites, and for transsexuals, research supports a bifurcation into homosexual and heterosexual subgroups (Blanchard, 1985). We shall attempt to explore this by asking our subjects about both their preferred sex partner, if any, and also their usual sex partner. Concerning the sexual preferences of transvestites, there is convincing data showing that at least 87% characterize themselves as heterosexual (Docter and Prince, 1997; Prince and Bentler, 1972). But despite this generalization, our own clinical experience has included many cases in which married, heterosexual transvestites report both intense fantasy and behavior, when cross-dressed, involving erotic interpersonal exchanges with men. Using a large sample of transvestites, we shall attempt to clarify this. Finally, little has been published concerning the impact of age upon transgender

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

Measures of Transgender Behavior

February 16, 2001

10:53

Style file version Nov. 19th, 1999

257

experience; we shall describe what we have found concerning the possible effects of aging. We have previously reviewed the assessment of transgender behavior from several perspectives (Docter and Fleming, 1992). Virtually all descriptions of such behavior, spanning transvestism and transsexualism, have recognized some overlap between these syndromes, especially concerning cross-gender identity (Benjamin, 1966; Ettner, 1999; Stoller, 1968, 1974). However, the factorial structure of the construct of transgenderism has not been adequately explored. One exception is the extensive series of reports by Blanchard and associates (e.g., Blanchard, 1985, 1989; Blanchard and Clemmenson, 1988). Of particular interest is Blanchard’s study of autogynephilia—the propensity to become sexually aroused by the thought or image of oneself as a woman. Using factor analysis and a small number of questionnaire items, he found that three autogynephilic subgroups could be identified in nonhomosexual gender-dysphoric men, all having the same “underlying disorder” of gender dysphoria, and all seeking sex reassignment as women. A key point is Blanchard’s finding that in these nonhomosexual men, sexual arousal was a major component both motivating cross-gender cognition and behavior, and providing powerful reinforcement for gender-deviant behavior. But unlike the present study, Blanchard was working with clinical intake questionnaire data that was not designed to explore other variables of greatest concern to us; he was mainly interested in formulating a typology for homosexual and nonhomosexual transsexuals, and in issues bearing on the clinical histories of these groups. In our 1992 report, we described four moderately correlated factors derived from the factor analysis of a 113-item transgender questionnaire, and we proposed that the identified factors were important components of transgenderism. These variables were identified as: Cross-Dressing Identity, Feminization of the Body, Sexual Arousal, and Social/Sexual Role. These factors were based on both exploratory and validation factor analyses. The present report is an extension of this earlier research. Summarizing, we sought to answer these questions (a) What is the factorial structure of transgender behavior and cognition? (b) Is there one or more higher order, overriding factor of transgenderism or gender dysphoria? (c) What dimensions of transgenderism may differentiate most clearly between transvestites and transsexuals? (d) What are the similarities which characterize transvestites and transsexuals? (e) What may be learned concerning both sexual partner preference and typical sexual behavior among transvestites and transsexuals? (f) Does age play a part in transgender behavior?

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

258

Style file version Nov. 19th, 1999

Docter and Fleming

METHOD Questionnaire A self-report questionnaire composed of 178 items was used to assess various beliefs and behaviors considered important in transvestism and transsexualism. The items were based on published descriptions of these transvestites and transsexuals, autobiographical reports, the suggestions of transvestites and transsexuals who were asked to critique earlier versions, and our own clinical experience. In the first phase of this project we used a 75-item questionnaire (Docter, 1988). This was expanded to a 113-item version (Docter and Fleming, 1992), which yielded four factors. Finally, we added 65 new items to our 113-item version to produce a questionnaire of 178 items. An exploratory factor analysis was then carried using 395 transgendered subjects, and the selection of the final 70 items was based on those items having loadings of at least .40, using a five-factor solution. It was this 70-item questionnaire which was used for the present factor analysis. Subjects Subjects were 516 biological males. They were self-described transvestites (88%) and transsexuals (12%), recruited as unpaid volunteers at several transgender conventions and through support groups in the United States and Canada. Our transvestite group was composed of periodic cross-dressers who dress fully as women; partial cross-dressers, that is, those who did not dress fully as women, were excluded. Our transsexual sample stated they were living continuously and permanently in the feminine gender role, with or without so-called sex reassignment surgery. Among the transsexual subgroup, 51% reported having received some form of sex reassignment surgery. Twenty-two percent of our subjects reported having participated in our earlier research; none had any clinical relationship with the researchers. The questionnaires provided for either anonymous responding, which was elected by fewer than 5%, or with an address. When an address was given, we provided a feedback report containing a subject’s scale scores, together with interpretive information. Demographic and Identifying Data The Mean age for transvestites was 45 (SD 11) with a range from 19 to 78 years; for transsexuals the Mean age was 44 (SD 10) with a range from 25 to 70 years. Eighty-two percent of the transvestites had married (transsexuals, 71%), and 45% of those who had married were currently living with their wives (transsexuals, 9%). Thirty-one percent of the transvestites had divorced (transsexuals, 58%), 6% were separated but not divorced (transsexuals, 4%).

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Style file version Nov. 19th, 1999

Measures of Transgender Behavior

259

All subjects were high school graduates, and 89% of both the transvestite and transsexual subgroups had attended at least one semester of college. Of those attending college, 90% indicated they had earned at least one academic degree, as follows: Associate in Arts, 19%, B.A. or B.S., 44%, M.A., 26%, and doctoral level, 11%. The B.A. degree was earned by 38% of the transvestites and 26% of the transsexuals. Ethnicity was distributed as follows: Caucasian, 90%, Afro-American, 5%, Hispanic, 4%, and Asian, 1%. Employment status was reported to be as follows: employed full-time, 84%, part-time, 3%, retired, 12%, and unemployed, 1%. The two subgroups were very similar in employment status. Geographically, the subjects were from all parts of the United States but most came from either the midwestern states (44%) or the western states (21%), with New England, the south, and the eastern seaboard underrepresented. This uneven sampling was due mainly to the location of transgender conventions and support groups from which participants were recruited. Seventy-one percent of our subjects had fathered at least one child. All except 12% were currently members of at least one transgender support or advocacy organization. Sixty-one percent said they had not participated in any prior version of this research, while 22% indicated they had done so; the remaining 17% were unsure. We also inquired about the attire worn by our subjects while they were completing the questionnaire: 63% said they were “dressed as a woman . . . ,” 27% “ . . . as a man . . . ,” and 10% in mixed-gender attire. Probable Sampling Biases Most of our subjects were better educated than the general population, and a high percentage had attended college and had received academic degrees. Taken as a whole, these individuals were better off socioeconomically than a hypothetical national sample of transgendered individuals. Such a sample would range from unemployed street persons to affluent professionals. Another source of bias was that all of our subjects volunteered to assist in this research. This self-selection probably contributed to the production of a sample which is well educated, has strong socioeconomic status, and is free of major mental health problems. Another source of bias may be that many transsexuals choose not to affiliate with support organizations similar to the ones from which our transsexuals were recruited; hence, they are not represented in this sample. Self-Description of Transgender Behavior Together with various identifying data, each subject responded to questions describing his/her transgender behavior. All transvestites described themselves as dressing fully as a woman and of doing so in sessions lasting several hours. Most transvestites said they did so a few times each month, while others ranged from

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Style file version Nov. 19th, 1999

260

Docter and Fleming

daily cross dressing to a few times per year. The transsexuals all stated they were living entirely in the cross-gender role without reversion to the masculine gender role. RESULTS Exploratory Factor Analysis A tetrachoric correlation matrix was formed using the 70 dichotomously scored questionnaire items. This matrix was factor analyzed using the principal factors method and the BMDP 4th program (Dixon, 1985). A scree test (Cattell, 1966) was employed. From an examination of the plot of the largest eigenvalues, the correct number of factors appeared to be four, five, or six. Each of these was then used in a factor analysis using an oblique rotation (direct quatimax). The solution found to be most interpretable was for five factors. A factor loading of .40 was used as the cut-off for item identification in composing each of the five scales. Initial Eigenvalues Prior to Rotation Prior to rotation, the initial eigenvalues, and the percent of variance explained were: Factor I: 29.9 (26.6% variance explained). Factor II: 1.77 (7.8%). Factor III: 1.0 (5.3%). Factor IV: .74 (3.5%). Factor V: .56 (1.9%). The total variance explained prior to rotation was: 45.1%. Five Factors The intercorrelation matrix for the five factors which were the product of our analysis is shown in Table I. Factor Scales and Interpretation Scale scores are the summation of questionnaire items comprising each factor. Our interpretation of the factors is based upon item content. The questionnaire Table I. Intercorrelations for Five Rotated Factors Factor

I

II

III

IV

I II III IV V

.27 −.37 .20 .03

−.28 .14 .17

.07 .23

.00

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Style file version Nov. 19th, 1999

Measures of Transgender Behavior

261

Table II. Means, Standard Deviations, and t Tests Comparing Transvestites and Transsexuals on Five Scales Factor I II III IV V

Identity Role Sexual arousal Androallure Pleasure

Transvestites N = 455

Transsexuals N = 61

Mean

SD

Mean

SD

t test

p

12.2 4.8 5.5 3.6 6.2

7.8 4.1 3.2 2.8 1.9

20.9 14.7 2.0 5.5 4.8

3.3 3.1 2.4 2.5 2.5

8.5 3.2 10.1 6.1 6.0

<.0001 <.0001 <.0001 <.0001 <.0001

items for the five factors are given below in an abbreviated form, together with our rationale for factor interpretation. Factor I: Transgender Identity This factor is comprised of 26 questionnaire items. For 516 subjects, the Mean was 13.3 (SD 7.9). Group Means are presented in Table II. This factor is interpreted as a measure of one’s self-perceived transgender identity—that is, one’s own view of himself/herself relative to masculine or feminine characteristics. Subjects having the higher scores on this factor not only perceive themselves as strongly feminine, but typically report an intense motivation to live entirely in the role of a woman through a permanent change in gender status—that is, a transsexual lifestyle. There is rejection of masculine gender attributes and a desire for both gender-role changes and surgical alteration consistent with becoming a male-to-female transsexual. This scale taps the private, subjective, and experiential attributes of one’s gender experience. Since this scale is measuring cognitions about one’s self, overt cross gender conduct is not being assessed; the focus is upon transgender identity. Low scores for this factor indicate greater congruity between one’s biological sex and one’s gender identity. The following items constitute Factor I. Item wording has been abbreviated where necessary.

Item content Would choose to live as a woman I wish I had been born a woman I am a woman in a man’s body I’d prefer to live as a full-time woman I would like to get rid of my male sex organs Would be happier if born a woman I always wished to be a woman My true gender is feminine

Loading .92 .86 .84 .83 .80 .79 .79 .77

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Style file version Nov. 19th, 1999

262

Docter and Fleming

Wished I could have the internal organs of a woman The “real me” is a woman I have received over 10 hours of electrolysis Believe I am a “true transsexual” I do not enjoy functioning as a man I think of my feminine name as my real name I am certain I am moving toward transsexualism I seldom want to wear men’s clothes I have always felt more like a woman than a man I consider myself to be a transsexual I shall someday (or I did) apply for sex reassignment Masculine identity is weaker than my feminine identity At some time, I shall (or do) live entirely as woman As a woman, I am expressing my “true self” I daydream of being a woman at least once per day Wished I could become pregnant I have spoken with an M.D. about taking estrogen I daydream of being a woman at least ten times per day

.77 .76 .75 .75 .73 .72 .72 .71 .71 .68 .64 .64 .64 .60 .60 .54 .52 .48

Factor II: Transgender Role This factor is composed of 18 items. For 516 subjects, the Mean was 6.0 (SD 5.2). This factor measures actual transgender role behavior. The most heavily loaded items describe various social behaviors. High scores on this factor indicate extensive actual transgender role conduct; low scores are indicative of infrequency or absence of such behavior. Item content I attend entertainment events in my feminine role I eat in restaurants in my feminine role I often shop appearing as a woman I have travelled by airplane in my feminine role I have a driver’s license with my femme picture I am able to “pass” as a woman in public places My feminine name is now my legal name I have checked into hotels or motels as a woman I have taken train trips . . . public transportation as a woman I have taken all-day auto . . . bus trips as a woman In feminine role, men have bought me drinks I have been a student . . . as a woman

Loading .87 .85 .81 .79 .78 .74 .68a .67 .66 .63 .61b .60

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Style file version Nov. 19th, 1999

Measures of Transgender Behavior

Developed passable style of speaking as a woman Relatives and friends know of my feminine identity I have taken female hormones regularly I have attended “straight” business meetings as a woman In feminine role, I have danced with a man Discussed with an M.D. possible cosmetic surgery a b

263

.58 .56 .56 .54 .50b .44

Also loads on Factor V. Also loads on Factor IV.

Factor III: Transgender Sexual Arousal This factor is composed of eleven items. For 516 subjects the Mean was 5.1 (SD 3.4). Nearly all of the items for this factor are based on the association of transgender behavior and the experience of sexual arousal, which is one of the best established generalities concerning transvestites. High scores on this factor are associated with a strong relationship between transgender behavior and sexual arousal, while low scores indicate the opposite.

Item content Wearing beautiful lingerie usually gives me sexual excitement Using beautiful makeup will make me sexually excited Some clothing is especially sexually arousing Wearing beautiful clothes brings greater sexual pleasure Become sexually excited because of my feminine appearance Feminine clothing (etc.) helps me have an orgasm I prefer to wear sexy hosiery and high heels I become sexually excited when dressing as a woman Reading of men becoming women makes me sexually excited When I’m tired, wearing feminine clothing will “perk” me up Wearing a garterbelt and stockings is more sexy than pantyhose a Also

Loading .72 .67 .64 .57 .56 .55 .52 .52 .51 .44a .44

loads on Factor V.

Factor IV: Androallure This factor is composed of nine items. For 516 subjects the Mean was 3.9 (SD 2.8). We have used the term androallure because seven of the nine items for this factor deal with sexual, affectionate, or social encounters between a transgendered (biologically male) person and another male. A single item pertains to sexual arousal with a woman, and one deals with sexual arousal to persons of either sex.

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

264

10:53

Style file version Nov. 19th, 1999

Docter and Fleming

None of these items is explicitly phrased to assess a history of sexual preference. High scores indicate intensity of erotic attraction or affectionate or sexual expression, between a transgendered person and a male, and low scores indicate a relative absence of such encounters. This is not a measure of preferred sex partner.

Item content When in the feminine role . . . attracted to both men and women When I’m wearing a sexy dress, I am more attracted to men In the feminine role, I’ve been kissed by a man In the feminine role, I have had a sexual encounter with a man While feminine . . . had sex encounter beyond kissing a man In the feminine role, I’m exclusively attracted to women Have a fantasy that I’m having sex with a man In the feminine role, I have danced with a man In the feminine role, men have bought me drinks a Also

Loading .76 .73 .69 .69 .68 .68 .66 .49a .42a

loads on Factor II.

Factor V: Pleasure This factor is comprised of nine items. For 516 subjects the Mean was 5.2 (SD 2.0). High scores on this factor are associated with nonsexual enjoyment, relaxation, feeling “perked up,” and joyful delight through transgender experience. Low scores are not associated with negative experience, but simply with less transgender pleasure.

Item content Being in the feminine role is more pleasurable than masculine Being in the feminine role is a super-pleasure for me I eat in restaurants in my feminine role Enjoy seeing pictures of myself in the feminine role Wearing feminine clothes makes me more “alive” or energetic Wearing something feminine will help when I feel “down” Feminine role makes it easier to express emotions My feminine name is now my legal name When I’m tired, wearing feminine clothing will “perk” me up a Also b Also

loads on Factor II. loads on Factor III.

Loading .72 .62 .59 .57 .53 .46 .43 .41a .40b

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Style file version Nov. 19th, 1999

Measures of Transgender Behavior

265

Scale Differences: Transvestites versus Transsexuals Transvestites and transsexuals were found to have significantly different Mean scores across the five scales, as shown in Table II. Overlapping Scores for Transvesites and Transsexuals Many transvestites have scale scores overlapping those of even the highest transsexuals, and vice versa. To examine this more closely, we established a cutoff score at the 67th percentile of whichever group scored highest on each of the five factors; we then defined overlap as the percentage of the second group scoring above this cutoff. Transvestites overlapped the higher-scoring transsexuals as follows: Factor I, 6%; Factor II, 19%; Factor IV, 46%. Transsexuals overlapped the higher scoring transvestites as follows: Factor III, 18%, and Factor V, 46%. Absence of Age Effect The transvestite and transsexual groups were independently divided into younger and older subgroups using the median as a division point. Using the t test, we found no significant differences between older and younger transsexuals for any of the five scales. For the transvestites, there were significant differences ( p < .002) for Factors IV and V. The younger transvestites obtained the higher scores for Androallure, while the older transvestites earned the higher scores for Pleasure. We also obtained Pearson product moment correlation coefficients between scale score and age for each of the groups and for each of the five variables; none of these correlations exceeded .30. Taken as a whole, our conclusion is that age effects may play a slight role in the determination of two variables for transvestites, and that age is unrelated to the scale scores of transsexuals. Preferred Sex Partners Our subjects were asked “With whom have you most strongly preferred to have sex during your adult life?” A separate question asked “Regardless of your preference, with whom have you usually had sex?” The results are presented in Table III. Second-Order Factor Analysis As an exploration, a second-order factor analysis was carried out based on the intercorrelation matrix for the five factors described above. The same previously

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Style file version Nov. 19th, 1999

266

Docter and Fleming Table III. Preferred and Usual Sex Partner

Adult female (%) Adult male (%) Equally preferred (%) No partner (%) None apply (%) a For

Transvestites N = 455

Transsexuals N = 58

Preferred

Usual

Preferred

Usual

69 7 10 5 9

53 6

47 19 17 4 13

40 21

a

34 7

a

27 12

Usual partner, this response option was not provided.

described analytic procedures were employed, including oblique rotation. Two factors emerged and the loadings upon these factors are shown in Table IV. The original factor of Sexual Arousal was most heavily loaded on Factor I. The pattern of loadings for the other three original factors supports the interpretation of this as a Transvestic Autogynephilia factor—that is, experiencing sexual arousal in association with the thought or image of oneself as a woman. For secondary Factor II, the highest loading (.57) was the primary factor of Androallure with moderate positive loadings for Identity and Role factors, and a very low loading (.17) for Sexual Arousal. We have named this factor Autogynephilic Pseudobisexuality. DISCUSSION Our results offer evidence for a five-factor model of transgenderism. These variables may help to clarify the structural dimensions of transvestism and transsexualism. The parameters considered here offer an alternative to the diagnostic or typological approach; at the least, they may help to refine such an approach. The scales we have developed are each more extensive than previously published measures of these variables. While transvestites and transsexuals live quite different gender lifestyles, their transgender experiences and erotic motives have many similarities. For example, 6% of our transvestites described feelings of very intense feminine gender Table IV. Factor Loadings for Two Secondary Factors Factor I II III IV V

Identity Role Sexual arousal Androallure Pleasure

Secondary Factor I Transvestic Autogynephilia

Secondary Factor II Autogynephilic Pseudobisexuality

−.40 −.35 .91 .09 .29

.34 .25 .17 .57 .03

Note. The correlation between the two secondary factors was −.20.

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Measures of Transgender Behavior

Style file version Nov. 19th, 1999

267

identity overlapping with the scale scores of most transsexuals. This is a confirmation of Blanchard’s finding (1989) that his three nonhomosexual autogynephilic subgroups showed less intense transgender identity than did his homosexual transsexuals. Many of our transsexuals reported sexual arousal associated with transgender ideation and behavior akin to what most transvestites said they experienced. This is also directly in line with Blanchard’s 1989 study. In an earlier report, Blanchard and Clemmensen (1988), found that in nonhomosexual transsexuals sexual arousal to transgender stimuli inversely covaried with gender dysphoria (r = −.56), and they noted that this sexual arousal is “relatively common” among heterosexual transsexuals (p. 430). Our findings do not support the model of a single over-arching construct, such as the gender dysphoria explanation for transsexualism. This simplification offers a one dimensional focus on what we see as highly complex, multidimensional cognition and behavior of the transsexual. Reliance upon a single construct, in our view, is simplistic. The same logic applies to the continuing oversimplification of transvestism, which is too often said to be little more than cross-dressing for the purpose of sexual arousal, while giving scant attention to other variables which play a part. While sexual reinforcers are almost invariably described in the history of the transvestite, the other variables highlighted here—Identity, Role, Pleasure and Androallure—are also important sources of reinforcement for most transvestites. Differences Between Group Means The scale score Means for transvestites and transsexuals differ significantly across the five variables, as shown in Table II. Since there are great differences in gender behavior of transvestites and transsexuals, these differences in scale score Means offer some support for the construct validity of our measures. Transsexuals scored highest for Factors I (Identity), Factor II (Role) and Factor IV (Androallure), which is what we had expected based on our earlier research (Docter and Fleming, 1992), while the transvestites scored the higher Means for Factor III (Sexual Arousal) and Factor V (Pleasure), which was also expected. While there is no surprise in these Mean differences, what is surprising is the extent of overlap in scale scores shared by the two groups.

Group Similarities in Scale Scores Overlap between the two groups was defined as that percentage of the group having the lower Mean scale score which exceeded the 67th percentile of the group having the higher Mean for a given variable. For all five factors, we found some overlap, ranging from a low on Factor I (Identity) of only 6%, to a high of 46% for

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

268

Style file version Nov. 19th, 1999

Docter and Fleming

Factors IV (Androallure) and V (Pleasure). Nineteen percent of the transvestites reported cross gender role behaviors (Factor II) overlapping with the highest twothirds of our transsexuals. These individuals said they behaved in transgender roles much as did the transsexuals, through participation in a variety of public events. At present, we do not know if such experience, perhaps combined with a high transgender identity, may be predictive of secondary transsexualism, but it is in line with the description of so-called marginal transvestites (Buhrich and McConaghy, 1977) considered more likely to opt for a transsexual lifestyle. While transsexuals often report little or no sexual arousal associated with their feminine expression of gender, we found 18% of this group exceeded the 67th percentile of the transvestite group. The distribution of Sexual Arousal scores for transsexuals were very negatively skewed, while for the transvestites, these scale scores were very evenly distributed. Only 15% of the transvestites described themselves as experiencing no sexual arousal associated with cross-dressing, while 75% of the transsexuals so stated. For scales I, II, and III, the amount of group overlap never exceeds 19%. In contrast, 46% of the transvestites exceeded the 67th percentile of the transsexual group on Factor IV (Androallure). High scale scores on Factor IV should not be interpreted as indicative of a homosexual preference, although many Androallure items deal with affectional, social, or sexual encounters between a transgendered person and a male. While affectional encounters with men are not rare in transvestites, the topic is seldom openly discussed in either their support-group meetings or in transvestite publications. Our cultural discomfort with gay-like behavior may play a part here. Perhaps the confidentiality and anonymity of our survey facilitated greater Androallure-type disclosures among our transvestite subjects. Finally, we found a 46% overlap of transsexuals for Factor V (Pleasure). Perhaps the only surprise here is that we did not have even more transsexuals reporting a special sense of elation, delight, relief from stress, and exhilaration associated with their transgender experience. Our interpretation of this is that many transsexuals do not experience anything emotionally unique about their change in gender; their transsexual lifestyle may simply be part of the ups and downs of life. In contrast, the transvestites report their periodic crossdressing episodes as uniquely pleasurable. Together with sexual reinforcers, the positive affective experiences described by our Pleasure variable constitute the most important reinforcers associated with their cross-dressing. Absence of Age Effect There is considerable reference in the literature concerning transvestites who, over the course of their lives, report lessening of the sexual significance of crossdressing (e.g., Brierley, 1979). We did not find such an age effect for our younger versus older transvestites when the Means for Sexual Arousal were compared. Our scale scores reflect significant Means differences for transvestites only for

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

Measures of Transgender Behavior

Style file version Nov. 19th, 1999

269

Androallure—wherein the younger men obtain higher scale scores, and Pleasure— wherein the older transvestites obtain the higher scores. While the age Means for these two scales differed, none of the correlations between scale scores and age exceeded .30 or −.30. This was also true for the transsexuals, who did not show any significant differences contrasting younger and older subjects for any of the five variables. Hence, except for the two-scale differences for transvestites, age does not appear to play a part relative to these variables.

Preferred and Usual Sex Partner For both groups there are some substantial differences between the declared preference of sex partner and the usual sex partner. About 30% of each group reported usually having solitary sex, while for both groups, only about 5% said they would prefer this. For the transvestites, 69% reported preferring a female partner, but this was the usual sex partner for only 53% of this group. While most transvestites are self-described heterosexuals, 6% reported a preference for a male sex partner, and 7% reported a male to be their usual sex partner. Very little is known about this homosexually or bisexually inclined subgroup of periodic cross-dressers. The 9% of transvestites who said that none of the response options applied to them gave a variety of reasons, such as medical complications, lack of interest, effect of medication, and lack of opportunity or experience. A surprising 47% of our transsexual subjects reported preferring a female sex partner, and 40% said this was their usual partner. Only 19% said they preferred a male partner, while 17% indicated a bisexual preference. This sample of transsexuals seems far more heterosexually inclined than we had expected. Perhaps our sample of transsexuals was overrepresented by the older, so-called secondary transsexuals who often have a heterosexual history. Those transsexuals who are usually having sex alone (27%) represent more than six times the percentage who indicate this as their preference (4%).

Factor Intercorrelations The intercorrelations among the five factors for the combined groups, following rotation, ranged from −.37 to .27, indicating that these variables are largely, but not entirely, independent. The inverse relationship of −.37 between Identity and Sexual Arousal was previously reported by Blanchard and Clemmensen (1988). The transsexual group, which scored higher on our Identity scale, typically scored lower on Sexual Arousal. Conversely, the transvestite group, which scored lower on the Identity factor are far more likely to report greater sexual arousal. The same relationship between scales is shown by the −.28 correlation between Role and Sexual Arousal, taking both groups as a whole. But it is less clear why a correlation of only .27 would be seen between Identity and Role. At first glance, one would expect that

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

10:53

270

Style file version Nov. 19th, 1999

Docter and Fleming

the more intense one’s transgender identity, the more extensive would be the transgender role, and that a substantial positive correlation would be seen. But this is not the case. For transvestites, the correlation was .32, and for transsexuals, −.13; Transgender Identity and Role share little variance for the transvestites and none for transsexuals. It is also of interest to note that only about 5% of the variance of Sexual Arousal is shared with Pleasure. These factors seem to be tapping two different domains of affective experience, the former highly sexual, and the other spanning positive but nonsexual affective experience. While much attention has been given to the reinforcing value of sexual arousal and masturbation in transvestites, comparatively little attention has been given to these nonerotic reinforcers. Second-Order Factor Analysis Two second-order factors emerged: For Factor I the highest loading variable from our earlier analysis was Sexual Arousal (.91); for Factor II the highest loading variable was Androallure (.57). Our interpretation of these factors is based on item content for each of these primary factors. It is clear that the dimension of sexual arousal is overwhelmingly represented in each of these secondary factors, but the factors differ in how sexual arousal is stimulated. For Factor I, which we shall identify as transvestic autogynephilia, the effective stimuli are thoughts and behavior associated with cross-dressing, and with the desire to imagine or visualize one’s self as a woman. Conversely, we have named Factor II autogynephilic pseudobisexuality—that is, the enhancement of one’s identity as a woman through garnering the attention, affection, or other indications of interest in one’s self as a woman, by a man. We are in debt to Blanchard (1989) for his extensive original work on the concept of autogynephilia. We have not used the terms heterosexual or homosexual to describe the item content of the scales sharing variance to form these secondary factors because these items cannot be read, simply, in terms of partner preference. There are, however, items which allow a respondent to indicate a different strength of erotic arousal associated with stereotypically feminine stimuli or masculine stimuli; such a preference may be independent of partner preference or of the usual sex partner, although in most cases an opposite sex/gender preference is reported. Further, concerning the Androallure scale, the quest of the transgenderist to be attractive to men, and to participate in either non-sexual interpersonal processes or various erotic foreplay does not necessarily equate to a homosexual partner preference. As contradictory as this may seem for heterosexual males, some transvestites and transsexuals appear to seek the attention of men, from talk to touching, as confirmation of their worth as “women,” rather than as a prelude to homosexual behavior. Once such gender and identity confirmation has been achieved, there may be little interest in pursuing actual sexual behavior with a male: hence, Blanchard’s term, pseudobisexuality (1989). None of this reasoning takes away from the fact that there are homosexually and bisexually inclined

P1: LMF/GDW/GEE/GEC

P2: FYJ/GAY

Archives of Sexual Behavior [asb]

QC: FTK

PP062-295247

February 16, 2001

Measures of Transgender Behavior

10:53

Style file version Nov. 19th, 1999

271

transvestites, and that in larger samples of transsexuals, a homosexual preference is commonly reported.

CONCLUSION We conclude that transgender cognition and behavior should be conceptualized as complex and multidimensional, and that the five variables presented here are fundamental constituent elements of transgender thought and behavior. We did not find a single overarching variable to help account for transgender behavior. Transgenderism is most accurately conceptualized as the product of many determinants, and very likely, to be independent of sexual partner preference. As an alternative to diagnostic classfication, we suggest the use of the five variables described here as a more complete approach to describing individual differences in gender experience and expression.

REFERENCES American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders, 4th Ed., American Psychiatric Association, Washington. Benjamin, H. (1966). The Transsexual Phenomenon, Julian Press, New York. Blanchard, R. (1985). Typology of male-to-female transsexualism. Arch. Sex. Behav. 14: 247–261. Blanchard, R. (1989). The concept of autogynephilia and the typology of male gender dysphoria. J. Nerv. Ment. Dis. 17: 616–623. Blanchard, R., and Clemmensen, L. H. (1988). A test of the DSM-III-R’s implicit assumption that fetishistic arousal and gender dysphoria are mutually exclusive. J. Sex. Res. 3: 426–432. Brierley, H. (1979). Transvestism: A Case Book, Pergamon, New York. Buhrich, N., and McConaghy, N. (1977). The clinical syndromes of femmiphilic transvestism. Arch. Sex. Behav. 6: 397–412. Cattell, R. B. (1966). The scree test for the number of factors. Multivar. Behav. Res. 1: 245–276. Dixon, W. J. (ed.) (1985). BMDP Statistical Software: 1985 Printing, University of California Press, Berkeley. Docter, R. F. (1988). Transvestites and Transsexuals: Toward a Theory of Cross-Gender Behavior, Plenum Press, New York. Docter, R. F., and Fleming, J. S. (1992). Dimensions of transvestism and transsexualism: The validation and factorial structure of the Cross-Gender Questionnaire. J. Psych. Hum. Sex. 4: 15–37. Docter, R. F., and Prince, V. (1997). Transvestism: A survey of 1032 cross-dressers. Arch. Sex. Behav. 6: 589–605. Ettner, R. (1999). Gender Loving Care, Norton, New York. Person, E., and Ovesey, L. (1974). The transsexual syndrome in males: II. Secondary transsexualism. Am. J. Psychother. 28: 4–20. Prince, V., and Bentler, P. M. (1972). Survey of 504 cases of transvetism. Psych. Rep. 31: 903–917. Stoller, R. J. (1968). Sex and Gender, Science House, New York. Stoller, R. J. (1974). Sex and Gender, II Aronson, New York.

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

11:28

Style file version Nov. 19th, 1999

Archives of Sexual Behavior, Vol. 30, No. 3, 2001

Men Sex Workers and Other Men Who Have Sex With Men: How Do Their HIV Risks Compare in New Zealand? Martin S. Weinberg, Ph.D.,1,4 Heather Worth, Ph.D.,2 and Colin J. Williams, Ph.D.3

Sex workers have their perspective on HIV transmission, claiming that in general they are more similar than different from other people in HIV status and the practice of safe sex. Such an assertion of similarity goes against public and professional opinion that prostitution is a major vector in the spread of AIDS. Taking the sex workers’ similarity claim seriously, this paper considers the conditions under which it would be valid. We focus on those factors that make a population more or less vulnerable to HIV and how they affect its spread into the sex work population. Data from New Zealand comparing men sex workers and other men who have sex with men is used to evaluate these ideas. Data partially support the hypothesis in that these two groups of men are similar with regard to their HIV status. We do find the sex workers to be different, however, in their being less likely to engage in safe sex practices. We provide an explanation for why this has not lead to their having a higher rate of seropositivity. KEY WORDS: male prostitutes; men prostitutes; sex workers; homosexuality; bisexuality; men who have sex with men; HIV; AIDS.

INTRODUCTION How do men sex workers (MSWs) and other men who have sex with men (OMSM) compare with one another in terms of HIV risk? One position—which we call the sex workers’ perspective—is derived from the writings of sex work 1 Department

of Sociology, Indiana University, Bloomington, Indiana. for Research on Gender, University of Auckland, Auckland, New Zealand. 3 Department of Sociology, Indiana University—Purdue University, Indianapolis, Indiana. 4 To whom correspondence should be addressed at Department of Sociology, Indiana University, 1020 E. Kirkwood Ave., Bloomington, Indiana 47405. 2 Institute

273 C 2001 Plenum Publishing Corporation 0004-0002/01/0600-0273$19.50/0 °

P1: LMD Archives of Sexual Behavior [asb]

274

PP062-295248

February 16, 2001

11:28

Style file version Nov. 19th, 1999

Weinberg, Worth, and Williams

activists and their academic allies (Chapkis 1997; Delacoste and Alexander, 1988). As Jenness (1993, pp. 126, 127) states: “[It] has disavowed its constituency’s standing as ‘different’. . . and focused on and promoted similarities between prostitutes and other groups in society, while rendering differences invisible.” The first part of this argument has been to repudiate the notion that prostitutes are people threatening their communities with contagion (cf. Alexander, 1995; Chapkis, 1997; Delacoste and Alexander, 1988; English Collective of Prostitutes, 1997; O’Neill, 1997; Pheterson, 1996). This point was often expressed by major spokespersons for sex workers at the 1998 International Conference on Prostitution–that in modern nations the rate of HIV infection is low among non–drug-using sex workers (cf. Kruhse-Mountburon, 1992; Overs and Longo, 1997; Plant, 1990) and not different from the rate in the general population. The second element of the sex workers’ perspective is that the argument above is true mainly because sex workers are conscientious about taking safe-sex precautions. Their position is that even though client demands for sex without condoms can lead sex workers to engage in unsafe sex (Bloor et al., 1992, 1993; Chapkis, 1997; McKeganey and Barnard, 1992), the proportion of less careful individuals is no greater than among nonworkers (cf. Davies and Feldman, 1997; Scambler, 1997). Sex workers further assert that they are more knowledgeable than other individuals about the facts of HIV transmission and even more likely to keep a medical check on their HIV status because sex is their occupation (statements made at the 1998 International Conference on Prostitution). Thus, as a matter of good sense, sex workers typically use condoms in their work. To say, then, that sex workers generally have been a major source of STDs and AIDS is to misunderstand the nature of prostitution. Rather, sex workers promote the image of themselves as careful, performing a service to society, being more similar to than different from other people, and being unfairly stigmatized (and in the process having negative auxiliary traits such as “disease carrier” attributed to them; cf. Scambler, 1997). There is quantitative research that supports the sex workers’ perspective with regard to HIV among MSWs. In Montreal, Shaver and Newmeyer (1996) find MSWs to be no more likely to report engaging in any anal intercourse or unprotected anal intercourse in noncommercial sex than are OMSM, and to be unlikely to engage in high risk behaviors with their clients. Also, they find MSWs are more likely to be tested and less likely to report being seropositive. Thus, Shaver and Newmeyer (1996, pp. 13, 15) conclude that there are “more similarities than differences” between the general group of men who have sex with men and the sex workers, and that “the data do not justify. . . fears about the risky practices of male hustlers.” In addition to the quantitative research, there are also qualitative studies that describe the ways in which sex workers maintain the practice of safe sex, and the factors related to risky practices (e.g., for men sex workers in Melbourne, see Browne and Minichiello, 1995, 1996; on the factors related to risky practices among men sex workers in the Netherlands, see De Graaf et al., 1994; for a similar

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

11:28

HIV-Risk Comparison among MSWs and OMSM in New Zealand

Style file version Nov. 19th, 1999

275

focus on women sex workers in The Netherlands, see Vanwesenbeeck, 1994, 1995; Vanwesenbeeck et al., 1994). There is research, however, that suggests a “more different than similar” viewpoint. Demographic results show MSWs as more likely than OMSM to be younger, lower class, and underclass minority members (Boles and Elifson, 1994, show this in Atlanta; Morse et al., 1991, in New Orleans; and, Pleak and Meyer-Bahlburg, 1990, in Manhattan)–statuses that are associated with a lesser likelihood of heeding health advice (cf. Plant, 1997). Further, they are more likely to be bisexual and heterosexual identifying men (Prestage and Hood, 1993, in Australia; Shaver and Newmeyer, 1996, in Montreal) who are less likely to be attached to the gay community and thus assumed to be less exposed to its safe sex programs (for the U.S., Doll and Beecker, 1996; Stokes et al., 1997; for Australia, Kippax et al., 1994). Thus, from the research, fear surrounding MSWs could be warranted because bisexual men have been found to be highly represented among sex workers, use condoms inconsistently, seldom disclose their bisexuality to their women partners, and report more multiple HIV risk behaviors than do exclusively homosexual men (Doll and Beeker, 1996; Morse et al., 1991; Stokes et al., 1997; also, cf. Von den Hock and Coutinho, 1991; Waldorf et al., 1900). They are, therefore, believed to be a conduit for HIV into the larger society. Our aim is to more clearly specify the conditions under which the sex workers’ formulation regarding HIV infection may or may not hold true. We propose the following sociological view. THEORY It seems clear that sex workers are at greater potential risk in contracting HIV because they are more likely to have a larger number of sex partners, engage in more frequent sex, and inject drugs (cf. Kaplan et al., 1987). Thus, for the sex workers’ perspective to be true, these behaviors must take place in locales that decrease a sex worker’s vulnerability to HIV infection (cf. Mann and Tarantola, 1996). We hypothesize that these variables include: The pool of infection. The rate of HIV infection is perhaps the most important element in creating a dangerous environment. When large proportions of people in a locale are infected, the vulnerability of others is increased. The economic level of the locale. The degree to which poverty exists is important as HIV risk is disproportionately concentrated in impoverished populations. Poverty also creates the alienation that can underlie a large IV drug culture. Locales that lack economic resources also will be unable to fund prevention programs, medical services and facilities, testing services, HIV-related education, and free condoms. The political element. Political elements are illustrated when, at the beginning of the AIDS epidemic in the United States, the conservative Republican

P1: LMD Archives of Sexual Behavior [asb]

276

PP062-295248

February 16, 2001

11:28

Style file version Nov. 19th, 1999

Weinberg, Worth, and Williams

administration ignored the mounting crisis. Such a refusal to recognize the problem can interfere with the free flow of information pertaining to HIV—for example, the reluctance to advertise condoms in the mass media, provide sex education, or give support to outreach programs. Sometimes those with HIV also may be subject to official or legal sanctions. Such discrimination against certain groups can lead to societal exclusion which exacerbates their vulnerability to the disease. In addition, laws that target certain behaviors can increase vulnerability to HIV. For example, the legal prohibition of certain drugs can block or hamper needle exchange programs, and police action against prostitutes and homosexuals make these persons more difficult to reach, thus discouraging them from developing their own outreach programs. Such outreach programs appear to be the most effective ones in controlling the health-related behaviors of sex workers (cf. Alexander, 1995; Overs and Longo, 1997). Other sociocultural factors. The more people are in a socially dependent position, the less power they have over their sexuality; thus, it becomes more difficult for them to insist on condom use. Also, vulnerability relates to both illiteracy and lack of education, which directly affect the spread of information about the disease and the ability to utilize safe sex practices. Finally, negative religious and societal attitudes toward sexuality can make openly dealing with the issue of HIV shameful, which can increase vulnerability. Given these considerations, we would suggest that the potential risks sex workers face from HIV infection will be reduced in certain locales and in such locales the likelihood that sex workers will be more similar than different from comparable persons in this regard will be high. This will be the case to the extent that the locale has (a) a small pool of infection, (b) an economy that maintains its population at a relatively high standard of living, (c) a government that recognizes the HIV problem and devotes resources to it, and (d) a sociocultural situation with values that do not exclude sex workers and drug users. These conditions should both decrease the vulnerability of its population to HIV infection and also make sex work safer, therefore increasing, in the case of this study, the similarity of MSWs and OMSM with regard to HIV-related phenomena.

THE CASE OF NEW ZEALAND New Zealand was chosen to examine these ideas because we believe it is a locale that is relatively low in vulnerability to HIV and thus its differential spread into the sex work occupation. We expect then that the sex workers’ formulation will hold true and that sex workers and comparable persons will be similar in rates of HIV and related phenomena. The cumulative incidence of AIDS per 100,000 population was used as an indicator of the pool of infection, which around the time of the study was 1.36 in

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

11:28

HIV-Risk Comparison among MSWs and OMSM in New Zealand

Style file version Nov. 19th, 1999

277

New Zealand (WHO, 1997). This is relatively low compared, for example, to the United States (13.81), Thailand (30.29), and African countries such as Zimbabwe and Nambia (79.28 and 165.51, respectively). Economically, New Zealand is a country with a relatively high standard of living and an encompassing middle-class. It has an economic level that has made it capable of providing a modern state-subsidized health system that has programs for HIV education and safe testing. There is a disparity in the economic position of people in New Zealand but not to the extent that creates the widespread estrangement among the poor that is fertile ground for the development of a large IV drug culture. Politically and governmentally, New Zealand has set up more liberal programs with respect to HIV prevention than other countries (e.g., the United States with its failure to have needle-exchange programs). There is also less interference with the free flow of complete information with regard to HIV. In addition, socioculturally, neither illiteracy nor a lack of education has interfered with the spread of information about HIV and the ability to use safe sex practices. New Zealand’s effectiveness in dealing with HIV is further aided by a culture that is less negative to sex and sexual minorities than many cultures, as seen in the Homosexual Law Reform Act of 1986 (which legalized consenting acts between men at an age comparable to heterosexuals), and the Human Rights Amendment Act of 1993 (which made it illegal to discriminate on the basis of HIV status). Further there is a general acceptance of homosexuals by the public (“Awareness of and Attitudes to Discrimination on Grounds of Sexual Orientation,” 1992). Because New Zealand does not have a large fundamentalist religious population or a strong “religious right,” there is a smaller proportion of people outraged at sexual minorities and commercial sex. This, combined with a value on privacy, has led to less social discrimination against sex workers and men who have sex with men, further reducing the estrangement of these populations that can increase vulnerability to HIV and its spread. New Zealand also is a relatively stable multiethnic society (although Maoris claim a loss of land and culture through colonial process and there are claims of prejudice by people from non-European backgrounds). Most of the population, however, does seem committed to inclusiveness. Finally, New Zealand has the politically organized presence of both effective homosexual organizations and sex worker organizations that provide HIV testing, education, and condom distribution as well as a drug and needle replacement program that works in conjunction with it. This program has obtained a lot of cooperation from sex workers and parlor owners and has been funded by the government. Given these features of New Zealand society, we suggest that it has an environment that is conducive to reducing vulnerability to HIV infection so that HIV-related differences between MSWs and OMSM will be small or absent.

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

278

11:28

Style file version Nov. 19th, 1999

Weinberg, Worth, and Williams

METHOD The hypotheses were tested with data from the New Zealand Male Call/Waea Mai, Tane Ma study carried out in 1996 by Heather Worth, Tony Hughes, and Clive Aspin. The study was hosted by the New Zealand AIDS Foundation and funded by the Health Research Council of New Zealand.

The Sample The target population of the survey was all men in New Zealand who have sex with men. In 1996, three staff were specifically hired to undertake the recruitment of participants. They contacted gay and bisexual community groups and individuals throughout the country who helped with recruitment in a variety of ways, such as distributing advertising materials, and holding meetings about the survey in their communities. As many different media as possible were utilized to advertise the study. This included advertisements in mainstream and gay magazines and newspapers, radio advertisements, and announcements on the primary 6:00 p.m. national TV news program as well as on the New Zealand Tonight Show. In the advertising and recruitment process, no reference to sex work or sex for money was made. A special effort was made to reach those men who have sex with men who are more difficult to locate: those who were not in major cities and those who were not attached to the gay community. The largest percentage of respondents (45%) heard about the study through the mainstream media, 40.7% through the gay press, 35.4% by word of mouth through friends, 28.2% through a gay TV program, and 27.3% through posters. The recruitment led 1852 men from every geographical area in New Zealand who have sex with men to participate in the HIV-related survey. Compared to the total New Zealand male population, the participants were more likely to come from major urban areas and less likely to be from smaller, and especially rural, areas. This reflects a general trend; many men who want to have sex with men move out of rural areas and small cities to live in the larger cities where such opportunities are greater.

The Procedure Advertisements provided a toll-free number for people to call who were interested in participating. Limitations are those associated with a self-selected sample of this type, for example, underrepresentation of those who are most estranged and being able to estimate what percentage of men have sex with men for money. However, we think self-selection problems would affect MSWs and OMSM equally in a country like New Zealand for the reasons presented in the section on their social environment.

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

11:28

HIV-Risk Comparison among MSWs and OMSM in New Zealand

Style file version Nov. 19th, 1999

279

Interviewers It was important to have interviewers who had the ability to elicit extremely personal information, to be sensitive to the needs of the respondents, and to detect hoax calls. Interviewers were selected by the investigators and given careful training. Interviewers participated in half-day training sessions at which they were taught how to handle difficult questions. Following the strategy used in an Australian study of men who have sex with men (Kippax et al., 1994), most interviewers were gay men recruited through the gay press. Women interviewers were hired, although respondents were offered the option of speaking to a man. The interviewers were chosen on the basis of their ability to ask sexually explicit questions in a nonjudgmental manner. Because of New Zealand’s large Maori and Pacific Island population, one Pacific Islander and two Maori bilinguals were also hired. The Interview Schedule/Measures The interview schedule was developed from the Australian Project mentioned above (Kippax et al., 1994). Because of cultural differences between Australia and New Zealand, the questionnaire was thoroughly pretested. The questionnaire was subsequently revised and retested based on the pretest and feedback from the gay community and sexual health professionals. The measures used for the present paper were as follows: Demographic information was obtained using the standard questions and answer categories used in census research in New Zealand. The ethnic and education categories, were, for example, NZ European, NZ Maori, Samoan, and Cook Island Maori. Sexual identity was ascertained by asking respondents if they personally identified with any of the following labels: Heterosexual, Straight, Bisexual, Gay, Homosexual, Queer, and indigenous terms such as, for example, Takataapui, a Maori term for “intimate friends of the same sex.” Attachment to the gay community was measured by directly asking if they see themselves personally as being part of the gay community. A multitude of questions were asked about sex practices with regular and casual men partners. The focus was on the degree to which the respondent engaged in unsafe anal and oral sex practices (e.g., anal intercourse without a condom and ejaculating inside, swallowing ejaculate). A number of questions were also constructed to look at unsafe sex with women, for example, the type of sexual relationship they were in with a woman (regular, casual), the number of women they had sex with, and frequencies of sex with women. Unsafe sex acts were specifically covered (e.g., for both casual and regular women partners, vaginal intercourse without a condom and ejaculating inside the woman, anal intercourse

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

280

11:28

Style file version Nov. 19th, 1999

Weinberg, Worth, and Williams

without a condom and ejaculating inside a woman). Also, information was gathered on whether they told sex partners who were women that they also had sex with men. With regard to unsafe drug use, respondents were asked how often they used needles for injecting illegal drugs and how often they had shared needles when injecting illegal drugs. In addition, respondents were asked if they ever had been tested for HIV, how many times they had been tested, and when was the last time. Most importantly, they were asked the latest test results. There was also a scale of 1–10 to gauge the respondents’ perceptions of their likelihood of infection. Finally, the question that determined whether the respondent would be placed in the MSW or OMSM group was whether or not he had been paid money to have sex with a man in the last 6 months. The sex work item reflected the often occasional rather than regular nature of sex work among men who have sex with men (cf. Prestage, 1994; Scambler, 1997).

Analysis Data were checked for inconsistencies and cleaned. The independent variable compared men who had ever been paid for sex in the last 6 months (N = 102) with those who had not (N = 1750). Cross-tabulations were run against the other variables. Since the differences that appeared on the items between the two groups were minimal, and any differences between the two groups always appeared to be in one answer location, usually the variables were simply dichotomized at this point of difference. For the measure of identity, however, because persons could use multiple categories, each category had to be considered by itself (scoring each person as Yes or No). Chi-square tests were employed to test for statistical significance. In the results section that follows, anything considered as a “difference” is statistically significant at least at the .05 level. The tables show the probability levels and the chi-square information for each run.

RESULTS Demographic Characteristics As found in other countries, in New Zealand the MSWs differed from the OMSM in terms of reports of their age, ethnicity, education, type of other paying jobs, and income. In comparison with the OMSM, the MSWs were younger, contained a higher percentage of minority members (in this case Maori), had less education, were more likely to be in blue collar jobs, and had lower incomes.

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

11:28

Style file version Nov. 19th, 1999

HIV-Risk Comparison among MSWs and OMSM in New Zealand

281

Identity and Attachment to the Gay Community There was no significant difference in the percentage of men who identified as heterosexual or straight in the two groups. However, a larger percentage of the MSWs than the OMSM identified themselves as bisexual and a smaller percentage as homosexual. Even though there was a difference in self-identification, the MSWs and the OMSM scored similarly on the measure of “attachment to the gay community.” There was only a 4% difference in the percentage falling in the “high” category (See Table I). Unsafe Sex Practices with Men The MSWs were less likely than the OMSM to report currently being in a regular sexual relationship with a man. On the other hand, given that they do sex work, as one might expect, more MSWs reported having sex with casual male partners: First, the MSWs reported a greater number of such partners both in the last 6 months, and in their lifetime. Second, the MSWs reported a higher frequency of casual sex with men (See Table II). Anal sex, potentially the most risky in terms of male transmission of HIV, was more likely to be engaged in with a man or a woman by the MSWs than by the OMSM. Of greater importance to the issue of infection, however, was that with a casual partner they were more likely to engage in unprotected receptive and insertive anal intercourse that included ejaculating inside the rectum. With regard to oral sex, the MSWs were also more likely to report swallowing their partner’s semen or ejaculating in their partner’s mouth in casual sex than were the OMSM. No significant differences were found between the two groups in such practices with regular partners. Table I. Comparison on Background Variables Measure

MSW

OMSM

χ2

df

p

Age < 30 Maori Less than college education Blue collar Income < NZ$1200 Heterosexual identity Straight identity Bisexual identity Homosexual identity Gay identity High gay community attachment

60.8% (102) 21.6% (102) 81.2% (101) 62.8% (76) 39.2% (102) 8.8% (102) 11.8% (102) 39.2% (102) 52.9% (102) 73.5% (102) 59.8% (102)

39.2% (1750) 8.5% (1745) 66.9% (1721) 38.0% (1397) 28.3% (1730) 6.2% (1759) 5.7% (1750) 26.7% (1750) 67.2% (1750) 80.0% (1751) 63.9% (1750)

17.765 18.215 8.286 18.068 5.019 .741 .213 16.439 9.436 2.860 .206

1 1 1 1 1 1 1 1 1 1 1

.001 .001 .01 .001 .05 .50 .70 .001 .01 .10 .70

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

282

11:28

Style file version Nov. 19th, 1999

Weinberg, Worth, and Williams Table II. Comparison on Dependent Variables Measure

MSW

OMSM

χ2

df

p

In a relationship with a man More than 20 male casual partners—last 6 months More than 200 male partners—lifetime Casual sex with men, more than 10 times a month Anal intercourse—last 6 months with a man or woman Receptive anal intercourse with a man—doesn’t always use condom and ejaculates inside rectum Casual partner Regular partner Active anal intercourse with a man—doesn’t always use condom and ejaculates inside rectum Casual partner Regular partner Swallowed partner’s ejaculate Casual partner Regular partner Ejaculated in male partner’s mouth Casual partner Regular partner Sex with women—last 6 months Didn’t disclose sex with men to women partners Injected illegal drugs—last 6 months Shared needle High on perceived “possibility” of infection Tested for HIV—last 6 months Tested multiple times Seropositive status

38.2% (102) 38.1% (97)

51.4% (1750) 9.9% (1748)

7.311 75.526

1 1

.01 .001

42.1% (95) 29.2% (89)

18.7% (1705) 17.7% (1254)

29.400 6.568

1 1

.001 .05

78.4% (102)

68.2% (1750)

4.246

1

.05

48.3% (29) 39.9% (18)

25.1% (345) 59.5% (444)

6.201 2.229

1 1

.05 .20

46.2% (39) 35.7% (14)

27.5% (404) 59.9% (456)

5.141 2.356

1 1

.05 .20

29.2% (89) 48.0% (25)

18.5% (1254) 48.0% (692)

5.474 0.000

1 1

.05 .99

51.7% (89) 44.0% (25) 31.4% (102) 66.6% (27)

32.1% (1254) 48.6% (717) 16.6% (1749) 75.5% (319)

13.404 0.000 15.639 1.456

1 1 1 1

.001 .99 .001 .30

14.9% (101) 26.6% (15) 36.4% (99)

1.7% (1746) 13.3% (30) 14.7% (1690)

63.866 .475 34.586

1 1 1

.001 .50 .001

59.9% (74) 88.9% (72) 2.8% (71)

31.2% (1229) 75.0% (1225) 4.4% (1204)

24.097 6.260 .882

1 1 1

.001 .05 .50

Unsafe Sex Practices with Women A significantly greater percentage of the MSWs than of the OMSM said they engaged in sex with women in the last 6 months. The number who reported engaging in potentially unsafe sex with women was too small (two, three, four, and ten for four different sex acts) to consider for significance tests. The percentage distributions, however, did not show a pattern of greater recklessness on the part of the MSWs. We were able to test for the statistical significance of a difference between the reports of the MSWs and OMSM as to whether they disclosed to their women sex partners that they had sex with men. There was no significant difference between the two groups.

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

11:28

HIV-Risk Comparison among MSWs and OMSM in New Zealand

Style file version Nov. 19th, 1999

283

Injected Drug Use A minority of the MSWs, but a significantly greater percent than the OMSM, reported injecting illegal drugs in the last 6 months. There was no significant difference in reports of sharing needles. Getting Tested and HIV Status The MSWs were more likely to think they might become infected with HIV. Also, they were more likely to report having been tested within the 6-month period before their interview and having been tested multiple times. There was, however, no significant difference in HIV status: 2.8% of the sex workers reported being seropositive, as did 4.4% of the OMSM. SUMMARY AND CONCLUSION In the introduction to this paper we described the “more similar than different” argument presented by sex workers. We tried to further develop their formulation by delineating the conditions under which it would hold true. We then suggested that New Zealand sustained such conditions. The first part of the sex workers’ position is supported: there were no significant differences between the MSWs and OMSM in reports of seropositive status. This is interesting in that dissimilarities appear between the groups that would lead one to expect a greater likelihood of HIV infection. As found in studies conducted in other countries (cf. Boles and Ellifson, 1994; Morse et al., 1991; Pleak and Meyer-Bahlburg, 1990; Prestage and Hood, 1993; Shaver and Newmeyer, 1996), the MSWs were younger, had lower income, were more likely to be in blue collar employment and from an ethnic underclass, and more likely to be bisexual—factors that studies suggest relate to higher risk (cf. Doll and Beeker, 1996; Kippax et al., 1994; Plant, 1997; Stokes et al., 1997). In addition, other factors that could add to risk were a greater number of casual male partners, a greater frequency of casual sex, and being less likely to be in a steady homosexual relationship. The MSWs were also more likely to report injecting drugs (although few said they shared needles). Most crucially, MSWs were more likely to report being unsafe in their anal sex with casual partners (as well as in their oral sex, not generally considered high risk by many in the AIDS field [cf. Dossier, 1990; Plant, 1997]). Yet, the above characteristics that add to risk did not result in a difference in their reported HIV status. We believe that the similarity in HIV status is explained by those characteristics of New Zealand that help to neutralize the possible health consequences of high risk. The most important of the contextual factors appears to be the small pool of infection. We believe that this, in turn, is related to a political system that

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

284

11:28

Style file version Nov. 19th, 1999

Weinberg, Worth, and Williams

is liberal, an economic structure without large pockets of abject poverty that are accompanied by an extensive IV drug culture, and sociocultural values that support an attitude of inclusiveness, a national health system and a great deal of public HIV education. Elsewhere, where there is a large pool of infection, this is usually associated with more poverty and illiteracy, a larger IV drug culture, less supportive government policies and laws, a poorer health system, and fewer AIDS education programs. In such locales, the sex worker’s greater vulnerability to HIV is more likely to be actualized in higher rates of seropositivity (for example, in Brazil, Cortes et al., 1989, found 44% of men sex workers to be HIV positive; among men from lower-class areas of New York City, Chiasson et al., 1988 found 53% of the sex workers to be seropositive). In conclusion, taking seriously the ideas of sex workers themselves led us to develop a viewpoint on HIV and carry out a study in New Zealand. While the findings support part of the sex workers perspective, they do not provide complete confirmation for it. New Zealand does draw attention to the local environment as a factor in the comparable HIV status of sex workers and others, and how this is possible even though their profiles of risk are dissimilar.

ACKNOWLEDGMENTS This research was partially supported by a grant from the Foundation for the Scientific Study of Sexuality. We wish to thank the Foundation for their support and the Archive’s anonymous referees for their helpful comments.

REFERENCES Alexander, P. (1995). Sex workers fight against AIDS: An international perspective. In Schneider, B., and Stoller, N. (eds.), Women Resisting AIDS: Feminist Strategies of Empowerment, Temple University Press, Philadelphia. Awareness of and Attitudes to Discrimination on Grounds of Sexual Orientation: A Survey of the New Zealand Population. (1992). Report prepared for the New Zealand AIDS Foundation. Bloor, M. J., McKeganey, N. P., Finlay, A., and Barnard, M. A. (1992). The inappropriateness of psychosocial models of risk behaviour for understanding HIV-related risk practices among Glasgow male prostitutes. AIDS Care 4: 131–137. Bloor, M. J., Barnard, M. A., Finalay, A., and McKeganey, N. P. (1993). HIV-related risk practices among Glasgow male prostitutes: Reframing concepts of risk behavior. Med. Anthropol. Q. 7: 152–169. Boles, J., and Elifson, K. (1994). Sexual identity and HIV: The male prostitute. J. Sex Res. 31: 39–46. Browne, J., and Minichiello, V. (1995). The social meanings behind male sex work: Implications for sexual interactions. Br. J. Sociol. 46: 598–622. Browne, J., and Minichiello, V. (1996). The social and work context of commercial sex between men: A research note. Aust. N. Z. J. Sociol. 32: 86–92. Chapkis, W. (1997). Live Sex Acts: Women Performing Erotic Labor, Routledge, New York.

P1: LMD Archives of Sexual Behavior [asb]

PP062-295248

February 16, 2001

11:28

HIV-Risk Comparison among MSWs and OMSM in New Zealand

Style file version Nov. 19th, 1999

285

Chiasson, M. A., Lifson, R. L., Stoneburner, W., Ewing, D., and Jaffe, H. W. (1988). HIV seroprevalence in male and female prostitutes in New York City. In Proceedings of the IV International Conference on AIDS, Stockholm, June. Cortes, E., Detels, R., Aboulafia, D., Li, X. L., Moudgil, T., Alam, M., Bonecker, C., Gonzaga, A., Oyafuso, L., and Tondo, M. (1989). HIV-1, HIV-2, HTLV-1 infection in high risk groups in Brazil. Aust. N. Z. J. Med. 320: 953–958. Davies, P., and Feldman, R. (1997). Prostitute men now. In Scambler, G., and Scambler, A. (eds.), Rethinking Prostitution: Purchasing Sex in the 1990s, Routledge, New York. De Graaf, R., Vanwesenbeeck, I., van Zessen, G., Straver, C., and Visser, J. (1994). Male prostitutes and safe sex: Different settings, different risks. AIDS Care 6: 277–288. Delacoste, F., and Alexander, P. (eds.). (1988). Sex Work: Writings by Women in the Sex Industry Virago, London. Doll, L. S., and Beeker, C. (1996). Male bisexual behavior and HIV risk in the United States: Synthesis of research with implications for behavioral interventions. AIDS Educ. Prevent. 8: 205–225. Dossier, P. (1990). The 3rd Epidemic: Repercussions of the Fear of AIDS, Panos Institute, London. English Collective of Prostitutes. (1997). Campaigning for legal change. In Scambler, G., and Scambler, A. (eds.). Rethinking Prostitution: Purchasing Sex in the 1990s, Routledge, New York. Jenness, V. (1993). Making it Work: The Prostitutes’ Rights Movement in Perspective, Aldine de Gruyter, New York. Kaplan, H. B., Johnson, R. J., and Simon, W. (1987). The sociological study of AIDS: A critical review of the literature and suggested research agenda. J. Health Soc. Behav. 28: 140–157. Kippax, S., Crawford, J., Rodden, P., and Benton, K. (1994). Report on Project Male-Call: National Telephone Survey of Men Who Have Sex with Men, Australian Government Publishing Service, Canberra. Kruhse-Mountburon, S. (1992). An evaluation of Australian legislation on prostitution. In Van de Vort, P., and Bruylant, P. M. (eds.), La Prostitution, Quarante ans Apres La Convention de New York, Ecole des Sciences Criminologiques Leon Cornil, Brussels. Mann, J. M., and Tarantola, D. J. M. (1996). AIDS in the World II: Global Dimensions, Social Roots, and Responses, Oxford, New York. McKeganey, N. S., and Barnard, M. (1992). Selling sex: Female street prostitution and HIV risk behavior in Glasgow. AIDS Care 4: 395–407. Morse, E. V., Simon, P. M., and Gaumier, H. R. (1991). A vector for transmission of HIV infection into the heterosexual world. Soc. Sci. Med. 32: 535–539. O’Neill, M. (1997). Prostitute women now. In Scambler, G., and Scambler, A. (eds.), Rethinking Prostitution: Purchasing Sex in the 1990s, Routledge, New York. Overs, C., and Longo, P. (1997). Making Sex Work Safe, AHRTAG, London. Pheterson, G. (1996). The Prostitution Prism, Amsterdam University Press, Amsterdam. Plant, M. (ed.). (1990). AIDS, Drugs and Prostitution, Routledge, London. Plant, M. (1997) Alcohol, drugs and social milieu. In Scambler, G., and Scambler, A. (eds.), Rethinking Prostitution: Purchasing Sex in the 1990s, Routledge, New York. Pleak, R., and Meyer-Bahlburg, H. F. L. (1990). Sexual behavior and AIDS knowledge of young male prostitutes in Manhattan. J. Sex Res. 27: 557–587. Prestage, G. (1994). Male and transsexual prostitution. In Perkins, R., Prestage, G., Sharp, R., and Lovejoy, F. (eds.), Sex Work, Sex Workers in Australia, UNSW Press, Sydney. Prestage, G., and Hood, D. (1993). Targeting non-gay attached homosexually-active working class men. Paper presented at the Social Aspects of AIDS Conference, London. Scambler, G. (1997). Conspicuous and inconspicuous sex work: The neglect of the ordinary and mundane. In Scambler, G., and Scambler, A. (eds.), Rethinking Prostitution: Purchasing Sex in the 1990s, Routledge, New York. Shaver, F. M., and Newmeyer, T. (1996). Men who have sex with men: A comparison of the sexual practices and risk-taking behavior of gay and bisexual men and male prostitutes. Paper presented at Sida, jeunesse et prevention. Au-dela du discours, des actions! McGill University, Montreal. Stokes, J. P., Vanable, P., and McKirnan, D. J. (1997). Comparing gay and bisexual men on sexual behavior, condom use, and psychosocial variables related to HIV/AIDS. Arch. Sex. Behav. 26: 377– 391. Vanwesenbeeck, I. (1994). Prostitutes Well-Being and Risk, VU University Press, Amsterdam.

P1: LMD Archives of Sexual Behavior [asb]

286

PP062-295248

February 16, 2001

11:28

Style file version Nov. 19th, 1999

Weinberg, Worth, and Williams

Vanwesenbeeck, I. (1995). Professional HIV risk taking, levels of victimization, and well-being in female prostitutes in the Netherlands. Arch. Sex. Behav. 24: 503–515. Vanwesenbeeck, I., van Zessen. G., de Graf, R., and Straver, C. J. (1994). Contextual and interactional factors influencing condom use in heterosexual prostitution contacts. Patient Educ. Couns. 24: 307–322. Von den Hock, J. A. R., and Coutinho, R. A. (1991). Homosexual prostitution among male drug users and its risk for HIV infection. Genitourin. Med. 67: 303–306. Waldorf, D., Murphy, S, Lauderback, D., Reinarman, C., and Marotta, T. (1990). Needle sharing among male prostitutes: Preliminary findings of the Prospero Project. J. Drug Iss. 20: 309–334.

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

February 23, 2001

8:53

Style file version Nov. 19th, 1999

Archives of Sexual Behavior, Vol. 30, No. 3, 2001

Negotiating Risks in Context: A Perspective on Unprotected Anal Intercourse and Barebacking Among Men Who Have Sex with Men—Where Do We Go from Here? Troy Suarez, Ph.D.1,2 and Jeffrey Miller, MSN, ACRN, APNP1

Recently, an alarming trend toward unprotected anal intercourse has emerged in men who have sex with men. A highly dangerous form of unprotected anal intercourse, barebacking—the deliberate and conscious choice to engage in anal sex without condoms knowing that there are risks involved—has received much attention in the gay press. This trend poses new challenges for HIV prevention. As the target population changes, prevention interventions must also evolve to remain effective. A review of the scientific literature on risk behaviors and the popular literature and websites devoted to barebacking suggests that many contextual factors influence the decision to engage in unprotected anal intercourse. This review examines the most salient contextual factors affecting risk behavior in gay/bisexual men. It also identifies four main cohorts and predominant contextual factors that appear to motivate unprotected anal intercourse in each. In answering the question “where do we go from here,” we conclude that contextual issues must be addressed in hybrid prevention interventions that include harm reduction, motivational interviewing, and traditional approaches. KEY WORDS: UAI; barebacking; prevention; HIV; MSM.

Because of intensive and targeted HIV-risk reduction interventions that emphasized the high risk of unprotected anal intercourse (UAI) and promoted consistent condom use, rates of UAI fell dramatically among homosexually active men throughout the 1980s and early 1990s. However, as HIV incidence rates declined 1 Center for AIDS Intervention Research, Department of Psychiatry and Behavioral Medicine, Medical

College of Wisconsin, Wisconsin. whom correspondence should be addressed at Ingenix Pharmaceutical Services, 1160 Parsippany Blvd., Parsippany, New Jersey 07054; e-mail: [email protected].

2 To

287 C 2001 Plenum Publishing Corporation 0004-0002/01/0600-0287$19.50/0 °

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

288

PP062-295249

February 23, 2001

8:53

Style file version Nov. 19th, 1999

Suarez and Miller

and leveled off among men who have sex with men (MSM), preventionists began to redirect their attention to other at-risk groups. Currently, less than 13% of empirically validated prevention interventions in the US are directed at men who have sex with men, even though MSM account for approximately 50% of current AIDS cases and remain the largest at-risk population (Sogolow et al., in press). Although assessing trends in risk behavior and HIV incidence is difficult given the poor quality of recent surveillance data in this population, current behavioral and epidemiological data indicate increased risk behavior with casual and anonymous partners both in the US and abroad (Centers for Disease Control [CDC], 1999a; Elford et al., 2000; Prestage et al., 1997; Schechter et al., 1998; Valleroy et al., 1998; van de Ven et al., 1998), increased incidence of sexually transmitted diseases (STDs) (CDC, 1997, 1999a,b, 2000), and, in at least one city where data are available, the possibility of increased HIV infection rates (San Francisco Department of Public Health, 2000). In addition to empirical support for increasing risk behavior in MSM, UAI in the form of “barebacking” has recently received much attention in the media. Barebacking—the deliberate and conscious choice to engage in risky sexual behavior knowing that there are risks involved—has garnered an almost cult-like following, with its own slang terminology, professional pornography, websites, and e-mail listservs. Generally, barebacking refers to UAI between casual and anonymous partners, but may also encompass oral sex with swallowing of the semen, and gloveless fisting. The term is less often used to connote UAI between committed partners. Because of its sensationalistic nature, it remains unclear whether individuals who identify as barebackers are the same individuals who failed to use condoms throughout the epidemic, or if this group includes individuals engaging in UAI for the first time. Nonetheless, the visibility of this group in the media has the potential to shift safer sex norms within the gay community as a whole (Wolitski et al., in press). The motivation for engaging in UAI with casual and anonymous partners may differ significantly from the motivation for engaging in UAI with regular partners. Whereas UAI between primary partners is heavily influenced by desires to express intimacy, trust, and love, the same behavior between casual/anonymous partners is most probably not affected equally by these same influences. The contexts that surround the individual, the sexual interaction, and the interpretation of the sexual interaction may then highly influence AIDS preventive behavior (Suarez and Kauth, in press). For the purposes of the present review, contextual issues include the cognitive, behavioral, and environmental factors that motivate one toward or away from AIDS preventive behavior. Examining risk behavior outside of these contexts provides insufficient information. Not all risk behavior decisions are irrational. Pinkerton and Abramson (1992) suggest that risky sex can be a “rational” decision if the benefits of risky sex (e.g., pleasure, intimacy) outweigh the perceived threat of AIDS (including perception

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

UAI and Barebacking in MSM

February 23, 2001

8:53

Style file version Nov. 19th, 1999

289

of risk and disease severity). Many factors can influence an individual’s decision whether or not to engage in risky sexual practices. For example, some men believe that risks vary as a function of factors such as partner characteristics (e.g., long term, monogamous seronegative partner), biology (e.g., viral load of the infected person), sexual practices (e.g., withdrawal, insertive or receptive UAI), and antiHIV measures (e.g., spermicides, condoms) (Suarez and Kauth, in press). These men may then believe they are reducing their HIV risk by engaging only in selected behaviors with particular partners. Although the efficacy of this strategy in reducing HIV risk has been debated by experts (e.g., Cook and Rosenberg, 1998; Ekstrand et al., 1993; Kippax et al., 1996; Pinkerton and Holtgrave, 1999; Wittkowski et al., 1998), many MSM perceive some of these methods to be effective. Resultantly, they use these factors to rationalize behavior and negotiate a comfortable level of risk (Suarez and Kauth, in press). Therefore, it is important to understand contextual influences in order to develop effective prevention programs that are salient in today’s environment. We have only just recently begun to understand the importance of contextual influences on sexual decision-making (Suarez and Kauth, in press). A small but growing empirical literature exists, examining how contexts affect sexual behavior. Therefore, much of the information on what issues are important and how they affect behavior must come from combining both the empirical literature and anecdotes from the popular press.3 A comprehensive review of the Anglo-American epidemiological, medical, nursing, psychiatric, and psychological literatures related to the determinants of UAI, as well as several websites and listservs devoted to barebacking, reveals the significance of several evolving contextual influences on risk behavior decision-making. Recently, the most widely discussed contextual factor affecting sexual decision-making has been the improved medical management of HIV, i.e., highly active antiretroviral therapies (HAART). Several studies in the US and abroad have demonstrated that the effectiveness of HAART has caused many men to reevaluate their adherence to safer sex practices (Dilley et al., 1998; Elford et al., 2000; Kelly et al., 1998; Murphy et al., 1999; Remien et al., 1998; Vanable et al., 2000). Some men believe that HIV-infected persons who are receiving HAART and have low or undetectable viral loads are less likely to transmit HIV to others, and are therefore safer as sexual partners (Kalichman et al., 1998; Suarez et al., 2000). Other men feel that HIV is now a manageable disease and no longer synonymous with certain death (Dilley et al., 1997; Kelly et al., 1998). Both of these beliefs may be associated with UAI in MSM (Murphy et al., 1999). 3 Although

predictors of UAI in gay/bisexual men have been thoroughly researched, little research exists on predictors of the current rise in UAI or into the phenomenon of barebacking. Because of the timeliness of this topic, we felt it important to integrate empirical works with current peer-reviewed conference presentations and anecdotal reports from the popular media. Integration of popular media accounts serves to humanize this new trend by adding personal accounts to the provided research.

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

290

PP062-295249

February 23, 2001

8:53

Style file version Nov. 19th, 1999

Suarez and Miller

The belief that HAART might reduce infectivity may not be totally erroneous. There is limited evidence that HAART reduces transmission risk (Pinkerton and Holtgrave, 1999). In particular, for many individuals taking certain medications, suppression of peripheral viral load to undetectable levels also results in suppression of seminal viral load (Barroso et al., 2000; Eron et al., 2000; Gupta et al., 1997; Vernazza et al., 2000) and a few studies examining the effects of antiretroviral therapy on heterosexual transmission of HIV have demonstrated that certain antiretrovirals can reduce HIV transmission during vaginal intercourse from an infected partner to an uninfected partner (Chirianni et al., 1992; Musicco et al., 1994). Further, peripheral viral load is positively correlated with infectiousness (Fiore et al., 1997; Lee et al., 1996; Quinn et al., 2000; Ragni et al., 1998; Royce et al., 1997). Still, the effect of HAART on transmission risk during UAI is not known, and not everyone with an undetectable peripheral viral load also has a corresponding undetectable seminal viral load. Moreover, drug resistance for virus found in the semen can occur before peripheral resistance is evidenced (Barroso et al., 2000). Therefore, disregarding safer sex practices based on assumptions of decreased infection risk could increase the risk of infection, perhaps with a strain that is resistant to current and possibly future medications. Another contextual factor that may work synergistically with HAART in reducing the fear of AIDS is AIDS burnout. Years of coping with HIV and changes to natural sexual behavior patterns can lead to burnout (Wolitski et al., in press). Hope that HIV may be cured or effectively maintained long-term can lead to increases in risky behavior (Kalichman, 1998). Ostrow and colleagues (2000) reported that burnout from years of exposure to prevention messages and trying to maintain safer sex practices was an independent predictor of UAI among HIVpositive MSM in four U.S. cities. Barebackers sometimes report fatigue with safer sex and the lack of meaningful prevention programs as a reason for barebacking. It is possible that improved treatments work synergistically with burnout to produce risk behavior in MSM (e.g., Murphy et al., 1999). Other contextual issues, although supported anecdotally, have less empirical support. For example, the eroticisation of HIV may play a large role in changing the community perception of AIDS, which may then lead to increasing risk behavior. Gone are references to AIDS as the grim reaper. Today’s images of persons living with HIV are of healthy, virile, and physically fit individuals. Although normalizing HIV is beneficial on many fronts, these images may serve to change social perceptions of HIV from a devastating killer to a less serious and manageable condition. Media accounts of barebacking may serve the same purpose. It is believed that barebackers represent a small but vocal group; however, the relatively positive media attention given to this group also has the potential to shift safer sex norms within the gay community as a whole (Wolitski et al., in press). Several of these factors together with many other issues might contribute to the noted increases in sexual risk behavior among MSM (McKirnan et al., 1996;

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

February 23, 2001

8:53

Style file version Nov. 19th, 1999

UAI and Barebacking in MSM

291

Suarez and Kauth, in press). We must better understand the contextual issues that drive risk behavior if we are to develop more effective and humanistic prevention interventions (Odets, 1994, 1996). If not, we are in danger of erasing the substantial prevention gains of the past. WHERE DO WE GO FROM HERE? CONTEXTUAL ISSUES, MOTIVATION, AND HIV PREVENTION Thus far, we presented the idea that men negotiate the level of risk that they are willing to take based on contexts. We also identified a few evolving contextual issues that may factor into these negotiations. Although scientific advances outwardly appear to strongly influence behavior other contextual factors also facilitate risk behavior. Based on this review, we identified several groups of men who differ on level of negotiated risk and predominate contextual influences. This (hypothetical) typology includes four groups of gay/bisexual men who engage in UAI, each of which may require different prevention approaches. Most contextual factors play a role to some extent in UAI among the four groups; however, the importance of each contextual factor varies between groups. Seroconcordant Couples The first group consists of men who engage in UAI within committed HIVseroconcordant relationships. Whether seropositive or seronegative, UAI in these dyads appears to be motivated by feelings of love and trust, and a desire to express emotional intimacy. For monogamous non–drug-using seronegative couples, the risk of HIV infection is infinitesimal (Kippax et al., 1993, 1996). Alternatively, HIV-positive couples may perceive condoms to be irrelevant because reinfection has not been empirically proved and may be extremely rare (NASTAD, 1999). Because these relationships are monogamous, the risk of introducing HIV, hepatitis, or other STDs is small. If one or the other partners commits infidelity, however, then UAI within such a relationship can be very risky. Because UAI is rooted in love, trust, and intimacy, behavior change in this group is highly unlikely—just as in heterosexual couples. The pleasure of UAI acts to reinforce this behavior in this group (Kelly and Kalichman, 1998). Rational Risk Takers The second group consists of those men whose behavior is based on a “rational” consideration of the risks of specific sexual acts. This group includes nonmonogamous couples with negotiated safety-like agreements, and HIVnegative individuals whose only UAI is insertive—all receptive anal intercourse is

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

February 23, 2001

8:53

292

Style file version Nov. 19th, 1999

Suarez and Miller

protected (e.g., Suarez and Kauth, in press; Wagner et al., 1998). Other individuals in this group might include men who rely on discussing HIV status with potential partners and only engaging in risk behavior with those that they believe are of a similar serostatus—“serosorting” (Suarez et al., 2000). Serosorting is not without complications, in that disclosure is not always accurate or honest (Cochran and Mays, 1990; Rowatt et al., 1998) and some individuals do not know they are infected. Moreover, those men who mislead potential partners about their status are also more likely to engage in risk behavior with discordant or unknown partners (Wolitski et al., 1998; Wulfert et al., 1999). Yet, many men continue to use this strategy to prevent infection. Many barebackers employ the strategy of serosorting. Several personal ads on barebacking websites explicitly state HIV status and the desired status of potential partners. Some have theorized that barebacking itself is a serosorting strategy in that barebacking is just a way for HIV-positive men to disclose their status without having to actually tell another they are HIV-positive (Hort, 2000). However, many barebacking ads are from men who claim to be HIV-negative. Nevertheless, many barebackers specify that they will only bareback with individuals of a similar serostatus. In this group, there is a very real possibility of HIV, hepatitis, or other STD infection. However, the risks may be reduced because of harm reduction activities that are put in place. These men view UAI as calculated risk. The behavior of these men may be more resistant to change because of comfort with their level of negotiated risk. For this group, the underlying contextual issues appear to be the weighing of scientific data on infectivity, the reinforcement value of UAI, and possibly AIDS burnout. For couples, intimacy is also a major contextual issue that motivates UAI. Irrational Risk Takers The third group consists of those individuals who deny their own risk or who use nonscientific or irrational information when making decisions regarding UAI. This group would include men who engage in receptive UAI with anonymous partners and also individuals who use faulty information, such as physical appearance to guess their partner’s HIV status. Many of these individuals base their decisions on faulty heuristics (e.g., “if God wanted me infected, I would be already” or “I am already infected and I don’t care about infecting anyone else”), or act out of poor self- and other respect (Dilley et al., 2000, Gold et al., 1990). The confirmation of status bias also can greatly affect an individual’s decision to engage in risk behavior. This bias allows individuals to convince themselves that their partner’s serostatus is the same as their own, based upon nonverbal cues (e.g., appearance, behavior of the partner, setting) (Suarez et al., 2000). For example,

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

February 23, 2001

UAI and Barebacking in MSM

8:53

Style file version Nov. 19th, 1999

293

an HIV-positive person might think, because a sexual partner did not suggest using a condom, that the sexual partner also must be HIV-positive. A seronegative person might interpret the same situation to mean that his partner also is uninfected. Like rational risk takers, individuals in this group appear to self-negotiate the level of risk they are willing to tolerate. The problem lies in the faulty heuristics used to gauge risk. The logic of this group appears to be convoluted and/or obscured by the heat of the moment, pure pleasure, intimacy issues, and sensation seeking (Ekstrand and Coates, 1990; Kelly and Kalichman, 1998). Our examination of websites also leads us to believe that alcohol and drug use heavily influence risk takers in this group. It appears that men in this group are mired in contextual issues that must be addressed prior to or simultaneously with HIV prevention efforts (Purcell et al., 1998).

YMSM The final group consists of young men who have sex with men (YMSM) who engage in UAI. Their lack of experience with the devastation of AIDS and their pessimistic view of the future may lead to risky behavior. Although YMSM do not have enough history with AIDS to be “burnt out” per se, these individuals may still adopt riskier sexual practices out of lack of respect for this potentially fatal disease. Seal and colleagues (2000) reported that YMSM feel pessimistic about aging as gay men and may view HIV as a way of escaping a dreaded future. Scarce (1999) reported that some YMSM with whom he talked actually set out to become infected, and some went so far as to select a “father” for their virus so that they could control how and when they got infected rather than leaving it to chance. Given the homophobic pressures that YMSM must manage when coming out, it is not surprising that these young men feel less than optimistic about their future. It appears that short of therapeutic intervention, only societal interventions aimed at reducing homophobia and promoting healthy and dynamic gay role models will help reduce fatalistic thinking in this group.

POSSIBLE PREVENTION DIRECTIVES FOR THOSE UNWILLING TO USE CONDOMS CONSISTENTLY The above cohorts have distinct intervention needs based on predominant contextual factors. As MSM continue to exhibit complex sexual patterns, HIV prevention must evolve to accommodate these trends. Previous prevention campaigns that stressed risk reduction through condom use or abstinence are becoming less effective as they have neglected to incorporate important aspects of sex and have been criticized for mechanizing and dehumanizing sexual behavior

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

February 23, 2001

8:53

Style file version Nov. 19th, 1999

294

Suarez and Miller

(McKirnan et al., 1996; Odets, 1994, 1996). Unfortunately, the very individuals in the greatest need of HIV prevention interventions are those least likely to attend traditional interventions (McKirnan et al., 1996). Therefore, new and innovative ways are needed of delivering prevention messages that will excite and motivate those people who have thus far been unwilling to adopt safer sex practices or who have returned to condomless sex after years of safer sex. A hybrid approach to prevention that includes elements of harm reduction, motivational interviewing, and traditional approaches may prove most fruitful. Changing Behavior in Unwilling Individuals A harm reduction approach instructs individuals to weigh the value of an activity against the costs of taking risks. This approach may provide better results for those unwilling to adopt the “condoms only” message (Odets, 1994, 1996). Although, historically, harm reduction campaigns have not been associated with sexual behavior, it is believed that this type of approach to sexual risk may provide benefits (Gold, 1995; Pinkerton and Abramson, 1992). A harm reduction approach to sex would suggest alternative risk reduction strategies that may afford “minimal” protection (e.g., early withdrawal, and vaccinations for hepatitis) or less than optimal protection (e.g., engaging in UAI only as the insertive partner). Even occasional condom use can greatly reduce the risk of HIV transmission (Pinkerton and Abramson, 1996). The advantage of a harm reduction approach for those unwilling to adopt condoms is fourfold. First, lasting behavior change is more likely because the client defines the self-negotiated strategy (Odets, 1996). Second, it meets the client where he is, not where the interventionist wants him to be, and then begins to move him towards a specified goal. Third, it provides tools for the individual to use if, and when, he does become motivated to change his behavior. Finally, it provides a nonjudgmental atmosphere from which change can occur. Judgmental approaches are unlikely to be successful in the long run and may add to the stigmatization and discrimination already faced by MSM, thereby increasing risk behavior. A harm reduction approach integrates well with motivational interviewing, which asserts that behavior is not constant and that with all maladaptive behavior there is some level of ambivalence (Denning, 1998). This ambivalence can be used as a tool to reduce resistance and motivate individuals to change risk behavior (AIDS Alert, 1999; Denning, 1998; Suarez and Kauth, in press). Just as with drug use, for some, sexual behavior is a way of coping. Therefore, those who are unwilling to give up risky sex must be approached similarly to those who use alcohol and drugs to cope. As a client begins to move towards a position of protecting himself, he will be more open to traditional efforts that focus on condom use (Suarez and Kauth, in press).

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

February 23, 2001

8:53

Style file version Nov. 19th, 1999

UAI and Barebacking in MSM

295

Changing Behavior in Motivated, but Habitually Risky Individuals Prevention interventions aimed at increasing self-worth and self-esteem may prove more effective for those individuals who are motivated to change their behavior but remain hampered by myriad contextual issues (Gerhart, 1999; Odets, 1994, 1996). Seal and colleagues (2000) provide a framework for interventions that go beyond what is traditionally thought of as an HIV prevention intervention. They suggest creating comprehensive sexuality programs that include the management of sexual arousal and the promotion of healthy and responsible sexual choices. They also suggest that effective interventions should be peer driven, focus on getting gay and bisexual men to care more about themselves, and include mentor programs for YMSM. Interventions should also incorporate decision-making between partners and should shift the focus from a binding intervention to one that enhances freedom and responsibility (Gerhart, 1999). Moreover, some individuals would benefit from longer-term therapy approaches that address long-standing personality factors that may also contribute to continued risk behavior (Suarez and Kauth, in press). Selective serotonin reuptake inhibitors (SSRIs) may also be helpful for reducing risk behavior in motivated individuals (Suarez, O’Leary, and Morganstern, in press). Because of their ability to reduce sexually compulsive behaviors (for review see Kafka, 2000), SSRIs may also be useful in reducing HIV transmission risk behavior (Kalichman et al., 1997). These medications have the added advantage of stabilizing mood which can also indirectly decrease risk behavior. Current research is looking at whether SSRIs are effective in reducing HIV transmission risk. These medications have demonstrated efficacy in reducing the total number of sexual partners, which is a HIV risk factor. See Suarez, O’Leary, and Morganstern (in press) for an extensive analysis of the use of SSRIs for HIV prevention. General Prevention Directives Interventions are needed that stress the limitations of current antiretroviral medications. These medications do not work for everyone, and even when they do work, they can cause deleterious side effects (e.g., body dysmorphic disorders related to HAART, i.e., lipodystrophy) and can be difficult to tolerate (Rabkin and Chesney, 1999). Interventions must also drive home the fact that almost 25% of newly infected persons are resistant to all three current classes of medications, and nearly 80% display resistance to at least one class of antiretrovirals (Voelker, 2000). This severely limits current and possibly future treatment options (Hecht et al., 1998; Wainberg and Friedland, 1998). Issues such as HAART and infectivity, risk associated with oral sex, and the correlation between viral load in the blood and semen are currently being investigated. It is better to say that the risk are unclear than to be discredited later.

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

February 23, 2001

296

8:53

Style file version Nov. 19th, 1999

Suarez and Miller

Although much attention has been directed toward helping HIV-negative MSM stay uninfected, relatively less attention has been given to the prevention needs of HIV-positive men. Medication advances are allowing many HIV-positive men to live longer and healthier lives, creating a need for primary prevention in this population (Kelly et al., 1998). Prevention messages should communicate the benefits of safer sex for people living with HIV, such as reducing the risk of STDs and the possibility of “reinfection” with another strain of HIV. Moreover, HIV prevention for people living with HIV needs to go beyond basic transmission issues to address other concerns that HIV-positive individuals must face (see Purcell et al., 1998). For example, prevention with this group should also address other health-related issues, such as barriers to HIV treatment adherence and access to healthcare. Because sexual behavior is heavily influenced by the context in which it occurs, interventions would be more effective if they helped HIV-positive men to manage the many relationship and individual issues that can lead to risky behaviors. Encouraging open discussion of serostatus is also important to dispel the confirmation of status bias (Suarez et al., 2000). Too often, MSM base risk behavior decisions on the attractiveness of the partner and the behaviors in which the partner is willing to engage (Gold et al., 1990). Making men aware of this bias and the fact that both HIV-positive and HIV-negative persons interpret an individual’s serostatus to be similar to their own when engaging in risky behavior, may help to reduce the effect of this bias. Finally, better surveillance and epidemiological data are needed. How risky is it for an HIV negative man to have insertive UAI with his HIV-positive partner who has an undetectable viral load? How risky is insertive UAI, or receptive oral sex for that matter? CONCLUSION UAI and barebacking may soon take their toll through increasing HIV infections, viral resistance, STDs, and rising treatment costs. Some level of ambivalence exists for most individuals who engage in risky behaviors. MSM must be motivated to change their risky behavior by directly dealing with their ambivalence to change. By integrating contextual factors into hybrid prevention campaigns that include harm reduction, motivational interviewing, and traditional approaches, the real-world effectiveness of interventions can be increased. This approach may also reach more MSM, not just those who are willing and able to adopt traditional safer sex practices. Many MSM are already using these harm reduction and negotiation strategies (e.g., Kippax et al., 1993, 1996; Gunther-Grey et al., 2000; Dilley et al., 2000); therefore, those behaviors already in place must be built upon to increase the impact of interventions.

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

February 23, 2001

8:53

UAI and Barebacking in MSM

Style file version Nov. 19th, 1999

297

ACKNOWLEDGMENTS This research was supported by Grant No. T32-MH19985 from the National Institute of Mental Health (NIMH) and by the NIMH center, grant, No. P30MH52776. Special thanks to Dr. Steven D. Pinkerton for his insightful comments, editorial assistance, and encouragement. We would also like to send special thanks to Seth Kalichman for his inspiration and assistance. Finally, we would also like to thank Drs. David Seal and Tim Heckman for their thought provoking and insightful comments, and Janice Sherman for her editorial assistance. REFERENCES AIDS Alert (1999). Behavior change model promotes HIV drug compliance, safe sex practices. AIDS Alert 14: 49. Barroso, P. F., Schechter, M., Gupta, P., Melo, M. F., Vieira, M., Murta, F. C., Souza, Y., and Harrison, L. H. (2000). Effect of antiretroviral therapy on shedding in semen. Ann. Int. Med. 133(4): 280– 284. Centers for Disease Control (1997, September 26). Gonorrhea among men who have sex with menselected sexually transmitted disease clinics, 1993–1996. Morb. Mort. Week. Rep. 46: 889–892. Centers for Disease Control (1999a, January 29). Increases in unsafe sex and rectal gonorrhea among men who have sex with men—San Francisco, California, 1994–1997. Morb. Mort. Week. Rep. 48: 45–48. Centers for Disease Control (1999b, September 10). Resurgent bacterial sexually transmitted disease among men who have sex with men—King County, Washington, 1997–1999. Morb. Mort. Week. Rep. 48: 773–776. Centers for Disease Control (2000, June 23). Gonorrhea—United States, 1998. Morb. Mort. Week. Rep. 48: 538–542. Chiranni, A., Perna, E., Liuzzi, G., D’Abbraccio, M., Bonadies, G., Paone, G., Foggia, M., Patricelli, V., and Piazza, M. (1992, Abstract No. PoC4530). Absence of anti-HIV seroconversion in heterosexual partners of HIV patients treated with zidovudine. Paper presented at the XI International Conference on AIDS. Cochran, S. D., and Mays, V. M. (1990). Sex, lies, and HIV. N. Engl. J. Med. 322: 774–775. Cook, R. L., and Rosenberg, M. J. (1998). Do spermicides containing nonoxynol-9 prevent sexually transmitted infections? A meta-analysis. Sex Transm. Dis. 25: 144–150. Denning, P. (1998). Therapeutic interventions for individuals with substance use, HIV, and personality disorders: Harm reduction as a unifying approach. In Session 4: 37–52. Dilley, J. W., Woods, W. J., and McFarland, W. (1997). Are advances in treatment changing views about high-risk sex? N. Engl. J. Med. 337: 501–502. Ekstrand, M., and Coates, T. J. (1990). Maintenance of safer sex behaviors as predictors of risky sex. Am. J. Pub. Health 80: 973–977. Ekstrand, M., Stall, R., Kegeles, S., Hays, R., DeMayo, M., and Coates, T. (1993). Safer sex among gay men: What is the ultimate goal. AIDS 7: 281–282. Elford, J., Bolding, G., Maguire, M., and Sherr, L. (2000). Combination therapies for HIV and sexual risk behavior among gay men. J. Acq. Imm. Defic. Syndr. 23: 266–271. Eron, J. J., Smeaton, L. M., Fiscus, S. A., Gulick, R. M., Currier, J. S., Lennox, J. L., D’Aquila, R. T., Rogers, M. D., Tung, R., and Murphy, R. I. (2000). The effects of protease inhibitors therapy on Human Immunodeficiency Virus Type 1 levels in semen (AIDS Clinical Trials Group Protocol 850). J. Infect. Dis. 181(5): 1622–1628. Fiore, J. R., Zhang, Y. J., Bjorndal, A., Di Stefano, M., Angarano, G., Pastore, G., and Fenyon, E. M. (1997). Biological correlates of HIV-1 heterosexual transmission. J. Acq. Imm. Defic. Syndr. 11: 1089–1094.

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

298

PP062-295249

February 23, 2001

8:53

Style file version Nov. 19th, 1999

Suarez and Miller

Gerhart, K. (1999, June). Beyond Barebacking: The complacency crisis. Genre 7: 47–51. Gold, R. (1995). Why we need to rethink AIDS education for gay men. AIDS Care 7: S11–S19. Gold, R. S., Skinner, M. J., Grant, P. J., and Plummer, D. C. (1990). Situational factors and thought processes associated with unprotected intercourse in gay men. Psychol. Health 5: 259–278. Gunther-Grey, C., Lyles, C., Sumartojo, K. H., Choi, K. H., Clark, L., Collens, C., Lin, L., O’Donnell, L., and Remafedi, G. (2000). Knowledge of partners HIV serostatus among men who have sex with men. Paper presented at the XIII International Conference on AIDS. Gupta, P., Mellors, J., Kingsley, L., Riddler, S., Singh, M. K., Schreiber, S., Cronin, M., and Rinaldo, C. R. (1997). High viral load in semen of human immunodeficiency virus type 1-infected men at all stages of disease and its reduction by therapy with protease and nonnucleoside reverse transcriptase inhibitors. J. Virol. 71: 6271–6275. Hecht, F. M., Grant, R. M., Petropoulos, C. J., Dillon, B., Chesney, M. A., Tian, H., Hellmann, N. S., Bandrapalli, N. I., Digilio, L., Branson, B., and Kahn, J. O. (1998). Sexual transmission of an HIV-1 variant resistant to multiple reverse-transcriptase and protease inhibitors. N. Engl. J. Med. 339: 307–311. Hort, H. K. (2000). Barebacking domino. Paper presented at the XIII International Conference on AIDS. Kafka, M. P. (2000). The paraphilia-related disorders: Nonparaphilic hypersexuality and sexual compulsivity/addiction. In Lieblum, S., and Rosen, R. C. (eds.), Principles and Practice of Sex Therapy, 3rd Ed., Guilford Press, New York, pp. 471–503. Kalichman, S. C. (1998). Understanding AIDS: Advances in Research and Treatment , 2nd Ed., American Psychological Association, Washington, DC. Kalichman, S. C., Greenberg, J., and Abel, G. G. (1997). HIV-seropositive men who engage in highrisk behaviour: Psychological characteristics and implications for prevention. AIDS Care 9: 441– 450. Kalichman, S. C., Nachimson, D., Cherry, C., and Williams, E. (1998). AIDS treatment advances and behavioral prevention setbacks: Preliminary assessment of reduced perceived threat of HIVAIDS. Health Psychol. 17: 546–550. Kelly, J. A., Hoffman, R. G., Rompa, D., and Gray, M. (1998). Protease inhibitor combination therapies and perceptions of gay men regarding AIDS severity and the need to maintain safer sex. AIDS 12: F91–F95. Kelly, J. A., and Kalichman, S. C. (1998). Reinforcement value of unsafe sex as a predictor of condom use and continued HIV/AIDS behavior among gay and bisexual men. Health Psychol. 17: 328– 335. Kelly, J. A., Otto-Salaj, L. L ., Sikkema, K. J., Pinkerton, S. D., and Bloom, F. R. (1998). Implications of HIV treatment advances for behavioral research on AIDS: Protease inhibitors and new challenges in HIV secondary prevention. Health Psychol. 17: 310–319. Kippax, S., Crawford, J., Davis, M., Rodden, P., and Dowssett, G. (1993). Sustaining safe sex: A longitudinal study of a sample of homosexual men. AIDS 7: 257–263. Kippax, S., Noble, J., Prestage, G., Crawford, J. M., Campbell, D., Baxter, D., and Cooper, D. (1996). Sexual Negotiation in the AIDS era: Negotiated safety revisited. AIDS 11: 191–197. Lee, T. H., Sakahara, N., Fiebig, E., Busch, M. P., O’Brien, T. R., and Herman, S. A. (1996). Correlations of HIV-1 RNA levels in plasma and heterosexual transmission of HIV-1 from infected transfusion recipients. J. Acq. Imm. Defic. Syndr. 12: 427–428. McKirnan, D. J., Ostrow, D. G., and Hope, B. (1996). Sex, drugs and escape: A psychological model of HIV-risk sexual behaviors. AIDS Care 8: 665–669. Murphy, S. T., Miller, L. C., Appleby, P. R., Marks, G., and Mansergh, G. (1999). Antiretroviral Drugs and Sexual Behavior in Gay and Bisexual Men: When Optimism Enhances Risk. Paper presented at the National Prevention Conference. Musicco, M., Lazzarin, A., Nicolosi, A., Gasparini, M., Costigliola, P., Arici, C., and Saracco, A. (1994). Antiretroviral treatment for men infected with human immunodeficiency virus type 1 reduces the incidence of heterosexual transmission. Arch. Int. Med. 154: 1971–1976. NASTAD (1999, September). Negotiated safety, barebacking, and more. HIV Prevent. Update 5: 1–9. Odets, W. W. (1994). AIDS education and harm reduction do gay men: Psychological approaches for the 21st century. AIDS Pub. Pol. J. 9: 1–15. Odets, W. W. (1996). Risk appraisals and HIV prevention. Focus 11: 1–4.

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

PP062-295249

UAI and Barebacking in MSM

February 23, 2001

8:53

Style file version Nov. 19th, 1999

299

Ostrow, D. G., Fox, K., Chmiel, J. S., Silvestre, A., Visscher, B. R., Vanable, P., Jacobson, L. P., and Strathdee, S. A. (2000). Attitudes towards highly active antiretroviral therapy predict sexual risk-taking among HIV-infected and uninfected gay men in the Multicenter AIDS cohort Study (MACS). Paper presented at the XIII International Conference on AIDS. Pinkerton, S. D., and Abramson, P. R. (1992). Is risky sex rationale? J. Sex Res. 29: 561–568. Pinkerton, S. D., and Abramson, P. R. (1996). Occasional condom use and HIV risk reduction. J. Acq. Imm. Defic. Syndr. 13: 456–460. Pinkerton, S. D., and Holtgrave, D. (1999). Combination antiretroviral therapies for HIV: Some economic considerations. In Ostrow, D. G., and Kalichman, S. C. (eds.), Psychosocial and Public Health Impacts of New HIV Therapies, Kluwer Academic, New York, (pp. 83– 107). Prestage, G., Grulich, A., Campbell, D., Kaldor, J., and Kippax, S. (1997, Abstract No. IS 113). Unprotected intercourse among HIV-positive men: The influence of combination therapies. Paper presented at the Annual Conference of the Australas Society of HIV Med. Purcell, D. W., DeGroff, A. S., and Wolitski, R. J. (1998). HIV prevention case management: Current practice and future directions. Health Soc. Work 23: 282. Quinn, T. C., Wawer, M. J., Sewankambo, N., Serwadda, D., Li, C., Wabwire-Mangen, F., Meehan, M. O., Lutalo. T., and Gray, R. H. (2000). Viral load and heterosexual transmission of human immunodeficiency virus type 1. N. Engl. J. Med. 342: 921–929. Rabkin, J., and Chesney, M. (1999). Combination antiretroviral therapies for HIV: Some economic considerations. In Ostrow, D. G., and Kalichman, S. C. (eds.), Psychosocial and Public Health Impacts of New HIV Therapies, Kluwer Academic, New York, pp. 61–79. Ragni, M. V., Faruki, H., and Kingsley, L. A. (1998). Heterosexual HIV-1 transmission and viral load in hemophilic patients. J. Acq. Imm. Defic. Syndr. 17: 42–45. Remien, R. H., Wagner, G., Carballo-Dieguez, A., and Dolezal, C. (1998). Who may be engaging in high-risk sex due to medical treatment advances? AIDS 12: 1560–1561. Rowatt, W. C., Cunningham, M. R., and Druen, P. B. (1999). Lying to get a date: The effects of facial physical attractiveness on the willingness to deceive prospective dating partners. J. Soc. Per. Rel. 16(2): 209–233. Royce, R. A., Sena, A., Cates, W., and Cohen, M. S. (1997). Sexual transmission of HIV. N. Engl. J. Med. 336: 1072–1078. San Francisco Department of Public Health (2000). The San Francisco Department of Public Health and AIDS Research Institute/UCSF Response to the Updated Estimates of HIV infection in San Francisco, 2000. Retrieved March 30, 1999 from the World Wide Web: http://hivinsite.ucsf. edu/ Scarce, M. (1999, February). A Ride on the Wild Side. POZ. Schechter, M., Strathdee, S. A., Martindale, S. L., Miller, M. L., Hogg, R. S., Woodfall, B., Sestak, P., and Schechter, M. T. (1998). Evidence of elevated HIV incidence and relapse to unsafer sex among young men having sex with men (MSM) in Vancouver, Canada. Paper presented at the XII International Conference AIDS. Seal, D. W., Kelly, J. A., Bloom, F. R., Stevenson, L. Y., Coley, B. I., and Broyles, L. A. (2000). HIV prevention with young men who have sex with men: What young men themselves say is needed. AIDS Care 12: 5–26. Sogolow, E., Peersman, G., Semaan, S., Strouse, D., Lyles, C., and the Prevention Research Synthesis Team (in press). CDC’s HIV AIDS Prevention Research Synthesis Project: Development, initial results, and future directions. J. Acq. Imm. Defic. Syndr. Suarez, T., and Kauth, M. R. (in press). HIV risk assessments in men who have sex with men. In Sessions. Suarez, T., Kelly, J. A., Pinkerton, S. D., Stevenson, Y. L., Hayat, M. J., Smith, M. D., and Ertl, T. (2000). The influence of a partner’s HIV serostatus and viral load on perceptions of sexual risk behavior in a community sample of gay and bisexual men. Manuscript under review. Suarez, T., O’Leary, A., and Morganstern, J. (in press). Selective serotonin reuptake inhibitors: Can the play a role in HIV prevention? In O’Leary, A. (ed.), Beyond Condoms, Kluwer Academic. Valleroy, L., MacKellar, D. A., Rosen, D., and Secura, G. (1998). Prevalence and predictors of unprotected receptive anal intercourse for 15–22-year old men who have sex with men in seven urban areas. Paper presented at the XII International Conference on AIDS, Geneva.

P1: LMD/GEC

P2: GCV/GCQ

Archives of Sexual Behavior [asb]

300

PP062-295249

February 23, 2001

8:53

Style file version Nov. 19th, 1999

Suarez and Miller

Vanable, P. A., Ostrow, D. G., McKirnan, D. J., Taywaditep, K. J., and Hope, B. A. (2000). Impact of combination therapies on HIV risk perceptions and sexual risk among HIV-positive and HIVnegative gay and bisexual men. Health Psychol. 19: 134–145. van de Ven, P., Campbell, D., Kippax, S., Knox, S., Prestage, G., Crawford, J., Kinder, P., and Cooper, D. (1998). Gay men who engage repeatedly in unprotected anal intercourse with casual partners: The Sydney Men and Sexual Health Study. Int. J. STD AIDS 9: 336–340. Vernazza, P. L., Troiani, L., Flepp, M. J., Cone, R. W., Schock, J., Roth, F., Boggian, K., Cohen, M. S., Fiscus, S. A., Eron, J. J., the Swiss cohort study (2000). Potent antiretroviral treatment of HIV-infection results in suppression of the seminal shedding of HIV. AIDS 14: 117– 121. Voelker, R. (2000). HIV drug resistance. JAMA 12: 169. Wagner, G. J., Remien, R. H., and Carballo-Dieguez, A. (1998). “Extramarital” sex: Is there an increased risk for HIV transmission? A study of male couples of mixed HIV status. AIDS Educ. Prev. 10: 245–256. Wainberg, M. A., and Friedland, G. (1998). Public health implications of antiretroviral therapy and HIV drug resistance. JAMA 279: 1977–1983. Wittkowski, K. M., Susser, E., and Dietz, K. (1998). The protective effects of condoms and nonoxynol-9 against HIV infection. Am. J. Pub. Health 88: 590–596. Wolitski, R. J., Reitmeijer, C. A. M., Goldbaum, G. M., and Wilson, R. M. (1998). HIV serostatus disclosures among gay and bisexual men in four American Cities: General pattern and relation to sexual practices. AIDS Care 10: 599–610. Wolitski, R., Valdiserri, R., Denning, P., and Levine, W. C. (in press). Are we headed for a resurgence in the epidemic among men who have sex with men? Am. J. Pub. Health. Wulfert, E., Safren, S. A., Brown, I., and Wan, C. K. (1999). Cognitive, behavioral, and personality correlates of HIV-positive persons’ unsafe sexual behavior. J. Appl. Soc. Psychol. 29: 223–244.

P1: FNN/GCR Archives of Sexual Behavior [asb]

PP062-295250

February 16, 2001

14:26

Style file version Nov. 19th, 1999

Archives of Sexual Behavior, Vol. 30, No. 3, 2001

Depersonalization, Self-Esteem and Body Image in Male-to-Female Transsexuals Compared to Male and Female Controls Uwe Wolfradt, Ph.D.,1,3 and Kerstin Neumann, M.D.2

Whether postoperative male-to-female transsexuals differ in regard to measures of self- and body image from a nontranssexual control group was investigated. A group of 30 postoperative male-to-female transsexuals and control groups of 30 males and 30 females completed self-report measures (depersonalization, selfesteem, gender identity traits, body image). Results showed that transsexuals and males scored higher on self-esteem and dynamic body image than the females did. No differences between the groups were found in terms of depersonalization and satisfaction. Transsexuals and females described themselves as more feminine than males. Regarding sex-role orientation, more androgynous subjects were found among transsexuals than in the control groups. General satisfaction is associated with feminine and masculine traits in transsexuals. Results are discussed in context of the function of these personality features for the identity development of male-to-female transsexuals. KEY WORDS: depersonalization; body image; self-esteem; male-to-female transsexuals; gender identity.

INTRODUCTION In recent years, transsexualism has been a topic at the center of interest in psychological research, in which the exploration of psychological functioning of transsexuals was often ignored (Midence and Hargreaves, 1997). The aim of our study was to investigate personality variables, such as depersonalization, self-esteem 1 Department 2 Department

of Psychology, Martin-Luther-University Halle, Germany. of Otorhinolaryngology, Head and Neck Surgery, Martin-Luther-University Halle,

Germany. whom correspondence should be addressed at Department of Psychology, Martin-LutherUniversity, D-06099 Halle, Germany; e-mail: [email protected].

3 To

301 C 2001 Plenum Publishing Corporation 0004-0002/01/0600-0301$19.50/0 °

P1: FNN/GCR Archives of Sexual Behavior [asb]

302

PP062-295250

February 16, 2001

14:26

Style file version Nov. 19th, 1999

Wolfradt and Neumann

and body image, and to compare postoperative male-to-female transsexuals (after a gender reassignment surgery and a voice operation) and male and female controls. In clinical psychological research, studies have been conducted to examine the gender identity and specific relative stable personality traits of transsexuals before gender reassignment surgery. Lutz et al. (1984) found that male-to-female transsexuals described themselves as more feminine than homosexual males did. Brems et al. (1993) investigated the relationship between sex-role stereotypes and psychopathology in male-to-female transsexuals in comparison with psychiatric patients and college students. Results based on the Draw-A-Person Test indicated that the transsexuals produced female drawings that were more feminine than either of the other groups. Buhrich (1981) reported that different groups of transvestites and transsexuals showed higher scores on neuroticism and introversion (measured with the Eysenck Personality Inventory, EPI) than the general population did. Johnson and Hunt (1990) examined 25 male-to-female transsexuals to determine whether introversion, depression, adjustment to work, and gender reorientation were associated with the typological variables androphilia, gynephilia, feminine gender identity in childhood, and age at onset of transsexualism. Subjects completed three scales of the Minnesota Multiphasic Personality Inventory (MMPI) and several sexual orientation measures to assess the typological variables. No significant relation was found between psychological disturbances and typological variables. Two significant relations were reported between social gender orientation (e.g., living as a woman) and androphilia, and between work adjustment and gynephilia. Brown et al. (1996) examined personality traits (NEO Personality Inventory) and sexual functioning (Derogatis Sexual Functioning Inventory) of 188 nonpatient male cross-dressers, who were classified as transvestites, transgenderists, or transsexuals based on self assessment and cross-gender activities. Results indicated that transsexuals scored higher on the Aesthetic facet scale of Openness to Experiences and reported lower sexual drive than the other groups. Overall, transsexuals and transgenderists had more psychiatric symptoms and a more pronounced feminine gender role and a poorer body image than transvestites. No significant differences among the groups were found in the five large personality domains of the NEO-PI. Summarized, the findings of previous research concerning the personality of transsexuals are heterogeneous and it is not clear if transsexuals have different personality traits than that of nontranssexual groups. In this context, personality traits reflect different central stable aspects of the identity, which were relatively insensitive for change and development. In this study, we focus on different specific variables, such as depersonalization, self-esteem, gender identity traits, and body image. These variables represent fundamental issues for assessing individual differences in personality. Depersonalization experiences represent a specific type of dissociation and can be described as “a persistent or recurrent feeling of being detached from one’s mental processes or body” (DSM-IV, American Psychiatric Association, 1994). Similar to dissociation experiences, depersonalization experiences may appear

P1: FNN/GCR Archives of Sexual Behavior [asb]

PP062-295250

February 16, 2001

Male-to-Female Transsexuals and Depersonalization

14:26

Style file version Nov. 19th, 1999

303

along a continuum of severity ranging from common (mild experience) transient symptoms in response to psychosocial stress to a chronic depersonalization disorder causing marked distress (Steinberg, 1991). There are several results concerning the dissociative experiences among transsexuals. Walling et al. (1998) reported that 10% of 64 transsexuals have high dissociation scores (measured on the Dissociative Experience Scale) indicating the possibility of significant dissociative disorders. These results were explained by emphasizing the importance of dissociative experiences (such as fantasy proneness, absorption) in coping with gender dysphoria. Devor (1994) interviewed 45 self-defined female-to-male transsexuals. 60% of whom reported about one or more cases of severe child abuse. According to Devor, transsexualism may be an adaptive extreme dissociative survival response to severe child abuse. Similar to dissociation, depersonalization can also be considered as a defense mechanism to reduce negative affectivity (Wolfradt and Engelmann, 1999). In Hartmann et al. (1997), persons with gender dysphoria disorders (androphilic and gynephilic males) have the highest scores on the subscales derealization/ depersonalization and negative body image on the ‘Narcissism Inventory’ (NI). Further findings about the association between gender identity disorders and reported higher dissociative experiences were presented in other studies (Coons, 1992; Saks, 1998; Schwartz, 1988). Some findings show that postoperative transsexuals have a disturbed body perception. Marone et al. (1998) found that in a Body Perception Test, in contrast to female-to-male groups with gender identity disorder, male-to-female groups had difficulties reintegrating various body areas into a single unit. They offer the explanation that anxiety plays an important role, leading to the inhibition of perception of the whole body. The aim of our study was to investigate the differences between postoperative male-to-female transsexuals and controls with regard to different personality traits such as depersonalization experiences, self-esteem, body image, and gender identity traits. It was hypothesized that transsexuals describe themselves as more feminine and report more depersonalization experiences than the control groups. Furthermore, it was hypothesized that transsexuals report a more negative body image and a lower self-esteem than the control groups. A further aim was to determine the relationship between satisfaction and the other personality traits.

METHOD Participants Ninety adults participated. The sample consisted of 30 postoperative male-tofemale transsexuals and 60 adults (30 females, 30 males) from Halle/Germany who were a control group. The mean age of all participants was 43 years (range 29–67 years). In the total sample, nearly all participants were employed in a wide range of

P1: FNN/GCR Archives of Sexual Behavior [asb]

PP062-295250

February 16, 2001

304

14:26

Style file version Nov. 19th, 1999

Wolfradt and Neumann

different jobs. All transsexuals received gender assignment surgery and a further operation to achieve a female voice by increasing the tension of the vocal cords. Voice operations were performed in the Department of Otorhinolaryngology, Head and Neck Surgery, at the University of Halle and were based on the technique as described by Isshiki (see Neumann and Berghaus, 1996). Voice operations were carried out in a range of 1–5 years after gender assignment surgery. Half (50%) of the transsexuals live without any partner (14.3% in the female group; 10.3% in the male group). Participants in this control group were matched with transsexuals according to age and occupational status. All subjects were volunteers and were not paid for their participation.

Procedure Participants filled in German versions of various questionnaires, including the scale for depersonalization experiences (SDPE) by Wolfradt (1998), the SelfEsteem-Scale (SES) by Rosenberg (1965; German Version: Ferring and Filipp, 1996), the Body-Image Questionnaire (BIQ) by Clement and Loewe (1996), a Gender Identity Trait Scale (GIS) by Altst¨otter-Gleich (1996) and were asked to answer a question concerning general life satisfaction. • The Scale of Depersonalization Experiences (SDPE) consists of 20 items that require a 5-point frequency format, from 1 (never) to 5 (very often). The SDPE reflects four clusters of depersonalization: disturbances in sense of self (e.g., “I have the feeling that parts of my body don’t belong to me”), self-awareness (e.g., “I observe myself as a stranger”), certainty of self (e.g., “I look into the mirror without recognizing myself really”), and derealization (e.g., “The world around me seems unfamiliar”). The internal consistency of the SDPE was α = .86. • The 10-item SES is a widely used method for assessing general self-esteem. Participants indicated their answers in a four-point format. Internal consistency of the SES was α = .78. • The BIQ consisted of 20 items assessing the dynamic body image (e.g., “I feel very fit” or “Sometimes I feel an extreme energy in myself ”) and rejected body image (e.g., “My body often annoys me” or “I am not satisfied with my body”). The answer format was from 1 (not true at all) to 5 (totally true). Internal consistencies for the subscale “rejected body image” was α = .73 and for the subscale “dynamic body image” α = .63. • The GIS consisted of 13 masculine traits (e.g., authoritarian, dominant, aggressive) and nine feminine traits (e.g., patient, understanding, sympathetic). Participants indicated their answers in a 5-point format, from 1 (not at all ) to 5 (extremely ). Based on an item analysis, only items with an item–total-correlation greater than rit = 0.30, six female traits

P1: FNN/GCR Archives of Sexual Behavior [asb]

PP062-295250

February 16, 2001

14:26

Style file version Nov. 19th, 1999

Male-to-Female Transsexuals and Depersonalization

305

(α = .71) and nine male traits (α = .80), were used for further analysis (see Loewenthal, 1996). • Finally a satisfaction item was presented: “Generally I am satisfied with my life,” from −3 (extremely dissatisfied ) to 3 (extremely satisfied ). RESULTS Mean Differences One-way-variance analyses among the groups were performed. The mean differences are shown in Table I. Contrary to our assumption, transsexuals did not report more depersonalization experiences than both control groups. However, the transsexuals and the control males achieved significantly higher self-esteem scores than the control females (F(2, 86) = 7.30, p < .01). The same results were found for the dynamic body image: transsexuals and males have a more dynamic body image than the females do (F(2, 87) = 7.33, p < .01). Furthermore, no significant difference in the rejected body image among the groups was found. With regard to the gender traits the following results emerged: Transsexuals and females described themselves as more feminine than males. The groups did not differ on masculine traits. Finally, no significant differences were found for life satisfaction. The Correlation Between Satisfaction and the Other Personality Traits Pearson–moment-correlations between satisfaction and gender identity traits, self-esteem, body image, and depersonalization were calculated for the three groups separately (see Table II). Table I. Mean Scores Among the Groups on Each Measure Controls

Self-esteem Body image Rejected Dynamic Gender identity traits Feminine Masculine Depersonalization General satisfaction

Transsexualsa

Femalesa

Malesa

3.44b

3.01a

3.37a,b

2.35a 3.77b

2.13a 3.16a

2.03a 3.48a,b

3.91b 3.29a 1.54a 5.70a

3.63a,b 3.07a 1.45a 5.37a

3.36a 3.26a 1.54a 5.40a

Note. Means having the same superscript (a or b) are not significantly different at p < .05 in the Scheff´e significant difference comparison. a n = 30.

P1: FNN/GCR Archives of Sexual Behavior [asb]

PP062-295250

February 16, 2001

14:26

Style file version Nov. 19th, 1999

306

Wolfradt and Neumann Table II. Zero-Order Correlations (r ) Between Satisfaction and the Personality Measure for Each Group Controls Transsexualsa Self-esteem Body image Rejected Dynamic Gender identity Feminine Masculine Depersonalization an

∗p

.17 −.18 .28 .63∗∗ .40∗ .01

r

Femalesa r

Malesa r

.44∗

−.14

−.18 .24

−.04 .49∗∗

.30 −.16 −.01

−.05 .23 .19

= 30. < .05; ∗∗ p < .01.

For the transsexuals, positive correlations between satisfaction and masculine as well as feminine traits were found. In the group of females, a positive correlation between satisfaction and self-esteem was found. In the group of males, general satisfaction correlated positively with dynamic body image. Different Sex Role Orientation Styles According to the sex-typing model devised by Bem et al. (1976), four different sex-role orientation styles can be identified: Those who are high in terms of both gender traits (masculine and feminine), those who are low in both, those who are high in terms of one and low in terms of the other, and those who show the reverse combination. Based on this taxonomy K -means cluster analyses were performed on the data, using the standardized scores of the masculine traits and the feminine traits as the dependent variable and subjects as the independent variable. The clustering algorithm is an iterative procedure that assigns cases to a specified number of nonoverlapping clusters (Hartigan, 1975). As a result, the cluster analysis revealed four groups which showed no significant difference with regard to age. The first group (n = 14) was labelled androgynous group and was characterized by high scores on masculine and feminine traits. The second group (n = 20) was called the undifferentiated group; the characteristic aspects are low scores on gender traits. The third group (n = 22) was named the masculine-typed group and consisted of those individuals who had low scores on feminine traits and high scores on masculine traits. Finally, the fourth group (n = 31) was named the feminine-typed group and was characterized by high feminine traits and low masculine traits. Table III shows the distribution of the members of the whole group among the sex-role orientation styles. This distribution reached significance (χ 2 = 24.75, p < .001). More than half of the transsexuals were in the androgynous group, more than half of the

P1: FNN/GCR Archives of Sexual Behavior [asb]

PP062-295250

February 16, 2001

14:26

Style file version Nov. 19th, 1999

Male-to-Female Transsexuals and Depersonalization

307

Table III. Distribution of the Groups in Regard to the Four Sex-Role Orientation Types Controls

Androgynous (%) Undifferentiated (%) Masculine-typed (%) Feminine-typed (%)

Transsexuals (n = 29)

Females (n = 30)

Males (n = 28)

71.4 35.0 13.6 29.0

7.1 40.0 22.7 51.6

21.4 25.0 63.6 19.4

females in the feminine-typed group, and more than half of the males in the masculine-typed group. The proportion of the group members is nearly equal in the undifferentiated group. DISCUSSION This study investigates the differences between postoperative male-to-female transsexuals and matched control groups of females and males in terms of different personality traits. Results confirm previous findings, namely, that transsexuals consider themselves as adjusted females, not as pathological males (e.g., Cohen et al., 1997). These results can be explained by the overall satisfaction with surgical voice operation (75% of transsexuals were satisfied). Our results indicate that transsexuals compared to females and males do not differ with regard to depersonalization and satisfaction. In comparison to other studies in which transsexuals had high dissociation scores (Walling et al., 1998) or had a high depersonalization score (Hartmann et al., 1997), our results speak in favor of a normal range for such phenomena. Depersonalization decreased in postoperative transsexuals, because the discrepancy between perceived body image and gender role is reduced. In this case, depersonalization as a defense mechanism against negative feelings that have their origin in this discrepancy is no longer necessary. The most interesting outcome of our study is that the transsexuals described themselves as similar to males with regard to self-esteem and dynamic body image, and similar to females with regard to feminine traits. These findings underlined the intermediate role of male-to-female transsexuals in various personality traits between males and females. Transsexuals are confronted with the developmental task to integrate feminine and masculine traits and behavior in order to find a new sexual role of self-understanding. Furthermore, after their operation, they are more focused on bodily sensations, like sexual feelings (Lief and Hubschman, 1993). The differences between transsexuals and females concerning their attitude toward the body showed that transsexuals behave like males in this respect. Many studies showed that females have more negative body image evaluations than males do (see Muth and Cash, 1997; Feingold and Mazzella, 1998). Contrary to previous research (Marone et al., 1998), the transsexuals did not perceive their body as more negative than did the control groups.

P1: FNN/GCR Archives of Sexual Behavior [asb]

PP062-295250

February 16, 2001

308

14:26

Style file version Nov. 19th, 1999

Wolfradt and Neumann

According to our results from the cluster analysis, transsexuals are more androgynous than the other groups. A broad research into sex-role orientation and personality showed that an androgynous style was indicative of greater psychosocial adjustment, such as satisfaction with life (Ramanaiah et al., 1995) and body (Ludwig and Brownell, 1999). Taylor and Hall (1982) emphasize that androgynous persons were more flexible in adapting to a wide range of masculine and feminine behavior than were the other groups. Furthermore, the correlational analysis indicates that high satisfaction in transsexuals is associated with the self-description of both masculine and feminine traits, while satisfaction in females is related to high self-esteem and in males to high dynamic body image. These findings emphasize the meaning of the gender identity traits for the self-image of transsexuals. It is important for postoperative male-to-female transsexuals to be accepted as females in the eyes of society. Satisfaction for females is strongly determined by positive self-evaluations, while males are more satisfied when a positive body evaluation is made. Our study examined only male-to-female transsexuals after their gender surgery change and an operation to achieve a female voice. It must be assumed that the point of time of questioning is important for the interpretation of data. Especially, male-to-female transsexuals are likely to be more dissatisfied with their life and more maladjusted before the operation than transsexuals after surgery with a positive result (see Lothstein, 1984). Pf¨afflin (1993) found that male-to-female transsexuals described themselves as both more feminine as well as more masculine after their gender surgery change than before. Further studies on the personality structure of transsexuals should concentrate more on the contribution of specific personality traits. It is necessary to find psychological measures which accurately assess on which dimensions postoperative transsexuals experience themselves as normal or deviant. However, only a long-term study may make it possible to find an answer to the question of the process of personality development in the transition period of preoperative to the postoperative life stage. REFERENCES Altst¨otter-Gleich, C. (1989). Theoriegeleitete Itemkonstruktion und -auswahl: Eine Modifikation des Einsatzes der Repertory-Grid-Technik dargestellt am Beispiel der Erfassung der Geschlechtsidentit¨at [Theory oriented item construction and selection: A modification of the use of the Repertory-Grid-Technique as shown at the exemple of the assessment of gender identity], Verlag f¨ur empirische P¨adagogik, Landau. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders, 4th Ed., American Psychiatric Association, Washington, DC. Bem, S. L., Martyna, W., and Watson, C. (1976). Sex typing and androgyny: Further explorations of the expressive domain. J. Pers. Soc. Psychol. 34: 1016–1023. Brems, C., Adams, R. L., and Skillman, G. D. (1993). Person drawing by transsexual clients, psychiatric clients, and nonclients compared: Indicators of sex-typing and pathology. Arch. Sex. Behav. 22: 253–264.

P1: FNN/GCR Archives of Sexual Behavior [asb]

PP062-295250

February 16, 2001

Male-to-Female Transsexuals and Depersonalization

14:26

Style file version Nov. 19th, 1999

309

Brown, G. R., Wise, T. N., Costa, P. T., Herbst, J. H., and Fagan, P. J. (1996). Personality characteristics and sexual functioning of 188 cross-dressing men. J. Nerv. Ment. Dis. 184: 265–273. Buhrich, N. (1981). Psychological adjustment in transvestism and transsexualism. Behav. Res. Ther. 19: 407–411. Clement, U., and Loewe, B. (1996). Die Validierung des FKB-20 als Instrument zur Erfassung von K¨orperbildst¨orungen bei psychosomatischen Patienten [The validation of the Body-ImageQuestionnaire (BIQ): A German-language instrument for assessing body image disturbances in patients with psychsomatic disorders]. Psychother. Psychosom. Med. Psychol. 46: 254–259. Cohen, L., Ruiter, C., de, Ringelberg, H., and Cohen-Kettenis, P. T. (1997). Psychological functioning of adolescent transsexuals: Personality and psychopathology. J. Clin. Psychol. 53: 187–196. Coons, P. M. (1992). Dissociative disorder not otherwise specified: A clinical investigation of 50 cases with suggestions for typology and treatment. Dissociation 5: 187–195. Devor, H. (1994). Transsexualism, dissociaton, and child abuse: An initial discussion based on nonclinical data. J. Psychol. Hum. Sex. 6: 49–72. Feingold, A., and Mazzella, R. (1998). Gender differences in body image are increasing. Psychol. Sci. 9: 190–195. Ferring, D., and Filipp, S.-H. (1996). Messung des Selbstwertgef¨uhls: Befunde zu Reliabilit¨at, Validit¨at und Stabilit¨at der Rosenberg-Skala [Measurement of self-esteem: Findings on reliability, validity, and stability of the Rosenberg Scale]. Diagnostica 42: 284–292. Hartigan, J. A. (1975). Clustering Algorithms, Wiley & Sons, New York. Hartmann, U., Becker, H., and Rueffer-Hesse, C. (1997). Self and gender: Narcissistic pathology and personality factors in gender dysphoric patients. Preliminary results of a prospective study. Int. J. Transg. 1: http://www.symposion.com/ijt/ijtc0103.htm. Johnson, S. L., and Hunt, D. D. (1990). The relationship of male transsexual typology to psychological adjustment. Arch. Sex. Behav. 19: 349–360. Lief, H. I., and Hubschman, L. (1993). Orgasm and postoperative transsexual. Arch. Sex. Behav. 22: 145–155. Loewenthal, K. L. (1996). An Introduction to Psychological Tests and Scales, UCL Press, London. Lothstein, L. M. (1984). Psychological testing with transsexuals: A 30-year review. J. Pers. Ass. 48: 500–507. Ludwig, M. R., and Brownell, K. D. (1999). Lesbians, bisexual women, and body image: An investigation of gender roles and social group affiliation. Int. J. Eat. Dis. 25: 89–97. Lutz, D. J., Roback, H. B., and Hart, M. (1984). Feminine gender identity and psychological adjustment of male transsexuals and male homosexuals. J. Sex Res. 20: 350–362. Marone, P., Iacoella, S., Cecchini, M. G., and Ravenna, A. R. (1998). An experimental study of body image and perception in gender identity disorders. Int. J. Trans. 2: http://www.symposion. com/ijt/ijtc0501.htm. Midence, K., and Hargreaves, I. (1997). Psychological adjustment in male-to-female transsexuals: An overview of research evidence. J. Psychol. 131: 602–614. Muth, J. L., and Cash, T. F. (1997). Body image attitudes: What difference does gender make? J. Appl. Soc. Psychol. 27: 1438–1452. Neumann, K., and Berghaus, A. (1996). Raising the medium speaking voice pitch by surgical means in male-female transsexuals. Arch. Oto-Rhino-Laryngology 253: 75. Pf¨afflin, F. (1993). Transsexualit¨at: Beitr¨age zur Psychopathologie, Psychodynamik und zum Verlauf [Transsexuality: Contributions to psychopathology, psychodynamics and to course], Enke, Stuttgart. Ramanaiah, N. V., Detwiler, F. R., and Byravan, A. (1995). Sex-role orientation and satisfaction with life. Psychol. Rep. 77: 1260–1262. Rosenberg, M. (1965). Society and the Adolescent Self-Image, Princeton University Press, Princeton. Saks, B. R. (1998). Transgenderism and Dissociative Identity Disorder—A case study. Int J. Trans. 2: http://www.symposion.com/ijt/ijtc0404.htm. Schwartz, P. G. (1988). A case of concurrent multiple personality disorder and transsexualism. Dissociation 1: 48–51. Steinberg, M. (1991). The spectrum of depersonalization: Assessment and treatment. In Tasman, A., and Goldfinger, S. M. (eds.), Review of Psychiatry, Vol. 10, American Psychiatric Press, Washington, DC, pp. 223–247.

P1: FNN/GCR Archives of Sexual Behavior [asb]

310

PP062-295250

February 16, 2001

14:26

Style file version Nov. 19th, 1999

Wolfradt and Neumann

Taylor, M. C., and Hall, J. A. (1982). Psychological androgyny: Theories, methods, and conclusions. Psychol. Bull. 92: 347–366. Walling, D. P., Goodwin, J. M., and Cole, C. M. (1998). Dissociation in a transsexual population. J. Sex. Edu. Thera. 23: 121–123. Wolfradt, U. (1998). The Scale of Depersonalization Experiences, Unpublished Manuscript, University Halle, Germany. Wolfradt, U., and Engelmann, S. (1999). Depersonalization, fantasies and coping behavior in clinical context. J. Clin. Psychol. 55: 225–232.

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

14:21

Style file version Nov. 19th, 1999

Archives of Sexual Behavior, Vol. 30, No. 3, 2001

Prevalence and Patterns of Child Sexual Abuse and Victim–Perpetrator Relationship Among Secondary School Students in the Northern Province (South Africa) S. N. Madu, D.Sc.,1,2 and K. Peltzer, Ph.D.1

An investigation into the prevalence and characteristics of child sexual abuse in the Northern Province (South Africa) was conducted. A total of 414 secondary school students in standard 9 and 10 in three representative secondary schools completed a retrospective self-rating questionnaire in a classroom setting. The questionnaire asked about childhood sexual abuse and the victim–perpetrator relationship. Results shows an overall (N = 414) child sexual abuse prevalence rate of 54.2%, 60% for males (N = 193), 53.2% for females (N = 216). Among them, 86.7% were kissed sexually, 60.9% were touched sexually, 28.9% were victims of oral/anal/vaginal intercourse. “Friend” was the highest indicated perpetrator in all patterns of sexual abuse. Many victims (86.7%) perceived themselves as not sexually abused as a child, and many (50.2%) rated their childhood as “very happy.” A call is made for more research, publicity, and campaigns in the area of child sexual abuse in the Province. KEY WORDS: child sexual abuse; prevalence; pattern; victim-perpetrator relationship; South Africa.

INTRODUCTION The prevalence of and problems associated with child sexual abuses have not been exhaustively researched in many parts of the world. Finkelhor and Browne (1986, pp. 143–179), after reviewing many publications formulated a model, a conceptual framework for the (initial and long-term) effects of childhood sexual abuse. They pointed at the effects to be the conjunction of the following 1 Department

of Psychology, University of the North, South Africa.

2 To whom correspondence should be addressed at Department of Psychology, University of the North,

Private Bag X1106, Sovenga 0727, South Africa. 311 C 2001 Plenum Publishing Corporation 0004-0002/01/0600-0311$19.50/0 °

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

312

14:21

Style file version Nov. 19th, 1999

Madu and Peltzer

trauma-causing factors: Traumatic Sexualization, Stigmatisation, Betrayal, and Powerlessness. Child sexual abuse is one of the social problems that calls for urgent attention in South Africa (Collings, 1992, 1993, 1995, 1997; Magwaza, 1994; Sonderling, 1993; Haffejee, 1991; Levett, 1989a,b). The prevalence among a sample of university female students of contact forms of sexual abuse (i.e. actual or attempted intercourse, oral or manual genital contact, sexual kissing, hugging, or touching) has been reported to be 30.9% (Levett, 1989a,b) and 34.8% (Collings, 1997). For both contact and noncontact forms of abuse (i.e. contact abuse plus exhibitionism, voyeurism, sexual threats or invitations), the prevalence has been reported to be 43.6% in female university students in South Africa (Levett, 1989a,b). For male university students, Collings (1991) reported a rate of 28.9% for contact and noncontact forms. The Child Protection Units of the South African Police Services identified during the first 6 months in 1996, a total of 19,805 cases of crimes against children less than 18 years of age, of which child sexual abuse were 7968 (40%) (rape: 7363, sodomy: 480, and incest: 125) (Piennar, 1996). Although the patterns of sexual abuse (i.e. sexual kisses, touches, oral, anal, or vaginal intercourse; and whether physical force was involved or not) and some characteristics, such as victim– perpetrator relationships, were considered, it has been established that they can vary from region to region (Fromuth and Burkhart, 1987; Peters et al., 1986; Finkelhor, 1979; Leth, 1994). Thus the present study in the Northern Province of South Africa. Bayley and Kings (1990) explained child sexual abuse to be when an adult or person significantly older or in a position of power interacts with a child in a sexual way for the gratification of the older person. This study limits itself to any contact form of sexual abuse (among the secondary school students) which took place before the age of 17 years. The perpetrator must be an adult or a person at least 5 years older than the child or a person in a position of power. The Northern Province has a population of 5.4 million inhabitants. Among them, 97.1% are blacks, 0.1% are coloured, 0.1% are Indians/Asians, and 2.7% are whites. Males comprise 45.7% and females 54.3%. Many inhabitants live in poor economic and medical conditions (Republic of South Africa, 1996; Health Systems Trust and Department of Health, 1997). In South Africa, if a child indicates that he or she has been sexually touched by an adult—where the adult is known—it is required by law that the information be revealed to the police, or to a Commissioner of Child Welfare or a social worker (Du Plessis, 1996). METHOD Participants Participants for this study were all standard nine and ten secondary (high) school students in three schools in the Northern Province. One school is situated

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

14:21

Style file version Nov. 19th, 1999

Prevalence and Patterns of Child Sexual Abuse in the Northern Province (S. Africa)

313

in a village (Graskop), one in a semiurban area (Mahwelereng), while the other is in an urban area (Pietersburg). Schools were randomly chosen and the representative character of each of them was confirmed by the Department of Education of the Province. Standard nine and ten secondary school students were chosen because the authors believe that they are mature enough to have the courage to report their sexually abusive experiences, and at the same time their ages are expected not to be too far above our operational maximum age for child sexual abuse (16 years). This would give them the maximum opportunity to remember what happened before they were 17. The total number of participants was 414: 193 (46.6%) males, 216 (52.2%) females, and 5 (1.2%) did not indicate their gender. Mean age was 18.5 years (SD = 2.18) and range 14–30 years; 15 (3.6%) did not indicate their age; 336 (81.1%) were blacks; 51 (12.3%) were whites; 10 (2.4%) were coloured; 11 (2.7%) were “other,” who decided to describe their skin colour in other ways, for example, “olive”; and 6 (1.4%) did not indicate their skin colour. A total of 207 (50%) live in villages, 107 (25.8%) in semiurban areas/towns, 94 (22.7%) in an urban area, and 6 (1.4%) did not indicate where they live. Instrument The instrument used is an anonymous, retrospective, self-rating child maltreatment questionnaire, an abbreviated and modified form of the Child Maltreatment Interview Schedule (Briere, 1993), which has the following components: 1. Questions on the demographic variables of the participants (gender, age, skin colour, and place of residence). 2. Questions on the (physical) contact forms of sexually abusive experiences of participants before the age of 17 years, with an adult or person at least 5 years older or a person in a position of power; who the perpetrator(s) was (were); and whether physical force was used. The patterns of contact sexual abuse considered were sexual kisses or touches, and oral, anal, or vaginal intercourse. 3. Questions on whether the participant perceives himself/herself as sexually abused as a child (i.e. before the age of 17 years) or not, and the overall rating of his or her childhood. The questionnaire includes both open-ended and close-ended questions, and in some closed-ended questions, multiple choice is possible. The second and third components of the questionnaire have been used by Raborifi (1997) in South Africa as part of a questionnaire for a study on the history of childhood abuse among female university students. Before use, we administered it to a group of 20 standard nine students in another school (not used for this study) to ensure that the students would understand the questions and that it would be

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

314

14:21

Style file version Nov. 19th, 1999

Madu and Peltzer

easy to administer. They were found to have no problem in understanding and answering the questions. Procedure The authors obtained approval from the University of the North ethics committee. Furthermore, permission for the research was taken both from the Department of Education in the province and from the principals of the schools. Principals of the schools took the ethical responsibility of informing and obtaining permission from the parents of the participants before hand, where necessary, and from the participants themselves. On the agreed dates with the schools, with the cooperation of the teachers, a research assistant distributed the questionnaire to all the school children in standard nine and ten in their classrooms. The children were first of all explained the purpose of the research, were allowed to ask questions about the research, and then asked to fill-in the questionnaires to the best of their knowledge. It was also made clear to them beforehand that those who did not want to participate should feel free to decline. A total of four (0.96%) students did not fill-in their questionnaires. Questionnaires were collected the same day they were distributed. The language of the questionnaire was English, which was also the language of its administration. All participants completed the questionnaires within 30 min even though 1 h was allocated for the exercise. In view of the fact that the questionnaire may have aroused some emotions, especially among the sexually victimised subjects, students were told to feel free to contact the researchers (through the telephone numbers and in the addresses provided to the students) in case of questions, counselling and/or psychotherapy. Alternatively, any clinical psychologist, psychotherapist, or counsellor available could be consulted. Moreover, it was agreed with the Department of Education that the result of the research shall be made available to them for use in planning preventive health care services in the province. The administration of the whole questionnaire was completed within 3 weeks in March 1998. RESULT Prevalence Of the total number of respondents (N = 414), the number who indicated any form of (physical) contact sexual abuse is 225. This is a prevalence rate of 54.2%. The number of male victims is 108, while that of female victims is 115. This is a prevalence rate of 60% for males and 53.2% for females. Two participants (0.9%) did not indicate their gender. The number of black victims are 172 (76.4%), that of whites is 39 (17.3% of the victims), that of coloured is 9 (4.0%), that of “others” who decided to describe

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

14:21

Style file version Nov. 19th, 1999

Prevalence and Patterns of Child Sexual Abuse in the Northern Province (S. Africa)

315

their skin colour in other ways, for example, “olive,” is 4 (1.8%), and 1 person (0.4%) did not indicate skin colour. This is a prevalence rate of 51.2% for blacks, 76.5% for whites, 90% for coloured, and 36.4% for others. The number of the victims living in villages is 110 (48.9%), in suburban areas 53 (23.6%), in urban areas 59 (26.2%); 3 (1.3%) did not indicate where they live. This is a prevalence rate of 53.1% for those living in villages, 49.5% for those in semiurban areas, and 62.8% for those in urban areas. Patterns of (Physical) Contact Child Abuse A total of 195 participants (86.7%) indicated that they were kissed in a sexual way; 30 (13.3%) indicated that the (sexual) kiss was done by force. A total of 137 participants (60.9%) indicated that someone touched their bodies in a sexual way or made them touch his/her sexual parts; 23 subjects (10.2%) indicated that the perpetrator used force. A total of 65 participants (28.9%) indicated that someone has ever had oral, anal, or vaginal intercourse with them or has placed their fingers or objects in the participant’s anus or vagina; 16 subjects (7.1%) indicated that the sexual intercourse was done by force. The Perpetrators Table I shows the frequency distribution of the perpetrator(s) who were indicated to have kissed the child in a sexual way. Table I shows that 63 subjects (32.3%) did not indicate the perpetrator; “friend” has the highest frequency, 87 (44.6%); followed by “mother,” 10 (5.1%); then by “stranger” with 9 (4.6%). Further Chi-square analysis to determine whether there is any significant difference in gender of those victimized (sexual kisses) by “friend” shows no significant difference: males = 44 and females = 42 (1 missing variable), χ 2 = 0.046,df = 1, p = ns. Table II shows the frequency distribution of the perpetrator(s) who were indicated to have touched the body of the participant in a sexual way before he or she was 17 years old, or who made the participant touch the perpetrator’s sexual parts. Table II shows that 7 (5.1%) did not indicate the perpetrator; “friend” was highest, 84 (61.3%); followed by “combination of professional and friend/extended family member/stranger/nanny,” 9 (6.6%); and then “stranger” and “combination of family member and friend/extended family member/stranger/nanny,” 6 (4.4%). Further Chi-square analysis to determine if there is any significant difference in gender of those victimized (sexual touches) by “friend” shows that there is no significant difference: males = 45 and females = 39 (1 missing variable), χ 2 = 0.428, df = 1, p = ns.

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

14:21

316

Style file version Nov. 19th, 1999

Madu and Peltzer

Table I. Frequency Distribution of the Perpetrator(s) Who Were Indicated to Have Kissed the Participants in a Sexual Way Before the Age of 17 Perpetrator

Frequency

Percent

Valid percent

Cumulative percent

Father Mother Brother Sister Uncle Aunt Friend Stranger Other professional Doctor Combination of family members Combination of family members & professional Combination of family members & friends/extended family/strangers/nanny Combination of family members & professionals & friends/extended family/strangers/nanny Combination of professionals & friends/extended family/strangers/nanny

1 10 1 2 4 2 87 9 1 3 1 2

0.5 5.1 0.5 1.0 2.1 1.0 44.6 4.6 0.5 1.5 0.5 1.0

0.8 7.6 0.8 1.5 3.0 1.5 65.9 6.8 0.8 2.3 0.8 1.5

0.8 8.3 9.1 10.6 13.6 15.2 81.1 87.9 90.9 90.2 91.7 93.2

2

1.0

1.5

94.7

.4

2.1

3.0

100.0

3

1.5

2.3

97.0

195

100.0

100.0

Total

Table III shows the frequency distribution of the perpetrators indicated to have had oral, anal, or vaginal intercourse with the participants or to have placed their fingers or objects in the participant’s anus or vagina. Table III shows that 7 subjects (10.8%) did not indicate the perpetrator; “friend” has the highest frequency, 40 (61.5%); followed by “stranger,” 5 (7.7%); and by “other professional,” 4 (6.2%). Further Chi-square analysis to find out whether there is any significant difference in gender of those victimised (oral, anal, or vaginal intercourse) by “friend” shows no significant difference: males = 15 and females = 25 (1 missing variable), χ 2 = 2.5, df = 1, p = ns. Other Questions Among those who answered the first question, 21 (9.3%) indicated that they perceived themselves as sexually abused as a child while 195 (90.7%) did not. Among those who perceived themselves as sexually abused as a child, 7 were males (i.e., 6.5% of the male victims) and 13 were females (i.e., 11.3% of the female victims). Among those who did not perceive themselves as sexually abused as a child, 99 were males (i.e., 91.7% of the male victims) and 97 were females (i.e., 84.4% of the female victims).

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

14:21

Style file version Nov. 19th, 1999

Prevalence and Patterns of Child Sexual Abuse in the Northern Province (S. Africa)

317

Table II. Frequency Distribution of the Perpetrator(s) Who Were Indicated to Have Touched the Body of the Participant in a Sexual Way Before He/She Was 17 Years Old, or Who Made the Subject Touch the Perpetrator’s Sexual Parts Perpetrator

Frequency

Percent

Valid percent

Cumulative percent

Sister Uncle Aunt Friend Stranger Babysitter/nanny Teacher Other professional Doctor Combination of friends/extended family members/strangers/nanny Combination of family members & professional Combination of family members & friends/extended family/strangers/nanny Combination of family members & professionals & friends/extended family/strangers/nanny Combination of professionals & friends/extended family/strangers/nanny

3 3 1 84 6 1 1 6 5 1

2.2 2.2 0.7 61.3 4.4 0.7 0.7 4.4 3.6 0.7

2.3 2.3 0.8 64.6 4.6 0.8 0.8 4.6 3.8 0.8

2.3 4.6 5.4 70.0 74.6 75.4 76.2 84.6 80.0 85.4

1

0.7

0.8

86.2

6

4.4

4.6

90.8

3

2.2

2.3

9

6.6

6.9

137

100.0

100.0

Total

100 97.7

Table III. Frequency Distribution of the Perpetrators Indicated to Have Had Oral, Anal, or Vaginal Intercourse With the Participants or Who Have Placed Their Fingers or Objects in the Participant’s Anus or Vagina Perpetrator

Frequency

Percent

Valid percent

Cumulative percent

Father Sister Uncle Aunt Friend Stranger Other professional Doctor Combination of friends/extended family members/strangers/nanny Combination of family members & professionals & friends/extended family/strangers/nanny Total

2 1 2 1 40 5 4 1 1

3.1 1.5 3.1 1.5 61.5 7.7 6.2 1.5 1.5

3.4 1.7 3.4 1.7 69.0 8.6 6.9 1.7 1.7

3.4 5.2 8.6 10.3 79.3 87.9 96.6 89.7 98.3

1

1.5

1.7

65

100.0

100.0

100

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

318

14:21

Style file version Nov. 19th, 1999

Madu and Peltzer

A total of 39 (17.3%) rated their childhood as “very unhappy,” 63 (28%) rated it as “average,” 113 (50.2%) rated it as “very happy,” and 10 (4.4%) did not indicate their rating. Among those who perceived themselves as very unhappy during childhood, 15 were males (i.e., 13.9% of the male victims) and 24 were females (i.e., 20.9% of the female victims). Among those who perceived themselves as average during childhood, 33 were males (i.e., 30.6% of the male victims) and 30 were females (i.e., 26.1% of the female victims). Among those who perceived themselves as very happy during childhood, 56 were males (i.e., 51.9% of the male victims) and 56 were females (i.e., 48.7% of the female victims). DISCUSSION The prevalence rates for contact forms of child sexual abuse among our participants (54.3% for all, 60% for males, 53.2% for females) are high. Looking at this result from the conceptual framework of Finkelhor and Browne (1986), it implies that many of these victims of child sexual abuse are likely to suffer psychological and behavioral problems as a result of traumatic sexualisation, stigmatisation, betrayal, and powerlessness. This view is confirmed by a review of the literature made by Beitchman et al. (1991) which shows that many adolescents with a history of childhood sexual abuse reported short-term sequelae such as sexual dissatisfaction, promiscuity, homosexuality, an increased risk for revictimisation, depression, and suicidal ideation or behavior. The prevalence rate among our female participants is higher than that reported by those of Levett (1989a,b, 30.9%) and those of Collings (1997, 34.8%). Two things may explain the difference in the rates obtained by Levett and Collings on the one hand, and that obtained by us. The area of our study is different (Northern Province). In this province, many parents work as migrant laborers—either in other provinces or at places far away from their homes. As a result, many children are left either alone at home during the weekdays after school or with nannies and grandparents, who may not give them proper care. Thus, they are vulnerable to sexual abuse from opportunistic perpetrators. The prevalence rate among our male participants (60%) is much higher than that reported by Collings, 1991, (28.9%). The prevalence rate among our male subjects (60%) is higher than that of our female subjects (53.2%). This is contrary to popular expectation and to other reported findings (Pilkington and Kremer, 1995), and for example, when the findings of Collings 1991 (28.9% for males) is compared with the findings of Collings 1997 (34.8% for females). Lodico et al. (1996) in their study among school-based adolescents found females were four times more likely to report sexual abuse than males. It may mean that our male participants felt freer or were more open in reporting their childhood sexual experiences than the females. Furthermore, it could also be that absence of the adult male (who is working as a migrant laborer) in many families and the frequent single parenthood in the society contribute to adult females abusing boys.

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

14:21

Style file version Nov. 19th, 1999

Prevalence and Patterns of Child Sexual Abuse in the Northern Province (S. Africa)

319

The above findings show higher prevalence rates than that in major studies of child sexual abuse in North America, where one-fifth to one-third of all women reported some sort of childhood sexual encounters with an adult male (Pilkington and Kremer, 1995) and a 20-countries study by Finkelhor (1994) who found a history of child sexual abuse in 7% to 36% for women and 3% to 29% for men. It is worth noting that the rates for blacks (51.2%), whites (76.5%), and coloured (90%) differ greatly. This gives room for speculation in the social and family lifestyles of the different groups. The social and familial tie among blacks, which is known to be strong and which therefore serves as a check for social misconduct, may account for the lower rate. The same social tie may have, as a control factor, also accounted for the lower rate of child sexual abuse among village dwellers (53.1%) when compared to that of urban-area dwellers (62.8%). Moreover, factor of residence has not been considered by other researchers in this field in South Africa. Prevalence rates for the different patterns of sexual abuse show that the actual genital intercourse (28.9%) is much lower than that which may be considered as “milder” forms of contact sexual abuse (sexual kisses, 86.7%, and touches, 60.9%). The rate at which force is being used by the perpetrators (13.3% for sexual kisses, 10.2% for sexual touches, and 7.1% for oral, anal, or vaginal intercourse) calls for attention and more in-depth study of rape among children in the province. Sexual abuse involving penetration, force or violence, and a close relationship to the perpetrator have been indicated to be the most harmful in terms of long-term effects on the child (Beitchman et al., 1991). In all three forms of sexual abuse considered in this study, “friend” was highest in the rank of perpetrators as compared to relatives (in parentheses): 44.4% (12.7%) for sexual kisses, 61.3% (9.5%) for sexual touches, 61.5% (10.2%) for oral, anal, or vaginal intercourse. Collings (1997) reported the victim–perpetrator relationship to be highest among “acquaintances” (40.4%). In agreement with our findings, we presume that “friend” would form the majority of what Collings called “acquaintance.” One may also speculate that many participants who indicated “friend” as the perpetrator may have decided to ignore our age criterion for perpetrators and indicated all forms of childhood sexual relationship with a friend or some of the participants may have written the word “friend” to mean an “acquaintance” (as some people in the area colloquially use it). This would raise not only the percentage of indication of “friend” as perpetrator but also the overall prevalence rate of child sexual abuse observed in the province. The high rate of “friends” as perpetrators is different from other studies wherein the highest rate of perpetrators was found among relatives, for example, among Mexican Americans, 43% were relatives and 37% acquaintances (Huston et al., 1995). This study has its limitations. Only standard nine and ten secondary school students were used as participants. This would limit the external validity of our findings. Moreover, noncontact forms of sexual abuse were not considered. There

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

14:21

320

Style file version Nov. 19th, 1999

Madu and Peltzer

are also other aspects of child sexual abuse that needs further investigation, for example, the exact age(s) of victims and perpetrator(s) at the time of victimization, and the gender of perpetrators. Knowledge of the gender of perpetrators would have made it possible for one to statistically match the gender of victims against that of perpetrators. Since some participants were victimized by both members and nonmembers of the family (at the same time or at different times), the authors did not investigate the number of those victimized by members and those by nonmembers of the family. Also the duration of victimization and secondary victimization (the psychological and behavioral effects of the victimization on the victims) were not investigated. The very high prevalence rates, especially among males, and the small percentage who felt abused, may be because of inclusion of many trivial incidents. This should be guarded against in future studies.

ACKNOWLEDGMENTS Thanks to the Department of Education of the Northern Province (South Africa), and the principals and students of the schools involved in this study for their support and cooperation. We are also grateful to the University of the North for providing a research grant for the study.

REFERENCES Bayley, C., and Kings, K. (1990). Child Sexual Abuse, Tavislock, London. Beitchman, J. H., Zucker, K. J., Hood, J. E., DaCosta, G. A., and Akman, D. (1991). A review of the short-term effects of child sexual abuse. Child Abuse Negl. 15(4): 537–556. Briere, J. N. (1993). Child Abuse Trauma, Sage, London. Collings, S. J. (1991). Childhood sexual abuse in a sample of South African University males: Prevalence and risk factors. S. Afr. J. Psychol. 21: 153–158. Collings, S. J. (1992). The process of victimization in childhood sexual abuse. Soc. Work 28(2): 1–6. Collings, S. J. (1993). Physically and sexually abused children: A comparative analysis of 200 reported cases. Soc. Work 29(4): 301–306. Collings, S. J. (1995). The long-term effects of contact and non-contact forms of child sexual abuse in a sample of university men. Child Abuse Negl. 19: 1–6. Collings, S. J. (1997). Child sexual abuse in a sample of South African women students: Prevalence, characteristics, and long-term effects. S. Afr. J. Psychol. 27(1): 37–42. Du Plessis, L. (1996). We won’t be beaten: A guide to the Prevention of Family Violence Act, NIPILAR, Pretoria. Finkelhor, D. (1979). Sexually Victimised Children, Free Press, New York. Finkelhor, D. (1994). The international epidemiology of child sexual abuse. Child Abuse Negl. 18: 409–417. Finkelhor, D., and Browne, A. (1986). Initial and long-term effects: A conceptual framework. In Finkelhor, D., Araji, S., Baron, L., Browne, A., Peters, S. D., and Wyatt, G. E. (eds.), A Source Book on Child Sexual Abuse, Sage, London, pp. 180–198. Fromuth, M. E., and Burkhart, B. R. (1987). Long-term psychological correlates of childhood sexual abuse in two samples of college men. Child Abuse Negl. 13: 533–542. Haffejee, I. E. (1991). Sexual abuse of Indian (Asian) children in South Africa: First report in a community undergoing cultural change. Child Abuse Negl. 15(1, 2): 147–151.

P1: FGA/GCR Archives of Sexual Behavior [asb]

PP062-295251

February 16, 2001

14:21

Style file version Nov. 19th, 1999

Prevalence and Patterns of Child Sexual Abuse in the Northern Province (S. Africa)

321

Health Systems Trust and the Department of Health. (1997). Health Care in the Northern Province, Kwik Kopy Printing, Durban. Huston, R. L., Parra, J. M., Prihoda, T. J., and Foulds, D. M. (1995). Characteristics of childhood sexual abuse in a predominantly Mexican-American population. Child Abuse Negl. 19(2): 165–176. Leth, I. (1994). Child sexual abuse: Sex differences in relations and context of abuse. Paper presented at the Tenth International Congress on Child Abuse and Neglect, Kuala Lumpur, Malaysia. Levett, A. (1989a). A study of childhood sexual abuse among South African University women students. S. Afr. J. Psychol. 19(3): 122–129. Levett, A. (1989b). Psychological trauma: Discourses of childhood sexual abuse. Unpublished doctoral thesis, University of Cape Town. Lodico, M. A., Gruber, E., and DiClemente, R. J. (1996). Childhood sexual abuse and coercive sex among school-based adolescents in a Midwestern state. J. Adolesc. Health 18(3): 211–217. Magwaza, A. S. (1994). Perception of family relationships in sexually abused children. Soc. Work 30(4): 390–396. Peters, S. D., Wyatt, G. E., and Finkelhor, D. (1986). Prevalence. In Finkelhor, D. (ed.), A Source Book on Child Sexual Abuse, Sage, Beverly Hills, CA. pp. 15–59. Pienaar, A. (1996). The child protection unit (CPU) of the South African police service. Focus Forum (HSRC/RGN) 4(3): 15–20. Pilkington, B., and Kremer, J. (1995). A review of the epidemiological research on child sexual abuse: Community and College student samples. Child Abuse Rev. 4: 84–98. Raborifi, M. M. (1997). History of childhood abuse in Northern Sotho female university students. A thesis submitted as a partial fulfilment of the Bachelor of Arts (Honours) Degree in Psychology, University of the North. Republic of South Africa. (1996). Statistics in Brief, Central Statics, Pretoria. Sonderling, S. (1993). Power of discourse and discourse of power in making an issue of sexual abuse in South Africa: The rise and fall of social problems. Crit. Arts 6(2): 1–26.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Archives of Sexual Behavior, Vol. 30, No. 3, 2001

BOOK REVIEWS Two-Spirit People: Native American Gender Identity, Sexuality, and Spirituality. By Sue-Ellen Jacobs, Wesley Thomas, and Sabine Lang. University of Illinois Press, Urbana, Illinois, 1997, 331 pp., $44.95 (cloth), $19.95 (paper). Reviewed by Frederick L. Whitam, Ph.D.1

This work is somewhat difficult to review because it is not a book in the usual sense, treating a single theme or presenting a set of interrelated ideas. It is, rather, a collection of papers and presentations funded by the Wenner-Gren Foundation for Anthropological Research held in Washington, D.C., in 1993 and Chicago in 1994, as well as a session at the meeting of the American Anthropological Association (AAA), entitled “Revisiting the ‘North American Berdache’ Empirically and Theoretically,” which took place in conjunction with the Washington meeting. The title of the AAA session describes the work’s contents more aptly than the published title. Perhaps the editors purposely avoided using the session title as the book title because it contains the word “berdache.” One of the main contributions of the book is the call for the elimination of the use of “berdache” by social scientists and the substitution of the term “two-spirit” people. The term “berdache” is of European origin imposed upon the social sciences and Native Americans by observers who were often hostile to transgendered people. The more appropriate term, “two-spirit” people, has emerged in recent years from within the Native American community and is used to refer not only to people called transsexual by the sex research community but also to people of homosexual orientation—lesbians and gay men. Jacobs favors the learning of Native categories of reality rather than imposing those developed in Western social and other sciences. While an admirable goal, it also suggests a point of view, perpetuated by some anthropologists, that Native American sexuality and gender arrangements are strikingly different from and perhaps even superior to those in larger American society or in other societies. While descriptions of terminology and custom in various Native American groups are important, the most striking impression emerging from this collection for the 1 Department

of Sociology, Arizona State University, Tempe, Arizona 85287-2101. 323 C 2001 Plenum Publishing Corporation 0004-0002/01/0600-0323$19.50/0 °

P1: FTK Archives of Sexual Behavior [asb]

324

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

sex researcher is how similar, not different, two-spirit people are to transsexuals, lesbians, and gay men—not only in larger North America but also in societies in other parts of the world. Despite differing cultural contexts and terminology, sex researchers will readily recognize familiar behavior. This work supports the notion of the universal occurrence of lesbian, gay, and transgendered people in all human societies with similar behavioral arrangements, despite somewhat differing cultural contexts (see, e.g., Whitam, 1983). For example, two-spirit people were often characterized by childhood cross-gender behavior. A Native American contributor, Michael Red Earth (Sisseton Dakota), reports that he was disinterested in farm chores but strongly attracted to his grandmother’s beadwork. He writes that his grandmother . . . supplemented the family income by doing beadwork. She would make beaded belts and medallions to sell through an Indian women’s craft cooperative. I was fascinated by her beadwork, and at an early age I was learning various beadwork techniques . . . Among our tribe, beadwork is considered to be women’s work. By expressing an interest in beadwork, I was making a declaration to my family that I was more interested in women’s work. I was choosing to be winkte even though I did not realize I was making a declaration, but thereafter people treated me differently. (p. 212)

Another contributor, Anguksuar [Richard LaFortune] (Yup’ik) reports that he played “Batgirl” in a childhood game based on comic strip characters (p. 218). There seems to be greater tolerance in Native American communities for lesbians than for gay men, an attitude which characterizes many societies. These accounts support the notion often reported by sex researchers that there are considerably more gay men than lesbians (see, e.g., Gebhard, 1972) and that there are more male-to-female than female-to-male transsexuals (see, e.g., American Psychiatric Association, 2000). Transgendered men tend to marry or partner with men who do not identify as gay, a point also discussed in mainstream sex research (see, e.g., Whitam, 1997). Aspects of this collection inadvertently suggest that Native American sexuality and gender arrangements are an integral part of human sexuality and gender arrangements characterized by the unique cultural settings in which they are acted out. The most interesting and valuable contribution of this work is that of contemporary two-spirit Native Americans themselves and the issues they raise. These persons are scattered all over the United States and Canada and come from various tribes. Some are of mixed descent. Among the issues raised by these writers are contemporary problems, such as the treatment of Native Americans—sometimes discriminatory—within the larger gay and lesbian communities, growing homophobia on the reservations, and the wide diversity in the treatment of two-spirit people in the many Native American tribes. An important theoretical question raised by several observers, but not fully answered, is to what extent more general American attitudes toward gay, lesbian, and transgendered people have changed traditional Native American attitudes toward two-spirit people.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

325

A problem for some of these scholars is that of terminology. While most sex researchers probably would agree that terminology referring to transgendered people is difficult at best and that the terminology used by Native Americans themselves should be understood, the invention of new terminology in English by Jacobs is puzzling. For example, she writes that “It is rarely understood, for example, that in some cultural traditions third-gender male-bodied people do not have sex with one another” (p. 12). She uses the term “female-bodied Native American” in place of simply “female Native American.” It is difficult to understand how the use of this idiosyncratic terminology is more helpful than “male-to-female transsexual” or “woman,” for example Jacobs’s introduction and lead article illustrate the failure of anthropologists to relate Native American terminology to the contemporary terminology rather widely accepted by sex researchers. Even more puzzling, perhaps, is the hostile attitude toward sex researchers expressed by Tafoya (Taos, Warm Springs), a well-known Native American psychologist. He endorses and quotes Irvine’s odd notion that: Sexology contributes to the collective sexual discourse of the medical and psychiatric professions, a discourse that is itself a means of social control. Categories of “natural” and “deviant” not only operate on the personal level to shape individual experience, but underpin the legal system as well. Whenever a personal preference becomes a “sexual dysfunction,” a “sexual deviancy,” or a crime it is a political decision often related to its status in the psychiatric community. (p. 197)

It is simply not true that sexologists (taken to refer to sex researchers) generally have favored criminalizing or taken a deviant model of homosexuality and transgender behavior. On the contrary, sex researchers historically (e.g., Magnus Hirschfeld, Havelock Ellis, and Alfred Kinsey) were at the forefront of debates on these issues calling for decriminalization and treating these behaviors as normal variations. Contemporary sex researchers generally follow this tradition and are among the strongest defenders of the rights of homosexuals and transgendered people. The most serious criticism of the anthropologists’ writing here is the tendency to cling to the notion of Native American sexuality and gender arrangements as special, unique, and even perhaps morally superior to those of other peoples. At the same time that Native American customs and attitudes should be carefully recorded, understood, and treasured there is a need to relate these customs and attitudes to those of human sexuality. Native American two-spirit people are indeed special and unique as are transgendered people everywhere. REFERENCES American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental Disorders, 4th Ed., text rev., American Psychiatric Association, Washington, DC.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

326

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

Gebhard, P. H. (1972). Incidence of overt homosexuality in the United States and Western Europe. In Livingstone, J. M. (ed.), National Institute of Mental Health Task Force On Homosexuality: Final Report and Background Papers, U. S. Department of Health, Education, and Welfare, Rockville, Maryland. Whitam, F. L. (1983). Culturally invariable properties of male homosexuality: Tentative conclusions from cross-cultural research. Arch. Sex Behav. 12: 207–226. Whitam, F. L. (1997). Culturally universal aspects of male homosexual transvestites and transsexuals. In Bullough, B., Bullough, V., and Elias, J. (eds.), Gender Blending, Prometheus, Amherst, NY.

Transmen & FTMs: Identities, Bodies, Genders, and Sexualities. By Jason Cromwell. University of Illinois Press, Urbana, Illinois, 1999, 201 pp., $42.50 (cloth), $19.95 (paper). Reviewed by Holly Devor, Ph.D.2

This volume is about making the realities of transmen and female-to-male transsexuals (FTMs) more visible. Cromwell uses the terms transmen and FTMs interchangeably to mean female-to-male transsexuals and female-bodied transgendered people who live as men. The main theme of the book is that a group of people whom Cromwell calls “medico-psychological practitioners” have controlled public discourse on female-bodied gender variance and that they have misunderstood, misrepresented, and distorted the reality of FTMs and transmen. Cromwell’s mission is to set the record straight by first exposing the problem and then providing a dose of reality from the perspectives of transmen and FTMs themselves. The book, a reworking of Cromwell’s doctoral dissertation in anthropology, is divided into 14 short chapters and an appendix, which are further subdivided into sections of a page or two, each beginning with short epigraphs. If this kind of format is to be successful, it requires of the author an extra effort at providing sufficient continuity to make it work. I found each section clear enough, but the mortar was not always strong enough to hold the bits together very well. In the opening two-page prologue and in “Excerpts from [Cromwell’s] Journey,” Cromwell establishes his credentials as a transman. Over the next several short chapters, Cromwell introduces some of the problems with the ways that “medico-psychological practitioners” fail to recognize that the physical state of a person’s body need not be the final arbiter of their gender. At the end of the third chapter, Cromwell sets out what struck me as his main message: Transpeople not only shake the foundations of the biological foundationalist and essentialist theories but also undermine them completely. They are social disruptions and as such are a threat to the social body. They are not like other people. Rather than allowing society to dictate who and what they are, they define themselves. (p. 43)

In other words, Cromwell sees transmen as serving a revolutionary function in society. By their very existence, they challenge fundamental social beliefs about 2 Department

Canada.

of Sociology, University of Victoria, POB 3050, Victoria, British Columbia V8W 3P5,

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

326

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

Gebhard, P. H. (1972). Incidence of overt homosexuality in the United States and Western Europe. In Livingstone, J. M. (ed.), National Institute of Mental Health Task Force On Homosexuality: Final Report and Background Papers, U. S. Department of Health, Education, and Welfare, Rockville, Maryland. Whitam, F. L. (1983). Culturally invariable properties of male homosexuality: Tentative conclusions from cross-cultural research. Arch. Sex Behav. 12: 207–226. Whitam, F. L. (1997). Culturally universal aspects of male homosexual transvestites and transsexuals. In Bullough, B., Bullough, V., and Elias, J. (eds.), Gender Blending, Prometheus, Amherst, NY.

Transmen & FTMs: Identities, Bodies, Genders, and Sexualities. By Jason Cromwell. University of Illinois Press, Urbana, Illinois, 1999, 201 pp., $42.50 (cloth), $19.95 (paper). Reviewed by Holly Devor, Ph.D.2

This volume is about making the realities of transmen and female-to-male transsexuals (FTMs) more visible. Cromwell uses the terms transmen and FTMs interchangeably to mean female-to-male transsexuals and female-bodied transgendered people who live as men. The main theme of the book is that a group of people whom Cromwell calls “medico-psychological practitioners” have controlled public discourse on female-bodied gender variance and that they have misunderstood, misrepresented, and distorted the reality of FTMs and transmen. Cromwell’s mission is to set the record straight by first exposing the problem and then providing a dose of reality from the perspectives of transmen and FTMs themselves. The book, a reworking of Cromwell’s doctoral dissertation in anthropology, is divided into 14 short chapters and an appendix, which are further subdivided into sections of a page or two, each beginning with short epigraphs. If this kind of format is to be successful, it requires of the author an extra effort at providing sufficient continuity to make it work. I found each section clear enough, but the mortar was not always strong enough to hold the bits together very well. In the opening two-page prologue and in “Excerpts from [Cromwell’s] Journey,” Cromwell establishes his credentials as a transman. Over the next several short chapters, Cromwell introduces some of the problems with the ways that “medico-psychological practitioners” fail to recognize that the physical state of a person’s body need not be the final arbiter of their gender. At the end of the third chapter, Cromwell sets out what struck me as his main message: Transpeople not only shake the foundations of the biological foundationalist and essentialist theories but also undermine them completely. They are social disruptions and as such are a threat to the social body. They are not like other people. Rather than allowing society to dictate who and what they are, they define themselves. (p. 43)

In other words, Cromwell sees transmen as serving a revolutionary function in society. By their very existence, they challenge fundamental social beliefs about 2 Department

Canada.

of Sociology, University of Victoria, POB 3050, Victoria, British Columbia V8W 3P5,

P1: FTK Archives of Sexual Behavior [asb]

Book Reviews

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

327

the immutable connections among bodies and genders and identities. Furthermore, according to Cromwell, FTMs are unique in their ability to step beyond the confines of social structures and definitions to create their identities and social statuses at will. Cromwell puts the central part of the book to two concurrent tasks. Firstly, he lambasts a wide range of anthropological and historical researchers for not properly understanding the transgendered nature and importance of female-bodied people of other times and places. Secondly, he reviews a range of historical and anthropological literature about gender variant female-bodied people with a goal to providing a more accurate and complete interpretation of the existing information. Although interesting and useful, he covers little new ground (see, e.g., Devor, 1997; Feinberg, 1996). In the next three chapters, Cromwell takes on the medico-psychological practitioners more directly. In these chapters, Cromwell points out how medicopsychological descriptions of FTMs leave no room for FTMs to be healthy and well-adjusted, but only failed women, maladjusted lesbians, or incomplete men. Cromwell explains this apparently willful blindness on the part of medicopsychological practitioners by making the broad assertion that the unorthodox genders of transmen “arouses something dreadful in male researchers’ psyches” (p. 116). Cromwell then illustrates by describing some of the ways that FTMs and transmen enjoy a sexuality that does not fit within the usual binary sexual orientation categories. Cromwell ends the main section of the book with a short critique of the process of the revisions which resulted in Version Five of the Standards of Care for Gender Identity Disorders by the Harry Benjamin International Gender Dysphoria Association (1998). He uses this account as an opportunity to once again highlight how wrong-headed the medico-psychological practitioners can be when it comes to FTMs and transmen, and how revolutionary and visionary transpeople can be by comparison. The book finishes with “Further Excerpts from [Cromwell’s] Journey,” an “Epilogue,” and an appendix about how he obtained the sample of individuals who participated in his research. The transmen and FTMs who provided the data on which Cromwell builds his arguments were drawn from Cromwell’s own social and personal connections. Four support groups were his main avenue of contact, two of which he cofacilitated and two of which were sponsored by a group which he cofounded. In addition to his participant observation at these groups, Cromwell reports that he conducted formal and informal interviews, administered a formal questionnaire, and spoke with more than 200 individuals. However, “only those individuals active in support groups, those who attended conferences, or those active on the Internet” (p. 12) were among the population from which Cromwell drew his sample and “no FTMs/transmen who were several years postphalloplasty agreed to talk [with him]” (p. 114). As a result, the book “does not include the voices of those who mainstream into society. Instead, it is about transmen and FTMs who, in varying

P1: FTK Archives of Sexual Behavior [asb]

328

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

degrees, are out as transpeople” (p. 14). Thus, Cromwell’s claim to be able to provide a more accurate picture of transmen and FTMs seems insecure. Were he to argue that his work provided a corrective lens on to a particular politicized segment of the FTM world, his case would be much strengthened. Unfortunately, Cromwell does not supply an overview or a copy of what questions were asked in any of his data-gathering situations. The reader is left to cobble together some idea of what was asked and answered from tidbits that Cromwell drops into the text at points where he uses a short quote or two to illustrate and support a claim that he is making about what transmen and FTMs are really like. We are never given any kind of systematic or comprehensive presentation of the data which resulted from his research. This makes it extremely difficult for readers to evaluate for themselves the strength of the case made by Cromwell. Furthermore, Cromwell ignores works by other authors which could provide corroborating data to support his own contentions. Cromwell thus weakens his case. I found this especially troubling because I agree with many of Cromwell’s contentions and have published data from my own extensive research in the area which he could have profitably used to strengthen his arguments (Devor, 1997). Ultimately, Cromwell has written a polemic rather than a piece of research. He offers very little information or theory that is new and he fails to make use of much of the more enlightened writing that has been published in the past few years. Cromwell’s strength is when he shows his anger at how his own and many men’s lives have been distorted and damaged by the ignorance and faulty assumptions made in much of the research about FTMs and transmen. I couldn’t agree more with most of his complaints and with many of the corrective assertions that he makes. Indeed, similar feelings comprise a very large part of what has motivated me in my own research with FTMs and female transgendered people and I have made some similar arguments myself. However, Cromwell fails to make his case convincingly on two accounts. In the first place, although the older literature to which he repeatedly refers is still influential, he does not take into account that in recent years many authors, both trans and nontrans, have taken up the kinds of perspectives that he advocates (e.g., Bullough et al., 1997; Denny, 1998; Ekins and King, 1996; Elliot, 2000; Halberstam, 1998; Prosser, 1998). The other major problem area in Cromwell’s work is in his odd use of data. The sample on which this work is based is, by his own admission, far from a cross-section of the FTM world, yet he seems to believe that the perspectives of these men can be effectively used to redefine what it means to be FTM. Perhaps they can. Only time will tell whether they are indeed a vanguard of what will become the norm in the future for FTMs and transmen. Furthermore, I have no quarrel with most of the assertions that he makes, but that is because I know from my own research that there is a segment of the FTM population who feel as Cromwell claims they do. The problem is that Cromwell does not provide enough of his own data, nor does he make good

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

329

use of anyone else’s, to demonstrate his points convincingly to less well-informed readers. If you want to know how politicized FTMs and transmen think and feel, despite all its weaknesses, this book provides a window into their world. It is not the only window and these points of view are not the only ones held by FTMs and transmen today. Indeed, transmen and FTMs are like everybody else in that they come in all varieties of humanity. The men who Cromwell represents best are those who pride themselves on being unique in their transness and in their queerness. Their voices represent a significant subgroup among transmen and one which needs to be heard. REFERENCES Bullough, V., Bullough, B., and Elias, J. (eds.). (1997). Gender Blending, Prometheus, Amherst, NY. Denny, D. (ed.). (1998). Current Concepts in Transgender Identity, Garland, New York. Devor, H. (1997). FTM: Female-to-Male Transsexuals in Society, Indiana University Press, Bloomington, IN. Feinberg, L. (1996). Transgender Warriors: Making History from Joan of Arc to RuPaul, Beacon Press, Boston. Ekins, R., and King, D. (eds.). (1996). Blending Genders: Social Aspects of Cross-Dressing and SexChanging, Routledge, London. Elliot, P. (2000). Review of Second Skins: The Body Narratives of Transsexuality. Arch. Sex Behav. 29: 397–400. Halberstam, J. (1998). Female Masculinity, Duke University Press, Durham, NC. Harry Benjamin International Gender Dysphoria Association (1998). The Standards of Care for Gender Identity Disorders (Fifth Version), Symposion Publishing, D¨usseldorf, Germany. Prosser, J. (1998). Second Skins: The Body Narratives of Transsexuality, Columbia University Press, New York.

Gender Reversals & Gender Cultures: Anthropological and Historical Perspectives. By Sabrina Petra Ramet. Routledge, London, 1996, 231 pp., $27.99. Travesti: Sex, Gender, and Culture among Brazilian Transgendered Prostitutes. By Don Kulick. University of Chicago Press, Chicago, 1998, 269 pp., $18.00. Reviewed by Lesley J. Gotlib, M.A.3

Gender and sexual identities that transgress the dominant Western dichotomy of female/male, woman/man have increasingly become the focus of inquiry among social scientists. The books reviewed here offer both compelling ethnographic and historical contributions to the growing literature on alternative, meaningful gendered and sexed subjectivities in diverse geographic regions and historical 3 Department

of Anthropology, University of Toronto, Toronto, Ontario, Canada M5S 1A1.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

329

use of anyone else’s, to demonstrate his points convincingly to less well-informed readers. If you want to know how politicized FTMs and transmen think and feel, despite all its weaknesses, this book provides a window into their world. It is not the only window and these points of view are not the only ones held by FTMs and transmen today. Indeed, transmen and FTMs are like everybody else in that they come in all varieties of humanity. The men who Cromwell represents best are those who pride themselves on being unique in their transness and in their queerness. Their voices represent a significant subgroup among transmen and one which needs to be heard. REFERENCES Bullough, V., Bullough, B., and Elias, J. (eds.). (1997). Gender Blending, Prometheus, Amherst, NY. Denny, D. (ed.). (1998). Current Concepts in Transgender Identity, Garland, New York. Devor, H. (1997). FTM: Female-to-Male Transsexuals in Society, Indiana University Press, Bloomington, IN. Feinberg, L. (1996). Transgender Warriors: Making History from Joan of Arc to RuPaul, Beacon Press, Boston. Ekins, R., and King, D. (eds.). (1996). Blending Genders: Social Aspects of Cross-Dressing and SexChanging, Routledge, London. Elliot, P. (2000). Review of Second Skins: The Body Narratives of Transsexuality. Arch. Sex Behav. 29: 397–400. Halberstam, J. (1998). Female Masculinity, Duke University Press, Durham, NC. Harry Benjamin International Gender Dysphoria Association (1998). The Standards of Care for Gender Identity Disorders (Fifth Version), Symposion Publishing, D¨usseldorf, Germany. Prosser, J. (1998). Second Skins: The Body Narratives of Transsexuality, Columbia University Press, New York.

Gender Reversals & Gender Cultures: Anthropological and Historical Perspectives. By Sabrina Petra Ramet. Routledge, London, 1996, 231 pp., $27.99. Travesti: Sex, Gender, and Culture among Brazilian Transgendered Prostitutes. By Don Kulick. University of Chicago Press, Chicago, 1998, 269 pp., $18.00. Reviewed by Lesley J. Gotlib, M.A.3

Gender and sexual identities that transgress the dominant Western dichotomy of female/male, woman/man have increasingly become the focus of inquiry among social scientists. The books reviewed here offer both compelling ethnographic and historical contributions to the growing literature on alternative, meaningful gendered and sexed subjectivities in diverse geographic regions and historical 3 Department

of Anthropology, University of Toronto, Toronto, Ontario, Canada M5S 1A1.

P1: FTK Archives of Sexual Behavior [asb]

330

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

moments. While demonstrating the myriad contexts in which expressions of gender are politically and culturally significant, these books are especially useful in their introductory quality to the ways in which gender and sex binaries can be problematized through social theorizing. The edited volume by Ramet is a compilation of historical and anthropological perspectives on the function of gender reversals over time and space. It is a book comparable to, although not as comprehensive as, an earlier collection by Herdt (1994) (see Pillard, 1999). It consists of 13 chapters, each providing introductory investigations into historical gender crossings in various geographical regions. In Chapter 1, Ramet calls attention to the notion of “gender culture,” which she understands to have derived from the literature on the social construction of gender. She equates this concept with gender ideology, which is defined as society’s understanding of what is possible, proper, and perverse in gender-linked behavior, and more specifically, that set of values, mores, and assumptions which establishes which behaviors are to be seen as gender-linked, with which gender or genders they are to be seen as linked, what is the society’s understanding of gender in the first place, and, consequently, how many genders there are. (p. 2)

Ramet effectively makes use of this notion of gender ideology as social control to bridge the diverse contributions in the volume, whose foci span the functions of gender reversals in ancient cultures and religious traditions, in theater and film, and as a mechanism of social order and mobility. Here I will offer a critical summary of some of those chapters whose material I found most captivating. As an anthropologist in the making, I found the anthropological/ethnographic contributions most intriguing. In “Traversing Gender: Cultural Context and Gender Practices” (Chapter 2), Bolin discusses gender variance in contemporary societies, emphasizing the importance of cultural contextualization in any attempt to understand the function of gender practices and how these practices are given cultural meaning. She makes use of a “five-form gender variance typology” based on ethnographic data through which she problematizes the Western gender paradigm. Most significant in this essay is Bolin’s call for further investigation into these cross-cultural examples of gender variation. This brief chapter allows only for an introductory look at the ethnographic record without any critical analysis. Bolin does succeed, however, in raising questions about the genital primacy of gender that persists in the North American ideology, as well as initiating an anthropological examination of contemporary transgender-identified individuals and communities in North America. Her argument with regard to cultural relativism is made clear and is one with which Ramet agrees, as noted in the introduction, where the editor makes evident her own commitment to transgender issues herself, being a male-to-female transsexual (footnote 69, pp. 19, 20). Among Native American populations, cross-dressing has been documented as early as the colonial period. Lang, a Native American anthropologist, has written extensively on female gender variation and the Berdache tradition, now referred

P1: FTK Archives of Sexual Behavior [asb]

Book Reviews

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

331

to in the literature as the Two-Spirit (Lang, 1998). In Lang’s contribution, entitled “There Is More Than Just Women and Men: Gender Variance in North American Indian Cultures” (Chapter 12), she argues that two-spirited individuals comprised a separate gender category from woman and man and continue to do so in areas where Western influence has not completely wiped the tradition out. Lang informs that the two-spirit tradition was for a long time interpreted as a way that homosexuality could be integrated into the Indian culture; however, she and many other scholars now explain that Native constructions of gender and sexual behavior differ substantially from the binary model; these individuals were/are not simply involved in gender reversals but are integral to a multiple-gender system which survives in Native ideology. Consequently, the terms transsexual and homosexual are not applicable in this case where gender identity does not exist in a direct linear relationship to physical sex (i.e., genitals). Becoming two-spirited is a matter of occupational preference and personality traits rather than an outcome of sexual preference or bodily characteristics. Lang provides a very useful foundation here to build upon in the study of Native North American gender variance, particularly in her critical perspective of the dominant two-sex, two-gender ideology in the modern West (cf. Whitam, 2001). Of all the anthropological papers in this collection, the most complex and fully engaging is that of Porter Poole, entitled “The Procreative and Ritual Constitution of Female, Male, and Other: Androgynous Beings in the Cultural Imagination of the Bimin-Kuskusmin of Papua New Guinea” (Chapter 13). In a descriptively rich and elaborate manner, Porter Poole explores “key images of androgyny in the cultural imagination of the Bimin-Kuskusmin of the West Sepik hinterland of Papua New Guinea” (p. 197). An earlier essay dealt with related gendered meanings and embodiments in this locale (Porter Poole, 1981). Here the author explains that among the Bimin-Kuskusmin, ideas concerning the androgynous are powerful in that they contrast sharply with the otherwise rigid, dichotomous male/female boundaries that exist in many sociocultural contexts maintaining substantial ideological importance. These images of androgyny are elaborated in understandings of procreation and maturation and contexts of myth and ritual where gender can be formed and transformed. Porter Poole begins his analysis with a critical consideration of the relationship between gender and personhood and how this permeates cultural environments and social contexts. He recognizes the complexity of gender and its ideological forces, as he does the need to examine the discursive traditions by which gender may be produced and then dismantled. There is a very eloquent discussion of the negotiation of gender ideology and the mechanisms through which one understanding of gender gains dominance over others and becomes embedded in certain social realms in everyday life. He proceeds to be critical of epistemologies that rely on simplistic assumptions about the nature of gender (i.e., its link to biological sex), a position supported by his own ethnographic data among the Bimin-Kuskusmin.

P1: FTK Archives of Sexual Behavior [asb]

332

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

The historical, political, and religious studies contributions to Ramet’s collection examine a wide range of locales though they share the common goal of deciphering the meaning of gender reversals in varied contexts. Ochshorn (Chapter 3) considers the case of Sumer between the 3rd and 1st millennium BCE where she explores the influence of more or less androgynous divinities in the everyday enactment of gender. Welch (Chapter 4) recounts the story in the Acts of Thecla, the culture of Thecla which emerged in the fourth and fifth centuries and explores the multiple interpretations of this gender reversal for both men and women of the time. Humes (Chapter 8) explores concepts of embodiment and gender modification in the Hindu and Buddhist religious philosophies where spiritual transformations from female to male are deemed necessary for ultimate salvation. Two of the essays I found most interesting are “Elena Alias Eleno: Genders, Sexualities, and ‘Race’ in the Mirror of Natural History in Sixteenth Century Spain” by Burshatin (Chapter 7) and “Gender, Power, and Spectacle in Late Imperial Chinese Theater” (Chapter 9), where Volpp writes about the representations of gender in novels and in their theatrical adaptations. She uses a play entitled The Male Queen to demonstrate the complex ways in which gender was imagined and enacted in Chinese theatrical performances. In summary, Ramet’s volume succeeds in achieving its aim—exploring the cultural diversity of gender reversals over time. The book itself is well organized and would provide a very useful teaching tool for gender/cultural studies. The volume by Kulick is an ethnography based on participant observation in the Brazilian city of Salvador. A travesti is described as a transgendered prostitute—a biological male who not only cross-dresses but adopts a female name and a feminine outward appearance, and attempts to approximate ideal Brazilian feminine bodily proportions through the injection of silicone into various body parts. Kulick reports that travestis do not self identify as women, although they do alter their bodies to look like them. They are homosexuals whose subjectivities demand an alternate understanding of the relations among gender identity, sexual identity, and sexual behavior from that which persists in the West. Kulick argues that travestis act to crystallize Brazilian configurations of these categories, rather than simply inverting them. He aims to elicit the extent to which ethnomethodological theory can be employed in understanding travesti subjectivity and to dislodge the presumed link between biological sex and gender. This book is divided into five sections, each exploring different aspects of travestis’ lives. First, Kulick contextualizes travestis’ worlds and engages the reader in the impoverished and violent content of their daily existence. He goes on to describe the process by which an individual becomes a travesti, emphasizing the role that masculine males play in substantiating a travesti’s sense of self. Here Kulick really underscores the distinction between those elements of travesti identity that are comparable to Western phenomena and those that are not. His clear understanding of the problematic insistence that nonsurgical, nondiagnosed

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

Book Reviews

8:14

Style file version Nov. 19th, 1999

333

gender-crossers are transsexual, as has been argued by anthropologists in the past, is significant. In a discussion of the different meanings of the labels transsexual and transgender, Kulick argues that acknowledging the differential concerns with gender identity and sexual activity are crucial. In the third chapter, Kulick outlines the nature of the travesti’s relationship to her heterosexual “boyfriend,” which is based on gift giving and a subjective affirmation of femininity. And he goes on to discuss prostitution as generating pleasure, the lack of family ties and, more theoretically, how the case of travestis offer insight into alternate gender and sex paradigms. While I lack the empirical fieldwork experience or regional expertise that previous reviewers of Kulick’s ethnography may have (Mott, 1999; Rebhun, in press), I can comment on some theoretical inconsistencies which I found troublesome in this book. Without delving too greatly into detailed summations of his argument, Kulick’s claim that travestis can be viewed as “constructive essentialists” is not well supported. In fact, this statement appears to contradict an earlier claim that travestis do not adhere to the genital primacy that the Western framework of gender upholds, and thus are not defined by their genitals. Further, Kulick’s model of gender identity in Brazil is confused by the lack of grounding it could have gained from a discussion of heterosexual relationships between feminine females and masculine males, for those readers who are not familiar with the cultural specifics. In accordance with some critical comments offered by Rebhun (in press), I found that the ethnographic material was not sufficiently supported by a historical engagement with Brazilian prostitution, male homosexuality, the concept of third gender/sex, or femininity in general in Brazil. Most surprising was the lack of discussion of the relation between wider economic issues and the significance of the consumption and exchange of material goods between travestis and their partners and families. Theorists concerned with the impact a “material culture” has on our understandings of gender and sex would demand an exploration of this in this context. REFERENCES Herdt, G. (ed.). (1994). Third Sex, Third Gender: Beyond Sexual Dimorphism in Culture and History, Zone Books, New York. Jacobs, S.-E., Lang, S., and Thomas, W. (eds.). (1997). Two-Spirit Peoples: Native American Gender Identity, Sexuality, and Spirituality, University of Illinois Press, Urbana, IL. Porter Poole, F. J. (1981). Transforming “natural” woman: Female ritual leaders and gender ideology among Bimin-Kuskusmin. In Ortner, S., and Whitehead, H. (eds.), Sexual Meanings, Cambridge University Press, Cambridge, England, pp. 116–165. Lang, S. (1998). Men as Women, Women as Men: Changing Gender in Native American Cultures, University of Texas Press, Austin, TX. Mott, L. (1999). Review of Travesti: Sex, gender and culture among Brazilian transgendered prostitutes. Available at [email protected]. Pillard, R. C. (1999). Review of Third sex, third gender: Beyond sexual dimorphism in culture and history. Arch. Sex Behav. 28: 408–409.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

334

8:14

Style file version Nov. 19th, 1999

Book Reviews

Rebhun, L. (in press). Perspectives on male homosexuality in Brazil [Review of Beneath the Equator: Cultures of Desire, Male Homosexuality, and Emerging Gay Communities in Brazil and Travesti: Sex, Gender and Culture among Brazilian Transgendered Prostitutes]. J. Gay Lesb. Stud. Whitam, F. L. (2001). Review of Two-Spirit Peoples: Native American Gender Identity, Sexuality, and Spirituality. Arch. Sex. Behav. 30(3): 323–326.

Blending Gender: Social Aspects of Cross Dressing and Sex Changing. By Richard Ekins and Dave King. Routledge, London, 1996, 157 pp., $17.95. Reviewed by Vern L. Bullough, R.N., Ph.D4

The cover blurb on this book is very misleading and off putting. It claims to be the first comprehensive treatment of the social aspects of cross-dressing and sex changing. In order to make such a claim the book ignores many researchers in the field. Some researchers like myself are mentioned and cited but the references are highly selective. Unfortunately, the articles and citations included are not up to date. Several of the original articles seem to end their account in 1980, as if nothing happened after that in terms of research. Alphabetical names of organizations are incorrectly titled. The information on the Erickson Education Foundation is distorted. An occasional article, such as by Buhrich (1976) is reprinted, and it might well have been an adequate description of an Australian cross-dressing club in the 1970s, but certainly is not today. Comparing the 1976 article with the current scene might have made for making some worthwhile comments, but the editors did not see fit to do so. On the other hand, there is some interesting material included, such as that of Farrer on the mention and discussion of cross-dressing in English newspapers and periodicals from the 1860s to 1914. There is also an appendix of hard to find short stories and novels in English featuring cross-dressing and sex changing. Rees’ personal account of “Becoming a Man: The Personal account of a Female-to-Male Transsexual” is insightful, but not radically different from other such accounts. Ekins, one of the editors, in his “The Career Path of the Male Femaler,” gives an interesting developmental model but it would have helped if he had read some of the work of Doctor or others who have dealt with some of the same themes. King, the second editor, summarizes the medical perspectives and the technology involved in transsexualism and emphasizes, as others have, that Christine Jorgensen was not the first. His reading of the Jorgensen case, however, is quite different from mine and it was not simply I believe the case of the Danish physicians regarding Jorgensen as a homosexual who desired castration and therefore should have it. He is, however, particularly good at looking at the psychiatric opposition to surgical change which in a sense found itself a prisoner of the psychiatric viewpoint of the 4 3304

W. Sierra Dr., Westlake Village, California 91362.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

Book Reviews

8:14

Style file version Nov. 19th, 1999

335

1920s. He also emphasizes the submergence of transvestism in the outpouring of literature on transsexualism. The study by Billings and Urban on the sociomedical construction of transsexualism was a valuable contribution in 1982 when it first appeared, but it was not without its critiques, and to include it in a book today without further discussion of the controversy between Money and his colleagues at The Johns Hopkins University and elsewhere implies that nothing has changed. In fact, the last citation to any literature on the topic is 1979 and the result is a totally misleading impression. This criticism of the lack of inclusion of current literature on the topic is reinforced by the attention given to Raymond’s (1980) The Transsexual Empire and her revised introduction to the 1994 edition of her book. There have been changes, some quite drastic, and Raymond’s book even in its first appearance was greatly exaggerated. Ekins contributes an interesting and valuable chapter on telephone sex and intimacy scripts as used by transgender individuals. The only other real reference to the current scene is the article by Whittle, “Gender Fucking or Fucking Gender,” which uses “queer” theory to look at current cultural contributions to theories of gender blending. There is no discussion of the efforts to set criteria for transsexualism by the Harry Benjamin International Gender Dysphoria Association and, in fact, the last word in the book is given to Raymond’s (1994) introduction to the reprint of her book. In sum, rather than the comprehensive overview of gender blending the book claims to be, it should be regarded as a rather eccentric potpourri of readings on transgenderism, many of them dated, but nonetheless historically valuable. Still, it gives the feel of some of the battles that were fought and which are still being fought. REFERENCES Buhrich, N. (1976). A heterosexual transvestite club: Psychiatric aspects. Aust. N. Z. J. Psychiat. 10: 331–335. Billings, D. B., and Urban, T. (1982). The socio-medical construction of transsexualism: An interpretation and critique. Soci. Probl. 30: 266–282. Raymond, J. T. (1980). The Transsexual Empire, The Women’s Press, London. Raymond, J. T. (1994). The Transsexual Empire, Rev. Ed., Teachers’ College Press, New York.

Women’s Sexuality Across the Life Span. By Judith C. Daniluk. Guilford Press, New York, 1998, 357 pp., $36.95. Reviewed by Rosemary Coates, Ph.D.5

I had a great deal of difficulty getting into this book because each time I started I was aware that it was very familiar territory. For readers who feel the same, I suggest 5 Centre

for Sexual Health, Division of Health Sciences, Curtin University, Selby St., Shenton Park 6008, Western Australia, Australia.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

334

8:14

Style file version Nov. 19th, 1999

Book Reviews

Rebhun, L. (in press). Perspectives on male homosexuality in Brazil [Review of Beneath the Equator: Cultures of Desire, Male Homosexuality, and Emerging Gay Communities in Brazil and Travesti: Sex, Gender and Culture among Brazilian Transgendered Prostitutes]. J. Gay Lesb. Stud. Whitam, F. L. (2001). Review of Two-Spirit Peoples: Native American Gender Identity, Sexuality, and Spirituality. Arch. Sex. Behav. 30(3): 323–326.

Blending Gender: Social Aspects of Cross Dressing and Sex Changing. By Richard Ekins and Dave King. Routledge, London, 1996, 157 pp., $17.95. Reviewed by Vern L. Bullough, R.N., Ph.D4

The cover blurb on this book is very misleading and off putting. It claims to be the first comprehensive treatment of the social aspects of cross-dressing and sex changing. In order to make such a claim the book ignores many researchers in the field. Some researchers like myself are mentioned and cited but the references are highly selective. Unfortunately, the articles and citations included are not up to date. Several of the original articles seem to end their account in 1980, as if nothing happened after that in terms of research. Alphabetical names of organizations are incorrectly titled. The information on the Erickson Education Foundation is distorted. An occasional article, such as by Buhrich (1976) is reprinted, and it might well have been an adequate description of an Australian cross-dressing club in the 1970s, but certainly is not today. Comparing the 1976 article with the current scene might have made for making some worthwhile comments, but the editors did not see fit to do so. On the other hand, there is some interesting material included, such as that of Farrer on the mention and discussion of cross-dressing in English newspapers and periodicals from the 1860s to 1914. There is also an appendix of hard to find short stories and novels in English featuring cross-dressing and sex changing. Rees’ personal account of “Becoming a Man: The Personal account of a Female-to-Male Transsexual” is insightful, but not radically different from other such accounts. Ekins, one of the editors, in his “The Career Path of the Male Femaler,” gives an interesting developmental model but it would have helped if he had read some of the work of Doctor or others who have dealt with some of the same themes. King, the second editor, summarizes the medical perspectives and the technology involved in transsexualism and emphasizes, as others have, that Christine Jorgensen was not the first. His reading of the Jorgensen case, however, is quite different from mine and it was not simply I believe the case of the Danish physicians regarding Jorgensen as a homosexual who desired castration and therefore should have it. He is, however, particularly good at looking at the psychiatric opposition to surgical change which in a sense found itself a prisoner of the psychiatric viewpoint of the 4 3304

W. Sierra Dr., Westlake Village, California 91362.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

Book Reviews

8:14

Style file version Nov. 19th, 1999

335

1920s. He also emphasizes the submergence of transvestism in the outpouring of literature on transsexualism. The study by Billings and Urban on the sociomedical construction of transsexualism was a valuable contribution in 1982 when it first appeared, but it was not without its critiques, and to include it in a book today without further discussion of the controversy between Money and his colleagues at The Johns Hopkins University and elsewhere implies that nothing has changed. In fact, the last citation to any literature on the topic is 1979 and the result is a totally misleading impression. This criticism of the lack of inclusion of current literature on the topic is reinforced by the attention given to Raymond’s (1980) The Transsexual Empire and her revised introduction to the 1994 edition of her book. There have been changes, some quite drastic, and Raymond’s book even in its first appearance was greatly exaggerated. Ekins contributes an interesting and valuable chapter on telephone sex and intimacy scripts as used by transgender individuals. The only other real reference to the current scene is the article by Whittle, “Gender Fucking or Fucking Gender,” which uses “queer” theory to look at current cultural contributions to theories of gender blending. There is no discussion of the efforts to set criteria for transsexualism by the Harry Benjamin International Gender Dysphoria Association and, in fact, the last word in the book is given to Raymond’s (1994) introduction to the reprint of her book. In sum, rather than the comprehensive overview of gender blending the book claims to be, it should be regarded as a rather eccentric potpourri of readings on transgenderism, many of them dated, but nonetheless historically valuable. Still, it gives the feel of some of the battles that were fought and which are still being fought. REFERENCES Buhrich, N. (1976). A heterosexual transvestite club: Psychiatric aspects. Aust. N. Z. J. Psychiat. 10: 331–335. Billings, D. B., and Urban, T. (1982). The socio-medical construction of transsexualism: An interpretation and critique. Soci. Probl. 30: 266–282. Raymond, J. T. (1980). The Transsexual Empire, The Women’s Press, London. Raymond, J. T. (1994). The Transsexual Empire, Rev. Ed., Teachers’ College Press, New York.

Women’s Sexuality Across the Life Span. By Judith C. Daniluk. Guilford Press, New York, 1998, 357 pp., $36.95. Reviewed by Rosemary Coates, Ph.D.5

I had a great deal of difficulty getting into this book because each time I started I was aware that it was very familiar territory. For readers who feel the same, I suggest 5 Centre

for Sexual Health, Division of Health Sciences, Curtin University, Selby St., Shenton Park 6008, Western Australia, Australia.

P1: FTK Archives of Sexual Behavior [asb]

336

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

that it is worth persisting. In this well-referenced book, Daniluk synthesizes the work of a broad range of past and current authors. Daniluk is not the first author to challenge the primacy of penile–vaginal intercourse as the one true measure of an individual’s sexual expression; however, she does give conviction to the message: “Despite the range of possible activities that could be construed as sexual, penile vaginal penetration is the one activity that has come to define a woman’s sexual status in our culture” (p. 11). In this book, Daniluk weaves theories, anecdotes, and research evidence into a comprehensive treatise on female sexuality. In so doing, she provides a rich texture to women’s sexual lives that is based on all our experiences. Although it might well be derived from a feminist philosophy, the way it is presented is not as hostile as some other authors’ is. Daniluk does not resile from confronting the issues that preoccupy feminist authors. She does, however, show women as dynamic participants in their socialization, rather than passive victims. Daniluk presents a systematic argument and posits some interesting questions through anecdotes from a wide range of women. The author derives information from her own sources and those of others, thus widening the demographics although, as Daniluk points out, there is a bias toward white, European, and American women. Set out in four parts, with parts two to four taking the reader through the life stages of women, Daniluk presents her summary in the opening chapter: “It is the contention of this book that many problems experienced by women related to their sexuality occur in attempting to accommodate and make sense of the physical, personal and interpersonal changes in their lives” (p. 19). Nevertheless, she constantly brings the reader back to the view that women can and do “resist the impositions of definitions and meanings that would serve to quell and repress women’s erotic life force” (p. 357). The first section, “Opening Pandora’s Box,” provides an overview of the evolution of the sexual self through the complex interaction with one’s environment. Although the metaphor of Pandora’s box is a clich´e, the reader should reflect upon its meaning. In Greek mythology, Pandora was the first woman, who Zeus sent to earth with a box full of evils. This was designed to counteract the blessings brought through Prometheus’ gift of fire. When Pandora opened the box, all the evils flew out, leaving only Hope within the box. Does Daniluk’s reference to Pandora imply that through exploring women’s sexuality we are disposing of evil or exposing it? Should Hope be left in the box or should that also be released? Daniluk exposes many of the myths and messages girls and women receive that lead them to have conflicting feelings and attitudes toward their body and their sexuality. Historically, womens’ sexuality has been described as fundamentally evil and something which is likely to lure men into wrongdoing. Aristotle’s notion that the female fetus is conceived through sperm from the left testicle led to the word sinister

P1: FTK Archives of Sexual Behavior [asb]

Book Reviews

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

337

(from the Latin sinistra, meaning left) being used to imply evil which originates in woman. Many of the world’s major religions tend to reinforce this view. These antecedents provide an explanation for ambiguities in current attitudes; however, it does not explain the origins of misogyny: “When a man marries it is no more than a sign that the feminine talent for persuasion and intimidation . . . has forced him into a more or less abhorrent compromise with his own honest inclinations and best interests” (Mencken, 1989). In this book, Daniluk exposes these attitudes and the influence they have on a woman’s sense of self. She considers the impact of these matters at each stage of a woman’s life. The critical socialization period during infancy and early childhood, the impact of menarche, pregnancy, motherhood, and menopause are explored with an emphasis on the dynamic interaction between the influences of biology, psychology, and society. In the penultimate section, in 11 of the 15 chapters, the author details “problematic meanings” in which she emphasizes those conflicting attitudes which may lead women to simultaneous feelings of bodily discomfort and pleasure. In 10 chapters, Daniluk ends on a positive note, providing practical suggestions for creating new meanings. It is the structure of its chapters that makes this book so readable. Daniluk is a little harsh in her criticism of Masters and Johnson in saying they have made no mention of sexual desire. Their work was intended only to examine the physiological responses to direct sexual stimulation and, in spite of recent criticism, it cannot be denied that their research represents a watershed in the study of human sexuality, as Daniluk herself acknowledges. Daniluk also criticizes sexologists in general for their phallocentricity; however, she is hoisted on her own petard when describing the excitement phase of the female sexual response cycle as “being analogous to penile erection” (p. 162) and also by mentioning the vagina on several occasions when vulva would have been more appropriate. I wonder how Daniluk defines “sexologist,” as the authors she cites in support of her argument are all scholars of human sexuality. Her criticisms may well have had some validity in the 1970s; fortunately, sexology as a multidisciplinary enterprise has continued to develop and one might well include Daniluk among its authors. In so doing, one might criticize her for returning to some of the old terminology, such as “frigid,” and some of the myths that today’s young adults would reject. For example on page 214 Daniluk lists 13 comparative statements, such as “Men are sexually aggressive, women are passive” and “Men never say ‘no’ to sex; when women say ‘no,’ they really mean ‘yes’.” Very few, other than a minority of judges, rapists, and their defence lawyers accept that old chestnut these days (Fischer and Good, 1998). Although it is not in Daniluk’s brief, it is worthy of mention here that some of these myths continue to be problematic for men who have not come to terms with modern women (Fischer and Good, 1998). Daniluk is, however, correct about the different levels of sexual guilt experienced by women vis-´a-vis men.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

338

8:14

Style file version Nov. 19th, 1999

Book Reviews

Despite these comments, Daniluk’s book is comprehensive. Its pages are full of useful information, contains interesting and illuminating anecdotes, and gives a positive perspective to women’s sexuality. She explores the perceptions and needs of women of all ages and under many circumstances, recognizing the particularities of stages, abilities, orientations, and life choices. The book is recommended for anyone who has an interest in women’s sexuality and women’s health. REFERENCES Fischer, A. R., and Good, G. E. (1998). New directions for the study of gender role attitudes. Psychol. Women Q. 22: 371–384. Mencken, H. L. (1989). The war between men and women. In Morgan, F. (ed.), A Misogynist’s Handbook, Jonathan Cape, London.

Talking Difference: On Gender and Language. By Mary Crawford. Sage Publications, Thousand Oaks, Caifornia, 1995, 207 pp., $22.95. Reviewed by Mary Gergen, Ph.D.6

Current versions of the differences between the ways men and women talk suggest that we might as well have come from different planets. We learn via television talk shows and popular self-help books that in order to get along better with our partners, our bosses, and our friends, we need to take a quick course in conversational translations. For example, when does “no” mean “yes” and “yes” mean “no” or does silence always mean assent? This volume critically examines the social science research on the question of whether women and men speak differently. Crawford’s focus is on conversations and the expansion of the question of difference to include queries such as the following: How are relations between men and women created and maintained in talk? How are power and status revealed in conversations? How is talking related to sexist, racist, or other discriminatory activities? The book is not a primer for improving one’s social life through conversations, but rather it is aimed at looking critically at the assumptions often made about men and women’s talk. Thus, scholars and students of conversation, especially in psychology, women’s studies, and communication are the audience for the text. At the same time, an adroit reading could provide some insights into one’s own linguistic behaviors and those of significant others, which could have a positive impact on one’s social relations. Crawford begins by pointing out some of the important limitations of all types of research that claim to find differences in the ways women and men talk. She 6 Department

of Psychology and Women’s Studies, Pennsylvania State University, Delaware County, 25 Yearsley Mill Road, Media, Pennsylvania 19063.

P1: FTK Archives of Sexual Behavior [asb]

PP062-[295252-57]

February 19, 2001

338

8:14

Style file version Nov. 19th, 1999

Book Reviews

Despite these comments, Daniluk’s book is comprehensive. Its pages are full of useful information, contains interesting and illuminating anecdotes, and gives a positive perspective to women’s sexuality. She explores the perceptions and needs of women of all ages and under many circumstances, recognizing the particularities of stages, abilities, orientations, and life choices. The book is recommended for anyone who has an interest in women’s sexuality and women’s health. REFERENCES Fischer, A. R., and Good, G. E. (1998). New directions for the study of gender role attitudes. Psychol. Women Q. 22: 371–384. Mencken, H. L. (1989). The war between men and women. In Morgan, F. (ed.), A Misogynist’s Handbook, Jonathan Cape, London.

Talking Difference: On Gender and Language. By Mary Crawford. Sage Publications, Thousand Oaks, Caifornia, 1995, 207 pp., $22.95. Reviewed by Mary Gergen, Ph.D.6

Current versions of the differences between the ways men and women talk suggest that we might as well have come from different planets. We learn via television talk shows and popular self-help books that in order to get along better with our partners, our bosses, and our friends, we need to take a quick course in conversational translations. For example, when does “no” mean “yes” and “yes” mean “no” or does silence always mean assent? This volume critically examines the social science research on the question of whether women and men speak differently. Crawford’s focus is on conversations and the expansion of the question of difference to include queries such as the following: How are relations between men and women created and maintained in talk? How are power and status revealed in conversations? How is talking related to sexist, racist, or other discriminatory activities? The book is not a primer for improving one’s social life through conversations, but rather it is aimed at looking critically at the assumptions often made about men and women’s talk. Thus, scholars and students of conversation, especially in psychology, women’s studies, and communication are the audience for the text. At the same time, an adroit reading could provide some insights into one’s own linguistic behaviors and those of significant others, which could have a positive impact on one’s social relations. Crawford begins by pointing out some of the important limitations of all types of research that claim to find differences in the ways women and men talk. She 6 Department

of Psychology and Women’s Studies, Pennsylvania State University, Delaware County, 25 Yearsley Mill Road, Media, Pennsylvania 19063.

P1: FTK Archives of Sexual Behavior [asb]

Book Reviews

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

339

asks important questions such as the following: Are the samples of the genders in research that claims difference really comparable? If doctors (mostly higher status men) and nurses (mostly lower status women), for example, are compared in terms of their conversational styles, is it gender or status that accounts for the findings of difference? What would happen to the results if we compared high school English teachers (mostly higher status women) with high school janitors (mostly lower status men)? Here, the notion of status eradicating gender differences is raised. Another issue Crawford deals with has to do with what causes gender differences in conversation. If women and men are different in the directness with which they express their desires, for example, is it having breasts or penises that account for it? What does biology have to do with forms of talk after all? Obviously, complex societal forces intervene between the biological category and the habits of language. Crawford is also critical of a fairly common form of advice that is derived from gender-difference research: Women should change their ways of speaking, not men. It is women who should become more assertive, more direct, or more explicit in expressing their desires. They should also develop a better sense of humor. This type of advice implicitly suggests that the male model of talking is the ideal and that women are defective; as a result, they suffer negative social consequences. Crawford is suspicious of such advice because of its sexist implications. In addition, research has shown that women are in a double bind if they do follow the advice because often when they emulate styles of talk associated with men, they are criticized for being unfeminine, bossy, or controlling. In each chapter, Crawford looks with a feminist eye at how various theories influence research on gender differences and its applications. She investigates assertiveness training, humor and humorlessness (why women are described as the latter), and the influence of cultural models of difference on language. In the chapter on the “two cultures” model of difference, she applies the view that women and men speak two different languages to issues related to rape and contends that this theoretical position has allowed men to be excused from the gravity of their crimes while increasing the negativity of the consequences for women. Crawford believes that such a model commits women to being the rape preventer. The potential victim learns that only if she can speak clearly and forcefully will she be able to ward off the rapist. Very little attention is paid to the responsibilities of the man to listen and respond to the voice of a woman, regardless of how assertive her style. Crawford argues that saying “no” (even if forcefully spoken) is not necessarily enough to protect a woman, and, if she is raped, she may blame herself for failing to successfully execute the rules of conversation now applied to her. The social forces that prevent women from exercising certain forms of protective behavior tend not to be recognized in this model. Societal pressures on both women and men for popularity, peer acceptance, and intimacy with others are invisible constructs from the individualistic theoretical perspective so often applied in recommendations to be assertive.

P1: FTK Archives of Sexual Behavior [asb]

340

PP062-[295252-57]

February 19, 2001

8:14

Style file version Nov. 19th, 1999

Book Reviews

For Crawford, language in its many uses is best described from a social constructionist perspective. Talking is the primary way in which we negotiate among others in social settings. It is an action-medium in itself. As such, conversation derives its meanings from the ongoing interplay of interlocutors who come together from different social positions. Her call is for research to focus more on the interactivity of conversation, with its structural constraints, and less on the minute details of individual bits of language, isolated from their natural settings, in which gender differences can be exposed. While she demonstrates a positive example of what she admires in research with a conversational analysis of her own, by and large her discussions throughout the book are drawn from research that is primarily designed according to the standards she criticizes. While acknowledging this, she defends the strategy as pragmatic given that studies based on experimentation and quantitative analysis have a high rhetorical value in the mainstream of psychology. I agree with her assessment, but I don’t entirely agree with her choice. Certainly, the lack of alternative research methods found in this book suggests either that there is little available other than these flawed data to rely upon or she has not wanted to risk using less acceptable qualitative studies of conversation study. I think the book would have been more coherent and challenging if she had introduced more innovative and ideologically sensitive methods from the mushrooming domain of qualitative research (e.g., narrative studies, collaborative research, and ethnographic endeavors). To paraphrase the famous line by Audre Lorde,“You can’t dismantle the master’s house with the master’s tools.” Crawford has taken a good crack at it all the same. I think the book is a worthwhile investment for the intended audience, given its thoughtful summaries of a large research realm. I would hope that in the near future new scholarly materials that are more attuned to the issues of empirical critique will be gathered and presented with the same sensitivity to feminist ideals, theoretical perspectives, and practical concerns as this book demonstrates.