Content uploaded by Kathryn Rebecca Klement
Author content
All content in this area was uploaded by Kathryn Rebecca Klement on Sep 28, 2016
Content may be subject to copyright.
Psychology of Consciousness: Theory,
Research, and Practice
Consensual BDSM Facilitates Role-Specific Altered States of
Consciousness: A Preliminary Study
James K. Ambler, Ellen M. Lee, Kathryn R. Klement, Tonio Loewald, Evelyn M. Comber, Sarah A.
Hanson, Bert Cutler, Nadine Cutler, and Brad J. Sagarin
Online First Publication, September 22, 2016. http://dx.doi.org/10.1037/cns0000097
CITATION
Ambler, J. K., Lee, E. M., Klement, K. R., Loewald, T., Comber, E. M., Hanson, S. A., Cutler, B.,
Cutler, N., & Sagarin, B. J. (2016, September 22). Consensual BDSM Facilitates Role-Specific
Altered States of Consciousness: A Preliminary Study. Psychology of Consciousness: Theory,
Research, and Practice. Advance online publication. http://dx.doi.org/10.1037/cns0000097
Consensual BDSM Facilitates Role-Specific Altered States of
Consciousness: A Preliminary Study
James K. Ambler, Ellen M. Lee,
and Kathryn R. Klement
Northern Illinois University
Tonio Loewald
Redwood City, California
Evelyn M. Comber
Northern Illinois University
Sarah A. Hanson
Georgia State University
Bert Cutler and Nadine Cutler
Tempe, Arizona
Brad J. Sagarin
Northern Illinois University
Researchers studying consensual bondage/discipline, dominance/submission, and sa-
dism/masochism (BDSM) have theorized that individuals pursue BDSM activities, in
part, due to the pleasant altered states of consciousness these activities produce.
However, to date, no research has tested whether BDSM activities actually facilitate
altered states. To this end, we randomly assigned 14 experienced BDSM practitioners
to the bottom role (the person who is bound, receiving stimulation, or following orders)
or the top role (the person providing stimulation, orders, or structure) for a BDSM
scene. Results suggest that topping was associated with an altered state aligned with
Csikszentmihalyi’s (1991) flow (measured with the Flow State Scale), and bottoming
was associated with an altered state aligned with Dietrich’s (2003) transient hypofron-
tality (measured with a Stroop test) as well as some facets of flow. Additional results
suggest that BDSM activities were associated with reductions in psychological stress
and negative affect, and increases in sexual arousal.
Keywords: sadomasochism, altered states of consciousness, flow, sexuality
Supplemental materials: http://dx.doi.org/10.1037/cns0000097.supp
Research suggests that a substantial minority
of people fantasize about or engage in bondage/
discipline, dominance/submission, and sadism/
masochism (BDSM) activities (Jozifkova &
Flegr, 2006; Moser & Levitt, 1987; but see
Richters, de Visser, Rissel, Grulich, & Smith,
2008, for lower prevalence rates when respon-
dents are asked about “B&D or S&M,” specif-
ically). A BDSM taxonomy developed by Ali-
son, Santtila, Sandnabba, and Nordling (2001)
James K. Ambler, Ellen M. Lee, and Kathryn R. Klement,
Department of Psychology, Northern Illinois University; Tonio
Loewald, Redwood City, California; Evelyn M. Comber, De-
partment of Psychology, Northern Illinois University; Sarah A.
Hanson, Department of Sociology, Georgia State University;
Bert Cutler and Nadine Cutler, Tempe, Arizona; Brad J. Sagarin,
Department of Psychology, Northern Illinois University.
This study was generously supported by a grant from
CLAW (http://clawinfo.org/) and the Community-
Academic Consortium for Research on Alternative Sex-
ualities (CARAS; https://carasresearch.org/), a donation
of space by the Arizona Power Exchange (APEX; http://
arizonapowerexchange.org/), and travel funds from
Northern Illinois University’s Presidential Commission
on Sexual Orientation and Gender Identity. We thank
Archer, Tess, Mick, and South for help with logis-
tics, Leslie Matuszewich for analyzing the saliva
samples, and John Skowronski for comments on the
manuscript.
Correspondence concerning this article should be ad-
dressed to Brad J. Sagarin, Department of Psychology,
Northern Illinois University, DeKalb, IL 60115. E-mail:
bsagarin@niu.edu
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Psychology of Consciousness: Theory, Research, and Practice © 2016 American Psychological Association
2016, Vol. 3, No. 3, 000 2326-5523/16/$12.00 http://dx.doi.org/10.1037/cns0000097
1
illustrates the wide range of behavior encom-
passed by BDSM, with categories of physical
restriction (e.g., bondage, handcuffs), adminis-
tration of pain (e.g., clothespins, spanking), hu-
miliation (e.g., verbal humiliation, gags), and a
category the authors refer to as hypermasculin-
ity (e.g., rimming, dildos). Although BDSM is
often conceptualized as inherently sexual,
Wiseman’s (1998) definition of SM from SM
101 recognized both sexual and nonsexual mo-
tivations for BDSM:
The knowing use of psychological dominance and sub-
mission, and/or physical bondage, and/or pain, and/or
related practices in a safe, legal, consensual manner in
order for the participants to experience erotic arousal
and/or personal growth. (p. 10)
Likewise, some BDSM practitioners incorpo-
rate sexual behaviors such as oral sex or sexual
intercourse into their BDSM activities, whereas
other BDSM practitioners engage in BDSM ac-
tivities that do not include these types of sexual
behaviors.
Early psychiatrists contended that BDSM
practitioners suffered from mental illness (e.g.,
Freud, 1938), but recent research refutes this
conclusion. For example, Connolly (2006) com-
pared BDSM practitioners to published norms
on 10 psychological pathologies. Compared
with non-BDSM practitioners, BDSM practitio-
ners exhibited lower levels of depression, anx-
iety, posttraumatic stress disorder, psychologi-
cal sadism, psychological masochism,
borderline pathology, and paranoia, although
the researcher reported similar levels of obses-
sive– compulsive disorder and higher levels of
dissociation and narcissism across groups. Sim-
ilarly, Wismeijer and van Assen (2013) com-
pared BDSM practitioners with non-BDSM
practitioners on major personality traits. They
found that compared with non-BDSM practitio-
ners, BDSM practitioners exhibited higher lev-
els of extraversion, conscientiousness, openness
to experience, and subjective well-being, and
lower levels of neuroticism, rejection sensitiv-
ity, and agreeableness (see also Hébert &
Weaver, 2014, for an examination of differen-
tial personality characteristics associated with
dominant BDSM practitioners and submissive
BDSM practitioners). Consistent with the per-
spective that BDSM does not stem from pathol-
ogy, Faccio, Casini, and Cipolletta (2014)
wrote,
In line with previous literature (Hoff, 2006; Richters et
al., 2008), our findings support the idea that BDSM is
a sexual interest or subculture attractive to a minority,
rather than a pathological symptom that may be de-
rived from past abuse or difficulty with “normal” sex.
(p. 761)
Thus, rather than deriving from pathology,
the motivation for BDSM may be rooted in
normal psychological functioning. Researchers
have forwarded a range of explanations for
BDSM behaviors including conceptualizing
BDSM as serious leisure (Newmahr, 2010), as a
method of increasing eroticism (Ortmann &
Sprott, 2013), as a common interest around
which meaningful communities develop (Gra-
ham, Butler, McGraw, Cannes, & Smith, 2015),
and as a method of temporarily escaping the
burdens of selfhood (Baumeister, 1988). Al-
though we find value in each of these explana-
tions, we focus on Baumeister’s theory because
it provides the most direct links to one of the
altered states of consciousness that the present
study investigates.
Baumeister’s (1988) escape from self-theory
argued that masochism, like many non-BDSM
activities (e.g., extreme sports, watching TV or
movies, playing games, using alcohol or drugs),
is one mechanism through which people can
escape from the sometimes burdensome self:
. . . High-level awareness can lead to anxiety and
discomfort under some circumstances. The require-
ments of making decisions under pressure or uncer-
tainty, of taking responsibility for actions that may
disappoint or harm others, of maintaining a favorable
public image and private image of self despite all
threats and challenges, and of asserting control over a
recalcitrant social environment can become oppressive
and stressful and can foster desires to escape. (p. 29)
According to Baumeister, many elements of
masochism, such as pain without permanent
harm or damage, bondage, diminished personal
power, and degradation facilitate escape from
the self. Consistent with this, interviewees in
Hébert and Weaver’s (2015) study of BDSM
practitioners identified psychological release as
one of the benefits of BDSM. Likewise, Pas-
coal, Cardoso, and Henriques (2015) found that
BDSM practitioners reported lower levels of
sexual dysfunction when engaged in BDSM
sexual activity compared to non-BDSM sexual
activity, possibly because “BDSM practices can
divert distress away from concerns about sexual
functioning” (p. 1059).
2 AMBLER ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Motivation to escape from the self may spur
people to seek pleasant altered states of con-
sciousness that may be experienced during
BDSM activities to facilitate escape (Newmahr,
2010; Pitagora & Ophelian, in press). This pos-
sibility is supported by the reports of BDSM
practitioners, who use the term subspace to re-
fer to a unique, subjectively pleasant, altered
state of consciousness that is sometimes expe-
rienced by the receiver of sensation (i.e., the
bottom) in a BDSM scene.
1
Although Baumeis-
ter’s (1988) theory only accounts for altered
states of consciousness in bottoms, BDSM prac-
titioners also describe a subjectively pleasant,
altered states of consciousness called topspace
that is sometimes experienced by the giver of
sensation (i.e., the top) in a BDSM scene. As
described by Newmahr (2010),
Bottoming is more likely to result in observable altered
consciousness than topping, but the community is also
more highly motivated to recognize altered states in
bottoms than in tops. The flow experience for bottoms
is most frequently called “subspace” or “bottom
space,” and has many other descriptors, including
“loopy,” “flying,” and “fried.” When the flow experi-
ence in topping is recognized discursively, these
phrases include “in the zone,” “grooving” or “in top
space.” (p. 327)
The goal of our study is to empirically test
whether BDSM behaviors facilitate altered sub-
jective experiences. In doing so, we link topping
and bottoming in BDSM scenes to different
altered states of consciousness. Consistent with
Newmahr (2010), we suggest that topping may
facilitate an altered state aligned with Csik-
szentmihalyi’s (1991) concept of mental flow.
Although Newmahr also linked bottoming to
flow, we suggest that bottoming may facilitate
an altered state more closely aligned with Di-
etrich’s (2003) ideas about transient hypofron-
tality. These concepts are reviewed in the sec-
tions that follow.
Flow
Csikszentmihalyi (1991) conceptualized flow
as an altered state of consciousness achieved
during optimal experiences. According to Csik-
szentmihalyi, flow contains nine dimensions:
“challenge-skill balance, action-awareness
merging, clear goals, unambiguous feedback,
concentration on task, sense of control, loss of
self-consciousness, time transformation, and
autotelic experience” (Jackson, Eklund, & Mar-
tin, 2010, p. 6), and is, subjectively, extremely
rewarding.
As noted earlier, Newmahr (2010) already
has forged a link between BDSM activities and
flow. She explained,
SM participants speak of their play in terms of ecstatic
experience, or what can be understood as flow. They
speak of weightlessness: of grooving and flying, of the
cessation of cognitive process and of the disappearance
of the world around them. (p. 328)
However, Newmahr also distinguished the ways
that tops and bottoms achieve flow, noting that
“Tops achieve flow through mental focus, par-
ticularly when engaged in activities that require
intense concentration, such as knife play, needle
play, and advanced bondage” (p. 328), whereas
“When bottoming, players experience flow as a
result of intense rhythmic sensation, sensation
or pain itself, unrelenting focus on a particular
task or concentrated effort to endure a sensation
or circumstance” (p. 328).
We similarly predict that BDSM will be
linked to the flow experience and that this link
will be especially strong for practitioners who
adopt the top role in a scene. We advance this
hypothesis because flow denotes optimal per-
formance on controlled, purposeful activities,
which are the types of activities that a top per-
forms during a scene, that is, directing and
controlling the person in the bottom role. In
contrast, participants who adopt the bottom role
in a scene temporarily cede power, control, and
decision-making to the top, although mecha-
nisms such as safewords enable the bottom to
reassert control if needed during a scene (Pi-
tagora, 2013), and abandoning these mental ac-
tivities does not seem to be consistent with the
flow experience described by Csikszentmihalyi
(1991).
Transient Hypofrontality
Participants in the bottom role may experi-
ence a distinct altered state of consciousness
1
Sagarin, Cutler, Cutler, Lawler-Sagarin, and Matusze-
wich (2009) define scenes as “defined periods of time dur-
ing which participants engage in a series of [BDSM] activ-
ities”, bottoms as “participants who [are] bound, receiving
stimulation, following orders, etc.,” and tops as “partici-
pants who provide stimulation, orders, or structure” (p.
187).
3BDSM AND ALTERED STATES
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
related to Dietrich’s (2003) transient hypofron-
tality hypothesis, which is also pertinent to an
array of subjective altered states of conscious-
ness including daydreaming, runner’s high,
meditation, and some drug highs.
The notion of transient hypofrontality rests
on two premises: (a) the brain possesses limited
resources that brain structures, systems, and ar-
eas compete for, and (b) the subjective experi-
ence of consciousness is a process. When activ-
ities such as exercise increase the demands on
brain areas responsible for basic sensory and
perceptual processes, autonomic nervous sys-
tem regulation, and motor output, the brain does
not receive additional blood flow to meet com-
peting demands (Dietrich, 2003). Instead, the
brain down-regulates certain regions to increase
blood flow to currently important areas of de-
mand. Research suggests that the frontal cortex
and the prefrontal cortex of the brain are con-
sistently down-regulated in this way (Del
Giorno, Hall, O’Leary, Bixby, & Miller, 2010;
Dietrich & Sparling, 2004).
Because the dorsolateral prefrontal cortex is
heavily active in working memory and sus-
tained attention (Cabeza & Nyberg, 2000), such
down-regulation can produce time distortions,
disinhibition from social constraints, and
changes in focused attention. Additionally, peo-
ple theorized to be experiencing transient hypo-
frontality report reductions in pain, decision
making activity, logic, and difficulty with mem-
ory, along with increased feelings of floating,
peacefulness, and living in the here and now
(Dietrich, 2003).
To examine a possible alignment of the ef-
fects of transient hypofrontality with subspace,
we asked workers (N⫽56) on Amazon’s Me-
chanical Turk (MTurk) who had experienced
subspace to describe the experience (see
Buhrmester, Kwang, & Gosling, 2011, for evi-
dence that “the data obtained [through MTurk]
are at least as reliable as those obtained via
traditional methods,” p. 3; though see Chandler,
Paolacci, Peer, Mueller, & Ratliff, 2015, for
concerns regarding the effects of MTurk work-
ers participating in multiple studies). Aside
from one respondent who interpreted subspace
from a Star Trek perspective, respondents’ de-
scriptions were consistent with characteristics
of transient hypofrontality:
With my girl, she was punishing me, and I felt a deep
sense of pain and pleasure, sort of a one with nature
and my environment, it was incredible.
It’s very dreamlike. I knew that I felt good, and I was
aware of my partner, but I was not paying attention to
anything else.
An almost trance like state. Pain didn’t exist in any
form.
Quotes from Wiseman (1998) are also consis-
tent with characteristics of transient hypofron-
tality:
Whenever I’m in bottom space I become sort of non-
verbal. It becomes hard for me to remember how to
talk. (p. 33)
I cannot relax unless I’m really tied up. (p. 134)
(from a professional dominant): I see a lot of engineers.
A heavy whipping is the only thing that brings them
out of their heads and into their bodies— out of their
thoughts and into the here and now. (p. 177)
In addition to these subjective alterations, a
state of transient hypofrontality would also be
expected to be associated with impairments on
objective tasks that involve the down-regulated
brain areas. One testable implication of this
possibility in the BDSM domain is that bottoms
would be expected to exhibit decrements in
performance on cognitive tests that assess the
ability to exert executive control during and
following engagement in a scene. For example,
after a BDSM scene, bottoms, but not tops, may
evince performance impairment on a Stroop
task (MacLeod, 1991).
Other Physiological and Psychological
Effects of BDSM Activities
Although, to our knowledge, no prior quan-
titative studies have examined whether BDSM
scenes facilitate altered states, previous research
suggests a number of other physiological and
psychological effects of BDSM activities. Sa-
garin, Cutler, Cutler, Lawler-Sagarin, and
Matuszewich (2009) had 58 experienced
BDSM practitioners complete surveys and pro-
vide saliva samples before and after participat-
ing in BDSM scenes. Saliva samples were used
to measure cortisol (a hormone associated with
stress) and testosterone (a hormone associated
with dominance and aggression). From before
to after their scenes, bottoms (but not tops)
showed increases in physiological stress, as
4 AMBLER ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
measured by cortisol, whereas both bottoms and
tops showed increases in relationship closeness.
Klement et al. (2015) administered surveys
and collected saliva samples from participants
engaged in a ritual involving temporary pierc-
ing and pulling against hooks placed through
the piercings. Ritual participants showed in-
creases in physiological stress, as measured by
cortisol, but reductions in psychological stress
and negative affect. Although the ritual did not
constitute a traditional BDSM scene, nor was it
conceptualized by participants as a BDSM
scene (participants perceived the experience as
more spiritual than sexual or sadomasochistic),
we anticipated that the physiological and psy-
chological effects might mirror those observed
among bottoms in BDSM scenes.
The Present Study
In the present study, we used an experimental
methodology to test the facilitative effects of
BDSM role on altered states of consciousness as
well as other physiological and psychological
variables. In the BDSM community, some indi-
viduals self-identify as switches. Switches
sometimes take on the bottom role and some-
times take on the top role in BDSM scenes. We
recruited seven pairs of switches who were will-
ing to be randomly assigned (by roll of a die) to
either the top or bottom role for a BDSM scene.
In contrast with prior studies, in which partici-
pants self-selected into the top or bottom role
(e.g., Sagarin et al., 2009), the random assign-
ment of switches to the top or bottom role
removes some of the ambiguity in interpreting
the results. That is, observed differences be-
tween tops and bottoms in the present study can
be attributed more confidently to the activities
performed by that role, rather than to the type of
person who chooses that role.
After each scene ended, participants com-
pleted a measure of mental flow. We predicted
that tops would report greater flow than bot-
toms. We also asked participants to complete
three Stroop tests: one prior to random assign-
ment, one just before their scene began, and one
just after their scene ended. We predicted that
after the scene, the performance of bottoms on
the test would be impaired relative to the per-
formance of tops. In particular, we predicted a
Role ⫻Time interaction such that bottoms
would show a greater increase in Stroop scores
from baseline to after their scenes than would
tops (higher Stroop scores represent worse per-
formance on the Stroop test).
Finally, based on Sagarin et al. (2009) and
Klement et al. (2015), we measured cortisol,
testosterone, psychological stress, self-other
overlap (a measure of relationship closeness,
Aron, Aron, & Smollan, 1992), positive and
negative affect, sexual arousal, and perceptions
of the scene. We predicted that both tops and
bottoms would show increases in self-other
overlap (consistent with Sagarin et al., 2009),
decreases in negative affect, and decreases in
self-reported stress from before to after their
scenes (consistent with Klement et al., 2015).
We also predicted Role ⫻Time interactions for
cortisol (consistent with Sagarin et al., 2009)
such that cortisol would increase from baseline
to after scenes for bottoms, whereas no change
would be evident for tops. With respect to tes-
tosterone, we predicted that bottoms would ex-
hibit decreases in testosterone from baseline to
just before their scenes, whereas tops would
exhibit increases, based on the idea that BDSM
practitioners might evidence anticipatory
changes in testosterone consistent with the
dominant or submissive role they are about to
adopt.
Method
Participants
Fourteen participants enrolled in the experi-
ment in seven pairs. Ten participants were
women; four were men. Participants ranged in
age from 23 to 64 years, M⫽40.86 years,
SD ⫽12.47 years. All participants were Cau-
casian. Ten participants reported their sexual
orientation as bisexual, three as heterosexual,
and one as gay. Seven participants indicated
they had no religion or were atheist, two re-
ported being spiritual or “woo,” and the five
others respectively reported their religion as
Buddhist, Hindu, Pagan, Christian, and nature-
based. All participants were experienced BDSM
practitioners, with experience ranging from 1 to
17 years, M⫽6.86 years, SD ⫽5.07 years.
Two pairs of participants were in a polyamorous
relationship, two pairs were in long-term dating
relationships, two pairs were friends, and one
pair had just met the evening of the study.
Participants reported they had performed be-
5BDSM AND ALTERED STATES
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
tween 18 and 1,500 prior scenes overall and
between 0 and 250 prior scenes with their study
partner.
Materials
Surveys. Participants responded to a total
of three paper surveys. The baseline survey
contained the Positive and Negative Affect
Schedule (PANAS, Watson, Clark, & Tellegen,
1988), measures (scale anchors are presented in
the results section) of self-reported stress and
sexual arousal, how much participants would
use the word “we” to describe their relationship
with their scene partner, the Inclusion of Other
in Self Scale (IOS, Aron et al., 1992, a one-item
scale consisting of seven pairs of increasingly
overlapping circles labeled “self” and “other”;
see Aron et al. for evidence of the reliability and
validity of the IOS including convergent valid-
ity with measures of relationship closeness and
intimacy), and demographics, including ques-
tions regarding previous BDSM experience.
The prescene survey was administered just prior
to the beginning of a scene and only included
the PANAS, self-reported stress, and sexual
arousal. Finally, the postscene survey included
the following: a retrospective PANAS (“How
did you feel at the end of your scene?”) and a
current PANAS (“How do you feel now?”),
retrospective and current psychological stress
and sexual arousal, use of the word “we” and
IOS, how participants felt the scene had gone,
how much the scene had been affected by the
random assignment to role, how spiritual, sex-
ual, sadomasochistic, and intense participants
found the scenes, and the Flow State Scale
(Jackson et al., 2010; Jackson & Marsh, 1996).
Cronbach’s alpha for positive affect was .89 in
Watson et al. (1988) and ranged from .71 to .91
in the present study.
2
Cronbach’s alpha for neg-
ative affect was .85 in Watson et al. Cronbach’s
alphas could not be calculated for negative af-
fect in the present study as some of the items
had 0 variance (e.g., all participants responded
“very slightly or not at all” for “ashamed”). Past
research supports the use of the PANAS in
repeated-measures designs (Brose, Voelkle,
Lövdén, Lindenberger, & Schmiedek, 2015;
Kennedy-Moore, Greenberg, Newman, &
Stone, 1992).
Saliva samples. Participants provided five
saliva samples over the course of the experi-
ment. Saliva samples were placed in an ice
chest immediately after collection. Subse-
quently, samples were stored in a household
freezer before being sent to Northern Illinois
University for analysis. Samples were shipped
on dry ice via overnight delivery and then stored
at ⫺65°C prior to analysis. Samples were ana-
lyzed for cortisol and testosterone (see Sagarin
et al., 2009, Study 2 for information on the
procedures for analysis; correlations for dupli-
cate assays in the present study were r(64) ⫽
.98, p⬍.001 for cortisol, r(63) ⫽.98, p⬍.001
for testosterone).
Flow. Flow was measured using the Flow
State Scale (Jackson et al., 2010; Jackson &
Marsh, 1996). This 36-item scale (1 ⫽strongly
disagree,2⫽disagree,3⫽neither agree nor
disagree,4⫽agree,5⫽strongly agree) mea-
sures nine dimensions of flow: Challenge-Skill
Balance (a feeling of balance between the de-
mands of the situation and personal skills), Ac-
tion-Awareness Merging (a feeling of being so
involved that one’s actions are automatic),
Clear Goals (a feeling of knowing exactly what
is needed in the situation on a moment to mo-
ment basis), Unambiguous Feedback (immedi-
ate and clear feedback that is seamlessly inte-
grated into the ongoing activity), Concentration
on Task at Hand (a feeling of being totally
focused), Sense of Control (an empowering
feeling of being in control and being free of the
fear of failure), Loss of Self-Consciousness
(concern for the self disappears), Transforma-
tion of Time (a change in how the passage of
time is perceived, either faster, slower, or a lack
of awareness), and Autotelic Experience (a feel-
ing of intrinsic reward, task is enjoyable in and
of itself). Each dimension is measured by four
items; participants were instructed to respond to
each item in terms of how they felt during the
scene. We included an “N/A” option for partic-
ipants who felt an item was not relevant to their
experience. Cronbach’s alpha for the Flow State
Scale was .80 to .92 in Jackson and Eklund
(2002) and .89 in the present study. Jackson,
Martin, and Eklund (2008) reported convergent
validity of the Flow State Scale with measures
2
The range of values for Cronbach’s alpha refers to the
Cronbach’s alphas calculated across the repeated measures
of each scale.
6 AMBLER ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
of intrinsic motivation, positive well-being, and
psychological distress.
The Stroop test. The Stroop test (Ma-
cLeod, 1991) was administered on an iPad and
consisted of 40 trials in which a character string
(red,blue,green,yellow,orxxxx) was displayed
in red, blue, green, or yellow font color (each of
the 20 combinations was presented twice). Par-
ticipants were instructed to ignore the semantic
meaning of the word and to respond to each by
pressing a button on the screen indicating the
font color (labeled in black type red,blue,
green, and yellow). No feedback was provided
for correct or incorrect answers. Stroop scores
reflect the difference between the mean laten-
cies on incongruent trials (e.g., red in green
type) and the mean latencies on control trials
(e.g., red in red type, xxxx in any color of type)
with higher Stroop scores indicating greater rel-
ative difficulty in responding to incongruent
trials. Trials with erroneous responses and tri-
als with latencies ⬎3SD from a participant’s
mean latency were removed (Linnman, Carl-
bring, Ahman, Andersson, & Andersson,
2006). Alternative analyses that include these
erroneous and ⬎3SD trials do not substan-
tively change the results. The Stroop test has
been used successfully in repeated-measures
designs, although past research is mixed on
whether the Stroop test shows practice effects
(see MacLeod, 1991, for a review).
Because our predictions specify a decrement
in Stroop performance for bottoms, practice ef-
fects would work against our hypothesis. Fur-
ther, Stroop results in the present study suggest
the absence of practice effects in our data.
Procedure
Participants were recruited through an-
nouncements at the host organization (the Ari-
zona Power Exchange), through e-mail to a
mailing list of individuals who had previously
expressed interest in BDSM research, and
through postings on FetLife. Recruitment mate-
rials explained that we were recruiting pairs of
participants who would be willing to be ran-
domly assigned to the top role or bottom role in
a BDSM scene to be performed with their study
partner. No restrictions were placed on the
scene activities, apart from the random assign-
ment to the top and the bottom. Thus, partici-
pants could choose the activities they wished to
perform, the length of their scene, and so forth.
Recruitment materials also explained that all
scenes would take place on the same evening at
the facility of the Arizona Power Exchange.
After reading the informed consent form and
providing verbal agreement,
3
participants com-
pleted (a) the baseline survey, (b) a practice
Stroop, and (c) a baseline Stroop and provided a
baseline saliva sample. Participants were then
randomly assigned to the top or bottom role for
their scene via a die role and a researcher was
assigned to each scene.
4
Participants were then
asked to wait at least 30 min prior to beginning
their scenes to allow for changes in cortisol or
testosterone to appear in the saliva. Some par-
ticipants began setting up equipment for their
scenes prior to 30 min passing, but all waited at
least 30 min to begin their scenes.
Just before participants began scene activi-
ties, the researcher administered the prescene
survey and the prescene Stroop and obtained a
prescene saliva sample. Participants then con-
ducted their scene. A separate researcher was
assigned to observe each scene (all participants
consented to observation, and all observations
were made within sight of the scene; thus, par-
ticipants could see the researcher observing
their scene), and the researcher’s priority during
the scene was to administer the study measures
and record timestamps of when those measures
were administered. However, the researchers
were also instructed to code whether certain
activities happened during the scene (yes/no).
Before the study began the researchers received
training on how to categorize BDSM activities
on a specific coding sheet. The training included
watching video clips of BDSM scenes and go-
ing over responses as a group. Additionally, the
researchers could reference the coding sheet
category definitions and examples at any time
while they were observing their assigned scene.
As soon as the primary scene activities
were over, the researcher administered a post-
3
The study was approved by Northern Illinois Universi-
ty’s Institutional Review Board. We requested and received
approval for a verbal informed consent procedure to protect
the anonymity of participants.
4
The number of simultaneous scenes prevented us from
assigning more than one researcher to a scene. All research-
ers were trained in the study’s methodology and observation
protocol. Further all researchers had practiced the observa-
tion protocol using video clips of BDSM scenes.
7BDSM AND ALTERED STATES
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
scene Stroop and obtained a postscene saliva
sample. Participants then began aftercare, if
they wished. Sagarin et al. (2009) define af-
tercare as “participant interaction after com-
pleting the main scene activities and before
leaving the play area” that, in their study,
“included gentle contact (e.g., hugging, cud-
dling), pleasurable stimulation (e.g., caress-
ing, massaging), and communication (e.g.,
talking, eye contact)” (p. 190).
Additional saliva samples were collected 20
min later and 40 min later. When participants
were ready, they completed the postscene sur-
vey, which included the flow measure.
Results
Missing Data
We obtained complete data on all measures
with three exceptions: (a) On the prescene
PANAS, one participant left one positive affect
item blank. This participant’s prescene positive
affect score was calculated using the other nine
items. (b) One participant did not answer the
question about the impact of random assign-
ment. This participant was excluded from the
analysis of that variable. (c) As detailed below,
a number of bottoms responded “N/A” on some
of the items on the Flow State Scale. For these
participants, dimensions of flow were calculated
using the nonmissing items. In one instance, a
participant responded “N/A” for all four items
within one dimension. For this participant, this
dimension was set as missing. Thereafter, the
total flow score and the two factors of flow were
calculated using the nonmissing dimensions. (d)
Saliva samples from two participants did not
allow us to obtain full measures of testosterone
and cortisol. These participants were excluded
from the corresponding hormonal analyses.
Characteristics of the Scenes
The duration of scenes varied from 37 to 79
min (M⫽57 min, SD ⫽15 min). Aftercare
occurred after six of the scenes and varied in
duration from 7 to 40 min (including 0 for the
scene with no aftercare, M⫽19 min, SD ⫽12
min). The postscene survey was administered
between 30 and 77 min after the end of the
scene (M⫽46 min, SD ⫽16 min). Research on
autobiographical memory and the fading affect
bias suggests that this delay should not impact
participants’ ability to accurately report their
affective state during the scene (Skowronski,
Walker, Henderson, & Bond, 2014).
All seven scenes included some form of gen-
tle, soft touching (feather duster, massage, back
scratching, tickling) and communication (con-
versation, laughter, moaning, whispering). Six
out of seven scenes ended with aftercare (cud-
dling on couch, hugging, talking and discussing
the scene, wrapping bottom in blanket). The
next most frequently observed activities were
seen in five scenes and included during care
(asking how things feel, asking if bottom needs
anything, offering water) and prescene prepara-
tion/negotiation (discussion of preferences,
preparation of toys, deep breathing together).
Four out of seven scenes included some form of
striking (flogging, caning, sharp slap with rod,
spanking). Three scenes involved some form of
bondage (shackles, chains), sexual activity
(contact with genitals), and playing with cloth-
ing (fetish dress, collars, leather gloves). Two
scenes involved pinching/pulling activities
(metal clamps, biting), and explicit control be-
havior (call and response commands). The fol-
lowing activities were only seen in one scene:
role or gender play (puppy play: roleplay in
which one or more people take on the roles of
dogs or puppies), sensory deprivation (blind-
fold), and aural stimulation (making thudding
sound of paddle against palm next to bottom’s
ear).
Perceptions of the Scenes
Two questions assessed scene quality. In re-
sponse to the question, “How well did the scene
go?” (1 ⫽very poorly,7⫽very well), both tops
(M⫽6.29, SD ⫽1.11) and bottoms (M⫽6.57,
SD ⫽.79) reported that their scenes had gone
well. Tops’ and bottoms’ ratings of how well
the scene went did not differ significantly,
t(12) ⫽⫺.56, p⫽.59, d⫽⫺0.29. Tops’ and
bottoms’ rating of scene intensity (1 ⫽not at all
intense,7⫽very intense) also did not differ
significantly (tops: M⫽5.21, SD ⫽1.63; bot-
toms: M⫽3.43, SD ⫽1.81), t(12) ⫽1.94, p⫽
.08, d⫽1.03.
Two questions assessed the impact of the
experiment on the scene. Participants were
asked how much being observed affected their
scene (1 ⫽not at all affected,7⫽very strongly
8 AMBLER ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
affected). Both tops (M⫽2.86, SD ⫽1.86) and
bottoms (M⫽1.71, SD ⫽0.95) reported little
effect of observation. Tops’ and bottoms’ rat-
ings of the effect of observation did not differ
significantly, t(12) ⫽1.45, p⫽.17, d⫽0.78.
Participants were then asked to rate the impact
of the random assignment on their scene (1 ⫽
strongly detracted from my scene,7⫽strongly
enhanced my scene). Tops (M⫽4.00, SD ⫽
1.41) and bottoms (M⫽4.50, SD ⫽1.23)
reported little impact. As with the effect of
observation, tops’ and bottoms’ ratings of the
impact of random assignment did not differ
significantly, t(11) ⫽.68, p⫽.51, d⫽⫺0.38.
Finally, participants responded to three ques-
tions assessing how spiritual, how sadomasoch-
istic, and how sexual they found their scenes to
be (anchors: 1 ⫽not at all,5⫽very).A2
(Role: top vs. bottom) ⫻3 (Characteristic: spir-
itual, sadomasochistic, sexual) analysis of vari-
ance was run to test whether tops and bottoms
differed in their characterization of the scene.
The role by characteristic interaction was non-
significant, but, nonetheless, showed a trend
toward significance, F(2, 24) ⫽2.89, p⫽.08,
p
2⫽.19, with tops perceiving the scene as
highest in sexuality (M⫽3.14, SD ⫽1.35) then
spirituality (M⫽2.86, SD ⫽1.35) and lowest
in sadomasochism (M⫽2.43, SD ⫽1.27) and
bottoms perceiving the scene as highest in sa-
domasochism (M⫽3.29, SD ⫽1.11) then
sexuality (M⫽2.71, SD ⫽1.60) and lowest in
spirituality (M⫽1.86, SD ⫽0.90).
Flow
Items on the 36-item Flow State Scale were
answeredona1(strongly disagree)to5
(strongly agree) scale with scores higher than 3
(neither agree nor disagree) indicating flow.
Participants were also given the option of re-
sponding “N/A” if they felt that an item was not
applicable for their experience. Across all 36
items (ignoring N/As), both tops (M⫽4.11,
SD ⫽0.36) and bottoms (M⫽4.11, SD ⫽0.55)
reported levels of flow significantly above 3,
t(6) ⫽8.07, p⬍.001, t(6) ⫽5.33, p⫽.002,
respectively. Tops and bottoms did not differ
significantly in their reported levels of flow,
t(12) ⫽0.00, p⫽1.00, d⫽0.00. However, an
examination of the N/A responses suggests
greater flow in tops than bottoms. In particular,
bottoms responded N/A on an average of 4.3
items (SD ⫽5.6), whereas tops responded N/A
on no items. Further, bottoms’ N/A responses
were most often on flow dimensions associated
with optimal performance (Challenge-Skill Bal-
ance, Clear Goals, and Immediate Feedback). In
contrast, bottoms provided no N/A responses on
flow dimensions associated with autotelic ab-
sorption (Loss of Self-Consciousness, Transfor-
mation of Time, and Autotelic Experience).
To further examine responses to the different
dimensions of flow, we calculated factor scores
based on the two factors that emerged in a study
on hook-pulls we conducted that also used the
Flow State Scale (Lee et al., 2016). The first
factor consisted of Challenge-Skill Balance,
Clear Goals, Unambiguous Feedback, Concen-
tration on Task at Hand, and Sense of Control,
and the second factor consisted of Action-
Awareness Merging, Loss of Self-Conscious-
ness, Transformation of Time, and Autotelic
Experience. We labeled the first factor optimal
performance and the second factor autotelic ab-
sorption. When we calculated these factors in
the present data, they showed a different pattern
in tops and bottoms. Specifically, tops showed
consistency across both factors of flow, show-
ing no significant differences between optimal
performance (M⫽4.12, SD ⫽0.57) and auto-
telic absorption (M⫽4.09, SD ⫽0.48), t(6) ⫽
0.12, p⫽.91. Bottoms, in contrast, showed
significantly lower levels of optimal perfor-
mance (M⫽3.69, SD ⫽0.90) compared to
autotelic absorption (M⫽4.58, SD ⫽0.41),
t(6) ⫽⫺2.66, p⫽.04. These results are con-
sistent with the N/A responses of bottoms,
which were most often on the optimal perfor-
mance dimensions of flow.
Stroop Scores
Table 1 and Figure 1 display the Stroop test
scores for tops and bottoms at baseline (prior to
random assignment to role), just prior to begin-
ning their scenes, and just after finishing their
scenes. Because the prediction regarding Stroop
scores involved only the baseline and postscene
Stroop tests, Stroop scores were analyzed twice,
once with prescene Stroop scores included and
once with prescene Stroop scores excluded.
For the first analysis, the Role (top vs. bot-
tom) ⫻Time (baseline vs. prescene vs. post-
scene) interaction was significant, F(2, 24) ⫽
6.69, p⫽.005, p
2⫽.36. The main effects of
9BDSM AND ALTERED STATES
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
role and time were nonsignificant, F(1, 12) ⫽
0.16, p⫽.69, p
2⫽.01, F(2, 24) ⫽2.69, p⫽
.09, p
2⫽.18, respectively.
For the second analysis, the predicted Role
(top vs. bottom) ⫻Time (baseline vs. post-
scene) interaction was significant, F(1, 12) ⫽
6.64, p⫽.02, p
2⫽.36. Examination of the
simple time of administration effects (baseline
vs. postscene) within each role showed a signif-
icant simple effect of time for bottoms, F(1,
12) ⫽10.62, p⫽.007, p
2⫽.47, but not for
tops, F(1, 12) ⫽0.15, p⫽.71, p
2⫽.01.
Cortisol
Table 2 displays cortisol levels for tops and
bottoms at baseline (prior to random assignment
to role), just prior to beginning their scenes, just
after finishing their scenes, 20 min later, and 40
min later. The omnibus Role (top vs. bottom) ⫻
Time (baseline vs. prescene vs. postscene vs.
20-min postscene vs. 40-min postscene) inter-
action was nonsignificant, F(4, 40) ⫽1.57, p⫽
.20, p
2⫽.14. The main effects of role and time
were also nonsignificant, F(1, 10) ⫽1.34, p⫽
.27, p
2⫽.12, F(4, 40) ⫽0.21, p⫽.93, p
2⫽
.02, respectively. The predicted Role (top vs.
bottom) ⫻Time (baseline vs. postscene) inter-
action was nonsignificant, but, nonetheless,
showed a trend toward significance, F(1, 10) ⫽
3.96, p⫽.08, p
2⫽.28. Simple effects within
each role showed nonsignificant simple effects
of time for bottoms, F(1, 10) ⫽2.97, p⫽.12,
p
2⫽.23, and tops, F(1, 10) ⫽1.09, p⫽.32,
p
2⫽.10.
Testosterone
Table 2 displays testosterone levels for tops
and bottoms at baseline (prior to random assign-
ment to role), just prior to beginning their
scenes, just after finishing their scenes, 20 min
later, and 40 min later. The omnibus Role ⫻
Time interaction was nonsignificant, F(4, 40) ⫽
0.35, p⫽.84, p
2⫽.03. The main effects of role
0
100
200
300
400
500
600
Baseline Pre-scene Immediate post-
scene
Stroop scores
Top
Bottom
Figure 1. Stroop scores for tops and bottoms. Stroop
scores reflect the difference between the mean latencies on
incongruent trials (e.g., red in green type) and the mean
latencies on control trials (e.g., red in red type, xxxx in any
color of type) with higher Stroop scores indicating greater
relative difficulty in responding to incongruent trials.
Table 1
Psychological Measures in Tops and Bottoms
Measure Role Baseline Prescene Immediate postscene Postscene
Stroop scores Top 180.2 (267.3) 273.9 (164.0) 131.6 (177.2) —
Bottom 92.3 (128.1) 83.2 (140.8) 504.3 (384.6) —
Positive affect Top 34.86 (4.56) 41.71 (5.41) 43.43 (4.79) 35.86 (6.39)
Bottom 38.43 (4.65) 40.25 (5.36) 33.29 (11.32) 33.71 (5.19)
Negative affect Top 13.57 (2.30) 11.43 (1.13) 10.57 (.98) 10.57 (1.13)
Bottom 14.86 (1.68) 13.71 (2.43) 10.14 (.38) 10.29 (.49)
Stress Top 1.71 (.76) 1.43 (.79) 1.00 (.00) 1.00 (.00)
Bottom 1.86 (.90) 1.43 (.54) 1.00 (.00) 1.00 (.00)
Sexual arousal Top 2.57 (.98) 3.14 (1.22) 3.57 (1.40) 2.86 (1.07)
Bottom 1.71 (.76) 2.57 (1.51) 2.86 (1.77) 2.14 (.90)
We Top 5.14 (2.34) — — 5.00 (2.31)
Bottom 4.14 (2.61) — — 4.29 (2.63)
IOS Top 3.14 (1.86) — — 4.14 (1.77)
Bottom 2.71 (1.89) — — 3.00 (2.24)
Note. IOS ⫽Inclusion of Other in Self Scale. Stroop scores reflect the difference between the mean latencies on
incongruent trials (e.g., red in green type) and the mean latencies on control trials (e.g., red in red type, xxxx in any color
of type). Immediate postscene measures were retrospective for affect, stress, and sexual arousal.
10 AMBLER ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
and time were also nonsignificant, F(1, 10) ⫽
1.53, p⫽.24, p
2⫽.13, F(4, 40) ⫽1.13, p⫽
.26, p
2⫽.10, respectively. Finally, the pre-
dicted Role ⫻Time (baseline vs. prescene)
interaction was also nonsignificant, F(1, 10) ⫽
0.39, p⫽.55, p
2⫽.04.
Other Psychological Effects
Table 1 contains descriptive statistics for
tops’ and bottoms’ levels of positive affect,
negative affect, stress, sexual arousal, the use
of “we” to describe their relationship, and
self-other overlap. For positive affect, partic-
ipants indicated the extent to which they were
feeling 10 positive affect terms (e.g., inter-
ested, excited) on a 5-point scale (1 ⫽very
slightly or not at all,2⫽a little,3⫽mod-
erately,4⫽quite a bit,5⫽extremely). For
negative affect, participants indicated the ex-
tent to which they were feeling 10 negative
affect terms (e.g., distressed, upset) on the
same scale. For stress and sexual arousal,
participants indicated the extent to which they
were feeling “stressed” and “sexually
aroused” on the same scale. Use of “we” was
indicated on a 7-point scale (1 ⫽I definitely
would not use the word “we” to describe my
relationship,7⫽I definitely would use the
word “we” to describe my relationship). Self-
other overlap was indicated on a scale dis-
playing seven sets of increasingly overlap-
ping circles labeled “self” and “other.”
Positive affect. For positive affect, there
was a significant main effect of time (baseline
vs. prescene vs. retrospective immediate post-
scene vs. later postscene), F(3, 36) ⫽4.00,
p⫽.02, p
2⫽.25, no main effect of role, F(1,
12) ⫽1.04, p⫽.33, p
2⫽.08, and a signif-
icant Time ⫻Role interaction, F(3, 36) ⫽
4.63, p⫽.01, p
2⫽.28 (see Figure 2). From
baseline to before the scene began, tops re-
ported a significant increase in positive affect,
F(1, 12) ⫽21.17, p⫽.001, p
2⫽.64,
whereas bottoms did not, F(1, 12) ⫽1.50,
p⫽.24, p
2⫽.11. From before the scene
began to immediately after the scene ended,
bottoms reported a significant decrease in
positive affect, F(1, 12) ⫽5.41, p⫽.04, p
2⫽
.31, whereas tops did not, F(1, 12) ⫽0.33,
p⫽.58, p
2⫽.03. Thereafter, from immedi-
ately after the scene ended to well after the
scene ended, tops reported a significant de-
crease in positive affect, F(1, 12) ⫽9.06, p⫽
.01, p
2⫽.43, whereas bottoms did not, F(1,
12) ⫽0.03, p⫽.87, p
2⫽.00.
Negative affect. For negative affect, there
was a significant main effect of time, F(3, 36) ⫽
27.51, p⬍.001, p
2⫽.70, no main effect of
role, F(1, 12) ⫽1.94, p⫽.19, p
2⫽.14, and a
significant Time ⫻Role interaction, F(3, 36) ⫽
3.39, p⫽.03, p
2⫽.22 (see Figure 3). Both tops
and bottoms reported decreases in negative af-
fect from baseline to postscene, but the decrease
in tops’ negative affect occurred from baseline
to prescene, whereas the decrease in bottoms’
negative affect occurred from prescene to im-
mediate postscene.
Table 2
Hormonal Measures in Tops and Bottoms
Measure Role Baseline Prescene Immediate postscene 20 min later 40 min later
Cortisol Top .113 (.075) .143 (.151) .090 (.043) .096 (.032) .099 (.056)
Bottom .144 (.093) .121 (.076) .189 (.099) .158 (.071) .135 (.057)
Testosterone Top 120.9 (73.0) 110.6 (44.4) 85.2 (49.5) 84.0 (38.4) 81.9 (45.7)
Bottom 186.8 (258.4) 129.9 (90.3) 115.3 (33.3) 158.8 (77.0) 96.4 (32.3)
Note. The SD for baseline testosterone for bottoms is due to an extremely high score for one participant (647.0). Exclusion
of this score does not substantively change the testosterone results.
10
15
20
25
30
35
40
45
50
Baseline Pre-scene Immediate
post-scene
Post-scene
Positive affect
Top
Bottom
Figure 2. Positive affect scores for tops and bottoms.
11BDSM AND ALTERED STATES
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Psychological stress. For psychological
stress, there was a significant main effect of
time, F(3, 36) ⫽8.92, p⬍.001, p
2⫽.43, no
main effect of role, F(1, 12) ⫽.04, p⫽.85,
p
2⫽.003, and no Time ⫻Role interaction,
F(3, 36) ⫽.08, p⫽.97, p
2⫽.01. Both tops and
bottoms reported decreases in psychological
stress from baseline to prescene and then from
prescene to postscene.
Sexual arousal. For sexual arousal, there
was a significant main effect of time, F(3, 36) ⫽
5.12, p⬍.01, p
2⫽.30, no main effect of role,
F(1, 12) ⫽1.62, p⫽.23, p
2⫽.12, and no
Time ⫻Role interaction, F(3, 36) ⫽.08, p⫽
.97, p
2⫽.01. Both tops and bottoms reported
increases in sexual arousal from baseline to
prescene and from prescene to immediately
postscene. Sexual arousal decreased thereafter
in both tops and bottoms.
“We.” For the use of the word “we” to de-
scribe their relationship with their scene partner,
there was no main effect of time (baseline vs.
postscene), F(1, 12) ⫽.00, p⫽1.0, p
2⬍.001, no
main effect of role, F(1, 12) ⫽.43, p⫽.53, p
2⫽
.03, and no Time ⫻Role interaction, F(1, 12) ⫽
.60, p⫽.45, p
2⫽.05. Both tops and bottoms
showed little change from before to after their
scenes in the use of the word “we” to describe
their relationship with their scene partner.
Self-other overlap. For the IOS, there was
a significant main effect of time (baseline vs.
postscene), F(1, 12) ⫽6.08, p⫽.03, p
2⫽.34,
no main effect of role, F(1, 12) ⫽.61, p⫽.45,
p
2⫽.05, and no Time ⫻Role interaction, F(1,
12) ⫽1.88, p⫽.20, p
2⫽.14. Both tops and
bottoms indicated greater self-other overlap
from baseline to postscene.
Discussion
Researchers studying consensual sadomas-
ochism (e.g., Baumeister, 1988; Newmahr,
2010; Pitagora & Ophelian, in press) have the-
orized that individuals pursue BDSM activities,
in part, due to the pleasant altered states of
consciousness these activities produce. How-
ever, to date, no research has tested whether
BDSM activities actually facilitate altered
states. To this end, we randomly assigned ex-
perienced BDSM practitioners to the top role or
to the bottom role for BDSM scenes and exam-
ined the impact of this assignment on measures
that reflect altered states. We hypothesized that
topping would facilitate an altered state aligned
with Csikszentmihalyi’s (1991) flow (measured
with the Flow State Scale), and bottoming
would facilitate an altered state aligned with
Dietrich’s (2003) transient hypofrontality (mea-
sured with a Stroop test).
Consistent with predictions, tops reported
high levels of flow. Unexpectedly, bottoms also
reported high levels of flow. In fact, contrary to
predictions, tops and bottoms reported nearly
identical levels of flow across the full scale.
However, an examination of the N/A responses
revealed that some bottoms (but no tops) indi-
cated that flow subdimensions associated with
agency (challenge-skill balance, clear goals, im-
mediate feedback) were not applicable to their
experience. Likewise, a comparison of the op-
timal performance and autotelic absorption fac-
tors of flow showed consistent flow across both
factors in tops but significantly lower levels of
optimal performance than autotelic absorption
in bottoms. These results suggest that topping
facilitates flow, whereas bottoming facilitates
pleasure and engagement but without the opti-
mal intentional performance characteristic of
flow.
A limitation of the present study is that flow
was measured only once after the scene was
over. Future studies that measured flow before
and after a scene would be better able to deter-
mine whether the scene activities facilitated
flow. In addition, the endorsement of N/A re-
sponses by bottoms on some flow items sug-
gests that the Flow State Scale might not be the
best scale to use to capture the relevant experi-
ences of bottoms. A final concern with the Flow
State Scale is that its questions are overt in what
they are asking. It is plausible that participants
10
15
20
25
30
35
40
45
50
Baseline Pre-scene Immediate
post-scene
Post-scene
Negative affect
Top
Bottom
Figure 3. Negative affect scores for tops and bottoms.
12 AMBLER ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
could have inferred what the scale is measuring
and could have provided answers corresponding
to what they believed the researchers were seek-
ing. Validation with an alternative, covert mea-
sure of flow would help address this concern.
The Stroop test seems less susceptible to this
type of strategic responding, as intentionally
manipulating a Stroop test would require artifi-
cially delaying either the incongruent trials (if
one wanted a higher Stroop score) or the control
trials (if one wanted a lower Stroop score).
Also consistent with predictions, bottoms,
but not tops, displayed short term reductions in
executive functioning consistent with transient
hypofrontality. These results are consistent with
Baumeister’s (1988) theory that bottoming fa-
cilitates an escape from self, and they suggest a
possible mechanism for this escape. Supporting
this interpretation, bottoms’ reported high levels
of Loss of Self-Consciousness on the Flow State
Scale (M⫽4.54, SD ⫽0.62, on the 1 to 5
scale).
Of course, transient hypofrontality is not the
only mechanism that can reduce Stroop perfor-
mance. For example, ego depletion has been
shown to decrease Stroop performance (Inzlicht
& Schmeichel, 2012). We believe that the tran-
sient hypofrontalilty hypothesis has the advan-
tage of parsimony: The descriptions of subspace
collected from MTurk respondents include ele-
ments that seem much more closely aligned
with transient hypofrontality than with ego de-
pletion (“one with nature and my environment,”
“very dreamlike,” “an almost trance like state”).
Furthermore, if bottoming consisted primarily
of willful, effortful endurance (e.g., enduring
the pain and humiliation while suppressing the
desire to use a safeword to end the scene), we
would expect bottoms to report increases in
psychological stress and negative affect. In-
stead, bottoms reported reductions in stress and
negative affect and rated their experience as
highly autotelic (4.79 out of 5 on the autotelic
dimension of the Flow State Scale). Nonethe-
less, a goal of future research might be to fur-
ther explore the mechanisms of transient hypo-
frontality and ego depletion as they apply to
BDSM activity.
These results help to illuminate one of the
reasons why people choose to engage in BDSM
activities: for the pleasurable altered states of
consciousness that appear to be associated with
these activities. However, we are not claiming
that BDSM is unique in facilitating these altered
states. Prior research on flow has identified a
number of activities that can facilitate flow
(e.g., athletics, music, yoga). Likewise, prior
research on transient hypofrontality has identi-
fied a number of activities that can facilitate this
effect (e.g., running, meditation, certain drugs).
Our study is, to our knowledge, the first to
provide empirical evidence that BDSM also fa-
cilitates these altered states. We are not claim-
ing, however, that BDSM is the only activity
that does so.
Additional Effects of Scenes
Prior research (Sagarin et al., 2009) reported
a Role ⫻Time interaction for cortisol such that
bottoms’ cortisol increased significantly from
baseline to just after the scene ended, whereas
tops’ cortisol did not change significantly from
baseline to just after the scene ended. In the
present study, the corresponding interaction was
nonsignificant, but, nonetheless, showed a trend
toward significance (p⫽.08). Coupled with the
findings that psychological stress (measured
prescene to postscene) decreased for both tops
and bottoms, we suggest that BDSM activities,
despite their physical stress, can be psycholog-
ically stress relieving for practitioners. Negative
affect also decreased from prescene to post-
scene for tops and bottoms, although the de-
crease occurred prescene for tops and during the
scene for bottoms. Additionally, both tops and
bottoms reported increases in sexual arousal
during their scenes.
Taken as a whole, BDSM activities appear to
facilitate subjectively enjoyable altered states of
consciousness, reductions in psychological
stress and negative affect, and increases in sex-
ual arousal. Furthermore, although the broad
pattern applies to both bottoms and tops, there
are role differences in the nature of the altered
states and the timing of the changes in affect.
Sample Size
In this section, we consider the implications
of our study’s small sample size. Our Nof 14 is
clearly low and notably lower than prior studies
of BDSM scenes (e.g., Sagarin et al., 2009, had
a total Nof 58). However, in contrast to prior
studies in which participants chose their roles
and provided fairly nonintrusive measures, the
demands of this experiment (a willingness to be
13BDSM AND ALTERED STATES
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
randomly assigned to role; physiological and
cognitive measures administered just prior to
scene commencement and just after scene com-
pletion but prior to aftercare) set a higher bar for
participant involvement.
Below, we consider whether the small sample
size increases the risk of four types of errors:
Type I error (the probability of a false positive
result), Type II error (the probability of a false
negative result), Type S error (the probability
that a significant result is in the wrong direc-
tion), and Type M error (the probability that a
significant result exaggerates the magnitude of
the effect).
Type I error. The Type I error rate of a
statistical test is held at a desired level based on
the critical value chosen for statistical signifi-
cance (typically .05), regardless of the sample
size (a smaller sample size requires a larger
effect size to achieve significance). Because our
methodology and analyses do not include the
types of questionable research practices that
inflate the Type I error rate (e.g., post hoc
exclusion of experimental conditions, post hoc
inclusion of covariates in analyses, etc.; John,
Loewenstein, & Prelec, 2012; Simmons, Nel-
son, & Simonsohn, 2011), the Type I error rate
of our significance tests is held at 5%. We
should note, however, that the separate signifi-
cance tests we performed on each of our mea-
sures combined with our choice to not correct
for alpha inflation (due to the a priori nature of
our predictions) increases the experiment-wise
Type I error rate.
Type II error. The Type II error rate of a
statistical test increases as the sample size de-
creases. This is where our small sample size
clearly carried costs: A number of potentially
interesting findings were not significant despite
the presence of moderate effect sizes.
Type S and Type M errors. Gelman and
Carlin (2014) provides a statistical technique
for estimating two additional risks: That signif-
icant results will have the wrong sign (a Type S
error), and that significant results will overesti-
mate the magnitude of the effect size (a Type M
error). We used Gelman and Carlin’s (2014)
technique to estimate the Type S and Type M
error rates for one of our primary findings: the
Role (top vs. bottom) ⫻Time (baseline vs.
postscene) interaction for Stroop scores.
Gelman and Carlin’s calculations use the stan-
dard error observed in the present study, but
they require an estimate of the population effect
size that is not taken from the present study (this
prevents a biased effect size estimate in the
present study from biasing the calculation of the
Type S and Type M errors).
Based on Cohen’s (1988) description that a
large effect would be obvious to a superficial
glance, we estimated a population effect size
corresponding to Cohen’s rule of thumb value
for large effects: d⫽0.8. The analyses pro-
duced an estimated Type S error rate of 0.006
and an estimated Type M error rate of 1.91.
From this, we can be fairly confident (99.4%
probability) that our significant Stroop effect is
in the correct direction, but our estimated effect
size probably overestimates the actual popula-
tion effect size by approximately a factor of 2.
Implications and Future Directions
As noted above, ego depletion represents a
plausible alternative explanation for the decre-
ments in Stroop performance in bottoms. In
addition, although the present data suggest that
consensual BDSM activities can exert positive
psychological effects (e.g., reductions in nega-
tive affect and self-reported stress), the present
data cannot speak to the mechanisms that
caused these effects. It is possible that these
effects stemmed directly from the BDSM activ-
ities, but it is also possible that they stemmed
indirectly via the increases in sexual arousal or
one of the other measured or unmeasured ef-
fects of the BDSM scenes. Future studies that
compared the effects of BDSM to the effects of
other types of sexual/intimate activities (e.g.,
non-BSDM sexual behaviors, massage) would
be of value in determining which effects (if any)
are unique to BDSM.
Because the postscene survey was adminis-
tered after participants engaged in aftercare, the
“right now” questions on the postscene survey
(the PANAS, self-reported stress, self-reported
sexual arousal, “we”ness, and the IOS) and the
40-min postscene hormonal measures were
likely impacted by aftercare. Future studies
could disentangle the possible impact of after-
care by administering these measures before
and after aftercare. In addition, it is not clear
whether certain questions were interpreted as
average across the scene or the peak during the
scene. For example, it is not clear whether scene
intensity represents the average intensity across
14 AMBLER ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
the scene or the peak intensity of the scene.
Further, the effects observed in this study are
almost certainly moderated by a variety of in-
dividual difference factors including a person’s
response to BDSM activities in general and with
a particular partner. If we randomly selected a
couple from the general population and asked
them to engage in BDSM behavior, for exam-
ple, their results would likely differ from the
results of self-selected individuals.
The observed scenes varied greatly in terms
of the activities that took place, the presence or
absence of sexual activity, the presence or ab-
sence of orgasm, the length of the scenes, the
decision whether to start setting up the scene
during the 30-min waiting period, the relation-
ship between the scene partners, and the gen-
ders, ages, sexual orientations, and other demo-
graphics of the participants. Unfortunately, the
present sample size precludes an examination of
whether certain scene elements (e.g., pain,
bondage, sexual activity, orgasm, length of
scene, timing of activities), relationship ele-
ments (e.g., level of experience, relationship
quality, relationship novelty, attraction, trust be-
tween scene partners), or gender or other demo-
graphic pairings are associated with observed
effects. Future research that tested these poten-
tial mediating processes and moderating vari-
ables and that measured additional personality
and psychopathology characteristics (particu-
larly with larger samples) would be of value. It
should be noted, however, that the variation in
scene activities and timing does not represent a
confound, as each scene provided both a top and
a bottom for analysis (and as a result, the vari-
ation in scene activities and timing were equally
represented in both levels of the Role indepen-
dent variable).
In addition, it would be of benefit for future
studies to ask participants directly whether they
had entered subspace or topspace, to administer
related measures such as a state scale of disso-
ciation, and to examine whether these subjec-
tive reports align with other measures of these
altered states. Likewise, it would be of benefit
for future studies to measure the effects of pri-
vate BDSM activity. Although all participants
in the current study consented to observation
and reported that the observation had little im-
pact on their scenes, it is possible that the ob-
servation impacted participants in a manner that
they did not perceive. Finally, future studies
could profitably examine whether the present
results replicate in individuals who show a
strong preference for the top or bottom role. It is
possible that individuals who exclusively take
on the top role or the bottom role might more
easily enter role-specific altered states of con-
sciousness. Alternatively, experience in role,
rather than exclusivity of role, might facilitate
altered states.
Conclusion
The present study provides preliminary evi-
dence that consensual BDSM activities facili-
tate role-specific altered states of consciousness,
with topping aligned with Csikszentmihalyi’s
(1991) flow and bottoming aligned with Di-
etrich’s (2003) transient hypofrontality as well
as some facets of flow. The results contribute to
a growing body of evidence that individuals
pursue BDSM for nonpathological reasons in-
cluding the pleasant altered states of conscious-
ness these activities are theorized to produce.
References
Alison, L., Santtila, P., Sandnabba, N. K., & Nor-
dling, N. (2001). Sadomasochistically oriented be-
havior: Diversity in practice and meaning. Ar-
chives of Sexual Behavior, 30, 1–12. http://dx.doi
.org/10.1023/A:1026438422383
Aron, A., Aron, E. N., & Smollan, D. (1992). Inclu-
sion of other in the Self Scale and the structure of
interpersonal closeness. Journal of Personality
and Social Psychology, 63, 596 – 612. http://dx.doi
.org/10.1037/0022-3514.63.4.596
Baumeister, R. F. (1988). Masochism as escape from
self. Journal of Sex Research, 25, 28 –59. http://dx
.doi.org/10.1080/00224498809551444
Brose, A., Voelkle, M. C., Lövdén, M., Linden-
berger, U., & Schmiedek, F. (2015). Differences in
the between-person and within-person structures
of affect are a matter of degree. European Journal
of Personality, 29, 55–71. http://dx.doi.org/10
.1002/per.1961
Buhrmester, M., Kwang, T., & Gosling, S. D. (2011).
Amazon’s Mechanical Turk: A new source of in-
expensive, yet high-quality, data? Perspectives on
Psychological Science, 6, 3–5. http://dx.doi.org/10
.1177/1745691610393980
Cabeza, R., & Nyberg, L. (2000). Imaging cognition
II: An empirical review of 275 PET and fMRI
studies. Journal of Cognitive Neuroscience, 12,
1– 47.
Chandler, J., Paolacci, G., Peer, E., Mueller, P., &
Ratliff, K. A. (2015). Using nonnaive participants
15BDSM AND ALTERED STATES
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
can reduce effect sizes. Psychological Science, 26,
1131–1139. http://dx.doi.org/10.1177/0956797
615585115
Cohen, J. (1988). Statistical power analysis for the
behavioral sciences (2nd ed.). Mahwah, NJ: Erl-
baum.
Connolly, P. H. (2006). Psychological functioning of
bondage/domination/sado-masochism (BDSM)
practitioners. Journal of Psychology & Human
Sexuality, 18, 79 –120. http://dx.doi.org/10.1300/
J056v18n01_05
Csikszentmihalyi, M. (1991). Flow: The psychology
of optimal experience. New York, NY: Harper
Collins.
Del Giorno, J. M., Hall, E. E., O’Leary, K. C., Bixby,
W. R., & Miller, P. C. (2010). Cognitive function
during acute exercise: A test of the transient hy-
pofrontality theory. Journal of Sport & Exercise
Psychology, 32, 312–323.
Dietrich, A. (2003). Functional neuroanatomy of al-
tered states of consciousness: The transient hypo-
frontality hypothesis. Consciousness and Cogni-
tion, 12, 231–256. http://dx.doi.org/10.1016/
S1053-8100(02)00046-6
Dietrich, A., & Sparling, P. B. (2004). Endurance
exercise selectively impairs prefrontal-dependent
cognition. Brain and Cognition, 55, 516 –524.
http://dx.doi.org/10.1016/j.bandc.2004.03.002
Faccio, E., Casini, C., & Cipolletta, S. (2014). For-
bidden games: The construction of sexuality and
sexual pleasure by BDSM “players.” Culture,
Health & Sexuality, 16, 752–764. http://dx.doi.org/
10.1080/13691058.2014.909531
Freud, S. (1938). The basic writings of Sigmund
Freud (A. A. Brill, Trans. & Ed.). New York, NY:
Modern Library.
Gelman, A., & Carlin, J. (2014). Beyond power cal-
culations: Assessing Type S (sign) and Type M
(magnitude) errors. Perspectives on Psychological
Science, 9, 641– 651. http://dx.doi.org/10.1177/
1745691614551642
Graham, B. C., Butler, S. E., McGraw, R., Cannes,
S. M., & Smith, J. (2015). Member perspectives on
the role of BDSM communities. Journal of Sex
Research. Advance online publication. http://dx
.doi.org/10.1080/00224499.2015.1067758
Hébert, A., & Weaver, A. (2014). An examination of
personality characteristics associated with BDSM
orientations. The Canadian Journal of Human Sex-
uality, 23, 106 –115.
Hébert, A., & Weaver, A. (2015). Perks, problems,
and the people who play: A qualitative exploration
of dominant and submissive BDSM roles. The
Canadian Journal of Human Sexuality, 24, 49 – 62.
http://dx.doi.org/10.3138/cjhs.2467
Inzlicht, M., & Schmeichel, B. J. (2012). What is ego
depletion? Toward a mechanistic revision of the
resource model of self-control. Perspectives on
Psychological Science, 7, 450 – 463. http://dx.doi
.org/10.1177/1745691612454134
Jackson, S., Eklund, B., & Martin, A. (2010). The
FLOW manual: The manual for the Flow Scales.
Queensland, Australia: Mind Garden, Inc.
Jackson, S. A., & Eklund, R. C. (2002). Assessing
flow in physical activity: The Flow State Scale-2
and Dispositional Flow Scale-2. Journal of Sport
& Exercise Psychology, 24, 133–150.
Jackson, S. A., & Marsh, H. W. (1996). Development
and validation of a scale to measure optimal expe-
rience: The Flow State Scale. Journal of Sport &
Exercise Psychology, 18, 17–35.
Jackson, S. A., Martin, A. J., & Eklund, R. C. (2008).
Long and short measures of flow: The construct
validity of the FSS-2, DFS-2, and new brief coun-
terparts. Journal of Sport & Exercise Psychology,
30, 561–587.
John, L. K., Loewenstein, G., & Prelec, D. (2012).
Measuring the prevalence of questionable research
practices with incentives for truth telling. Psycho-
logical Science, 23, 524 –532. http://dx.doi.org/10
.1177/0956797611430953
Jozifkova, E., & Flegr, J. (2006). Dominance, submissivity
(and homosexuality) in general population: Testing of
evolutionary hypothesis of sadomasochism by Internet-
trap-method. Neuroendocrinology Letters, 27, 711–718.
Kennedy-Moore, E., Greenberg, M. A., Newman,
M. G., & Stone, A. A. (1992). The relationship
between daily events and mood: The mood mea-
sure may matter. Motivation and Emotion, 16,
143–155. http://dx.doi.org/10.1007/BF00995516
Klement, K. R., Lee, E. M., Ambler, J. K., Hanson,
S. A., Comber, E., Wietting, D.,...Sagarin, B. J.
(2015). Extreme rituals in a BDSM context: The
physiological and psychological effects of the
“Dance of Souls.” Manuscript submitted for pub-
lication.
Lee, E. M., Klement, K. R., Ambler, J. K., Loewald,
T., Comber, E., Hanson, S. A.,...Sagarin, B. J.
(2016). Altered states of consciousness during an
extreme ritual. PLoS ONE, 11(5), e0153126. http://
dx.doi.org/10.1371/journal.pone.0153126
Linnman, C., Carlbring, P., Ahman, A., Andersson,
H., & Andersson, G. (2006). The Stroop effect on
the internet. Computers in Human Behavior, 22,
448 – 455. http://dx.doi.org/10.1016/j.chb.2004.09
.010
MacLeod, C. M. (1991). Half a century of research
on the Stroop effect: An integrative review. Psy-
chological Bulletin, 109, 163–203. http://dx.doi
.org/10.1037/0033-2909.109.2.163
Moser, C., & Levitt, E. E. (1987). An exploratory-
descriptive study of a sadomasochistically oriented
sample. Journal of Sex Research, 23, 322–337.
http://dx.doi.org/10.1080/00224498709551370
Newmahr, S. (2010). Rethinking kink: Sadomasoch-
ism as serious leisure. Qualitative Sociology, 33,
16 AMBLER ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
313–331. http://dx.doi.org/10.1007/s11133-010-
9158-9
Ortmann, D. M., & Sprott, R. A. (2013). Sexual
outsiders: Understanding BDSM sexualities and
communities. Lanham, MD: Rowman & Littlefield
Publishers.
Pascoal, P. M., Cardoso, D., & Henriques, R. (2015).
Sexual satisfaction and distress in sexual function-
ing in a sample of the BDSM community: A com-
parison study between BDSM and non-BDSM
contexts. Journal of Sexual Medicine, 12, 1052–
1061. http://dx.doi.org/10.1111/jsm.12835
Pitagora, D. (2013). Consent vs. coercion: BDSM
interactions highlight a fine but immutable line.
The New School Psychology Bulletin, 10, 27–36.
Pitagora, D., & Ophelian, A. (in press). Therapeutic
benefits of subspace. Electronic Journal of Human
Sexuality.
Richters, J., de Visser, R. O., Rissel, C. E., Grulich,
A. E., & Smith, A. M. A. (2008). Demographic
and psychosocial features of participants in bond-
age and discipline, “sadomasochism” or domi-
nance and submission (BDSM): Data from a Na-
tional Survey. Journal of Sexual Medicine, 5,
1660 –1668. http://dx.doi.org/10.1111/j.1743-6109
.2008.00795.x
Sagarin, B. J., Cutler, B., Cutler, N., Lawler-Sagarin,
K. A., & Matuszewich, L. (2009). Hormonal
changes and couple bonding in consensual sado-
masochistic activity. Archives of Sexual Behavior,
38, 186 –200. http://dx.doi.org/10.1007/s10508-
008-9374-5
Simmons, J. P., Nelson, L. D., & Simonsohn, U.
(2011). False-positive psychology: Undisclosed
flexibility in data collection and analysis allows
presenting anything as significant. Psychological
Science, 22, 1359 –1366. http://dx.doi.org/10
.1177/0956797611417632
Skowronski, J. J., Walker, W. R., Henderson, D. X.,
& Bond, G. D. (2014). The fading affect bias: Its
history, its implications, and its future. In J. M.
Olson & M. P. Zanna (Eds.), Advances in experi-
mental social psychology (Vol. 49, pp. 163–218).
New York, NY: Academic Press/Elsevier. http://
dx.doi.org/10.1016/B978-0-12-800052-6.00003-2
Watson, D., Clark, L. A., & Tellegen, A. (1988).
Development and validation of brief measures of
positive and negative affect: The PANAS scales.
Journal of Personality and Social Psychology, 54,
1063–1070. http://dx.doi.org/10.1037/0022-3514
.54.6.1063
Wiseman, J. (1998). SM 101: A realistic introduction
(2nd ed.). Gardena, CA: Greenery Press.
Wismeijer, A. A. J., & van Assen, M. A. L. M.
(2013). Psychological characteristics of BDSM
practitioners. Journal of Sexual Medicine, 10,
1943–1952. http://dx.doi.org/10.1111/jsm.12192
Received July 30, 2015
Revision received April 19, 2016
Accepted June 12, 2016 䡲
17BDSM AND ALTERED STATES
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
A preview of this full-text is provided by American Psychological Association.
Content available from Psychology of Consciousness: Theory, Research, and Practice
This content is subject to copyright. Terms and conditions apply.