Sexual Presence: A qEEG Analysis of Sexual Arousal to Synthetic Pornography
Patrice Renaud 1 2, Sarah Michelle Neveu 1 2 3, Joanne-L. Rouleau 4 & Christian Joyal 5
1 Université du Québec en Outaouais, St-Jérôme, Québec, Canada
2 Institut National de Psychiatrie Légale Philippe-Pinel, Montréal, Québec, Canada
3 Université du Québec à Montréal, Montréal, Québec, Canada
4 Université de Montréal, Montréal, Québec, Canada
5 Université du Québec àTrois-Rivières, Trois-Rivières, Québec, Canada
Abstract:
The concept of sexual presence (SP) is explored experimentally by comparing how synthetic and non-synthetic pornography elicit psychophysiological responses in male heterosexual participants. Using quantitative electroencephalography (qEEG), penile plethysmography (PPG) and oculometry, it was shown that distinct psychophysiological profiles are associated with SP gradients. Furthermore, SP is positively correlated with genital and cerebral sexual arousal. Some EEG counter-intuitive results might be interpreted as a form of uncanny valley effect.
Keywords:
Sexual presence, synthetic pornography, uncanny valley, penile plethysmography, oculometry, quantitative electroencephalography
Full Article:
Sexual Presence: A qEEG Analysis of Sexual Arousal to Synthetic Pornography
As anticipated by Lombard and Jones (2004, 2013), understanding the role of presence in pornography will grow in importance in the years to come. According to Market Watch: “By 2025, such adult content is forecast to be a $1 billion business, the third-biggest virtual-reality sector, after videogames ($1.4 billion) and NFL-related content ($1.23 billion’’ (Booton, 2015). Renaud, Fontanessi and Benbouriche (2014) suggested the following definition of sexual presence (SP), “SP as a “second order” technologically mediated experience, is defined as a psychophysiological state of sexual arousal, including a subjective erotic perception, whose content and extent are determined by the interplay between individual psychobiological predispositions, idiosyncratic past experiences, and what is sexually afforded by a mediating technology”.
In a forensic context, it has been demonstrated that virtual agents can be used to prompt sexual and perceptual motor responses distinguishing between teleiophiles (individuals sexually attracted to adults) and pedophiles (Fromberger et al. 2012; Renaud, Rouleau, Granger, Barsetti, & Bouchard, 2002; Renaud et al., 2011; Trottier et al., 2014). Furthermore, Fromberger et al. (2015) recently reported results showing that high-immersive virtual environments tend to enhance realism ratings as well as ratings of sexual attractiveness towards three-dimensional human sexual stimuli.
Moreover, anterior alpha asymmetry in the frontal area of the brain as measured with EEG has been associated with approach behavior and motivation towards appetitive stimuli (Gable & Harmon-Jones, 2008; Wacker, Chavanon, Leue, & Stemmler, 2008). Prause and her colleagues (2014) have shown how a preponderance of left-hemisphere activation over the right one is an index of cerebral sexual arousal in men and women. These authors compared neutral and sexually motivating erotic films and found a positive relationship between self-reported mental sexual arousal and alpha asymmetry in their participants.
The research in neurorobotics coming from the Saygin Lab tends to support the contention that stimuli depicting human-like appearances prompt specific brain activations (Urgen et al., 2013). Saygin’s team has shown to their participants a robot (i.e. without human-like appearance), an android (i.e. a robot with human appearance) and an actual human being while measuring theta EEG power (4-8 Hz) in the frontal cortex (F3 and F4). A stronger theta power observed in the robot condition was interpreted as reflecting reciprocal exchanges between perceptual processing and long-term memory, especially during retrieval of items from semantic memory. To sum up, accessing and processing information about the robot’s appearance was more demanding than when it was done for the human-like agents. In a similar experiment relying on reaction-time measures, the same team arrived to the conclusion that the robot agent captured attention differently when compared to its human-like equivalents (Li, lorendo, Miller, Ishiguro, & Saygin, 2015). Moreover, the Saygin Lab puts forward the idea that these results might also help to explain the uncanny valley effect (Mori, 1970/2012). The uncanny valley refers to the fact that the observation of a human-like agent, depending on its degree of physical similarity to human appearance, can evoke subjective responses in a nonlinear fashion, with more realism not necessarily equating with more positive feelings or sense of presence (see Ho & McDorman, 2017; MacDorman, Green, Ho, & Koch, 2009). The eerie feeling associated with the uncanny valley effect is known to hinder empathy towards synthetic characters (McDorman, Srinivas, & Patel, 2013). Empathy is a complex response that usually refers to the ability to feel what another person subjectively feels (Yang, Decety, Lee, Chen, & Cheng, 2009). This ability presumably stems from a proper theory of mind, that is to say a specific reading and valid emotional responses towards what is exhibited by others. Following the four-factor model, empathy comprises four distinct functions: (1) perspective taking, i.e the tendency to adopt the subjective point of view of others; (2) cognitive fantasizing, i.e. the proneness to immerse ourselves into fictional situations; (3) empathic concern, i.e. the inclination to feel compassion or sympathy for others; and (4) personal distress, i.e. the bias toward feeling anxiety or fear in response to others’ distress (Davis, 1983; Chrysikou and Thompson, 2016; Joyal, Neveu, Boukhalfi, Jackson, & Renaud, 2018). Furthermore, it is known that the suppression of the alpha rhythm recorded at the sensorimotor cortex (also known as the mu rhythm; 8-13Hz) is associated with empathy and that the absence of this suppression was observed among clinical populations showing a lack of empathy (Jackson & Decety, 2004; Joyal et al., 2018; Oberman et al., 2005; Woodruff, Martin, & Bilyk, 2011; Yang et al., 2009). Urgen and colleagues (2013) report a significant mu suppression effect towards robots and human-like agents. More generally speaking, spatial presence, as defined loosely as “the sense of being in an environment” (Steuer, 1995), and as the background into which social presence, and in particular SP, can happen, has been assimilated to brain activity in the parietal lobe and the dorsolateral prefrontal cortex, both in alpha and theta rythms (Baumgartner et al., 2006, 2008; Kober, Kurzmann & Keuper, 2012). Through alpha suppression or desynchronization, the parietal lobe plays a central role in spatial processing, visuo-motor control as well as in experiencing the simulated context from an egocentric stance (Kober, Kurzmann & Keuper, 2012; Planck, Müller, Onton, Makeig, & Gramann, 2010). Activation in the dorsolateral prefrontal cortex is more readily associated to an allocentric experience and thus to a lesser feeling of spatial presence.
Goals and Hypotheses
The aim of the study is to better understand the psychophysiology of SP in heterosexual males by comparing how different levels of subjective SP induced by real (photo-realistic) and synthetic pornography translate into psychophysiological states. The study relies on a triangulation strategy in using simultaneously converging measures of sexual arousal and interest, namely sexual plethysmography, oculometry and electroencephalography.
We propose the following hypotheses with respect to the psychophysiological responses coming with the subjective experience of sexual presence (SP):
H1. Levels of SP will come with distinct brain activation patterns, both in terms of cerebral regions and electrical oscillations recorded using EEG.
H2. Levels of SP will come with distinct perceptual motor patterns related to sexual interests.
H3: SP will be positively correlated with sexual arousal.
H4. SP evoked by sexual stimuli will be modulated by perceptual motor patterns involved in the extracting of visual cues.
Method
Participants
Forty-one (41) male adults (mean age of 34.3 ± 8.7, 21 to 54 year-old) were recruited among the general population (online ads). Twenty-five of these were single. They were all right-handed heterosexuals. They reported not having neurological or psychological problems. Twenty (20) were Caucasian, 7 were African-Canadian, 6 were Arab-Canadian, one was Latin-Canadian (from South-America) and one was Indian-Canadian (from India). As for pornography using habits, one never used it, 11 participants viewed it less than once a week, 12 viewed it once or twice a week and 11, three or more times a week. Six (6) participants were excluded from the analyses because of no sexual responses. Pornography using habits data were used in correlation analyses.
Materials
Pornographic videos. For the study we selected a 165 sec soundless excerpt from a commercial heterosexual pornographic movie with two Caucasian protagonists in their twenties. This excerpt contained scenes of oral sex and vaginal penetrations. To create synthetic equivalents to this excerpt, we used 3DXChat from the adult Web Site Sex game devil (http://sexgamedevil.com/). This application allows the customization of pornographic 3D animations, choosing from a library of characters, sexual attributes as well sexual behaviors. The characters generated by this application are emotionless at the facial level. Using Adobe special effects, we then created a cartoonish equivalent the Sex game devil animation, i.e. a black and white less defined version of the latter. Hence we had three videos, the photo-realistic one, the virtual and the cartoonish (see Figure 1). A photo-realistic and non-sexual humoristic video clip was presented at the outset to establish a baseline line level.
Penile Plethysmography. Penile plethysmography (PPG) measures variations of blood volume in the penis and is used particularly to assess sexual arousal (Freund, Sedlacek, & Knob, 1965; Laws & Marshall, 2003); PPG is associated to hypothesis H3. Circumferential PPG requires the wearing of a thin mercury-in-rubber strain gauge around the shaft of the penis. This gauge is a small rubber tube filled with mercury forming a ring. During an erectile response, the gauge stretches and changes in the mercury column produces variations in electric conductibility (Limestone Technologies, DataPac USMTM 16-bit data polygraph acquisition instrument). The erectile response value obtained in each condition is expressed in mm of stretching and corresponds to the difference between the highest value recorded during the condition and the baseline value measured from the start of the same condition. In clinical sexual plethysmography, a penile response of more than 3.00 mm is deemed significant, that is departing from a state of flaccidity.
Quantitative EEG (qEEG). Quantitative electroencephalography (qEEG) gives access to the spectrum of wavelengths characterizing specific responses distributed over the scalp as they are unfolding in time (Tong & Thakor, 2009). Contrary to the evoked potential method, qEEG is more adapted to complex and longer scenarios such as those used in video games and virtual reality applications.
A 64 electrodes EEG cap (Acticap) was used with a Brain Amp amplifier (all from Brain products ©, Germany), installed following the 10-20 convention. Recorder and Analyzer2 software were used to record and analyze the EEG data, respectively (Brain products). Artefacts and eye-movements were manually removed from the EEG digital tracing by two trained specialists following inter-rater agreement. The sample rate was of 500 Hz and the impedance was kept under 25 Ω (in accordance with the Acticap system specifications). Electrodes were referenced at Fpz and the recordings were re-referenced with the means of left and right mastoids. Fast Fourier Transforms were applied to the data to obtain power (mV2) of frequencies (1 to 30 Hz) in each 1 s epoch (Oberman et al., 2005). The mean frequency for each condition was then analyzed. More specifically, the power of the Alpha (8-13 Hz) and the Theta (4-8 Hz) bands was isolated from electrodes Oz (occipital area), P7P3PzP4P8 (parietal area), C3CzC4T7T8 (sensorimotor area), and FP1FP2F7F3F4F8 (frontal area). In order to control for inter-individual differences in power emission, the ratio of the power during the resting state/experimental condition was computed and transformed into their natural logarithm (to normalize data distribution (Oberman et al., 2005). This computation gives an index of alpha activity suppression, the latter being associated with a concentration of activity in the brain area being probed. Frontal asymmetry scores were computed by subtracting the log-transformed alpha power of the right frontal hemisphere area (F8F4FP2) from the corresponding log-transformed alpha power of the left frontal hemisphere (F7F3FP1) area. A higher score on this index indicates relatively greater alpha power in the left hemisphere; this index sexual arousal is associated to hypothesis H1.
Eye-tracking. For oculomotor measurements, we used a binocular dark pupil tracking Tobii TX-300 model (23 in. monitor) with a gaze sampling frequency of 300 Hz, a gaze sampling variability of less than 0.3%, an accuracy of 0.4° and a precision of 0.14°. The editing of areas of interest (AOI) was used to track eye movements relative to specific portions of the animated stimuli. In order to extract the number of fixations as well as the mean fixation duration happening therein, the face and the genital areas of the female and male characters were tracked that way in each sexual video; these two indexes are related to hypothesis H2 and H4 .
Sexual presence: self-report measure. Ten questions were asked after each condition to assess subjective sexual presence (SSP) in a self-reported fashion. Three questions were about the level of realism experienced and seven others about the level of involvement in the depicted sexual situations; each question had to be answered using a seven-point Likert scale: 1=Not at all, 4=Moderately and 7=Completely. Level of realism: To what degree did you feel a sexual arousal similar as one felt in a real sexual situation? To what extent the visual quality of the sexual scenes appeared realistic to you? To what degree the graphic quality of the sexual scenes hindered your sexual response? Level of involvement: To what extent did you have the feeling of witnessing live sexual intercourses taking place in front of you? To what degree were you under the impression of being beside individuals having a sexual intercourse? To what extent did you have the feeling of being one of the individuals having a sexual intercourse? To what degree did you identify yourself with one of the individuals having a sexual intercourse? To what extent did you feel what were sexually feeling the characters? To what degree were you under the impression of actually taking part of the sexual action? To what extent did you have the impression of being outside the sexual action taking place in the scene? Total SSP is the sum of both sub-scales. Questions appeared on the screen sequentially and were answered using a keyboard. SSP scales are related to hypotheses H1, H2, H3 and H4.
Protocol
After arriving at the Virtual Reality Applications in Forensic Psychiatry laboratory at the Philippe-Pinel Institute of Montreal, participants were briefed on the purpose of the study and received a detailed description of the assessment process and methods used to record their physiological reactions. They were administered the Immersive tendency questionnaire (Witmer & Singer, 1998) and were asked questions about their sexual life and habits. They were then set up with an EEG cap and left alone. The experimenter moved to an adjacent room to monitor the progress of the experiment. The two rooms were linked by an intercom system allowing two-way communications throughout the procedure.
Participants were then asked to put on a penile gauge according to standard practice in penile plethysmography, i.e. around the middle part of the penis shaft. A 2-min resting state period, one-minute eyes-closed and another eyes-opened, was then given to the participants to assess their EEG base level without stimulation. Subjects were then presented with a 9 points calibration screen which allowed the experimenter to adjust the infrared ocular tracking system for gaze-direction monitoring. Before each trial, a fixation cross appeared in the middle of the screen. When participant’s gaze was on the cross, the experimenter started the trial. A return to erectile baseline was also required before any stimuli were presented.
Participants were instructed to simply attend to the video freely and allow their physiological responses to occur naturally. Between each sexual video, participants were administered the sexual presence questionnaire. Sexual stimuli were presented in a counterbalanced order chosen randomly for each participant. A baseline measure for EEG and PPG was taken at the outset with the presentation of a nonsexual humoristic video.
Before leaving the laboratory, participants were debriefed. The entire procedure lasted approximately 60 minutes and at the end the participants received a monetary compensation. This experiment was conducted in conformity with the ethical guidelines of the Philippe-Pinel Institute of Montreal.
Statistical Analyses
Four sets of statistical analyses were completed in this study. First, a repeated-measures MANOVA was done from the self-reported measure of SP (SSP) to make sure that we had a significant SP gradient across our three sexual videos conditions (Cartoon, Virtual, Photo-Realistic). Second, PPG and qEEG were analyzed using a MANCOVA with the resting state 8-13Hz occipital response as the covariate. Four (4) SP conditions were tested in this analysis, i.e. the non-sexual baseline level (the humoristic video) and the three sexual videos. As a third set of analyses, we performed a repeated-measures MANOVA with the oculomotor measurements (number of fixations and mean fixation duration in AOIs), using a within-subjects two-factors model based on SP levels (3; Cartoon, Virtual, Photo-Realistic) and Sex (the characters’ sex; i.e. female or male). Finally, a series of bivariate correlations between the dependent variables were computed.
Results
Self-reported sexual presence (SSP)
Self-reported sexual presence (SSP) is significantly different following the sexual videos presented (F(4, 31) = 17.12, p < .001, Λ = 0.312, partial η2 = 0.688). For the total score of SSP, according to pairwise comparisons, the Photo-Realistic video brings about a significantly higher SSP than the Virtual and the Cartoon’s, and the Virtual video leads to a higher SSP when compared to the Cartoon video (Cartoon, M: 20.5, SD: 7.6; Virtual, M: 24.9, SD: 11.8; Photo-Realistic, M: 36.9, SD: 12.2: F(1, 34) = 32.40, p < .001; F(1, 34) = 57.06, p < .001; F(1, 34) = 5.72, p < .05).We thus have a significant SSP gradient across the three SP conditions (see Figure 1d). Furthermore, we get the same pattern with the level of realism subscale (see Table 1) while the level of involvement subscale comes out differently with a statistically non-significant difference between Cartoon and Virtual videos (see Table 1).
Penile plethysmography (PPG) and quantitative electroencephalography (qEEG)
For the second set of analyses, the MANCOVA also yields significant effects for SP conditions in PPG and qEEG (F(15, 15) = 27.25, p < .001, Λ = 0.035, partial η2 = 0.965). More specifically, all sexual stimuli prompt a significant sexual arousal response; this sexual arousal is statistically different from the one obtained from the baseline video (F(1, 34) = 32.40, p < .001; F(1, 34) = 57.06, p < .001; F(1, 34) = 5.72, p < .05; F(1, 34) = 5.72, p < .05). In turn, however, the three sexual videos do not differ significantly from each other in that respect, even though a clear tendency can be seen in Table XX and Figure 2.
As explained previously, anterior alpha asymmetry, i.e. more activity in the left hemisphere compared to the right one, has been suggested to index approach behavior and motivation, and especially cerebral sexual arousal (Gable & Harmon-Jones, 2008; Prause et al., 2014). As can be seen in Figure 3, this asymmetry index presents a profile similar to the one found with PPG (see Figure 2) while being more specific than the latter in significantly distinguishing between the Cartoon video and the Photo-Realistic one; the difference between the Virtual and the Photo-Realistic one is at the cusp of significance with a p of 0.053 (see Table 2). Otherwise, despite significant differences in SSP, and as with PPG, the Cartoon and the Virtual videos are not statistically different on that matter, prompting similar levels of asymmetry in frontal alpha.
QEEG results in alpha and theta oscillations are presented from the parietal to central and frontal areas of the brain. First, at the parietal level, we can observe in both brain rhythms that all conditions differ from each other, with the Photo-Realistic video, that is the condition with the highest level of SP, displaying the strongest suppression effect (see Table 2 and Figure 4), followed by the Cartoon video and the Virtual one. The parietal activity during the Virtual video stands out clearly with a much less important suppression, i.e. with more activity expressed in a less constrained fashion. All three sexual conditions differ significantly from the non-sexual condition, most probably reflecting attention processing and sensory integration differences.
As mentioned previously, alpha suppression in the sensorimotor cortex is especially linked to empathy responses. Results show, however, that the non-sexual, the Cartoon and the Photo-Realistic conditions do not differ significantly from one another in sensorimotor alpha suppression. Only the Virtual video again stands out significantly (see Table 2 and Figure 4). This condition brings about less alpha suppression than the other two sexual stimuli and the baseline video. Although no explicit hypotheses were made about it, theta activity in the sensorimotor cortex is reported for completeness. Here, the strongest suppression is found in the Cartoon condition followed by the Photo-Realistic one and the non-sexual. Again, the Virtual video differs markedly showing the largest synchronization response, compared to the others.
Theta power in the frontal area has been shown to increase in the processing of non-human visual appearances, especially in distinguishing humanoid from non-humanoid robots (Urgen et al., 2013). In our results, the power of the theta oscillations recorded in the frontal area is significantly higher with the Virtual video compared to all the other experimental conditions (see Table 2 and Figure 4). Theta power in the Cartoon condition is not statistically different from the one found during the viewing of the baseline video. Also, the Photo-Realistic video prompts a significantly higher theta power compared to the Cartoon and the non-sexual conditions. Results obtained in alpha are less specific than in theta, only the Virtual video condition differs significantly from all the others in that frequency.
Oculomotor measurements
As a final set of analyses, we now turn towards the oculomotor responses acquired during the viewing of the sexual stimuli according to their SP level and the sex of the characters portrayed. The MANOVA reveals significant effects of the SP factor (F(8, 27) = 9.14, p < .001, Λ = 0. 270, partial η2 = 0.730), the Sex factor (F(4, 31) = 63.24, p < .001, Λ = 0.270, η2 = 0.891), as well as the interaction between the latter (F(8, 27) = 6.06, p < .001, Λ = 0.358, η2 = 0.642). Table 3 and Figure 5 offer a synthetic view of these results. The Sex main effect is of interest: participants (all heterosexual males) focused more on the face and genital areas of the female character. They showed more and longer fixations in the genital area, as well as longer fixations in the face area of the female character.
The breaking down of the above-mentioned interaction effect gives us a clearer picture of how levels of SP might be linked to the perceptual motor extraction of information in critical areas of sexual stimuli (see Table 4). Two repeated measures ANOVA using SP level factor alone, one with the female character and the other with the male, give significant clues on that matter (F(8, 27) = 7.30, p < .001, Λ = 0.316, η2 = 0.684; (F(8, 27) = 9.08, p < .001, Λ = 0.271, η2 =0.729). A series of within-subject pairwise comparisons shows further how these variables relate (see Figure 5).
Levels of SP affect oculomotor responses to the female character in the following way: the number of fixations found in the face and female genitalia AOIs are significantly higher in the Photo-Realistic video condition compared to the Cartoon and Virtual’s, the latter two being statistically indistinct at this chapter. In regards to the mean fixation duration in these same AOIs, the Cartoon condition displays the longest duration, followed by the Virtual and the Photo-Realistic ones. These differences are statistically significant. Thus, as level of SP increases, it seems that the time spent scrutinizing critical cues of the sexual object decreases while the number of extraction point increases.
The picture with the male character is different and less straightforward. Contrary to what is happening with the female character, the number of fixations in the face AOI are significantly lower in the Photo-Realistic video condition compared to the Cartoon and Virtual’s, the latter two being statistically indistinct at this level. The number of fixations in the penis AOI is in turn significantly higher in the Photo-Realistic video condition compared to the Virtual and the Cartoon’s, this number in turn being significantly higher in the Virtual video condition than in its less defined version. Finally, the mean fixation duration in the male face AOI is significantly longer in the Cartoon video condition, followed by the Virtual and then the Photo-Realistic one. No significant differences are found for the mean fixation duration at the male genital level across SP level conditions.
Correlations
Finally, interesting correlations help us better understand these results. First, total SSP and its Realism scale correlate positively with penile erection as measured with PPG (r(32) = .341, p < .05; r(32) = .377, p < .05), which denotes that subjective erotic perception induced by technological means, and especially the impression of sexual realism, appears to be modulated by genital arousal. Cerebral sexual arousal also correlates positively with perceived sexual realism; this is so for frontal asymmetry in alpha and perceived realism (Realism scale of the SSP), respectively in the cartoon and the virtual video conditions (F8/F7: r(30) = .458, p < .01; F4/F3: r(30) = .450, p < .01) .
We also obtain strong positive correlations between fixation durations in the female AOIs (face and vulva) and alpha activity, respectively in parietal and central regions (face: r(30) = .446, p < .05; r(30) = .421, p < .05; vulva: r(30) = .495, p < .01; r(30) = .393, p < .05); the longer the perceptual motor activity dwells in these AOIs, the weaker the alpha suppression we get, i.e. a pattern of higher brain waves synchronization. Conversely, as the number of fixations in the male face AOI increases, the alpha suppression in parietal and central regions also increases (r(30) = .469, p < .01; r(30) = .361, p < .05; r(30) = .341, p < .05); as a reminder, alpha (or mu) suppression (or desynchronization) in the central cortex (sensorimotor area) is associated with empathy responses. Thus, the experience of SP can possibly be explained, at least in part, by the identification with the same sex character (remember that participants are heterosexual males) while disengaging at that level with the desired sexual object. This identification process could also require some kind of same-sex genital association as exemplified by this positive correlation between hemispheric asymmetry in the frontal part of the brain (FP2/FP1) and the time spent fixating the penis AOIs (r(30) = .395, p < .05), more asymmetry being correlated with a higher cerebral sexual arousal or more generally with approach behavior or motivation. As a corollary to these results, we have found positive correlations between SSP questionnaire items pertaining to identification with the characters and eye movements, in the Virtual video condition. Total fixation duration and fixation durations in the female and male face AOIs correlate with the responses to 1) To what degree did you identify yourself with one of the individuals having a sexual intercourse? and 2) To what extent did you feel what were sexually feeling the characters? ( (r(35) = .452, p < .01); (r(35) = .375, p < .05; (r(35) = .406, p < .05); (r(35) = .438, p < .01; (r(35) = .368, p < .05; (r(35) = .395, p < .05)).
Also, a strong and highly significant negative correlation between frontal theta activity (F3/F7) and alpha (or mu) suppression in the central region was found in the Virtual video condition (r(34) = .640, p< .001). As more perceptual and semantic memory processing is required (possibly because of the uncanniness of the stimuli), less cerebral activity related to empathy responses is recorded. In the same vein, as frontal theta activity (F4/F8) increases, frontal asymmetry in alpha diminish (r(34) = .420, p< .01). So, more processing effort seems to correspond to less cerebral sexual arousal. This result was found for the Virtual condition and we obtained a similar correlation with the Cartoon’s, but nothing significant in that matter came out with the Photo-Realistic one. To be noted, because we speak of correlations here, these results can also be understood the other way around, namely, more sexual cerebral arousal comes with a lesser cerebral response akin to the one possibly associated to the feeling of uncanniness. This last point will be discussed in the next section.
Finally, individual differences and past sexual experiences might also play a role in the experience of SP as shown by two other positive correlations, one between the total score at the immersive tendency questionnaire (Witmer & Singer, 1998) and the time spent scrutinizing the genital area of the female character (r(33) = .356, p < .05), and the other between pornography using habits (weekly number of viewings) and the focusing on female genitalia, i.e. the number of fixations in the corresponding AOI (r(33) = .414, p < .05).
Discussion
Our first hypothesis appears to be partially corroborated in that we are obtaining distinct EEG profiles following SP gradients. Without any surprise the low-level visual processing of the erotic material comes with specific brain activation in the parietal region. Since the three SP conditions are not composed of the same colors, textures and layout, this was expected. The Virtual condition, however, stands out markedly on that matter. The same peculiarity is observed for the Virtual condition when empathy and perceptual processing of human appearance are considered (mu suppression in the sensorimotor and theta activity in the frontal area). This may well be attributable to the uncanniness of this condition. Indeed, closer to reality than the cartoon’s because of its better image definition and colors, the Virtual condition is possibly dwelling in the famous Mori’s valley (Mori, 1970, 2012). In support of this interpretation, the extra perceptual and memory processing associated with uncanniness is strongly associated with less empathy and less brain sexual arousal. This is in phase with what we know about mu suppression and empathy (Jackson & Decety, 2004; Oberman et al., 2005; Woodruff, Martin, & Bilyk, 2011; Yang et al., 2009), as well as with the conclusions of Urgen et al. (2013) on how we react to borderline human appearance.
On the other hand, frontal alpha asymmetry almost perfectly matches sexual arousal as indexed by erectile responses (see figure 2 and 3), and these are both positively correlated with the subjective feeling of SP (Hypothesis 3; figure 1d). This is in line with the results of Prause et al. (2014). This result is also true for the Virtual video condition which does not go on its own way at that level. It is as if the more purely sexual dimension of SP was dissociated from the strangeness of an erotic stimulus which human avatars are equivocal. Or is it maybe that the uncanny valley effect is simply incompatible with sexual attraction? A different explanation, akin to the Freudian thesis of The Uncanny (1919), is that angst related to uncanniness (Unheimlichkeit) would emerge in a familiar situation (Heimlich) where an affect is repressed because of its instinctual underpinning. This repression would of course be right away of a sexual nature in the case of sexual presence. In the Freudian lingo, we would say that partial instincts aimed at partial objects, leading to an aesthetic experience suffused with strangeness. In a more modern terminology, some elements of the sexual gestalt of the stimulus were sexually appealing despite others. This leads us to Hypothesis 2 stating that levels of SP would come with distinct perceptual motor patterns related to sexual interests.
Our results clearly show that perceptual extraction dynamics of critical sexual cues changed following SP levels, especially for the preferred sexual object, namely the female character (with heterosexual males). Furthermore, a sexual stimulus closer to reality seems to require less processing time and a more fluid pattern of analysis, possibly because of its well-defined low-level visual features, or because of the less ambiguous and uncanny appearance of its protagonists. We presented elsewhere an explanation of perceptual extraction dynamics in VR, and especially with critical sexual cues (Renaud, Chartier & Albert, 2009; Renaud et al., 2010). In the Gibsonian perspective of ecological psychology and affordance-based perception, it can be said that extracting perceptual information for sexual purposes varies according to how well affordances are attuned to the person’s sexual goals and preferences.
In the same vein, we have hypothesis 4 stating that SP evoked by sexual stimuli would be modulated by perceptual motor patterns involved in the extracting of visual cues. Correlations between subjective identification to the characters, perceptual motor activity (especially at the facial level) and mu suppression are indeed quite interesting. There seems to be there a beginning of explanation of how identification to and embodiment in synthetic characters take place in a context of simulated sexual content. The illusion to participate by proxy in sexual activities seems to be based on specific perceptual-cognitive and brain processes. How these processes organize themselves and mutually constrain each other remain to be established.
Some limitations should be considered in this study. First, our sample is relatively small and reproducing these results with more participants is a must. Second, to completely validate our model of SP, female participants will have to be enrolled in the future; the present scientific contribution is nevertheless valuable because it lays the ground of a new methodology and validate central concepts related to SP. This will be done in the next phase of our research project. As for the sexual stimuli that were used, they are not completely equivalent across the conditions since we could not animate the synthetic characters at the facial level. Finally, our assessment of SSP relies on a questionnaire whose validation is ongoing.
Conclusion
The results obtained are coherent with the definitional elements of SP proposed at the outset (Renaud, Fontanessi & Benbouriche, 2014). However, what became clear with this study is that the understanding of the concept of SP is tightly linked to the uncanny valley effect. The deepening of the understanding of this conceptual proximity will be unavoidable in the future. From a Darwinian point of view, the fact that an individual can prepare for breeding with an illusion, and an illusion without valid genetic characteristics, is somewhat confusing… Confusing and possibly a way to shed light on the very concept of presence itself by looking at conflictual motivations at play when facing VR driven illusions.
The next steps will thus consist initially in better controlling the uncanny valley effect in a sexual context, with participants of both sexes, as well as according to gender differences. Recent progress in measuring the uncanny valley effect will certainly help achieving this (Ho & McDorman, 2017). Similarly, the role of immersion level will have to be understood, at the first and third person, as SP experienced in these conditions might be differently lived. Finally, we believe that the dynamics and interconnection of the brain components involved in SP should be probed using analytical strategies such as coherence analysis (Sporns, 2012; Tong & Thakor, 2011). This way, the tracking of the phase shifts in SP psychophysiological processes will be more integrated.
Deepening understanding of SP is necessary for practical reasons related to the growing place of cybersexuality in society. As stated earlier, it is also necessary from a fundamental point of view because it can help exploring the foundations of presence differently.
References
Baumgartner, T., Speck, D., Wettstein, D., Masnari, O., Beeli, G. & Jäncke, L. (2008). Feeling present in arousing virtual reality worlds: prefrontal brain regions differentially orchestrate presence experience in adults and children. Frontiers in Human Neuroscience, 2, 1-12.
Baumgartner, T., Valko, L., Esslen, M. & Jäncke, L., (2006). Neural correlate of spatial presence in an arousing and noninteractive virtual reality: An EEG and psychophysiology study. Cyberpsychology & Behavior, 9, 30-45.
Booton, J. (2015, July 15). Porn industry’s billion-dollar new frontier. Retrived from http://www.marketwatch.com/story/how-the-future-of-virtual-reality-depends-on-porn-2015-07-15.
Bouchard, S., St-Jacques, J., Robillard, G. & Renaud, P. (2008). Anxiety increases the sense of presence in virtual reality. Presence: Teleoperators and Virtual Environments 17, 376-391.
Chrysikou, E. G., and Thompson, W. J. (2016). Assessing cognitive and affective empathy through the interpersonal reactivity index: an argument against a two-factor model. Assessment 23, 769–777. doi: 10.1177/1073191115599055
Davis, M. H. (1983). Measuring individual differences in empathy: evidence for a multidimensional approach. J. Pers. Soc. Psychol. 44, 113–126. doi: 10.1037/0022-3514.44.1.113
Freud, S. (1919). The ‘Uncanny’. The Standard Edition of the Complete Psychological Works of Sigmund Freud, Volume XVII (1917-1919): An Infantile Neurosis and Other Works, 217-256.
Freund, K., Sedlacek, F., & Knob, K. (1965). A simple transducer for mechanical plethysmography of the male genital. Journal of the Experimental Analysis of Behavior, 8(3), 169-170. https://doi.org/10.1901/jeab.1965.8-169
Fromberger, P., Jordan, K., Steinkrauss, H., von Herder, J., Witzel, J., Stolpmann, G., … & Müller, J. L. (2012). Diagnostic accuracy of eye movements in assessing pedophilia. The Journal of Sexual Medicine, 9, 1868-1882.
Fromberger, P., Meyer, S., Kempf, C., Jordan, K., & Müller, J. L. (2015). Virtual Viewing Time: The Relationship between Presence and Sexual Interest in Androphilic and Gynephilic Men. PloS one, 10(5), e0127156.
Gable, P. A., & Harmon-Jones, E. (2008). Relative left frontal activation to appetitive stimuli: Considering the role of individual differences. Psychophysiology, 45, 275-278.
Ho, C. C., & MacDorman, K. F. (2017). Measuring the Uncanny Valley Effect. International Journal of Social Robotics, 9(1), 129-139.
Jackson, P. L. & Decety, J. (2004). Motor Cognition: A new paradigm to investigate self-other interactions. Current Opinion in Neurobiology, 14, 259-263.
Janssen, E. (2011). Sexual arousal in men: A review and conceptual analysis. Hormones and Behavior, 59, 708-716.
Joyal C.C., Neveu S.M., Boukhalfi T., Jackson P.L. & Renaud P. (2018). Suppression of Sensorimotor Alpha Power Associated With Pain Expressed by an Avatar: A Preliminary EEG Study. Front. Hum. Neurosci. 12:273. doi: 10.3389/fnhum.2018.00273
Kober, S., Kurzmann, J. & Neuper, C. (2012). Cortical correlate of spatial presence in 2D and 3D interactive virtual reality: an EEG study. International Journal of Psychophysioly, 83, 365-374.
Laws, D. R., & Marshall, W. L. (2003). A Brief History of Behavioral and Cognitive Behavioral Approaches to Sexual Offenders: Part 1. Early Developments. Sexual Abuse: A Journal of Research and Treatment, 15, 75-92.
Li, A. X., Florendo, M., Miller, L. E., Ishiguro, H., & Saygin, A. P. (2015, March). Robot form and motion influences social attention. In Proceedings of the Tenth Annual ACM/IEEE International Conference on Human-Robot Interaction (pp. 43-50). ACM.
Lombard, M., & Jones, M. T. (2004). Presence and sexuality. In Proceedings of 7th international workshop on Presence (pp. 28-35).
Lombard, M., & Jones, M. T. (2013). Telepresence and Sexuality: A Review and a Call to Scholars. Human Technology, 9, 22-55.
MacDorman, K., Green, R., Ho, C.- C. & Koch, C. (2009). Too real for comfort? Uncanny responses to computer generated faces. Computers in Human Behavior, 25, 695-710.
MacDorman, K. F., Srinivas, P., & Patel, H. (2013). The uncanny valley does not interfere with level 1 visual perspective taking. Computers in Human Behavior, 29, 1671-1685.
Mori, M., MacDorman, K.F. & Kageki, N. (2012). The uncanny valley. IEEE Robotics and Automation, 19, 98-100. (Original work published 1970 in Japanese)
Oberman, L. M., Hubbard, E. M., McCleery, J. P., Altschuler, E. L., Ramachandran, V. S., & Pineda J. A. (2005). EEG evidence for mirror neuron dysfunction in autism spectrum disorders. Cognitive Brain Research, 24(2), 190-98.
Plank, M., Müller, H.M., Onton, J., Makeig, S. & Gramann, K. (2010). Human EEG correlates of spatial navigation within egocentric and allocentric reference frames. In: Hölscher, C., & al. (Ed.), Spatial Cognition VII. Springer, Berlin Heidelberg, pp. 191-206.
Prause, N., Staley,C. & Roberts, V. (2015). Frontal alpha asymmetry and sexually motivated states. Psychophysiology, 51, 226–235.
Renaud, P., Fontanessi, L., & Benbouriche, M. (2014, March). Sexual presence as a dimorphic phenomenon. In Proceedings of the 15th Conference of the International Society for Presence Research (pp. 167-172).
Renaud, P., Bouchard, S., & Proulx, R. (2002). Behavioral avoidance dynamics in the presence of a virtual spider. IEEE Transactions on Information Technology in Biomedicine, 6, 235-243.
Renaud, P., Chartier, S., Rouleau, J.-L., Proulx, J., Goyette, M., Trottier, D., . . . & Bouchard, S. (2011). Using immersive virtual reality and ecological psychology to probe into child molesters’ phenomenology. Journal of Sexual Aggression, 19, 102-120.
Renaud, P., Chartier, S., & Albert, G. (2009). Embodied and embedded : the dynamics of extracting perceptual visual invariants. In S. Guastello, M. Koopmans, & D. Pincus (Eds.), Chaos and Complexity in Psychology: The Theory of Nonlinear Dynamical Systems. New York: Cambridge University Press, 177-205.
Renaud, P., Goyette, M., Chartier, S., Zhornitski, S., Trottier, D., Rouleau, J. L., et al. (2010). Sexual Affordances, Perceptual-motor Invariance Extraction and Intentional Nonlinear Dynamics: Sexually Deviant and Non-deviant Patterns in Male Subjects. Nonlinear dynamics, psychology, and life sciences, 14(4), 463-462.
Renaud, P., Rouleau, J.-L., Granger, L., Barsetti, I., & Bouchard, S. (2002). Measuring sexual preferences in virtual reality: A pilot study. Cyberpsychology & Behavior, 5, 1-9.
Renaud, P., Trottier, D., Nolet, K., Rouleau, J.-L., Goyette, M., & Bouchard, S. (2014). Sexual self-regulation and cognitive absorption as factors of sexual response towards virtual characters. Cyberpsychology, Behavior and Social Network, 17, 241-247.
Sporns, O. (2012). Discovering the human connectome. MIT Press, Cambridge MA.
Steuer, J., 1995. Defining virtual reality: dimensions determining telepresence. In: Biocca, F., Levy, M.R. (Eds.), Communication in the Age of Virtual Reality. Lawrence Erlbaum, Hillsdale, NJ, pp. 33-56.
Tong, S. & Thakor, N.V. (2009). Quantitative EEG analysis methods and clinical applications. Artech House: Norwood, MA.
Trottier, D., Renaud, P., Rouleau J.-L., Goyette, M., Saumur, C., Boukhalfi, T., & Bouchard, S. (2014). Using immersive virtual reality and anatomically correct computer-generated characters in the forensic assessment of deviant sexual preferences. Virtual Reality, 18, 37-47.
Urgen, B.A., Plank, M., Ishiguro, H., Poizner, H. & Saygin, A.P. (2013) EEG Theta and Mu oscillations during perception of human and robot actions. Frontiers in Neurorobotics, 7(19), 1-13.
Wacker, J., Chavanon, M. L., Leue, A., & Stemmler, G. (2008). Is running away right? The behavioral activation-behavioral inhibition model of anterior asymmetry. Emotion, 8, 232-249.
Witmer, B.G and Singer, M.J. (1998) Measuring Presence in Virtual Environments: A Presence Questionnaire. Presence: Teleoperators and Virtual Environments, 7, 225-240.
Woodruff, C. C., Martin, T., and Bilyk, N. (2011). Differences in self- and other-induced Mu suppression are correlated with empathic abilities. Brain Res. 1405, 69-76. doi: 10.1016/j.brainres.2011.05.046.
Yang, C. Y., Decety, J., Lee, S., Chen, C., & Cheng, Y. (2009). Gender differences in the mu rhythm during empathy for pain: An electroencephalographic study. Brain Res. 1251, 176-184. doi: 10.1016/j.brainres.2008.11.062.