Our present models of the human sexual response have been developed and refined from earlier proposals. The first four-phased word/text model for the human sexual response was that of Moll [8] as shown in Figure 4-1. His four descriptive phases were non-specific in terms of the various genital structures involved thus were applicable to both females and males. Havelock Ellis [9] oversimplified the model into just two phases, namely, “tumescence ” (the swelling of genital tissues) and “detumescence” (their resolution to baseline), a model that had poor explanation of the facts, viz., little consilience. Van de Velde [10], in his popular book Ideal Marriage, published diagrams that used a continuous graphic line to represent the sexual arousal (accumulated tension) of the female (and male) from their basal state ascending to a peak at which orgasm occurred and then rapidly falling away back to the basal state. He acknowledged, in a footnote, that earlier authors had also published such curves. The weakness of his line graphs was that they did not identify the nature of the two graphic axes although they could be surmised to be the degree of sexual arousal (vertical or ordinate y-axis) and its duration (horizontal or abscissa x-axis). Reich [11], a pupil of Freud and a controversial sexologist, published a graphic line, unisexual model showing a slow ascent to an apex (orgasm) followed by a rapid descent that described “the orgastically satisfying sexual act pertaining only to the course of a few, typical naturalistic phase and modes of behaviour” that lasted for 5–20 min. After the “forepleasure” (now renamed as foreplay), he divided up his line graph into five distinct naturalistic phases (penile penetration, phase of voluntary control of excitation, phase of involuntary control of muscle contractions, sudden and steep ascent to climax, orgasm). Dickenson [12] modelled the human sexual response as various types of coital scenarios plotting the line graphs of male and female “sensations” graphically against time in minutes. He acknowledged, like van de Velde [10], that similar graphs of coitus had been published by a number of other authors but without the actual timings (see Dickenson [12] for references). Whether the “sensations ” were central or peripheral or a combination was not described. In some respects, the early models adopted an elementary “nomothetic” approach , that is, establishing the common generalisations of the sexual arousal. Much later was the development of “idiographic” models (see biopsychosocial models of Basson, the Sexual Tipping Point and the Sexual Man models below) where the underlying “zeitgeist” was to characterise what makes sexual arousal individually unique. The methodologies used by the two approaches differ in that the nomothetic uses surveys, experimentation with measurements, statistical evaluations and analysis, psychometric testing and quantitative methodology, while the idiographic uses interviews, case studies, “focus” groups and qualitative methodology (e.g. Grounded Theory, observation, open-ended questions, iterative study design).

Wenger, Jones and Jones [13] published their simple, triphasic model of the human sexual response in their book Physiological Psychology in 1956 (Figure 4-2). Its three phases were suggested to be mediated by first the parasympathetic nervous system and then the sympathetic returning to the parasympathetic for resolution. The model was never seriously endorsed probably because:

Masters and Johnson [14] were aware of the publications of all four authors and incorporated their concepts into their own descriptive unisex model that had both graphic and text modes of presentation. The major difference was that they used their empirical findings from their laboratory physiological and observational studies to create and establish their model unlike previous authors’ mental conceptualisations. In their concise, sequential, text model of sexual arousal based on their laboratory observations, they arbitrarily categorised four phases described as excitation (E), plateau (P), orgasm (O) and resolution (R) often referred to by its acronym as the EPOR model (Figure 4-3). While the concepts of the excitation, orgasm and resolution phases are practically self-explanatory, that of the plateau phase is not. According to the descriptive accounts of Masters and Johnson [14], in the female plateau phase, mottling occurs on the breast skin, the colour of the engorged minor labia changes to a pink or bright red, the clitoral shaft and glans retract beneath the clitoral hood (retraction reaction), the outer third of the vagina becomes distended with venous blood that reduces the size of the entrance to the vaginal cavity (the orgasmic platform) and the full elevation of the uterus (vaginal tenting) is accomplished. In the male, a number of features occur, namely, a pre-ejaculation of secretion from Cowper’s (bulbourethral) gland, a small increase in the size of the penile coronal glans, maximum testicular elevation to the perineum and enlargement of the testes by vasocongestion.

The model, while conceptually useful, has not been without criticism [15, 16] although it is often overlooked that Masters and Johnson themselves stated “in apologia” that:

Robinson [17] argued convincingly that the plateau phase was misnamed as the sexual excitement did not actually plateau but was still rising towards the orgasmic climax and that all the changes described occurred simply in the late part of the excitement phase. There appeared to be no reason to create a different phase allowing its abandonment. The newer version of the model thus became the EOR model. Despite this, authors who apparently are not familiar with the various developments still include the plateau phase in the model. Another peculiarity is that the original EPOR model when described against the “circular” Basson model (see section below) has been named as the “linear model ” as it is the assumption that each EPOR phase occurs linearly from the previous phase. Yet Masters and Johnson [14], in the heading of the first chapter of their book, called their model “the sexual response cycle”, and their Figures 1.​1 and 1.​2 have legends the “male sexual cycle ” and the “female sexual cycle ” presumably because after the orgasm the resolution phase could return the individual’s status to basal. An important feature of the Master and Johnson’s model was that the females were able to have multiple, sequential orgasms unlike the males (see Figure 4-5).

Major weaknesses of the EPOR model were its overly strong genital focus and a lack of a phase of being aware of a desire for sexual activity. Two sexual therapists, Helen Kaplan [18, 19] and Harold Lief [20], reported that some of their female patients had an absence of a desire to undertake sexual activity even with their loved ones. She surmised that there must a period before the so-called excitation phase when the individual felt a need to undertake sexual activity and named this phase the “desire phase ”. The model (Figure 4-3) then became the DEOR model, viz., desire (D), excitation (E), orgasm (O) and resolution (R). It should be noted that Kaplan created this phase simply from her work with patients and never surveyed “normal” subjects to ascertain whether they actually experienced such a “desire phase” before partaking of sexual arousal activity. She was the first to identify disorders of sexual desire as a distinct entity. When Garde and Lunde [21] surveyed normal Danish women who were orgasmic, they reported that some 32% had never experienced “spontaneous” desire. In a later random sample survey of American women, Michael, Gagnon, Laumann and Kolata [22] found that some 33% of women (one out of three) answered “yes” to the question “During the last 12 months has there ever been a period of months or more when you lacked interests in sex?” Levin [23] suggested that these surveys indicated that some 33% of orgasmic women did not appear to have a desire phase preceding their excitation phase. Laqueur [24] summarised the earlier literature that questioned “the very existence of a female sexual desire”. No scientific epidemiological information about the incidence of female sexual desire was available in the nineteenth century. Kaplan and others automatically placed the desire phase before the excitation phase because at the time it was conceived that only one presentation of sexual desire was activated. While the “spontaneous” or endogenous desire has to be placed before the excitation phase, Levin [23] questioned the position of the desire phase in the model and suggested that a second desire phase could well be positioned during the excitation phase (Figure 4-3). This concept which is now called a “reactive desire” phase (D2) activated by sexual arousal per se compared to the “proactive desire phase” (D1) (Figure 4-1) has been accepted by many as a significant advance on the previous DEOR model (Basson, 16). Its shortened, acronymic form is the D ₁ D ₂ EOR model.

Other authors produced criticisms of the Masters and Johnson [14] model. Hoon [25] voiced his dissatisfaction in relation to its definitional and sequential reliability as there was no interobserver agreement about the changes observed, while Guttman scaling was not applied. Tiefer [15] questioned many aspects of the study from a clinical and feminist point of view as she argued that it was based on biased subject selection, experimenter bias and biases in methods. Basson [16], in her analysis, tabled a number of the major flaws in its conceptualisation and offered a new model of female sexual arousal (see below). Morrow [26] criticised it from a sociological perspective, while Levin [27] highlighted some unexamined mechanisms and some incorrect features involved in the sexual arousal processes and proposed specific modifications and corrections. It should be remembered, however, that despite all the criticisms the basic model, albeit with some modifications, has survived for more than 50 years, and many of the descriptions of what happens during human sexual arousal are still those of Masters and Johnson.

Whipple and Brash- McGreer proposed a circular model of the female sexual response composed of the four stages of Reed’s (1998) Erotic Stimulus Pathway, namely, seduction (desire, attraction), sensations (excitation and plateau), surrender (orgasm) and reflection (resolution) [29]. The phases in brackets represent the phases of the EPOR model in juxtaposition to the four new phases. The authors suggested that if the sexual experience was agreeable and satisfying, it would have a positive feedback to undertake a further “seduction” experience. Apart from the circularity, the four new phases made little or no impact to the development of the arousal literature.

The next significant development in modelling the female sexual response was that of Rosemary Basson [16, 30], a medical sex therapist working in Canada. Based on her clinical experiences treating women with sexual difficulties, she conceptualised a qualitative biopsychosocial model that has become known as the “circular model” of the female sexual response. While she accepted that women in new sexual relationships usually followed the DEOR model of sexual responses, she argued that women, especially those in long-term relationships, are not always activated to undertake sexual activity prefaced by a desire phase (Figure 4-4). She proposed that many started from a position that she called “neutrality” (i.e. basal) and that subsequent sexual stimulation then activated their sexual desire. Moreover, they did not necessarily undertake the sexual activity to obtain sexual release but more for other rewards or gains such as intimacy needs and feelings for the partner. The model has become very popular as an alternative to the DEOR , but its general applicability to all groups of women has been questioned (see section on empirical testing below).

With the publication of different models of female sexual arousal, the problem of evaluating their validity becomes critical. One obvious way of empirically testing their validity is to ask women which model best fits their own sexual behaviour/activity. The study of Sand and Fisher [31] was the first to empirically test the endorsement of the EPOR, DEOR and the biopsychosocial models in a community sample of women (registered nurses, n = 133) with and without sexual dysfunction as to what model best fitted their sexual experience. Their possible dysfunctional aspects were assessed using the validated Female Sexual Function Index (FSFI; Wiegel, Meston and Rosen [32]). They found that equal proportions endorsed the EPOR, DEOR and biopsychosocial model, but those with the lowest FSFI score, indicating possible sexual difficulties, chose the latter model. They argued that this showed heterogeneity of women’s sexual response and that the biopsychosocial model best reflected women with sexual concerns.

Since the Sand and Fisher [31] study, a number of other investigations on how women with and without sexual dysfunction endorse the various sexual response models have been published, and there is now significant controversy over which groups of women endorse the different models. Giles and McCabe [33] used an anonymous online survey that was completed by 404 women. They concluded, like Sand and Fisher, that the linear model more accurately represented the sexual responses of women without sexual dysfunction, while a modified circular model of Basson best fitted that of women with female sexual dysfunction. Hayes [34] undertook a systematic review of papers that compared the linear and circular models published since 1990. Of the 898 studies identified of which 13 met the inclusion criteria, only two compared the linear and circular models with limited evidence that most women identified with the linear model although some aspects of the female response fitted the circular model. Basson’s [35] recent, but delayed, response to such studies was “that the FSFI questionnaire used to assess sexual concerns although validated reflected a non-evidence based conceptualization of the sexual response simplified as a linear entity of discrete sequential phases, beginning with the desire at the outset of sexual activity, arousal that is focussed on genital events rather than the subjective excitement”. Other reviews of female sexual response models can be found in Wylie and Mimoun [36] and Perelman [37].

More recently, Giraldi, Kristensen and Sand [38], using an online study, investigated further which sexual response model represents a large cross-sectional sample of sexually active Danish men and women endorsed. In the case of the women, 34% chose the EOR model (Kaplan), 28% the linear EPOR model (Masters and Johnson 14), 25.6% the circular Basson model and 12.5% none of the models presented. Those women that showed sexual dysfunction (assessed by the FSFI) and most importantly also suffering distress significantly related to the circular Basson model. They concluded that women with no sexual dysfunction who were satisfied with their sexual life endorsed the linear EPOR model. This study was criticised by Basson, Correia, Driscoll, Laan, Toates and Tiefer [39] on a number of aspects. First, questioning why the authors only chose to ask about one type of sexual experience, and secondly the descriptions of the models used was questioned and whether the women understood the important aspects of the model, and it was suggested that the women could have been confused by the descriptions and chose the simplest formulation. It was argued that the “way in which the items tapping into sexual desire are worded, use of the FSFI and IEF (Index of Erectile Function) biases against participants who rarely acknowledge non-triggered desire”. Those more likely to endorse non-sexual desire will score higher on the desire domains of these two questionnaires. Given that there are no questions that tap into responsive desire persons with more of the latter will of course score in the “dysfunctional range”. The rebuttal by Giraldi, Kristensen and Sand [40] discounted the second criticism as they claimed they employed the wording used by Basson herself to describe the models. Their counter against criticism of the use of the FSFI and the IEF was that while they were not perfect they were the best validated questionnaires to use. Furthermore, they also assessed sexual distress and satisfaction with sexual life, and the women with a low FSFI score and high distress endorsed the Basson model. This result was in line with other quoted epidemiological studies. The dispute remains unsettled.

This unisexual, biopsychosocial behavioural and cultural model developed by Perelman [41] embraces both mental and physical aspects of sex. It is based on the concept that both mental and physical (MAP) factors can either “activate” or “inhibit” sexual arousal and in this respect has similarities to the dual control model. The obvious difference is that the Sexual Tipping Point (STP) model is usually portrayed using a graphic model of a classic two-pan weighing balance. One pan is weighted by inhibiting or negative factors of sexual arousal while the other pan by excitatory or positive factors. Whichever weighting is the heaviest upsets the balance to the negative or positive side of sexual arousal. The model was trademarked in 2013 and assigned to the MAP Education and Research Fund to allow it to freely distribute Sexual Tripping Point resources (explanatory video and related animations, publications and presentations) worldwide. A detailed history of the STP model development can be accessed on the Internet at www.​maped.​org/​history/​ [42].