Engagement with Female-oriented Male–Male Incest Erotica: A Comparison of Sinophone and Anglophone Boys’ Love Fandom

ABSTRACT

Boys’ Love (BL) is a global by-and-for women genre of youth culture focused on male-on-male sexuality and romance. Incest relationships are not uncommon in BL yet there is no research on what kinds are of most interest to the audience, and with which sub-demographics, and none offer an intercultural comparison. We address this lacuna analyzing data from the largest BL audience survey to date in both Anglophone (N = 1715) and Sinophone (N = 1922) regions. CATPCA reveals a strikingly similar component structure across the two cultures with a preference hierarchy descending from non-blood relationships, blood intergenerational, then brothers. For both regions, it is the avid fans who tend to consume incest material while, for the Sinophone, it is the less socially empowered who appear most engaged: women, the other-gendered, and the nonheterosexual. Moreover, Sinophone fans are more concerned about legal issues than are Anglophone. Other subtle cultural differences suggest Sinophone BL fans focus on family rules and roles and the Anglophone on the intimacy of brotherly bonds. As young women have increasing opportunity to create and consume sexuality explicit material geared to their particular tastes and needs, our study provides important information to inform debates about the forms, functions, and legislative context around pornography.

Pornography is a social, legal, and intensely gendered issue. Sexually explicit material is subject to public scrutiny and legislation, almost universally, and is often perceived as a masculine space in which women are denigrated and exploited (Ezioni 2020). However, increasingly, young women, are open creators and consumers of pornography (Attwood, Smith, and Barker 2019). Boys’ Love, a.k.a, BL, yaoi, danmei, amongst other designations, is one such phenomenon being a by-and-for women genre of youth culture that caters for a taste in male-on-male sexuality and romance (McLelland and Welker 2015). It presents in many formats including manga, fan fiction, fan art, light novels, computer games, by amateurs and professionals, in concrete and virtual form, and is truly global. BL has two independent roots in the early 1970s: Japanese shonen-ai (literally ‘boy love’) manga and Anglophone slash fiction, mainly in the US, developed from sexualized fan fiction about Kirk and Spock from the TV series Star Trek (Kustritz 2003).

BL culture took-off in the Anglophone West during the 1980s (Madill 2018) and gained traction in Sinophone regions (Mainland China, Macau, Hong Kong, Taiwan) in the late-1990s (Wei 2014). These are regions of world importance with very different social and ideological outlooks. The different contexts are reflected in their overarching approach, and specific impasses, in relation to sexually explicit material, youth culture, sexual diversity, and gender. Female-oriented male-male erotica sits at the apex of this complex and fraught constellation, complicated by the sheer mass of BL material available, mostly on the internet, and the relative youth, and gender, of its audience. That is, recent research on the demographics of the BL audience confirms that it is predominantly female: 88% in Sinophone and 75% in Anglophone regions, and relatively young: mean age of 20 years in Sinophone and 23 years in Anglophone regions (Madill and Zhao forthcoming).

In Mainland China, by far the most populated Sinophone region, all formal publications are reviewed and registered by the related government department and must receive a permit before publication. Thus, most self-published material, as is most BL, is illegal (Yang and Xu 2017). Moreover, authorities do not allow materials with sexual contents except where it is considered to be high literature or classical arts (Wang 2007). Hence, BL is vulnerable when authorities implement online ‘clean-ups’ during which platforms are instructed to delete problematic material. For example, in April 2018 the official regulation account of social media platform Sina Weibo announced that it would eliminate all BL manga, fiction, and other sexual materials related to homosexuality. This lead to more than one hundred accounts being locked and an uncountable number of post deleted (Chang and Tian 2020). In fact, authorities in Mainland China have taken an increasingly antipathetic position toward homosexuality and, as a result, also BL. For example, the China Television Drama Production Industry Association issued the General Rules for the Contents of Television Series on December 31, 2015 which prohibited the portrayal of non-heterosexuality or ‘immorality’ including homosexuality, bisexuality, transgenderism, extramarital relationships, wife swapping, and other activities associated with sexual liberty.

Researchers, such as Zheng (2017), are concerned that BL may have a bad influence on young people’s sexual identity and views about gender and the basic assumption in Chinese society is that sexually explicit materials are harmful and immoral. This puts BL fandom in Mainland China at risk, and between 2011 and 2017 there were at least five arrests for uploading BL pornography online (for background see Yang and Xu 2016). Moreover, a Chinese sexuality scholar – Professor Yinhe Li – who supported BL authors and was critical of the limits placed on freedom of expression on Weibo was banned from posting online for months. Within this context, it is not surprising, possibly, that there is hardly any research on engagement with pornography in Mainland China.

In Anglophone regions, there are many and variable restrictions on sexually explicit material covered by obscene publications and pornography laws, which attempt to keep up with internet technology. Arguably, the main foci in recent years have been legislation on ‘revenge’ porn, child pornography, and ‘extreme’ pornography such as that involving real animals (Perry 2019). This includes the aborted attempt in Australia on internet filtering of ‘child abuse material’, which, partly because of manga’s cute-and-youthful visual stylization, would have captured much BL (McLelland 2012). Generally, the Anglophone West has been reasonably tolerant of mainstream ‘soft porn’, although sexually explicit material has divided ‘sex positive’ and ‘anti-porn’ feminists in a bitter ideological battle (Popolla 2020). Today, Western women’s engagement with pornography is increasingly normative, at least amongst younger age groups (Attwood, Smith, and Barker 2019).

In its focus on male-male sexuality, BL explores a vast range of scenarios from the gently implicit, through to romance, the sexual explicit, BDSM themes, and ‘fetishistic’ niche taste. Possibly surprisingly, incest relationship material is not uncommon (Zsila and Demetrovics 2017), perhaps particularly within Chinese BL/danmei (Chao 2016). Classic early progenitor BL manga Song of the Wind and Trees (Takemiya 1976-1984) is credited with introducing incest themes (Bollmann 2010). We focus on incest themes in this article because incest is highly charged, socially taboo, and proscribed in some form in most societies (Turner and Maryanski 2005). Hence, young women’s apparent interest in male-on-male incest fantasy is intriguing. More pragmatically, we have the largest dataset to date which includes information on engagement with BL incest material in two world regions with differing ideologies of the family. Although, of course, Sinophone and Anglophone regions are very diverse, they can be differentiated broadly in terms of a focus on, respectively, the extended versus the nuclear family, and on patriarchally-structured (Jankowiak and Moore 2016) versus increasingly single-mother led, blended, and inconstant family environments (Djundeva, Dykstra, and Emery 2019). Offering some evidence for the impact of the different ideologies of the family in these two regions, research on Harry Potter BL fan fiction in which Harry and Draco are married demonstrates that they reflect basic cultural differences. That is, in the Chinese material the pair were much more likely to live in an extended family, for it to be important that the marriage is endorsed by the wider family, and to produce children, especially boys (Madill, Zhao, and Fan 2018).

In terms of research on the incest content of BL, the literature is small, and most focuses on Chinese danmei. Xu and Yang (2013) argue that father–son incest stories in China “showcase a feminine attempt to reorder the power structure in the family by means of eros” (p.30). Tatlow (2014) reports online an interview with BL researcher Katrien Jacobs in which they suggest that father–son danmei in which the son is dominant is “a veiled call for social and political change [because] (i)n Confucian culture, the relationship between father and son mirrors that of ruler and servant, so subvert the father and you subvert the state.” Chao (2016) includes a discussion of incest in what they call the ‘grotesque eroticism’ of danmei in which women are described as “(p)laying with and subverting the notion of a masculine parameter in a patriarchal institution” (p.72). There is little research on incest BL in other contexts, but McLelland (2011) cites Japanese academic Yukari Fujimoto as suggesting that, at least some, incest BL may have a healing function for women who are often the victims of such sexual abuses. Another Japanese researcher, Nagaike (2012), has drawn on psychoanalytic theory by appealing to women’s repressed incest fantasies, and Foster (2015) posits the internalization of a lost and taboo object of love in his consideration of slash fiction in the US.

There is no research on what kind of incest relationships are of most interest to the BL audience, and with which sub-demographics, and none offering an intercultural comparison on this topic. Hence, we ask: Which incest sexual content do Sinophone and Anglophone participants particularly enjoy in BL and does this differ between cultures? and Are there sub-demographics in Sinophone and in Anglophone culture who enjoy particular forms of incest sexual content in BL and do these forms relate also to enjoyment of particular storylines and concern with legal issues? These are interesting questions given that we know relatively little about young women’s engagement in pornography, and particularly pornography implicating taboo family themes.

Materials and methods

Ethical approval was obtained from the School of Psychology Ethics Committee, University of Leeds, UK.

Data collection

Our 43-question survey has five sections: demographics, BL materials, feelings about BL, social relationships, and other erotic materials. Responses are, in the main, on a five-point likert-scale and some questions include an open-text response box. The English-language version opened in November 2014 and was promoted via relevant internet forums, social network websites, and by e-mailing anime and manga clubs. The version in Simplified Chinese opened in March 2015 and was promoted via universities in China, leaflets distributed with products from an online Chinese BL shop, and manga events in China at which a paper copy was provided for immediate completion and transferred online for analysis (N = 200). Data were collected until November 14, 2018. Although BL is an evolving genre, there is no reason to believe that incest material has changed significantly over this 4-year period. The Sinophone survey N = 1922 and the Anglophone survey N = 1715, the latter including only participants who noted their first language to be English. Demographic information is in Table 1.

Table 1. Demographics by region: age, gender, sexual orientation, nationality and country of current residence

Sinophone (N=1922)		Anglophone (N=1715)
Age in years
mean19.75 range 12-37 SD 3.05		mean 22.92 range 11-70 SD 7.09
Gender
11.1% of sample ‘male’ 1.1% of sample ‘other-gendered’		16.6% of sample ‘male’ 8.8% of sample ‘other-gendered’
Sexual orientation
66.5% of sample ‘heterosexual’ 64.1% of males ‘heterosexual’ 67.3% of females ‘heterosexual’ 25.0% of other-gendered ‘heterosexual’		31.1% of sample ‘heterosexual’ 9.9% of males ‘heterosexual’ 39.2% of females ‘heterosexual’ 2.6% of other-gendered ‘heterosexual’
Nationality	Current residence	Nationality	Current residence
Chinese (N=1883)	Mainland China (N=1693)	USA (N=774)	USA (N=1109)
Hong Kongese (N=38)	Hong Kong (N=198)	UK (N=331)	UK (N=333)
Taiwanese (N=1)	Macau (N=4)	Asian (N=122)	Asia (N=63)
	UK (N=12)	Australian (N=73)	Australia (N=96)
	USA (N=5)	Canadian (N=64)	Canada (N=83)
	Other (N=10)	New Zealand (N=12)	New Zealand (N=14)
		Mexican (N=8)	South Africa (N=3)
		South African (N=3)	Other (N=6)
		Other European (N=20)
		Other (N=5)

The Sinophone sample is younger and of a narrower age range than the Anglophone, possibly reflecting the shorter period during which BL has been popular in Sinophone regions. The Sinophone sample is more heterosexual and female than the Anglophone, commensurate with the strong heterosexual hegemony and traditionalism of Sinophone culture (Junren 2013), given that BL is thought of as a woman’s genre. Most Sinophone participants are from Mainland China. Anglophone participants broadly reflect the number of native English speakers in the larger Anglophone countries (Lyons 2017).

Creation of two new variables

First, responses to What is your sexual orientation? were recoded to two categories – ‘heterosexual’ and ‘nonheterosexual’ – the latter including responses bisexual, homosexual/lesbian/gay, not sure, and other (Sinophone survey) and bisexual, homosexual/lesbian/gay, polysexual/pansexual, not sure, and other (Anglophone survey). This managed the fact that, for cultural reasons, option polysexual/pansexual was included only in the Anglophone survey. Second, cumulative percentage to the question How intensely are you a BL fan? suggested the validity of dividing participants meaningfully into two groups: casual fans (not at all, a bit, quite: cumulative percentage Sinophone 49%, Anglophone 35%) and avid fans (a lot, extremely: cumulative percentage Sinophone 51%, Anglophone 65%) (Figures 1 and 2).

Figure 1. How intensely are you a BL fan? – Sinophone survey % response bar graph

Figure 2. How intensely are you a BL fan? – Anglophone survey % response bar graph

Data analysis

To address research question 1, we analyzed the multiple response survey question: Which ‘incestual’ sexual relationships do you particularly enjoy in BL? (Select all that apply or leave blank if you don’t particularly enjoy incestual sexual relationships in BL). Participants were able to select as many of the response options as they wanted from: father/son (seme/uke), son/father (seme/uke), brothers, uncle/nephew (seme/uke), nephew/uncle (seme/uke), each of which had four sub-types: blood, adoptive, step, and in-law (i.e., in total 20 options). Hence, each option was coded as individual binary-nominal type question (1 = no, 2 = yes). To clarify, the seme/uke distinction in BL indicates a 'top/bottom' or 'active/passive' relationship role.

Categorical Principal Component Analysis (CATPCA) was used to inspect underlying structure in the form of ‘components’ for each cultural group. CATPCA is a data reduction technique used when the data is categorical, or variables of mixed measurement, and does not assume a linear relationship between data (Manisera, Dusseldorp, and van der Kooij 2010). To perform CATPCA, the SPSS 21.0 Categories Module was applied. This reduces a data set – consisting of many variables with complicated correlational patterns – to a smaller number of relatively uncorrelated components each of which consists of relatively highly inter-correlated variables. Deciding the most meaningful number of components involves consideration of eigenvalue which indicates the amount by which a component has been stretched in a geometrical transformation. The elbow point of the scree plot (eigenvalue by component) and percentage variance accounted for (PVAF) are the most important indictors of fit for the principal components (i.e., most meaningful reduction of the data set with regard to semi-independent inter-correlations of variables) (Linting and van der Kooij 2012).

To address research question 2, we explored if a sub-demographic is associated with each incest content CATPCA component within each cultural group. Demographic variables of interest are gender, sexual orientation, and fandom intensity. Participants were identified who had endorsed all of the highest loading variables on each of the components. Where more than three variables were highly endorsed, only the top three were selected. First, using endorsement of key variables, we calculated the percentage of Sinophone and Anglophone participants engaging with each CATPCA incest component by gender and sexual orientation. Second, the Chi-square test of independence was used to determine if any of three categorical demographic variables – gender (male, female, other), sexual orientation (heterosexual, nonheterosexual), and fandom intensity (casual, avid) are associated with each incest component within cultural group.

We also explored the extent to which any sub-demographic associated with each incest content CATPCA component within each cultural group cohered in terms of liking for BL sexual storyline type and concern over legal issues with regard to sexual content. Hence, third, because the data are not normally distributed, the nonparametric test Mann–Whitney U was used to test for differences in median on the likert scale storyline variables – I like BL stories that are mostly romance and no sex, I like BL stories with ‘implicit relationships’, I like BL stories with explicit sex – and likert scale variable To what extent do you worry about legal issues in relation to the sexual content of BL? within cultural group in relation to endorsement of each incest component. The result of all comparisons undertaken are reported and, where significant, effect size is provided – Cramer’s V (Chi-square) and Pearson’s r (Mann–Whitney) – where trivial effect is less than .1, a small effect .1 to .29, a medium effect .3 to .49, and large effect .5 and above (Cohen 1988). The demanding significance level of p < .001 was set throughout.

Results

CATPCA of incest relationships – Sinophone data

The scree plot of the Sinophone data demonstrates that eigenvalues start to level-off (i.e., the ‘elbow’) after the second component (Figure 3). Although it is important to inspect the scree plot, Kaiser’s criterion is the most common technique in deciding the number of meaningful components and it recommends retaining all with an eigenvalue above 1. Field (2009) indicates that the Kaiser criterion is reliable when simple size above N = 250 and the Chinese sample is N = 1922. This means that the decision was made to retain components 1–4. (Component 5 has an eigenvalue of .843.)

Figure 3. Scree plot of Sinophone responses to Which ‘incestual’ sexual relationships do you particularly enjoy in BL?.

The total PVAF across the four components account for 72.25% of total variance in response of the Chinese participants to Which ‘incestual’ sexual relationships do you particularly enjoy in BL? (Table 2). This is much higher than the 60% of explained variance suggested for a ‘good’ factor analysis by Hair et al. (2018). Unrotated, component 1 accounts for the largest PVAF at 51.78%, with components 2–4 accounting respectively for 8.70%, 6.02%, and 5.65%. Cronbach’s alpha is a measure of internal consistency, with a coefficient of .70 or higher considered acceptable in the social sciences. So, unrotated, only component 1 has good internal consistency.

Table 2. Variance accounted for in the four components retained in the Sinophone responses to Which ‘incestual’ sexual relationships do you particularly enjoy in BL?

Component	Unrotated Solution			Rotated Solution^a
	Cronbach’s Alpha	Eigenvalue	PVAF %	Cronbach’s Alpha	Eigenvalue	PVAF %
1	.951	10.357	51.783	.928	4.313	21.566
2	.448	1.741	8.704	.931	4.302	21.512
3	.178	1.204	6.019	.904	3.385	16.924
4	.122	1.131	5.645	.886	2.432	12.158
Total	.980^b	14.432	72.159	.980	14.432	72.159

^aRotation Method: Varimax with Kaiser Normalization.

^bTotal Cronbach’s Alpha is based on the total Eigenvalue.

Without changing fit, an orthogonal, varimax rotation can produce the most structurally simple solution in which each variable loads as highly as possible on only one component. A simplified solution was, indeed, produced through rotation for the Sinophone data. In the rotated solution: (i) all four components retain an eigenvalue above 1; (ii) PVAF is redistributed in such a way as to maintain the strength of component 1 while improving that of the three smaller components: 21.57%, 21.51%, 16.92%, and 12.16%, respectively, and; (iii) Cronbach’s Alpha is now above .7 for all four components indicating that all now have an acceptable level of internal consistency. For ease of comparison with the Anglophone component structure, from now on the Sinophone components will be numbered 1a, 1b, 2, and 3.

Loading of each variable on each component is indicated by the Pearson correlation, the value of which ranges between −1 and 1. Components derived from CATPCA are interpretable in light of the variables loading most highly on each (Table 2). The variables loading most highly on component 1a are adoptive father/son, adoptive son/father, adoptive uncle/nephew, step uncle/nephew, adoptive nephew/uncle, step nephew/uncle. The similarities of the variables loading on component 1a suggest that it denotes enjoyment of ‘Adoptive’ incest relationships in BL. The variables loading most highly on component 1b are stepfather/son, in-law father/son, stepson/father, in-law son/father, in-law uncle/nephew, in-law nephew/uncle. The similarity of the variables loading on component 1b suggests that it denotes enjoyment of ‘Step/In-law’ incest relationships in BL. Because of the similarity of variables loading on components 1a and 1b – and to make comparison easier with the Anglophone analysis below – 1a and 1b are clustered under a higher-order label denoting enjoyment of ‘Non-Blood’ incest relationships in BL. The variables loading most highly on component 2 are blood father/son, blood son/father, blood brothers, blood uncle/nephew, and blood nephew/uncle relationship. The similarity of variables loading on component 2 suggests that it denotes enjoyment of ‘Blood’ incest relationships in BL. The variables loading most highly on component 3 are adoptive brothers, stepbrothers, and in-law brothers. The similarity of variable loading on component 3 suggests that it denotes enjoyment of ‘Brother’ incest relationships in BL.

CATPCA of incest relationships – Anglophone data

The scree plot of the Anglophone data demonstrates that eigenvalues start to level-off (i.e., the ‘elbow’) after the second component (Figure 4). Although it is important to inspect the scree plot, Kaiser’s criterion is the most common technique in deciding the number of meaningful components and it recommends retaining all with an eigenvalue above 1. The Kaiser criterion is reliable when simple size above N = 250 and the Anglophone sample is N = 1713. Hence, the decision was made to retain components 1–3. (Component 4 has an eigenvalue of .924).

Figure 4. Scree plot of Anglophone responses to Which ‘incestual’ sexual relationships do you particularly enjoy in BL?.

The total PVAF across the three components account for 72.44% of total variance in response of the Anglophone participants to Which ‘incestual’ sexual relationships do you particularly enjoy in BL? (Table 4). This is a much higher than the 60% of explained variance suggested for a ‘good’ factor analysis. Unrotated, component 1 accounts for the largest PVAF at 55.22%, with components 2–3 accounting respectively for 9.39%, and 7.83%. Cronbach’s alpha indicates that, although three components are retained on the basis of eigenvalue, unrotated only component 1 has good internal consistency.

Table 3. Comparison of rotated component structure of Sinophone and Anglophone incest relationship CATPCAs

	Sino	Anglo	Sino	Sino	Anglo	Sino	Anglo
	1a	1	1b	2	2	3	3
Variable	Non-blood Adoptive	Non-blood	Non-blood Step/In-law	Blood	Blood	Brother	Brother
Blood father/son	.101	.112	.229	.790	.833	.053	.209
Adoptive father/son	.724	.489	.174	.238	.508	.223	.353
Step father/son	.471	.579	.535	.276	.452	.170	.337
In-law father/son	.171	.659	.769	.300	.331	.140	.316
Blood son/father	.163	.276	.300	.767	.812	.020	.071
Adoptive son/father	.738	.648	.234	.205	.451	.170	.171
Step son/father	.494	.701	.541	.291	.394	.120	.197
In-law son/father	.207	.754	.773	.315	.266	.106	.205
Blood brothers	.108	.090	.084	.620	.392	.578	.737
Adoptive brothers	.490	.253	.078	.049	.107	.699	.855
Step brothers	.315	.263	.251	.095	.104	.771	.872
In-law brothers	.126	.358	.491	.136	.075	.719	.800
Blood uncle/nephew	.301	.375	.255	.737	.690	.150	.175
Adoptive uncle/nephew	.794	.752	.266	.151	.279	.201	.268
Step uncle/nephew	.547	.795	.537	.209	.221	.269	.266
In-law uncle/nephew	.290	.815	.759	.215	.182	.249	.261
Blood nephew/uncle	.344	.426	.278	.727	.707	.124	.069
Adoptive nephew/uncle	.809	.808	.266	.180	.258	.183	.165
Step nephew/uncle	.574	.857	.535	.212	.222	.219	.163
In-law nephew/uncle	.313	.873	.745	.218	.141	.218	.170

Notes: Highest loading per variable in bold; variables that differ between groups or anomalous are shaded

Table 4. Variance accounted for in the three components retained in the Anglophone responses to Which ‘incestual’ sexual relationships do you particularly enjoy in BL?

Component	Unrotated Solution			Rotated Solution^a
	Cronbach’s Alpha	Eigenvalue	PVAF %	Cronbach’s Alpha	Eigenvalue	PVAF
1	.958	11.043	55.215	.949	7.180	35.901
2	.492	1.878	9.391	.921	3.806	19.032
3	.380	1.566	7.831	.902	3.501	17.504
Total	.980^b	14.487	72.437	.980	14.487	72.437

^aRotation Method: Varimax with Kaiser Normalization.

^bTotal Cronbach’s Alpha is based on the total Eigenvalue.

A simplified solution was produced through rotation for the Anglophone data. In the rotated solution: (i) all three components retain an eigenvalue above 1; (ii) PVAF is redistributed in such a way as to maintain the strength of component 1 while improving that of the two smaller components: 35.90%, 19.03% and 17.50%, respectively; and (iii) Cronbach’s Alpha is now above .7 for all three components indicating that all now have an acceptable level of internal consistency.

Components derived from CATPCA are interpretable in light of the variables loading most highly on each (Table 3). The variables loading most highly on component 1 are stepfather/son, in-law father/son, adoptive son/father, stepson/father, in-law son/father, adoptive uncle/nephew, step uncle/nephew, in-law uncle/nephew, adoptive nephew/uncle, step nephew/uncle, and in-law nephew/uncle. The similarity of variables loading on component 1 suggests that they denote enjoyment of ‘Non-Blood’ incest relationships in BL. The variables loading most highly on components 2 are blood father/son, adoptive father/son, blood son/father, blood uncle/nephew, and blood nephew/uncle. The similarity of variables loading on component 2 suggests that they denote enjoyment of ‘Blood’ incest relationships in BL The variables loading most highly on component 3 are blood brothers, adoptive brothers, stepbrothers, and in-law brothers. The similarity of variables loading on component 3 suggests that they denote enjoyment of ‘Brother’ incest relationships in BL.

Sub-demographics associated with CATPCA components

In order to explore if a sub-demographic of each data set is associated with each incest relationship CATPCA component, participants were identified who had endorsed all of the highest loading, ‘key’ variables on each of the components (Table 5).

Table 5. Key variables in order of importance for each incest relationship CATPCA component by culture used to identify engaged demographic

	Component 1a Key variables		Component 1b Key variables	Component 2 Key variables	Component 3 Key variables
Sinophone	Adoptive nephew/uncle Adoptive uncle/nephew Adoptive son/father		In-law son/father In-law father/son In-law uncle/nephew	Blood father/son Blood son/father Blood uncle/nephew	Step brothers In-law brothers Adoptive brothers
Anglophone		In-law nephew/uncle Step nephew/uncle In-law uncle/nephew		Blood father/son Blood son/father Blood nephew/uncle	Step brothers Adoptive brothers In-law brothers

Where more than three variables are highly endorsed, only the top three variables were selected. Number of participants endorsing all key variables of a component were identified and the percentage of Sinophone and Anglophone participants engaging with each CATPCA incest component reported by gender and sexual orientation (Table 6). A higher percentage of Sinophone participants engage with incest relationship in BL than do Anglophone participants for component 1 (Non-Blood Relationships) and component 2 (Blood Relationships.) The pattern is reversed for component 3 (Brother Relationships).

Table 6. Percentage of Sinophone and Anglophone BL fandom engaging with each CATPCA incest relationship component by gender and sexual orientation

C	% of Sinophone (N = 1922)						% of Anglophone (N = 1715)
	Total	Male	Female	Other	Het.	Non-het.	Total	Male	Female	Other	Het.	Non-het.
1a	21.3% (N = 409)	1.2% (N = 23)	19.9% (N = 382)	0.2% (N = 4)	12.5% (N = 240)	8.8% (N = 169)	12.1% (N = 207)	1.1% (N = 17)	9.7% (N = 167)	1.2% (N = 21)	3.7% (N = 64)	8.4% (N = 143)
1b	10.5% (N = 201)	0.7% (N = 13)	9.6% (N = 184)	0.2% (N = 4)	5.2% (N = 99)	5.3% (N = 102)
2	10.3% (N = 198)	0.9% (N = 18)	9.1% (N = 175)	0.3% (N = 5)	5.0% (N = 97)	5.3% (N = 101)	5.7% (N = 98)	0.7% (N = 12)	4.3% (N = 74)	0.7% (N = 12)	1.3% (N = 23)	4.4% (N = 75)
3	25.2% (N = 484)	1.5% (N = 29)	23.2% (N = 446)	0.5% (N = 9)	12.3% (N = 236)	12.9% (N = 248)	30.7% (N = 527)	4.6% (N = 79)	23.0% (N = 395)	3.1% (N = 53)	8.0% (N = 138)	22.7% (N = 389)

Note: C means ‘component’.

The Chi-square test of independence was used to determine if any of three categorical demographic variables – gender (male, female, other), sexual orientation (heterosexual, nonheterosexual), and fandom intensity (casual, avid) are associated with each of the incest components within cultural group. We are particularly interested in engagement with ‘blood brother’ incest BL, so tested for associations also with the sub-variable. The Mann–Whitney was used to test for differences in median on the likert scale storyline variables I like BL stories that are mostly romance and no sex, I like BL stories with ‘implicit relationships’, I like BL stories with explicit sex, and likert scale variable To what extent do you worry about legal issues in relation to the sexual content of BL? within cultural group in relation to endorsement of each of the four components (Table 7).

Table 7. Summary statistical results for Sinophone and Anglophone demographic, storyline and legal concern associated with each CATPCA incest relationship component and the variable blood brother relationships

		Sinophone		Anglophone
Component/Variable	Variables	Effect size	Direction	Effect size	Direction
1a Non-Blood Adoptive; Non-blood	Gender	Small	Male↓ Female↑	/	/
	Sexual orientation	/	/	/	/
	Fandom intensity	Small	Casual↓ Avid↑	Small	Casual↓ Avid↑
	Romance no sex	/	/	/	/
	Implicit relationship	/	/	/	/
	Explicit sex	Small	↑	Small	↑
	Legal concern	Small	↑	Trivial	↑
1b Non-Blood Step/In-law	Gender	/	/
	Sexual orientation	Small	Het↓ Non-het↑
	Fandom intensity	Small	Casual↓ Avid↑
	Romance no sex	/	/
	Implicit relationship	/	/
	Explicit sex	Small	↑
	Legal concern	Small	↑
2 Blood	Gender	/	/	/	/
	Sexual orientation	Small	Het↓ Non-het↑	/	/
	Fandom intensity	Small	Casual↓ Avid↑	Small	Casual↓ Avid↑
	Romance no sex	/	/	/	/
	Implicit relationship	/	/	/	/
	Explicit sex	Small	↑	/	/
	Legal concern	Small	↑	/	/
3 Brother	Gender	Small	Male↓ Female↑ Other↑	/	/
	Sexual orientation	Small	Het↓ Non-het↑	/	/
	Fandom intensity	Small	Casual↓ Avid↑	Small	Casual↓ Avid↑
	Romance no sex	/	/	/	/
	Implicit relationship	/	/	/	/
	Explicit sex	Medium	↑	Small	↑
	Legal concern	Small	↑	Trivial	↑
blood brother	Gender	Small	Male↓ Female↑ Other↑	/	/
	Sexual orientation	Small	Het↓ Non-het↑	Small	Het↓ Non-het↑
	Fandom intensity	Medium	Casual↓ Avid↑	Small	Casual↓ Avid↑
	Romance no sex	/	/	/	/
	Implicit relationship	/	/	Trivial	↑
	Explicit sex	Medium	↑	Small	↑
	Legal concern	Small	↑	Trivial	↑

For both the Sinophone and Anglophone participants, it is the avid, as opposed to casual, fans who most like incest relationships in BL across the board. However, in the Sinophone sample, female participants better enjoy Non-Blood Adoptive and Brother incest relationships than male participants, the latter also particularly enjoyed by other-gendered participants. Moreover, in the Sinophone sample, nonheterosexual participants better enjoy all but the Non-Blood Adoptive incest relationships than heterosexual participants. For the Anglophone group, the only demographic variable differentiating interest in incest material, is that nonheterosexual fans particularly like blood brother relationships. The Sinophone, but not the Anglophone participants, are concerned about legal issue in relation to all BL incest material. Finally, participants who like explicitly sexual stories also tend to enjoy incest relationship material, although this pattern does not hold for Anglophone participants with respect to the Blood component.

Discussion

Our study is the first to explore and to provide robust and striking evidence that, when it comes to incest content, two major regions of the global BL fandom with very different sociopolitical contexts have a characteristic pattern of tastes, while identifying some subtle but important cultural differences.

There are some clear similarities between the Sinophone and Anglophone groups. First, there is a preference hierarchy: non-blood relationships are most popular, followed by blood intergenerational, followed by brothers, which are a close intragenerational relationship. That is, the closer the incest relationship becomes, the smaller the group with which it is popular. Hence, although incest relationships appear deviant, the group of fans expressing a liking for this material still demonstrate preference for the more distant forms by both genetics and generation. Second, the fact that all correlations between incest variables are positive indicates that, when incest material in BL is enjoyed, all types of incest relationships are generally liked. Although BL materials often deal with explicit sexuality, it is usually considered to be predominantly a romance genre (Pagliassotti 2010). Hence, it is likely that incest is a narrative technique for forging close and intimate relations rather than always portraying deliberately deviant sexuality. For example, in intergenerational incest stories such as father–son relationships, a more normative peer relation is often implied where there is visual portrayal of the protagonists, such as in the very popular manga Kirepap by Takagi Ryou.

There are also some important differences between the Sinophone and Anglophone groups. In terms of variables, the differences between the Sinophone and Anglophone participants are minimal and consist in the placing within component structure of adoptive father/son and blood brothers. In the Sinophone data, adoptive father/son sits comfortably in Component 1: Non-Blood Adoptive, whereas in the Anglophone it weighs most highly, anomalously, in Component 2: Blood. This is likely a statistical artifact, and adoptive father/son correlates well with both Component 2: Blood (.508) and Component 1: Non-Blood (.489) (Table 3). In the Sinophone data, blood brothers sits in Component 2: Blood, whereas in the Anglophone it sits in Component 3: Brothers. Hence, the cultural groups stress differentially the two aspects of ‘blood brother’ incest. Mirroring this, while the Anglophone components cluster all forms of brother incest relationships, the Sinophone differentiate blood brothers. In a strongly patriarchal culture, such as in Mainland China, the father’s bloodline is highly important and, hence, non-blood family members are clearly distinguished (Santos and Harrell 2017). In contrast, the Anglophone participants may be more focused on the close, obligated, peer relationship typified by brothers and less on whether-or-not this is a blood relationship.

As a group, Sinophone participants distinguish Non-Blood Adoptive and Non-Blood Step/In-law incest relationships while, as a group, the Anglophone combine these into one Non-Blood incest component. This difference may be connected to the strong family hierarchy and structure which is traditional in Sinophone cultures which differentiates very clearly family position and role (Zang and Zhao 2017). Moreover, Anglophone participants understand adoptive father/son incest to be like a blood relationship while, in the Sinophone data, it clusters with other non-blood relationships. In terms of cultural meanings, this difference in perceived closeness of adoptive father/son relationships is likely reflected in, and hence illuminated through, adoption traditions and process in the two regions.

In most Anglophone countries, adoption (outside of step-families) is via authorized institutions and involves a long approval process, for example, in the UK, of about 6 months (https://www.gov.uk/child-adoption/early-stages-of-adoption). Adoption is also relatively well accepted and common involving, for example, 2% of children in the US (https://www.parents.com/parenting/adoption/facts/). Moreover, in Anglophone regions, adoption tends to occur at a relatively young age such that, in the UK, 78% of children adopted in 2018 were aged 4 years or less (https://www.statista.com/statistics/536897/age-of-adopted-children-england-uk/) and, in the US, about 60% of children adopted via private organizations are placed within a month of birth (https://www.safy.org/). Hence, the acceptability, frequency, and young age of adopted children could mean that adoptive child–parent relationships appear little different from blood child–parent relationships for Anglophone participants. In contrast, adoptions more commonly occur between relative families than between strangers in Mainland China although, for this reason, accurate data are unavailable. Even so, Wei and Lei (2012) argue that adoptees can find it difficult to build a strong relationship with either their adoptive or blood parents. Moreover, the very fact of intra-family adoption in Mainland China makes the distinction ‘adopted’ and ‘blood’ (interpreted as first-degree relative) particularly important.

It is possibly not surprising that adoptive and in-law uncle/nephew or nephew/uncle relationship: that is, the Non-Blood incest relationships appear reasonably acceptable for both the Sinophone and Anglophone participants. However, interestingly, while the Sinophone differentiate ‘adoptive’ and ‘in-law’ relationships they are clustered together by the Anglophone. Differentiation in the type of non-blood family incest may be particularly important in Sinophone culture for traditional reasons. Although powerful clan traditions require men to continue their family line, in poor and remote environments it can be difficult to find a wife and girls can be adopted into a family as wives-in-waiting. This custom, called ‘Tongyangxi’, remained popular until relatively recently and still exists in some areas of Mainland China (Jiang 2010; Wolf and Huang 1980). Hence, a sexual relationship between adoptive relatives, as opposed to in-laws, could have very different social resonances for a Sinophone and Anglophone audience.

We interrogated our data for sub-demographics in Chinese and in Anglophone Culture who enjoy particular forms of ‘incestual’ sexual relationships in BL and if these forms relate also to enjoyment of particular storylines and concern with legal issues. Incest material in BL is enjoyed best by avid fans and/or those who like explicitly sexual stories. It may be unsurprising that liking the more controversial material, such as explicit sex and incest, go together. However, compared to casual fans, avid fans are also more likely to be familiar with the tropes and themes of BL, including the need to create intimacy situations in these stories. For example, the Japanese manga Sumanai!! Masumi-kun by Kaname Itsuki, about a father-son-in-law relationship, created an intimacy situation by having the daughter/wife leave abruptly on the day of the wedding. More generally, it is usual for BL to soften incest themes through contextualization, minimizing age difference in intergenerational incest through narrative and, where relevant, visual means, and romantic plots lines.

Demographic differences are more apparent in the Sinophone group such that female participants better enjoy Non-Blood Adoptive and Brother incest relationships, the latter also particularly enjoyed by other-gendered people, and nonheterosexual participants particularly enjoy all but the Non-Blood Adoptive incest relationships. In essence, it is the less socially empowered Sinophone demographics who appear most drawn to different facets of BL incest material: women, the other-gendered, and the nonheterosexual.

The Non-Blood/Adoptive and Brother relationships liked by Sinophone women, both imply a natural intimacy and might be close to the trope of ‘childhood sweethearts’ in romance fiction, which is considered a woman’s genre (Radway 1982), while offering particularly compelling plot dilemmas. In East Asian romance stories, childhood sweethearts play a central role with themes of growing up together, developing a close friendship, and evolution of a physical attraction. For example, the famous Tang Dynasty Chinese poet Lǐ Bái wrote the Song of Changgan which conveys the sentiments of a wife missing her husband and recalling their time together in childhood (Holyoak 2007). Even incest relationships play an important role in East Asian romantic stories. For example, Thunderstorm by Cao Yu is one of the most important plays of Chinese modern literature and associated with the relationship between a son and his stepmother (Chen 2010). Moreover, Autumn in My Heart, which first aired in 2000, is an extremely popular and influential South Korean television series about a romance between brother and sister which drew a massive home audience as well as huge ratings in Mainland China, Hong Kong, Taiwan, and Japan (Kim 2014). In the same way, as suggested in BL, incest stories may be attractive to a mass audience in East Asia for the romance of overcoming barriers and this may resonate, in particular, with people who face stigma or are relatively disempowered in their love life.

Interestingly, blood brother relationship had the most and strongest associations with the variables explored and is the incest relationship for which the Anglophone audience most mirrors that of the Sinophone: that is, avid fans, those who like stories with explicit sex, who have concerns about legalities, are nonheterosexual and/or, if Sinophone, are female or other-gendered. The fact that brothers mirror the central relationship form in BL: i.e., young male peers, may explain why blood brother relationships bring together the core features of each culture’s audience for incest material.

Finally, Sinophone fans are more concerned about legal issue in relation to all kinds of BL incest material than are Anglophone fans. This is, perhaps, explainable by the fact that, Anglophone regions, fictional incest material is not per se illegal whereas in Mainland China all sexually explicit material is illegal (Wang 2007).

This is the first study to generate the data that allows exploration, and with large samples, of the types of incest material most engaged within grassroots, yet global, sexual entertainment by-and-for young women. Moreover, we are able to explore this in relation to two very different cultural settings and determine the sub-demographics within each most engaged with different kinds of incest material. We demonstrate a striking similarity in component structure across Sinophone and Anglophone participants while evidencing minor differences interpretable as cultural divergences. In summary, Sinophone BL fans focus on family rules and roles and the Anglophone on the implications of intimacy afforded by brotherly bonds. For both groups, it is the avid fans who tend to consume incest material while, for the Sinophone, this includes also more people who face social stigma and/or are relatively disempowered in terms of who and how they love. As young women have increasing opportunity to create and consume sexuality explicit material geared to their particular tastes and needs, it is important to understand what these are in order to inform debates about the forms, functions, and legislative context around pornography.

Declaration of Interest Statement

The authors have no conflicts of interest to declare that are relevant to the content of this article.

Additional information

Notes on contributors

Anna Madill

Anna Madill is Chair of Qualitative Inquiry in the School of Psychology, University of Leeds, UK. She is a Fellow of the British Psychological Society and of the Academy of Social Sciences. She co-founded and chaired (2008-11) the BPS Qualitative Methods in Psychology Section, is on the editorial boards of Qualitative Psychology and Qualitative Research in Psychology. The authors’ online fandom survey can be found in Chinese at http://www.sojump.com/m/3989081.aspx and in English at https://leeds.onlinesurveys.ac.uk/blfandomsurvey. Anna’s projects include research funded by the British Academy on Boys’ Love manga. She tweets @UKFujoshi and her work can be found at https://leeds.academia.edu/AnnaMadill.

Yao Zhao

Yao Zhao completed her doctorate in the School of Psychology, University of Leeds, UK. Her thesis is titled ‘Understanding Yaoi/Boys’ Love Chinese Fandom’ and she was supervised by Professor Anna Madill. Yao was born in Kunming in Yannan Province in South West China and obtained her Bachelor of Science in Psychology at the Australia National University.