Humanities and social scientific research methods in porn studies

Abstract

Porn studies researchers in the humanities have tended to use different research methods from those in social sciences. There has been surprisingly little conversation between the groups about methodology. This article presents a basic introduction to textual analysis and statistical analysis, aiming to provide for all porn studies researchers a familiarity with these two quite distinct traditions of data analysis. Comparing these two approaches, the article suggests that social science approaches are often strongly reliable – but can sacrifice validity to this end. Textual analysis is much less reliable, but has the capacity to be strongly valid. Statistical methods tend to produce a picture of human beings as groups, in terms of what they have in common, whereas humanities approaches often seek out uniqueness. Social science approaches have asked a more limited range of questions than have the humanities. The article ends with a call to mix up the kinds of research methods that are applied to various objects of study.

Full text

Humanities and social scientific research methods in porn studies
Alan McKee
Abstract
Porn studies researchers in the humanities have tended to use different research methods from
those in social sciences. There has been surprisingly little conversation between the groups
about methodology. This article presents a basic introduction to textual analysis and statistical
analysis, aiming to provide for all porn studies researchers a familiarity with these two quite
distinct traditions of data analysis. Comparing these two approaches, the article suggests that
social science approaches are often strongly reliable – but can sacrifice validity to this end.
Textual analysis is much less reliable, but has the capacity to be strongly valid. Statistical
methods tend to produce a picture of human beings as groups, in terms of what they have in
common, whereas humanities approaches often seek out uniqueness. Social science
approaches have asked a more limited range of questions than have the humanities. The
article ends with a call to mix up the kinds of research methods that are applied to various
objects of study.
Keywords
pornography, methodology, quantitative, qualitative, textual analysis, statistical analysis,
reliability, validity
Introduction
In 2011, Associate Professor John D Foubert of Oklahoma State University wrote in the
journal Sex Addiction and Compulsivity that ‘It is difficult to find a methodologically sound
study that shows a lack of some kind of harm when men view pornography’ (Foubert, Brosi
and Bannon 2011, 213-214). In the area of porn studies, methodology – the study of research
methods – is particularly important. Foubert’s claim refers to quantitative research in the
discipline of Psychology. But there exist other research methods in other Disciplines which
produce quite different forms of knowledge. Foubert’s comment suggests that he is not
familiar with the variety of research methods available to researchers in Porn Studies, and
this is not uncommon. There has been surprisingly little discussion between researchers who
1

use statistical research methods and humanities researchers who use approaches like textual
analysis about the ways in which their research methods function, the different ways in which
they produce knowledge, and the implications of these epistemological differences for our
understandings of pornography as a phenomenon. This article aims to contribute to just such
a discussion.
How to read this article
This article takes a slightly unusual approach. It starts by presenting a basic introduction to
one of the key methods of data analysis from the humanities – textual analysis – and one of
the key methods of the social sciences – statistical analysis - so that readers by the end of the
article will, hopefully, have an understanding of each of them. The risk with such an approach
is that many readers will already be consummate practitioners of textual analysis – for them
that section will feel like something targeted at undergraduate students rather than something
suitable for an academic research journal. But hopefully, for those readers, the discussion of
statistical methods might be original and useful. Conversely, for other readers who have been
practising statistical methods of data analysis for thirty years, the section of the article
introducing those methods will feel embarrassingly familiar. But hopefully, for them, the
section on textual analysis will be illuminating. It is my hope that by the end of the article
there will be few readers who feel that they have not learned anything about research
methods: and an increased number of readers who feel that they now have a basic
understanding of both textual and statistical modes of analysis. I then go on to identify a
number of differences between humanities and social scientific approaches to porn studies;
and make a call to mix-up our research methods and objects of study.
Why write about research methods?
I’ve been researching and publishing on pornography since 1997. In those sixteen years I’ve
done many different things with pornography. My first degree was in Film studies, and I
began by subjecting pornographic films to the same kinds of textual analysis that I had been
trained to apply to film noir, musicals and German expressionism (McKee 1997). Then I
applied for, and won, my first research grant, to study pornography in Australia from the
perspectives of production, content and consumption. This lead me to explore a series of
other approaches to pornography – including quantitative analyses of the texts of
2

pornography alongside surveys and interviews to gather data about the ways in which it was
made and the ways in which it was consumed (McKee, Albury and Lumby 2008). I moved to
QUT, a University which led the way in Australia in analyses of cultural policy, and so found
myself exploring what would happen if you tried to take a governmental policy-studies
analysis to pornography (McKee 2001). I had always been interested in entertainment more
generally, and my interest in the aesthetic systems employed in the evaluation of
entertainment by its consumers led me to edit a collection which included a chapter exploring
the aesthetic system of pornography (McKee 2007). As I explored the ways in which it is
possible to study pornography, and the ways in which different methods of data gathering and
data analysis produced different kinds of information – indeed, in some ways, produced
different objects of study, and different meanings of the word ‘pornography – I published
some of my insights into these issues (McKee 2009).
And so it was with delight that I received an invitation from Clarissa Smith and Feona
Attwood to contribute to this journal an article about research methods for studying
pornography. I have written before about some of the philosophical differences underlying
social scientific and humanities approaches to studying pornography (McKee 2009) and so in
this article I take a slightly different approach. As I suggest above, there has been surprisingly
little conversation between porn studies researchers from different disciplines about their
practices of data gathering and analysis. In this article I wanted to provide at least some sense
for practitioners of textual analysis and of statistical analysis of what it is that the other does.
In the genre of the academic journal article it isn’t possible to provide an overview of every
possible research method and its relationship to porn studies (I did try that in the first draft of
this article, resulting in a piece that discussed twenty seven different methods and was over
10,000 words long, dismissed by referees as unreadable and the editors of this journal as
unpublishable). It is for this reason that I chose to provide a brief introduction to textual
analysis and statistical analysis, to illustrate the use of these methods in porn studies, and to
take them as a starting point for a discussion of the implications of these different approaches.
Textual analysis
Textual analysis is one of the key research methods of the humanities. Despite this fact it’s
difficult to find a straightforward description of what textual analysis, or a step-by-step guide
as to how to do it. The humanities have not traditionally been rigorous in reflecting on or
3

accounting for their own research methods. I was once asked to write a guide to doing textual
analysis (see McKee 2003) and in the course of writing it I realized that there are a series of
different forms of textual analysis in the humanities which – at least in my experience – are
not generally made explicit as research methods. So in this article I draw out what seem to me
to be four common types of textual analysis – even if they are not usually recognized as such.
I should emphasise that these are not familiar or settled research methods. This is exploratory
writing.
i) Textual analysis (ideological): The first of my forms of textual analysis is ideological.
Such an approach looks for hidden ideologies in a text – such as patriarchy, racism,
heteronormativity, and so on. This form of textual analysis is characterized by a lack of
interest in the surface level of what texts appear to be saying, and also a lack of interest
in what interpretations audiences say they make of texts. This approach also tends to
look for negative interpretations of a text – no matter how positive a text might appear
on the surface, ideological textual analysis aims to find a negative reading (Albury
2009, 648). This approach to the study of pornography has been extremely popular. In
one example, Jensen and Dines conducted an ‘interpretive analysis’ of fourteen
pornographic videos, at each point looking for a negative interpretation of material. For
example, they found that there were no videos in their sample that included rape scenes
– and they describe this finding in the following way:
While some pornographic videos portray women as reluctant or prudish, in need
of being coaxed or coerced into having sex, the women in the videos in our
sample never said no and were always immediately ready for sexual activity … In
short, virtually all women in the videos were portrayed as ‘nymphomaniacs’
(Jensen and Dines 1998, 73)
At each point where a number of interpretations are possible – for example, the fact
that there is no coercion of women in the sample of videos analysed could be
interpreted as a sign of women’s consent or agency – ideological textual analysis seeks
a negative interpretation.
ii) Textual analysis (poststructural). The second form of textual analysis I propose is
poststructural – by which I mean an analysis of a text that makes an informed guess
about the meanings of that text made by the audiences who consume it (I explain why I
use the term 'poststructural' for this form of textual analysis in McKee 2003, 9-13).
4

This form of textual analysis is interested in surface meanings, believing that these
provide us with useful information about how populations make sense of their world
(as to the question, if we’re interested in the meanings that audiences actually make of
texts, we don’t just interview them to find out rather than doing textual analysis? see
Textual Analysis: A Beginner’s Guide, pages 83-89). In an instance of this kind of
textual analysis Margaret Henderson analyses two lesbian-produced Australian
pornographic magazines, Wicked Women and Slit. She considers the texts themselves,
the genre in which they operate, the industrial context of their production, their likely
audiences and the wider cultural context, and uses this information to produce situated
interpretations of the texts to support her argument that these magazines ‘put
pornography in the service of lesbians to de/mystify lesbian flesh: to show lesbians the
real and unreal relations of the lesbian and her kind’ (Henderson 2013, 178).
iii) Textual analysis (appreciation): this form of textual analysis is most common in the
disciplines of literary studies, film studies and visual arts. Writers taking this approach
celebrate the text (often understood as a work of art), talking about its beauty or other
aesthetic achievements. In academic studies of pornography it’s rare to find this form
of textual analysis applied to pornographic texts themselves. Mark McLelland’s chapter
‘The best website for men how have sex with men’, which takes the form of an
appreciation of a sexually explicit website, might be a taken as an example of this
category:
I enjoy CFS primarily as a discursive space – it offers me visual, but importantly,
narrative pleasure. It is a subversive space – as insulting to mainstream
heterosexual norms as it is to a new homonormative gay orthodoxy that sees gay
‘liberation’ in assimilating those very norms. In other words, CFS is fully sick
(McLelland 2007, 83)
iv) Textual analysis (exegesis): my final category of textual analysis (and I emphasise
again that these are not settled and broadly accepted categories – I developed them
when I returned to textual analysis as a method of analyzing data and tried to
understand exactly what academics do under that rubric) is exegesis. When using this
approach the writer explains the ideas that are put forward by a text. This approach is
most commonly used by academics in the analysis of books by other academics.
However, it is increasingly being applied to non-academic texts (as in Thomas
McLaughlin’s exegesis of the intellectual work of blues songs - McLaughlin 1996).
5

Kobena Mercer has taken this approach to pornography, for example, explaining the
intellectual work of Robert Mapplethorpe’s gay nudes:
… the shocking modernism that informs the ironic juxtaposition of elements
drawn from the repository of high culture – where the nude is indeed one of the
most valued genres in Western art history – can be read as a subversive
recording of the normative aesthetic ideal. In this view it becomes possible to
reverse the reading of racial festishism in Mapplethorpe’s work, not as a
repetition of racist fantasies but as a deconstructive strategy that lays bare
psychic and social relations of ambivalence in the representation of race and
sexuality (Mercer 1991, 186-187)
This approach may seem similar to textual analysis for appreciation. I would argue that
appreciation can take a number of forms, while exegesis is a subset of appreciation that
is particularly interested in the ideas that are offered by a text.
This approach can also include writing about pornography which refers only to
previous academic or activist writing about pornography but does not gather any new
data from outside those genres. For example, Ullen publishes an article where he
engages with the writing of philosopher Rae Langton, who claims that speech act
theory helps us to understand how pornography works. Ullen clarifies her ideas, and the
ideas of John Austin about speech act theory, in order to argue that Langton’s claim is
unconvincing (Ullen 2013).
Statistical analysis
Quantitative research is a key aspect of social scientific porn studies, and the use of statistical
analysis is a key element of quantitative research. Generally speaking the social sciences are
more rigorous than the humanities in their approach to research methods. Methods are
explicitly spelled out and taught to undergraduate students. It is expected in most pieces of
social science research that you will make explicit what methods of data gathering and
analysis were used to produce your results. The four forms of statistical analysis presented
below are likely to be familiar to most social scientists:
i) Statistical analysis: regression analysis. Regression analyses are used to quantify the
relationships between variables – particularly how a change in an independent
variable will affect a dependent variable. This has been the most common statistical
6

approach to studies in pornography, typically used to investigate whether exposure to
pornography is related to attitudes towards women or attitudes towards sex. Braun-
Courville and Rojas surveyed 433 adolescents, and then subjected the data to
regression analysis. They write that:
Binary logistic regression analyses revealed that adolescents exposed to [Sexually
Explicit Websites] were significantly more likely to have multiple lifetime sexual
partners, more than one sexual partner in the last three months, used alcohol or
other substances at last sexual encounter, and ever engaged in anal sex (Braun-
Courville and Rojas 2009, 159)
The biggest issue for pornographic research using regression analysis is the difference
between correlation – two things happen at the same time – and causality – one thing
causes the other to happen. Psychological research into pornography using statistical
methods has consistently confused correlation with causality. For example, Braun-
Colville and Rojas find that consuming pornography is related to various other sexually
permissive acts. And like all researchers who do cross-sectional surveys they state
explicitly that they can’t say anything about causality:
We are unable to establish whether exposure to sexually explicit materials leads
to engagement in sexual behavior or whether those individuals who partake in
more high-risk sexual behaviors also have a tendency to seek out sexually explicit
Web sites (Braun-Courville and Rojas 2009, 161)
But despite this explicit acknowledgement, they write their article – as do many
psychologists using statistical approaches to the effects of pornography – from the
assumption that it is pornography that is causing the other sexual behaviors, throughout
the paper making statements about ‘the Internet’s impact on adolescent sexual attitudes
and behaviors’ (Braun-Courville and Rojas 2009, 156) and claiming that ‘prolonged
exposure [to pornography] can lead to … sexually permissive attitudes’ (Braun-
Courville and Rojas 2009, 158). They do this even though, as they themselves note,
‘Whether visiting sexual explicit Web sites leads to engagement in high-risk sexual
behaviors or vice versa cannot be established from this study’ (Braun-Courville and
Rojas 2009, 160).]
ii) Statistical analysis: factor analysis and structural equation modeling. These forms of
statistical analysis are employed in exploratory work and allow researchers to
7

understand the structure of the data or reduce numbers of variables. For example Hald
surveyed 699 young heterosexual Danish adults to find out if there were gender
differences in situational, interpersonal and behavioral characteristics of pornography
consumption. He found that four variables of pornography consumption were highly
correlated – ‘average time of use per week, frequency of use, pornography
consumption when having sexual activity on one’s own and exposure patterns of
pornography within the last twelve months’ (Hald 2006, 580). He conducted factor
analysis on the four variables to see if they could be combined into a single measure –
‘Pornography consumption’ – and found that it was meaningful to use the single
measure.
iii) Statistical analysis: cluster analysis. Cluster analysis is used to organize respondents
into groups. Lottes surveyed 395 people to see if they formed coherent groups in
regard to their attitudes towards abortion, biological differences between the sexes,
pornography and what constitutes sexual normality (Lottes 1985). Applying a cluster
analysis she found that the sample fell into four groups. The first group had the most
permissive sexual attitudes - including towards pornography – were less likely to
believe that differences between the sexes were biological and reported the lowest
frequency of religious service attendance. The second group gave least support to
egalitarian gender role attitudes but ‘weak to high’ support to sexually permissive
views, favourable views on pornography, and weak support to abortion and
homosexuality. The third group reported a ‘moderate’ rate of religious service
attendance, had the highest support for egalitarian gender roles, low rates of
acceptance for pornography, and moderate support for abortion and homosexuality.
The fourth group had the highest rate of religious service attendance, the highest mean
age, low support to egalitarian gender role attitudes, and consistently sexually
restrictive attitudes. It was the only group that consistently disapproved of both
pornography and homosexuality and ‘attributed gender behaviorial differences more
to heredity than social conditioning’ (Lottes 1985, 417).
iv) Statistical analysis: analyses of variance. T-tests and ANOVA (analysis of variance)
are used to detect differences between groups of respondents. Lo and Wei surveyed
2628 college and high school students to see if the third person effect in relation to
Internet pornography (the belief that other people are more vulnerable to negative
effects than oneself) is mediated by gender. They asked respondents to ‘estimate the
8

likely negative effects of “surfing pornographic websites” on moral values, attitudes
towards the opposite sex, sexual knowledge, sexual attitudes and sexual behavior’ (Lo
and Wei 2002, 21). Applying a t-test to the results they found ‘female respondents
were more likely than male respondents to perceive other male students to be more
negatively influenced by Internet pornography’ (Lo and Wei 2002, 23-24).
Reliability and validity
As I noted above, the humanities have tended to be less rigorous in their methodology than
have the social sciences. Some concepts drawn from the vocabulary of social sciences
methodology can usefully be applied to humanities methods – and to the differences between
humanities and social sciences approaches to porn studies. The terms ‘reliability’ and
‘validity’ will be familiar to any readers trained in the social sciences – although their
application to a humanities method like textual analysis might not be. As for readers trained
in the humanities – it is perfectly possible to get through an entire program of research
training in the humanities (as I did) without ever being introduced to these terms (although
this might be less true for younger researchers – certainly the research training our PhD
students in my Creative Industries Faculty get these days is more rigorous than anything I
was ever exposed to).
And so, for those readers like me who may not have a familiarity with these terms, a brief
summary - reliability describes the extent to which a method will produce the same data each
time it is applied, regardless of who is doing the analysis. Validity describes the extent to
which data actually describes what’s happening in the situation that is being studied. Some
research methods are more ‘reliable’ than others, while some are more ‘valid’. Sometimes
these two characteristics can work together, but at others they can come into conflict. Take
the example of a statistical content analysis of pornography (where you count the number of
times a certain thing happens in pornography) compared with a textual analysis (where you
describe in words what happens in pornography). Content analysis has a high reliability –
once you have created a definition of what you’re counting, then, using the same collection of
pornographic movies, any researcher can come up with the same results. By contrast, textual
analysis relies more on individual expertise: if you were exploring the question ‘What are the
power relations in 1984 gay porn video Powertool?’ using textual analysis, two different
9

researchers could come up with two quite different answers, based on their knowledge of the
genre, audiences, visual language, and so on.
But, at the same, while content analysis can have high reliability it can also risk low validity.
The process of counting allows for little flexibility or nuance, and so once you have your
definition in place you have to keep on counting – even if the interpretations that you end up
making do not match the ways in which a text is interpreted in the real world. So if you are
counting aggression in pornography, for example, once you have your definition – say ‘any
purposeful action causing physical or psychological harm to oneself or another person’
(Bridges et al. 2010, 1072) – then you have to count everything that might fall under that
definition. This would include consensual spanking for example – where one person asks that
another spanks them, and then says ‘I love that. It’s so sexy’. In a content analysis using the
above definition, this would have to be counted as an instance of aggression, even if it
doesn’t match up with what many people are concerned about when they say they want to
reduce violence in pornography. Content analysis doesn’t allow for such caveats - you just
find out that ‘On the whole, the pornographic scenes analyzed in this study were aggressive;
only 10.2% (n = 31) of scenes did not contain an aggressive act’ – even if most of them were
consensual (Bridges et al. 2010, 1075).
Common factors vs uniqueness
A second point about the different kinds of knowledge produced by a humanities approach
like textual analysis compared with a social scientific approach such as statistical analysis
concerns their different orientations towards similarity and difference. Large scale surveys
can provide a good sense of what large populations have in common but they’re not much
good at letting you understand the individual idiosyncrasies of how particular groups or
people make sense of the world – they favour commonality rather than uniqueness. On the
other hand, textual analysis is often applied to texts that are idiosyncratic rather than
representative – texts that show how things could be done differently, that are surprisingly
transgressive or creative or insightful. In those texts it is their very lack of representativeness
that is prized.
The kinds of questions asked
10

Finally it’s worth making explicit that research methods can only answer the questions that
the researcher asks. This is so obvious that it hardly bears mentioning – except for the
concomitant argument that therefore no research method can be objective. I’ve written before
about the different understandings of the concept of objectivity in the humanities and social
sciences (McKee 2009, 631-635). In the humanities it is now generally accepted that no piece
of research can be objective, in the sense of considering every possible point of view about an
area of study. By contrast, much social science work remains committed to the belief that
researchers can be properly impartial and let the facts speak for themselves. In reconciling
these approaches it’s useful to think about the questions that have been asked about
pornography. There exist hundreds of articles that use statistical methods to attempt to
discover whether consuming pornography causes men to have negative attitudes towards
women. But I haven’t come across any articles that use statistical attempt to discover whether
consuming pornography leads to more open levels of communication about sexuality, or to a
better level of acceptance of one’s sexual identity, or higher levels of sexual agency. The
studies that explore the question of whether consuming pornography causes men to have
negative attitudes towards women is good social science that gathers and analyses its data is
appropriate ways. But the research tradition in pornography can never be ‘objective’ in a
more everyday sense of that word because it can never ask every possible question about its
area of study. There will always be possible approaches, issues, voices and concerns that are
excluded from the research. Much of the statistical work on pornography has focussed on the
homogenous and the unitary – what is the (single) effect that pornography has on everyone
who consumes it?
By contrast, textual analyses of pornography have tended to address a wider range of
questions –exploring not only the ways in which pornography might represent women in
ways that are sexist or violent, but also representations of domesticity in amateur
pornography (Albury 1997), the porn star as brand (Nikunen and Paasonen 2007),
pornography and breastfeeding (Giles 2002) among many other topics. This is not to say that
there is any essence of textual analysis that makes it more suited to answering a wide range of
questions: it is rather that, historically, statistical forms of analysis in porn studies have
tended to focus on a more rigidly confined series of questions.
Conclusion
11

Associate Professor John D Foubert, following his contention that ‘It is difficult to find a
methodologically sound study that shows a lack of some kind of harm when men view
pornography’ (Foubert, Brosi and Bannon 2011, 213-214) goes on to use a t-test on survey
results to find that ‘men who saw mainstream pornography scored significantly higher on
self-reporting likelihood of raping and likelihood of committing sexual assault than men who
did not see mainstream pornography during the last 12 months’ (Foubert, Brosi and Bannon
2011, 221-222). But this is only one aspect of thinking about the relationships between
pornography, individual consumers, and the cultures and society within which it is consumed.
Clarissa Smith, for example, has suggested in her qualitative research a whole range of
responses that pornography might provoke – such as boredom, disappointment or
embarrassment (Smith 2002). At this moment in the history of porn studies I’m not aware of
any statistical work that attempts to measure the extent and importance of those responses to
the genre. But wouldn’t it be interesting to find out? What becomes clear from studying the
different research methods that are applied to pornography is that porn studies can benefit
from conversations about methodology across disciplines, and from more creative mixes of
research methods with objects of study. We are starting to see statistical methods applied to
original topics – such as the experiences of women who appear in pornography (Griffith et al.
2013). But there still remain many issues that have never been subject to statistical analysis,
and many aspects of textual analysis that could benefit from the rigour about methodology
typically applied in the social sciences. We have barely scratched the surface of the work to
be done – which is wonderful news to report in the first issue of a brand new journal devoted
to the area of Porn Studies.
Acknowledgments
This research was funded by an Australian Research Council Discovery grant, ‘Young
people, sex and the media’, 2012-2014
My grateful thanks for Andrew Ward for his insights into the range of statistical methods
available. All responsibility for errors in understanding the material remains my own.
References
Albury, Kath. 2009. "Reading porn reparatively." Sexualities 12 (5): 647-653. doi:
10.1177/1363460709340373.
12

Albury, Katherine. 1997. "Homie-Erotica: Heterosexual Female Desire in the Picture." Media
International Australia (84): 19-27. Accessed September 6, 2013.
http://qut.summon.serialssolutions.com/link/0/eLvHCXMwY2BQSEpKAXZvEk0skyw
MkoEpAtSETTQCLc4xN0o1NUpCGWxDKs3dhBiYUvNEGZTdXEOcPXRBG9rjCyBn
LsQbWJiYAFsaZqDmBWi_pBgDbyJoCXheCXirWIo4A0tJUWmqOKgIFQcaJ87AEW
FpFOQXYRkA4QrBuHrF4O1MeoUl4sASGxzbuoZ6pgAzSyvq. doi:
<http://search.informit.com.au.ezp01.library.qut.edu.au/documentSummary;dn=0844
32660575059;res=IELLCC>.
Braun-Courville, Debra and Mary Rojas. 2009. "Exposure to sexually explicit websites and
adolescent sexual attitudes and behaviors." Journal of Adolescent Health 45 (2): 156-
162. doi: 10.1016/j.jadohealth.2008.12.004.
Bridges, Ana J, Robert Wosnitzer, Erica Scharrer, Chyng Sun and Racheal Liberman. 2010.
"Aggression and sexual behavior in best-selling pornography videos: a content
analysis update." Violence Against Women 16 (10): 1065-1085.
Foubert, John D., Matthew W. Brosi and R. Sean Bannon. 2011. "Pornography viewing
among fraternity men: Effects on bystander intervention, rape myth acceptance and
behavioral intent to commit sexual assault." Sexual Addiction & Compulsivity 18 (4):
212-231. Accessed 14/02/2012. http://dx.doi.org/10.1080/10720162.2011.625552.
doi: 10.1080/10720162.2011.625552.
Giles, Fiona. 2002. "Fountains of love and loveliness: in praise of the dripping wet breast."
Journal of the Motherhood Initiative for Research and Community Involvement 4 (1).
Accessed September 9, 2013.
Griffith, James D, Sharon Mitchell, Christian L Hart, Lea T Adams and Lucy L Gu. 2013.
"Pornography actresses: an assessment of the damaged goods hypothesis." Journal
of Sex Research. doi: 10.1080/00224499.2012.719168.
Hald, Gert Martin. 2006. "Gender differences in pornography consumption among young
heterosexual Danish adults." Archives of Sexual Behavior 35 (5): 577-585. doi:
10.1007/s10508-006-9064-0.
Henderson, Margaret. 2013. "Pornography in the service of lesbians: the case of 'Wicked
Women' and 'Slit' magazines." Australasian Journal of Popular Culture 2 (2): 159-
182. doi: 10.1386/ajpc.2.2.159_1.
Jensen, Robert and Gail Dines. 1998. "The content of mass-marketed pornography." In
Pornography: the production and consumption of inequality, edited by Gail Dines,
Robert Jensen and Anne Russo, 65-100. New York and London: Routledge.
Lo, Ven-hwei and Ran Wei. 2002. "Third-person effect, gender and pornography on the
Internet." Journal of Broadcasting and Electronic Media 46 (1): 13-33. doi:
10.1207/s15506878jobem4601_2.
13

Lottes, Ilsa L. 1985. "The use of cluster analysis to determine belief patterns of sexual
attitudes." Journal of Sex research 21 (4): 405-421. doi:
10.1080/00224498509551278.
McKee, Alan. 1997. "Penetration and power." Media International Australia (84): 7-18.
McKee, Alan. 2001. "A cultural policy argument for government subsidies of gay porn
production in Australia." In Queer in the Twenty-First Century: The body - queer and
politic, edited by Keith J Atkinson and Justin J Finnerty, 118-137. Fortitude Valley,
Qld: Gay and Lesbian Welfare Association.
McKee, Alan. 2003. Textual Analysis: a beginner's guide. London: Sage.
McKee, Alan. 2007. Beautiful Things in Popular Culture. Malden, MA: Blackwell Publishing.
McKee, Alan. 2009. "Social scientists don't say titwank." Sexualities 12 (5): 629-646.
McKee, Alan, Kath Albury and Catharine Lumby. 2008. The Porn Report. Melbourne:
Melbourne University Press.
McLaughlin, Thomas. 1996. Streets Smarts and Critical Theory: Listening to the vernacular.
Madison: University of Wisconsin Press.
McLelland, Mark. 2007. "The best website for men who have sex with men:
cruisingforsex.com." In Beautiful Things in Popular Culture, edited by Alan McKee,
79-86. Malden, MA: Blackwell Publishing.
Mercer, Kobena. 1991. "Skin Head Sex Thing: racial difference and the homoerotic
imaginary." In How Do I Look? Queer film and video, edited by Bad Object-Choices,
169-210. Seattle, Washington: Bay Press.
Nikunen, Kaarina and Susanna Paasonen. 2007. "Porn star as brand: pornification and the
intermedia career of Rakel Liekki." Velvet Light Trap (59): 30-40.
Smith, Clarissa. 2002. "'They're ordinary people, not aliens from the planet Sex!': the
mundane excitements of pornography for women". In Journal of Mundane Behavior.
Ullen, Magnus. 2013. "The solipsism of pornography: speech act theory and the anti-porn
position." Sexuality & Culture 17: 3210347. doi: 10.1007/s12119-012-9154-1.
14