Content uploaded by Alon Lischinsky
Author content
All content in this area was uploaded by Alon Lischinsky on Aug 28, 2017
Content may be subject to copyright.
Distant reading intimate encounters:
a big data approach to online erotica
Dr Alon Lischinsky and Dr Kat Gupta
Oxford Brookes University, UK
alischinsky@brookes.ac.uk || kat@mixosaurus.co.uk
@alischinsky || @mixosaurus
Why pornography?
•“Prolonged consumption of pornography distorts
perceptions of sexuality […] breeds discontent with the
physical appearance and the sexual performance of
intimate partners [and] increases men's propensity for
committing rape.”
Zillmann (1986:26–7)
•“Pornography […] portrays stereotypical views of women
as passive."
Rosen & Christman (1997:210)
Why pornography?
•“Pornography is valuable. It sends out messages of comfort
and rebellion. It says: Lust is not evil. The body is not hateful.
Physical pleasure is a joyful thing and should not be hidden
or denied.”
Califia (2000:52)
•“One of the most common informal sources of sexual
information was pornography in its various forms.
Pornography was usually described as one of the only
resources available.”
Kubicek et al. (2010:250)
Why written pornography?
•“The visceral energies directed ‘at’ pornography
by social forces concentrate on its visual
aspects […] While pornographic books and
magazines are also mass-produced and
consumed, they are less visible as targets of
critique or analysis because the prevailing figure
for pornographic consumption lies so squarely in
the enthralled spectatorship of the eye.”
Wicke (1995:75)
Current research into pornography
•Historical, literary and aesthetic
perspectives on the development of genres
•Gay, lesbian and queer pornography
•Transgressive pornography (c.f. Porn
Studies (2017) special edition on gonzo
porn)
Literotica
•Free site aimed at amateur writers
•Largely image-free
•Stories are submitted through a form or
email and are reviewed before posting
–Process takes 3-7 days
–Authors choose the category most suitable for
their story
Discourse
•Discourse as reflecting/constructing a
social reality
•This gets complicated with pornography
Discourse
•Discourse as reflecting/constructing a social reality
•This gets complicated with pornography
•Complex relationship between imagined acts, what
may be fantasised about but not actually wanted in
real life, the writer and reader's sexual experience,
the role of acting, socially salient sex acts…
Fantasy and reality
•Awareness of legal restrictions and
realities
–Does not allow stories involving bestiality,
mutilation, snuff and underage sexual
encounters
Fantasy and reality
•Awareness of legal restrictions and
realities
–Does not allow stories involving bestiality,
mutilation, snuff and underage sexual
encounters
•Stories obviously based in fantasy
–Encounters with werewolves, vampires,
demons and aliens
Fantasy and reality
•Awareness of legal restrictions and realities
–Does not allow stories involving bestiality,
mutilation, snuff and underage sexual encounters
•Stories obviously based in fantasy
–Encounters with werewolves, vampires, demons
and aliens
•Unrealistic depictions of sex
Literotica corpus
•31 genres
–anal, BDSM, celebrities, chain stories, erotic
couplings, erotic horror, exhibitionist & voyeur,
fetish, first time, gay male, group sex, how to,
humor & satire, illustrated, incest/taboo, interracial
love, lesbian sex, letters & transcripts, loving wives,
mature, mind control, non-erotic,
nonconsent/reluctance, nonhuman, novels &
novellas, reviews & essays, romance, sci-fi &
fantasy, text with audio, toys & masturbation,
transsexuals & crossdressers
Literotica corpus
•31 genres
•Top 1000 stories for each genre (incl.
multi-chaptered stories)
•1,385,891,072 tokens
Keywords
•By genre
•Reference corpora
–BNC
–All of literotica
•Top 50 (or fewer) keywords
Keywords
•Potentially 1,550 unique keywords
Reference corpus No. unique keywords
Literotica genres and BNC-written
Number of
genres
No. of
keywords
Keywords
!"
" ##$%&#&&#
#%!
' ( )%##
* ##&+&&%!
' """ %%# #
* ( % ### ! !&,
##!!!
' ( )%" %%%%
###" !+
" -&&&,
###!!!#
!!!#
#
"%%####!"
+&%%
&&&&# +
+ %#+!
!"!"!!+
Literotica genres and BNC-written
Number of
genres
No. of
keywords
Keywords
I
!"
" ##$%&#&&she
#%!
' ( )%##
* ##&+&&%!
' """ %%# #
* ( % ### ! !&,
##!!!
' ( )%" %%%%
###" !+
" -&&&,
###!!!#
!!!#
#
"%%#he##!"
+&%%
&&&&# +
+ %#+!
!"!"!!+
Pronouns
Pronouns
•Transitive uses of fuck take female
patients almost three times as frequently
as male ones
•A majority of the instances of transitive
fuck with female agents corresponds to
uses of insertable sex toys
Lischinsky (2017)
Literotica genres & Literotica-all
Number of genres Number of keywords
31 1 (user error?)
5-9 83
2-4 419
1 612
metadiscourse
metadiscourse
•How to and Reviews & Essays
•Characterised by words about writing, reading
and characterisation e.g. characters, readers,
stories, writer
•Overlaps with the genre of Letters &
Transcripts due to use of second person
pronouns and direct address (your, yourself)
non-porn components
non-porn components
•Celebrities, Chain stories, Erotic horror,
Fantasy, Non-erotic, Nonhuman, Romance
•Characterised by greater attention to setting,
description and (sometimes) fighting
•Supernatural elements (demon, wolf, vampire)
•Historical fantasy (castle, forest, lord,
princess)
distinctiveness
distinctiveness
•Anal and Toys & masturbation
•Limited overlap with each other (beads,
jelly) but for the most part, do not have
shared keywords
•Describing sex acts that do not involve
other people and/or non-PIV sex
control and power(?)
control & power(?)
•BDSM, Exhibitionist, Fetish, Nonconsent-
Reluctance, Transsexual-Crossdressers
•BDSM, Exhibitionist and Fetish: humiliation,
humiliated
•BDSM and Nonconsent-Reluctance: cuffs, gag,
restraints, wrists, tied
•BDSM and Fetish: bondage, obey, spank
•Fetish and Transsexual-Crossdressers: heeled,
latex, panty
control & power(?)
•Some spurious connections e.g. sir is
shared by Letters & Transcripts and
BDSM
–signalling power and/or distance?
What comes next…
•Written pornography is not homogenous
What comes next…
•Written pornography is not homogenous
•Attention to stylistic detail and
distinctiveness
wolves
scary wolves…
wolves
scary wolves… …or sexy wolves?
What comes next…
•Written pornography is not homogenous
•Attention to stylistic detail and distinctiveness
•Do the same categories emerge with
unsupervised categorisation?
–MDA
–Cluster analysis
–Topic modelling
References
•Blondel, V. D., Guillaume, J. L., Lambiotte, R., & Lefebvre, E. (2008). Fast
unfolding of communities in large networks. Journal of Statistical
Mechanisms. 2008.
•Califia, P. (2000). Public Sex: The Culture of Radical Sex (2nd ed.) San
Francisco: Cleis.
•Kubicek, K., Beyer, G., Iverson, E., Kipke, M. (2010). In the Dark: Young
Men’s Stories of Sexual Initiation in the Absence of Relevant Sexual Health
Information. Health Education & Behavior, 37(2), pp. 243-263.
•Lischinsky, A. (2017). Doing the naughty or having it done to you? Agent roles
in erotic writing. Porn Studies
•Rosen, D. & Christman, J. (1997). Toward a New Model of Sexuality. In:
Soble, K. (Ed). Sex, Love and Friendship (pp. 199-214). Amsterdam: Rodopi
•Zillmann, D. (1986) Effects of Prolonged Consumption of Pornography. Paper
written for the Surgeon General's Workshop on Pornography and Public
Health. Arlington, VA.