Moral panic stifles useful dialogue on social media “trolling”

Today’s Guest Hoyden Dr Peter John Chen is a Lecturer in the Department of Government and International Relations at the University of Sydney. You can follow him on Twitter.

Content note: discussion of sexist and misogynist slurs,
includes quoted examples.

Current discussion in the mainstream press of social media trolling has considerable relevance for the conduct of Australian public discourse. Since before the election of Julia Gillard in 2010 I have observed a tendency for a very gendered treatment of the Prime Minister that goes beyond the natural interest in the first elected female leader of this country.

In my analysis of the 2010 election (in Simms and Warhurst, 2011) I observed that the media tended to focus on her having special responsibilities for maintaining and repairing relationships with other political figures (such as Kevin Rudd and Tony Abbott) that no male politician would be expected to do. From that point on there has been a marked decline in the discursive treatment of the PM about which Hoydenizens are very well aware.

Thus the recent “discovery”of the anti-social aspects of the social media by the commercial news organisations did appear to have some promise, but if its fixation on celebrity scorn and encouragement of tighter government regulation of social media has the hallmarks of a classic moral panic that’s likely to move through the issue-attention cycle quickly and take the oxygen out of a more useful dialogue about civil discourse.

This is unfortunate because the pseudonymitypseudo-anonymity offered by social media does reduce some of the social controls on public behaviour around civility and the notion that reciprocal respect is necessary for meaningful dialogue. If the debate about online language (and the wider discourse about gender and public life) is to have value, we need to take stock of the extent and nature of the behaviour concerned. This will allow any corrective or regulatory discussions to be a little more grounded than they have been to-date.

From work I’ve been doing for a forthcoming book on new media and Australian politics, I have some useful data that may partially inform this discussion in the form of Facebook wallposts from 600 Australians collected before this recent debate took off (late 2011). In recent days I’ve reanalysed this dataset to shed some light on the treatment of women in the social media space.


The research was a simple content analysis of the 600 users’ wall posts. On average each user had 24 posts captured (for a total of 14,069 posts). Stratified sampling based on locality was employed to match the population and gender distribution.

In the coding frame two key concepts were coded: Misogyny, drawing upon Armstrong’s operationalised definition (2001) to include references to assault, rape and murder; and sexism (“any phrase that may be interpreted to be treating men and women differently, simply on the basis of their sex”) based on the work of Boxill, et al to include references to social, workplace/professional and sexual instrumentalism (1997: 114-5). In addition, misandry was included in the coding frame.

Given the nature of language, even in the comparatively simple communicative environment of Facebook wallposts, the ability to definitively identify and recognise these three concepts is limited. A conservative approach to coding was taken where the reference appeared unambiguous in nature. A different coder would have made different decisions (the frequent use of “cunts” and “bitches” to describe men by men was not included, but it could be argued that this use of derogatory slang to denigrate men through comparison with women/women’s genitalia is an endemic form of sexism).


The findings are interesting, but possibly not surprising.

At the most extreme end of the spectrum, there were very, very few misogynistic comments identified (0.021% of posts, all from men). An example of this would be the group “cutting your mum’s car breaks[sic] when she cooks a bad dinner”.

The vast majority of identified content was in the form of sexist posts. These took two forms. The first (0.092% of all posts) were sexist statements about women’s social position (an example of which being “babe I promise to listen to you if you shut your piehole, permanently”). The second, and far more prevalent was statements about women as sexual objects or their sexual instrumentalisation (0.391% of all posts). These latter posts ranged from the classification of women as sexually available (sluts) to more graphic expressions (i.e. “The thing I love most is cumming on your face, suck it bitch!”)

While the quantum of posts of this nature is comparatively small overall, it is quite marked that almost 11% of all posters in the sample group made one or more of these posts in the sample period. While men where the most likely to post this content (13.5%), women were also well presented (8%) as regulators of the gender.

The role that social media plays in this is significant, beyond simply the pseudonymity offered by being behind a computer screen. Importantly the majority of sexist statements are not personal utterances, but come from the membership of Facebook groups. This shows the role of social proof (support for individual bahaviour assumed from others’ participation in the same) in driving along these types of sexist attitudes: very, very few of the posts were direct articulations by the sample group members but liking or joining groups with sexist titles and/or objectives.

Interestingly, in comparison to the recent controversy around broadcaster Alan Jones declaring that women are “destroying the joint”, there were no sexist statements in the sample that focused on the professional competence of women in general terms. Here we see that advances in the professional standing and workplace participation of women appear to be at odds with the pornification of society in the saw sexist attitudes are expressed.

Proving its a man’s world, misandry was not seen in the sample (though, stretching the definition to it’s widest, one post may have qualified).

How does this compare with other “hate” online? In the sample I also coded racists statements to see the extent to which we see casual racism online. Here Australia compares comparatively well, particularly given the recent history of moral outrage against asylum seekers and a general back-pedalling on multiculuralism. Only 0.037% of posts contained casual racism (negligibly higher from men than women), with less than two percent of the 600 people sampled engaging in this types of statement in the sample group.

References cited

Armstrong, Edward, 2001, “Gangsta misogyny: A content analysis of the portrayals of violence against women in rap music, 1987-1993”, Journal of Criminal Justice and Popular Culture, 8(2): 96-126.

Boxill, Ian, Claudia Chambers, Eleanor Wint, 1997, Introduction to Social Research: With Applications to the Caribbean, Kingston: Canoe Press.

Chen, Peter John, 2011,”The new media and the campaign”, in M Simms and J Warhurst (eds), Julia 2010: The Caretaker Election, Canberra: ANU e-press.

Categories: culture wars, ethics & philosophy, gender & feminism, parties and factions

Tags: , , , , ,

7 replies

  1. The findings are interesting, but possibly not surprising.

    I suspect that’s why there’s been so few comments on your post, Peter. None of us are surprised, although it’s always useful to see one’s gut perceptions confirmed by actual data.
    There was nothing especially surprising in Media Watch’s more general take on the Twitter Troll Panic of 2012 either, but that’s also worth seeing (or reading the transcript) to get some actual facts – celebrities being abused by anonymous haters on twitter is contemptible, but it’s a very small fraction of the much larger cyberbullying problem which mostly affects teens at the hands of people they go to school with…

  2. I’ve been busy and not had time to read much online.
    My biggest issue is what appears to be misuse of the term “pseudonymity” in the analysis. I didn’t think it was a difficult concept, and I’m afraid that I’m judging it as either lack of care or familiarity with common issues of online behaviour.

    This is unfortunate because the pseudonymity offered by social media does reduce some of the social controls on public behaviour around civility and the notion that reciprocal respect is necessary for meaningful dialogue.

    The role that social media plays in this is significant, beyond simply the pseudonymity offered by being behind a computer screen.

    It’s particularly weird seeing it in a discussion of Facebook, which has a policy of no pseudonymity (which is not particularly enforced AFAICT).

  3. Good story from Julie Posetti, includes some commentary from Wendy Bacon:

    Important Media Shift story on #twitter, #trolls & handing over account info to police by Posetti
    — Wendy Bacon (@Wendy_Bacon) September 22, 2012

  4. @aqua is correct, it should have read “pseudo-anonymity”.

  5. Thanks!
    One thing I’ve noticed online is when someone enters a group (shared online social space) and behaves as though it’s a pseudo-anonymous space when the “locals” think it’s a space of people known to each other with expected behaviour standards. It was really common on USENET, but you still see it in blog comment sections. It’s much rarer in “real life”, people seem better at picking up the cues about which rules apply.

  6. Re.: #6
    I think that’s right – I guess there’s a question of how “static” the space is. This makes some places (like a G+ circle or a FB group) probably more likely to develop rules than say a temporary hashtag community on twitter. There’s an interesting analogy about interpersonal incivility in the real world and places of high mobility (like transport interchanges).
    See: Phillips, Tim and Philip Smith, 2003, “Everyday incivility: Towards a benchmark”, The Sociological Review, 51(1): 85-107.

%d bloggers like this: