Rape, Attack, Abuse, Assault: Representations of Women in the Far-Right using Corpus Linguistics

edl word cloud


In this post, I’m using a corpus of postings on Facebook by supporters of the EDL to study how women are constructed in far-right discourse. I’m using Sketch Engine, and in particular Word Sketch, a function which provides detailed collocational information of search words.

women and children

This post focuses on the words woman and women. When I looked at collocates which are the object of woman / women, the following table was obtained:

object of women


The collocates appear to suggest a high degree of negativity associated with women, a great deal of violence and oppression as the collocates include: rape, treat, attack, stone, cover abuse, assault, oppress, hate, marry, insult, see. Other collocates not shown above include force and beat, which further empathize the discourse prosody of violence, oppression and victimization which is associated with women in discourses produced by members and supporters of far-right ideology

In this post, I’ll  focus on the concordance lines of the more frequent collocates.







The concordance lines present a certain amount of context to the collocational information. In the first set of concordance lines there appears to be a discussion as to whether white men rape women, or whether this can be an issue for which Muslim men are solely responsible. The notion that women are victims appears to be accepted. The second set of concordance lines construct women as being treat badly by Asian men, and the third set depicts women as being defenseless and victims of physical attack.


muslim women


Word Sketch can also provide information of the modifiers of search words. The most common modifiers of women are as follows:

modifier of women

white women 2

muslim women 2

Both white women and Muslim women are constructed as victims. White women are constructed as victims of rape, whereas Muslim women appear to be constructed as the victims of oppression by Muslim men.

The depiction of women as victims, primarily of violence by Muslim men and Muslim culture, positions male supporters of the EDL as defenders of women and consequently of values and traditions of English culture.




Comparing EDL Discourses Using Word Clouds

For this post I’ve used an online visualization software called ‘TagCrowd’ available at http://tagcrowd.com/


I’m interested in how texts created by EDL leaders differ from those of members who post on their Facebook page and used  a visualization software to help with this.

The first image is the word cloud created from a corpus of EDL texts.

edl word cloud

The second image was created using texts from EDL supports who posted on the groups Facebook page.

edl-fb word cloud


By comparing and contrasting these two images, it is possible to observe discourse topics, although not what is being said about the particular themes. It appears that both corpora contain discourses related to islam, muslims, edl, british, halal country and government. There appears to be differences also. Certain words caught my attention among the many differences between the two images. The EDL corpus contains discourses related to topics such as extremism, radical, sharia, terrorism and community. Such words were not significant in the Facebook corpus. However, the Facebook corpus contained discourses which were not prominent in  the EDL corpus. These include racist, hate, white, grooming, scum, kids, girls children, women and sex.

Although it is not possible to gain conclusive evidence of the discourses present in the corpora, it does appear that the EDL hierarchy are more concerned with issues related to Islamic extremism and terrorism. The corpus of Facebook text contains discourses related to family, race, and conflict (abuse, bad, grooming, hate, idiot, kill, racist, rape, scum).

Although the text contain fundamental similarities, there are also significant differences.