Untitled Document

Alison G. Smith (Ed.), The Relationship between Rhetoric and Terrorist Violence London: Routledge, 2013.

128 pp. ISBN 978-0-415-82360-9. £ 85.- (hbk)

Reviewed by Mark Dechesne

The systematic analysis of the content of political texts holds considerable promise when it comes to predicting terrorist violence. That is the basic premise behind The Relationship between Rhetoric and Terrorist Violence, edited by Allison Smith, a social scientist at the Science and Technology Directorate of the US Department of Homeland Security. The volume is a compilation of the approaches by some of the leading  experts in the area of Linguistic Content Analysis. In the volume, they discuss their strategies to analyse a single set of texts from two violent, and two radical but non-violent organisations, in order to identify potential predictors of violent action.

The idea for the volume arose from earlier research carried out by the editor. In her earlier work, Smith analyzed the political exclamations of 13 political organisations known to be violent, and compared these exclamations with those of matching radical but non-violent organisations. She coded the texts for in-group and out-group affiliation and power motive imagery, and found that these implicit motives differed between the violent and nonviolent political organisations.

In this volume, Smith has created a similar setup. Documents (such as speeches, interviews, and articles) from the violent Central o Core Al-Qaeda  and Al-Qaeda in the Arabian Peninsula (AQAP) are compared with the non-violent Hizb ut-Tahrir and the Movement of Islamic Reform in Arabia. For this volume, however, a number of leading experts were invited to report on their attempts to scrutinize the documents. They were asked to focus on two issues: 1) The use of Linguistic Content Analysis (LCA) to identify violent as opposed to non-violent political participation; and 2) the use of LCA to predict terrorist activity. The experts provide an interesting overview of the diversity of available LCA techniques, ranging from theory based versus data driven, and automated versus human coding.  

Despite this diversity, the basic principle of the LCA method is applied throughout the volume. The words and sentences of the political documents used are scanned and matched with indicator words that are derived from theory or prior empirical research. Jamie Pennebaker of the University of Texas, for example, reports on the match of the words in political documents with words derived from his earlier research on Linguistic Inquiry and Word Count (which includes categories with associated occurrences or co-occurrences of words, e.g.  the category “social-emotional style” is associated with the co-occurrence of personal pronouns and affective words). Sanfilippo, McFrath, and Whitney (Pacific Northwest National Laboratory) infuse automated Frame Analysis with theoretical insights from the social scientific literature on violence (regarding e.g. moral disengagement) and scan the documents for indicators that are derived from this literature. Hart and Lind (University of Texas) apply DICTION, a program containing ‘dictionaries’ of semantically clustered words. Hermann and Sakiev of the Moynihan Institute of Global Affairs at Syracuse University use Profiler Plus software to match the documents with criteria for seven traits that are thought to characterize political leaders. Stephan Walker (Arizona State University) uses the same software to match the words in the political documents with dictionaries he specified as part of the Verbs in Context System. David Winter (University of Michigan) takes a less formal approach with human coders who rated the texts for consistencies with his taxonomy of human social motives of Achievement, Affiliation, and Power. Lucian Gideon Conway III and colleagues (University of Montana) also use human coders to scan the documents for differences in elaborative and integrative complexity across parties and across violent and non-violent episodes. Finally, Peter Suedfeld and Jelena Brcic (University of British Columbia) trained two scorers to judge the extent to which the political documents reflect human values, including e.g. Self-Direction, Universalism, Benevolence, Conformity, and Security.

The reports on findings are brief but contain a vast amount of insightful results. This is no surprise given the size of the dataset and the diversity of variables employed in their examination of the documents. Several researchers aptly note that by chance alone a substantial number of significant differences between violent and non-violent organisations are to be expected. But there are certainly more differences reported in the reports than mere chance would prescribe. Fortunately, a concluding chapter by Conway and Conway summarizes several findings that hold across the various chapters and are therefore particularly noteworthy for our understanding of the difference between violent and non-violent political rhetoric. Across authors and methods, terrorist rhetoric is found to be of lesser complexity, to come with greater emphasis on affiliation, to stress issues of control and power, while remarkably, violent and non-violent organisations do not differ in their hostility against their adversaries, only in the methods they use to target them. Of further interest, expressions of achievement and optimism were in some analyses associated with greater propensity towards violence while in others with lesser violence. Predicting a terrorist attack turns out to be more difficult than differentiating violent from non-violent political expression. Conway and Conway cautiously conclude from their meta-analysis that a general exaggeration of a terrorist organisation’s typical style signals an impending attack. For example, if a violent political organisation already exhibited lesser complexity, this characteristic will become more salient just prior to an attack. 

In conclusion, this is a highly valuable academic treatment that provides an overview of the high potential of Linguistic Content Analysis for predicting likely terrorist attacks. Given its high potential, it is perhaps unfortunate that the contributors only briefly describe their research methods, and above all, that there is little general consideration of the value and limitations of LCA for terrorism analyses. Nonetheless, in an age where the potential of “Big Data” and  associated  algorithms are increasingly recognized for the predictive capability of a great variety of socially significant phenomena, readers will find this book a state-of-the-art overview of the methods and heuristics used to analyse large amounts of political data to identify potentially violent forms of political participation.

About the Reviewer: Dr. Mark Dechesne is Senior Researcher at the Centre for Regional Knowledge Development at Leiden University’s The Hague Campus. Prior to this appointment he was a researcher in the field of psychology with START at the University of Maryland.

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.


Perspectives on Terrorism is  a journal of the Terrorism Research Initiative and the Center for Terrorism and Security Studies

ISSN  2334-3745 (Online)

Disclaimer, Terms and Conditions