Discourse Analysis and Media Attitudes: The Representation of Islam in British Press. Paul Baker, Costas Gabrielatos, and Tom McEnery. Cambridge, UK: Cambridge University Press, 2013. 294 pp. $99 hbk. $79 ebk.DOI: 10.1177/1077699015569232fSince Edward Said launched ship of post-colonial studies with his touchstone study Orientalism and its accessible sidekick Covering Islam, it has become a scholarly commonplace that media represent Muslims negatively. The popularization of term Islamophobia in mid- to late- 1990s and attacks of September 11 have provided plenty of news copy for scholars to examine, and representations they discover are often distorted, monolithic, and preoccupied with themes of conflict. So frequently has this result been upheld that spurious representations of Muslims in media have acquired status of a given.The team from Lancashire, England, that put this book together have done field a service in not treating subject as a given but subjecting it to exhaustive and rigorous scrutiny. The trio-two are professors at University of Lancaster, and third lectures at Edge Hill University-use corpus-driven linguistic analysis to examine how British press talks about Muslims and Islam. They analyze these data at a macro level, querying words most typically linked with their topic and assessing how those words and their context change over time. Guided by results, they ask more specific questions of data, teasing out language of reporting on belief, gender, collectivization, and (particularly salient for U.K. tabloids) receipt of government benefits.Consonant with established line from literature, authors conclude that association of Islam with conflict looked suspiciously high. They offer a few prescriptions, including an invitation for better reflection as to whether religious orientation of people in stories is actually salient to reporting.What sets this discourse analysis apart from its predecessors is its completeness: The authors' sample is not a selection of newspapers, accommodating political stance of outlet (left-leaning, right-leaning), format (tabloid, broadsheet), or a narrow period; neither, having aggregated all articles that mention topic, do they thin sample by random or purposive selection. Their corpus includes all major daily newspapers and all articles published therein that refer to Islam and Muslims between 1998 and 2009. Although authors allow that some articles may have slipped through their net, corpus they compiled contains 200,037 articles.That weight gives them authority to make distinctions they do-for example, that left-leaning broadsheets contain the most balanced reporting on Muslims, or that tabloids associate Islam and Muslims with terror far more than broadsheets. Again, these seem like truths we already know, but numbers and analysis give concrete shape to impressionistic judgments: In their corpus, tabloids used terror 22,188 times, compared with 22,032 in broadsheets. …