Monthly Archives: January 2019

Fairclough’s Critical Discourse Analaysis: A practical discussion

I’ve received enough questions from graduate students about how to apply critical discourse analysis to examining a particular text or set of texts that I’ve decided it’s time to write these thoughts down in one place so I can direct students here rather than endlessly repeat myself in emails.

One option for doing critical discourse analysis is to apply Fairclough’s (1992a, 1992b, 1995) model, which he refers to as a “three-dimensional model” (1992b) encompassing “analysis of context, analysis of processes of text production and interpretation, analysis of text” (p. 213).

For better or worse, many students find Fairclough rather dense reading, so here I would like to break what is rather complex reasoning down into what I hope is simpler language. Basically, the way I approach Fairclough’s categories is that they are concerned with what people say (the textual), how they say it (the social/context), and the implications of what they say (the practice/processes of text production). One of the reasons Fairclough’s writing is so dense is that these are all interrelated, and so he goes to pains to note that they can’t really be dealt with in isolation, but that rather each has to be considered in light of how it is influenced by and influences the other dimensions of analysis.

I find that talking about these different dimensions of critical discourse analysis in the abstract leads to an endless rabbit hole of speculative thinking. So, rather than going there, I would prefer to try to put them into practice with some examples. A common pastime of CDA and graduate students is analysis of the news, so below I’ve presented the headlines featured on the Wikipedia pages (as of the time of writing) for three UK-based newspapers, two tabloids and one broadsheet:



The Guardian: Students drafted in to plug NHS gaps (

For those wondering (like I was), according to Wikipedia, broadsheet refers to a newspaper printed in the largest available format, while tabloid refers to a newspaper printed in a smaller format. For students writing master’s level assignments, I would generally recommend you seek out a more academic reference than Wikipedia; one of the benefits of blogging is that I don’t have to write to the academic standard expected of other kinds of work, like postgraduate assignments and manuscripts for publication. A good place to start for finding an academic reference concerning broadsheet and tabloid newspapers would be a Google Scholar search for tabloid and broadsheet.

There are several things that can be said about the textual aspects of these headlines, but one striking difference is that the tabloids use all caps while The Guardian uses sentence caps. That’s a textual observation. The social implications of this is that the tabloid headlines can be interpreted as shouting, a marked form of written language, while The Guardian is using an unmarked form of presentation for its headlines. That marked/unmarked shouting as all caps and normal conversation as sentence caps is another observation that in an assignment or paper written for publication would likely require an academic citation to back up the claim. Blake (1999) might be a good go-to here. These are my Google Scholar search results.

There’s more that can be said about font size, page placement, relative size on the page, etc., but I think most of those characteristics would point toward the basic conclusion I’ve already drawn; Tabloids appear to be shouting their headlines while The Guardian (or, the broadsheet newspaper) is more engaged in a kind of unmarked conversation. Although this is a hypothesis that would need to be tested through further analysis in a postgraduate assignment that would hopefully include analysis of the actual articles in addition to their headlines.

So what are the implications of this, or the discourse practice insights that can be drawn from this observation? Well, please don’t consider this a complete analysis, but I think one conclusion that can be drawn is that the tabloids are seeking to sensationalize the news, perhaps because they rely on newsstand sales for their income more heavily than a broadsheet newspaper (a hypothesis that could perhaps be probed by visiting the newspapers’ websites and looking at their reporting of their subscription rates). Sensational headlines may tempt people to buy the tabloid newspaper to read about whatever headline happens to make them the most interested or upset. Another possible implication is that the tabloids view news as something to anger and upset readers, while the broadsheet is oriented more toward informing readers. Implications here could also be to push newsstand sales. There is also the implication of the evil other as an archetype in the tabloids; the ‘Germans’ (a foreign other) and ‘mental patients’ (people who are not ‘normal’) are portrayed as doing violence to a shared assumption regarding a ‘normal’/unmarked ‘English’ (likely ‘white’) way of life, and these tabloids are both using this way of seeing the world to appeal to their readership and sell newspapers and are also perpetuating this way of seeing the world by portraying it as the ‘unmarked’/preferable way of describing/understanding the current social reality. In other words, the writer(s) and the reader(s) of the tabloids see their normative way of life as the preferable/only ‘correct’ way of life, and the stories told through the tabloids help to prop up that way of seeing the world and reassure the people who believe that way that it’s the ‘correct’ way to believe.

I doubt these are all new insights, and so in a postgraduate assignment, the author’s job would be to see what’s been written before about broadsheet versus tabloid newspapers and to consider what from the existing literature their analysis could help to reinforce and what insights their analysis could add.

That these categories are interrelated comes through when considering the headlines being written in other languages. Japanese doesn’t have a textual equivalent to all caps; that’s a phenomena of the English language, and so that textual aspect of the language is intimately interlinked with its social implications. Japanese can of course do textual shouting, but it tends to be through visual/formatting changes, such as underlining or boldface font. Similarly, it would be preposterous to think of Der Spiegel printing the headline GERMANS DECLARE WAR ON OUR £; that helps to show how the normative assumptions prevalent in the UK are textually realized in the language used; ‘our £’ is both a textual decision to represent the UK currency and also an expression of common UK national identity.

A caveat here for those interested in doing something new in the field is that analysis of news discourse has largely been ‘done’. If you really are interested in investigating something new, then considering how the issues discussed here come up in production of, for example, memes or other emerging online forms of discourse would perhaps point you in the direction of newer territory. In the production and dissemination of a meme, there are much larger questions at play; while news sources, both tabloids and newspapers, report who produced the text, and the how is generally understood, something like the following meme, captured on Facebook, brings a lot of ambiguity with it as to who produced it, for whom, and for what purposes:

Example of a meme captured from Facebook

There is a lot going on textually in the meme regarding assumptions of shared identity and ways of seeing the world that I don’t really feel like getting into here. It also reads starkly similarly to the tabloid headings above. What I would like to discuss here is an issue inherent in the way Facebook works and that I feel is of particular importance in any attempt at analyzing it. Namely, who is behind the ‘Be My Voice’ page is inherently unknowable (unless you have the broad behind-the-scenes access to IP addresses and accounts that Facebook does). It could be based in any number of countries, and the purpose behind its existence could be quite different from the reasons why individual Facebook users choose to share the images posted to its page; this has emerged rather broadly as an issue with regard to election coverage of the 2016 US presidential election, but is a pervasive feature of internet discourse more generally. Furthermore, in an AI curated, individual interest-based feed space such as Facebook, who sees and doesn’t see the meme, or who its audience is, is not a straightforward issue to analyze. While it wouldn’t be an easy task, seeking to tackle these kinds of issues and questions in an assignment would certainly be contributing to knowledge in the sense of charting new territory in applied linguistic analysis.


Blake, G. (1999). Managers at work: Email with feeling. Research Technology Management, 42(6), 12–13. Retrieved from

Fairclough, N. (1992a). Discourse and Social Change. Cambridge: Polity Press.

Fairclough, N. (1992b). Discourse and text: Linguistic and intertextual analysis within discourse analysis. Discourse & Society, 3(2), 193–217.

Fairclough, N. (1995). Critical Discourse Analysis. London: Longman.