Using Atlas-ti for Textually Oriented (Critical) Discourse Analysis Part 1 of 3: Atlas-ti as an analytical writing space

Screen Shot 2017-08-04 at 12.15.11I’ve always liked the idea of using computer assisted qualitative data analysis software (CAQDAS) – such as Atlas-ti – when doing the textual analysis part of a CDA project but in reality I would often became frustrated with it. Now, though, I think I’ve finally found a way of using Atlas-ti which suits (at least my way of) doing textually oriented discourse analysis.

To use this kind of software, one loads documents (word, documents, pdfs – and now sound files, picture files, and videos) in to the software and then ‘code’ sections of them – highlight a piece of text and give it a name: social actor, topos of authority, LIFE IS A JOURNEY metaphor, for example.

When it came to actually writing up my analysis and interpretation of the texts, though, I would find that I ended up putting relevant quotations from my data into a word document and, pretty much, re-analysing and interpreting these segments of data again as part of the writing process – it turns out that writing is part of the analytical process for me – and it seemed quicker to dispense with the CAQDAS and just cut straight to the analytical writing stage.

What hadn’t ‘clicked’ with me was that the software can be used as a ‘space’ for writing as well as coding and managing your data – this is probably taken for granted by many experienced users of CAQDAS but is something I had entirely missed.

The key to this is using memos. Atlas-ti includes the facility to write memos which can be linked to each other and to quotations you’ve identified and to codes that you have used. These memos can be a spaces for writing the analysis, writing about concepts, and importantly for me, keeping a research diary. Even better, PDF files of journal articles can also be loaded into a project and this academic literature can also be coded – or just referred to – whilst in the analytical process.

Why is this important? It may not be for everyone, but I find that being able to see my text, and quotations, as well as being able to quickly call upon literature that I know I’ll refer to as part of the final article – all side by side – is really useful.

For me, the apparently trivial act of switching out of the software in which I’m marking up my text, looking for another file, going into my reference library for some academic literature or making a note of how I want to express something breaks me out of that analytical zone. Holding all of these various aspect in one software ‘space’, is far less disruptive than keeping them as separate files that I have to access with different software apps and – in the end – using Atlas-ti as a complete analytical writing space is becoming more productive for me.

What is CDA? 1: Theory

In a previous post I wrote about CDA as a method, but here I want to address theory: CDA is a theory as well as a method. The importance of theory is that it informs the analytical, interpretive and explanatory/theory-building method of CDA.

A prime theoretical distinction in this theory is between text and discourse.

In CDA, the language used in a social event is a ‘text’ – either spoken or written. Social events are particular interactions between people, and these interactions can be between just two individuals or tens of thousands.

Defining features of events are that they are finite and unique: they occur, then end, never to be repeated in in quite the same way again.

Language-use can also have properties which appear and reappear in social practices. Social practices are patterns and conventions of behaviour which participants produce, reproduce and, sometimes, transform across social events in different places and at different times. In CDA, the language patterns and conventions associated with a social practice is ‘discourse’.

This distinction gives users of CDA a way of thinking about and interpreting language in use. It gives rise to analytical questions over the extent to which the properties and features of a text are unique to the event in which the text has been produced. Is the feature beyond the familiar patterns and conventions of the social practices within which the event seems to be taking place? Are the features of the text entirely within the familiar patterns and conventions of the social practice? Or, more likely, is there an interplay of features being beyond and within convention.

Using this distinction, we can begin to interpret texts. Is the speaker or writer using the conventions expertly and to the advantage of some social actor or other? Alternatively, have they been unable to overcome language convention and fallen into following convention to the disadvantage of themselves or those they represent?

The theoretical distinction between text and discourse – not unique to CDA, but crucial to it nonetheless – is an important basis for the analytical and interpretive method of CDA.

Analysis of Intertextuality: CDA and Nudge

Reframing obesity: a critical discourse analysis of the UK’s first social marketing campaign

Critical Policy Studies, 2016

Critical Policy Studies recently published a very interesting and important new article by Jane Mulderrig. The article is a critique of the UK government’s ‘Change4Life’ anti obesity social marketing campaign. The article shows how ‘scientific claims about obesity are recontextualized, simplified, and distorted in this campaign’.

The interest for CDA:Method, though, is the use of CDA and especially the article’s focus on:

patterns of intertextuality, legitimation, and representation, I investigate how this advert recontextualizes and simplifies particular understandings of obesity, presenting individualized solutions to what I argue is a complex and collective social problem.

Its great to see an analysis of a policy implementation using CDA which also sets out to include the analysis of intertextuality as part of the analysis. As an example, Mulderrig shows how an advert draws on scientific articles and reworks them in the context of the social marketing campaign:

This advert launches the campaign by defining the nature of the policy problem it aims to address. It does so through a historical narrative that essentially locates the source of the problem in modern consumer lifestyles. It presents to the public for the first time some of the key assumptions underpinning the C4L strategy about the causes, health risks, and solutions to societal obesity. The intertextual origins of these health claims can be traced to a series of policy texts and scientific reports.

Ultimately, the intertextuality does not serve us well:

the abstract, child-like cartoon genre of the advert facilitates a simplification and distortion of scientific research on obesity and helps obfuscate even further the complex environmental and political economic causes of obesity, and in particular, its correlation with increasing social inequality.

But it also raises questions about the manner in which the practice of social marketing appears to be being articulate with the practice of scientific research – simplification may be acceptable, but distortion is worrying.

The use of Order of Discourse in CDA

‘Order of Discourse’ is a key concept for CDA. It is, according to Fairclough ‘a particular combination or configuration of genres, discourses and styles which constitutes the discourse aspect of a network of social practices’ (2001: 220) and is important because it is ‘the social resource which enables and constrains interaction’ (Chouliaraki and Fairclough, 1999: 63).

However, a search of Critical Discourse Studies for articles that use the concept turns up just 18 results (listed below). Why might this be?

A closer look shows that even for these articles the concept is one that is generally seen as a given rather than as something that has been taken as an object of study.

Could we take ‘order of discourse’ as an object of study and analysis? The kind of question I have in mind is ‘what is the order of discourse for practices of UK policy making?’ To ask this sort of question is to come at discourse from a direction which is different to that ordinarily taken in studies which undertake CDA. The usual practice is to analyse texts for the genres, discourses and styles which are worked together in that text and to extrapolate from text to order of discourse. If we find legal, political, literary language in a text then we say that those social practices are being worked together in the text. We might criticise the fact that an educational text includes language from a consumerist discourse and extrapolate that practices of education are being colonised by practices of marketisation. This kind of approach is good for identifying change, but it may also hold an assumption that the order of discourse and practice that is being colonised was OK to begin with. What if the problem with the practice in which we are interested is of a more long-standing nature than contemporary colonisation?

Berry (2004)(

Luke (2004) Owen (2004)(

McKenna (2004)(

O’regan (2006)(

Maingueneau & O’regan (2006)(

Fernback (2007)(

Vannini & Mccright (2007)(

Hart (2008)(

Lindgren (2009)(

Prinsloo (2009)(

Wodak & Fairclough (2010) (

Montesano Montessori (2011)(

Gunders (2012)(

Millar (2013)(

Morrish & Sauntson (2013)(

Polyzou (2015)(

Shoshana (2016)(

Zhu, Ren & Han (2016)(

Review of Discourse and Democracy

9780415872355Nice review of my book, Discourse and Democracy in Critical Policy Studies:

“In order to achieve its critical and emancipatory goals, CDA explicitly engages with other disciplines, bringing theoretical concepts from relevant fields into transdisciplinary dialogue with linguistic concepts to support a critical exploration of the research problem. Discourse and Democracy offers an excellent, book-length example of this approach which will be of particular interest to scholars seeking to operationalize detailed text analysis in critical policy research”.

CDA as Method

This blog is about method in CDA. This might not seem controversial – CDA studies do, after all, employ method, PhDs, books and articles include methodological discussion and there are numerous textbooks which describe method in CDA.

But there has been a trend recently – seen both at CDA meetings and conferences and in published articles – in which CDA is described as ‘an approach’ or a ‘movement’, and these descriptions are paired with a specific denial of CDA as method. This description of CDA as a ‘movement’ was used by some contributors to the CDA 20+ meeting a couple of years ago in Amsterdam and, in to pick up one example in print, Baker et al., say this:

We understand CDA to be an academic movement, a way of doing discourse analysis from a critical perspective, which often focuses on theoretical concepts such as power, ideology and domination. We do not view CDA as being a method nor are specific methods solely associated with it.

In my view, this position is a precarious one to take. Indeed, the contradiction in the quote above illustrates this, for what is ‘method’ if not ‘a way of doing…’?

The bulk of academic research follows a very general process: selecting and establishing some background knowledge on the topic (usually involving a review of literature), identifying what you want to do (rationale), and adopting procedures for doing it (specific method).

At this very general level the CDA ‘method’ is to pick a topic which involves a social problem. As Fairclough, Mulderrig and Wodak have written, ‘CDA addresses social problems’. The general CDA ‘method’ is to adopt a rationale in which discourse is seen as part of the problem and discourse analysis as a way of addressing the problem through interpretation and explanatory critique.

But it is at the level of adopting procedures for doing text analysis that the denial of a CDA method is being levelled. To return to Baker et al., the quote above continues:

Instead, it [CDA] adopts any method that is adequate to realize the aims of specific CDA-inspired research.

Here, I suspect that ‘method’ is being used to refer to a number of things including methods of data collection, methods of data analysis, and methods of interpretation, explanation and theory building.

I suspect, too, that individual studies rarely neglect to adopt method in whichever of these categories is relevant to the study and that the broader question is whether or not there is a recognisable ‘CDA method’ in data collection, analysis, interpretation, explanation and theory building across the body of work in CDA. My inclination is to hold that there is. Research relies on method. If one is going to do CDA then one must view CDA as having method.

I am also inclined to take the position that CDA is not a movement but an academic theory of discourse in social life and a method of research which operationalises that theory.

It is, however, the ongoing purpose of this blog to work through the questions of CDA as method more carefully.