Multimodal Critical Discourse Analysis

Andrea Mayr

Halliday’s work has also contributed many of its analytical tools for the kind of linguistic analysis carried out in Critical Discourse Analysis (CDA). CDA is probably the most comprehensive attempt to develop a theory of the inter-connectedness of discourse, power and ideology. The term ‘critical’ principally means unravelling or ‘denaturalizing’ ideologies expressed in discourse and revealing how power structures are constructed and negotiated in and through discourse. CDA research specifically analyses institutional, political, gender and media discourses which ‘testify to more or less overt relations of struggle and conflict’ (Wodak 2001: 2). Because of its solid analytical foundation, Halliday’s work helps CDA practitioners to ground concerns about power and ideology in the detailed analysis of language. Both fields also share the view of language as socially constructed: language both shapes and is shaped by society.

Although the general thrust in CDA has been towards the analysis of linguistic structures, more recently there has been a visual turn inspired by scholars who have incorporated visual images into concepts of discourse and have moved towards broader multimodal conceptions (Kress and van Leeuwen 1996; Machin 2007). This extension of CDA into visual semiotics also has its origins in early Hallidayan theory which maintains that language is only one semiotic resource out of many and that several forms of representations, linguistic and non-linguistic, are used in the construction of discourse. For example, while political and ideological views of newspapers can be expressed in the choice of different vocabularies (e.g. ‘resistance fighters’ vs. ‘insurgents’) and different grammatical structures (e.g. active vs. passive constructions), visual structures in the form of images just as much can convey ideological meanings. Applying some of the linguistic principles found in SFL, Multimodal Critical Discourse Analysis (MCDA) therefore shows how images, photographs, diagrams and graphics also work to create meanings communicated by a text, which are often more implicit or indirect than language.

The work of Kress and van Leeuwen (1996) in particular has developed a set of tools derived from SFL that allows us to study the choices of visual features as well as lexical and grammatical choices in language. One of these tools is social actor analysis (van Leeuwen, 1996), a linguistic and visual inventory of the ways we can describe and classify people and some of the ideological effects that these classifications can have. According to van Leeuwen, people can be personalized or impersonalized, represented as specific individuals or as generic types. Certain naming strategies therefore foreground aspects of a person’s identity while backgrounding others. To illustrate this, let us briefly look at media representations of young people which often construct them as a problem. For example, in the following headlines taken from British (tabloid) newspapers


Hoodies to be banned from shopping centre

Yobs rule streets

Hoodie bike yobs attack teacher

Teenage mother was shot down by hoodies.


We find that by impersonalizing young people as ‘hoodies’ and ‘yobs’, they are “objectivised” and turned into generic types. A tabloid newspaper may refer to them as ‘hoodies’ throughout an article, obscuring who exactly these young people are who may have experienced poor education, have had few opportunities and have little to gain from and contribute to mainstream society.










Visually, often accompanying images are chosen with stereotypical and generic portrayals of these ‘hoodies’ on housing estates. In the image below, we see three generic ‘hoodies’, with one looking straight at the viewer. Kress and van Leeuwen would call this a ‘demand’ image act, drawing on Halliday’s (1985) notion of speech acts. Just like the language, the imagery used – the generic picture of threatening young men in hooded tops and tracksuits – foregrounds their deviance and backgrounds structural reasons. Images like this are not meant to document but symbolize a certain type of young people. Language and image are used to form part of wider media discourses on young people that are ideological in the sense that they create patterns of inclusion (‘teenage mother’ as one of ‘us’) and exclusion (‘yob’ and ‘hoodies’ as others) and direct attention away from the links between poverty, lack of opportunity and deviance that often characterize the lives of these young people.

This brief discussion has served to illustrate the important role Halliday’s ideas have played in research conducted in CDA. Looking at language and images from a combined SFL and CDA approach means that we can begin to grasp how language and visuals are chosen in a given situation and also to suggest why certain communicative choices are made and not others.

Be the first to comment

Leave a Reply

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.