3.9 Data analysis
3.9.3 Critical discourse analysis
According to Bhaskar (1998: 238), the critical realist starting point is the analysis of how meanings are constructed in social practice. To complement the analytical framework discussed above, I therefore used CDA as an analysing strategy to see how new academics constructed meaning by drawing (tacitly or explicitly) on a range of discourses in HE to legitimise, repudiate, enhance, enable, challenge, and even resist the material and ideational systems embedded in their academic reality. Given the critical and social realist ontological and epistemological stances adopted in this study, CDA was a natural choice as a data- analysing strategy because it is compatible with both of these, and because most of the data generated and gathered in this study constitutes textual representations of a semiotic kind. However, this study is less concerned with linguistic, textual or micro features of the
discourses (Jacobs 2014) and more interested in their discursive ability as mechanisms to enable or constrain (Fairclough 2003) the exercise of agency among new academics in HE. The term ‘discourse’ is understood and drawn upon in different ways by discourse analysts, but critical discourse analysis as a ‘clump’ or set of ideas at the level of the Real that has causal powers to affect social practice (Fairclough 2000; Gee 1992; Kress 1985) is significant for this study. According to Boughey (2012b), the text is experienced at the Empirical Level; it emerges at the level of the Actual, and it is a structuring mechanism at the level of the Real through SEPs (such as language, and class), CEPs (in discursive ideas and beliefs) and PEPs (in the way agents draw on them). Discourses as sets of ideas which ‘clump’ together in language or other sign systems may be stable, central or marginal, or not (Ibid.) in an open and intransitive system.
In using CDA in the analysis, I took care to uphold the realist characteristics and features of agents, the process, and the texts as independent and autonomous entities. In recent work, Fairclough (2003; 2010), moving towards a critical realist perspective on CDA derived from Bhaskar (1998), asserts that social events are constituted through the intersection of two causal powers – social practices and social agents. The overriding objective in using CDA is to give precise accounts of social change by looking at the ways and extent to which social events are produced through discourse (Fairclough 2005). Research in CDA is thus related to the practical problems of social life, with the objective of an ‘explanatory critique’ (Bhaskar 1998), based on the discoveries of social practices and from searching out methods for their solution (Chouliaraki and Fairclough 1999).
In examining how new academics mediate their structural and cultural contexts through discourses, it was important to instate them as active agents. Social agents actively work these ‘resources’ to create texts out of them, rather than simply instantiating them (Papa 2008). They ‘texture’ the texts through the use of devices such as hedging and modalising, manifested in the causal powers of new academics in this study, because it involved them as agents with causal powers drawing upon social structures (including language) and practices (including orders of discourse) in producing texts (Fairclough et al. 2004).
The most influential approach in CDA comes from Systemic Functional Linguistics (SFL), which analyses the relationship between language and its social functions (Halliday 1994). The importance of the text in relation to its context (Ibid.) is what distinguishes discourse analysts from critical discourse analysts (Fairclough et al. 2004). The ‘critical’ aspect of CDA is concerned with the truthfulness of texts, their production, and their interpretation, while a
‘genre’ perspective of discourse analysis (Bazerman 1997) relies on understanding texts as they are used for socially-recognised purposes. I found that the discourses drawn on by new academics in this study are closely interrelated with the context of HE, in which they are embedded.
It is equally important to see discourses not just as being activated by agents but capable of structuring mechanisms themselves. Discourses are seen as generative mechanisms of the various dimensions of social life, including physical, chemical, biological, economic, social, psychological and linguistic chemistry, possessing distinct structures, with effect in the events and experiences of agents (Fairclough et al. 2004). In this study, I was particularly interested in how dominant and marginalised discourses in HE contributed to the emergence of sub-discourses at the socio-interaction level (see Chapter Four). According to Kincheloe, Steinberg and Gresson (1996: 30), ‘discourses as tacit rules regulate what can and cannot be said, who can speak, and who must listen, whose educational perspectives are scientific and whose are unlearned and unimportant – in short it [sic] defines what is thinkable’. The data in this study, emanating from multiple sources and generated through multiple modalities and methods, provided a rich canvas for CDA, given its adherence to critical realist terms and its commitment to analysing texts as part of broader contextual realities. This was crucial in this study, where textual meaning was interrelated but not determined by the context of the university and HE in general. There was recognition of new academics’ ability to ‘work’ textual resources, often unknowingly, in ways that shaped their choices, placing them in an active role and in a dialectical relationship with discourses.
The discourse aspect of a social order is called an ‘order of discourse’, which is the way in which a range of genres, discourses and styles are networked (Kress 1989). An order of discourse usually encompasses a range of sub-discourses. In the example of the ‘quality discourse’, these would include discourses of performativity, accountability, and marketisation, and so on (Quinn 2006). The orders of discourse represent the totality of the discursive practices of a social domain and how they are articulated together (Fairclough 2005: 53). Orders of discourse can be envisaged as existing at three levels of realisation: situational, institutional, and societal (MacDonald and O’Regan 2009). In Chapter Four, orders of discourse, together with their sub- and associated discourses at different levels of realisation, are identified and discussed, as they pertain to the structuring mechanisms in this study.
substantive and meta-theoretical knowledge as well as the realist schema for data analysis discussed above (Table 7). Moving from a ‘soft eye’ gaze, informed by an implicit knowledge of the field through my immersion in it as academic staff developer, I used critical discourse analysis to lean on my trained ‘academic’ gaze (Maton 2001) in the HE field to enhance the analytical process (discussed already) of coding, interpreting, critiquing, exemplifying, abstracting, and retroducing, against the backdrop of discourses and HE substantive literature. I used CDA explicitly to maintain the tension between text and context, driven by critical realist disposition to get to underlying causal effects of discourses in order to explain how new academics, as primary agents, mediate the structural and cultural conditions that they confront. Aware that these accounts were fallible because of their many interpretations in an open system, I used CDA cautiously so that it did not over-determine the narrative construction of either the river-of-life story or the photovoice story enacted and embodied by the new academics as active agents. I remained committed to letting the data speak for itself as much as it could (Maton 2001). The actual presentation of the analysis of the social and socio-cultural interaction at T2–3 in Chapter Five is based on the adaptations I made in the adoption of the critical realist schema used in this chapter.