Postmodernism as a Theoretical Framework for Learner Autonomy Research

The multidimensional nature of current conceptualisations of learner autonomy combined with the need to access both individual and social constructions presents both ontological and epistemological challenges when researching a pedagogy for learner autonomy. This paper will discuss the advantages and challenges to exploring the development of learner autonomy and its implication for practice from a postmodernist and social constructivist perspective, based on a qualitative case study of the development of advanced specialist and nonspecialist foreign language learners and their teacher as a learner practitioner-researcher, in a large Higher Education Institution in England. It will show, in particular, how such methodology facilitates a dynamic research design, providing an opportunity to adapt and use existing theories whilst maintaining a strong emphasis on the learner's 'voices'. It will conclude with a set of recommendations together with the limitations of such an approach.

epistemological challenges. Learner autonomy and its development have been studied using a variety of paradigms and theoretical frameworks ranging from positivist quantitative approaches such as Fazey and Fazey's (2001) study of the autonomy of first year undergraduates, to narrative qualitative ones (e.g. Benson, 2006;Karlsson & Kjisik, 2009).
Although Curtis (2004) has associated postmodern theory with practical applications for promoting learner autonomy, no studies appear to have researched the development of learner autonomy using a postmodern orientation.
This paper aims to examine the advantages and challenges of using a postmodernist approach to exploring the development of learner autonomy, in undergraduate specialist and non-specialist language learners, studying advanced level French modules i within a large UK Higher Education Institution, within a social constructivist perspective, from a practitioner researcher's perspective. After some consideration of the study's context and of the learner autonomy construct, the paper will discuss the benefits and inherent challenges of using postmodernist methodology, within a multifaceted and multi-dimensional theoretical framework. It will also discuss briefly how it was applied in practice, using a few illustrative extracts from the study's data analysis, although a more detailed account is clearly beyond the scope of this paper. The paper will conclude with some recommendations for using postmodernism as a research design as well as consideration of the limitations of such an approach.

Context of the Study
The research focused on a case study of undergraduate specialist and non-specialist students studying French at the advanced stages of the University Language Scheme (ULS) -Stage 5 (CEFR ii B2/C1) and Stage 6 a or b (CEFR C1).
The ULS, as an Institution-wide Language Programme (IWLP), offers a choice of language modules to students of other disciplines (non-specialists) as well as core language modules to students majoring or minoring in Languages (specialists), alongside other specialisms such as International Business, Tourism or TESOL, in six stages of languages, loosely based on the CEFR, from ab-initio (A1) to degree standards (C1/ C2). Although the ULS offers seven languages, only Spanish, French, German and Italian are available at the two most advanced levels due to insufficient demand in other languages.
Together with the degree provision, the ULS modules have a vocational/ applied orientation. At the advanced levels, the scheme had three key aims: • the development of language skills to enable students to function in both a social and business environment, including mediation (Route 6b only); • the acquisition of general knowledge of the country/countries where the target language is spoken, together with the development of students' awareness of its/their society, traditions, customs and business culture; • the development of autonomous language learners.
Although the language study on the ULS could be described as general rather than discipline specific, Stages 5 and 6 module designs allowed students enrolled in other courses to focus part of their language study on their subject specialism through the production of a subject-specific portfolio, report and presentation. In the case of students majoring or doing a minor in Languages at ULS 6, the portfolio involved the development of more specialist language skills such as negotiation, translation and interpreting. To encourage planning and reflection, each portfolio included a planning record and self-evaluation (normally in L2 although some were written in L1). ULS 5 and 6a students were also encouraged to keep a research diary in L1 or 2. A summary of the assessment programme can be found in Table 1.  (1997) model considering the inter-relationships between the learning environment, the teacher, and the student. The main attraction for me was that both models were the result of comprehensive case studies/ literature reviews rather than single studies, and had therefore encompassed a broad range of perspectives and findings from different methodologies.
Benson identifies control over cognitive processes as probably "the most fundamental level" (2001, p. 87). His model covers three key areas for control over cognitive processes: • attention-active engagement with linguistic input, involving conscious apprehension and awareness of specific aspects of the language; • Metacognitive knowledge at task level encompassing any evidence of: decision to carry out the task, decisions about content, progression, place and time of learning, the selection and use of cognitive strategies, and the criteria selected for evaluation; • reflection-any form of reflection on the language, the learning process, their role within that process, (pp. 86-87).

Social level
William and Burden's (1997) social model of the teaching and learning process emphasised the role of the learning environment and that of learner-teacher interactions with this environment. I added the learner to learner dimension as shown below (see Figure 1) to highlight the role of peers in the process, in line with the importance accorded to peer to peer learning in existing literature in the field and beyond (e.g. Boud, Cohen, & Sampson, 2013;Kohonen, 2002;Little, 2000).

The importance of metacognition
Whilst attention and reflection were clearly important, the literature in the field of learner autonomy suggested that the development of metacognitive knowledge (Benson, 2001(Benson, , 2011Wenden, 1995Wenden, , 1998 was perhaps the determining factor in the development of autonomous learners. My initial focus was therefore metacognition and its development within a social context.

Other studies
Past and recent studies on learner autonomy have employed a wide variety of methodologies. Psychometric approaches using and/or developing rating scales as part of a study of the development of autonomy were popular with many researchers (e.g. Cotterall, 1995;Guglielmino, 1978;Lai, 2001). More recently, mixed methods have been adopted as a variation to more positivist methodologies (see Cooker, 2015or Tassinari, 2015. Constructivist qualitative practitioner action research involving a strong reflective/reflexive element gained popularity with authors such as Champagne et al. (2001) and Karlsson (2008). Complexity theory seeking to grapple with the multi-faceted and multi-dimensional aspects of the construct has also gained ground as a theoretical paradigm (see for instance Dörnyei, Macintyre & Henry, 2015). Narrative and highly inductive qualitative stories have also been used to privilege participants' voices (Benson, 2006;Benson & Cooker, 2013). In my own study, I was seeking a way that would enable me to represent all the voices, including the teacher's, without favouring any particular world view at the outset. I also sought to reflect the social dimension of the research and its close relation to practice, within a social constructivist paradigm. My main concern was to avoid framing participants' responses within a pre-existing theoretical framework. Although accessing voices suggested a phenomenological/ interpretivist methodological approach, concerned with discovering the interconnected whole of the phenomena under investigation (Denzin & Lincoln, 2005;Maykut & Morehouse, 1994), rather than a positivist one looking at component variables to understand the whole phenomena, this approach needed to be articulated within a theoretical framework that could accommodate the various perspectives and constructions emerging from all the 'voices', including student learners, myself as the learner-practitioner researcher as well as theorists/ authors from the literature, without privileging any.
I was also aware that my own conceptualisation (and attraction to the concept in learning and teaching) was strongly influenced by the ideological and political dimension of learner autonomy reflected in work such as Pennycook (1997) together with critical pedagogies originating from the work of Freire (1974). Using critical theory as a basis for my study, however, would have cast participants' responses within a broader context of resistance or conflict. I was therefore seeking an alternative to a critical theory lens.

Postmodernism as a research paradigm
Postmodernism tends to be described in relation to its antonym, Modernism, which is underpinned by a belief in the power of science to deliver unified and legitimate knowledge (Sarup, 1993). In contrast, Postmodernism tends to reject "metanarratives" which offer this unified and monolithic view of the world (Lyotard, 1984). Stronach & MacLure (1997) identify deconstructive approaches to meaning as a key feature of postmodernist thinking.
Deconstruction defined by Johnson (1994) as "opening up meaning as a question, as a non given, as a bafflement" (p, 39) challenges assumptions about the stability of meaning and points to a plurality of perspectives or constructions of the world we live in. The postmodern approach therefore entails a pluralist perspective and multiple 'truths' contingent to specific settings. Haber (1994, p. 13) asserts that Postmodernism is both "committed to" and "constitutive of difference". Although postmodernism is not a unified theory, all postmodern theorists recognise-"the arbitrary and conventional nature of everything social-language, culture, practice, subjectivity, and society itself" (Best & Kellner, 1991, p. 20). Despite the challenges of its inherent relativism (Hargreaves, 1994) and frequent criticism of its excessive relativity (Hill, McLaren, Cole, & Rikowski, 1999), a postmodern stance can provide new insights into complex and dynamic phenomena, in a situated context (Brown & Jones, 2001;Stronach & McLure, 1997). Although Hargreaves (1994) advises against postmodernism as a theoretical position in educational research because of it relativism, he recognises the methodological benefits of adopting such a stance as part of a research project.
In the context of accessing learners' own voices, without pre-empting their perspective on their context and prior experience, postmodernism seemed to provide a useful, albeit challenging, interpretive framework. In order to guard against excessive relativity, I found that Moderate Postmodernism that is a version of postmodernism which accepts cumulative knowledge (Roseneau, 1992), including non-postmodernist as well as postmodernist works, offered rigour (i.e. in-depth theoretical consideration) without rigidity (e.g. imposition of a metanarrative). Fox (2000) for examples identifies three key principles guiding postmodern research, particularly in relation to practitioner research/action research namely: "Knowledge is local and contingent; the research is constitutive of difference" (p. 19) meaning that the research question should neither "close down or limit the ways in which the subjects will be understood or conceive of themselves", i.e. participants' responses should not be framed at the start of the research; and "the theory should be related to practice" suggesting that 349 research questions and "their theoretical consequences will be of direct practical relevance to practice" (p. 20).
There were, therefore, four key reasons for adopting a postmodernist research approach in my study: • the need not to privilege individual or groups of individuals' constructions of the world reflecting a 'pluralist perspective'; • the situated nature of the study focusing on the participants' constructions of reality within their specific learning context; • my wish to move away from macro theories, such as critical theory, as a theoretical framework to a 'micro logical/theoretical' level, aligned with the study's participants' personal theory and closer to practice; • the dynamic nature of the research, and the possibility of change at a micro level working within the constraints imposed by the wider society, rather using a critical theory of resistance or conflict approach, in order to seek an ideal practice (Brown & Jones, 2001) or an absolute solution to the development of autonomy.
The political and ideological nature of the concept of autonomy as Pennycook (1997, p. 35) highlights as central cannot be denied. Clearly, the psychological dimension of autonomy does not operate in a vacuum, independently of the cultural and social values of the broader society (Pennycook, 1997, p. 44). Brown and Jones (2001) mention "critical pedagogy in a postmodern world" because we need to recognise that these values will vary between individuals and communities depending on their context and prior experience.

Questions
The key questions posed in the case study relating to the development of autonomous language learners were: 1) What do undergraduate language students believe about learning and their role within the process?
2) How do these beliefs relate to current conceptualisation of learner autonomy in

Dealing with language
A postmodernist research design entails giving due consideration to language and the plurality of meaning implied in this theoretical orientation. This means problematising the language used in participants' responses through disassociating words and their meanings, i.e.
participants may ascribe different meanings to the same word.
This meant taking a poststructuralist rather than a structuralist perspective i.e. moving away from Saussure's (1910-11) structuralist theory of language to poststructuralist plurality of meanings and shifting meanings. Signifiers and signified are no longer stable but continually separating to reattach themselves in "new combinations" (Derrida as cited in Sarup (1993), p. 33). The signifier and signified are still closely linked but the signifier has supremacy over the signified. In other words, there is not one privileged interpretation of a written or an oral discourse. Whilst Brown and Jones (2001, p. 98) emphasise the intolerance of poststructuralist theories relating to 'accurate expositions' or 'conclusive definitions', Spivak (1980) points to the usefulness of a poststructuralist stance in potentially unsettling and questioning "our mindset in the way we account for what we see" (p. 75). This will impact on the way the data can be collected and analysed as clarifying meanings and identifying themes, for instance, has to involve all participants. This was particularly true for Phase 1 which dealt with the students' perspective as opposed to Phase 2 which was solely concerned with the practitioner-researcher perspective.
Postmodernism and a social constructivist approach Candy (1991, p. 252) broadly describes constructivism as "a cluster of approaches which hold that knowledge cannot be taught but must be constructed by the learner". Linking constructivism with psychological theories of learning, Benson (2001) citing Paris and Byrnes (1989, p. 170) stresses that "knowledge is produced through socially conditioned processes of interpretation" (p. 36). The social constructivist approach to knowledge and its acquisition fitted the postmodernist /poststructuralist stance adopted in the study, in that it implied a local and individual/ group construction of the social context, including the relativity of the language used by participants to articulate these constructions. Maykut and Morehouse (1994)  Phase 2 data collection process and data analysis extracts). Maykut and Morehouse (1994, p. 145) identify the "trustworthiness" of the research as an indication of validity. Credibility can be assessed via a detailed description of the research process and its expected outcomes (Maykut & Morehouse, 1994;Nunan, 1992).

Scope and limitations of the research
Transparency, particularly a clear audit trail, will enable readers to "walk through the work from beginning to end" (Maykut & Morehouse, 1994, p. 145). Given its postmodern orientation, how this study's findings resonate with the readers' experiences and perceptions of 'reality', i.e. its face validity (Krueger, 1994, p. 32) helps 'validate' its results as well as its process of enquiry. In relation to a postmodernist approach, Smith and Hodkinson (2009, p. 38) point out that external evidence cannot force or coerce individuals' agreement to "see the social and educational world in the same way". The reader needs to be taken through the research process and be convinced by the author's arguments and evidence.

Data collection methods and analysis for Phases 1 and 2
As stated earlier, the research was subdivided into two Phases: • Phase 1 which focused on accessing learners' 'voices' through focus groups, i.e. the student learners' and through the learner practitioner-researcher's reflective diary as well as other 'voices' from the literature, leading to a revised construct; • Phase 2 which analysed 28 self-evaluation reports and five detailed research diaries, drawn from the same learners as the Phase 1 focus groups' participants, including the learner practitioner-researcher, for evidence of 'autonomy in practice', in order to assess the usefulness of the revised construct as well as the autonomy of all the learners involved. The outcome of Phase 2 would determine further intervention through curriculum and assessment design at the advanced stages on the ULS.

Outline of Phase 1 data collection process and data analysis extracts
Although a detailed account is beyond the scope of this paper, this section gives a brief outline of the data collection process and some illustrative extracts from the analysis together with the key findings, in order to show the approach's application in practice, within this particular study.

Focus group process
Phase 1 used focus group described by Krueger (1994, p. 19) as group interactions aimed at identifying feelings, ways of thinking and perceptions which may differ between participants. Focus groups are distinct from group interviews in that they include participants with common characteristics. They are repeated several times, with similar participants, to identify themes and patterns in perception (Krueger, 1994, p. 19).
During Phase 1 of the study, four focus groups of between two (group 2) and five participants (group 3) met between one and three times over one academic year (see O'Leary, 2010), for a duration of one to one and a half hours as recommended by Krueger (1994). Student volunteers, all experienced language learners, were grouped by language stage (Stage 6-Groups 1 & 3 and Stage 5-Groups 2 & 4). Group members were in the same class with me as their teacher, with the exception of Group 4 (3 students).
The focus groups discussions were carried out in line with standard ethical consideration. I produced an interview guide before each group meetings with a summary of the outcomes of previous meetings and/or other focus groups as applicable, included at the start of each meeting (Arskey & Knight, 1999;Banister, Burman, Parker, & Taylor, 1994;Krueger, 1994). In order to access their constructions of the learning process both generally and with regards to languages, participants were asked to: identify the characteristics of a good teacher/language teacher; a good learner/language learner; describe successful and unsuccessful learning experiences both generally and for languages; reflect on their own acquisition of language proficiency and outline their plan for future language learning after University. These questions were kept sufficiently broad to cover their learning/language learning experience in HE, rather than focusing on the last two years when I taught most of them. This was intended to mitigate the inequalities in the power relationship, as I taught some of them. Krueger (1994) suggests that the researcher may wish to modify the question structure and timing to get "confirmation of emerging themes and ideas" (p. 145). He points out that "typically the first focus group yields a considerable amount of information", in this case Group 1, however "each additional focus group produces decreasing amounts of new insights" (Krueger, 1994, p. 135). I used summaries created with the participants to develop the next interview guide for the following group meetings. Some minor modifications were made to the order of items within the interview guides as the study progressed. In particular, learner and teacher were interchanged to try and elicit more data on the learner in Groups 2, 3 and 4.

Recording and analysing the data
In relation to the data collection, Krueger (1994, pp. 143-144) suggests five possible options: 1) a transcript-based analysis; 2) a taped-based analysis using an abridged transcript based on careful listening of the tape; 3) a note-based analysis; 4) a debriefing session and summary comments at the conclusion of the focus group; and 5) memory-based analysis where an oral account is presented. I opted for a mixture of the second and third options, with some further adaptation to suit the poststructuralist approach to language: part of the analysis and the 'debriefing' elements of the third option took place with the participants as co-constructors/ co-researchers in relation to the preliminary findings, as recorded on the flip charts. I used the tape to produce partial transcripts of detailed interactions to complement the thematic analysis done with the participants, during the meetings.
I conducted the interviews in their L1 (English) using a 'brainstorming' approach with each participant's contributions recorded on the flip chart, in the participants' own words, for instance Approachability (Scott v , Group 1, meeting 1) and Self-motivated (Lesley, Group 3). I recapped the student's responses, throughout the interview, using the statements written on the flip charts to further clarify meanings and identify emerging themes with the students as the following extracts from Group 3, meeting 1 shows:  I also asked students to rank each theme in order of importance at the end of the session. I noted all differences of opinion on the flip chart reflecting the plurality of perspectives rather than trying to reach a consensus.

Findings
As illustrated by the extracts from flip chart summaries in the Appendix, the data from the Focus groups identified: the importance of the affective dimension, particularly the student-teacher relationship; the centrality of the teacher in the learning process in their mind; the high expectation of the teacher's subject expertise; and their recognition that effective learning depends on the learners' ability to work independently together with 356 an awareness of the benefits of collaboration/peer support.

The literature review
Based on the main themes identified in Phase 1, the review of existing literature focused, more particularly, on theories and case study research relating to affect, including motivation, and cognition/metacognition both from individual and collaborative perspectives, together with the role of the teacher within the learning process.
Although some of the responses were not unexpected, I found the importance attributed to affect in the participants' responses, particularly in relation to the teacher's role, particularly puzzling. The affective dimension is not usually considered in relation to learner development/ training at these levels. The findings from Phase 1, including motivation and the importance of collaboration, led to a revised definition and model which gave 'affect' a stronger emphasis, as shown in Figure 2.

Phase 2 outline of data collection and extracts from the data analyis
Based on the revised model of control over cognitive processes, I analysed twentyeight self-evaluation reports (students produce only one per portfolio-based assessment) and five learner diaries from the student learners (diaries were not compulsory and only five students completed theirs) as well as entries from my own research diaries, as a learner practitioner-researcher, for evidence of control over cognitive processes, in line with Benson (2001Benson ( , 2011, with affect now within a separate category: • control over cognitive processes using the broad categories of attention, metacognitive knowledge at task level and reflection; • emotional intelligence or affect, including evidence of addressing their own anxiety, encouraging themselves, awareness of their own emotional temperature as well as evidence of empathy and cooperation with others. The analysis, although done purely from the perspective of the (learner) practitioner researcher contained a number of challenges, particularly from a language perspective. Some of the diaries were written in the target language (L2 for the learners) and translated, adding further layers of interpretation. This was mitigated by including the raw data in the appendices but, as in L2, the written statements may not have accurately reflected the meaning intended by the learners. There was no opportunity to discuss the analysis with the student participants, in order to explore potential interpretations as they had already left.

Self-evaluation reports and Learner diaries
The findings suggested that, in practice, the student learners displayed some degree of autonomy. Reliance on tutor support and subject expertise appeared to be more limited than expected, although there was a discrepancy between the diary entries and the selfevaluation in this area, with most diaries referring to the tutor. A few extracts below illustrate the data analysis: Attention Although perhaps most difficult to identify, 25% of the self-evaluation reports included a focus on particular linguistic aspects such as specific items of vocabulary:

Learner-researcher practitioner diary
My own diaries were analysed using the same categories as for the students. In all 27 entries between 10 and 20 pages long were analysed. The extracts gave a flavour of the context of the research as I 'lived' it / perceived it such as this extract on reflection demonstrates: Reflection as in the case of the students is wide-ranging. In addition to reflection on concepts and theories, my research diaries included some observations/ commentaries on the data collection process: The mix of the group was very different from the previous ones-[..] they were all reasonably cosmopolitan-an 'independent' group who were confident of their language ability. They also must have learnt an awful lot of the language whilst abroad and yet were still prepared to assert the centrality of the teacher in the learning process" (DOL 10) Whilst Phase 2 concluded that there was some clear indication of the students demonstrating their autonomy in relation to other cognitive processes, evidence of control over the affective dimension was very limited. With the exception of one or two of the Learner-researcher practitioner's diary entries, there was no evidence in relation to cooperation or other relational aspects, despite the importance expressed by the focus groups' participants who were the authors of the majority of the self-evaluation reports.
These findings suggested the need for collaborative activities and for paying more attention to the affective dimension of the learning process, through creating more opportunities for collaboration with peers and encouraging students to articulate the emotional dimension of their learning experience. This led to the addition of peer feedback and collaborative learning activities such as group interpreting as part of an e.portfolio (see O'Leary, 2014).

Conclusion
As highlighted in the above discussion of the methodology and the brief summary of its outcomes, I found that a moderate postmodernist approach, albeit challenging, can yield useful insights into a phenomenon through avoiding closing down or limiting the ways in which participants can be understood or 'conceive of themselves' at the start of the research process. Despite the inherent relativity of postmodernism and the ever shifting meaning of language from a poststructuralist perspective, the approach enabled qualitative data to be questioned/ explored in different ways, facilitating understanding rather than coconstruction of meaning in respecting individual differences. It also created an opportunity for my own voice to be heard openly through the analysis of my own diary entries and reflection sections, rather than covertly.
The 'voices' in this study unsettled the primacy of metacognition in the development of autonomy in my own understanding, opening up new considerations relating to 'affect'.
The change of emphasis from my own perspective on autonomy and the focus of my research is reflected in the final broader title of 'Developing autonomous language learners within the HE curriculum: a postmodern and social constructivist perspective' chosen for the study.
The outcome influenced my own practice (and hopefully that of some of my colleagues/ readers), particularly in relation to affective considerations which now forms an integral part of class activities and assessment preparation with some degree of success (see O'Leary, 2014). The relativism inherent in a postmodernist research design together with its limited context, however, means that this study's findings need to be constantly challenged in 'conversation' with other and newer studies and theories, rather than simply adopted.

Christine O'Leary is a Principal Lecturer in Languages and Cultures at Sheffield Hallam
University (UK) where she currently heads the Languages and Cultures Subject Group. Her research interests include learner development and autonomy, assessment as/ for learning and Higher Education pedagogies. She has presented and published papers in these areas for the last 20 years.