As I was reading and replying to students’ email journals every week, what struck me was their consistently emotional tone. This was certainly not dry data. It was literally bursting with feelings. As I have argued elsewhere (Sampson, 2016a; see also Radford, 2007), I find that action research offers the possibility of looking back historically after the collection of data in order to explore new directions in the researcher’s understandings. It was an interest in investigating the situated, emotional nature of my own EFL classrooms that prompted me to re-analyse participant journals from the perspective of L2 study feelings.
Looking at Levels: Describing the Whole and the Parts
I started with a seemingly straightforward quest to look at the collected data and describe the feelings and connected object foci to which writing in the participant journals pointed. I will not dwell in too much detail on this initial re-analysis, but instead refer readers to Sampson (2020). Suffice to say, I carried out routine thematic content analysis, without predetermined categories, of different feelings and object foci using the qualitative data management application NVivo. Entries were additionally coded to ‘week’ codes in order to facilitate an examination of the ebb and flow of feelings across the semester. This first stage of analysis involved a process of quantifying the qualitative data to gain a sense of the kinds of emotions and the extent of their experience by my students (Onwuegbuzie & Daniel, 2003).
Analysis revealed a kaleidoscope of feelings. Confirming my initial intuition of the emotional tone to the journals, an incredible 94% of collected responses touched upon feelings in some way. A total of 10 feelings frequently emerged across the entries of learners in the two classes: seven pleasant (or ‘positive’) feelings – a sense of achievement, enjoyment, gratitude, interest, admiration, excitement, surprise – and three unpleasant (or ‘negative’) feelings – disappointment, a sense of difficulty and anxiety. In congruence with past research (Garrett & Young, 2009; MacIntyre & Vincze, 2017), and supporting Pavlenko’s (2013) call for an expansion of research away from an obsession with L2 anxiety, there was a far greater incidence of pleasant feelings. As the most prevalent, a sense of achievement was apparent in a remarkable 44% of entries. Moreover, the greatest spread of students (45 of 47) wrote about enjoyment. In terms of unpleasant emotions, disappointment (31% of entries) and a sense of difficulty (27% of entries) were more prominent than anxiety (18%). All in all, suggesting the range of feelings experienced in the L2 classroom, over 30 different students mentioned eight of the feelings at least once. Even the remaining two feelings were mentioned by around 20 different students.
However, complexity thinking fosters an understanding that in looking only at the averaged, generalized whole, we may lose sight of the experience of any one learner. As Morin (2006: 6) describes, ‘knowledge of the parts is not enough, the knowledge of the whole as a whole is not enough’ but we must attempt ‘to comprehend the relations between the whole and the parts’. Teaching, even more so than other ‘helping’ professions like nursing and social work, requires a focus at one and the same time upon individuals as well as a group as a whole (Urdan, 2014). In order to more adequately represent the whole (class group, whole semester) and the parts (individual participants, specific lessons), I employed a tool known as multiple threading (Davis & Sumara, 2006). In their original exposition, these researchers used this tool to illustrate how often and to what extent individual voices or ideas contribute to an overall text, such as a research paper or dramatic performance. Multiple threading ‘involves the presentation of several narrative strands’ in which ‘some may be only brief phrases or single images that punctuate the text, and strands may overlap or interlink at times’ (Davis & Sumara, 2006: 162). I adapted this tool to visually represent not only how often, but also in what ways individual students joined the overall ‘feeling narrative’ of their class group on any specific day, and over the semester. The multiple threading used weeks across a horizontal axis and individual learners along a vertical axis. As such, I allocated a ‘square’ to each learner for each week of data collected. Emergent from analysis of the qualitative data, I then counted the number of different feelings experienced by a learner for a particular week. The square at the intersection of learner-week was then divided as evenly as possible based on this count, and different shading or hatching applied to represent these feelings (see Figure 3.1 – names are pseudonyms). (One word of caution here is that squares with a larger area devoted to a specific feeling do not imply that this feeling is stronger.)
Figure 3.1 Multiple threading of learners’ reference to feelings by lesson
The multiple threading representation is somewhat of a visual overload which requires more effort – or allows more freedom of insight – on the part of the observer. Nevertheless, using multiple threading from a complexity perspective reminds us to maintain our understanding of the ways in which individual students all contribute to the whole of the feeling narrative of their class.
I recognize that multiple threading is partial, as proponents of complexity understand all representations of data and ‘knowing’ to be (Davis & Sumara, 2006). I do, however, find it offers potential as one tool to preserve a sense of the whole and the parts, in terms of learner-class and lesson-semester. It allows us to zoom in on the feelings of an individual student in one particular lesson, or trace their emotional trajectory across the semester; concurrently, we can contrast the feelings of multiple students in the same class during a lesson or across time. By examining any week in Figure 3.1, we can gain a sense of the ambivalent nature of feelings perceived by individual students and across students in any given lesson. That is, the multiple threading aligns with Boudreau et al. (2018) in hinting that there is a great deal of interpersonal variation in how feelings evolve over the course of classroom experiences. Multiple threading moreover allows us to maintain sight of the longer timescale of the semester: Some learners tended to mention positively-valenced feelings more consistently across the semester, other students more negatively-valenced feelings, while still others revealed a relatively even mix. In line with Garrett and Young (2009: 221), it suggests a focus on how feelings were ‘modified by new experiences … over time’ and emerged in different ways within and across lessons. Finally, we can also use the multiple threading matrix to gain a sense of the general emotional orientation of the class as a whole. By assigning contrasting colours or hatching for more pleasant or unpleasant feelings, we can make a rudimentary analysis of the emotional valence dynamics of the whole. This said, we must remain cognizant that more pleasant feelings across the class group does not necessarily equate with constructive action tendencies or motivation, with the reverse also being true (Méndez López & Peña Aguilar, 2013; Pinner, 2016). Notwithstanding, I feel that a multiple threading does a reasonable job of representing the kinds of implicit understandings that teachers develop of the general emotional climate dynamics of a class group.
Describing the Emergence of Feelings over Interacting Timescales
My take on complexity as a philosophical and research position aligns with Kuhn (2007: 299 – emphasis mine) in understanding that ‘it is more useful to have evocative rather than prescriptive descriptions,’ such that research based on complexity thinking ‘is utilized for exploring possibilities rather than prescribing relationships and processes’. Ushioda’s (2009: 217) argument that much research to date has intended to ‘uncover rule-governed psychological laws that explain’ instead of ‘explor[ing] the dynamic complexity of personal meaning-making in social context’ sums up well a fundamental difference in perspective. Complexity thinking runs in opposition to ‘the principle of simplicity [which] either separates that which is linked (disjunction), or unifies that which is diverse (reduction)’ (Morin, 2008: 39). As Morin (2008: 5) also proposes:
Complexity is a fabric (complexus: that which is woven together) of heterogeneous constituents that are inseparably associated: complexity poses the paradox of the one and the many. … Complexity is in fact the fabric of events, actions, interactions, retroactions, determinations, and chance that constitute our phenomenal world.
I therefore wanted to further