‘An interesting paper but not sufficiently theoretical’: What does theorising in social research look like?

Abstract

This article explores the concept of theorising in social research: what is theorising; how does theorising look to those doing it and how can it be explained as a practice. The article draws on different sources, including accounts of theory and theorising in the literature as well as first-hand reporting by academics. A view is reached of theorising as a personal undertaking involving a commitment to comprehend the world. It is argued that theorising is driven by the identification of interesting problems (practical or theoretical or both) and a motivation to solve them. Theorising involves a shift of awareness from subsidiary to focal and such a shift allows a more abstract level of understanding. Theorising tends to be an individual undertaking which draws on personal knowledge, but it is only made possible by psychological tools which have been handed down from the past, ones which need to be accessed, understood and adapted. Theorising requires a disciplined creativity in which the imagination is channelled into finding solutions that are compatible with observed data.

Introduction

I once received feedback on a paper I had submitted to a journal in which the reviewer stated that it was ‘an interesting paper but not sufficiently theoretical’. I did not understand the point being made as I felt I had theorised by modelling my findings, as it happens by using a kind of network analysis. However, I had clearly missed something important as far as the reviewer was concerned. I needed to theorise more deeply but what would that mean and how could I do it? I sidestepped the question and explained to the editor that the article was scholarly and this proved an acceptable defence. However, confusion about theory has stayed with me over the course of my academic career and I wanted to do something to address it. In particular, I wanted to understand two things: ‘What is meant by theorising?’ and ‘What does theorising look like in practice?’

The article covers my reflections on these questions. It is structured around six interlocking episodes: Understanding theory and theorising; Theorising in use; First person accounts of theorising; Working towards the concept of theorising; Refining the concept of theorising; Final reflections on theorising. Those concerned with theorising have often argued for a narrative, or at least an insider, approach to reporting so as to make clear the personal nature of the activity and to enable the reader to track the process of discovery (e.g. Clegg, 2012). In keeping with this appeal, my article is a first-person account of an exploration of theorising. It is, appropriately, an account of theorising about theorising.

Understanding theory and theorising: what is the problem?

The problem is both theory and theorising are open to varying interpretations and are difficult to describe. For example, Merton (1967) sees sociological theory as referring to logically interconnected sets of propositions from which ‘empirical uniformities’ can be derived and empirically tested. This appears at first to be quite clear but Merton recognises that the process of theorising is not linear and may consist as much of serendipity as design; there is no single method to follow and no single version of theory. Recent writers have further stressed the varied nature of theory and theorising. For example, both Abend (2008) and Krause (2016) offered a list of, respectively, seven and six associations with theory. As with Merton, these covered the identification of patterns within the data, including associations between variables and possible causal relationships, but theory could also provide a more general picture as to how data fit together. For Abend and Krause, theorising might also consist of applying concepts to new data and developing new concepts taking account of observations and previous concepts. Theory can also be normative (what ought to happen) and can have a meta dimension so that theorising can be thinking about theory and those who propose theoretical frameworks. If a single underpinning idea about theory is Martindale’s (2013) ‘theories seek to explain things’, then the idea of explanation has to be understood in very broad terms.

Given the above, the task of teaching and learning to theorise can appear opaque for new researchers. Kiley (2015) illustrates this point by citing a student looking back on the experience of being examined: ‘people kept asking me about my theoretical perspective but I didn’t have a clue what they were talking about’ (p. 57). Further, the concept of theorising seems to shift not only across fields but by position in that field so that, for example, practitioner researchers might expect a theoretical contribution to have implications for practice in ways in which more ‘disciplinary’ researchers might not (Evans, 2002).

Differences in perspectives on theorising were confirmed for me in another source of evidence about theorising: a project on teaching theory in which I had participated (Hammond et al., 2016a, 2016b). Here, I found that for some research students theorising was tied to methodologies such as coding and data analysis or designing conceptual models to be tested through experimental methods, but for others theory was located at a distance both from data and from procedures for collecting data. Grounded theory was raised in many of these discussions as, unlike some other approaches, key texts referred explicitly to theorising. These texts further set out an extraordinarily clear guide to the process of generating theory (Glaser and Strauss, 1967), even if this process was disputed among grounded theorists themselves as well as researchers more broadly. Research students in my project felt vulnerable when offering theoretical insight and they felt particularly susceptible to ‘theory games’. For example, they tended to believe that others would ‘get one over on them’ if they had misunderstood Bourdieu, Foucault or other major thinkers. They were also concerned about the limited range of resources on which they were asked to draw and wanted theory to provoke exploratory (to use the language of Mercer, 1995) discussions rather than disputational talk over correct or incorrect interpretations of the great thinkers. There was a discipline dimension here so that, for example, those who were located in self-styled theoretical fields felt particularly vulnerable about their understanding of theory, in comparison to those in more practice-based ones. As I talked about the project to colleagues, I discovered similar feelings of confusion and at times anxiety about theory.

At this stage then the problem seemed clear. At a cognitive level theory as a concept was difficult to comprehend as it was multi-dimensional, while at a level of practice discussion of theory could be prescriptive and defensive. I had also learnt that the difficulties I had experienced in understanding theorising were not just private or particular to myself, others experienced something similar. This was a start but so far I had tended to focus more on theory rather than the act of theorising. I decided next to look at how the term theorising was being used in the literature.

Theorising in use

What do researchers claim to be doing when they are theorising? To address this question I carried out a simple web search for publications with the word ‘theorising’ in the title field and I accessed 18 of the most cited papers. (Details of these papers are included in Table 1 in the uploaded Supporting File.) This was a rough and ready, even naïve, approach but it succeeded in giving me an overview of the term in use. If language use can be understood as a game, here was the word theorising being played.

A common theme in the papers was that theorising involved finding a perspective or ‘lens’ through which a problem should be viewed and an invitation to see the problem in a new or revised way. For example, in discussing corporate social reporting, Adams (2002) argued that while past research had explored factors affecting reporting, this had meant that ‘internal contextual factors’ had been underplayed and should be given more focus. Other writers, too, saw that the methodological problems in their fields were leading them to a greater focus on agency (see Connell, 1985; Popay et al., 1998; Williams, 1995). Couldry (2004) in looking at text analysis argued that researchers should pay less attention to ‘proving’ the impact of text on a reader and more to what people did when engaging with a text and why they did it.

Sometimes a new approach was needed as the object of study had changed. For example, Cassarino (2004) suggested that its vastly increased scale had called for new approaches to research into global migration and, in the same vein, Chatterton and Hollands (2002) argued that cities had changed markedly and new approaches were needed to understand those changes. A recurring theme in the papers was to expand the frame of reference by loosening discipline boundaries (as in Clegg, 2001, writing about gender and computing) and/or by introducing new resources to the field (as in Williams, 1995, who drew attention to the relevance of Bourdieu in respect of research into the ‘class, health and life-styles debate’). There was, however, little on the process of discovery in these papers, though in discussing feminist theorising Connell (1985) gave a sustained consideration as to what counted as evidence in social research and Livholts (2009) explored problems of positionality and reflexivity.

None of the reviewed papers fully defined theorising, even though the term was used in the title of each. Rather, in nearly all cases, theorising could be summarised as involving three steps: Identifying a gap (both in terms of data about a topic as well as epistemological underpinning); Offering an alternative and explaining that alternative (often in terms of broadening the field); Justifying the alternative logically and empirically (even if the relationship of the data to the theoretical frameworks was not always clear). However, the process of theorising was underexplored. This was not surprising as most academic writing is concerned with output (theory) rather than the process (theorising), something that has obvious value for efficient reporting. However, by ‘blackboxing’ the process of discovery outputs can be mistaken as inevitable or necessary and it is easy to forget that they are the product of human interpretation.

My review of past papers was then helpful, but only to an extent. I needed to dig deeper into the practice of theorising and here, prompted by a colleague, I found Swedberg (see, for example, Swedberg, 2012, 2015, 2016a) a useful point of reference in work. I précis the key points from Swedberg (2012) below:

• Theorising is distinctive to theory. Theorising is the process that leads to theory, theory is built, theorising is the process of building. Theorising focuses on discovery rather than justification; justifying comes later.

• Theorising is a personal undertaking, which draws on one’s own resources and on one’s own ideas and experiences.

• Theorising can take many different forms but often calls for a different way of thinking which is more intuitive, less procedural, than other steps in the research process.

• Observation is critical to theorising by which Swedberg means concentrating on a phenomenon, ‘staying with it; and trying to understand it’. Observation can draw on many different resources, but theorising is not confined by the data.

I felt this was moving the inquiry forward. Swedberg was providing a particularly helpful distinction between theory and theorising and this strengthened my belief that many of the papers I had reviewed earlier, while useful, were not really about theorising at all. I also felt freed by Swedberg to present my inquiry as a personal undertaking and to be open about its iterative and serendipitous nature.

Yet Swedberg was not the last word on theorising and critical, albeit generally supportive, replies have come from Abbott (2012) and Carleheden (2016) among others. What particularly struck me about Swedberg was that although he himself was a sociologist he was not very sociological in discussing theorising. In part, I welcomed this as I identify myself as an educationalist rather than a sociologist, but it did mean that in presenting theorising as a personal undertaking it would be easy to miss the extent to which it was supported by engagement with publicly available resources – in Swedberg’s case intellectual resources such as Pragmatism and the work of Peirce. I felt a further tension lay in the idea of theorising as both intuitive and a process of logical reasoning. In fact Swedberg used several ideas to describe theorising (it is playful, abductive, tacit, it operates at a subconscious level, it involves free association and it is deeply observational) but these were not synonymous. My final reflection on Swedberg was that there was a lot on the process of theorising but less than I expected about his own process of discovery. I decided that my next step would be to access more personal accounts of theorising.

First-person accounts of theorising

There is relatively little reference to personal knowledge in discussing theorising though ethnography is an exception as the researcher’s own inquiring experience is the instrument of data collection. This means that reflexivity is taken very seriously by ethnographers (see Davies, 2014; Rossing and Scott, 2016; Stewart, 1998 and many others). For example, Davies (2014), in writing about children’s play, pays close attention to her own feelings and how these affect her interpretation of the events she is observing. At one point she notices her:

extreme exclusion anxiety on behalf of Tom (a child in her study). I see something that hurts. His exclusion has become unbearable to watch. I catch myself falling into a habitual interpretive pattern of individualizing the problem and blaming Jonathan (another child in the study). I want someone to tell Jonathan this is not how the children behave here.

This is important as it shows how an emotional response, in this case to a child’s exclusion, led to an initial theorisation as to why something was happening even if, as Davies later reveals, she was not bound by those feelings.

A concern for emotion and positionality, as well as biography and autobiography, has been widely shared, too, within feminist methodology (e.g. Middleton, 1993; Rowbotham, 2015) and questions of reflexivity have become more mainstream in discussing qualitative methodology (e.g. Huisman, 2008), including action-oriented inquiry (e.g. Winter, 1989). When it comes to Pragmatism, influential for Swedberg, North American writers found it natural to draw on their own experiences and write from the first person. This can be seen, for example, in Dewey (1910) when discussing thinking and in James’s (1907) rather convoluted story of what he learnt from a picnic in the country. In more broad-based social science the most cited personal reflection on theorising is that of Mills (1959), see, for example, Burawoy (2010) and Gane and Back, (2012), though the reissued Eco (2015) is becoming popular too. Indeed Mills (1959: 202) gave, and illustrated, a particularly useful metaphor for theorising as the ‘shuttling back and forth’ between data and theoretical concepts. Both Mills and Eco stressed the meticulous use of notebooks and paying close attention to both the data and to theoretical resources. For both, questions of identity were tightly bound up with their work and Mills, in particular, embraced the role of an outsider from what was for him an otherwise compromised practice. Implicit in both was that through scholarship it was possible to arrive at a level of objectivity that could be distinguished from everyday or straight narrative reporting. However, scholarship did not rule out the role of chance and indeed misunderstanding. For example, Eco recalled telling a sceptical reviewer that he could remember the exact location of a quote in a book by Abbot Vallet, as it was something that had been important to him in writing his own thesis. Eco (2015) finds the book and:

I … open it once again with a certain trepidation, look for the equally feted page, which I find with its beautiful exclamation point in the margin. I show the page and then I read him the excerpt that has helped me so much. I read it, I read it again and I am astonished. The Abbot Vallet had never formulated the idea that I attributed to him; that is to say he never made the connection that seemed so brilliant to me, a connection between the theory of judgements on the theory of beauty.[] And for more than twenty years I had been grateful to the old abbot for something he had never given me. I had produced the magic key on my own. (p. xxv)

This kind of misremembering intrigued me as it was something I, along I imagine with many others, had experienced though I had always been keen to dismiss it as an error. But what Eco shows is the inevitability of the distortion and indeed its value. Abbot Vallet had enabled Eco to access an important idea even if he had misinterpreted what the Abbot had written.

Reading Eco and Mills convinced me that I needed to access first-hand accounts of theorising. This was something I was able to do as I had carried out a series of long interviews with academics, at different stages in their academic careers, at my university (see Table 2 in the Supporting File) for the purpose of adding resources to a teaching project on theory. I listened again to the interviews, coded and aggregated my responses (see Table 3 in the Supporting File) and produced a narrative account. I was struck by the interest these colleagues showed in the project and the resonance their stories had with Swedberg, at least in respect to the idea of theorising as a personal undertaking. There were three things in particular that stood out when reflecting on the interviews: theorising involved problem solving; it involved abstracting from the mass of data; and it drew on other resources.

It seemed that nearly all attempts to engage with theory and theorising began by noticing a problem and, in the process, becoming aware of a difficulty or gap in the field. For example, one colleague saw a problem of accountability in non-governmental organisations (NGOs) and wanted to explore whether it was possible to build and test a framework which balanced conflicting demands in ethical ways. Another had a career-long interest in understanding destructive behaviour and noted the difficulty this posed for his discipline in which it was readily assumed that all behaviour was rational. For another, theory in the field of representative democracy was not reflecting the reality of social practice. As he put it:

I had a strong intuitive sense, I can’t put it better than that, that existing theory was too institutionally fixed and did not represent practices on a day-to-day basis. … and I just felt that existing theories were institutionally fixed and uncritical and I had to do something new. (Colleague D)

These tensions raised pressing theoretical, and, at least for some, practical questions, but they also posed intrinsically interesting problems. As colleague F explained, ‘I find different topics puzzle pieces’. As with all puzzles solutions did not come easily. Rather, commitment was needed to persevere with the puzzle and live with the discomfort that caused. I was struck by how my colleague explained that she felt ‘uprooted’ until she could solve the puzzle, and in her own doctorate research she spent ‘almost sleepless nights as I felt I had the data and they said this and they said that. But I did not have a theory’.

Colleagues needed to give these puzzles their full attention and they needed to take responsibility for offering solutions. Such a responsibility could expose them in front of others as it called for a degree of contrariness, or at least innovation, and this required inner conviction. The purpose of solving the puzzle was to present something that captured a story about the context they were exploring and something that could provide an explanation as to what was happening or at least the possibilities contained within that context.

My second reflection was that theorising meant focusing on what was important, in other words abstracting something from the data in order to explain what was happening. Theorising was a huge challenge as it involved looking closely at the data (in the broad sense of the term, for not all were empirical social researchers), but at the same time looking beyond the data and what was not given within the data. As colleague B put it, ‘The data do not speak for themselves, we have to say something more’. An essential point was that theorising provided a way of integrating varied sources of data that had previously been seen as disparate or without pattern. Theorising often felt like a sudden breakthrough and many were surprised how suddenly the essence of a problem was captured, when everything (or everything that was being attended to) seemed to fit around an overarching idea. Theorising, one explained was:

to suddenly make sense of it. You read it in a different way. Your main focus … at the beginning is on getting the key points but when you find a theory that works you are thinking not only do I understand this but this is really helping me to do this and useful … rather than … I am fighting against it. (Colleague F)

These moments of breakthrough might be experienced as, to some degree, spontaneous but they did not come out of the blue, they were often the product of a long engagement with data. A story that particularly struck me was that of a colleague who had spent a long time in the Moscow Red Army archives taking notes from documents written in the 1930s. Eventually, he came across an account of how the system of control could be thwarted. As he put it:

It was almost like reading a little play. This person, a Ministry of Defence officer, went to a factory and asked how much did this tank cost you to make. This guy (the officer) is a colonel and they (the factory management) won’t tell him. They say it is a military secret. So he went to the top and asked them (to tell the factory management) not to be so stupid but they still would not tell him. It made no difference. (Colleague C)

What the story illustrated for this researcher was that there was a principle at stake which trumped common sense and even political pressure within an authoritarian regime and that was the principle of secrecy. In his story, a factory manager had ‘got one over’ the system by evoking secrecy. My colleague explained that once he had grasped the importance of secrecy he could see it in all the accounts he had read. In his words, he ‘decided to spend the next few years looking at the system through the lens of secrecy’. The metaphor of theorising as a lens on the data is perhaps overused but it did seem to fit here; secrecy did not represent the truth of a situation but it was a way of looking at and integrating otherwise disparate events.

My colleagues recognised that the process of theorising was a subjective one and that others might prefer to look at a problem in a different way. However, they believed that theorising was subject to checks, for example, consistency and coherence in reasoning and, to varying degrees, all argued that theorising should be rooted in data. As one explained:

I have heard some colleagues saying I am concerned myself with theory and this has always puzzled me: why can’t I value them both (data and theory)? Data without theory is like a box with a hole in it, it needs something to bounce off. (Colleague B)

This colleague was not proposing a correspondence view of reality, rather she was offering a check on the credibility of theory in a way that had echoes of Mills’s ‘shuttling back and forth’ between data and framing of data.

My third key theme was that there were resources, principally texts, on which to draw even if theorising was a personal undertaking. These resources offered guidance in that they suggested key issues to look for, but they did not provide off the peg solutions. As colleague B continued:

theorising is not accepting (theory) but changing it. I have been accused of not using theory in the pure sense as I will always use some of it but I make a new version…

There was in several accounts a balance to be struck between respect for the field and lack of deference, as one put it ‘you should not neglect that tradition, you need to understand it’. Texts were a tool for theorising and in examining texts one had to look for the language employed and look for specialised meanings across fields.

Working towards a concept of theorising

I had read around the literature, looked at the usage of the term and had first-hand accounts of theorising. I felt able to reach an initial view of theorising. Three things now seemed clear.

First, theory could be understood as the goal of theorising, but theory, as seen in discussing Abend and Krause, carried a range of meaning. It was helpful, however, to see an integrating role for theory, one that involved, in Merton’s words, ‘identifying patterns and regularities’ in the data (texts, images as much as primary data), such identification having an explanatory power. ‘Explanatory power’ for some implied causality but might refer, as often in qualitative inquiry, to capturing perspectives on phenomena and identifying the consequences of agency.

Second, as seen in the section on ‘theorising in use’, a contribution to theory, or a contribution particularly valued by academia, would be to explain a phenomenon using new concepts, or better a new phenomenon using new concepts, in a way that was seen as compatible with data and held together by logical argument. Such a contribution might raise questions of epistemology, including the relationship between frameworks and data, and, at a meta level, a stance on the role of human agency alongside a view of the meaning of data itself (Brinkmann, 2014).

Third, theorising was daunting as it seemed to require a certain way of thinking which was difficult to capture and explain. Theorising appeared in Swedberg and in the interviews with colleagues to be driven by restlessness and dissatisfaction with how a problem was understood at present. This was followed by a corresponding satisfaction when alighting upon a solution.

A further observation was that even if theorising was difficult to structure and explain there were some helpful suggestions as to how it could be supported. For example, all my colleagues pointed to the importance of an engagement with literature, not just a literature associated with a particular field but a wider reading. Reading broadens the range of possible perspectives on theory, notwithstanding disputes as to when and where theory should make an appearance in social inquiry. Of course it would be impossible to read all there was, and earlier it was seen that some research students were inhibited from theorising by a fear of not having read widely enough or misinterpreting what they had read. However, this is an exaggerated fear if we are to follow Eco.

At this stage I felt happy with the view of theorising I had arrived at, but not everything was resolved. In particular, I was unsettled by my failure to address what seemed a foundational question: ‘If theorising involved a change in a state of understanding then how was it possible to go from one level of understanding to another?’

The answer to this question is often constructed around the idea of abduction, particularly in qualitative research (see Brinkmann, 2014; Charmaz, 2006; Hammersley, 2016; Tavory and Timmermans, 2014 and so on). Abduction is seen as bridging inductive and deductive ways of thinking by drawing on our capacity to make patterns out of repeated observation. Peirce is credited with offering the first take on abduction when recalling his investigation of the theft of his watch (the account is found in Sebok and Umiker-Sebok, 1981 among other locations and Swedberg tells the story well, too). Peirce found himself perplexed; he ‘knew’ who the thief was but he did not have/could not articulate the grounds for his suspicion. This did not stop him following the suspect and it emerged that Peirce was right in his ‘guess’. This story captures the different kind of thinking implied in theorising, a hunch, an obsession in following that hunch and the sense of everything falling into place once the solution is found. The story resonated for me as it chimed with the accounts of theorising I had accessed and showed that theorising was in part a matter of hunch, even guesswork. However, it did not seem to me entirely satisfactory. In particular, it did not explain what lay behind a hunch in the first place or why for Peirce grounds for suspicion were so difficult to access. I needed to take a step back.

Refining the concept of theorising

It occurred to me that the question as to how was it possible to go from one view of an event to another felt like one asked by educationalists and, in search of an answer, I returned to some of the education literature with which I was familiar.

I went back, first, to work by Michael Polanyi. Polanyi was not himself an educationalist, rather his key concern was to critique the positivist version of natural science (e.g. Mullins, 2006). However, it was his ideas of focal attention that found a home in education research (e.g. Edwards, 2014; Eraut, 2000) and indeed, like many others, I was encouraged to read Polanyi when first joining an education faculty. In a line of argument made familiar via Popper, Polanyi saw theories as hypotheses about the world that were put forward before verification and, as in Swedberg, for him theorising preceded justification (Polanyi, 1958: 35). For Polanyi, knowing was personal in that it involved intention. This perspective ties in with the views of phenomenologists such as Heidegger and Kierkergard but Polanyi’s particular take was to identify two types of awareness: focal and subsidiary. A focal awareness was on the goal of an activity (Polanyi, 1958: 57); a subsidiary one on the particulars.

The point for Polanyi (and perhaps in contrast to Peirce) was that these two kinds of awareness cannot be comprehended simultaneously. As an illustration, Polanyi cited his own bilingualism and suggested that if he were to focus on the content of a text (a focal awareness) he might not notice in which language it was written (a subsidiary awareness). Indeed, were he to suddenly focus on subsidiary features his focal attention would become disrupted and he would lose a grasp of the content just as, in a later example, an actor might experience stage fright if he/she focused on the subsidiary (the script) and not the focal (the role). As Polanyi sums up ‘all particulars become meaningless if we lose sight of the pattern which they jointly constitute’.

The shift of focus from subsidiary to focal might be experienced as a breakthrough as it represented a qualitative change in our understanding. Such a shift might also appear as intuitive or spontaneous, but this was only because the subsidiary focus had become subordinated (Polanyi, 1958: 11). The term subordinated is worth drawing attention to. Polanyi almost implied there were alternating perspectives on data (he was after all consciously influenced by Gestalt psychology), but it was more accurate to say the particulars become subordinated (to use an analogy ‘parked’) and, hence, tacitly held. It is very difficult to explain the process of reaching judgements as tacit knowledge is not easy to access – in gaining a focal awareness we have stopped attending to the details which we were once so intent on focusing upon.

Polanyi’s idea of shifting focus, in particular the changed character of thinking once particulars had been ‘subordinated’, helped go a long way to explain the sense of ‘everything falling into place’, which colleagues in my study experienced. Reading Polanyi had also helped me understand why Eco, earlier, had lost focus on what the Abbot Vallet had actually written; why, in Sennett (2008: 74), excellent workshop practice might die with the ‘master’; or why Peirce found his grounds for suspicion so difficult to access. As Polanyi put it, we know more than we can say.

How far Polanyi offered a general theory of knowing is an open question and there are other aspects of his theory, for example, the focus on skillful performance and the importance of apprenticeship (Polanyi, 1958: 56–57), which there is not space to pursue here. However, his account of knowing gave me a useful point of reference for understanding changes in perspective. He also provided a convincing picture of theorising as a personal undertaking, and why theorising has to involve a search for meaning.

However, I was also aware that Polanyi was less illuminating in putting theorising into a social context. Cognition was seen as personal even if there was recognition that systems which ‘foster and satisfy’ intellectual passions could survive only with the support of a society which respected those values (Polanyi, 1958: 216–227). For a more complete perspective on social context, I turned again to education literature and the social psychology tradition expressed most clearly in Vygotsky and taken up in, for example, ideas of distributed learning (Salomon, 1993), situated learning and Activity Theory (Cole and Engerström, 1993), and Valsiner’s (1997) three zones. Vygotsky was particularly important in that while we commonly think of language as a tool to express ideas, which of course it is, for Vygotsky the relationship between thought and language was complex. Thinking was mediated by language, and as Vygotsky (1930) put it:

The point about conceptual tools is that they modify the entire course and structure of mental functions by determining the structure of the new instrumental act, just as the technical tool modifies the process of natural adaptation by determining the form of labor operations. It recreates, reconstructs the whole structure of behavior just like a technical tool recreates the entire system of labor operations.

Vygotsky’s special concern was language in its broadest sense (Vygotsky, 1930). He was interested in how we moved from direct experience of the world to gaining internal control over higher order concepts (see Derry, 2008). Indeed Vygotsky created a distinction between everyday and scientific concepts, but these terms could be misleading as non-technical language had remarkable explanatory potential. This, for me, was illustrated in the story of the Red Army archives reported earlier. In this case a breakthrough had occurred when a researcher noticed a concept, that of secrecy, that could be harnessed to make sense of a range of reported experiences. Secrecy, at first sight, appears a rather prosaic term but it turns out to be a particularly subtle one, affording a focus on behaviour (social actors behaving secretively) as well as a social phenomenon (a secretive society) which existed, or at least seemed to exist, outside of the agency of the individual actor. Thus, a term in everyday use can turn out to have the potential to unlock our understanding of events once we can identify it, put it to use and understand its subtlety.

In fact language seemed key throughout my exploration of theorising and a further insight from Vygotsky was that language should be seen as social before it became intrapersonal (Cole and Wertsch, 1996; Vygotsky, 1978: 128). This seems particularly apt as it is our ability to internalise language from social interaction, and put it to creative use in new situations, which turns out to provide a useful analogy for the activity of theorising. Indeed this view of language allows us to balance theorising as the work of the autonomous individual with a ‘practice based’ perspective (Gherardi, 2000) in which language is constructed in communities of practice. If we only look at autonomy, we might idealise the research process and its ‘beguiling’ rhetoric (see Bourdieu, 1991; Johnson et al., 2000); if we only look at practice, including distortions at the structural (Scott, 2000) and ideological (Burton, 2015) levels, we might lose sight of autonomy altogether. We need a way to understand theorising as about individual cognition within a distributed context and Vygotsky provides part of the answer.

Final reflections on theorising

I set out to understand what theorising entailed. From all that I had read and heard first-hand, theorising was the search for explanation. This search could take various directions, and could assume different positions on causality and ontology. However, common to each attempt at theorising was identifying ‘patterns and regularities’ and offering an explanation for what was happening. Theorising enabled a contribution to theory, and a contribution that was particularly valued by academia was one that helped address a gap by offering a new type of explanation, one that took account of a changing context.

My next question was ‘What did theorising look like?’ From Swedberg, and from the Pragmatic tradition he cited, theorising looked like puzzle solving. It seemed to be fuelled by the identification of interesting problems (practical or theoretical or both) and a motivation to solve them. Among my colleagues, theorising was often experienced as a breakthrough and generated the sense of achievement associated with having solved a puzzle. There was more than one way to solve the puzzle, but whatever solution was offered needed to be supported by marshalling of data and logical argument.

I then asked how was it that in theorising we went from one level of understanding to another, one which was more abstract and more integrative of the data. This triggered a reawakened interest in Polanyi. From Polanyi I had come to see theorising as a shift from subsidiary to focal awareness, but only once the significance of the particulars of an event had been understood. This shift required a subordination of the particulars and this meant that theorising was particularly difficult to describe.

Theorising was throughout seen as a language activity and as such took place at both a social and an individual (or internal) level. This interplay between social and internal was captured strongly in the social psychology perspective, deriving from Vygotsky. As Cole and Engerström (1993: 1) suggest, ‘everything has been thought of before; the task is to think of it again in ways that are appropriate for one’s circumstances’. Theorising then was only made possible by psychological tools which had been handed down from the past, but these tools needed to be accessed, understood and creatively adapted.

Throughout my inquiry I found sources which stressed the spontaneous or unexpected nature of theorising. However, theorising should not be ‘black boxed’ or treated as unexplainable; the steps taken during an inquiry can be identified and held up to reflection. Indeed, one step typically involved engagement with literature – though when this should happen remained an open question. The literature offered frameworks for describing and explaining events and these needed to be understood. These frameworks might be firmly established within one’s field of research or they might have been constructed in other contexts and ‘borrowed’. In either case the task of theorising involved the transfer of concepts and their subsequent adaption. In this sense, theorising involved participation within a community of scholars but this might be an ‘imagined community’ and it was surprising how rarely colleagues spoke of theorising as a directly experienced collaborative activity.

Theorising was rarely linear and I have tried to reflect its iterative nature in the structure of this article. As such, there is no end point to theorisation and in my case I could see that I was left with new questions to tackle. For example, I had argued that individual thought was only made possible by socially mediated language but had I underplayed how much of our cognition was embodied, given that I had sometimes illustrated this embodiment? Second, I had described the outcome of theorising as provisional knowledge and believed, along with my colleagues, that theories required both warrant and backing. I rejected the correspondence view of social theory but I was very far from rejecting explanation. Others have complained of the ‘intellectualist bias’ in social research, which for Bourdieu et al. (1991) referred to a tendency to seek explanation above ‘uncertainty and indetermination’. Perhaps I had fallen into this trap? There were perspectives on theorising that I had left hanging too. Should theorising, as implied in both Bourdieu and Mills, be considered a craft (Burawoy, 2010)? Should we work towards a more sociological account of theorising at least in the sense of providing a critique of the distorting effects of gender and class on what counts as knowledge? If theorising is a language game as earlier suggested might we think more seriously about Wittgenstein (Standish, 1995)?

My final reflection revisits the feedback (‘a good paper but not very theoretical’) that triggered this investigation. What advice might a more helpful reviewer have offered? On the basis of my investigation, she/he could have acknowledged that concerns about theorising for its own sake and the playing of theory games are widely held in practice-based fields. She/he could have gone on, nonetheless, to ask whether there was a broader concept, made apparent by wider reading, that captured what was important in my findings. At the same time, she/he might have warned against placing excessive trust in existing frameworks for these were produced in another context for another purpose. Any framework leaves gaps and I could have been advised to work hard at understanding those gaps and their consequences for our shared understanding of social phenomena. My reviewer might have reminded me too of the value of a reflective learning approach when it came to theorising, and drawn attention to a particularly helpful four-stage model offered by Swedberg, (2012, 2016b), paraphrased here as: observe and choose something interesting to investigate further; name the central concept; build out the theory (using metaphors, comparisons; diagrams); complete the tentative theory and explanation. Finally, she/he might have reminded me of the value of theory for integrating findings and for contributing to a social research discourse community.

Supplemental Material

supplementarymethinn – Supplemental material for ‘An interesting paper but not sufficiently theoretical’: What does theorising in social research look like?

Supplemental material, supplementarymethinn for ‘An interesting paper but not sufficiently theoretical’: What does theorising in social research look like? by Michael Hammond in Methodological Innovations