Imagination and Theory Building

Having spent months, on and off, observing and participating in the lives of primary school classrooms, I returned to my desk in a small post-graduate office. I’d been looking forward to this: after months of feeling like all I could do was respond to the onward rush of words and bodies as children and teachers managed the arrival, sitting, standing, talking, writing, playing, telling, and telling-off that makes up a school day, now, now, I was going to be able to think. And to help me think, I had the field notes, interviews, pictures, and stories created by children, teachers, and me. I was going to find something in all that to write about.

It was an immediately intimidating proposition. Ten thousand words of notes, over one hundred pictures and the same number of stories, three and a half hours of interviews, and more. Well, I thought, nothing will happen if I just sit here; I should do something.

Remembering advice given in the qualitative research methods texts I had studied as an undergraduate, I made a set of file cards. Onto each, I stuck a picture drawn by a child in response to the task ‘a time you used your imagination’.

I then set about trying to make sense of these pictures, looking carefully and laying them out in piles and rows on my desk, and sometimes on my floor. In doing so, I was making patterns of similarity and difference. I was using these cards to transform a set of pictures drawn by a hundred different children, some of whom shared a classroom and others who would never meet, into categories of imagination that made sense in and as theories. I was hoping that through my actions the correct theory of what imagination is would emerge.

In fact it was theories – plural – that emerged. I could pile these cards in many ways, all of which made sense in terms of previous theoretical positions and that seemed accurate in terms of classroom life. I noticed this as first one, then another, way of piling the cards made me feel discomforted. I did not wish to endorse some of these ways of piling nor the theories that they implied. I made one theory, then another.

In this chapter I hope to make the case for four interlinked contentions. My first contention is the easiest. It is that the researcher’s imagination is always, to some extent, a bodily imagination. This is to remember that humans are not minds floating above the world, but are inextricably involved in it with flesh and fingers. Doing data analysis by physically moving around concrete objects made this particularly obvious. But if we think of the other forms of work we do while analysing information – for me a matter of drawing mind maps, listing ideas complete with arrows and cross-outs, typing and rearranging with the cut and paste function – we begin to see that all the work we do before a piece of work seems finished is likewise bodily. We do not, or only rarely, find a way to imagine the connection between this piece of information and that piece, without at some stage interacting in a bodily way with the stuff of the world.

So this is contention number one: the imagination necessary to make any claims about new information is done with the help of our bodies. This opens us to a second contention, again made obvious by the concreteness of this particular data set. If the researcher’s imagination is bodily then there are multiple possibilities for how they might analyse their data. This means there are multiple theories they can make that will give order and shape to their information. I will soon explore this idea more with the help of Annemarie Mol, John Law, and Kieran Egan. These scholars’ work illuminates why multiplicity occurs and gives us good reason to believe that it is not, in itself, a problem. Instead is it is a natural consequence of our general embodied relationships with the world.

Their work does something else too. It is based on the idea that intervention should be the goal of scholarship, not just representation. We are inevitably intervening so we’d better be careful how. This brings us to my third contention: the problem with the ways I find to pile my data is not that they are bad representations of the world or of what goes on in the world. In fact, each seems correct in certain ways and is used to structure school life. However, each carries with it implications that are distasteful to me for ethical reasons. Each makes me discomforted when I think of what this would mean for particular children if applied back into classrooms as the single truth. So the problem with each way of piling my data is that as theories they would be bad interventions. They are embedded in, and would further embed in schools, a worrying politics. I will discuss this contention in the third, and biggest, section of this chapter through a discussion of three ways to pile my data.

This brings us to my fourth and final contention. I argue, with Helen Verran and Andrew Pickering, as well as Law and Mol, that we need a different type of politics to think better about data analysis. In this final section of the chapter I will introduce a way to frame this new politics funnelled through a final way I found to pile my data.

Justifying Multiplicity

I will start with my second contention that there are multiple ways to theorise a data set, hoping that my first will be reinforced throughout. In this section, I provide three ways of understanding the multiplicity of theories. One, Mol’s account, finds multiplicity in the fact of our enactment of reality. With this I defend multiplicity as being part of how we live. The second, Law’s, finds multiplicity in our ways of understanding and communicating about the world. I use this to explain where my particular stories about imagination come from. The third, Egan’s, I use to hint at the possibility that we will face the problem of multiplicity more often as time goes on.

Justifying Multiplicity One: A Consequence of Practice

Let us go further into why my practice of methodological imagination (my ways of imagining as I did research, analysis, and writing) might have been generating these multiple accounts. What was happening and should I be alarmed?

The work of Annemarie Mol helps us to see one way into this problem. In her book, The Body Multiple (2002), she looks at the practices that make atherosclerosis differently in different parts of one Dutch hospital. She saw patients with the same disease and suffering in various ways being sent to different parts of the hospital for diagnosis and treatment. How their diseased legs and their suffering were dealt with depended on whether they were looked at by the eyes of the walking therapist, the X-rays of clinicians, or the microscopes of pathologists. With the staff’s various embodied and technologically enhanced visions for seeing legs, cells, blood vessels, and their tools for treating these – walking therapy, cutting, medication – they made different atheroscleroses. They saw, understood and treated different diseases although all ‘really’ were atherosclerosis.

In essence, Mol’s argument is that because we enact or practise reality (of diseased bodies in her case) we can and do enact it differently in different times and places. This is crucially not an argument about perspectives – that people see or understand the world differently. Rather, she says that we are able to enact things differently and deal easily in practice with the multiplicity that ensues. Multiplicity does not worry Mol for she believes it to be a feature of our lives that we cope with as an everyday matter. She is concerned instead with how we manage multiplicity without noticing. How, she asks, are the different versions of the one disease, made by practice in operating theatre and outpatients’ clinic, able to hang together? What happens when the different versions are brought into one place? Or, to put it another way, how do we draw a picture of the multiple realities that exist, caused by different practices, but separated by only small spaces, times, or priorities? Her answer is that we find ways to spatially distribute, include within wholes, and place in hierarchies (Mol 2002). For Mol, multiple ways of performing a thing lead to multiple versions of that thing available to us. Following Mol, we would understand the multiplicity of my theories to be a natural consequence of embodied theory making.

Justifying Multiplicity Two: Interpellation into Modes of Ordering

In the article, ‘On the Subject of the Object’, John Law provides another way to understand the multiplicity of data analysis (Law 2000). Again this is a consequence of practice, but one that comes by being pulled into already established theories. He talks of interpellation into modes of ordering. This is to use Louis Althusser’s picture of state authority and ideology. Interpellation for Althusser means to find oneself part of the relations of power as soon as action is taken to pull you in. A policeman calls ‘hey you’, and you, in turning, become subject to him; subject to the material and ideological structures of the state (Althusser 1977: 160–165). By using this notion, Law refers to the fact that certain ways of enacting reality might call to us, ‘hey you’, making us in the moment of our acknowledgement subjected to their power (Law 2000: 13–15; see also Butler 1995; Verran 2001: 101–104).

It is this way when I am piling my cards. These theories do not come from nowhere, but from the talk of teachers and the writings of various theorists. It feels as if knowing these theories and the resonances they have in schools makes me receptive to the calls of pictures to be fitted into those patterns. It feels that, in knowing the relations of parts to whole in these various theories, each picture becomes interpellated into its place as soon as the theory comes to my mind.

Law illustrates the multiplicity of potential interpellations by telling the story of writing, or trying to write, the history of a military plane. This account became impossible for him to complete, and he writes to understand why this might have been. Part of his explanation is that he, his embodied mind, was being pulled into the various and mutually exclusive accounts and desires circulating around the aircraft. Should he write a ‘plain history’, a view-from-nowhere story about the historical trajectory of the craft? Should he write a policy history, detailing why the plane had been decommissioned and recommending ways to avoid such future losses? This was what many of his informants seemed to be expecting, but this was not what he wanted to add to the world. Should he write in an ethical mode, telling a story of the iniquities of the bomber in relation to the military industrial complex? This was a story deeply embedded within him already from exposure to the leftist political movements of the 1970s and 80s. Or perhaps he should write in an academic mode, telling the story in the particular disembodied language of the professional sociologist, a group to which he belonged. Maybe he should write this as an aesthetic story, telling of the power and even beauty of the machine that had called him, years before, to study it? These were some of the paths he felt pulled down, some stories he felt compelled by persons and objects to tell. But they were incompatible. Eventually he told the story of the aircraft in several ways, including this one of interpellation into multiple accounts (Law 2000: 19–23; see also Law 2002a; Law 2002b). Instead of hiding the multiplicity and the personal, ethical conflict he faced, he made it the subject of his enquiry. There is a suggestion implicit here. If he experienced multiplicity in writing his data, perhaps so do others. Perhaps it is a consequence of the way the world is, not of our inadequacies as clear thinkers. This suggests that we should talk about how to deal with multiplicity in theory making.

Justifying Multiplicity Three: A Consequence of Ironic Understanding

Kieran Egan suggests a different way to think about multiplicity: as ironic understanding. His work is useful because it suggest that we in the twenty-first century might be getting better at recognising and dealing with explicit multiplicity. Egan tells a history of ways of thinking. In it we have now reached a stage of literacy that finds us distrusting the completeness of language and yet still participating in it. Just think of a person doing a Google search, dealing with the many pages of text that are flattened together into equality, but some more accurate and some more useful to their task. That person copes with the partiality and multiplicity of the results with their critical and selective gaze – their ironic understanding.

To understand what Egan means by ironic understanding and what it has to do with multiplicity, we need to draw in his larger theory of thinking. In his 1997 book, The Educated Mind, Egan expands and clarifies his argument that children’s learning ideally recapitulates the changes in understanding that are visible in the historical record of Western civilisation. His argument rests on the idea that kinds of understanding are linked to our facility to use language. We are more capable of different thinking when we have the tool of written language than when we have only oral language, for example. Egan explains that

[k]inds of understanding are just the ways the mind works when using particular tools. All the kinds of understanding are potential or embryonic in all minds, along with an indeterminate range of other kinds of understanding that are so little evoked in our cultural environments that we hardly can recognize them (Egan 1997: 176).

Importantly, we retain the ability to perform each stage of thinking even when a new stage is achieved. Some losses of attention, skill, and interest are always to be experienced, but these can, and should be, minimised. Education should, he argues, be conceived as the appropriate encouragement of the ‘kind of understanding’ suitable to the stage a class has reached. I would like to remind readers at this point that Egan’s are intentionally made, as are the broadest claims and the biggest pictures of incredibly complex histories and learning experiences. His basic scheme, massively simplified, is as follows:

Pre-language – Somatic Understanding

Oral language – Mythic Understanding

Early written language – Romantic Understanding

Sophisticated written language – Philosophic Understanding

Distrust of language itself – Ironic Understanding

Egan adds another column to this table. He suggests that we can understand Western intellectual history as the process of Western culture moving through these stages (the idea that children recapitulate the history of Western thought as they learn has been developed and defended throughout Egan’s work. See Egan 1990; 1992; 2000). In The Educated Mind, he uses figures in Greek intellectual history that stand as having made the leaps required to move from one to the next. Mythic understanding, he suggests, is ubiquitous in cultures without written language. It remains so, in the stories we are all familiar with from our childhoods and in many of our other stories (on television, for example). Romantic understanding can be seen in Herodotus’s History. This he explains as being history interested in truth and extremes of human culture, biology, and feats. During romantic understanding, children might be drawn to the Guinness Book of World Records, to heroes, and to human agents as causes of historical changes. Philosophic understandings are those buildings of systems, those searching for laws, and cravings for certainty that Thucydides stands for, and which dominated intellectual endeavour in the nineteenth century. And ironic understanding is figured according to Egan by the ever-questioning Socrates, who was always denying the certainty of his knowledge even while he clearly did know much.

Over the twentieth century and into the twenty-first, as texts have proliferated and their authority and certainty questioned, ironic understanding has spread. To some extent this was brought into common consciousness by the post-modern concern with the limits of language. With ironic understanding we become more used to dealing with multiple accounts, their truths, and their differences. Perhaps (and this is me, not Egan), we have come to be more accepting of multiple accounts as we have become more used to the ‘democratic’ flattening of texts and authority with our new technologies: iPads, YouTube, Google searches, Internet news (for a related argument about flattening, see Law 2008). For Egan, ironic understanding does not need to flounder in the sense of uncertainty that this recognition brings. Instead, like Verran, he finds the process of being weaned away from certainty to be a good thing. For him, ironic understanding should be greeted with the ‘flexible, buoyant recognition of a multi-vocal world, both within and without’ (Egan 1997: 155).

Multiple Theories

With this optimistic vision of multiplicity, let us return to the picture of me piling cards – making theories and thus ontologically different ‘imaginations’. Well, what did these piles look like? What theoretical arguments was I slotting into? And what would the implications of each be?

Pile One: Imagination and Generalisations

In one attempt at piling, tempting for its binary nature and apparent simplicity, I found I could build two main groupings – a tree with two branches. In one pile would go all the pictures that seemed to tell only about personal experiences, taking quite literally the question ‘a time you used your imagination’. These could then be sub-grouped into categories that I could gloss as absorption of self into other media (TV, movies, books, computer games); recall of past personal event (for example, imagining the fish to be caught while the line was in the water); projection of self into possible future (winning a swimming trophy); and projection of self into myth (riding a dragon).

Into the second core pile would go those that seemed to reflect on what imagination is, showing children themselves abstracting from what they had done, seen, or heard about imagination, and generalising about imagination as a concept. This pile could be broken into two further subgroups. In one went pictures that suggested imagination and creativity were the same. For example, several show groups of children doing art, carrying the implication that ‘we always imagine in art’. The other group pictured various ways of crossing conceptual boundaries, suggesting a generalisation that saw imagination as a faculty that looks upon the categories of the world and enables their humorous mixing (‘a monkey living in a house and a human being in a tree’). This distinction between the personal and the generalising did seem to work – most pictures were easy to categorise somewhere and all were possible to categorise. However, the system as a whole I found discomforting and I soon realised why: I had been interpellated into a building a Piagetian theory.

Jean Piaget worked in the mid-twentieth century building theories about children’s cognitive development that had much to do with the late nineteenth century, and everything to do with the Euro-American conviction that science was the ultimate human achievement (Kitchener 1986. For a powerful critique of Piaget, see Walkerdine 1988). This rested on three assumptions we ought to be wary of. The first, coming out of the metaphysics that separates the world from knowers, assumes that scientific representations of the singular truth about the world are possible. The second follows on from this, taking as obvious that these truths are the highest goal for individual knowers and society alike. The third sees knowing as primarily involving disembodied minds (see Mol 2002: 153–157). In Piaget’s version, children’s cognitive function (their ability to think) would gradually develop towards this single goal of scientific knowing. Piaget described his aim as ‘to study the origins of the various kinds of knowledge, starting with their most elementary forms, and to follow their development to later levels up to and including scientific thought’ (quoted in Kitchener 1986: 4, emphasis added).

Fig. 3 Imagination as the Particular Experience of Absorption of Self into Harry Potter
Image reads: ‘(Harry Potter) A Gryffindor table in the great hall with desserts/Hogwarts’

Fig. 4 Imagination as the General Experience of Reading any Book
Images to show imagination as particular and generalised experience

What would be wrong with accepting that this holds true in the primary schools I participated in? Please note that I am making an unfamiliar move here. I am not suggesting that Piaget’s work is necessarily wrong. It is likely that children who are better at generalising will be better at science, and this capacity tends to increase with age. Perhaps also, science is the most sophisticated form of human thinking. Nor am I suggesting that Piaget’s theories are unimportant in what goes on at primary schools, for it is certainly part of the tacit knowledge that teachers bring to their classroom practice. Indeed, on occasion teachers do re-enact the Piagetian division of children’s intelligence based on their ability to theorise. In one lesson, for example, children made Plasticine symbols of the story they had just heard. The teacher used this as a diagnostic tool, pointing out to me the most ‘advanced’ children – those who had generalised the best (Catholic School, author field notes, 6 September 2007).

My discomfort sprang rather from the realisation that I would be re-enacting Piaget’s theory wholesale, looking backwards to find a theory to place on top of the practices I had observed. I would be imposing a previous picture of reality suitable perhaps for its time and place, upon the practices of quite a different time and place. This would, firstly, shut down enquiry. There would be nothing much more to say, apart from that some children are better at generalising than others. But given this theory was developed by a scientific thinker, and being used by someone trained in analytic if not scientific thought, all this would say is ‘these children are like me’. I am good at generalising, I am intelligent, therefore anyone like me is intelligent and anyone unlike me – i.e. bad at generalising – is not.

More, this arrogance would devalue the pictures (and hence development) of many of the children, including all those at Steiner school. A major part of Steiner educational theory argues that children are not able to think abstractly until the age of fourteen (Harwood 1958; see also chapter four of this work), and certainly the Steiner children I got to know were never encouraged to abstract while I was there. To adopt this form of theorising would be to suggest that Steiner children were at a lower cognitive level than others, despite their many other abilities and their general physical and mental equality with the rest of their age cohort. This would be single-minded and unfair – amounting to an IQ test of my own devising that marked some children high and others low.

It would also be bad for my research. I would be fixing myself, and my school participants, in (and as doing) a single reality – developing scientific thinking. I would then be unable to see what is clever, interesting, and important about their pictures. This would not adequately show the complexities of practice; complexities that were evident as I saw I could pile my cards differently.

Fig. 5 ‘Lying in a Bath of Money’

Fig. 6 ‘Winning a Swimming Competition’
Images to show imagination as type: self-fantasy

Pile Two: Imagination and Typologies

Un-piling and re-piling I found a new pattern. This saw the cards divided into eight piles, each of which had children doing different types/methods of imagining. So children could be seen as doing imagination as: memory, self-fantasy, mythic fantasy, absorption, memory/reflection on creativity, memory/reflection on absorption, transcending the limits of the situation, and humour/play with categories.

This piling also worked in that cards seemed to fit somewhere (sometimes in two groups) and none were left out. But again I realised I was drawing on a theoretical position, and again I was discomforted by what I would be doing if I let this patterning rest.

Building typologies is commonplace in data analysis and theorising, and indeed is a useful technique for thinking with. In the case of imagination, typologies have been necessary for trying to unravel the obvious complexity and multiplicity that greet us when we start to think about how the word is used in our lives. Imagination can be used to talk about creativity, fantasy, lies, goal making, re-creating vivid memories, planning, and many other things. But, as Raymond Williams says, ‘there is a problem in using not just the same word but the same concept’ for all the various manifestations of imagination (Williams 1991: 267). Moreover, this variety seems to call out for ordering. In a classic phenomenological version, Sartre (1948) gives an outline of the types of imagination as he experienced them. And more recently, McGinn stated that ‘it is difficult to approach systematically, to impose order on the various manifestations of imagination. That is what I have attempted in this book’ (McGinn 2004: iv).

In themselves, these works need not be worrying if they remained reflections on their authors’ own imagination – a topic everyone is uniquely able to speak about. There is a problem, however, in assuming that one’s own mental processes are like those of all others. This might be the case, but it is far from certain, and even more uncertain when one is talking about people living in different times and places from oneself. And, of course, in research interested in practice, minds are not even particularly relevant: action is.

Moreover, typographies seldom rest at division into difference: they are easily turned into hierarchies. Thus, Sartre distinguishes between good and bad imaginations; Williams argues that only one type of imagination really deserves the name; and McGinn devises a spectrum of imagination as shades of increasing sophistication (Sartre 1948; Williams 1991; McGinn 2004). This temptation is not incidental to the typographical approach; rather it occurs as a result of the search to reduce complexity – what is the best imagination, what is imagination ‘really’? This is intrinsic to universalism: there might be multiple types or multiple meanings of imagination, but generally, one is more truly imaginative or more truly good.

Typologies of imagination are also built into school days. Just as art is done in a different room from mathematics, creative expression is kept out of maths, and mental representations of the abstract are kept out of art. Making up an imaginative game is to be done only in the playground and at break times, not in the more serious class time. Drama, the embodiment of another character, is a Friday afternoon treat.

It is exactly this tendency to divide one imagination from another that I wish to avoid. Not only does it collapse the multiplicities and complexities of practice into confusing examples of some single typographical system (even if my own), but this would also tempt me into blindness when it comes to things that don’t seem to fit, or to force them into fitting. It would also claim that other minds are like my own and divide things in the same way. My ways of dividing imagination are, however, a product of my specific times and places – books read, schools attended, friends talked to.

Pile Three: Imagination and Meaning

This would not do. So again I tried, looking, thinking and piling, making patterns emerge. In the theory I played out this time, children are seen as having each drawn a meaning of imagination. Each has told me what (perhaps among other things) imagination means to them. These too could be grouped, with more categories needed and more pictures that seemed to fit in more than one category. I found pictures that could be read as saying imagination is something that helps us play; enables us to create stories and pictures; gives us historical empathy; lets us transcend the frustrations of our situation; and many others. I felt much more comfortable now – I was giving agency and voice to my participants – but still I was perplexed and doubtful.

The theoretical format I am adopting here is one which an anthropologist might use. In grounded theory for example, a qualitative researcher would attempt to gain an understanding of the perspectives that their research subjects had – to find what meanings the world had for the people being studied. As Kathy Charmaz, one of the most prolific anthropologists to write about grounded theory as well as use it, puts it, grounded theory ‘celebrates first hand knowledge of empirical worlds […][it] aims toward interpretive understanding of subjects’ meanings’ (Charmaz 2000: 510). This is a good approach in terms of the ethics of fieldwork – respecting the ‘world views’ of participants.

But it has to be used carefully, for it gets things the wrong way around. If we engage in looking at practice, we see that meaning doesn’t cause practice: it is caused, enacted, and changed in practice (for more on this idea see chapter two). Imagination, this view tells us, is not something out there, a real and fixed part of people’s minds and lives, waiting to be represented. Nor is imagination simply a concept constructed by children – or more conventionally, constructed and taught by the adults around them.

Instead it is something that is enacted in practice – sometimes, perhaps usually, resonating with the definitions that people will speak or write, not because they are ‘correct’ or because they are speaking for their language or culture, but because material arrangements, habits of word and gesture, and the organisation of school time allow certain practices to occur and be prioritised.

Fig. 7 Imagination as Meaning Play with Categories
Image reads (with spelling corrected): ‘I am drawing a picture of a monkey living in a house and a human living in a tree instead of a monkey in a tree and a human in a house’

Fig. 8 Imagination as Meaning Transcending the Frustrations of a Situation by Dropping a Chair on an Annoying Sister
Images to show imagination as having meaning

Looking at these pictures as simply telling us about meaning would give us two options. Either we would have to see them as better or worse accounts of the universal meaning of imagination, or as better or worse representations of the many equally valid socially constructed meanings of imagination. It would ignore the practices the children had depicted, casting them as simple scene-setting or as explanatory devices. It would see the materials and differing teacher instructions as limits to the cross-school comparability of the pictures, not as practices of interest in themselves. Our task would be to see through the pictures into what the children were really trying to say: which meaning of imagination were they trying to articulate? Where did the ‘real’ meaning of imagination, purified of technology and practice, rest for each child?

This approach would furthermore fall into the metaphysics that Verran has tried to make explicit and thereby refuse. We are very used to taking for granted the tripartite division between the world, representations of the world (knowledge), and knowers in whose minds knowledge is made and the world judged. If we took the search for meaning as our methodological basis, we would be making this tripartite division. We would have to see me, a researcher, removed from and judging the participants in my study (children and teachers). I would be separate from my world of knowers, seeing them from a distance. Those knowers, the children and teachers, would be assumed to see ‘imagination’ from a distance as either a single object seen from different perspectives, or as a cultural construct. They would be separate from their object of knowledge – ‘imagination’. I would be separate from the children and teachers, who would be separate from imagination. I would be looking at them looking at imagination. This would give me the role of maker of judgement, ignoring my own participation in the enactments of imagination at these schools and the fact that I am enacting a version of truth suitable to a particular university discipline at a particular time and place. It would fail to meet Addelson’s standards for ethical double participation (1994: 160–182).

Pile Four: Imagination and the Politics of Becoming

So with these worries sensed but unarticulated, I decomposed the piles again, sitting with the whole lot and simply looking. Suddenly striking this time were the pictures that didn’t seem to quite fit my expectations – those that surprised and confused. There was one that showed two children and a house, captioned ‘when I was doing a dance with my friends’. A practice of embodied imagination, perhaps. Another was of a computer screen with a maze-like game on it; imagining with a computer, a child made cyborg. A third showed a road, trees, and snake but not as a scaled or representative view but which (to me) captured a feeling. Imagination as bodily affect felt when going bush with dad?

These three pictures, and others too, showed practices and linked them to imagination. They were an expression of imaginations that had emerged in doings, and which are not contained in what observers and theorists say about imagination.

What would I be claiming about imagination if I took this piling seriously? All these pictures are about children’s bodily interactions with parts of the human and non-human world. They seem to be about the new and surprising products of these interactions, be they dances, understanding of the structure of a computer game, or the emotions conjured by being in the wilds of Australia. So these pictures show imagination as the unpredictable outcomes of human and non-human interactions. These may only last a moment, but can stay with us in memory.

Fig. 9 Imagination as a Feeling
The image reads: ‘This picture is about when me and my dad go into the bush I imagine’.

Fig. 10 Embodied Imagination
The image reads: When I was doing a dance with my friends

Images to show imagination as the emerging new

When examined in this light, the other pictures made sense also. They too were about interactions with other humans and/or non-humans, and they were about the outcomes of these interactions. The difference was that these others pictured interactions and outcomes that were more familiar. They told us about outcomes we have already cemented in our culture as being about imagination. Art, we are used to thinking, is about imagination; so too is the fantasy of winning a swimming competition, and pretending it is oneself who is riding the dragon from the story. In other words, these pictures hid the fact that they were about the outcomes of unpredictable interactions, because we have seen and articulated those interactions before. Having done so does not mean that we would have been able to predict it, just that when it occurs we know what to call it.

To theorise imagination in this way is, just as before, to invite consequences. Now they are for how we should do academic work. They are consequences that implicate a type of politics. I will follow Andrew Pickering and call this a politics of experimentation. By this he means to ‘imaginatively and critically explore the open-ended spaces of the world’s possibility’ (Pickering 2008: 13). Doing this kind of politics well requires engaging in a new and difficult way of seeing and talking that Verran, Mol and Law, and Star can help us with.

This politics, I suggest, has three main elements. The first, articulated best by Verran, is that we should engage with what is new and momentary. These can be recognised more easily at some points than others: for Verran what is of interest are the moments when people with different ontological commitments meet and try to work together. She calls these ‘postcolonial moments’ (Verran 2002: 730; Verran 2004). What comes out of such moments are the type of open-ended possibilities that Pickering speaks of. So too with imagination – we should work to notice the outcomes when people work with humans, places, and technologies that are unfamiliar. The unfamiliar demands an imaginative response if people are to find ways to live and work with them. The outcomes of these interactions, however, might be novel, fleeting, and difficult to grasp.

The second element in Pickering’s politics is to engage in what Mol and Law call ‘ontological politics’. To some extent, I will treat this as a caveat for the first part of our critical arsenal. This term was first coined by Mol in 1999 and later adopted by Law (2004). By ontological politics they mean that we must be aware that ‘reality does not precede the mundane practices in which we interact with it, but is rather shaped within these practices’ (Mol 1999: 75). Ontological politics helps us recognise that the agents involved in imagining are already being shaped by the reality they live with, and the practices this reality is made by. Their choices about what to do with their imaginations are already limited by their practical lives. Teachers, for example, have to choose to follow the curriculum to some extent. Children have to choose to accept this as their learning, or face punishments. PhD candidates have to choose to write in ways that are accepted by supervisors and examiners, and authors in ways that their publishers and publics will like. This means that the possibilities of the world are not so open-ended as we might hope, but the ways in which they are closed are particular to differently practising agents.

Finally, and with Susan Leigh Star, to do Pickering’s politics we must be ready to attempt to articulate that which is difficult or impossible to articulate, including when we follow the conventions of academic writing. Star explores an aspect of this problem in a beautiful piece about the experience of pain. Her suggestion is that underlying much of science and social science is the aim to generalise experience. But there are experiences that one can only know in the particular. One aim of social science therefore should be to tell of experience in such a way that it is generalisable, that captures a collective experience, by choosing sometimes to tell of the particular and momentary: ‘Responding to experience means letting generalization and specificity be in dialectic in our writings and biographies’ (Star 1998: 144). This is what Pickering asks in calling for imaginative exploration of the world’s open-ended possibilities. Sometimes the one, used imaginatively, can tell us about the many.


I began this chapter by arguing for the bodily nature of data analysis and theory building. This, I have said, naturally leads to multiplicity. Multiplicity, I have gone on to say, is not in itself a problem. Rather it is something we should think carefully about how to deal with. In this case, I have argued, multiple theories of imagination are not wrong. Each represents some of what goes on in classrooms and makes sense in broader theoretical traditions. But each has ethical and political consequences that we need to be conscious of if we are to make good decisions about which to endorse, particularly if we think our writing might have consequences for what happens in schools.

In this case I have argued that a politics of experimentation, enacted by theorising imagination as the unpredictable outcome of human and non-human interactions, is the most ethical. This is partly because (I believe) it is the most accurate: it gels with the metaphysics I follow throughout this book that tells reality as emergent in practice. But it is also the most ethical because it returns a type of integrity to the work of social science. By focusing on what is new, on what is possible and not possible for particular agents, and by striving to find new ways to articulate the experience of us all, the politics of experimentation is one that engages with people and things struggling to speak, act, and imagine anew.