Rethinking children’s agency: Power,assemblages,freedom and materiality

Abstract

This article attempts to rethink agency for childhood studies, drawing on Foucault’s theorisations of power, Deleuze and Guattari’s concept of assemblage, Bennett’s vital materialism and Grosz’s account of Bergson’s conception of freedom. I argue that (1) agency is ambivalent, that is, it has no intrinsic ethical value; (2) agency is not a property of individual children but happens within assemblages; (3) it is analytically useful to distinguish between more routine and more inventive tendencies of agency; and (4) agency arises in the relations between the organic and the inorganic, as life actualises the virtual potential of matter for indeterminacy. These ideas contribute to ongoing debates about agency within the field, and connect these debates with wider questions about children’s relations with materials and nonhumans.

Introduction

No one really knows what human agency is, or what humans are doing when they are said to perform as agents. In the face of every analysis, human agency remains something of a mystery. (Bennett, 2010: 34)

This article contributes to recent attempts to rethink agency within childhood studies. It does so via four linked arguments. I begin by expanding on critiques of agency in the field, arguing that childhood studies could benefit from viewing agency as (1) ethico-politically ambivalent and (2) not the property of individual subjects, but something that arises from the relations within heterogeneous assemblages. These arguments draw on Foucault’s theorisations of power, Deleuze and Guattari’s concept of the assemblage and Bennett’s vital materialism. I then develop these arguments further via Grosz’s Bergsonian account of subjectivity, agency and freedom. These ideas suggest that (3) it may be helpful to recognise a spectrum of children’s agencies, ranging from the more routine to the more inventive, and (4) children’s agencies arise not only through the relations of assemblages but also, more specifically, in the margin of indeterminacy produced by the relations between organic life and inorganic matter, understood not as distinct realms but as opposing tendencies.

I want to work towards a conception of agency that intensifies its specificity rather than flattening it out into more general concepts such as action, activity or movement; that recognises both the radically transformative potential of agency and its propensity to be channelled through established patterns and structures; and that recognises the complementary roles played by both living beings and non-living materials in constituting agency. The theorisation I am outlining also resists the temptation to see certain kinds of agency as intrinsically superior to others, instead recognising the ambivalence of all forms of agency.

The theorists on whose work I draw in this article do not always sit easily together, but their approaches do have certain resonances with each other, and I have tried to tune into these. There are well-known affinities between Foucault, Deleuze and Guattari, and between Bergson and Deleuze, while Bennett makes extensive use of Deleuze and Guattari’s concept of assemblage and Bergson’s vitalism. I have therefore made connections between these thinkers and their ideas, but at the same time, there is no aspiration to forge them into a coherent whole. For Bennett, in particular, the Bergsonian distinction between organic and inorganic does not go far enough in establishing agency as something immanent to matter, and I try to address this critique at the end of the article.

Rethinking children’s agency 1: power and assemblages

In the new social studies of childhood, the idea that children are active social agents became a familiar mantra from the 1990s onwards. This position worked against ingrained tendencies across research, policy and practice to treat children as intrinsically lesser beings and helped to advance the case for children’s rights (Esser et al., 2016). Initially, agency was not subject to much conceptual scrutiny in the new social studies of childhood: ‘The agency of children as actors is often glossed over, taken to be an essential, virtually unmediated characteristic of humans that does not require much explanation’ (Prout, 2011: 7).

Over the last decade or so, however, critical thinking about agency has been gathering momentum in the field:

… a counter-movement is emerging which does not reject the importance of studying children as social actors who are exercising agency but which is critical of the tendency in childhood studies to treat children’s agency in a celebratory, uncritical, a-theoretical, non-relational, locally-bound and non-reflective manner (Huijsmans, 2011: 1308).

This body of work has included attempts to critique, refine or redefine agency in various ways (Esser et al., 2016; Hackett et al., 2015; Oswell, 2013; Tisdall and Punch, 2012), including in relation to questions of scale (Ansell, 2009), children’s voices (Komulainen, 2007; Kraftl, 2013), children’s eating habits (Eßer, 2017), children living on the streets (Bordonaro, 2012; Davies, 2008), children living through armed conflict (Seymour, 2012), participatory methods (Gallacher and Gallagher, 2008; Holland et al., 2010), schools and early years institutions (Guo and Dalli, 2016; Teague, 2014), institutional practices of inclusion (Dalkilic and Vadeboncoeur, 2016) and child protection social work (Ackermann and Robin, 2016), to name just a few examples.

These debates raise the question of whether there is anything distinctive about the agency of children, as compared to human agency in general (see for example Punch, 2002). More profoundly, following posthumanist accounts of agency, some of this work begins to consider whether and how the agencies of children and humans ought to be distinguished from the agencies of other kinds of beings such as other animals, plants, technologies or minerals. To give a preliminary answer to these questions, the argument that I am developing in this article draws on post-structuralist thinking to emphasise the specificity of particular agencies as they flow within particular assemblages, and the relational quality of agency, as something that happens in the relations between the different bodies within a given assemblage. In other words, the question of whether children’s agency is distinct from that of adults is unanswerable, because children’s agency is not a property of children but an effect arising within relations between children and various other kinds of beings. Children will exercise agency in radically different ways in different assemblages. Depending on what assemblage is being analysed, agency might be observed flowing between children and adults, children and objects, children and machines, children and animals or among any combination of these and other kinds of bodies. The analysis of children’s agency, from this point of view, requires close attention to ‘where and how, between whom, between what points, according to what processes, and with what effects, power is applied’ (Foucault, 2007: 2).

These processes are what any analysis of children’s agency needs to address. The aim of this article is thus to develop some conceptual tools that can sharpen attention to how children are operating within agentic assemblages. There is no prospect, within such an approach, of making general statements about the agency of children per se. It is possible, however, to examine specific assemblages in which children play a central role – families, childcare settings, schools, public and private spaces, and so on – and the techniques, practices and relations through which, within these assemblages, children’s agency is elicited, expressed, channelled, restricted and so forth.

The concept of assemblage I am using is derived from Deleuze and Guattari (1987). It provides a more open-ended alternative to understanding children as situated in socio-cultural contexts. Assemblage is an approximate, and arguably inaccurate, translation of the French word agencement, indicating an arrangement or layout of heterogeneous elements (Nail, 2017). Assemblages, on this view, are collections of bodies in relation with one another. These bodies may include children, adults, materials, technologies, spaces, bodies of ideas or discourses – in principle, any kind of body. This kind of heterogeneity is also part of Foucault’s concept of dispositif, usually translated as apparatus, which he used to refer to networks of power relations between different bodies. Despite the Foucaultian basis of this article, I have chosen to dwell more on assemblage than on apparatus, because the latter has a specificity that is not ideal for thinking about agency. The apparatus, for Foucault, is a strategic configuration of power-knowledge that responds to some perceived social need or problem (Agamben, 2009). We can think, for example, of child development or child protection systems as dispositifs in Foucault’s sense. Assemblages can also be strategic and large scale, but equally they can be more tactical, temporary, small-scale formations. Assemblage is thus more widely applicable to any set of relations in which children are involved.

Assemblages, in Deleuze and Guattari’s use of the concept, have several key features. First, they are defined by contingent relations between a set of bodies. Second, the relations of an assemblage are in a process of continual becoming; any apparent persistence is an effect of repetition rather than sameness. Third, assemblages are not unities. Unlike organic bodies, they are not self-contained systems in which all the parts form a coherent whole, where each part has a defined and essential function without which the whole cannot exist. Rather, assemblages are multiplicities of related elements, which can be reconfigured or combined with new elements. Deleuze and Guattari repeatedly refer to assemblages as machinic, indicating this capacity for reconfiguration and coupling. Within an assemblage, different permutations and combinations of elements can produce different effects, and machinic assemblages can also be plugged into other assemblages, again producing unforeseen results. The assemblage of child will operate differently when coupled with the assemblage of family, as compared to the assemblage of school, for example, or the assemblage of neighbourhood.

Hooking up Deleuze and Guattari’s concept of assemblage with Foucault’s ideas about power has the potential to advance the critiques of children’s agency outlined above. Thinking with these concepts amplifies two issues in particular, which I will discuss in turn below: the first concerns how agency is valued in childhood studies; the second concerns how agency is attributed to individual children.

In the new social studies of childhood, there is a tendency – by no means universal, but common enough to be identifiable – to view agency as a force that can liberate children from structural constraints, and as such something that should be more fully acknowledged, valued and encouraged. This view of agency is evident, for example, in the discourse advocating participatory and co-production methods in research with children (e.g. Kellett et al., 2004). While helpful as part of an emancipatory politics, the promotion and celebration of children’s agency can sometimes verge on romanticising children’s actions, making it more difficult to see how agency is ethically ambivalent. If we understand agency as a type of micro-power, then, following Foucault (1983), it can be defined as actions that affect other actions. Such actions do not have any intrinsic ethical value. The ethics of power depend on how power is being exercised, through which bodies, through which relations, apparatuses and techniques. Power may have effects of liberation, domination, subversion, colonisation, persuasion, participation, co-operation, intensification or attenuation of conflict, and so on.

Typologies of agency often come with values attached, either implicitly or explicitly. Davies (1994), for example, distinguishes between subversive and transformative agency and implies that the latter is more valuable for empowerment. From a post-structuralist perspective, such claims are difficult to make outside of a specific empirical situation. As a form of power, agency is too ambivalent to have specific ethical functions attributed to it. The same technique of power may flip from being liberatory to being oppressive depending on how it is deployed, and within what assemblage, such that it is impossible ever to guarantee liberation. Childhood studies research has documented how discourses of agency and voice, ostensibly designed to facilitate inclusion and empowerment, may actually function in a regulatory way, enforcing norms about how children should behave, and excluding those whose bodies do not or cannot conform to the dominant model (Dalkilic and Vadeboncoeur, 2016; Komulainen, 2007; MacLure et al., 2010).

As for the issue of attributing agency to children, Foucault’s work provides compelling analyses of how human subjectivity, far from being an essential part of human nature, is produced through various technologies of power. From this perspective, agency cannot be attributed to children as a pregiven capacity. It is rather an effect of the development of a type of selfhood that enables a person to understand themselves, and be understood by others, as capable of decisive action. For Foucault, we become subjects of power by first being subject to power (Butler, 1997). This type of selfhood is especially prevalent in liberal democracies, where state power works by shaping, eliciting and harnessing the agency of subjects to regulate themselves. Foucault’s (1992) term for this mode of power was governmentality. Childhood is a life stage where subjects are in their early stages of formation, and as such are targeted with particular intensity by the apparatuses of governmentality (as evident in the apparatus of early intervention for example).

The subject, on this view, is formed through such a profound ingraining of biological–cultural–historical–material patterns that it is difficult, in some cases perhaps impossible, for selves to then break free from or operate outside of these patterns. Yet, the social and political forces of liberal humanism place so much value on the individual that there is a constant imperative to attribute agency to people: to see subjects as the main source of action and meaning, as ultimately in charge of, and responsible for, everything significant that happens. Such narratives of individual human mastery downplay the role of the assemblage, the myriad ‘memories, intentions, contentions, intestinal bacteria, eyeglasses, and blood sugar … the plastic computer keyboard, the bird song from the open window, or the air or particulates in the room, to name only a few of the participants’ (Bennett, 2010: 23). In childhood studies, part of the difficulty with agency is that it has been used in ways that attribute significant actions to children, despite the fact that in any significant action, there are always many bodies and forces at work, both internal and external to the analytical unit of ‘child’.

Foucault’s thought reminds us that claims to agency are never neutral descriptions. They are forms of power-knowledge that forge subjects. Encouraging thought to locate agency within an individual human actor is an implicitly humanist move, silencing the role of other kinds of beings, bodies and forces in making things happen. As Eßer (2017) suggests, childhood studies has been ‘too focused on an individualistic concept of agency … Starting from this criticism, there has been a shift from a substantialist and isolated view of children’s agency towards more connected and related forms of agency’. (p. 286) This shift reflects wider developments in social theory such as actor network theory, new materialism, post-humanism and vital materialism, all of which set out ontologies in which agency arises within heterogeneous assemblages, rather than being a capacity located within individual bodies. Bennett (2010) puts it like this:

… an actant never really acts alone. Its efficacy or agency always depends on the collaboration, cooperation, or interactive interference of many bodies and forces. A lot happens to the concept of agency once nonhuman things are figured less as social constructions and more as actors, and once humans themselves are assessed not as autonomous but as vital materialities. (p. 21)

It should be emphasised that this line of thinking does not discount the possibility that children might be important mediators, nodes, conduits or centres of agency. Such functions can only arise, however, within the relations of a larger assemblage:

The task becomes to identify the contours of the swarm and the kind of relations that obtain between its bits. To figure the generative source of effects as a swarm is to see human intentions as always in competition and confederation with many other strivings, for a human intention is like a pebble thrown into a pond, or an electrical current sent through a wire or neural network: it vibrates and merges with other currents, to affect and be affected. This understanding of agency does not deny the existence of that thrust called intentionality, but it does see it as less definitive of outcomes. (Bennett, 2010: 32)

These critiques raise the question of whether agency should be entirely abandoned as a concept. Ingold (2006), for instance, argues that it is a mistake to attribute agency to people or objects, as though it were a property lodged within bodies, and instead advocates an animist ontology, in which beings are understood as lines of movement. From this perspective, the world is not a static container for various objects, but a shifting meshwork woven by the lines of movement of beings, in their unfolding relations with each other. For Ingold, what matters is movement, not agency.

Yet, perhaps, agency has not yet outlived its usefulness for childhood studies, despite the problems outlined above. Adults who spend time researching, working with or parenting children often remark on how children routinely act decisively, in ways that appear to have significant effects on the assemblages of which they are part. These acts are remarkable because, compared to many other actants in the assemblages of which they are part, children have limited access to structural resources such as money, status and language, limited physical power, a relatively smaller repertoire of learned skills and so on. Despite these structural constraints, even small babies find ways to command adult attention, express preferences, refuse things they do not like, get more of what they want and constantly stretch the limits of their own capabilities. We could say that children exercise agency despite their vulnerabilities, but another way to put this would be that they exercise agency through their vulnerabilities. Vulnerability in this sense is not a lack, but following Butler and colleagues (2016), a radical openness to being affected by events, which is fundamental to the ability of life to feel, grow, change and act, but which also necessarily places bodies at increased risk of harm. The intensity of this relation between children’s vulnerabilities and their capacities for action is sociologically compelling. While children’s decisive actions always happen in relation with other kinds of beings and objects, often children appear to be key players in shaping these relations – never the only players, but certainly players worth taking seriously. In many cases, children find ways to exercise something that looks like agency despite or against the dominant orientations of the power relations within an assemblage.

Having rehearsed some of the difficulties with the concept of agency in childhood studies, and suggested that the concept of assemblage might provide a helpful way of rethinking agency, the next section of this article develops this rethinking further, drawing on Elizabeth Grosz’s Bergsonian account of subjectivity and freedom. For childhood studies, I argue that these ideas are helpful in suggesting that (1) agencies may be more routine or more inventive, and that these are very different tendencies, and (2) agency can be understood as arising in the margin of indeterminancy produced through relations between the organic and the inorganic.

Rethinking children’s agency 2: freedom and materiality

Writing about feminism, materialism and freedom, Grosz states that her interest is not in articulating how women can become ‘free from’ external forms of power such as patriarchy, but rather in how women might intensify their ‘freedom to’ act in ways that make the world different. This thinking can be transposed to help think about children, focusing attention not on how to free them from the structures of adult oppression, how to ‘give them’ power, rights and a voice, but on how children can and do exercise forms of freedom. The concepts of agency and freedom, while related, are clearly not interchangeable, and it is freedom in which Grosz appears to be most interested. As I will show later in the article, however, how Grosz defines freedom can help to distinguish between agencies that are more routine and those that are more inventive. Her discussion of freedom is thus helpful for refining the analysis of agency.

Grosz draws heavily on Bergson to make three key arguments. The first is that freedom is best understood not as a property of subjects but as a quality of acts. For Bergson, free acts are those that spring from the subject and express the whole of the subject. Such acts ‘are integral to who or what the subject is’ (Grosz, 2010: 144), but for Bergson – as for Foucault, Deleuze and Guattari – subjects are always becoming, so free acts cannot express the identity of a coherent, stable self. Rather, free acts are precisely those acts which can be appropriated by, and incorporated into, subjective processes of becoming. If we define agency as this process of free actions, flowing through subjects, whose effects of change are then ‘retroactively integrated into the subject’s history and continuity’ (Grosz, 2010: 146), then agency is not just an event in which a becoming subject acts – it is the very event of that becoming itself, an event that changes both the acting subject and the world: ‘Free acts are those which both express us and which transform us, which express our transforming’ (Grosz, 2010: 146).

Grosz emphasises that this kind of freedom is not about choice between a range of alternative options that are known in advance and then selected by an agent. Bergson rather thinks of freedom in a way that is similar to Nietzsche’s idea of the will to power, as the intensive quality by which certain acts make a difference, rather than the movement of choosing from among different options:

It is not a freedom of selection, of consumption, a freedom linked to the acquisition of objects but a freedom of action that is above all connected to an active self, an embodied being, a being who acts in a world of other beings and objects. (Grosz, 2010: 147)

In this line of thinking, it is less that subjects exercise agency and more that agency is the process by which subjects are transformed. This idea can be expanded on via Foucault’s (1984) work on what he referred to as techniques or technologies of the self, defined as:

… those intentional and voluntary actions by which men (sic) not only set themselves rules of conduct, but also seek to transform themselves, to change themselves in their singular being, and to make their life into an oeuvre that carries certain aesthetic values and meets certain stylistic criteria. (pp. 10–11)

This idea of subjects shaping themselves may seem like a radical departure from Foucault’s earlier work on how subjects are formed by apparatuses of power external to them, but Butler (1997) describes how these two forms of power are in fact intimately linked to a fundamental ambivalence of agency: that although the agency of the subject presupposes the subject’s subordination, this agency exceeds the subordinating power which enabled it:

Power considered as a condition of the subject is necessarily not the same as power considered as what the subject is said to wield. The power that initiates the subject fails to remain continuous with the power that is the subject’s agency. (p. 12)

Indeed, many of the subjective capacities that neoliberal apparatuses of biopower try to drum into children – the ability to take responsibility for their actions, comply with rules, self-regulate and self-manage – are aimed at producing subjects who co-operate by governing themselves. In so doing, however, these processes of subjection produce subjects who are also well-equipped to transform themselves in ways that are not tied to the agenda of the governmental apparatus. It is this latter kind of self-transformative agency that, for Foucault – and I think for Grosz too – constitutes our best hope of freedom.

The second key feature of Grosz’s (2010) Bergsonian account of freedom is that free acts are far less common than the everyday forms of agency that are often described in the childhood studies literature:

Freedom is thus the exception rather than the rule in the sense that it can function only through the ‘autonomy’ of the living being against a background of routinized or habituated activity. It is only insofar as most of everyday life is accommodated through automatism, by a kind of reflex or habit, that free acts have their aesthetico-moral force and their effects on their author or agent. (p. 148)

In this respect, Bergson’s notion of freedom is significantly different to the techniques of the self, which operate precisely by modifying habits. In the case of Greek care of the self, Foucault (1986) shows how these techniques involved daily work on mundane matters such as diet, exercise and regimen. Similar ideas can be found in Nietzsche (2001), for whom exercising the will in a way that shapes one’s own character requires ‘long practice and daily work’ (p. 164), and Lefebvre (1991, 2004), who sees repetitive everyday actions as an important and often overlooked way through which change can happen. Such everyday agencies can be contrasted with Foucault’s (2001) account of parrhesia, the Greek term for the practice of speaking frankly and truthfully in circumstances where it is risky to do so. Parrhesia is not everyday speech but a speech act charged with an unusual intensity of courage, which as a result has the potential to transform both the speaker and the interlocutors, and as such is closer to what Grosz, after Bergson, means by freedom.

These ideas could all be analytically useful in childhood studies, to help differentiate between different kinds of agencies, relations and assemblages. In particular, it might be beneficial to distinguish between the following:

1. Routine agencies that function through norms, patterns of response, reflexes, habits and skills. Such routine agencies happen wherever bodies or forces act in ways that develop or reproduce conventions, comply with rules or follow patterns. Routine agency is not slavishly deterministic; even with ingrained patterns of action, for something to constitute agency, there needs to be room for a degree of indeterminacy and improvisation. But routine agency never wholly contravenes, overturns or reinvents established rules.

2. More exceptional forms of agency that are animated by the kind of freedom that Grosz is interested in. These inventive agencies break established patterns and habits, redirect forces and significantly reconfigure relations. They are unexpected eruptions that disturb the status quo, usually in situations where there is something at stake or an element of risk. Bodies through which this kind of inventive agency flows are transformed by it, and in turn transform the assemblages in which the body participates.

No doubt, there are agencies that do not fall neatly into either of these two categories; they might be better understood as different tendencies rather than discrete categories. There is some similarity here with Klocker’s (2007) idea of thin and thick agencies, but those concepts relate to the extent of the range of choices available to children. Recall that for Bergson, freedom is not a question of choice but rather of the intensity of indeterminacy in a given assemblage. It is this intensity of indeterminacy that may be helpful in distinguishing between different kinds of agency. It is also important to avoid setting up a value hierarchy, with inventive agency seen as intrinsically superior to routine agency. For example, Guo and Dalli (2016) show how in a New Zealand early years setting, ‘small’ agencies, based on fitting in with routines and established language structures, enabled immigrant Chinese children to develop a sense of belonging. All kinds of agencies may serve all kinds of ethico-political functions. The aim of distinguishing between different kinds of agency is not to privilege some forms over others, but to provide better analytical purchase on the multitude of forms of action often lumped together under the rubric of agency.

To show how this distinction between routine and inventive agencies might be useful for empirical work in childhood studies, take the example of recent research showing that young children are capable of becoming proficient and enthusiastic users of digital technologies such as iPads (e.g. Flewitt et al., 2015; Lynch and Redpath, 2014). Within the new social studies of childhood paradigm, such findings would appear to be evidence of children’s agency. Yet, the devices in question have been deliberately engineered to capture human attention so as to generate surplus value. When children follow the routine functionality built into iPads and other consumer technologies, their actions are not free in Bergson and Grosz’s sense of the term. The technology is eliciting and shaping children’s action in a way that habituates them into its wider programme – for example, by producing subjects who willingly generate data that can be used for algorithmic processing, to serve the functions of what Deleuze (1992) calls societies of control. Grosz’s account enables us to distinguish clearly this kind of agency from transformative forms of agency. To repeat my earlier argument about the ethico-political ambiguity of agency: this is not to denigrate routine agencies, which may have both beneficial and harmful effects. Nevertheless, it is analytically and politically useful to be able to distinguish these routine agencies from events in which less scripted forms of action take place, breaking out from the designed parameters of technology.

The inventive tendency of agency can be seen on the more rare occasions when children and young people repurpose assemblages in unforeseen ways. In relation to mobile technologies, one example is the ‘Teen Buzz’ ring tone that appropriates the sound of Mosquito devices. The Mosquito is a sonic technology used to discourage young people from loitering in public spaces. It works by generating high-frequency buzzing that can produce discomfort. High-frequency noise disproportionately affects younger people, whose hearing range typically extends higher than that of older people as a result of age-related hearing loss (Saladin, 2014). The device thus plays on auditory physiology to exercise spatial power, circumscribing territory in a way that explicitly and deliberately enacts age-based discrimination. But the sound was also infamously repurposed as a ring tone for young people’s mobile phones, nicknamed ‘Teen Buzz’, enabling them to detect incoming text messages in a way that is less audible to adult surveillance, in school classrooms for example. The ring tone can be heard – or not heard, depending on the condition of your auditory physiology – at the Mosquito Ringtones website: http://www.freemosquitoringtones.org/.

Teen Buzz is an inventive and ingenious misuse of technology. Again, it is important to emphasise that these qualities do not make it intrinsically ethically ‘good’; it could easily be used to help propagate bullying and abusive messages. The point is that the appropriation of the Mosquito sound represents a greater degree of freedom in the sphere of technology than routine use of devices. It plugs together two technological assemblages – the mobile phone and the Mosquito Device – to actualise previously hidden potentials within them.

The third key argument in Grosz’s account of Bergson is that freedom arises from relations between the organic and the inorganic, between life and matter. For Bergson, the organic and the inorganic are related tendencies rather than separate realms – Bennett (2010: 76) refers to them as strivings, propensities and leanings that only exist in relation to each other. The tendency of organic life is characterised by what Bergson called élan vital, which is a drive towards movement, activity, change, surprise and the division of life into new forms: ‘The task of élan vital is to shake awake that lazy bones of matter and insert into it a measure of surprise’ (Bennett, 2010: 78). The inorganic is the opposite tendency, a drive towards stability and durability, providing life with enough regularity to persist, to anchor itself and ‘perform habitual actions with a measure of some guarantee of efficacy’ (Grosz, 2010: 151). In Deleuzian terms, we could say that the organic is what produces difference while the inorganic produces repetition – with difference and repetition understood as mutually constitutive rather than opposed. An obvious example relating to childhood is the way that a school classroom, as a formation of inorganic matter, can persist over time in a relatively stable form, while the children who live through that classroom, appropriating its materials for learning, grow, change and move on.

Again, on this account, freedom is not a property that resides in subjects or indeed in any kind of being, object or body. Freedom rather arises through the relations between living beings and materials, insofar as these relations produce indeterminacy. The inorganic materials of the universe contain within them the virtual potential for indeterminacy, and organic life actualises these virtual potentials by appropriating and reworking materials in ways that maximise indeterminacy. This movement of indetermination, for Bergson, is the force of life. Life in turn renews inorganic matter and its potential for indeterminancy though posthumous processes of decomposition, the organic always eventually returning to the inorganic from which it arose. Returning to the different tendencies of agency discussed above, more routine forms of agency stick closer to the contours of the inorganic, working within a narrow margin of indeterminacy, while more inventive forms of agency push matter further, yielding a greater degree of indeterminacy.

Bennett (2010) argues that while Bergson’s thinking here points in the right direction, it does not go far enough, because it continues a long tradition of rendering matter inert. She finds this framing insufficient for developing a politics that recognises the power of things and the distribution of agency throughout heterogeneous assemblages. Without getting into a detailed conceptual discussion of this criticism, I want to suggest that the organic/inorganic distinction may be useful in childhood studies for analytical reasons, particularly when it comes to engaging children, practitioners and childhood studies students in thinking about their relations with the material world in ways that might initially seem counter-intuitive. If we want to rework childhood studies’ analyses to argue that, for example, children’s eating is shaped by the agencies of food (Eßer, 2017), or that groundwater is an agent that affects children’s relations with place (Horton and Kraftl, 2018), then starting with a Bergsonian distinction between the inorganic and the organic – understood as tendencies in productive tension with each other, rather than mutually exclusive categories – might help to make sense of the fact that there are clearly different kinds of agency involved in these situations, and that these agencies may not be straightforwardly comparable.

Latour’s way of dealing with this problem is to describe nonhumans as actants rather than agents (e.g. Johnson, 1988), while Ingold’s (2006, 2011) solution is to let go of the concept of agency in favour of animacy, and let go of the concept of objects in favour of things. These alternative concepts are more capacious than agency in encompassing multiple forms of action flowing through multiple kinds of bodies, materials and forces, but they come with the risk of reducing analytical specificity. Starting with the assumption that everything is animate, that everything is acting or moving, is not particularly helpful for distinguishing between different kinds of action or movement. Bergson’s distinction between organic and inorganic tendencies offers an orientation that neither sets humans apart from all other beings nor completely flattens out ontology. A focus on the relations between the organic and the inorganic allows a recognition of the significant differences between these two tendencies, while also acknowledging their mutually constitutive relations, and without privileging one over the other. From this perspective, it is through the relations between the organic and the inorganic that agency happens, and thus it is to these relations that analyses of children’s agency ought to attend.

Conclusion

In this article, I have attempted to rethink agency in a way that responds to critiques of the concept in childhood studies. Informed by Foucault’s accounts of power, Deleuze and Guattari’s concept of assemblage, Bennett’s vital materialism and Grosz’s account of freedom in Bergson, I have set out a version of agency that tries to intensify the specificity of the concept, rather than flattening it out into something more general such as movement, animacy or action. I have made four arguments:

1. Agency is ambivalent. It does not have intrinsic ethical value. The ethics of agency can only be determined by examining specific instances of specific kinds of agencies – how they are being exercised, through which bodies, through which relations and assemblages and with what effects.

2. Attributing agency to individual children obscures more than it reveals. Agency is better thought of as a quality of acts that happen within heterogeneous assemblages. This line of thinking does not discount children’s agency or intentions. Rather, it seeks to analyse these things as forces that emerge from, and contribute to, larger assemblages.

3. It may be helpful to think of different tendencies of agency, from more routine to more inventive. Routine agencies function by establishing or reproducing patterns of response, as when children follow conventions, rules or habits. Inventive agencies are more unexpected, erupting in ways that break established patterns, redirect forces and reconfigure relations. They transform both the bodies through which they flow and the assemblages in which those bodies participate. In keeping with principle 1, neither of these kinds of agency can be considered intrinsically superior. Both can serve all kinds of ethico-political functions.

4. The distinction made by Bergson between the organic and the inorganic helps to displace humanistic accounts of agency, while still differentiating between life and matter. Again, the organic and the inorganic are to be understood as tendencies rather than discrete categories. Organic life tends towards movement and transformation, seeking to maximise indeterminacy through its relations with materials. Inorganic matter tends towards stability and durability, enabling life to persist. The two tendencies are always related and, once again, neither should be privileged over the other.

I offer these principles not as an attempt to finally answer the question of what agency is in childhood studies but in precisely the opposite spirit: to throw some additional ideas into the mix of debates about children’s agency, to help shift attention onto what agency does, how it flows in relations, and to provide analytical resources for exploring children’s relations with other kinds of bodies, forces and materials.