Theories of power and social change. Power contestations and their implications for research on social change and innovation

ABSTRACT

This paper proposes a meta-theoretical framework for studying power in processes of change and innovation. Power is one of the most contested concepts in social and political theory. This paper discusses seven prevailing points of contestation: Power over versus power to, centred versus diffused, consensual versus conflictual, constraining versus enabling, quantity versus quality, empowerment versus disempowerment and power in relation to knowledge. The paper reviews how different scholars have dealt with abovementioned points of contestation and identifies how different theories of power can be translated into specific empirical questions to systematically explore power in processes of social change and innovation.

KEYWORDS:

• Power
• social change
• sustainability transitions
• social innovation

1. Introduction

There is an increasing attention for processes of social change towards more sustainable and just societies. This attention is manifested in various interdisciplinary research fields, including sustainability transitions research (e.g. Geels & Schot 2007; Grin et al. 2010, Markard et al. 2012, Loorbach et al. 2017) and social innovation studies (e.g. Moulaert et al. 2013, 2017; Cajaiba-Santana 2014; van Moore et al. 2015, van der Have and Rubalcaba 2016, Westley et al. 2016, Avelino et al. 2017, 2019, Pel et al. 2020). These fields of research on social change – and their respective conferences, journals and conceptual models – co-evolve with government policies at various levels (municipalities, national departments, European Commission, United Nations, etc.). While they differ in their language and foci, they share an interest in change and innovation as drivers for societal improvement, and an (implicit or explicit) belief in human knowledge and agency to change the world for the better.

In this enthusiasm-for-social-change, the ‘dark’ and ‘unintended’ effects of social change and innovation often tend to be underemphasised, as well as the fierce power struggles and inequalities that come with it. In response, several scholars have set out to include an explicit attention for power and politics in their discussion of innovation and social change (e.g. Swyngedouw 2005; Voß et al. 2009, Moulaert et al. 2007, Meadowcroft 2009, Avelino and Rotmans 2009, 2011, Grin 2010, Smith and Stirling 2010, Kern 2011, Hoffman 2013, Hess 2013, Geels 2014, Scoones et al. 2015, Boonstra 2016, Brandsen et al. 2016, Avelino et al. 2016, Avelino & Grin 2017, Ahlborg 2017, Teasdale et al. 2020). These endeavours demonstrate various challenges for more profoundly couching research on social change and innovation in social and political theories of power. This paper discusses these challenges and sets out to broaden and deepen the theoretical basis for studying the implications of power for research on social change and innovation.

Power is one of the most contested concepts in social and political theory. Definitions are manifold and highly diverse, ranging from power as actor-specific resources used in the pursuit of self-interests (Weber in: Fuchs 2001) to power as the capacity of a social system to mobilize resources to realize collective goals (Parsons [1967]2002). This is why Lukes ([1974]2002, p. 45) contended that power is an ‘essentially contested concept’, one of those concepts which ‘inevitably involve endless disputes about their proper uses on the part of their users’, also adding that ‘to engage in such disputes is itself to engage in politics’. According to Haugaard (2002, p. 3), however, power is not so much an ‘essentially contested concept’ but rather a ‘family resemblance concept’¹. Any attempt to capture the ‘essence’ of the word will exclude aspects that might be essential in a given context. Rather than trying to capture the essence of power in one, all-encompassing definition, the challenge is to construct a local language that is suitable in a specific context.

Indeed, the challenge of this paper is to present a conceptual power language that can be used in the context of social change research. This is easier said than done, because it seems that the concept of power is more often applied to explain a lack of change rather than change itself. There are a number of exceptions, where power is conceptualised as the capacity to change and ‘to act otherwise’, as eloquently formulated in the following definition of power: ‘In its most general sense, power is (…) the “can” which mediates the desired or intended outcomes of social actors and the actual realization of these outcomes in their daily social practices’. (Davis et al. [1991] 2002, p. 214). However, the large majority of explicit power theories tend to privilege stability over change. Even more agent-based theories of power can be unsatisfactory in terms of conceptualising (the possibility for) change (Stewart 2001:16). Giddens, for example, characterized power as being ‘generated in and through the reproduction of structures of domination’ (Giddens [1984]2002, p. 160). Even though Giddens has often been either applauded or criticized for privileging agency, the author’s interpretation of power remains in fact narrowly defined in terms of dependence on and domination of structures. As Stewart (2001:16, emphasis added) expresses it: ‘in spite of Giddens’ formal commitment to possibilities of “making a difference”, it effectively makes power a function of the distribution of resources, subject only to actors’ capabilities to draw upon such resources effectively (…) [Giddens specification of power] makes socially transformative capacity substantially dependent upon “existing” structures of domination’. The privileging of stability over change is also manifested in multiple power typologies that are either resource based (e.g. Mann [1986]2002) or based on a vertical distinction between different levels of aggregation (e.g. Clegg [1989]2002). These typologies privilege stability over change, in the sense that they often focus on (the distribution of) existing resources, and/or on the relation between actors and existing structures, at a specific point in time.

This can be related to the lack of a dimension of time in much of social science theory. The lack of the temporal dimension in social theory has been addressed by various authors, who claim that ‘political science in particular, but also social sciences more generally, have become increasingly decontextualized’ and that ‘a prime form of this decontextualization was the loss of an explicit theoretical treatment of time’ (Pollit 2008:7). If time is ignored, it logically follows that stability is privileged over change, for static moments in time are by definition ‘stable’. When taking the dimension of time into account – as is inherent to social change research (or at least should be) – the occurrence of change (and novelty) becomes a fact. This is not to say that change is to be privileged over stability, rather that they should (initially) be treated equally. To what extent stability supersedes change – or vice versa – should be an empirical question, rather than being precluded in a theoretical conceptualization of power.

This paper proposes to critically reconsider how existing power theories relate to change and to formulate empirical questions on the relations between power and processes of social change and innovation. This is done by first carefully considering the different ‘family members’ of the power concept, as discussed in social and political theory. Rather than making a futile attempt to provide an overview of all power interpretations, the paper discusses seven prevailing points of contestation in academic debates on power:

1. Power ‘over’ versus power ‘to’

2. Centred versus diffused

3. Consensual versus conflictual

4. Constraining versus enabling

5. Quantity versus quality

6. Empowerment versus disempowerment

7. Power = knowledge versus power ≠ knowledge

As the essence of power cannot be captured, it follows that the essence of its contestation can also not be fully captured. The dichotomies mentioned above overlap with one another, and can thus not be perfectly distinguished, nor can different theories of power be fully categorized in these terms. Several perspectives on power attempt to overcome at least one of the abovementioned dichotomies; by pointing out a third issue beyond a bipolarity (e.g. Lukes), by turning a dichotomy into a duality (e.g. Giddens), or by integrating them as different dimensions in a multi-levelled framework (e.g. Clegg).

For each of the above points of contestations, this paper synthesises how different scholars (e.g. Arendt [1979]2002, Parsons [1967]2002, Lukes [1974]2002, Giddens [1984]2002, Clegg [1989]2002, Foucault [1977]2002, 1980, 1982, Thomas and Velthouse 1990, Flyvbjerg 1998, Haugaard 2002) have dealt with the abovementioned points of contestation, what we can learn from them, and, most importantly, what these points of contestation imply for understanding and empirically investigating the role of power in research on social change processes. Rather than ‘choosing sides’ within these power debates or attempting to ‘solve’ them, the challenge is to acknowledge the different dimensions of these power contestations, and on that basis distil empirical questions that can be used to systematically and meaningfully explore the role of power in processes of social change and innovation, remaining sensitive to various dimensions of power as discussed in the literature.

The structure of this paper is as follows. Each of the abovementioned seven power contestations is introduced in a separate section. Each section starts with a discussion of how the contestation relates to power theories in the literature, followed by a short reflection on implications for research on social change and innovation, including the formulation of empirical research questions to be asked about power in processes of social change and innovation. For the sake of readability, the fields of research on sustainability transitions and social innovation are referred to as (research on) ‘social change and innovation’ or just ‘change’ in short.

2. Power ‘over’ versus power ‘to’

Even if one holds that power is exercised rather than possessed (Foucault [1977]2002), the question remains what is exercised; is it a capacity ‘to’ act and achieve something, or is it a social relationship in which A exercises power ‘over’ B? In this regard Morriss points out that power is derived from the Latin word potere – ‘to be able’ – and claims that in philosophical and linguistic terms, power ‘is always a concept referring to an ability, capacity or dispositional property’ ([1987]2002, p. 283). According to Morriss, ‘everything that needs to be said about power can be said by using the idea of the capacity to effect outcomes’ (Morris [1987]2002, p. 299). This definition of power as a capacity to act, does not exclude conflictual or violent dimensions: ‘if we are interested in the “conflictual aspect” of power, we can very easily look at someone’s power to kick others around, or their power to win conflicts’ (ibid). Nevertheless, theories that focus on power as a capacity (as in e.g. Parsons and Arendt) are criticized for ignoring the relational or oppressive aspects of power ‘over’ others (Lukes [1974]2002), or for ‘fail[ing] to account for individuals or groups in the community who, though they do not exercise power, nonetheless have power, in the sense that many people try assiduously to anticipate their reactions’ (Dahl [1968]2002:20 in reference to Bachrach and Baratz 1962).

The previous citation points to another question, i.e. to what extent power exists before its exercise; does an actor already have power if he or she has the potential to exercise it, or can an actor only be considered powerful once he or she actually exercises power? According to Barnes, power is both ‘a potential or capacity which may or may not be used’, as well as something that ‘is possessed’ ([1988]2002, p. 125), the relevant point being that power always ‘resides in the social context and outside its possessor’ (ibid:127). Or as Clegg puts it; people ‘possess power only in so far as they are relationally constituted as doing so’ (Clegg [1989]2002, p. 257). The same could be said about the exercise of power; that it resides in the social context and outside of its exerciser, and that people only exercise power in so far as they are relationally constituted in doing so. Thus it seems that power ‘over’ and power ‘to’ are not mutually exclusive; both can be ‘possessed’ and ‘exercised’, and both are ‘relationally’ constituted in some way or another.

Some authors have proposed to nuance and elaborate the distinctions between ‘power over’ and ‘power to’, by e.g. elaborating power with as a third type, distinguishing between power over (coercion and manipulation), power to (resistance and empowerment) and power with (cooperation and learning) (cf. Partzsch 2015). The argument made is that ‘there are situations in which power is neither attributed solely to A nor to B, but to both’, however that ‘power with is not exercised independently from power over dimensions’, and that we need to analyse how these different types of power are intertwined so as to remain attentive to conflicts of interests and values that often remain invisible or hidden (Partzsch and Fuchs 2012).

2.1 Power ‘to’ versus power ‘over’: implications for research on social change and innovation

While ‘power over’, ‘power to’ and ‘power with’ refer to different perspectives on power, they can also be understood as different manifestations of power. An important starting point for understanding change is acknowledging how these different manifestations of power can play a role in processes of change. For instance, A and B can contribute to or resist change (power to), and/or A can coerce or hamper B to change (power over), and/or A and B can cooperate for or against change (power with). Moreover, all three manifestations of power can be objects of change, in the sense that change and innovation initiatives can explicitly strive to challenge certain forms of oppression and domination (power over), and/or strive to empower people through capacity building (power to) or collaboration (power with). Relevant empirical questions to be asked about social change and innovation are:

• Power over: Who is exercising power over whom? How are which structures of domination/oppression/dependence changed or (re)produced?

  • How does change (re)construct (new) structures of domination, oppression and dependency? Which existing dependencies, dominations and oppressions are reinforced?

  • How and to what extent do change initiatives aim to overcome/resolve/improve existing structures of domination? How and to what extent are existing power dependencies challenged and ‘independencies’ created?

• Power to: How is power exercised for/through/against change?

  • How do actors exercise power to enable or resist change?

  • (To what extent and how) does the change contribute to capacity-building in terms of power to?

• Power with: How do actors collaborate in the exercise of power for/against change?

  • What are the power coalitions that enable/resist change?

  • (How and to what extent) are existing power coalitions (re)produced/challenged by/for/through change?

3. Centred versus diffused

Another classical debate on power is the one between ‘pluralists’ and ‘elitists’. One side emphasized that elites possess power over society, while the other side stressed that political power concerns a struggle between plural interest groups. While Dahl ([1968]2002] criticized the ‘ruling elite model’ by pointing out that political power comes from broad decision-making processes, Bachrach and Baratz (1962) referred to the ‘second face of power’ to emphasize how elites are capable of predetermining agenda-setting before and outside the open process of decision-making, for instance by keeping certain issues off the agenda (also referred to as ‘non-decision making’). In addition, Lukes ([1974]2002) introduced a ‘third face of power’, referring to processes of preference-shaping. Therein certain groups shape the interests and preferences of other groups, as such not even having to keep issues ‘off the agenda’, as these issues are prevented from emerging in people’s minds in the first place.

This debate relates to Foucault’s notions that the hierarchical, centralized notion of power as a King would have it or any other judicial-political entity is an outdated legacy from the past, and to Mann’s distinction between authoritative power and diffused power. While authoritative power ‘comprises definite commands and conscious obedience’, diffused power ‘spreads in more, spontaneous, unconscious, decentred ways throughout a population, resulting in similar social practices that embody power relations but are not explicitly commanded’ (Mann in: Stewart 2001:25). An essential trait of diffused power is ‘normalization’, i.e. the belief that certain practices are ‘moral’ or in the ‘common interest’, which relates back to Lukes’ preference-shaping, and to various discursive interpretations of power as found in Foucauldian analyses.

3.1. Centred versus diffused power: implications for research on social change and innovation

Both academic and public discourses on social change and innovation involve various empirical cases as well as conceptual notions of diffusion or decentralisation of power. Such decentralisation is often implicitly or explicitly assumed to be ‘good’ and to lead to more ‘equal’ or ‘just’ power relations. Examples are renewable energy innovations and the associated (possibilities) for decentralisation of energy production, which may enable community-led energy cooperatives. Theories of power can be useful to question how, to what extent and under which conditions, such change and innovation contribute to a decentralisation of power relations.

However, power theories also provide insights on how processes of power decentralisation can also involve the (re)construction of old or new power inequalities and oppressions (e.g. if and when community-led energy cooperatives primarily involve affluent actors with a high social capital, possibly – and often unintendedly – excluding others). Moreover, power theories help remind us that social change and innovation processes do not necessarily strive for the decentralisation of power. On the contrary, they can also manifest in a ‘re-centralisation’ or ’re-enforcement’ of centralised power (e.g. when renewable energy is co-opted by existing energy multinationals). As such, it is particularly pertinent to study the following empirical questions about power in social change and innovation processes:

• (How) are the three faces of power manifested in processes of change?

• How and to what extent is power diffused, (de)centralised and/or recentralised by/for/through change?

• Are there ‘ruling elites’ or ‘centres of power’ that hamper and/or enable change, and if so, who/where are they? Which other groups are involved, and who is excluded?

• How and by whom is the agenda of change decided? Which issues are kept off the agenda?

• How are preferences and interests underlying processes of change shaped? What are related processes of normalization?

4. Consensual versus conflictual

Debates on power often revolve around the question whether power is consensual or conflictual (Haugaard 2002). This relates to the question whether power is distributive or collective. In the distributive model, power is ‘zero-sum’, i.e. gained by one actor at the cost of another actor. In the collective model, actors can enhance their joint power, as is the case in Parsons’ earlier mentioned definition of power as the capacity of a societal system to achieve collective goals (Parsons [1967]2002, p. 93), or in Arendt’s interpretation of power as ‘the human ability not just to act but to act in concert’ 1979]2002, p. 137). These models are ‘consensual’ in the sense that both Parsons and Arendt position consensus as a necessary condition of power. Parsons compared power to money, claiming that its meaning can only survive as long as society supports it, and that power diminishes when it is used illegitimately (similarly to processes of inflation). According to Arendt, violence can destroy power, but ‘is utterly incapable of creating it’, and ‘power and violence are opposites; where the one rules absolutely, the other is absent’ 1979]2002, p. 143). This starkly contradicts with Mann’s characterization of violence as ‘the most concentrated, if bluntest, instrument of human power’ ([1986]2002, p. 177).

Distinguishing violence from power does not necessarily take away the conflictual, physical, or oppressive dimensions of power. Quite on the contrary, the ability of oppressing without blunt violence is regarded by some as the essential characteristic of power. As Foucault puts it: ‘subjection is not only obtained by the instruments of violence or ideology; it can also be direct, physical, pitting force against force, bearing on material elements, and yet without involving violence; it may be calculated, organized, technically thought out; it may be subtle, make use neither of weapons nor of terror and yet remain a physical order’ (Foucault [1977]2002, p. 191). According to Foucault, ‘power is a form of pacification which works by codifying and taming war through the imposition of particular knowledge as truth’ (ibid: 185). This resonates with Lukes’ preference-shaping, which challenges the Weberian premise of power as influence in spite of resistance. The capacity to make resistance dissolve – by shaping preferences – is understood as a decisive moment in the exercise of power. The distinction between ‘consensual’ and ‘conflictual’ power is therefore a tricky one, as power is characterized by some as the ability of blurring this distinction, either by turning a conflictual situation into a consensual one, or by preventing conflict from emerging in the first place. Here it is also important not to equate ‘conflictual’ with ‘power over’ or ‘consensual’ with ‘power to’. As Haugaard (2012, p. 37) points out:

“The idea that in principle not all exercises of power over constitute domination shifts the line between the consensual and conflictual traditions. It has generally been assumed that consensual traditions concerned power to, while the conflictual, power over. Thus, power over equates with domination. What always was slightly anomalous in this interpretation was that the main consensual theorists (Parsons 1963, Arendt 1970, Barnes 1988, Searl 2007) clearly see themselves as writing about power over, as well as power to”.

4.1. Consensual versus conflictual: implications for research on social change and innovation

The most important lesson to be learnt from the power literature is to be aware of conflicts that may be ‘hidden’ behind seemingly consensual processes, but also the other way around, to acknowledge the consensual forces that in the end may give rise to conflict. On the one hand, social change and innovation research should pay (more) attention to conflictual (elements of) social change. At the same time, social change research should also be aware that (1) power struggles and oppression are also manifested in (seemingly) consensual processes, and (2) conflictual processes are not necessarily oppressive and may help to challenge structures of domination and oppression. Or in other words: the ‘dark sides’ of social change and innovation have various layers. These ‘dark sides’ do not necessarily refer to the conflictual aspects: conflict also has bright sides, and consensus also has dark sides. Empirical questions to ask about social change and innovation include:

• How are both consensus and conflict manifested in processes of change?

  • To what extent is power for/against change consensual and/or conflictual?

  • How is the consensus and/or conflict on power for/against change produced?

• (How) are conflicts and contestations ‘hidden’ under seemingly consensual processes and vice versa: (how) is there consensus ‘hidden’ under seemingly conflictual processes?

  • How and to what extent are conflicts/contestations being obscured or pacified?

  • How and to what extent are consensus/coalitions being obscured or disrupted?

• How and to what extent are consensus and conflict experienced as oppressive and/or emancipatory?

  • How and to what extent is the consensus for/against change oppressive?

  • How and to what extent is the conflict for/about change emancipatory?

5. Constraining versus enabling

Power plays an important role in the agent-structure debate. The point of contestation is whether power lies mostly on the ‘agent side’ (as that which enables actors to make a difference), or on the ‘structure side’ (as that which predetermines and constrains the behavioural options of actors). Within certain debates, ‘power and structural constraint are theorized as opposite ends of a continuous spectrum. At one end of the spectrum social relations are contingent (…) whereas at the other they are determined (…) at the contingent end there is power (A could have acted differently) and, at the determined end, there is structure (A had no possibility of acting differently)’ (Haugaard 2002:38, emphasis added). For some, power is inherently agent-centric, as formulated by Lukes: ‘Human agents, whether individuals or collectivities, have power or are powerful within structural limits, which enable and constrain their power. (…) It they are so structurally constrained or determined that they are unable to act otherwise than they do, then they are powerless to do so, and so they are powerless, not powerful’ (Hayward and Lukes 2008, p. 12). In contrast, Foucault has analysed power as an inherently non-subjective phenomenon that it is exercised by structures and through actors, contending that ‘individuals are the vehicles of power’ (Foucault 1980, p. 101). Foucault demonstrates how certain mechanisms ‘automize’ and ‘disindividualize’ power and how a material or ideological structure can be used for ‘creating and sustaining a power relation independent of the person who exercises it’ ([1977]2002, p. 196).

Giddens criticizes Foucault for not relating power ‘to a satisfactory agency and knowledgeability as involved in the “making of history”’ (Giddens [1984]2002, p. 160).² Giddens’ own theory of structuration aims to overcome the polarity between structure and agency, by theorizing how structures are both enabling and constraining, and how agents make use of these structures in their daily practices, power being the capacity of agents to draw on these structures to achieve outcomes (Giddens [1984]2002). Agency may be human but according to Clegg this is not necessarily the case; agency may be organizational rather than human, and it can also be exercised by a computer decision-making system (Clegg [1989]2002:250, 270–71). With his ‘three circuits’ of power, Clegg offers a multi-levelled model to theorize power as a complex interplay between agency, rules of the game at the organizational level, and structures of domination at the societal system level (consecutively referred to as ‘relational’, ‘dispositional’ and ‘structural’ power).

5.1. Constraining versus enabling: implications for research on social change and innovation

The main implication for social change research is to acknowledge different manifestations of power as being both enabling and constraining. First, the constraining of one group of actors usually involves the enabling of other actors, and vice versa. Therefore, whether a particular act of power can be characterized as enabling or constraining in relation to change and innovation, depends on the empirical starting point and taken perspective in a given analysis. For instance, while the invention of a new surveillance technology may be ‘constraining’ for a certain group of actors, it does nevertheless ‘enable’ another group of actors to implement order and surveillance for a given purpose. Moreover, there might be a certain willing compliance of actors to be constrained (see Haugaard 2012). The two interpretations most common in social change research is that structural power (power exercised by/through structures) is (1) an object of social change (i.e. structure is what is to be innovated/transformed) or (2) a constraint for social change. However, structural power can also be an enabler for social change. Examples are the introduction of a new legal form that recognises the social enterprises as a legal entity in and of itself that combines for-profit and non-profit logics and is taxed accordingly. Once implemented and enforced, this can be understood as the structural exercise of power that enables change and innovation. This also means that one should avoid entirely equating ‘structure’ with ‘constraint’, and ‘agency’ with ‘enablement’. On this basis, I propose the following empirical questions to be asked about change and innovation:

• How are both structure and agency manifested in processes of change?

  • How and which human and non-human agents and structures in social change are (perceived to be) exercising power?

• Who/what is enabled and who/what is constrained in/for/by the process of change, how and by whom or what?

  • How and which human and non-human agents are (perceived to be) enabled by what exercise of power?

  • How and which human and non-human agents are (perceived to be) constrained by what exercise of power?

• How/to what extent is structural power manifested as: (1) an object of social change (to be changed), (2) a constraint for social change, and/or (3) an enabler for social change?

6. Quantity versus quality

Notions of power often come with quantitative connotations of having more or less power. Even when distinguishing between power over and power to, both have quantitative implications for thinking about power relations and (in)equalities. Indeed, most of the power literature has focused primarily on two type of power relations: (1) A has/exercises power over B, and/or (2) A has/exercises more/less power than B to achieve something. In my own work on power, I have argued that there is a third type of power relation that tends to be overseen: (3) A and B exercise a different type of power to/over (Avelino and Rotmans 2009). On that basis, I have proposed a typology of power relations and dynamics to analyse processes of social change (Avelino and Wittmayer 2016 – see Table 1 below). Therein the main distinction is not between ‘to’ and ‘over’, but rather, between ‘over’, ‘more/less’ and ‘different’, and between the distinct ways in which all those types of power relations can manifest.

Table 1. Typology of power relations & dynamics (adopted from: Avelino and Wittmayer 2016)

Type of Power Relations	Types of Power Dynamics
Power over	A depends on B but B also depends on A = > A and B have power over each other	A depends on B but B does not depend on A = > B has power over A	A and B do not depend on each other = > A and B have no power over each other
	mutual dependence	one-sided dependence	independence
More/less power	A exercises more/less power than B, but A and B have similar, collective goals	A exercises more/less power than B, while A and B have mutually exclusive goals	A exercises more/less power than B, A and B have independent co-existent goals
	cooperation	competition	co-existence
Different power	A’s and B’s different power exercises enable and support one another	A’s and B’s different power exercises restrict, resist or disrupt one another	A’s and B’s different power exercises do not (significantly) affect one another
	synergy	antagonism	indifference

One way to distinguish between different types of power is according to the resources that are mobilised. Think for instance of Mann’s ([1986]2002) classical distinction between ideological, economic, military, and political sources of power. Another way to distinguish between different types of power to/over, is to focus on the nature of the power exercise in relation to stability and change. Often, power is presented as something that is, in principle, neutral in relation to change or stability: if one ‘has’ power, one can choose to use it either for pushing change or for defending the status quo, depending on one’s interests. However, I have proposed to acknowledge ‘power-to-change/create-something-new’ as something that exists as a qualitatively different phenomenon from ‘power-to-reproduce-the-existing’. More specifically, I proposed a distinction between reinforcive, innovative and transformative power (Avelino 2017). Reinforcive power is the capacity to reinforce and reproduce existing structures and institutions. This relates to the so-called ‘structural’ interpretations of power (section 5), but I have proposed to call it reinforcive rather than ‘structural’ or ‘institutional’, so as to emphasize that also this type of power can be exercised by actors, rather than only by structures or institutions.

Innovative power is the capacity to create new resources. The creation of new resources is an act of power in and of itself, in the sense that the creation of a new resource (e.g. electric vehicles in combination with renewable energy technologies) can make actors employing them less dependent on existing resources (e.g. fossil fuels), and thereby less dependent on existing physical structures and dominant actors that own and control those existing resources (e.g. the oil industry). If the creation of a new resource can disrupt existing power relations, it can be viewed as an exercise of power in and of itself. This concept of innovative power has been inspired by Arendt’s notion of natality, i.e. the human capacity to be original and create something new (1994, in: Gordon 2002). Arendt defines power as the human ability not just to act but to act in concert, emphasizing visibility and plurality as conditions of power (1958, in: Gordon 2002). Visibility and plurality distinguish innovative power from notions such as ‘invention’ or ‘creativity’. A new resource is powerless if it is not visible to plural actors.

Transformative power is the capacity of actors to challenge, alter and replace structures and institutions, by developing (re)new(ed) structures and institutions, be it a (re)new(ed) legal structure, physical infrastructure, economic paradigm or religious ideology. The idea of ‘new’ structures and institutions should be understood in terms of ‘renewed’ and ‘reconfigured’, where it is not about creating entirely new things out of nothing, but rather a matter of significantly challenging, altering and/or replacing existing institutions and structures by recombining, transposing and reinventing specific elements (Haxeltine et al. 2017, Avelino 2017, Pel et al. 2020). As emphasized by Sewell (1992, p. 27), ‘the same resourceful agency that sustains the reproduction of structures also makes possible their transformation – by means of transpositions of schemas and remobilizations of resources that make the new structures recognizable as transformations of the old’. Rather than viewing the capacity to reproduce as being the same capacity as the capacity to transform, I argue that the renewal of structures is an inherently different capacity then reinforcing structures. While the two capacities might be combined by one actor, they do not necessarily coincide, and one capacity manifests in different acts than the other (e.g. challenging, altering, questioning versus complying, following, and imitating).

6.1. Quantity versus quality: implications for research on social change and innovation

The challenge for research on change and innovation is combining both quantitative and qualitative understandings of power. In the same way that power over, to and with are all relevant, it is necessary to acknowledge the diversity of power relations following from power over/under, more/less power and different kinds of power. These three different types of power relations can coincide with another, but not necessarily. The fact that A (e.g. multinational energy company) exercises more power than B (e.g. renewable energy research group) in absolute terms, does not necessarily mean that A has power over B, nor vice versa. Especially when B exercises a different type of power that A cannot, for instance, thereby achieving a certain level of independence from A. Acknowledging this, together with how power can be exercised in different ways, is quite essential for doing research on change and innovation, as it helps to understand how and why change and innovation can persist in the first place, even in the face of structural constraints. Whether or not change and innovation occur is not solely dependent on whether or not actors enabling change are exercising more power than the actors obstructing change (although this is part of the story). It is also a matter of understanding how each type of power relation can manifest in different ways and come with multiple power dynamics, ranging from mutual dependence, one-sided dependence or independence, to cooperation, competition and co-existence, to synergy, antagonism or disconnection (see Table 1). Based on these overall observations concerning quantitative and qualitative power relations, the following empirical questions about social change and innovation can be formulated:

• How and to what extent are what kinds of power exercised for/through/against change, by and over whom? How and to what extent are which resources mobilised for change?

● Power over

• ○ Which actors are exercising what kind of power over others in/for/against change?

• ○ How and to what extent are which actors (in)dependent on/from each other in/for/through processes of change?

● More/less power

• ○ Which actors are exercising more/less power in/for/through change? Who has more/less access to which resources?

• ○ How and to what extent are which actors cooperating, competing and/or co-existing in/for/through processes of change?

● Different power

• ○ What different types of resource-based power (e.g. ideological, economic, military, political) are being exercised in processes of change, how and by whom?

• ○ How and to what extent are reinforcive power, innovative power and transformative power being exercised, how and by whom?

7. Empowerment versus disempowerment

We can find yet another variation of the contestation over power as enabling versus power as constraining, in the discussion of empowerment and disempowerment. Like power, empowerment has no agreed-upon definition and is often used to capture ‘a family of somewhat related meanings’ (Thomas and Velthouse 1990, p. 666). Generally speaking, empowerment refers to a person’s belief that ‘he or she can direct (…) events towards desired ends’ (Elmes and Smith 2001:34). While literature based on earlier research still defines empowerment in terms of delegating decision-making authority (Boje & Rosile 2001:93), recent research has defined empowerment in more psychological terms, where empowerment is related to processes self-determination and intrinsic motivation (Spreitzer et al. 1999:511, Conger and Kanungo 1988, Thomas and Velthouse 1990, Avelino et al. 2020).

As more and more organisations were involved in empowerment programs, the 90’s were seen as ‘an era of empowerment’, in which empowerment of people was ‘clearly emerging as the organizational revolution of the 1990s’ (Gandz & Bird 1996:383). Empowerment was believed to speed up decision-making in organizations, increase employee loyalty and overall productivity (ultimately resulting in higher profits) (Juhl et al. 1997:103), as such being a necessary condition for organizations to compete in the modern era of increased globalization, competition, downsizing, and acceleration (Quinn & Spreitzer 1997:37). This ‘empowerment movement’ faced various forms of criticism, being characterised as the ‘emperor’s new clothes’ (Argyris 1998: 98); an ‘elusive’ concept (Quinn & Spreitzer 1997:37) with a ‘deceptive allure’ (Eccles 1993: 13). As various empowerment programs failed, authors highlighted the mistaken assumptions on which they were based. One of the things they emphasize is that power cannot be shared or delegated, but only attained and exercised from within: ‘We can confer authority; but power or capacity, no man can give or take (…) Power is not a pre-existing thing which can be handed out to someone, or wrenched from someone’ (Follet in: Boje & Rosile 2001:90, 102). In this view, power is a self-developing capacity and it is thus impossible to empower others in terms of ‘giving’ others power. One might be able to create a context that is more enabling, but ultimately people ‘must choose to be empowered’ and ‘efforts that assume an empowered [individual] is a passive recipient of a brilliant program design are doomed. Empowered people empower themselves’ (Quinn and Spreitzer 1997: 41).

Some of the fiercest critics of empowerment draw on critical theory and argue that practices geared at empowerment actually re-enforce current relations of power, and worsen the patterns of domination and dependence. While much of the mainstream management research refers to ‘the transitive use of the verb: to grant or bestow power’, critical theorists use ‘the reflexive usage: to gain or assume power over someone else’ (Hardy & Leiba-O’Sullivan 1998:475). From such critical perspective, the empowerment terminology ‘reduces conflict by emphasizing consensus and cooperation, through such terms as “associates”, “team members”, “players”, and “coaches”’, and this ‘reduce[s] the necessity of having to use more visible or coercive forms of power to ensure organizational goals are met and to quell resistance’ (ibid:466). These critical perspectives are sceptical about empowerment programs that do not address ‘material conditions of domination’ (Boje & Rosile 2001:93–94). Moreover, critical theorists argue that relations of power depend on ‘one’s location in the system’, and that one cannot alter these relationships at the interpersonal level without changing the system (Boje & Rosile 2001:111). Besides this impossibility of empowering others without-changing-the-system-as-a-whole, critical theorists warn that attempts to empower someone else ‘creates a dependence relationship which, by definition, is disempowering’ (Hardy and Leiba-O’Sullivan 1998:469), and that ‘bestowing power only reinforces the dualism of powerful-powerless, thus ultimately maintaining the superior position of the powerful’ (Boje & Rosile 2001:102).

7.1. Empowerment versus disempowerment: implications for research on social change and innovation

Critical perspectives on power teach us that managerial perspectives on empowerment have many problematic connotations. Using ‘empowerment’ in social change and innovation research requires awareness of that ‘historical discursive legacy’ of the empowerment concept. The (dis)empowerment paradox implies that ‘empowerment’ processes in change and innovation often come with (unintended) ‘disempowerment’ consequences, and that it is necessary to remain critical and attentive to both sides of that same coin. An additional insight from empowerment literature is that (dis)empowerment concerns should not only be with ‘social change agents’ but also with other actors affected by the social change (e.g. government officials or conservative groups).

Based on these critical insights on empowerment, I propose to study processes of social change and innovation in terms of (dis)empowerment (Avelino 2017) On the one hand, we can use insights from e.g. organisational psychology to analyse how actors become empowered to contribute to social change, and how the interpretative, psychological processes underlying such agency for change are facilitated, for instance through participatory methods (e.g. Avelino 2009) or translocal networks (e.g. Avelino et al. 2020). On the other hand, we can also employ critical perspectives on empowerment to critically analyse change and innovation interventions and discourses in terms of their unintended disempowering affects (e.g. Avelino et al. 2019).

(Dis)empowerment in relation to change can refer to (1) (dis)empowerment as a result of change, (2) (dis)empowerment as constraining/enabling change and/or (3) process of (dis)empowerment as a process of change in itself. By consistently using the term (dis)empowerment, I emphasise its two-sided nature and stress the importance of remaining critical to the unintended ‘dark’ side of attempts to change the world for the better. Moreover, (dis)empowerment is intertwined with existing power relations. Understanding (dis)empowerment in/for social change requires and understanding of existing power relations in the societal contexts in which social change takes place. Empirical questions to be asked about process of change and innovation are:

• Who is (dis)empowered, by whom or by what?

• What are the (un)intended (dis)empowerment consequences of social change?

• How and to what extent is (dis)empowerment manifested as:

  • an intentional outcome of change (empowerment as an end)

  • a constraining/enabling factor in change (empowerment as means)

  • a manifestation of the social change in itself (empowerment as process of change)

8. Power = knowledge versus power ≠ knowledge

The relation between power and knowledge is one of the most contested in social theory (Garcia 2001). According to Bourdieu ‘the power to impose and to inculcate a vision of divisions, that is, the power to make visible and explicit social divisions that are implicit, is political power par excellence’ ([1989]2002, p. 142). Or in other words, by developing and communicating knowledge about society, one is exercising power. Barnes even defines power as ‘the distribution of knowledge’ within society, claiming also that ‘to possess power an agent must be known to possess it’ (Barnes [1988]2002, p. 126). The author explicitly distinguishes knowledge from individual belief, stating that ‘every individual in a society may be in error about some aspect of social power (…), so that none of them truly knows where power lies overall, and yet power will lie, necessarily, by its nature, where it is known to lie’ (ibid:126). In a way, both Bourdieu and Barnes argue that knowledge defines power. This, however, is different from saying that power defines knowledge. To what extent power defines knowledge, is an age-old discussion, illustrated by notorious debates between ‘Habermasians’ and ‘Foucauldians’, and reminiscent of the differences between positivistic paradigms and post-positivism.

Although many would agree that power can distort knowledge, the point of debate is whether there still exists such a thing as ‘truth’ or ‘knowledge’ that is ‘free’ of power, i.e. whether it makes sense to speak of knowledge that is not defined by power. When Lukes argues that dominant groups shape perceptions and preferences in such a way that ‘dominated’ groups are ‘unaware’ of their ‘real interests’ (also referred to as ‘false consciousness’), he is suggesting that there is such a thing as ‘real’ interests, i.e. an ‘objective truth’, that can be known and distinguished from ‘false’ and ‘imposed’ interests (Haugaard 2002, p. 39). In contrast, Foucault argues that ‘we should admit rather that power produces knowledge (…) that power and knowledge directly imply one another; that there is no power relation without the correlative constitution of a field of knowledge, nor any knowledge that does not presuppose and constitute at the same time power relations’ ([1977]2002, p. 192). Giddens, however, argues that ‘Foucault’s rehabilitation of the concept of power (…) is achieved only at the cost of succumbing to a Nietzschean strain in which power is seemingly prior to truth’ ([1984]2002, p. 160). Foucault and his ‘followers’ (e.g. Flyvbjerg 1998), are often criticized for their claim that power is always prior to knowledge, truth, or rationality.

Knowledge does not only have a ‘cognitive but also a performative significance’ (Barnes [1988]2002123). Because, as Bourdieu points out, ‘the categories of perception, the schemata of classification (…) the words, the names which construct social reality as much as they express it, are the stake par excellence of political struggle, which is a struggle to impose the legitimate principle of vision and division’ (Bourdieu [1989]2002239). This means that by constructing and communicating knowledge, one is exercising power, not only in terms of mobilizing knowledge, but also in terms of influencing how other actors mobilize all sorts of resources. As formulated by Haugaard: ‘physical power is derived from a knowledge and manipulation of physical objects, while social power is based upon knowledge and membership of social systems’ (2002: 113, in reference to Barnes [1988]2002) and: ‘what enables actors to reproduce structure is their knowledge of social life’ (ibid: 148, in reference to Giddens [1984]2002).

With power and knowledge comes responsibility. This is why scholars like Flyvbjerg not only emphasise the importance of understanding existing power relations and the intertwined nature of power and rationality (1998) but also call upon researchers to take a normative stances in relation to those power dynamics. More specifically, Flyvbjerg (2001, Flyvbjerg 2004) has taken up the notion of phronesis to argue for ‘making social science matter again’ by calling on researchers to dare make normative judgements about what ‘ought’ (not) to be, by answering the following ‘phronetic’ research questions: (1) where are we going; (2) who gains and who loses, and by which mechanisms of power; (3) is this development desirable; and (4) what, if anything, should we do about it?

8.1. Power prior to knowledge versus power as knowledge: implications for research on social change and innovation

Understanding power in the context of social change, requires an understanding of how knowledge in that context is utilised. Moreover, those studying social change are often part of – or have an impact on – that context. The very concept of, discourse on, and research about, social change and innovation (‘transitions’, ‘sustainability’, ‘social innovation’), is in itself an exercise of power, and has power implications. Knowledge development and communication about/on social change is an exercise of power in and of itself. For instance, defining a ‘transition’ or ‘social innovation’ and subsequently assessing that one form of change is ‘more’ or ‘less’ transformative than another, may have serious power consequences. Such knowledge, in turn, could be (ab)used to exercise power in/over processes of change and innovation.

• What kind of knowledges, discourses, ideologies and normativities underly the processes of change, implicitly or explicitly?

• How is knowledge of and discourse on change co-evolving with power dynamics in the change processes?

  • How is the knowledge in the context of change and innovation organised, for and by whom?

  • How and to what extent is knowledge about change and innovation changing, and how does that change involve what kind of shifting power relations?

• How and to what extent is knowledge mobilised as:

  • an object of change

  • an instrument for enabling/constraining change?

9. Synthesis & Conclusion

This paper has started from the observation that there is an increasing attention on processes of social change and innovation towards more sustainable and just societies, as exemplified by public and scientific discourses on ‘sustainability transitions’ and ‘social innovation’. These fields of research share an underlying notion of change and innovation as drivers for societal improvement, and an (implicit or explicit) belief in human knowledge and agency to change the world for the better. In this enthusiasm-for-social-change, the ‘dark’ and ‘unintended’ effects of social change and innovation tend to remain underemphasised, as well as the fierce power struggles and inequalities that come with it.

In response, this paper has set out to demonstrate the diverse ways in which social and political theorists have approached the notion of power, and what this could imply for research on social change and innovation in terms of asking empirical questions about the diverse roles and dimensions of power in processes of social change. A total of seven power contestations in the literature were identified. For each of these points of contestation, this paper discussed how different scholars have dealt with the abovementioned points of contestation, what we can learn from that, and, most importantly, what these points of contestation imply for empirically investigating power in the context of social change and innovation. Rather than ‘choosing sides’ within these power debates or attempting to ‘solve’ them, I have proposed to acknowledge the different dimensions of these power contestations, and on that basis distil empirical questions that can be used to systematically explore the role of power in processes of change and innovation from different perspectives. A succinct summary of all empirical questions for each contestation is given in Table 2 below.

Table 2. Power contestations and questions for research on social change and innovation

Power Contestations	Questions about social change & innovation (hereafter “change”)
Power ‘over’ < > power ‘to’ (e.g. Dahl, Parsons, Foucault, Morris, Davis, Giddens, Arendt, Gordon, Stewart)	Power over: Who is exercising power over whom? How are which structures of domination/oppression/dependence changed or (re)produced? Power to: How is power exercised for/through/against change? Power with: How do actors collaborate in the exercise of power for/against change?
Centred < > diffused (e.g. Dahl, Bachrach & Baratz, Lukes, Mann, Foucault, Gramsci)	How are the three/four faces of power manifested in processes of change? How is power diffused, (de)centralised and/or recentralised in/by/for change? Who is included and excluded? How & by whom is the agenda of change decided? Which issues are kept off the agenda? How are underlying preferences shaped?
Consensual < > conflictual (e.g. Parsons, Arendt, Mann, Haugaard)	How are both consensus and conflict manifested in change? Which conflicts are ‘hidden’ under seemingly consensual processes? How and to what extent is consensus oppressive and conflict emancipatory (and vice versa) in processes of change?
Constraining < > enabling (e.g. Foucault, Giddens, Clegg, Davis, Arendt, Hayward & Lukes)	How are both structure & agency manifested in change? Who/what is enabled and/or constrained by change and how? How/to what extent are which structures (a) an object of change (to be transformed), (b) a constraint for change, (c) an enabler for change?
Quantity < > quality (Mann, Sewell, Arendt, Avelino)	How and to what extent are what different kinds of power exercised for/through/against change, by and over whom? Which actors are exercising more/less power in/of/through change, and how? Who has more/less access to which resources? How do power relations/dynamics manifest in change (e.g. cooperation, (in)dependence, competition, co-existence, synergy, antagonism)?
Empowerment < > disempowerment (e.g. Boje & Rosile, Hardy & Leiba-O’Sullivan, Follet)	Who is (dis)empowered in/by change, by whom or by what? (How) is (dis)empowerment manifested in change as (a) intentional outcome (empowerment as end), (b) constraining/enabling factor (empowerment as means), or (c) object/type of change in itself?
Knowledge as < > prior to power (e.g. Bourdieau, Flyvbjerg, Lukes, Foucault, Barnes)	Which knowledges, discourses, ideologies underly the process of change? How is knowledge on change co-evolving with which power dynamics? How is knowledge mobilised as (a) an object of change, or (b) an instrument for enabling/constraining change?

Acknowledging the many contestations of power also means that it is impossible to agree on one single definition of power. However, broadly speaking, we can think of power dialectically as the (in)capacity of actors to mobilise means to achieve ends. Hence, the challenge for social change and innovation research is to approach the notion of power dialectically and broadly in terms of the human (in)capacity to change what ‘is’ into that which humanity thinks ‘ought to be’. This comes with a complex set of dimensions, in which capacity by one actor at one level can imply incapacity elsewhere, and in which power is both enabling and constraining. This includes both power to and power over, both centralisation and diffusion, both conflict and consensus, both enablers and constraints, both quantity and quality, both empowerment and disempowerment.

This paper has emphasized how understandings of power are contested and debated. Across these contestations, however, there are several elements of power that theorists agree on. One such agreement is that power is relationally constituted and that it ‘resides in the social context’ (Barnes [1988]2002, p. 127). This means that when the social context changes, power relations are bound to change as well, and that as such, changing power relations form an inevitable dimension of social change and innovation. This is especially the case for research fields on ‘sustainability transitions’ and ‘social innovation’, where changing social relations are emphasized as being a characterising aspect of transformative change (Loorbach et al. 2017, Pel et al. 2020). As such, an overarching question to be asked about power in research on change and innovation, a question that cuts across all seven contestations, is: how are processes of change and innovation transforming and/or (re)producing existing and/or new power relations? The power contestations and diverse empirical questions formulated in this paper serve to unpack and specify the different dimensions of this broader question.

Another recurring theme across the different contestations of power and associated questions, is how power can be a means towards achieving change, as well as end of change in itself. There is a considerable amount of research on innovation and ecological sustainability that tends to treat power primarily as a process issue i.e. instrument for change (e.g. how power is/can be exercised to achieve e.g. ecological or economic goals) (Avelino 2011, 2017). However, as discussed in the introduction, there has been an increasing interest in the political aspects of social change and innovation, which includes attention on power as a more substantive issue, possibly an end of change itself. I have earlier argued (ibid) that discussions about social change and innovation should always be accompanied by critical questions about the intended and unintended consequences of change and innovation regarding the improvement and/or exacerbation of problematic power relations (inequality, oppression, exclusion, exploitation and so on). The formulation of power questions in the table above provide a basis to study the role of different power dimensions in processes of change and innovation, in an empirical and analytical manner. However, what is not yet included in this overview, concerns a systematic discussion of how normative political theories of power can be used to question the power-ethical issues underlying processes of change and innovation. As such, an interesting avenue for future research would be to explore the implicit and explicit normativities and directionalities of social change and innovation (and related discourses on e.g. ‘sustainability’, ‘justice’ and ‘inclusivity’) from a political theory point of view, and to relate these to the seven power contestations as identified in this paper.

The purpose of this paper, and its overview of seven power contestations and related questions, is not to argue that social change and innovation researchers who want to engage with the issue of power should necessarily integrate all these power dimensions and questions – and the vast theoretical literature underlying them – in their own empirical research. What I do argue, however, is that each and every dimension of power in and of itself can and should be approached dialectically in the context of empirical research. For instance, when looking for empirical evidence of how ‘power over’ is constraining social change, there is a certain ‘dialectic duty’ to also consider empirical evidence for how ‘power to’ is enabling social change. Likewise, there is a dialectic responsibility to systematically question the future power implications of social change interventions, which compels us to focus part of our power analysis on unravelling and deconstructing political contradictions, paradoxes and ironies that often lie beneath discourses on social change and innovation towards more ‘sustainable’ and ‘just’ societies. This is especially about highlighting unintended side-effects, such as how empowering some often leads to disempowering others, or how policies to achieve ecological sustainability goals on the short or middle-long term can lead to unsustainable social relations and imbalanced power relations in the long term. However, when highlighting these unintended disempowerment consequences of social change interventions, a dialectic approach invites us to also acknowledge the potentially intended empowering effects of that same intervention. By highlighting the diversity of power contestations in the literature, I hope to entice social change and innovation researchers to acknowledge and study both the capacities and incapacities for social change, including the hopes and aspirations, as well as the concerns and fears, of both powerful and powerless individuals in ongoing processes of social change.

Author biography

Flor Avelino works as a researcher and lecturer on in the politics of sustainability transitions and social innovation at the Dutch Research Institute for Transitions (DRIFT) at the Erasmus University of Rotterdam. She specialises in the role of power and empowerment in processes of social change and has a particular empirical and personal interest in translocal networks and social movements that strive towards more just and sustainable communities and societies.

Acknowledgments

This article is based on research carried out as part of the Transformative Social Innovation Theory (“TRANSIT”) project which was funded by the European Union’s Seventh Framework Programme (FP7) under grant agreement 613169. The views expressed in this article are the sole responsibility of the authors and do not necessarily reflect the views of the European Union.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Flor Avelino

Flor Avelino works at the Dutch Research Institute for Transitions (DRIFT) as senior researcher and lecturer in the politics of sustainability transitions and social innovation, and as associate professor at the Department of Public Administration & Sociology (DPAS) at the Erasmus School of Social & Behavioural Sciences (ESSB, Erasmus University of Rotterdam). With a background in political science, she specialises in power theories and has a particular interest in understanding how people and networks are (dis)empowered to contribute to change and how power relations are being challenged and reproduced through translocal processes of social innovation and sustainability transitions.

Notes

1. A typical example of a ‘family resemblance concept’ is the word ‘game’: its meaning inherently depends on the context in which it is used. The ‘playfulness’ of a card game played at home starkly contradicts with the ‘seriousness’ of a political game. All possible meanings of the word ‘game’ partly overlap and partly contradict each other, hence making it impossible to agree on one all-encompassing definition.

2. Although Foucault is often criticized for his ‘death of the subject’, Haugaard claims that this is contestable (Haugaard 2002: 209). Interpreting what Foucault meant by power can be considered a separate debate in itself, as authors frequently accuse each other of either misunderstanding or neglecting parts of Foucault’s work (e.g. Aladjam 1995, Borch 2005, Garcia 2001, Heiskala 2001, Infinitio 2003, Thompson 2003). This is further complicated by the fact that there is quite some difference and even contradiction between the ‘early Foucault’ and the ‘later Foucault’, and because at least one of the ‘Foucaults’ has explicitly emphasized that he does not aim to present a theory nor a model of power, but rather a ‘toolbox’ for studying power.