Making “Mini-Me’s”: Service-Learning as Governance of the Self in a South African Context

This article presents an attempt to examine my own service-learning practices through the use of the conceptual tools of Michel Foucault, in particular his notions of governmentality and power. The article views the development of service-learning in South Africa and our current practices as operating within a regime of truth, and it considers service-learning as an apparatus for constructing particular kinds of subjects. From a broad conceptual lens, the article moves to the analysis of an interaction during a critical reflection process in service-learning in an attempt to examine actual practices and how these may produce different subjectivities. The article is an attempt to encourage other practitioners to reflect on their own practices, uncover their assumptions, and ask how things could be otherwise.


Introduction
stated that it is "not that everything is bad, but that everything is dangerous, which is not exactly the same as bad. If everything is dangerous, then we always have something to do." This article presents a Foucauldian analysis of servicelearning practices that is not an attempt to pronounce judgement, but rather to open up ways in which it could be dangerous. Butin (2010, 18) warned "service learning is not safe. It is anything but safe." If it is dangerous, then we have work to do.

The Catalyst
A few years ago, I was invited to attend student service-learning presentations at a sister campus of our university. As the students were presenting, it struck me how much they sounded like my colleague who had been the course instructor. They used her frameworks of understanding; they highlighted the issues that were most pertinent to her; they used her most familiar terms, and some even spoke with her passion for the issues confronted in the course. At that moment, I became aware of how, as servicelearning practitioners, and educators in general, we direct our students' learning towards our priorities and in terms of our frames of reference. This was a worrying moment, as I was attracted to service-learning because it seemed to offer an alternative to conventional teaching practice, which appeared to me to be more prescriptive. The "openness", flexibility, and "student-led" nature of service-learning resonated with my desire for a less authoritative approach to teaching.
Having already been practising service-learning for a number of years, I wondered how many "mini-me's" I had produced over the years, and considered the dangers of this kind of production. After all, my intentions were noble and good, and who could dispute that we need socially aware and active students to help challenge the inequitable system? But what if this powerful tool of service-learning ended up in the wrong hands? What nefarious goals might students be indoctrinated towards? Harkavy and Benson (1998, 12) argued that the revolutionary mantra for servicelearning should be "overthrowing Plato and instituting Dewey". They argued for the democratisation of education, as opposed to the elitist Platonic approach. They rejected Plato's dualistic notion of separating "pure" theory and applied practice, and instead recommended Dewey's instrumental learning through inquiry. This argument is appealing; after all, as educators, we want our students to learn through active engagement with "genuine dilemmas and perplexities" (Harkavy and Benson 1998, 16). But, in considering the questions generated by my experience above, I wondered how far the service-learning field had come in terms of overthrowing Plato. He stated the purpose of education as "the process of drawing and guiding children towards that principle which is pronounced right by the law and confirmed as truly right by the experience of the oldest and the most just" (Plato 1980; The Laws 659). Were we, in continuing not to examine our own assumptions and philosophies, deciding what should be pronounced and confirmed as truly right, and guiding our students towards those principles? I became concerned that service-learning may be the "emperor's new clothes", and was reminded of a cynical friend of mine who always enquired after my work by asking how the social engineering project was going. I was particularly concerned as my experience in service-learning over the years had revealed the power of the pedagogy in rendering students vulnerable, and thus more open to enquiry and learning.
I wondered how to bring a critical perspective to my work to encourage this level of critical reflection.

My Positionings
As a lecturer in the discipline of psychology, I have experience in a variety of different learning contexts, from "normal" lecture-based classes with hundreds of undergraduate students, to individual supervision of students' clinical case work. Service-learning is part of this teaching portfolio, together with a strong interest in community psychology.
The scare quotes around "normal" are intentional, since, in the literature, servicelearning has often been positioned as counter-normative. For example, Jeff Howard wrote "service-learning is not for the meek" (1998,28) as it moves beyond traditional classroom-based teaching and learning practices. This counter-normative nature of service-learning has become part of its identity, and practitioners are often regarded as pioneers for promoting shared responsibility, active learning opportunities for students, more egalitarian approaches to learning and coping with the unpredictability of working in the real world (Ash and Clayton 2009).
As will become apparent in this article, this distinction between normative and counternormative is central to my concern regarding the construction of "mini-me's". Foucault (Simon and Foucault 1971) argued that while a seminar format may appear to be more open and egalitarian, it is less "honest" than the lecture. The lecture system, with a professor behind his/her desk and no opportunities for student discussion, has, what he calls, [a] crude honesty, provided it states what it is: not the proclamation of a truth, but the tentative result of some work which has its hypotheses, methods and which therefore can appeal for criticism and objections: the student is free to uncover its blunders. (Simon and Foucault 1971, 199-200) In contrast, the seminar, with its apparent respect for freedom, is far more dangerous: [B]ut don't you think that a professor who takes charge of students at the beginning of the year, makes them work in small groups, invites them to enter his own work, shares with them his own problem and methods-don't you think that students coming out of this seminar will be even more twisted than if they had simply attended a series of lectures? Will they not tend to consider as acquired, natural, evident and absolutely true what is after all only the system, the code and the grid of the professor? Isn't there the risk that the professor feeds them with ideas much more insidiously? (Simon and Foucault 1971, 199) Foucault's observations troubled me, particularly as I had been one of the original participants in the CHESP (Community Higher Education Service Partnership) initiative, and thus deeply immersed in the normative frameworks of service-learning. As Butin (2010) argued, there is a strong normative framework in service-learning, which allows practitioners to privilege particular models and goals. There are prescriptions on more or less effective ways to do service-learning; there are step-bystep guides of how to design a course; there are comparisons of different types of service-learning and service-learning research, with some regarded as better than others, and there are exemplars and best practices. These are all sanctions, techniques, part of the apparatus of service-learning as a regime of truth. There are "experts" in the field who pronounce the truth of service-learning, and are revered and respected (professors, pioneers, winners of awards). Thus, ironically, the operations of the power/knowledge nexus are apparent in this egalitarian educational endeavour. In addition, there are international associations and conferences where participants receive recognition and acclaim, and participate in creating and recreating this regime of truth. There are prizes for emerging scholars and for those who have contributed the most in the field. 1 Given my concerns regarding "insidiously twisting" students, I searched for those who may share my fears. There is excellent and informative research regarding critical service-learning (Mitchell 2008), critical reflection for optimal outcomes through service-learning (Ash and Clayton 2009), and the promotion of community benefit (Stoecker and Tryon 2009). This literature questions service-learning practices and outcomes. Morrison's (2015) description of her own experience in service-learning research was particularly revealing. She argued that we need to be reflexive in our service-learning research, aware of our stance and how that serves to construct particular kinds of truths: "who is the I that shapes the eye? and How does the I shape the eye?" (2015, 54). The critical eye I was looking for would "trouble" the water that we servicelearning practitioners swim in, so that I could become aware of the water. I thus went in search of a conceptual framework to supply a lens through which I could examine service-learning processes and practices. Foucault's tools were most useful for this kind of conceptual work. Michel Foucault (1926-1984 was a French postmodern, poststructuralist philosopher whose books, articles, lectures and interviews have sparked much controversy by challenging the Western philosophical tradition. Through his work, he tried to demonstrate that what is viewed as absolute and universal is frequently the product of historical development. His perspective opens up new possibilities of being, by revealing that current reality is not a given, and challenges us to think about how things might be different. Foucault's "ontology of the present" involves investigating who we are today, and how that has been constructed by a) the forms of knowledge (discourses) that we have of ourselves, b) political forces and how we are controlled through disciplinary practices, or c) the relationships we have with ourselves (McHoul and Grace 2002, viii). Foucault tried to alert us to the ways in which things could have been, and can be, otherwise.

Conceptual Framework
I found Michel Foucault's notion of power as a productive force and as an act, not a possession, helpful in trying to understand what might be occurring in service-learning (Foucault 1980b, 119). When I explained to a former student that I was considering using a Foucauldian perspective, he exclaimed "but we felt so agentic in service-learning". Understanding power as a transaction helps to explain this sense of agency, while at the same time being governed in some way (Foucault 1990).
Being governed, by ourselves or others, is central to the concept of governmentality.

Governmentality
The issue of governmentality emerged in Foucault's work during the late 1970s where he focused on the "problematic of government", or how people are governed in modern societies (Smart 2002, xiv). Governmentality can be understood as "the whole range of practices that constitute, define, organise and instrumentalise the strategies that individuals in their freedom can use in dealing with each other" (Foucault 2000, 300). Governmentality is concerned with both the conduct of the population, and how we conduct ourselves. In everyday life, our conduct is managed by experts in various institutions (e.g. the family, medical personnel, psychologists, marketers) who have authority as a result of their expertise, which is accorded the status of truth. Dean (2010, 20) explained that an analysis of governmentality involves examining those practices "that try to shape, sculpt, mobilise and work through the choices, desires, aspirations, needs, wants and lifestyles of individuals and groups". Governmentality presupposes subjects who are free to choose to respond in a variety of ways, and it attempts to mould these choices to secure the ends of government.
Also, key to governmentality is the regime of truth within which these processes are constructed and exercised. As Foucault observed, Truth is a thing of this world: it is produced only by virtue of multiple forms of constraint. And it induces regular effects of power. Each society has its regime of truth, its "general politics" of truth: that is, the types of discourse which it accepts and makes function as true; the mechanisms and instances which enable one to distinguish true and false statements; the means by which each is sanctioned; the techniques and procedures accorded value in the acquisition of truth; the status of those who are charged with saying what counts as true. (Foucault 1980b, 131) Foucault used the term "apparatus" or "dispositif" to refer to the network of structures and processes that are employed to maintain power relations and to promote a particular regime of truth. He explained: What I'm trying to pick out with this term is, firstly, a thoroughly heterogeneous ensemble consisting of discourses, institutions, architectural forms, regulatory decisions, laws, administrative measures, scientific statements, philosophical, moral and philanthropic propositions-in short, the said as much as the unsaid. Such are the elements of the apparatus. The apparatus itself is the system of relations that can be established between these elements. (Foucault 1980a, 194) Foucault further explained that an apparatus has a dominant strategic function and emerges in response to an "urgent need" (Foucault 1980a, 195). This specific strategic response is rationalised over time and turned into a technology of power in other situations (Rabinow and Rose 2003, 11). As Nicoll and Fejes (2009) explained, an apparatus is not put in place by any particular interest group, but is rather the outcome of the confluence of dispersed activities and ideas that then operate as a strategy.
Governmentality is concerned with how we are governed, both at the strategic level, as in regimes of truth and the dispositif, and also with how we govern ourselves.

Pastoral Power
One of Foucault's expositions of power was in the form of pastoral power. This understanding of power was most congruent with my experiences in service-learning, especially in respect of my relationships with my students. According to Foucault, this form of power stems from Judeo/Christian traditions and is concerned with the relationship of the shepherd and the flock. It is through the care of others that they are dominated, by instilling in them the need to care for themselves. Pastoral power is a beneficent power-Foucault emphasised "pastoral power is, I think, entirely defined by its beneficence; its only raison d'être is doing good, and in order to do good" (Foucault 2007, 172).
Pastoral power is analogous to the complex reciprocal relationship of the shepherd and the flock, and revolves around salvation, obedience and truth. The shepherd is responsible for her/his flock and accountable, not only for their actions but also for their thoughts and attitudes. The duty of the shepherd is the salvation of the flock-even to the point of self-sacrifice. The shepherd guides and protects her/his flock in order to ensure their well-being. S/he maintains a vigilant surveillance over each individual and the whole flock, as s/he will have to account for them. The flock in turn is required to submit to the guidance of the shepherd, to whom they owe total obedience. This individualised submission to the shepherd is necessary to arrive at a state of obedience. In terms of the problem of truth, the shepherd needs to teach the flock, through the examination of their conscience (spiritual direction), the truth about themselves.

Confession
The relationship of submission and obedience that characterises pastoral power highlights the importance of the technology of confession. Indeed, the shepherd needs to know the minds of the flock, and confession is central to the workings of these power relations. The importance of confession is its usefulness as a technology of individualisation: "The truthful confession was inscribed at the heart of the procedures of individualisation by power" (Foucault 1990, 58). Through confession, individuals are actively involved in self-governance-they are obligated to tell the truth about themselves and act upon that truth. Techniques of the self are those actions that are employed to govern the self in accordance with that truth. Thus, confession serves as a technology to engage individuals (us) in defining themselves in accordance with social norms (subjection), in the guise of a liberatory practice, which helps us to unlock the truth about ourselves.

Freedom and Resistance
Although we can never be outside power, in discussing the exercise of power, Foucault emphasised the importance of freedom. He proposed that power could only be exercised "over free subjects, and only insofar as they are free" (1982,780). Foucault emphasised the centrality of freedom to the operations of power. He also differentiated between freedom and liberation and domination. He was primarily interested in the practices of freedom, in how subjects adopt strategies and tactics to free themselves from the constraints placed upon them by regimes of truth. This freedom is not about being left alone but about "re-making ourselves into what we would like to be: freedom for, not just freedom from" (May 2011, 79;italics in original). It is about choosing by whom we wish to be governed and to what end (Foucault 2007).
Given this quick and rudimentary overview of some of Foucault's ideas, the next section tries to provide a brief outline of the "genealogy" of service-learning in our context.

Service-Learning in South Africa (Briefly)
Service-learning in South Africa was mainly driven by the CHESP project (in 2000), which was funded, in part, by the Ford Foundation and the Department of Education. CHESP sought to increase community engagement at South African higher education institutions by selecting "pioneer" universities and then training and encouraging core group members to promote the initiatives in their contexts (e.g. higher education institutions [HEIs], service-organisations, communities). 2 While the CHESP initiative funded more than 100 service-learning courses across eight HEIs, many activist academics had been engaged in community-based initiatives prior to this, and during the apartheid years.
The CHESP project occurred in the context of Department of Education (DoE) mandates to transform higher education (i.e. Education White Paper 3 [DoE 1997] and the Higher Education Quality Committee [HEQC 2001]). Thus, South African HEIs were fertile ground for the CHESP initiative-Department of Education policies, international donors and experienced academic activists ensured that the programme was successful during its tenure. Since 2007, when the funding period ended, there has been a lull in these kinds of activities (Stanton and Erasmus 2013). Although many academics still continue with service-learning programmes in a variety of disciplines, these are largely unfunded and unsupported by the HEI. Research and publications regarding service-learning in South Africa are still robust, and debates regarding its origins, relevance and effectiveness continue.

Service-Learning through a Foucauldian Lens
Foucauldian approaches generate many questions. This exercise does not aim to provide the answer; Foucault would argue that that is not possible. Instead, I am interested in what questions can be asked and with what effect. The objective of this exercise therefore is to make explicit our service-learning discourses and practices, in order that we may have new insights and perhaps adjust our strategies. Critical for this exposition is the notion that we are always operating within a field of power/knowledge and can never be outside it; that is, "there is no absolute outside" (Foucault 1990, 94).
While other authors (Gilbride-Brown 2011) have employed Foucault to reflect on aspects of the service-learning endeavour as a "transformative regime of truth" (Olson 2015, 42) or as located within the governmentality of the university (Preece 2016), there does not seem to have been any analysis of service-learning as a regime of truth (barring Butin, below), or as a dispositif. This is attempted below. Foucault (1980b) pointed out that each social formation has its own regime of truth, its own politics of truth, and notions of what counts as true. Dan Butin is one of the few service-learning researchers who have employed a Foucauldian perspective. He observed that "service learning, as any other educational reform model, has its own blind spots, its own unacknowledged and unexamined assumptions, and its own impositional narratives" (Butin 2006, 1). If liberatory practices often impose their own regime of truth (as Foucault claimed), there is a need to examine more closely what is happening in service-learning.

Service-Learning as a Regime of Truth
The discourses evident in the service-learning literature have to do with transformation, mutual benefit, partnerships, egalitarian approaches, and community empowerment. Different factions within service-learning privilege certain discourses, with gentler approaches speaking of citizenship and learning outcomes for students, and more radical approaches calling for activism and student transformation. There is a service-learning vocabulary of terms that practitioners in the field participate in and create. Researchers talk of civic engagement, social justice, critical reflection, reciprocal relationships, and stages of learning, which all construct a form of "truth" about service-learning. There is talk of different forms of knowledges, local knowledge and disciplinary knowledge (i.e. stemming from a particular academic discipline), which each have their own power/knowledge nexuses. Once again, depending on the approach favoured, these differing forms of power/knowledge are privileged or subjugated.
As a regime of truth, service-learning also has various elements and networks of structures and processes, which constitute an apparatus that is employed to direct and maintain power practices. Three elements of this ensemble-policy, notions of participation and empowerment, and critical reflection-are discussed below.

Service-Learning as an Apparatus
Foucault explained that an apparatus usually emerges in response to an urgent need and serves a strategic function. Stanton and Erasmus (2013) compared the emergence and operation of service-learning in the United States of America (USA) and South Africa (SA). Stanton noted that, in the USA, service-learning emerged as a grassroots activist movement in favour of educational reform to ensure higher education institutions "thr [ew] open their windows" to their communities (Stanton, Giles, and Cruz 1999, 132). This bottom-up movement has subsequently been supported by government initiatives and corporate funding. Erasmus noted the importance of South African government policy in driving a community engagement agenda (Stanton and Erasmus 2013). This policy imperative is discussed below.

Policy Imperatives
Miller and Rose (1990) highlighted the function of policy as a programme for reforming reality. They highlighted that policies are evidence of the belief that reality can be directed differently and more effectively. The Department of Education's White Paper 3 (1997) outlined "A Programme for Higher Education Transformation", which offered the optimistic perspective that higher education institutions can, and should, contribute to the reconstruction of South African society through community service programmes. It is important to consider the context of this policy, which was published soon after the first democratic election, and in the light of the ANC's reconstruction and development programme. The White Paper highlighted the problematic context the apartheid years created and emphasised the "urgent need" for reconstructing "domestic social and economic relations to eradicate and redress [the] inequitable patterns of ownership, wealth, and social and economic practices" (DoE 1997, 9). As an instrument of social reform, higher education institutions were further tasked with producing students who are socially responsible and aware.
From a Foucauldian perspective, the role of this policy in trying to develop a society that is different to that which existed prior to 1994 is clear. The arguments for action are convincing, drawing on notions of injustice, inequality, and disparity to highlight the urgent need for reform. The policy even outlines the mechanisms for enacting this change-through community service programmes and the production of certain kinds of students. Following the development of the policy, more structures were put in place to ensure the implementation of the policy. The Council for Higher Education and the Higher Education Quality Committee created criteria for institutions to demonstrate their compliance with the policy imperative.
An addition to the ensemble, in South Africa, was the creation of the CHESP initiative, which sought to assist with the implementation of the mandate in higher education.
CHESP's power/knowledge nexus was strengthened by state and overseas donor funding, access to US academics/faculty and service-learning resources, and persons of influence in South African education. Through offering grants and immersion training (for selected universities), CHESP provided incentives for already committed individuals to strengthen their activism in the field of community engagement. As a governmentality project, CHESP promoted a particular model of community engagement, through service-learning in the context of a community-higher educationservice partner arrangement (Stanton and Erasmus 2013). Although this model did not suit all contexts and forms of community engagement (Mitchell and Humphries 2007), compliance was encouraged through reward (grants for service-learning modules at universities) and through participation in a community of practice (CHESP core groups), where discourses were created and reinforced during contact sessions. In Miller and Rose's (1990, 4) terms, allies were enrolled and subjected to a process of "interessement". CHESP core group participants were expected to take their learning back to their institutions and promote uptake of service-learning through various microlevel activities, for example, seminars, workshops, consultations, and community meetings. (Mitchell, Trotter, and Gelmon 2005).
Through the policy imperative, and the structures created to support the implementation of the policy, higher education institutions and particular academics/faculty were responsibilised. Rose's (1999, xxiii) notion of "government through freedom" is relevant here. Rose (1999) highlighted the interplay between being governed by the state and governing oneself, taking cognisance of opportunities for contestation and resistance. As highlighted earlier, many academics who responded to the call for service-learning programmes were already involved in community engagement or activist initiatives. Their participation may have served self-interest (resistance or freedom) in terms of receiving funding and support for work that they were already engaged in and responsibilised towards, as well as contestation, by continuing to pursue models of engagement that did not necessarily fit the CHESP model (cf. Mitchell, Trotter, and Gelmon 2005). Foucault [1980c] advises us not to focus on), the power effects of the policy for the transformation of higher education (and its subsequent structures for implementation) have resulted in forms of government over the community engagement activities of higher education institutions and their staff. This governmentality is expected to further permeate into the kinds of students these institutions and practices produce.

Mechanisms of Participation and Empowerment
At, what could be perceived as, the opposite end of the spectrum of coercive power practices are the notions of participation and empowerment that are central to the service-learning endeavour. Participation and empowerment are generally considered tools for more democratic and egalitarian practices and favourable outcomes. They are appropriated in various domains because of their seemingly neutral or even beneficent nature. But, even these seemingly innocent tools are places where power effects are exercised.
Student and community participation is key for the success of any service-learning endeavour. Writings on partnerships in service-learning have underscored the importance of mutually beneficial and empowering relationships between community partners and their higher education partners and students. In discussion of best practices, care is taken (Bringle, Hatcher, and Games 1997) to underline how community partners need to be consulted at every step of the process. Service-learning, by definition, requires students to participate in community-based activities as part of their academic curriculum. The participation of communities is also assumed in the creation of the service-learning activity-communities, at the very least, need to participate by hosting students at their sites.
The idea of participation per se is not problematised in the service-learning literature. Participation in partnerships and the dynamics of these relationships between communities and higher education institutions have been reflected on (Mitchell and Humphries 2007). Some authors have also described students' varying levels of engagement in service-learning over the course of a module (cf. Kiely 2004). However, there does not seem to have been any reflection on the assumption of participation in service-learning.
Assumptions about participation position it as a neutral activity (Masschelein and Quaghebeur 2005), and that it encompasses choice and individuals exercising their freedom to choose. After all, students choose to enrol for a service-learning course; they choose to be active in a community; the community partner chooses to have an arrangement with the university, and so on. From a Foucauldian perspective, participation can be regarded as a form of governmentality, whereby participation prescribes a certain way of being; it promotes a particular kind of subjectivation and a way of demonstrating one's freedom. Masschelein and Quaghebeur (2005, 68) explained it as follows: "participation acts upon individuals by getting them to act in and on their own interests, by getting them to act as self-determining, self-controlling, self-reliant, competent and autonomous actors". Paradoxically, in participating, one is constructing oneself through techniques of the self. In other words, when we think we are freely participating, we are actually subscribing to the governing mechanisms involved in being a participating subject.
In the main, the service-learning literature has not problematised the notion of participation (of students, communities, faculty). Participation is a fundamental prerequisite for the pedagogy to function (the subjects involved in the process need to act in some way) and many researchers have failed to recognise that participation (as an act of self-governance) itself is "part of an operation of power, governing people to behave themselves in a particular determined way" (Quaghebeur, Masschelein, and Nguyen 2004, 154). It is important to remember that Foucault does not assign judgement or evaluation to these practices; rather he asks how things could be otherwise. Asking service-learning practitioners how practice would be different if participation were not taken for granted exposes the operations of power in the endeavour.
Similarly, empowerment is also assumed to be a beneficent practice. How can there be anything wrong with ensuring another also has 3 power? Service-learning aims to empower both students and communities to effect change. Service-learning literature has emphasised the numerous benefits to students that service-learning claims to yield. These benefits of skills, knowledge, attitudes and behaviour are considered empowering.
A brief sojourn into other fields can help to problematise this. Ellsworth (1989) and Gore (1990) both found themselves caught up in the circularity of being part of the apparatus they wished to challenge and change. In her article "Why Doesn't This Feel Empowering?", Elizabeth Ellsworth (1989, 298) argued that critical pedagogical concepts and practices such as "'empowerment,' 'student voice,' 'dialogue' and even the term 'critical'-are repressive myths that perpetuate relations of domination". She described the radical educator as one who attempts to help students recognise injustice and empower them to act against oppression, while at the same time transforming her understanding in response to her students. Her experience in running an anti-racism class demonstrated the impossibility of neutralising oppression in classroom practices, even when trying to employ a critical pedagogical approach as an alternative to more normative practices. She explained that she failed in her attempts to radicalise the educational process, and observed that "[c]ritical pedagogues are always implicated in the very structures they are trying to change" (Ellsworth 1989, 310).
Jennifer Gore (1990) used a Foucauldian lens to reflect on critical and feminist pedagogical practices and found herself in a similar dilemma. Her main concerns centred on perceptions of power and agency. When the teacher is positioned as the "empower-er" it implies an omnipotent agentic position, where the teacher is able to share her power or give it away. "To em-power suggests that power can be given, provided, controlled, held, conferred, taken away" (Gore 1990, 9). Gore contended that empowerment discourses set up an agent (teacher) who empowers others (students), resulting in distinctions between "us" and "them". She argued that when the focus is on others, we may neglect to examine our own role in the conditions we seek to change: In attempts to empower others we need to acknowledge that our agency has limits, that we might "get it wrong" in assuming we know what would be empowering for others, and that no matter what our aims or how we go about "empowering", our efforts will be partial and inconsistent. (Gore 1990, 15) 3 In Foucauldian terms, power cannot be possessed or "had"-it can only be exercised.
In the end, Gore (1990, 10) concluded that empowerment is better conceptualised as "the exercise of power in an attempt (that might not be successful) to help others to exercise power".
Our service-learning practices of dialogues between students and academic staff, between community members and academic staff, and between students and community members, have the intention of being empowering. After all, one does not effect real change without ensuring those who are responsible for it have a sense of agency. Yet Gore's (1990) dilemmas become ours when we recognise the circularity of participating in perpetuating those power effects. By way of very simple explanation: We use our authority (in the nicest way possible, e.g. a team talk) to instruct our students (others) that their interactions with communities should be empowering. What we model for them is the use of power effects to convince others to change. My argument here is not that we should avoid the exercise of power through these techniques, but rather that we should recognise them for what they are, instead of assuming they are neutral and beneficent practices. We are never outside power (Foucault 1990). Foucault (1980c) advised that we should study the exercise of power at its extremities, as this is where it produces real effects. The critical reflection process in service-learning is where this kind of micro-analysis of instances of power is possible.

The Process of Critical Reflection
A central part of the service-learning apparatus is the critical reflection process. While critical reflection is common in other educational endeavours, and involves the same technologies of the self, it is positioned as key to the learning process in service-learning (Eyler and Giles 1999). The extent and reach of critical reflection depend on the model of service-learning employed. Service-learning endeavours differ vastly, and can be located on a continuum. So once-off visits to communities and international immersion experiences may both be categorised as service-learning, although they are likely to have different processes, aims and expectations. Models that aim for transformation in students are likely to offer much more intensive experiences for students than superficial exposure or awareness visits.
Reflection and critical reflection take a variety of forms in service-learning, from faceto-face small discussion groups, to online discussion boards, peer conversations and interviews, critical incident journals, reflective essays, and learning logs, among others. Reflection can be oral or written; it can be individual or group based; and it can be for assessment or personal purposes. The common aspect to these reflection activities is that they require the individual to make the private public, even if that public is an internalised other. From a Foucauldian perspective, it involves the individual in technologies of the self, in a form of telling the truth about the self and constituting that self. From this perspective, these various tools can be understood as different forms of self-regulation and self-governance.
From a governmentality perspective, reflection can be viewed as a technique of the self, where through internalising the other, the student subjects her/himself to scrutiny and self-governance, always aiming to achieve the desired norm. From the perspective of pastoral power, reflection can be viewed as confession-making the private available for public scrutiny, appraisal and judgement, requesting absolution and shaping oneself into the desirable subject.
Foucault's conceptualisation of the confessional is useful here. Service-learning places the academic in a position of authority, which allows her/him to insist on the students submitting to some kind of confessional process (reflective tool). Devas (2004) noted that this position of authority allows the authority to decide how the confession should take place, what needs to be confessed, and what will count as truth. These power effects are strengthened when evaluation (or assessment for credit) is allocated to the product of confession. Foucault (1990) highlighted that the potential for the extraction of truth is strengthened by the intimacy of the relationship between the confessor and the authority/listener. The service-learning literature abounds with extracts from students' journals or reflective essays where they confess and construct deep aspects of their selves to the reader (see, for example, Bursaw 2013;Carrington 2011;Kiely 2004). They make public thoughts, attitudes, behaviours, responses, and feelings in raw ways, where they expose their shadows and their light. I would posit that it is the intensity of the service-learning experience that makes this kind of intimacy and confession possible. Service-learning aims to throw students into disequilibrium, to have them experience disorientating dilemmas, as the assumption is that these will lead to transformational learning (Kiely 2004;2005;Mezirow 1997). These experiences of dissonance often result in students requesting assistance from a more expert other, the shepherd, who also provides guidance and emotional support.
My argument is that often service-learning intentionally places students in contexts which render them vulnerable and more malleable and open to forms of governance, including governance of the self. The academic/authority/shepherd participates in this process by making her/himself accessible, through confessing her/his own shortcomings, and her/his own dedication to the process of becoming. The intensity of these power effects will differ in various contexts and different forms of servicelearning, with, for example, a written one-page descriptive reflection demanding less of the participants than an immersive group experience, but I contend the strategies and tactics are the same. The one-page reflection still requires the private be made public, submission to an authority for approval or redemption, obedience and submission on the part of the student, and at least some self-regulation is enforced.
These differences in the power effects alert us to the possibilities of freedom and resistance. From a Foucauldian perspective, power and resistance co-constitute each other; there cannot be the operation of power without the operation of resistance and the existence of free subjects (Foucault 1982). Foucault was interested in the strategies and tactics that subjects adopt to free themselves from the power effects of regimes of truth-how subjects make themselves into what they would like to be. Within the service-learning field, we need to examine how students (community members/faculty) resist the forms of subjectification they are exposed to as well as how they use their experience to practise their freedom by experimenting with alternative ways of being. Though some would argue that freedom is another form of being governed (Rose 1999) and an illusion of choice (Graham 2007), it is worth attending to the freedom practices that the participants employ within a regime of truth.
Within the service-learning sphere, it is possible to imagine that student resistance would be evident in refusals, silences, or even alternative courses of action. For example, a student may refuse to make submissions-either at all, or in the form required-or to work in a particular community or in a particular way, or to participate in group or online discussions. Resistance may also be evident in silences and withdrawal, either in class or on paper or in community sites. Resistance may also be evident in proposing and motivating for alternatives. Practices of freedom may be evident in students taking on different identities to those proffered by the regime of truth, challenging established practices, realising the rules of the game (Macfarlane and Gourlay 2009), and complying in order to achieve credit or praise. Participants may practise their freedom by choosing to participate in a manner that best suits them, in Foucault's words by asking, "By whom do we consent to be directed or conducted? How do we want to be conducted? Towards what do we want to be led?" (Foucault 2007, 264). Likewise, communities and academics may also resist the subjectivities constructed in a particular regime of truth and, in practising their freedom, choose to be otherwise.
It is easy to be lured by an exploration of motives and intentions in this kind of analysis, or to make judgements about what is good or bad. Neither of these avenues of enquiry yields useful results, as these results always depend on whose version of the truth is privileged. It is therefore more useful to examine what people actually do-their practices.
As children we used to say "one, two, skip a few, ninety-nine, one hundred". The reader should bear this in mind as I now leap to provide some "evidence" of how I used the notion of the confessional in my own service-learning practice.

A Micro-Analysis of Practice
This section bypasses discussions of methodology, explorations of the contradictory subjectivities constructed in the context of a service-learning course (available from the author), and skips to an analysis of what actually happened in the critical reflection session (verbal discussions) in one of my classes. It is useful for the reader to know that the excerpts presented are from a postgraduate course in community psychology, which five students elected to participate in.
As described above, Foucauldian critiques of critical reflection draw attention to the possibility that this taken-for-granted everyday practice is actually a device to promote governmentality, both through the pastoral power practices of the facilitator and the technologies of the self, which bring one's conduct in line with what is desirable within a particular regime of truth. The excerpts presented here are an attempt to explicate two of the strategies and tactics used in the talk. They focus on the talk in context. I utilised Wetherell's (1998) strategy of considering the contexts of conversation, alongside the repertoires and positioning made available in the talk (a form of Foucauldian discourse analysis).
Upon working with and analysing the recordings and transcripts, it became apparent that preserving relationships in the service-learning process was paramount for all the participants. As there were only six participants in the course, the intimacy of the relationships was intensified. Employing Foucault's notion of pastoral power, it is the relationship between the shepherd and the sheep that effects change, that produces subjects who are able to care for themselves and others. As facilitator, it was therefore vital to promote this kind of relationship with the students. Two examples of the way in which this was achieved are discussed below.

Humour
What was noticeable upon immersion in the data is the frequency of laughter within the sessions and the group. There are many studies on the purposes of humour and laughter in interaction. Jefferson's (1979) studies on laughter demonstrated that it is deployed to manage interactions, and is not spontaneous and involuntary, as is often presumed.
Similarly, in the current study, I found that humour and laughter were used in a variety of different ways in the talk, to diffuse tension, to soften discipline, to ease disagreement, and to indicate solidarity.
In the excerpt below, humour was used to diffuse feelings of helplessness and frustration, both with the context, and with my inability (unwillingness) to provide solutions.
825 Group: OOOOOOOOOOOOOOH (Session 4) In the extract, Lisa made a plea for assistance, which I refused to offer. Humour was used to manage the tension this evoked. In lines 812-814, Lisa expressed her sense of helplessness and uncertainty at what can be achieved in a context of structural inequality ("societal structures are (.) poverty racism" [line 813]). My response to her statement that she had "lots of questions" was "Great" (line 815), in other words, celebrating her uncertainty and confusion. Elle's "as well" in line 816 appears to be a statement of agreement with Lisa's uncertainty. Elle identified with Lisa's positioning. My unexpected response of "Great" (as opposed to a more likely sympathetic stance) was not well received by Anna whose verbal "BUT I WANT THE ANSWERS:: (.)CAROL" and her non-verbal banging of hands on the table, reminiscent of a recalcitrant child, were softened with her laugher (lines 817; 818) and with her exaggeration of the behaviour, which indicated that the "tantrum" was deliberate.
Humour was used here to diffuse the tension between what the students wanted me to provide (answers-as could be the norm in other psychology courses), and what I was able and willing to offer (my own ignorance-"whatever made you think] I had answers" [line 819]). My not being able to provide solutions, combined with my pleasure at their questions and uncertainty ("Great"), had the potential for dissatisfaction towards me. This was negotiated through the use of humour by Anna and me. Anna stated that the literature did not hold answers either ("and you can read millions]" [lines [821][822]), demonstrating further frustration. In her statement, she deflected the blame for their uncertainty away from me, indicating that even millions of readings could not provide the answers they were looking for. Elle added to the group frustration by suggesting that they already knew the answers ("The answers are within you" [lines [823][824]). This statement was also an act of collusion with me, as it advised the other students not to look to me for solutions. The group responded with a combined and loud "OOOOOOOOOOOOH", a joking, mocking expression in response to Elle's statement, and her collusion with me.

Self-Disclosure and Leading by Example
Foucault (2007) emphasised that the example that the shepherd sets is not one of perfection. It is important that the shepherd makes her/his failings known to the flock so that they may learn through her/his mistakes. Part of the exposure of such imperfections is through repentance and humility. The data revealed many occasions where I shared my own examples of experiences of working in communities, and my failures.
722 Lisa: Like when you went to that (.) the story that you told [us (.) that you 723 went to] the school ((talking to Elle)) 724 Anna: [Oh (.) that's toilet] 725 Lisa: And the school and the bucket and so many kids (.) and even though (.) What is interesting about this talk is that the roles of shepherd and sheep were reversed. In the talk, the students tried to offer me reassurance about my failed intervention. I had previously told them about the difficulties of intervening in a system and disrupting relationships, and I had used an example from my own experience to relate this lesson. Here Lisa referred back to that story (where the children had a bucket in the corner for a toilet) and reassured me, reminding me "you are this one person" (line 726). I refused her reassurance with "[b]ut I didn't make any difference" (line 727). There was then a rapid exchange of further reassurance and refusal ("Lisa: But how do you know that you didn't make a = Carol: = I didn't make any di nothing's happened nothing's changed") (lines 729-729), whereupon Lisa changed her tactic to commending my effort: "[b]ut still you made the effort" (line 730). She was joined by Mary and Anna in trying convince me that having tried was better than no effort at all. My overall scepticism was evident in my comment "[Let's keep] believing that (.) no I'm not being sarcastic" (lines 735-736), as I had to reassure them that I was not being sarcastic when I spoke about maintaining belief (keeping the faith).
This piece of talk tried to deal with a problem of disillusionment. The students assumed the role of shepherd here, comforting and reassuring me regarding my failed attempt at an intervention. Together, they took care to commend me for my efforts. The message that was conveyed here was that one must persevere against the odds, and that even one person can make a difference. They carefully used questioning "how do you know" (line 728) and suggestions of alternative outcomes to try to convince me that my attempts had not been futile. Thus, the confession of the failing of the shepherd indeed evokes renewed commitment to the cause from the flock. In their reassurance of me, they produced an argument for themselves to remain dedicated in the face of failure.
As a tactic, self-disclosure models the desired behaviour and encourages self-governing confessional practices. Modelling creates a norm for the flock to follow. This sets the standard for the behaviour of the sheep, which they are expected to achieve through techniques of the self, disciplining themselves to achieve the aspirations of the shepherd and supporting each other as each sheep struggles towards the goal. I hope that these two excerpts have managed to demonstrate, firstly, how the relationship is prioritised because it is through this that the effects of pastoral power are most powerfully achieved and maintained. Within this particular service-learning context, the strategies and tactics seemed to converge around the goals of promoting learning, preserving the relationship, fostering solidarity and promoting group cohesion. From the perspective of pastoral power, cohesion ensures the flock stays together and that the members are obedient to their calling. Solidarity is also about remaining true to the cause and operating within group norms. In addition to humour and laugher, this was also promoted through other tactics (not reported here). By rendering the work significant and worthy of attention, it ensured the sheep remained focused on their goals. A further goal in this service-learning context was to encourage commitment to "the work", ensuring appropriate and responsible community engagement was sustained. Self-disclosure and modelling by the shepherd can also be understood as a strategy for ensuring renewed or sustained commitment, where the sheep assist the shepherd to remain true, even when the shepherd fails. This modelling of the desired behaviour and confessional practices can also be seen as encouraging self-governance.

Conclusion and Implications for Research and Practice
I have found that conceptualising service-learning as an apparatus, and understanding the elements of that apparatus, has promoted my awareness that I am part of this apparatus, and provides cause for me to reflect on my practices in this context. I hope that this article encourages others to do the same.
Likewise, considering the service-learning field as a regime of truth also exposes how certain activities come to be normalised and promoted as "good", and how people and processes come to be reified and viewed as beyond question. Foregrounding this through a poststructural lens should hopefully remind us of our (radical) roots and prevent us from becoming complacent about our everyday ways of being.
I imagine that those who are drawn to service-learning because of its possibilities for resisting and challenging the status quo in the academy will be dismayed at my analysis of my practice. I would argue, however, that this study provides opportunity for hope, that if we regard students as active participants in the service-learning process, if we provide space for resistance and freedom, students may be more enabled to choose the kinds of subjects they wish to become. It is my hope that precisely because we are drawn to service-learning, we would be open to exploring these spaces.
Lastly, if we recognise that we are always within power and we can never be outside of it, and that power is a productive force evident in the interactions between people, then we can be liberated to identify how power flows in the service-learning context, and what we can do to mediate that flow. Being aware that we are not engaging in a neutral practice, and that we are always promoting some kind of agenda (even when we may not be aware of that agenda), requires us to engage in our own self-governance. It demands that we reflect on our practices, that we confess the "truth" to ourselves, and that we shape our future selves and practices around that truth.
If service-learning practitioners can embrace this kind of lens, it opens up the possibility of problematising our other taken-for-granted practices: assumptions regarding community participation, voice, benefit (cost); considering student or systemic resistances to service-learning; exploring the power effects described above across different models of service-learning; further critiquing the goals or outcomes of servicelearning and other educational practices. If we consider that our assumptions and everyday practices may be dangerous, then we have work to do.