Interactional power : observing and identifying power in interaction analyses of adult education situations depending on power notions and data types 1

Since the 1970s, various aspects of power have been at the focus of theoretical and empirical adult education research. Despite the actual interest in political and discursive aspects of power, this article emphasizes the importance of interactional studies when observing and identifying power based on various types of data. As for German interaction studies, three phases can be distinguished, characterized by a) observations of failed participation based on records of classroom behaviour, b) the identification of mutual power negotiation in classroom and counselling situations based on transcriptions, and c) the identification of the power of physical settings in adult education classrooms and in counselling sessions based on visual data. It is presumed that observing/identifying power in adult education classrooms and counselling sessions generally depends not only on the notions of power underlying the studies but also on the data types produced and the methods applied for their interpretation. In addition, the question is raised whether the identification of power can be considered a power practice used by adult education researchers.


Introduction
From its very beginnings, adult education has been legitimized to a large extent by pointing out that large parts of the adult population were deprived of power: the power to participate in politics and society, the power of self-determination, and the power of intellectual autonomy.According to Elias and Merriam (1980), the diverse philosophies that influenced the theory and practice of power struggles in adult education can been identified as liberal, progressive, humanistic, and radical.One of the main functions of adult education, therefore, seemed to be to compensate for social, educational, and individual disadvantages (Pöggeler, 1975; Olbrich, 2001 2 ).Even today, the existence of disadvantaged groups is used to justify adult education and lifelong learning -although not necessarily in opposition to other justifications, such as enabling individuals to adapt to societal, technological, and economic changes (cf.Kraus, 2001).
The traditional focus on the empowerment function of adult education was shaken by the suspicion that adult education itself might be an instrument of power.The idea is not entirely new.In Germany it can be traced back to 1872, when Wilhelm Liebknecht delivered his famous speech, 'Knowledge is power -power is knowledge', in which he pointed out that non-political bourgeois education reinforces the political and social conditions oppressing the workers, who used to be the main target group of adult education in the nineteenth century.So it is possible to draw a line from early Marxist theory to the idea that the adult education classroom 'is a duplication of the existing societal relations of power replete with hierarchies and privileges conferred along lines of gender, race, class, sexual orientation and other status markers' (Johnson-Bailey & Cervero, 1997, p. 243), on the one hand, and current ideas of governmentality as influenced by the works of Michel Foucault (cf. Fejes & Nicoll, 2008) on the other.
The critical focus on social injustice and inequity, as well as the modern concept of governmentality, has had a remarkable influence on empirical adult education research.This affects research on participation in adult education, including questions of access and inclusion (see e.g.Sargant, Field, Francis, Schuller & Tuckett, 1997;Jackson, 2011), and critical analyses of official documents on adult education and lifelong learning (see e.g.Edwards & Nicoll, 2001;Brine, 2006) following the Foucauldian approach to lifelong learning as intrinsic to contemporary technologies and strategies of power.
In view of the remarkable change that policy discourse has undergone in recent years (cf. Wildemeersch & Olesen, 2012), adult education research turns out to be more interested in the political and discursive aspects of power, and less in the interactional ones.Interactional perspectives that emphasize symbolic interactionism, constructivism, and performance theory, and focus on micro-level phenomena like the negotiation of status and power, collaborative meaning making, and bodily practice.
While research on teaching and learning in adult education based on observation and tape-recording of classroom sessions seems to have lost its significance in recent years, interest in interactional research has been rekindled by new forms of (technologybased) teaching, the increasing importance of counselling and guidance in adult education, and new ways of recording (visual) data.This change of data is accompanied by a shift of focus to different forms and manifestations of power, which will be discussed in detail below.The present article, therefore, aims to demonstrate the (ongoing) importance of research into interactive power in adult education situations, both in the classroom and in counselling sessions, and to draw attention to the influence exerted not only by theories of power but also by various types of data used in empirical research.To this end, I will present an outline of the history of empirical research on interaction in adult education classrooms and, to a minor extent, in adult education counselling sessions since the 1970s, with special emphasis on the type of data collected and interpreted.

Observing and identifying power in studies of adult education interactions
In view of the fact that empirical research data are always rooted in national contexts and influenced by local infrastructures, it seems reasonable to concentrate on studies conducted to a particular cultural and historical practice.That is why I will focus almost exclusively on (West) German 3 studies of adult education interactions (for an overview see Nolda, 2010) when reviewing the various aspects of power observed or identified 4 .

Observing failed participation based on records of classroom behaviour
The early 1970s saw a series of studies devoted to opening the 'black box' of the adult education classroom as a result of growing political interest in adult and continuing education as a fourth sector of the public education system, various political documents and laws fostering adult education in the German states, and the establishment of departments of adult education at the universities 5 .Research in this field was largely based on tape recording and methods examining the verbal interaction between participants 6 .
One of the first studies on adult education classroom interactions in Germanspeaking countries (Schalk, 1975) explored the impact of language barriers on discussions between lower-class and middle-class participants in Austrian community adult education centres (Volkshochschulen).Transcripts of the discussions were analysed according to the categories of elaborated and restricted language code use (e.g. the use of hypotactic or paratactic sentence structures) developed by Bernstein (1971), who asserted a direct relationship between social class and language.Schalk reported that middle-class adults, when engaging in discussions with lower-class adults, were able to switch from their own elaborated code to the restricted code of the lower-class participants (at least when speaking about concrete topics), whereas members of the lower class were not capable of making such an adaptation.The study showed how speakers using only the restricted code were excluded from discussions about more abstract topics, which could be seen as an argument for using and teaching the elaborated code in adult education classes.
A study conducted by Siebert and Gerl (1975) aimed to develop a didacticmethodological theory of adult education based on precise knowledge of actual teaching events in adult education to check the implementation of adult education strategies and postulates, and to develop tools for analysing and planning adult education.The study design was based on the conviction that adult education could promote the 'democratization of all social sections' if it were possible to enhance the ability of students to actively take part in courses.This referred, on the one hand, to the selection of suitable learning subjects and, on the other hand, to the establishment of forms of interaction enabling learners to articulate their learning interests and influence their learning processes (Ibid.,p. 24).Interactions in adult education classes in German community adult education centres were observed by classifying the contributions of learners and teachers according to their didactic function, distinguishing between content orientation and process orientation.The authors of the study found that the majority of classes observed were content-and teacher-orientated, that most teachers presented themselves as experts, that students preferred to ask teachers (and not other students), and that students showed more interest in the acquisition of knowledge, whereas teachers were more interested in the problematization of knowledge (cf.Siebert, 1975).The design of this study was influenced by the interaction analysis technique developed by Ned Flanders (1970), consisting of an 'objective' and systematic observation of classroom events, especially along the qualitative and quantitative dimensions of teachers' (mostly dominant) verbal behaviour in the classroom.
Another study from that time by Weymann (1977) combined educational and sociological issues.For example, it analysed whether those who failed in other educational systems really got a second chance in adult education, what might be the reason if they didn't, and what sort of pedagogical conclusions should be drawn to improve the situation.Referring to Bernstein's code theory and the theory of metacommunication, Weymann applied categories such as intentionality, reciprocity, digitality, analogy, and dominance.He measured how much and how often participants spoke, and how teachers dealt with students' contributions.As a matter of fact, it was found that personal and evaluative statements by lower-class students were only reluctantly accepted by teachers belonging to the middle classes.Students, on the other hand, were mostly reluctant to meet the teachers' demands for providing critical summaries of the groups' learning processes (Ibid.).
Thus the relevant finding of this study was that it identified the dominant (verbal) behaviour of teachers and the inefficacy of their efforts to encourage socially and educationally deprived students to verbalize critical attitudes.The same is true of a study focusing on paid educational leave (Bildungsurlaube) organized by trade unions (Kejcz, Monshausen, Nuissl, Paatsch & Schenk, 1979-1980).Observation protocols and tape recordings of classroom interaction were analysed to check whether the principle of learner or participant orientation (Teilnehmerorientierung 7 ) was realized.Researchers wanted to know in detail whether teachers responded to students' experiences, how competencies were distributed among participants, whether classroom participants agreed about contents, and how participants' interpretative patterns (Deutungsmuster 8 ) were discussed.Researchers found evidence that participants were often talking at cross-purposes and that misunderstandings were mostly ignored.Observation protocols and tape recordings suggested that social injustice was, in a way, duplicated in adult education: the mechanisms that were found to prevent participation in adult education were similar to those that prevented participation in society 9 .
Studies of this period are characterized by an approach that attributed power to a group (members of the middle classes/teachers) and then scrutinized adult education classrooms for evidence of whether or not members of the superior group (the middle classes/teachers) enabled the participation of members of the subordinate group (members of the lower classes, students).Power, or lack of power, was conceptualized as being basically stable.By measuring the quantity and distribution of certain items of verbal behaviour, the exertion of power through individual speakers or a group of students was made 'visible'.This was achieved on the basis of observation protocols and orthographic transcriptions of spoken language, which were however not always included, or only partially included, in the study reports.

Identifying mutual power negotiation in classroom and counselling situations based on transcriptions
The critical impulse of dedicated adult educators committed to fighting social injustice in the 1970s and 1980s was attenuated by the establishment of academic adult education and the necessity to adopt the impartial norms of science.As a consequence, methodological accuracy and quality came in focus, and researchers did not necessarily feel obliged to directly advance the quality of adult education practice or combat exclusion.
In the 1990s, the rise of the qualitative paradigm and symbolic interactionism led to studies that replaced the attributive notion of power by a relational one.Their work was based on the assumption that power relations between students and teachers were produced interactively.Power was not seen as being equally distributed between partners but as being essentially dependent on the existence of partners (cf.Luhmann, 2003) and above all as being inherent to any interaction.And most importantly, social differences and hierarchies were not understood as given but as mutually produced and even changed by interaction.That is why written reports included meticulous, fulllength transcripts-a methodological constraint taken from conversation analysis, enabling readers to examine and review the researchers' interpretations (cf.Psathas, 1995).
The idea that hierarchies were not given beforehand but had to be claimed and negotiated even applied to interactions in which adult students were examined by teachers.A study of foreign language examinations demonstrated the ways in which examiners as well as examinees try to impose their will on one another and reach an agreement about this.The only (or preferable) way to trace these interactions is via meticulous transcripts marking silences, slips of the tongue, overlapping speech, and intonation curves (Nolda, 1990).The data were interpreted following, on the one hand, the principles of conversation analysis, strictly considering the sequential order of interactions and even minute details.On the other hand, the interpretation was following the principles of 'objective hermeneutics' 10 , aiming at the utterance meaning as distinct from the speaker's meaning.This was accomplished especially by discussing different or even controversial readings of certain passages.
The same method was applied in a study concerning classroom behaviour in liberal adult education (Nolda, 1996).Based upon line-by-line analyses of classroom sessions, categories were generated that referred to aspects of power such as reacting to the institutional lack of power, indirectly exerting power by organizing lessons, or selfpresenting and establishing group identity.Claims for power could be identified both on the part of teachers and on the part of students.Teachers, for instance, preferred to present themselves as experts who often tried to ignore opinions other than their own, whereas students sometimes used lessons as a stage to present their knowledge and themselves as morally superior persons.
That some findings of earlier studies were confirmed and others questioned was above all the result of a more scrupulous and methodologically controlled analysis.But it was also in part due to the fact that adult education itself had changed.We should keep in mind that empirical research of this kind does not automatically produce final descriptions of adult education irrespective of the time and place where the data were gathered.To a certain extent, interaction analyses can therefore be used as an instrument for recognizing and defining characteristics of a certain period, a certain field, or a certain type of adult education institution.
The way teachers deal with adult learners' interpretative patterns was studied by Schüssler (2000), who carried out a detailed analysis of two adult education classroom lessons: one in which these patterns were made explicit by the teacher and another in which the teacher avoided confrontation with them.The teacher's intervention (in a vocational training setting), which may be seen as an act of power -though meant as an offer to improve students' self-awareness and autonomy -often caused resistance.So power became evident both when maintaining these patterns and when questioning them.The concept of interpretative patterns and the andragogical claim to intervene or take them into account, discussed widely in contemporary German literature (cf.Arnold, 1985), led to analyses of the way in which patterns of interpretation influence interaction in adult education classrooms.Dealing with interpretative patterns was shown to be a special form of power negotiation in the adult education classroom.
Another facet of interactional resistance was described by Gieseke and Robak (2000), who provided a detailed interpretation of a videotaped seminar for persons working in adult learning.Seminar participants resisted the deductive models presented by the teacher; the teacher resisted the wish of the participants to discuss their daily work.Both did so in a rather indirect manner.So the interaction was characterized by an alternating dominance of participants talking and the teacher explaining theoretical models.
In the 1990s, guidance and counselling 11 , as a generally learning-oriented process, became an important part of adult education practice and a widely discussed topic in the literature on adult education (Projekt SOPEK, 1991;Eckert, Schiersmann &Tippelt, 1997).Disse (2005) reconstructed a consultation session that was part of a compulsory training programme for unemployed persons financed by the Federal Employment Office; Müller (2005) examined counselling sessions for people seeking advice about adult education offers; Maier-Gutheil (2009) analysed consultation services for people planning to launch a new firm.
Like learning situations, counselling situations are characterized by genuinely asymmetrical relationships between (seemingly) superior teachers/counsellors and (seemingly) subordinate students/clients.Interaction studies therefore served to produce a microanalysis of the subtle dynamics of power going on in semi-public adult education classrooms and non-public counselling sessions.
Based on their meticulous analysis of accurate, full-length transcripts, Disse (2005) and Maier-Gutheil (2009) found that, faced with the danger of becoming objects of administrative measures, clients are definitely able to resist -in an interactive mannercounsellors' power claims, and that counsellors in turn interactively deal with the resistance of clients.Müller's study underlines the differences between three types of counselling -information-oriented, situation-oriented, and biography-oriented counselling -showing that it depends on the specific type of counselling whether criteria such as length of utterances have to be considered as signs of power (claims).
In the above mentioned studies, the use of meticulous transcripts made it possible to reconstruct, on a micro level, subtle interactional practices of claiming, maintaining, and resisting power and to enhance adult educators' knowledge about the interactional dynamics of classroom sessions and counselling situations.

Identifying the power of physical settings in adult education classrooms and in counselling sessions based on visual data
At the beginning of the twenty-first century, the social and pedagogical relevance of the dimensions 'space/environment' and 'body' was increasingly acknowledged by theorists and researchers in education and the social sciences (cf.Ecarius & Löw, 1997;Langer, 2008), accompanied by a growing interest in visual data (cf.Pilarczyk & Mietzner, 2005;Dinkelaker & Herrle, 2009).
Visual data like photographs and videos made aspects of power visible that had hitherto been neglected, especially the impact of the classroom as a physical setting and the non-verbal behaviour of teachers and students.Special attention was paid to the teaching and learning materials and to the persons being present: their looks, posture, facial expressions, and gaze movements.
This tendency was strengthened by renewed interest in the ethnographic approach (cf.Hünersdorf, Mader & Müller 2008), which tried to capture the complexity of classroom life (cf. Watson Gegeo, 1997), including spatial relationships: 'We learn about power in adult education by studying the micro-dynamics of particular learning groups in particular classrooms (the gestures, body postures, seating arrangements, facial tics and phrases that learners and teachers commonly utter)' (Brookfield, 2005, p. 126).
Although rooms used for adult education lessons are not always specifically prepared for this special use, and the time teachers and learners spend in these rooms is much shorter than in school or university, they do show underlying concepts of teaching and learning and of the power relations inherent in them.This refers not only to settings suitable for lecture-style teaching but also to settings suitable for group discussions, such as chair circles which place each participant in a state of constant visibility.In accordance with ideas proposed by Foucault and the theory of governmentality, an arrangement like this can be seen as an instrument that makes students 'prisoners of a power situation fostered by themselves' (Klingovsky, 2009, pp. 161-126.).Perhaps unlike school classrooms, rooms used for adult education lessons are not only signs of a 'pathic power' (cf.Schultheis, 2007) to which teachers and students have to surrender.They also suggest certain appropriations -suggestions teachers and students are to a certain degree free to act on.Identifying the ways in which teachers and students deal with the spatial structure of classrooms may give interesting insights.Based on videos and room sketches, strategies of adaptation, change, or avoidance (e.g. by rearranging the furniture, crossing borders, taking a seat near to or far from others) could be traced (cf.Nolda, 2006).
Videos can show in detail how teachers and students deal with artefacts like blackboards, computers, or computer-based presentations.The ubiquitous PowerPoint in particular has been the object of video-based interaction research (cf.Schnettler & Knoblauch, 2007).The question of power first of all involves asking which persons exclusively or primarily use media such as black and white boards, flip charts, and projectors.Of nearly equal importance is the way in which these persons use them: writing on a board, for instance, may indirectly support and enforce the opinion held by the person having access to the board (cf.Kade, 2014).
Analysing non-verbal phenomena in interaction analyses-strongly recommended by Goffmann ( 2005)-presents specific challenges.For example, it is almost impossible to identify a superior physical habitus 12 based on videos showing adult learners in classrooms during lessons in which only little freedom of bodily action is allowed, or during lessons especially designed to train bodily actions, which therefore cannot be counted among the 'natural' qualities of a person.Instead, it makes sense to identify exactly how teachers create distance and proximity to students, how students react, form axes of interaction together, or exclude others (cf.Kendon, 1973), and how those who are excluded react to being excluded and how power can be established by the visible unwillingness to communicate (cf.Herrle & Nolda, 2010).
Gestures like raising a finger, whether isolated or accompanying speech, are not automatically signs of exerting power, but have to be looked at in their contexts and in connection with other behaviour occurring simultaneously.That is why the methodologically controlled analysis of photographs and stills/frames, in which -in line with Panofsky's (2006) model -researchers distinguish between the pre-iconographic and the iconographic level, trying to define a person's 'habitus' (cf.Bohnsack, 2008), is a way to gain insights that go far beyond the possibilities offered by participant observations or transcriptions of recorded speech.On the other hand, the movements of persons and their speech are often indispensable for understanding the situation in which a gesture is embedded (cf.Streeck & Knapp, 1992).
Video analyses of adult education classrooms comprising liberal adult education, foreign language lessons, vocational trainings, physical education, and the like (see e.g.Herrle, 2007;Simon, 2008;Kade, Dinkelaker & Herrle, 2009;Karisch, 2010;Schindler, 2009;Kade, Nolda, Dinkelaker & Herrle, 2014) demonstrate the 'power' of the visual approach in detecting non-verbal ways of exerting and fighting interactional power.One of the main aspects explored in these analyses is eye movement.The relevance of eye movement, emphasized early on by some representatives of conversation analysis (cf.Goodwin, 1979;Heath, 1997), does not only relate to the behaviour of teachers seeking to get or intensify students' attention but to all persons present in the classroom.Power dynamics can be detected by tracking deviations from the expected direction of vision (mutual and synchronous looking at interactions partners).Avoiding eye contact, for instance, might be meant and understood as a subtle form of denying recognition (cf.Schaffer, 2008) or as a gesture of submission (cf.Dinkelaker & Herrle, 2009).Unlike language-based data types, which are derived from participant observation or recordings, videos as original data can be watched as long and as often as possible so that even minute details like short and unspectacular movements, gestures or gaze movements can be captured in a way that would have been impossible in the real situation.In addition, exchange of glances and gestures between students, which are normally out of teachers' sight, can be used for the interpretation of comments on power claims.
Another way of identifying interactional power relations with the help of visual data is by analysing official photographs showing learning or counselling situations.At first glance, they represent -unlike tape and video recordings by researchers -the way institutions want themselves or their work to be seen by the public or by their clients.With the help of scrupulous analyses it is nevertheless possible to detect 'hidden' or even officially denied aspects of power.Wilke (2011), in a study using the documentary method for analysing visual data developed by Bohnsack (2008), showed how the idea of lifelong learning is visualized in documents placed on the website of the European Commission.As a matter of fact, one of the photographs he analysed showed rather elitist (academic) learning settings, where older people are presented in a way that marks them as belonging to the middle classes but not really belonging to the academic setting.So the idea of (social and generational) inclusion championed in EU documents seems to be in part contradicted by the photograph, which reveals a paternalistic attitude officially denied.Furthermore, the idea of lifelong learning is visualized as the passive reception of information, seemingly without the possibility of actively participating or interacting.
Even in studies based on tape recordings, additional visual data can offer valuable insights.In a study on interaction in adult education counselling (Stanik, forthcoming), photographs of counselling situations used by institutions in their online selfpresentations were analysed in order to confront visual self-presentations with the findings of the empirical research based on audiotapes.One of the photographs of a counsellor and an advisee shows that it is only the counsellor who has access to a computer and various information materials neatly positioned behind her back.The counsellor thus appears as a representative of an institution that seems to provide wellstructured information but is not directly involved in discussing problems that might occur with the client.It is the spatial power of access to information that is visualized indicating the dimension of knowledge power.Both photographs can be seen as unwitting demonstrations of (subtle) institutional power that is denied in official statements by the institutions in question.

Conclusions
As shown above, identifying power by means of interaction research depends not only on different notions of power but also on the data that are used or produced and analysed.Referring to attributional notions of power, analysts of written reports on classroom behaviour found out that the verbal strategies of middle-class teachers prevent learners from participating.Referring to relational notions of power, analysts of spoken language revealed interactive self-presentations and power dynamics by dealing with learners' interpretation patterns.Based on pictures and stills, rooms and artefacts can be understood attributionally as means of exerting power; based on videos, they can be interpreted as challenges that can be met in various ways.The careful analysis of photographs and stills offers insights into the subtleties of non-verbal power, questioning the idea that interaction and interactional power is predominantly a verbal, rationally controlled phenomenon 13 .With the help of audio-visual media, the way power is claimed and/or questioned through gazes can be tracked in (almost) every detail.
It is no question that research on various aspects of power in adult education has been strongly influenced by diverse theories of power.The example of the development of interaction research in adult education shows that the impact of the type of data used, and of the method of interpretation applied, is equally important.So the observation and detection of power also depends on technical progress and the development of new methods.
To say that certain individuals or groups exert or question power is, in a way, itself a practice of power, because it demonstrates the ability to look behind the surface of behaviours and actions that usually go unnoticed.The question whether theorists of power and adult education researchers contribute to the detection of hidden or subtle power mechanisms in adult education situations or in education generally (cf.Bilstein, Keiner, Ecarius & Wimmer, 2007) is therefore also a question of the hidden power of theorists of power and adult education researchers themselves.That comprises not only moral objections to the exertion of power by practitioners.The question that arises is: to what extent is the identification of power in educational situations by theorists and researchers more or less strongly influenced by their suspicion or assumption of its existence -that is to say, to what extent are researchers simply searching for affirmations of what think they already knew?
Theories and studies demonstrating that power is exerted by someone or something hitherto or usually considered beyond suspicion attract attention -a temptation which is possibly hard to resist.In order to clarify these doubts, it would be useful if researchers generally disclosed not only the notion of power they adhere to but also their reasons for choosing a certain method and a certain data type, outlining the implications of these choices.This means that a clear distinction has to be made between general notions of power and empirical evidence of power relations based on specific data types.
So in addition to the finding that adult education is a site of power, it can be assumed that adult education research involves power practices, too.This assumption is in accordance with Foucault's notion of the all-pervasive nature of power (cf.Foucault 1991) and is informed by Luhmann's (2003) insight into the productive power of observations of the second order.Notes 1 Modified, extended, and updated version of Nolda (2010).
2 For an overview of critical approaches to education in general adult education from an international perspective see Westwood, 1996.  3 For an overview of German adult education, its historical development, legal basis, institutions, and so on, see Nuissl and Pehl, 2004.  4 The article might therefore also serve to support the dissemination of research and scholarly writing across language barriers (cf.Fejes & Nicoll, 2013). 5The relevance of the establishment of professors in adult education for research in this area is underlined by Fejes & Nicoll, 2013.  Early studies in adult classroom interaction were based entirely on notes written during lessons (cf.Seitter, 2010). 7For a detailed analysis of the principle of Teilnehmerorientierung and its current meanings in German adult education, see Holm, 2012.  The term Deutungsmuster, referring to schemes of perceptions and meanings that prefigure the understanding of the world, was introduced to German empirical social science by Ulrich Oevermann (1973/2001) and goes back to the lifeworld philosophy of Alfred Schütz. 9The same model was applied to the observation of interactional power in adult education classrooms resulting from gender differences (see Hoverstadt, 1997). 10For an overview of the concept of Ulrich Oevermann's 'objective hermeneutics', see Reichertz, 2004.  1Because the major concern of guidance is with the decision as a product, and because counselling is more concerned with the process of decision-making (cf.Potter, 1996), both counselling and (modern) teaching mainly aim at enabling and facilitating. 12A person's physical habitus or -as Foucault puts it, the hexis -shows his/her systematic relation to and his/her position in the social world (Bohn, 1991) 13 An illuminating example is provided by Pielarczyk and Mietzner (2000), who analyzed an official photograph from the late 1980s showing a school class in the German Democratic Republic, in which a political ritual was performed in a way that contradicted the enthusiasm officially required and formulated in the caption.