education policy analysis archives

: This conceptual article examines how consultants use a mundane policy device, the powerpoint presentation, to manage education policy relations between international lenders and education ministries in the global South. The article theorizes presentations as socio-material assemblages that combine consultants, software, visualization conventions, presentation styles


Mundane Technologies of Educational Policy Consultancy: Presentations and Interpretive Control
The meetings themselves soon took on a pattern. I would show a powerpoint presentation and take the Chief Minister through the Roadmap actions; we'd pause on the Reds and I would try to explain why it was Red and what we planned to do about it. . . . We standardised the graphs to make them instantly recognisable from one stocktake to the next. . . . We introduced the "heat maps" which showed the Chief Minister which districts were on track and which weren't. . . . The Chief Minister loved the heat maps: "I shall sleep with these under my pillow," he said. (Barber, 2013, p. 62) The use of technologies in education policy has become a major research area, with much of the emphasis being placed on developments in datafication, surveillance, algorithmic policy, and platformization (e.g., Grimaldi & Ball, 2021). But older, mundane technologies, technologies that have become "soft and often invisible (that is, forgotten or not recognized as technologies)" (Czarniawska, 2004, p. 775), remain important, especially for controlling the meanings of information as it flows across extended networks. Critical perspectives on education policy suggest that such meanings mutate as they move and that network nodes can be sites of interpretive conflict. This article examines how transnational education policy consultants draw on a particular mundane technology-the powerpoint presentation-to manage such meanings and tensions.
Along with other researchers we use powerpoint as a generic term to refer to a relatively standardized and widely distributed presentation system that includes not only software, which may be Microsoft's or an alternative with similar features (e.g., LibreOffice Impress, Google Slides) as well as large ensembles of communicative forms with which the software has become entangled, such as standardized practices of slide construction (e.g., Knoblauch, 2013).
Powerpoint is a ubiquitous element of state and corporate practice (e.g., Kernbach et al., 2015, p. 292;Stark & Paravel, 2008, p. 32), perhaps even constitutive of consultancy as a profession (Schoeneborn, 2013(Schoeneborn, , p. 1778. It is an infrastructural element of policy in the sense that it is woven into the conventions of policy fields and learned as a condition of participation, and it enables participants from different socio-cultural backgrounds to share information and coordinate action across space and time "in a natural, ready-to-hand fashion" (Star & Ruhleder, 1996, p. 114). Like other technologies, powerpoint presentations congeal and mask biases, rules, and dispositions about whose knowledge counts and what forms it can take. They are a device through which the meanings of policy information can be controlled. This is primarily a conceptual article interrogating how this device is constituted and put to work by consultants to manage the meanings of information in education projects in the global South. By calling it conceptual, we mean that instead of analyzing the formation, movement, or implementation of a particular policy, we focus on the features, functions, and uses of powerpoint that consultants draw on to structure relations of epistemic authority in education policy processes. Thus we anchor our arguments in a range of evidence gleaned from the multiple literatures, internet sources, and interviews with local experts in a South Asian country where consultants oversaw a long-term, internationally-funded education project. Transnational consulting of this sort has been described as a form of "outsourcing" in which the models and programs of external experts displace "the long-term work of civil servants and even politicians" (Ylönen and Kuusela, 2017, p. 242;cf. Gunter & Mills, 2017). But outsourcing is not quite accurate. Consultants perform key parts of their work inside local settings during short, periodic visits. They depend on local experts as "important interlocutors between international and national policy elites, and as gatekeepers to both" (Kamruzzaman, 2017, p. 43;Ocampo & Neu, 2008). To understand these relations we focus less on how consultants help policy move (e.g., Ball, 2016) than on how they maintain control of policy that's already been brought to ground. How do consultants manage interaction and public interpretations of evidence in meetings when they don't speak the local languages and don't have first-hand knowledge of the situations they're talking about? How do they use powerpoint presentations to structure the visibility and legibility of information? How do these dynamics maintain epistemic hierarchies that privilege consultants' views over those of local experts? How are presentations used to maneuver local experts into public acquiescence to consultants' interpretations?
The standard answers to such questions-that consultants draw authority from credentials, reputation, and technical skill, or that local actors are socialized to accept the consultants' assumptions and analytic practices -understate the skepticism of local experts and their tensions with consultants (e.g., Koch, 2020), and ignore the performative strategies and technologies that consultants use to forestall or dampen such tensions.

Data and Method
This conceptual analysis forms a "constitutive argument" (Pacewicz, 2022) about the uses of presentations in structuring the epistemic relations of consultants and clients in transnational educational consultancy. As Pacewicz (2022) explains, constitutive analyses are "analytical descriptions about the makeup and useful categorization of social phenomenon, which commonly focus on existence (i.e., 'X exists'), casing (i.e., 'X is a case of Y'), or categorization (i.e., 'X consists of subcomponents A and B')" (p. 933). Such arguments logically precede causal analyses inasmuch as they attempt to clarify what it is that needs to be explained. In the present case the ubiquity of powerpoint presentations removes the need for an existence argument. Instead we proceed with what Pacewisz calls casing and categorization. As a form of "casing" we examine how presentations can be understood as technologies of epistemic authority. As a strategy of categorization we put forward a model of three key subcomponents of powerpoint presentations: the use of visual coding schemes, the ritualization of the presentation format, and the mobilization of embodied linguistic and performative capacities that enable presenters to dominate interaction (cf. Pacewicz, 2022, pp. 945-946).
The way we go about this also draws on a kind of abductive analysis that begins with "anomalous and surprising empirical findings" and then systematically interrogates them with "multiple existing sociological theories" (Timmermans & Tavory, 2012, p. 169). One anomaly that caught our attention was that despite its prominent role in policy events and its acknowledged importance as a tool of consultants' work, powerpoint remains relatively neglected in education policy analyses. A second anomaly was the fact that local experts, though they had deeper and more contextualized knowledge of local policy implementation and its effects, did not voice their disagreements with the consultants' interpretations. We interrogate these findings with theories that examine powerpoint presentations as forms of visual or attentional control, as ritual, as extraterrorial sites, as enactments of post-colonial linguistic ideologies, and so forth. As Timmermans and Tavory (2012) put it, each theoretical engagement "abstracts and highlights different aspects of the phenomenon, rendering it comparable to different phenomena and turning it into a generalization that then can be linked to other fields and theories" (p. 177). The multiple theories work both as ontological acids for breaking down familiar ways of thinking about powerpoint and consulting relations, and as elements of an analytic toolkit for highlighting essential features of the presentations. As is common with qualitative research there is no neat separation of theory from method or findings: they are woven throughout the article.
This kind of analysis does not construct theory solely through inference from data. Rather, the data suggest analytic "puzzles" (Mears, 2017)-for example, why use visual codes if they confuse clients?-and provide evidence of the range and variation of phenomenon. For this reason, we looked for material from diverse cases as a "maximum variation" strategy of selection (Flyvbjerg, 2006). Some of that case material comes from a set of semi-structured narrative interviews (Seidman, 2004) with 12 current or former members of a state agency responsible for mediating relations between international funding organizations such as the World Bank and the education ministry of a South Asian country. Access to the site was made through the first author's personal connections. The interviewees' activities range from liaison work to managing aspects of the implementation of multi-year, internationally-funded education projects. We cannot describe the agency in more detail without making its identity apparent and violating the confidentiality of our participants, all of whom are referred to here by pseudonyms. It is important, however, to note that the interviewees describe a situation in which consultants have been working with the local education agency for an extended period of time. It is possible that the uses and forms of presentations in such contexts would differ from those, for example, in the initial stages of a consulting relation.
These interviews initially drew our attention to the significance of powerpoint presentations. We then extended the range of cases by examining accounts of powerpoint uses written by consultants (as in the introductory quote from Barber, 2013), as well as descriptions of powerpoint presentations in corporate settings (e.g., Alvesson, 2011), internet discussions of slide construction, online archives of slide sets prepared by consultants, and of course the large academic literature on powerpoint, discussed in sections to follow. To find this material we conducted systematic websearches to locate examples of consultants' uses of powerpoint in publicly available slides, presentations, and reports. We began by searching for material online related to the particular projects described by our interviewees. We then used simple searches with the Google search engine, using as search terms "powerpoint" with the names of particular consulting firms (e.g., McKinsey), education agencies, and reform projects. We also used the filters or "prefix operators" of the search engine to identify additional material. Finally, we used the keywords "powerpoint" and "presentations" with Google Scholar as well as the websites of journals in education policy, international relations, and organizational sociology to explore academic literatures on the topic.
We do not claim that the materials thus assembled represent the full range of powerpoint presentations in educational consultancy. Many presentations are not recorded and many slide sets are not saved, and those that are may not be publicly available, although some become accessible when posted to internet sites by participants or released through lawsuits or government investigations (some consultants now anticipate such releases by inserting disclaimer slides, for example, "this material was used by McKinsey & Company during an oral presentation; it is not a complete record of the discussion" [Masnick, 2021]).
An additional problem for researchers is that, as was the case here, the slides one can obtain contain data, charts, maps of states or provinces, or pictures of individuals that identify the country, consultants, and agencies involved. In most cases it is not feasible to alter this material in ways that would both convey its actual forms and uses while at the same time maintaining the confidentiality of participants. As a result, to maintain confidentiality and protect participants, we draw on published or publicly accessible materials to illustrate certain points, and draw empirical examples from the published literature as well as from interviews. We begin by describing the kind of consulting we are focusing on, then take up the analytic puzzle of how powerpoint presentations work as technologies of interpretive control.

Consultants and Informational Control
Policy consultants can be defined as "non-state, private-sector, profit-driven actors that are nevertheless involved in the policy process through (usually) contractual arrangements with state agencies" (Prince, 2012, p. 195). It follows that transnational consultants do these things across national borders, moving and managing flows of ideas, money, people, and practices. Although contracted to the state, they are often hired at the behest of the funding organization. In addition to helping plan, monitor, and evaluate funded projects such consultants function as "reputational intermediaries" (Stone, 2004) who speak the funders' language, "know how to play by the donors' rules," and can be depended upon to generate informational products on schedule in the proper form (Kamurzzaman, 2016, p. 52). Their presence implies that things will move smoothly: that as money is disbursed, projects will be put in motion, targets will be achieved, and data will be collected, analyzed, and circulated back to the funding organization.
But the smoothness of these relations may be illusory. Interactions among consultants, funders and state agencies are complex in part because the state is complex. As Bok and Coe (2017) complain, "whenever the state is acknowledged by policy mobility scholars-generally infrequently -it tends to be treated as a structural, functional backdrop to be negotiated, rather than a dynamic coalition of actors" (p. 53). These dynamic coalitions form governance ecologies full of "social contradictions and political struggles as well as internal conflicts and rivalries among its branches" (Jessop, 2010, p. 49). Transnational consultants forge different kinds of relations with different elements inside such ecologies. Their ability to move and manage policy depends not only on the support of political decision-makers, but on their ability to shape the interpretation of program data and marginalize sources of alternative interpretation, in particular the staff within state agencies (Ocampo & Neu, 2008). Variously labeled "local/national experts" (Koch, 2020), "national development experts" (Kamruzzaman, 2017), "local development practitioners" (Olwig, 2013), implementation agents, or the "implementariat" (Peters, 2020), such staff often have training in Western universities and work experience with international organizations (Broome & Seabrooke, 2012). This positions them to perform the "interpretive labor" needed to translate between local institutions and consultants and global funders. The concept is Graeber's (2012), who argues that "within relations of domination, it is generally the subordinates who are effectively relegated the work of understanding how the social relations in question really work," and in particular of figuring out how those at the top think (p. 118). Peters (2020) extends this concept to argue that the "implementariat" in development relations (a category that includes those we are calling local experts) is "tasked . . . to understand what is important to development's elite and act accordingly" (p. 65). A key part of this work centers on management and evaluation activities in which local experts "are meant to communicate the progress of their work to their institutional superiors" (Peters, 2020, p. 65). These communicative roles make the local experts examples of "symbolic intermediaries who manage the work of inscription and representation at the ground level, and "then convey. . . those representations to a professional"-a consultant or World Bank analyst-"who operates on the representations to synthesize a more complex symbolic product" such as a written report or powerpoint presentation (Barley, 1996, p. 421).
Such work requires us to modify the standard picture of official statistics. The usual account describes them as products of a progressive simplification of information in which data collected by street-level or lower-status workers is "manipulated, parsed, and moved upward" through a process that "removes assumptions, discretion and ambiguity…. The premises behind the numbers disappear" (Espeland & Stevens, 2008, pp. 421-22). That characterization obviously does apply here in part. Premises disappear from slides and reports as information is moved across state or institutional boundaries and up the hierarchy to the funding organizations. Within the state, however, the premises remain part of the collective knowledge of local experts who manage the translation of data into indicators. Some, at least, will be present at meetings where consultants present and interpret that data for political decision-makers. The local experts' deeper knowledge of local contexts positions them to challenge and raise questions about consultant's claims and to offer alternative interpretations (Kamruzzaman, 2017, p. 43;Koch, 2020, p. 2;Ocampo & Neu, 2008;pp. 87-88; on corporate contexts, see Alvesson, 2011Alvesson, , pp. 1654Alvesson, -1655. Consultants need to forestall such interventions. Powerpoint presentations are a device for doing this. Presentations enable consultants to present interpreted material to both decision makers and secondary audiences of local experts in ways that forestall unanticipated comments and interpretive challenges from the latter.

Powerpoint Presentations as Policy Device
Consultants obviously draw on many communicative resources. The most important of these are tools of modeling and calculation that can be used to render political issues as technical problems. But consultants also need technologies for making programs visible to decision-makers and for controlling the interpretation of the visible. Visualization media are ordering devices that play central roles in the "material and technological conditioning and structuring of experience, agency and interaction" (Beverungen et al., 2019, p. 624). Powerpoint is the most widely used of these, and defines the visual standard for presentations. Our work departs from other research on powerpoint presentations in several ways. First, much of that work focuses on corporate or governmental settings in which presenters and audiences share cultural and linguistic (and often educational) backgrounds. Our work focuses on powerpoint presentations in education policy encounters between people who speak different languages and have histories of harshly unequal relations (for example, as colonizers and colonized). Second, much of the research focuses on the semiotic potentials of slides (e.g., Kernbach et al., 2015;Tufte, 2006) or on strategies of slide construction used in particular presentations for organizational or public events (e.g., Bourgoin & Muneisa, 2016;Kaplan, 2011). Our emphasis, by contrast, is on the performative uses of powerpoint in periodic, repeated events embedded in relatively long-lasting (multi-year) consulting relations in which "meaning making unfolds over multilevel conversations and repeated interactions with visuals" (Knight et al., 2018, p. 897). Finally, unlike much of the research, we focus is not on the reactions of a primary audience of decision-makers, but on the effects of the presentations on secondary audiences of local experts who might, but do not, question consultants' claims or provide alternative interpretations of their own.
We begin conceptualizing powerpoint presentations by looking at them as policy instruments or devices. To start with a minimal definition, these consist of standardized, easily movable artifacts or objects coupled to sets of symbols or conventions for simplifying and representing a wide range of content domains; these artifact-symbol assemblages are embedded in widely shared but flexible performance conventions that allow participants from diverse organizational or socio-cultural backgrounds to jointly if unequally interact. Definitions become more specific by elaborating on these basic features. For Le Galès (2016) a "policy instrument" is a device that is both technical and social. It structures specific social relations between public authorities and those it is addressed to, according to the representations and meanings it conveys. Instrumentation refers to the set of problems posed by the choice and use of instruments (techniques, methods of operation, devices) that allow government policy to be made material and operational (p. 513) Powerpoint presentations fit only loosely into this definition. Presentations structure relations not only between public authorities and their others but among the varied groups, consultants, political leaders, and local experts, inside the authorities. Policy instruments in Le Galès' sense gain some of their efficacy from the fact that their material and operational forms stretch across time and space. But unlike formulaic written reports (Stirrat, 2000, p. 42) that, once distributed, are difficult to modify, the visual forms used in powerpoint slides are malleable and their performed interpretations have a short, dissipative half-life and can be over-written in subsequent presentations.
Another and perhaps better approach, borrowed from economic sociology, is to conceptualize powerpoint presentations as elements of larger ensembles of policy technologies-as policy devices "made up of human bodies but also of prostheses, tools, equipment, technical devices, algorithms, etc" (Callon 2005, p. 4). In economic sociology a device is a "material and discursive assemblage that intervenes in the construction of markets" (Muniesa et al., 2007, p. 2; also see Callon & Muniesa, 2005). We adapt this idea to conceptualize a policy device in transnational consulting as such an assemblage that intervenes in the policy circuits translating funding into project information and project information into more funding. Devices do this by detaching entities or events (e,g,, a child reading) from lived processes (e.g., by turning them into test scores), configuring these events or entities in a calculative space (e.g., a spreadsheet showing reading test scores), manipulating and transforming them into more condensed representations (tables or other infographics comparing scores across space and time), and finally, extracting or mobilizing them in forms that are "able to leave the calculative space and circulate elsewhere in an acceptable way (without taking with it the whole calculative apparatus)." for example, powerpoint slides (Callon & Muniesa, 2005, p. 1232. Consultant assemblages that include such devices can do different things than consultant assemblages from which such devices are absent. By the same token, the effects that can be produced through the presentation software's features and standardized formats depend on how they're interfolded with consultants, audiences, and institutions in particular performances. Software and slides "are not static backdrops to action; they partake in action" (Cochoy et al., 2016, p. 5).
The general ontological position here is that agency and life are woven through webs of humans and non-humans. The idea of devices is a way to conceptualize how such webs are organized to act across broad expanses of space and time. Funding agencies, for example, enroll national education systems in distributed webs of influence through loan schemes and networks of elites, consultants, and technicians. The focus on devices is less concerned with the architecture of such policy networks or the mechanisms (e.g., conferences) through which they are assembled than on the work involved in holding them together once they're formed. The relations represented by lines and nodes network diagrams are moving processes that have to be periodically channeled and contained through events like the "now-ubiquitous" monitoring and evaluation meetings (Peters, 2020, pp. 65-66) that consultants use both to keep political decision-makers informed about programs. Powerpoint presentations are devices for structuring participation and controlling interpretation in such meetings. They combine humans with software, slides, indicators, data, and settings in ways that reflect not only the affordances of the software but widely disseminated techniques for using the software to create visualizations and perform them in certain kinds of communicative events. The following sections examine two key elements of the assemblage: the uses of visual coding of information and the presenters' embodied performances of information.

Visual Coding
Software such as powerpoint have built-in, "preformatted features" (Kernbach et al., 2015) that reflect their designers' user models. Over time, however, the software can become tightly associated with particular contexts and uses. Standardized presentation formats emerge through a kind of institutional isomorphism as slide templates and examples of actual presentations become available in books or on the internet, and new users copy or appropriate the styles of established firms. Standardization is also spurred by the development of prescriptive literatures on slide design and presentation, and the growth of cadres of media specialists who coalesce around orthodoxies of best practice (Hertz et al., 2015). Such standardized formats (e.g., Galletta, n.d.) enable consultants to create presentations quickly and to re-use slides or slide structures in different settings. Some firms "maintain libraries of 'ready-to-use' powerpoint slides that have been proven to work, enabling them to provide 'instant insights" (Chong & Bourgoin 2020, p. 87). Others have developed their own presentation brands (e.g., "McKinsey-style powerpoints") marked by characteristic elements such as traffic light and heat map semaphores (e.g., PresentationPoint 2015). Finally, such templates and "overused default options" form elements of a generic presentation style, such that "most presentations created for diverse purposes in different industries around the globe look and feel exactly the same" (Kernbach et al., 2015, p. 306). This stylistic isomorphism in turn influences audience expectations. Presentations with standardized formats and slide construction "are granted more legitimacy than other texts" (Kaplan, 2011, p. 327).
As Tufte (2006) argues, the conventions of powerpoint presentations embody a "distinct cognitive style" particularly suited to the collapse of information as it moves up a hierarchy: The structure and defaults of the software promote simplification and the erasure of context and ambiguity. The process is similar to the production and mobilization of indicators in funded policy.
Consider the slide in Figure 1. Using Merry's (2016, p. 15) vocabulary, "English" and "Math" are names for "composite" indicators (representing complex phenomena of reading, writing, manipulating numbers). These are built out of count indicators keyed to a tightly bounded behavior event (e.g., the number of children matching pictures and words on a test). The effect is a simplification that buffers political decision-makers "from the very empirical phenomena over which [they are] reputed to have mastery" (Barley, 1996, p. 5). The presentation overwrites questions about what matching words to pictures (which words? what pictures?) means in relation to one's ability to speak, read, or write a language, what performance on double-digit addition in a one-off test says about numeracy, and why 80% is so much better than 75%. Instead of asking such questions to situate the program in local or national contexts, the powerpoint toolkit allows consultants to make the program visible in the abstract space of the projection screen and thus make political (and Tufte would say, moral) issues visible as technical problems.
Visibility, however, is not the same as legibility. Powerpoint allows consultants to make data visible while keeping its interpretation under their control. As Gunter and Mills (2017) suggest, presentations are moments of epistemic power relations: When an electronic slide is displayed in a workshop or presentational pitch, in which the consultant is engaging in knowledge production, there is an exercise of power taking place in how contractualism is 'lived' through the exchanges about what is presented as known, what it means, and what actions are deemed necessary, and how this renders other forms of knowledge unthinkable and unsayable. (p. 67)

Color-Coded Indicator Data
This may well be the aim, but "unsayable" does not imply "unthinkable," and the difference is consequential. It is because other interpretations are thinkable and can, in principle, be supplied by local experts, that devices of epistemic mediation such as powerpoint are needed to make "other forms of knowledge" difficult to articulate in presentation events. Elements of the powerpoint device assemblage, in particular certain ways of visually coding information, play a key role in this work.
By visual coding we mean the use of standardized icons, semaphores, and colors to present data. The ostensible aim is to make the meanings of the data obvious or easily legible to the viewer: a red flag means the data show the situation is bad, a green one signals that things are going well. Viewers are positioned as receivers of judgement -the consultant has already decided what the data mean --rather than analysts or interpreters in their own right. The ability to discriminate among icons and codes, however, isn't automatic. For example, one of the local experts we interviewed who worked as a liaison with consultants, Ahmed, suggested that semaphores like traffic lights and heat maps actually generated interpretive problems: Traffic lights concept is introduced by them [the consultants]. At first we were confused why there were traffic lights in education. Heat map was also introduced by them .. . . On heat map all the colors are so light and similar you cannot tell which district is complete red and which district is less red [less than complete red]. . . . These things are extravaganza. Each set of semaphores, in Ahmed's view, entailed a learning process that could be especially difficult for the elite political actors in the audience: The current CM [the Chief Minister who had started about 10 months previously] will need at least two years to understand these things, like what is a heat map. If the CM changes then the new CM will also need at least two years to understand these things.
This is probably exaggeration, but the broader literature on powerpoint provides illustrations of other instances in which such semaphores have created confusion rather than legibility (e.g., MacDougall, 2020). What is undeniable is that they make audiences dependent on presenters for understanding how such semaphores are actually connected to data. Slide presentations create interactional dependencies and concentrate interpretive authority in the presenting consultants. The ability to "see" data is stretched across devices linking multiple participants in ways that make some reliant on others for perception and understanding. Presentations are structures of opacity as well as visualization tools. As Robles-Anderson and Svensson (2016) put it, "slides bind knowledge to performance rather than to representation. They are not designed to provide audiences with evidence that speaks for itself" (p. 21). This use of powerpoint is arguably a form of governance by numbers, but it operates through an optic that allows presenters to reveal numbers while simultaneously overwriting them with semaphores and color-codes. If numbers are shown on slides, they are pushed into small print at the side where they become things-already-interpreted rather than materials to be inspected and analyzed. Consultants are open about these strategies, arguing that "'Chief Ministers or Prime Ministers are busy people. They don't want to look at scatter graphs and distributions and lots of dots on a page'" (Barber, quoted in Gold, 2016). Semaphores or colors, in other words, forestall analysis, questions, and interpretation in presentation contexts. Thus the prominent consultant and government advisor Sir Michael Barber (2013) explains how the color-codes used in slides for monthly powerpoint presentations about program performance-green for success, red for failing, amber in-between-were based on his judgement and could be changed at his discretion as a mechanism of control. Preliminary color codings were shown to local administrators before formal presentations to the Chief Minister, and as "no official wanted to risk a Red (of any shade) if it could be avoided," he could pressure administrators to take certain actions: "then, and only then, could the colour change" (p. 59). According to Ahmed, the interviewee quoted above, consultants could use such interpretive control to forestall alternatives explanations: "If the donor is against a project/policy, they just need to put red and Chief Minister question the policy, and the Secretary and the Minister do not want the red traffic light as their jobs depend on Chief Minister support."

Performance Events
The slides and visual codes constitute only part of the device assemblage. The performative work of the presenter also plays a key role in interpretative regulation. Powerpoint slides can be constructed as stand-alone texts, but when fashioned for presentations much of their meaning depends on the ways consultants structure and enact their performance. This depends on the performer/consultants' ability to control not just the visual coding, but also the scheduling of presentations, the pacing of performances, the language of the presentation, and the participant structures of the event -all of these become elements of the device assemblage.
The usual approaches to analyzing embodied policy performances (Rai et al., 2021) underplay the role of the kind of extra-somatic technologies (software, slides, presentation screens) that powerpoint brings to the forefront. In addition, they focus primarily on political performances in relatively open events such as parliamentary settings, conferences, and trade fairs, or events in the "media spotlight" (Rai, et al., 2021). We add to this literature by considering the role of extrasomatic technologies and by pointing to the importance of performance in political spaces where one can close the door, restrict the attendance, and orchestrate seating and participation formats.

Controlling Settings
The presentation setting matters in several ways. First, the lighting of rooms, the arrangement of furniture, and the seating of participants are ways of focusing attention and signaling status and participation rights (Gunter & Mills, 2017;Kernbach et al, 2015). Of less recognized importance are the ways that periodic visits by transnational consultants (one local actor referred to them as "Airport people") and meetings in closed-door rooms with restricted access generate a special kind of self-contained albeit short-lived policy space. Presentations are ways of generating a transitory space of interaction that encode expectations about appropriate forms of visual representation, communicative practice, and epistemic performance. Local experts may be socialized into the cultural practices of the consultants (Broome & Seabrooke, 2012); there may even be forms of "epistemic homophily" (Ball, 2016, pp. 553), but differences in communicative style, language, background, and political and emotional commitment remain. Powerpoint in transnational consulting doesn't just "detach things from other things and attach them to other things" (Muniesa et al., 2007, p. 4); it also detaches the presentation event from its geographical setting and the local cultural context. One can think of this as a kind of cultural extraterritoriality in which the presentation events of transnational educational consultancy are "positioned outside of the sovereignty and jurisdiction that surrounds them" (Weizman, 2005). Powerpoint presentations turn meeting rooms in countries around the world-if only for the time-span of the meeting-into standardized event spaces in which the forms of culture controlled by consultants are authoritative and other forms are marginalized or excluded. Such events are extraterritorial not just in the sense of being outside the political sovereignty of the state in which they are physically located, but also in the sense of being decoupled from local cultural and communicative practices: like a moving stageset, they provide an infrastructure through which participants with different cultural and communicative repertoires can act out the performance and reception of certain kinds of knowledge.

Controlling Pace
Within such settings performance dynamics push audiences towards reliance on the presenter for an understanding visual material. The potential for "relentless sequentiality" in presentations (Tufte, 2006)-a powerpoint deck might consist of a 100 slides-is less a bug than a feature of the performance system. It allows presenters to control the pacing of the presentation and to use speed as a means of display both their virtuosity of limiting the opportunities of the audience to question or intervene in the presentation. In one example from the literature, Powers (2016) describes a 4-hour powerpoint presentation on AIDS policy in South Africa: The audience squinted at slides that passed quickly with little explanation . . . It was hard to follow because of the rapid pace, and powerpoint slides were changed before their information could be digested. The room was filled with hushed whispers between people trying to explain to one another points raised. I scrambled to jot down notes fast enough to keep up with the presentation. Nomfusi, who sat to my left, shook her head and sighed, whispering, "She is speaking too fast." (p. 243).
Even when the audience is fluent in the language being used the slides show allows presenters to control both the rate at which the visual information moves on the screed and the speed of the verbal presentation. Unlike a written text or handout that can be re-read or studied, the slides are a moving one-way array. All of this makes participation more difficult, though obviously higher ranking ministers could in principal intervene.

Controlling Language
Speed of speech aside, presentations fold authority into the presenter's command of a "dominant" language, usually English, the "lingua franca of the aid system" (Roth, 2019, p. 41). This status is a reflection both of Britain's past colonial rule and of the current dominance of European and North American institutions in global governance. By performing presentations in English, consultants-intentionally or not-pull these histories and structures into the event. Language dominance is presupposed in the construction of the policy device (as air is presupposed in the construction of an internal combustion engine). It functions here as a mechanism of cultural closure, making it difficult for local experts to articulate disagreement or engage in argument for fear of embarrassing themselves in front of powerful officials. As Ali, a monitoring specialist with the agency explained, this centrality of English kept certain speakers quiet and certain issues from being raised: There are language barriers between donors and us.
[The] IO brings English language speaking peoples, and some people [who are responsible for the project] struggle to communicate with them especially in meetings because they fear that they may embarrass themselves in front of [their boss]. After meeting they would say that they disagree with English speaking people projects, but they cannot explain that to English speaking people.
Because the consultant can monopolize visual attention with slides and drive meetings forward through speech and slides, such participation constraints go unmentioned. The problem is not just linguistic facility, however. Powerpoint events rest on language ideologies that "locate linguistic phenomena as part of, and evidence for, what they believe to be systematic behavioral, aesthetic, affective, and moral contrasts among the social groups indexed" (Irvine & Gal, 2000, p. 37). In situations where English consultants work in former British colonial territories, the powerpoint presentation mixes in the colonial histories of language dominance and summons a cultural microfield in which the manners, accent, speech repertoires, and deportment of white European consultants with elite backgrounds give them dominant positions. English functions as "a hierarchyenforcing agent" (Waheed, 2020, p. 172). No conscious intent on the consultant's part is required. Husnain, a data and monitoring specialist, explained that "there's a culture . . . where someone from a foreign country, especially someone white says something, and everyone's like 'Oh, yeah, that's definitely correct. That's the way to go.' So, I guess we're still colonized in that sense." If there's "an imperialism of thought" in which "particular cultural models and categories [are] being imposed upon and often accepted in the developing world" (Stirrat, 2000, p. 41), one way this imposition is accomplished is by routing discussions through academic registers and elite accents that are markers of "high social status and professional competence" (Raza, 2015, p. 95). The cultural models "imposed" are not just substantive policy models of how schools systems should work. They are also models of how information is to be presented and how its meanings are to be made visible. As Ansorge and Barkawi (2014) suggest, utile forms such as powerpoint, or "media with a standardised layout that make knowledge available for particular purposes" (p. 5), have both "world-ordering" (framing) and "world-making" uses: "Even banal utile forms, concerned with the most mundane matters, can instantiate world views that legitimate and naturalise Western presence and policy in the non European world" (p. 12).

Participant Structures
The naturalization of Western presence and policy is stabilized through ritualized participant structures. "Participant structure" refers to the ways that such things as speakers' rights, discursive turn-taking, control of the floor, expectations for audience behavior, and the like are organized for particular speech events (Goodwin & Goodwin, 2005). Powerpoint presentations in a variety of contexts impose participant structures that severely limit audience roles and participation rights. They "become a pure ritual in which the audience expects a vivid show of slides while sitting passively and listening to what the speaker has to say" (Kernbach et al., 2015, p. 306). The concept of 'ritual' has multiple definitions in policy research (e.g., Aalberts et al., 2020). Here we use the idea of "political ritual" to refer to a high degree of formalization in how participants are expected to act: "there is 'artificiality' in certain types of behavior adopted by the principal protagonists, behavior expressive of respect, meditation, emotion, etc" (Abélès, 1988, p. 393). Powerpoint slides become ritual technologies when they serve as "the focalizing elements around which the various sequences of the rite are structured" (Smith, 1984, p. 5;quoted in Halloy, 2015, p. 361), and thus function as "mind traps" or "snares for thought" (Abélès, 1988;Halloy, 2015), or perhaps more accurately, as traps for preventing the public expression of unsanctioned thoughts. As Bloch (1975) suggests of a certain kind of political language, "the features of articulation have been . . . rendered arthritic and so the possible answers are dramatically reduced" (pp. 19-20). These ritualized events involve "displays of virtuosity" that position local participants as passive members of an audience that is "is not there to be convinced or educated"-let alone offer alternative interpretations-but to "watch a show" (Collins, 1988, pp. 728, 729;Stark & Paravel, 2008). This is a form of "aesthetic governmentality" (Ghertner, 2015) in which accountability and evaluation are enacted through performances keyed to aesthetic expectations about how presentation settings, slides, and presenters should look, speak, and act. In such virtuoso displays, "uncertainty is eliminated. . . . Displays of virtuosity are about lock in. The fix is in, outcomes are locked up, contending interpretations are locked out" (Stark & Paravel, 2008, p. 48). Ahmed uses similar language to describe presentations as "fixed": the consultant's "team prepares him [the presenter] thoroughly for presentation. . . . All the relevant justifications are presented to him. All the people sitting in that gathering [the Chief Minister meeting] fix the match beforehand." This does not mean that questions aren't asked, but that their main function is to demonstrate that the speaker knows the right way to ask questions: "Question and answer sessions aren't primarily about eliciting information. They are ritual means of determining whether or not the speaker is a member of the group" (Robles-Anderson & Svensson, 2016, p. 23). Remaining silent in such circumstances can be just as effective as speaking for signaling group affiliation.

Conclusion
Our focus has been on one aspect of transnational education policy consulting-the management of globally extended information circuits-and one mechanism of interpretive control, the powerpoint presentation. Funded projects require the production of certain forms of information at the local level. The information is then moved in two directions: back to the funding agencies, where it provides accountability for money lent, and back into local implementation arenas, where it can be used to guide future program efforts. Consultants play critical roles regulating information flows in both directions. Our focus here has been on the flow in one direction: back to the local context. Drawing on published descriptions of powerpoint presentations, internet resources, and interviews with local experts in one setting, we have argued that powerpoint presentations in such contexts are devices that enable consultants to control the flow and interpretation of program information. We also argued that tensions can arise around the ways local experts are positioned in these information circuits. Consulting relations differentiate actors in the nodes of policy networks, producing at least two factions: a primary audience of political decisionmakers, usually with some sort of "chief minister" at the top, and secondary audiences of local experts. Consultants need the decision-makers to accept their interpretations, and among other things this means they need to forestall or mute potential disagreements or alternative interpretations from local experts. Power operates not just through ways of "seeing" the world (Scott, 1998) but through ways of showing it and making others watch quietly.
Our aim was to make constitutive arguments about the nature of the phenomenon: how should we think of powerpoint presentations in these contexts? Our claims are that presentations are devices that enable consultants to establish closed spaces in which they can control the interpretation of information by rendering project data in visually coded forms-semaphores, color codes, flags and heat maps-and by drawing on language ideologies and participant structures that ritualize discourse within the presentation context. These coded materials focus the participants' attention and make consultants' interpretations easily visible (e.g., red signals bad problems) while simultaneously rendering the reasoning behind the interpretations opaque. This epistemic control is buttressed by the consultant' ability to control the pace at which information becomes visible, the language in which it is revealed and discussed, and the participant structures of the presentations. Rather than a form of outsourcing knowledge work, powerpoint is a device through which consultants generate visual and performative ritual in a way that "crowds out local knowledge" (Broome & Seabrooke, 2012, p. 6).
In a more theoretical register, we argued that presentation devices were heterogeneous assemblages of people, tools, symbols, and actions that can intervene in the circuits translating funding into project information and project information into more funding. They detach information from local settings, move it around, and reassemble it for different audiences. Presentations can thus help stabilize relations inside the nodes of policy networks and reinforce the linking ties that bind them to other nodes.
As a conceptual argument there are clearly limitations to our work and a need for more empirical research. First, there is a need to trace the relations of powerpoint slides to other texts generated through consultancies, such as the deliverables that flow back to funding organizations. Powerpoint slides are peculiar devices in that, as noted, they can destabilize the meanings of numbers by coding the same numbers with differently colored semaphores. Once published in an official document, however, data related to indicators and deliverables are difficult to alter and remain relatively stable as they move. How and through what processes does the information shown on slides undergo this transformation? What are the relations between these types of texts? Second, as noted earlier our interview data are drawn from a setting in which consultants had long-standing relations with the local ministry. It's possible that presentations change form and function across the life-span of a funded project. Thus we need to track how consultants' repertoires of communication and information management shift across the course of long-term projects. Third and related, we might expect presentations to change in relation to the beginnings and endings of funding periods. Fourth, although we argue and the literature shows that there is substantial standardization of powerpoint slides and presentation forms, it possible that different consulting groups, staffed by consultants from different cultural, linguistic, and educational backgrounds, produce different performance styles using these standardized materials. The forms these differences take and their consequences for interpretive control are matters for further exploration. Finally, we might ask how the epistemic relations generated through the types of presentations we describe influence the capacity and institutional memory of local or national education agencies in the states receiving funds. Do they harden internal divisions within national ministries (Peters, 2020, p. 66), inspire work-arounds, or leave local experts isolated and cynical?