Researching Education Through Actor-Network Theory

Tara Fenwick, Richard Edwards

• Publisher: Wiley-Blackwell
• Year: 2012
• ISBN:
  9781118275863
• Edition: 1
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

Researching Education Through Actor-Network Theory offers a new take on educational research, demonstrating the ways in which actor-network theory can expand the understanding of educational change.

• An international collaboration exploring diverse manifestations of educational change
• Illustrates the impact of actor-network theory on educational research
• Positions education as a key area where actor-network theory can add value, as it has been shown to do in other social sciences
• A valuable resource for anyone interested in the sociology and philosophy of education

1

Devices and Educational Change

Jan Nespor

Things are thick with power relations and politics. (Bijker, 2007 p. 115)

This paper examines two cases of device-mediated educational change. One involves a computer-assisted interactive video module that provided a half-hour of instruction for a university course, the other an assistive communication device that proved a supposedly retarded pre-school child to be intelligent. Both were created in the mid-1980s, in sites roughly 30 miles apart.

The interactive video was viewed as a success and won support for its makers; the assistive communication device was rejected and its maker cautioned against repeating such work. Two decades later there are no records or organizational memories of the devices (save those of their makers), but both, I’ll argue, were key events in processes of significant organizational transformation.

I have two aims in examining these processes. One is to shed light on the roles of devices in organizational transformations initiated by middle-level workers such as technicians and teachers. I’ll argue that device mediated changes, rather than the predefined outcomes of planned efforts, or the products of activity systems organized around explicit objectives, are effects of non-linear processes arising out of improvisations that ‘continuously generat[e] new results’ (Abbott, 2005a, p. 402) across the worker’s career. Devices are key to these improvisations: They shape change by slowing things down (orienting work around devices that don’t exist yet and require indefinite development processes), or speeding things up (creating devices that seem to do in a short span what otherwise requires long, complex interaction); they re-shape relations among organizations by enrolling allies, or weaken organizational boundaries by making them vulnerable to formerly excluded claimants. Finally, devices can be used to reorganize agency itself in core organizational activities—shifting the location or attribution of who does what, shifting participants from one actor category to another, or creating new categories of agents. Each of these uses appears in the cases described below.

The second aim of the paper is to develop theoretical tools for analyzing such change processes. Although I draw ideas from several fields, the basic perspective taken derives from actor network theory (ANT), ‘a disparate family of material-semiotic tools, sensibilities, and methods of analysis that treats everything in the social and natural worlds as a continuously generated effect of the webs of relations within which they are located’ (Law, 2009, p. 141). The next section outlines how I want to engage this perspective.

Devices and Distribution in Actor Network Theory

ANT tells us that we are what we are by virtue of our associations—the ways ‘our’ identities, thoughts, and actions are produced and spread through people, things, situations, and structures (Law, 1994, pp. 100–101; cf. Lave, 1988; Hutchins, 1995). The idea is not that there are no differences between people and things, but that they are not and cannot be separated. We can move through different settings, use different artifacts and tools, and interact with other people in myriad ways, but we can’t get outside such relations. Detachment and de-contextualization can only be accomplished through re-attachment and re-contextualization. As Munro (1996) puts it, you can only go from one network configuration to another—‘one is never traveling out from a place (the core self) and then returning … the only movement is one of circulation: around and around from figure to figure’ (pp. 263–264).

As the term ‘actor network’ implies, people and networks are thus ‘co-extensive’ (Callon & Law, 1997, p. 169). Even our thoughts are network effects:

Parts of our selves extend beyond the skin in every imaginable way … Our memories are in families and libraries as well as inside our skins; our perceptions are extended and fragmented by technologies of every sort. … When we use the shorthand ‘individual’ or ‘individual cognition’, we are thus only pointing to a density. (Star, 1995, pp. 19–20; Latour, 2005, p. 211)

Devices are necessarily central to any account in which ‘the social’ is thus ‘materially heterogeneous’ (Callon & Law, 1997, p. 167). What this means is that we have to treat interaction as involving not just physically co-present humans but artifacts and environments which congeal past actions—‘new hybrid social-and-material practices are constrained and enabled by equally hybrid preexisting practices’ (Law & Singleton, 2000, p. 766)—and mediate the ongoing transactions of people widely separated in time and space. From this perspective, agency, the ‘capacity to act and to give meaning to action’ (Callon, 2005, p. 4), is not a monopoly of bounded human individuals but instead a ‘relational effect’ (Law, 1994, p. 100; Callon & Law, 1995, p. 502), possible only by virtue of the fact that people ‘hook up’ (Latour, 1999, p. 18) with institutions, buildings, landscapes, discourses, artifacts, microbes, and the rest, all of which are also networks or ‘assemblages’ (Cooper, 1998). ‘Action, including its reflexive dimension that produces meaning, takes place in hybrid collectives comprising human beings as well as material and technical devices, texts, etc.’ (Callon, 2005, p. 4).

The emphasis in ANT is less on the structure of such hybrids than the movement of their constitutive associations across times and spaces. Networks are treated not as stable structures in static landscapes but as contingent effects of ‘translations’—the term ANT practitioners give to the ‘displacement, drift, invention, mediation, the creation of a link that did not exist before and that to some degree modifies two elements or agents’ (Latour, 1994, p. 32). In some cases translations become relatively stabilized and generate organizational structures, infrastructural categories, and well-defined paths for getting things done. But stability can be difficult to achieve and sustain, and often doesn’t take a form foreseen by those who accomplish it.

ANT is less a theory of such processes than a set of assumptions and conceptual tools for studying them. Individually these assumptions are not unique to ANT (e.g. Ingold, 2000, pp. 304–5; Lang, 1993; Hutchins, 1995; Fuchs, 2001), but ANT’s ways of formulating and combining them have an entailing incompleteness that pushes us towards empirical engagement. As Law (2009) explains, describing ANT in the abstract ‘misses the point because it is not abstract but is grounded in empirical case studies. We can only understand the approach if we have a sense of those case studies and how these work in practice’ (p. 141).

At the same time, these case studies should have a heavy theoretical recoil. Each involves a translation of ANT itself, pushing the uses of its tools and methods, challenging its sensibilities. The two cases examined here, for example, raise at least the following questions.

First, how are associations among people and things accomplished? There are different ways to ‘circulate’, ‘delegate’, ‘sum’, and ‘shift down’, and Latour (1996b) acknowledges that ANT ‘is an extremely bad tool for differentiating associations. It gives a black and white picture not a colored and contrasted one’ (p. 380). A speed bump, to take one of his examples, may be ‘full of engineers and chancellors and lawmakers, commingling their wills and their story lines with those of gravel, concrete, paint, and standard calculation’ (Latour, 1994, p. 41), but the question is how and why ‘commingling’ happens—for example, how important to a given outcome is the sequencing of assembly, the pacing of composition, the specific mix of the elements associated, whether a given element is essential to the mix or open to substitution, and whether the associations are reversible or easily changed. Do associations and delegations come slowly and incrementally, allowing different kinds of uses at different stages as a device takes form (or as different versions of a device are produced), or do commitments come together all at once (the organization bets on a particular product)? Are commitments large at the outset or do they gradually build? How does one translation relate to a preceding sequence of translations (e.g., Latour, 1996a, p. 91; Law & Callon, 1992, p. 52)?

Second, if body, agency, and mind are distributed networks, how are they made to look like discrete, bounded entities? It may be true that we commonly ‘localize agency as singularity—usually singularity in the form of human bodies’, through ‘attributions which efface the other entities and relations in the collectif, or consign these to a supporting and infrastructural role’ (Callon & Law, 1995, pp. 502–503). But ANT does not allow us to naturalize such attributions or take them for granted. We know that people do not always individuate agency in this fashion (e.g. Strauss, 2007; Comaroff & Comaroff, 2001), and the persistence of institutions for producing individual subjects (e.g., Foucault, 1979) and promoting public narratives of ‘singularity’ (Somers & Block, 2005) suggest that such attributions are unstable and require continuous effort. It is as plausible for people to see cognition and agency as distributed or ‘stretched out’ across people and things as to think it intra-individual.

Third, terms like ‘artifact’ and ‘non-human’ (Barron, 2003) are inclusive by design, but as such deflect attention from questions of who makes (or can make) a certain kind of device, who controls use of the device, how access to it is organized, who supplies the power for it, what kinds of products it makes, how it moves, and how it is made visible to different observers (cf. Kirsch & Mitchell, 2004).

Fourth, ANT’s analyses of technological change often bracket focal devices (as I do here)—an aircraft, a self-coupling train car—and treat the networks and translations out of which they percolate as being about those devices (I try not to). The obvious point is that such processes are often about other things besides or instead of the focal object. The speed bump may not be what the engineers and chancellors were making, but something that was made because or in spite of other things they did. This may not be relevant to an ANT account of ‘speed bumps’, but it may be what matters to the people involved. If we take ‘identity’ to mean ‘an actor’s experience of a category, tie, role, network, or group, coupled with a public representation of that experience … not private and individual but public and relational’ (Tilly, 2002, p. 75), organizations are not only about ‘work’, ‘activity systems’—or devices—but also about people and groups making identities through articulations with devices.

Finally, the individuals ANT takes as points of entry in tracing networks are usually high-status participants—officials, administrators, ‘engineers, technicians, and technocrats’—situated in ‘their own separate world’ (Latour, 1996a, p. vii). ANT foregrounds activity defined from these standpoints rather than, for example, those of mid-level workers doing things administrators might not care about or approve of. The two cases that follow, by contrast, focus on how middle-level workers in organizations thread connections among devices and different fields of action, and how the devices simultaneously reshape the workers’ organizational worlds. I argue that in these processes the organizations themselves change—though not necessarily as the narratives attached to the devices imply. As Ferguson (1990) notes in another context, ‘it may be that what is most important about a “development” project is not so much what it fails to do but what it does do; it may be that its real importance in the end lies in the “side effects” ’ (p. 254; but contrast Hirschman, 1991, pp. 39–42).

Method

The first case examined below, in which a university-based technologist creates an interactive video teaching device, is based on data drawn from a 1997–2004 study of the introduction of computer-mediated instruction at an American research university (Nespor, 2006). That work involved interviews with administrators and professors (the main sources of data used here), interviews with students, documentary analysis, and classroom observation. The discussion of the preschool teacher’s creation of an augmentative communication device is drawn from an extensive ethnographic study of special education practices across time (1989–1991, 2005–2008). In particular I use interview materials (from 1989 and 2005) and documentary materials tracing the work of a teacher, B. The paper also draws on B’s own dissertation account of this work, written in 1992.¹

Foregrounding the work of two people has drawbacks. As Pierson (2004, p. 141) notes, using individuals as entry points for analyzing organizational change favors a focus on ‘particular kinds of actors—entrepreneurs, “skilled social actors”, and “losers” ’. Czarniawska (2009) adds that such research easily falls into the fallacy of ‘post hoc, ergo propter hoc: when an institution has been established, people who were involved in establishing it are seen as decisive for its establishment’ (pp. 438–439). This paper may be guilty on both counts, but I would argue that these two protagonists deserve the attention not only as the earliest makers and users in their organizations of the kinds of educational technologies described, but because these early efforts ramified over the decades into major changes in the organizations, and because both had influence well beyond their organizations on the how such devices came to be used by others. Computer-mediated instruction would have come to the university, and assistive communication devices to the public schools, regardless of whether these two particular people had been involved, but without focusing on individuals we have few ways of understanding or even observing change processes structured across biographical and career-length time frames.

Little ‘Demos’: Technology and Organizational Identity

Teaching in formal educational settings assumes a web of relations linking teachers, students, schools, and content disciplines. Changing teaching involves changing the translations that generate this web. One way to do this is to work on single elements—for example, train better teachers or create better curricular materials. The approach taken in the case below, by contrast, was to use devices to reconfigure the web of relations connecting university teaching to external networks such as professional organizations and technology corporations.

J, the protagonist of this section, played a key role in this effort. He had arrived at the university in 1971, with the job title of ‘producer-director’, ‘to ...

Table of contents

• Cover
• Educational Philosophy and Theory Special Issue Book Series
• Title page
• Copyright page
• Notes on Contributors
• Foreword
• Introduction
• 1 Devices and Educational Change
• 2 Translating the Prescribed into the Enacted Curriculum in College and School
• 3 Unruly Practices: What a sociology of translations can offer to educational policy analysis
• 4 ANT on the PISA Trail: Following the statistical pursuit of certainty
• 5 Assembling the ‘Accomplished’ Teacher: The performativity and politics of professional teaching standards
• 6 Reading Educational Reform with Actor-Network Theory: Fluid spaces, otherings, and ambivalences
• Index