How are Realities Enacted by Data? Analysis of the Practices and Knowing of Professionals Working with Data in light of the ActorNetwork Theory.
| Autor | Pra, Raquel |
1 Introduction
The data generated from the use of digital technologies such as computers, smartphones, and social networks are thought to be the new oil, since they generate significant profits for the companies that extract, store, and analyze them (Zuboff, 2021). Making decisions based on data such as big data, instead of intuition, is thought to be a capacity that distinguishes the most successful companies. This finding is based on the belief in the neutrality, objectivity, and veracity of the data (Boyd & Crawford, 2014; Esposti, 2014; Iliadis & Russo, 2016).
The organization and hierarchization of data and the relationship between those data are established through architectures present in computer systems, known as databases. These have become a source of power and knowledge, as they are used to elaborate predictive models to represent populations (Ansorge, 2011) and prevent and act on human behavior (Zuboff, 2021). These models are elaborated based on algorithms, which are mathematical formulas that perform computational tasks, such as cross-checking data. Because of that, the knowledge that emerges from these databases is called algorithmic (Fuchs & Chandler, 2019).
The interrelationships between data, algorithms, and organizations are epistemological, as they produce knowledge for decision making, as well as supporting the supply and improvement of products and services. But they are also ontological, as they produce models that are thought to be representative and capable of enacting certain realities (Glaser et al., 2021; O'Neil, 2016). The predictive models generated in this association are used to determine whether someone will be granted a loan to buy a house or not and what rate they will pay for it; to define if someone will advance or not in the stages of a selection process; and whether they will be fired or not (O'Neil, 2016).
Considering this context of increasing use of data to create mathematical models capable of predicting future behaviors and helping in business decision making, this theoretical-empirical article aims to understand how realities are enacted through data. For that, we employ Actor-Network Theory (ANT) as a theoretical-methodological approach to identify and describe the practices and knowing of two groups of professionals that work with data.
Within the field of organizational studies, research tends to emphasize the role of guidance by data and digital technologies in process optimization, without taking into account the situational arrangements that can result in unexpected effects (Glaser et al., 2021; Trittin-Ulbrich et al., 2020). Digital technologies such as algorithms are not immutable, independent, and neutral, as usually understood in studies, since they emerge and cause effects on specific configurations (Glaser et al., 2021).
According to Cooren (2020), there are no independent entities with inherent characteristics and the organizational studies fail by continuing to treat the material world as something tangible and distinct from the world of affects. After all, every organizational phenomenon is material and relational. The assumptions and objectives that form the basis for the selection, analysis, and understanding of data, as well as the narratives that are formed based on them, are not neutral and objective as positivist studies argue, since they derive from political choices reflected in routines and ways of organizing (Dourish & Cruz, 2018; Labatut et al., 2012; Ratner & Gad, 2019; Vesa & Tienari, 2020).
Considering this contextualization, we highlight as a research contribution to the area of organizational studies our shedding light on the configurations of human and non-human elements, situational and emerging, through which organizations and realities are enacted, based on data practices. We intend, through this exposure, to problematize naturalized understandings in the area of Administration regarding the objectivity, neutrality, and veracity of the world models created based on data, by demonstrating that the knowledge generated emerges in specific configurations. Considering those compositions, these models can lead to process optimization, but also to unexpected effects, through errors occurring in the projection of scenarios.
Initially, we present the Actor-Network Theory and its contributions to the understanding of the configurations of human and non-human elements that enact organizations and realities. In the methodological framework we describe how we operationalize ANT as a method throughout the research, to identify the practices and knowing of the data professionals. Then, in light of ANT, we describe and analyze the interrelationships between practices, knowing, and the enactment of realities. Finally, we present the concluding remarks of the study.
2 Actor-Network theory
The Actor-Network Theory, known in the beginning as the Sociology of Translation, forms the umbrella approach known as Practice Theory or Practice-Based Research (PBR) (Gherardi, 2012). Despite the diversity, some understandings about the nature of practices are common to the PBR approach: (1) holistic and qualitative --practices are formed by a set of activities that acquire sense and make them a unit, a way of enacting throughout the whole organization; (2) temporary--they persist in time, by being repeated several times, becoming a normal way of doing things. However, this reproduction is not mechanical and varies depending on the elements and conditions that affect its re-edition; (3) legitimacy--they are socially recognized through their negotiated normative aspects, involving ethical, esthetic, and technological aspects; and (4) they are ways of ordering the world and organizational environments in a temporary and precarious way, activating a network involving human and non-human elements.
ANT is a theory and method of French origin, originating from science and technology studies in the social context (Mol, 2010). Its first wave was centered on concepts such as actor, network, translation, and technoscientific practice. The second one was based on the notion of process through the concept of enactment (Camillis et al., 2020) and broadened the field of studies beyond technoscience. More than a theory or method, ANT is understood as a practice: a way of doing and engaging in the world (Farias et al., 2020).
The reality for ANT is not exterior, singular, or definitive, but the result of ever-emerging associations between human and non-human elements, such as language (semiotics), organizations, objects, and animals, which mark its relationist ontology (Law, 1999). Knowledge is similarly perceived as a result of the "successful alignment of human and non-human elements ('heterogeneous engineering') and the human capacity to produce effects over the world" (Nicolini et al., 2003, p. 19).
By shedding light on the action of non-human elements, ANT seeks to retrieve the complexity and heterogeneity that constitute reality, relegated up to a certain point in sociological studies, portraying the tensions that exist between agency and structure, actor and network (Latour, 2012; Law, 1999). Respecting the ontological differences between humans and non-humans, both are symmetrically considered as capable of agency, that is, of generating effects on a network of relationships (Camillis et al., 2016).
For a non-human to receive that nomenclature, this element cannot be passive, as it should be capable of making the difference in the network (Camillis et al., 2016). Non-human elements can act as mere intermediaries, when they merely transport without transformation, or as mediators, when they present an active role, modifying and distorting contents, producing and reproducing the social in its multiple forms (Latour, 2012).
The theoretical perspective adopted in this research fits the second wave of ANT, post-ANT or ANT and After, as denominated by Law (1999). It is a reformulation derived from criticisms directed at the concept of translation, by naturalizing, simplifying, and stabilizing the ways of ordering the world, suggesting that it occurs in a particular, prescriptive, and nonproblematic way (Alcadipani & Tureta, 2009). The term enactment comes to be used, indicating that reality is in constant transformation and continuity, where stability is an exception.
2.1 ANT, knowing, enactment of realities
The term enactment reinforces that ordering is a constant process derived from practices and relationships (Law, 1999). For ANT, all things are understood as enactments, that is, effects that are continually produced in relationship networks (Fenwick & Edwards, 2010). In other words, ANT does not deny stability, but demonstrates that it is not fixed and coexists with the chaos of reality.
Mol (1999, 2002) clarifies that, as reality is enacted, it is multiple, since it is not possible to characterize it as unique or definitive, and it is constituted based on the variety of historically, culturally, and materially located practices. Another term proposed by the author that reinforces the question of naturalizations is political ontology, used to demonstrate that the condition of possibilities is never given, as it is a permanent production process. Therefore, facts and possibilities are produced and negotiated, emerging in certain contexts. That is, reality does not precede but rather is shaped by practices, and it can always be of another form.
By studying knowing through ANT, this is understood as an enactment, an effect of the alignment of heterogeneous elements in a network and not simply an individual and cognitive process or a social realization (Fenwick & Edwards, 2010). Just like organizations, knowing is relational, dynamic, and provisional, enacted in daily life and in the realization of practices, based on networks of heterogeneous relationships and the subjects' experiences (Bussular & Antonello, 2018).
This vision comes close to the concept of...
Para continuar a ler
PEÇA SUA AVALIAÇÃOCOPYRIGHT GALE, Cengage Learning. All rights reserved.
