Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/sage/the-limits-of-computation-a-philosophical-critique-of-contemporary-big-Gs0kFSEcra
- Publisher
- SAGE
- Copyright
- Copyright © 2022 by SAGE Publications Ltd, unless otherwise noted. Manuscript content on this site is licensed under Creative Commons Licenses.
- ISSN
- 2053-9517
- eISSN
- 2053-9517
- DOI
- 10.1177/2053951718811843
- Publisher site
- See Article on Publisher Site
Abstract
This paper reviews the contemporary discussion on the epistemological and ontological effects of Big Data within social science, observing an increased focus on relationality and complexity, and a tendency to naturalize social phenomena. The epistemic limits of this emerging computational paradigm are outlined through a comparison with the discussions in the early days of digitalization, when digital technology was primarily seen through the lens of dematerialization, and as part of the larger processes of ‘‘postmodernity’’. Since then, the online landscape has become increasingly centralized, and the ‘‘liquidity’’ of dematerialized technology has come to empower online platforms in shaping the conditions for human behavior. This contrast between the contemporary epistemological currents and the previous philosophical discussions brings to the fore contradictions within the study of digital social life: While qualitative change has become increasingly dominant, the focus has gone towards quantitative methods; while the platforms have become empowered to shape social behavior, the focus has gone from social context to naturalizing social patterns; while meaning is increasingly contested and fragmented, the role of hermeneutics has diminished; while platforms have become power hubs pursuing their interests through sophisticated data manipulation, the data they provide is increasingly trusted to hold the keys to understanding social life. These contradictions, we argue, are partially the result of a lack of philosophical discussion on the nature of social reality in the digital era; only from a firm metatheoretical perspective can we avoid forgetting the reality of the system under study as we are affected by the powerful social life of Big Data. Keywords Big Data, data analytics, end of theory, computational social science, data-driven science, epistemology Introduction scientific discovery. The more data there is the more The term ‘‘Big Data’’ is used to describe the volume of discoveries can be made’’ (Rosling, 2010). Some have information produced through the use of technologies pointed to a ‘‘fourth paradigm’’ for science, as new like mobile devices, positioning systems, and online ser- algorithmic, computational, and analytical tools pro- vices—‘‘[i]n a digitized world, consumers going about duce ‘‘gold’’ from this data resource (Bell et al., 2009; their day—communicating, browsing, buying, sharing, Hey et al., 2009). searching create their own enormous trails of data’’ (Manyika et al., 2011: 1). The increasing use of digital services has given social scientists unprecedented access Department of Sociology, University of Amsterdam, Netherlands Department of Sociology and Work Science, University of Gothenburg, to previously unimaginable data; traces of the lives, Sweden dreams, and feelings of hundreds of millions of people. This seems to bring great promises for social Corresponding author: scientific work, as the ‘‘data deluge ... is leading us to Petter To¨rnberg, Department of Sociology, University of Amsterdam, an ever greater understanding of life on Earth and the Postbus 15508, 1001 NA, Amsterdam, Netherlands. Universe beyond .. . [it may] transform the process of Email: k.p.tornberg@uva.nl Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http:// www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Big Data & Society But Big Data is also associated to a number of sci- Computational Social Science, towards a new compu- entific challenges, that have led to the competing view tationally based perspective by underlining the limita- that ‘‘the glittering promise of online data abundance tions of traditional quantitative methods. As we will too often proves to be fool’s gold’’ (Karpf, 2012: 652). argue, their response implies a particular social ontol- Big Data has brought a development that is simultan- ogy, which focuses on relations and sees social struc- eously enticing and vexing: a veritable ‘‘siren-song of tures as patterns emerging from underlying local abundant data’’ (Karpf, 2012) that is causing research- interaction. ers to flock to the study of phenomena identifiable in This paper looks at what this approach leaves out. growing Big Data, and ignore phenomena not so By weaving together a series of theoretical views, the inscribed. They are met by a host of difficulties, some paper outlines the epistemic limits of the emerging com- of which are a natural part of new technologies to putational paradigm. We bring back into the contem- which we have yet to grow accustomed, but there are porary discussion the views that dominated the social also issues that run deeper than the mere settling of dust sciences in the early days of the digital era—a time (e.g. Boyd and Crawford, 2012: 20). Big Data is not when digitalization was seen not as something that only different in quantity, but also in quality, and it would make society more amenable to formal methods, seems that the new shapes of data do not always fit but rather as the precise opposite: as part and parcel of into the holes of old theory. This has resulted in it the processes of postmodernity. It was seen as part of a bringing fundamental issues to the fore: the questioning development towards increasing openness to society as of epistemological assumptions, discussions of the val- a system, thereby limiting the usefulness of quantitative idity of disciplinary divides, critique of methodological approaches. We argue that these theoretical points still monism, and the rejection of long-trusted simplifica- hold and should be taken into account within the cur- tions (Kitchin, 2014). Throughout the sciences, similar rent relational perspective, but that they also need to be tensions can be seen as the data deluge causes long brought up to date with more recent developments of submerged epistemological questions to float to the digital technology, focusing on power and the role of surface. digital platforms. While the traditional variable-based approaches to We begin by looking at the impact that Big Data has social science have struggled with the new forms of had on contemporary social science. data, approaches with their roots in the natural sci- ences have stepped forth to meet the tide. A central Contemporary digital research: player in the multitude of new approaches is Computation and relations Computational Social Science, located at the ‘‘inter- section of the social and computational sciences, an Despite its name, size is arguably not the most defin- intersection that includes analysis of web-scale obser- ing feature of ‘‘Big Data’’ (e.g. Boyd and Crawford, vational data, virtual lab-style experiments, and com- 2012): the concept rather describes a set of parallel putational modeling’’ Watts, 2013:1. This natural developments in various disciplines, whose common scientific approach has brought with it a range of denominator is an increasing proliferation of data methods to allow dealing with the complexity of sets that have proven difficult to fit into existing para- mass-interaction (Conte et al., 2013). The renegoti- digms. In the computer industry—first to feel the ation of demarcations between the natural and social effects of this development—quantity was indeed sciences following from this development seems to in among the primary issues, since traditional tools, part be leading to a renewed naturalism, sometimes such as relational databases, proved incapable to referred to as the ‘‘end of theory’’ (Anderson, 2008) deal with new demands emerging from large-scale sys- and exemplified by Lev Manovich’s (2016) view that tems (Manovich, 2011). But in the social sciences, the ‘‘Digital is what gave culture the scale of physics, emerging problems associated to Big Data were differ- chemistry or neuroscience. Now we have enough ent: as Boyd and Crawford (2012) observe, some of data and fast enough computers to actually study the data sets understood as examples of ‘‘Big Data’’ the ‘physics’ of culture’’. In this new naturalism, soci- (e.g. some Twitter studies) are significantly smaller ety is subsumed not under the traditional Cartesian- than sets understood as ‘‘traditional’’ data (e.g. Newtonian paradigm, but instead under the metathe- census data), pointing to the fact that the data quan- ory of particles and flows, and analogues such as tities in themselves are not the issue—even huge quan- ‘‘avalanches and granular flows, flocks of birds and tities of structured census data are relatively easy to fish, networks of interaction in neurology, cell biology process using traditional tools. and technology’’ (Ball, 2012: ix). Instead, the use of the term ‘‘Big Data’’ seems to This situation paints a picture in which Big Data has point toward qualities of the data, associated to a taken parts of social science, in particular fields like deep failure of traditional approaches. In other words, To¨rnberg and To¨rnberg 3 the impact of Big Data is not seen as merely methodo- argues, has the potential to transform our understand- logical—rather, in the words of Boyd and Crawford ing of our lives, organizations, and societies. Other (2012), they are associated to ‘‘a profound change at scholars have argued that ‘‘a new kind of social sci- the levels of epistemology’’ p.665. ence’’ is needed (e.g. Christakis, 2012) (a call referring When it comes to the methods that structure the Wolfram’s (2002) declaration of ‘‘a new kind of sci- data, the transition is perhaps best headlined as a ence’’, i.e. Complexity Science), to respond to the fun- shift from mathematically organized data to algorith- damental changes that Big Data brings in its wake. This mically organized data (see also Tornberg, forthcom- new science is seen as the answer to the crisis of the old ing). While survey data is constructed for processing approach of empirical sociology, through a supplanting through variable-based analysis, requiring pre-com- of surveys and interviews with data mining and GIS partmentalized data designed to be palatable for a sci- analysis (Savage and Burrows, 2007). This is seen as entific perspective that sees the social world through a bringing a redrawing of the disciplinary boundaries lens of averages and variances, data extracted from by, as Watts (2007) argues, resolving the issues that digital technologies tends to be structured by and for has made the social sciences ‘‘less successful’’ than the algorithmic processing, implying indexed data struc- physical and biological sciences in providing explana- tures and traversable networks (Mackenzie, 2012; tory and coherent theoretical accounts of, for example, Marres and Weltevrede, 2013). New data therefore the complexities of collective social behavior (see e.g. tends to be poorly suited for statistical analysis; it Bajec and Heppner, 2009; Dorigo and Stu¨ tzle, 2010; often comes in small chunks, spreading and diffusing Helbing et al., 2005; Johnson, 2002). Thus, the new in complex and constantly transforming networks, data is seen as enabling a convergence between the without clearly defined bounds. The social ontology social and natural sciences under a new approach and that digital technologies operationalize is not focused ontology (Christakis, 2012). The difference in social on the summing up of a population in fixed categories, ontologies that are operationalized by the old and but rather on the individuals and their dynamic connec- new data in many ways corresponds to the contrast tions and interactions (Castellani, 2014; Uprichard, between ‘‘complicated’’ and ‘‘complex’’ systems 2013). This implies no longer producing data by depart- (Morin, 2007). This similarity is not incidental: the ing from the aim of a whole, implicitly assumed to be Santa Fe school of Complexity Science developed lar- gely around the study of, and with, computers and the sum of its parts, but rather departing from the parts and their location within a data structure. algorithms, in which the dynamics of computer This has been taken to suggest that while census data models of mass-interactive systems were studied under is produced for scientific analysis, Big Data is a ‘‘nat- labels such as ALife, Agent-Based Modeling and urally occurring by-product’’ (Edwards et al., 2013; Cellular Automata (Galison, 1997). It was largely this Kitchin, 2014), constituted by traces of ongoing social study that became the foundation for the theory of processes rather than something produced for scientific complexity, describing both an ontological category consumption. This ostensible rawness is taken to mean and an approach (e.g. Mitchell, 2009). It is for this that the ontology revealed by Big Data is in fact the reason to little surprise that Big Data seems similar to true, relational nature of the social world that had pre- this ontological category and responds well to a similar viously been concealed by the survey method. As Allen methodology, as they carry the same basic social ontol- Barton (1968: 1) puts it, ‘‘the survey is a sociological ogy within their structure, having been shaped by the meat grinder, tearing the individual from his social con- same type of methods and technologies. text and guaranteeing that nobody in the study inter- What, then, is implied by this ontology of complex- acts with anyone else in it’’. ity? According to Complexity Science, complicated sys- This idea of rawness is seen within the emerging tems have large attribute-rich components, with simple Computational Social Sciences as providing the foun- and limited interaction, while complex systems typically dation for a new approach to studying the social world, have many, simple components interacting in sophisti- which is argued to have the potential to solve many of cated ways (Andersson and To¨ rnberg, 2017). If the sociology’s deep-rooted problems. Alex Pentland refers structure of the components of an automobile is an to this as ‘‘sociology of the 21th century’’ in Manovich, example of the former, the fluid organization of a 2011:464, and says that digital data ‘‘give us the chance flock of birds is an example of the latter. Because of to view society in all its complexity, through the mil- the structure of digital trace data, the social ontology of lions of networks of person-to-person exchanges’’. The complexity has increasingly become the implicit or even new data allows us to navigate ‘‘data sets without explicit foundation for many of the empirical compu- making the distinction between the level of the individ- tational approaches to social science, and the relation- ual component and that of the aggregated structure’’ and interaction-focused field of Complexity Science has (Latour et al., 2012: 590), which, Lazer et al. (2009) become a powerful impetus in the development of the 4 Big Data & Society new social scientific disciplines around Big Data (see complexity to constituting an implicit foundation in e.g. Conte et al., 2013; Jungherr, 2015). many of the tools and approaches used for social sci- The complexity approach has proven highly cap- entific research. For instance, complex network analysis able of analyzing many types of systems that have has become widely used within parts of mainstream otherwise been impenetrable to formal approaches social science, focusing on how relations/interactions (e.g. Mitchell, 2009). Complexity Science is centered on the micro-level lead to the formation of higher on the core use of formal models of mass-interaction, level social patterns (e.g. Strogatz, 2001). The influence focusing not as much on social facts and aggregate of complexity thinking, and the linking of macro explanations, but rather on the emergence of aggre- dynamics to individual behavior, is also exemplified gate pattern. This means putting the finger exactly on by many theories and concepts that are increasingly the limits of aggregate measures, since emergence is used by social scientists even outside of digital methods, the very opposite of aggregation (e.g. Wimsatt, 2007: including terms such as ‘‘threshold effects’’ or ‘‘tipping- 274–276); the whole is different from the sum of its points’’, ‘‘power-laws’’, ‘‘preferential attachment’’, and parts (see Anderson, 1972). so on. For a sociologist, this view of the social world is The increasing prevalence of the complexity perspec- more reminiscent of Tarde’s notion of imitation than tive is often argued to be not only the result of differ- Durkheim’s concept of social facts (Candea, 2010; ences in the ‘‘rawness’’ of the data and what it can To¨ rnberg, 2017), but while Tarde departed from reveal of the social world, but also to reflect actual theory and was criticized for lack of method, changes in the nature of social interaction. There may Complexity Science developed largely because of, and be certain merits to such claims: just as researchers are on the foundation of, new methods (Helin et al., 2014). shaped by the social life of Big Data, so are its users. The epistemological perspective of complexity Differences that are often emphasized is that digital relates to the generally accepted view in the various social life seems more quantitative, regular, predictable sciences dealing with complex collective behavior, that (as illustrated by the successes of platform and data there exist some fundamental differences between the analysis companies that subsist precisely on predictive individual and the aggregate levels (Calhoun, 2002; analysis), which is argued to motivate a more natural Knorr-Cetina and Cicourel, 1984). Traditionally, in scientific approach to the data. Two such lines of argu- ment in particular can be identified. the social sciences, the existence of levels have often been assumed, and questions have focused mainly on First, social media is seen as having brought with it issues like whether the micro or macro level is the suit- increased quantification of social interaction, since, for able level of analysis, or if the two would be possible to instance, the number of likes that a post has received on ‘‘reconcile’’ using some higher-level theory. In practice, Facebook requires no additional operations to be this has primarily been handled through disciplinary quantified. While traditional data on social interaction and methodological separations, leaving the question requires transcription by researchers, a process that lifts of the emergence of structure from individuals to the difficult questions about the interpretation of inton- road-side. Complexity Science, in contrast, focuses ations, pauses and subtle facial expressions, the users explicitly and almost exclusively on this question of platforms like Twitter and Facebook seem to have (Erdi, 2007; Mitchell, 2009). already done the work of encoding their messages into The Computational Social Science introduced by a quantified and standardized format. In other words, e.g. Lazer et al. (2009) constituted to certain degree a our social life has thus become more natively quanti- reboot or re-appropriation of the term: the fied, more ordered and structured, as we increasingly Computational Social Science of the previous decade use numbers and codified data to navigate our social was part of Complexity Science and was never linked world, flattening the exchange of meaning into num- to large-scale data, but rather approached society bers, written words, and a choice from a predefined mainly through simulation in general, and agent- set of smileys. Such native quantitativity has historic- based modeling in particular. While a certain rift ally implied a tendency for an increased focus on quan- between the data-focused and the simulation-focused titative approaches also for the study of these systems, Computational Social Science can be identified, they as is exemplified by the strong mathematical orientation were, and are, strongly connected through a common of mainstream economics (Sayer, 1992). perspective on society as a relationally and dynamically Second, social media is seen as having brought with complex system. it an increased prevalence of synchronistic behavior, With the rise of Big Data, Complexity Science seems taking the form of cascades of similar ‘‘viral’’ actions. to be increasingly experiencing an ‘‘obliteration by This has in part been argued to be the result of the ways incorporation’’ (Merton, 1968: 27), as the perspective that social media are designed to bring forth certain transitions from an explicit focus on emergence and type of behavior, by enticing and bringing to the fore To¨rnberg and To¨rnberg 5 our reactions and instincts, thus undermining the In these discussions, the notions of digitalization and agency we have over our own minds (Alter, 2017). dematerialization were connected to the larger contem- Our technological sophistication is thus said to have, porary discussions around terms like ‘‘postmodernity’’, ironically, brought us away from reflexive agency and ‘‘liquid modernity’’, ‘‘late capitalism’’, and ‘‘acceler- closer to reactive, animalistic and instinctual behavior ation’’. Analyses of the implications of digital technol- well-described by analogies such as ‘‘avalanches and ogy can be found in a range of strands, from granular flows, flocks of birds and fish’’ (Ball, 2012: Baudrillard’s (1994) simulacra, Jameson’s (1991) cultural ix). The predictability of such online behavior is analysis of late capitalism, Beck’s (1992) portrayal of de- argued to have enabled the successes of platform and structuration in the Risk Society; in Giddens’ (2002) data analysis companies that subsist precisely on pre- imagery of a disordered runaway world; in Bauman’s dictive analysis. (2000) liquid modernity in which ‘‘flows’’ replace the In summary, the combination between a social determinate social structure and cultural systems; in life that is more reactive, instinctive and natively Archer’s (2014) morphogenic society where morphogen- quantified, and an understanding of Big Data as some- esis increasingly dominated over morphostasis. thing fundamentally new, raw and natural, is stirring to The common denominator of these views is in many life again the old corpse of naı¨ve naturalism, whose his- ways the precise opposite of the epistemological conclu- torical refusal to lie down was noted already by sions taken in the current debate on digital media: these Bhaskar (1978). Because of the methods and algorithms scholars saw the digital technology as being part of a that have molded this new digital data, this naturalist late modernity ‘‘uncontrollable and quintessentially ontology takes the shape of complexity. This idea of a kaleidoscopic in form’’ (Archer, 2014: 1). As blurring of the boundary between the natural and Archer emphasizes, this means that just because a social world, and suggested taming of society’s ‘‘vex- social phenomenon (institution, role, group, belief or atious nature’’ (Dahrendorf, 1968: 23), may not practice) continues to bear the same name, ‘‘it cannot always be explicit and openly articulated, but is automatically be regarded as being ‘the same’’’ (p.6), nonetheless apparent in the way a majority of the scho- and continuously stable. Digitalization and dematerial- lars within the computational approach the social ization were thus seen as part of the processes of post- world. modernism in that it constitutes the dissolution of an In the following sections, we will look at the limits of impediment to the pace of change. This is part of the this new naturalism, by scrutinizing its implicit assump- larger process of modernity, in which, ‘‘instead of tions about the social world. We will approach these inhabiting a stable world of objects made to last, limitations through a contrast with the discussions at human beings found themselves sucked into an accel- the early era of digitalization. erating process of production and consumption’’. (Arendt 1958: xiv) According to these scholars, digitalization can thus Early digital research and beyond: be understood as yet another step or phase of this tran- Liquidity and postmodernity sition, in which capitalism has, as Jameson (1991) In the early discussions on the implications of digital argued, reached its purest form. Through digitalization, technology, pre-dating the age of ubiquitous social this process has finally melted the very materiality of media and digital platforms, digital technology was pri- technology, permitting all that is solid to melt into air; marily viewed through the lens of dematerialization: the or, in this case, source code. ‘‘It is as though we had transition of technology from atoms to bits (Mitchell, forced open the distinguishing boundaries which pro- 1996; Negroponte, 1996). The focus in this literature tected the world, the human artifice, from nature [...] was on the social implications of the possibilities for delivering and abandoning to them the always threa- rapid change brought on by digital technology: through tened stability of a human world’’ (Arendt, 1958: 126). the Internet, technological changes can be distributed In this perspective, the stability of the social world is to billions of users within seconds, and the reactions of connected to the very materiality of technology: since these users can be instantly evaluated. According to material change tends to be slow, technologies have these early scholars, the digital is not limited by the provided a relatively solid foundation for social pat- constraints of the material world: it has left behind terns to lean on (Elder-Vass, 2017). For instance, a building can remain standing for hundreds of years the sculpturing of hard matter for the fluidity of elec- trons and software. Through this, technology and contribute to propagate the social context in was argued to having reduced its function as a stabilizer which it was constructed. Thus, in the understanding of social structures, which implies that the social of this literature, it is precisely this stabilization that is context and the basis for interpretation become more undermined by the dematerialization brought by digital fluid. technology. 6 Big Data & Society As Hayles (1999) points out, this new instability is broader feedback processes of innovation, in particular brought into our very language, and the ways we inter- the evaluation of how new innovations affect the social pret the world. To analyze this, Hayles build upon web in which they become part. Sophisticated data ana- Lacan’s concept of ‘‘floating signifiers’’, adding that lysis, A/B testing, and instantaneous evaluation of the they through digital technology also begin to flicker: social practices evolving on digital platforms enable our words become unstable, their meaning amorphous platform owners to shape their users’ behavior with and constantly transforming. In Hayles (p.52) words, unprecedented precision and control. The feedback information technologies ‘‘fundamentally alter the rela- loop between evaluation and innovation (described by tion of signified to signifier[,] thus carrying the instabil- e.g. Lane, 2016) has become increasingly rapid, as tech- ities implicit in Lacanian floating signifiers one step nology owners have precise and detailed data on how further.’’ their products are taken up in a larger sociotechnical Taken together, the view proposed by these early crit- context. ical scholars can be understood as digitalization bringing These two factors have meant that the fluidity and increasing openness (in the sense of ‘‘open systems’’ in e.g. capacity for rapid change of dematerialized technology, Bhaskar, 1978) to society as a system by enabling rapid theorized by the scholars of the early days of the technological change, in turn bringing what Lane and Internet, have not only played into a postmodern cul- Maxfield (2005) call ‘‘ontological uncertainty’’: an ture of late capitalism, but has also been channeled into increased propensity for qualitative change. This propen- new forms of power for the owners of technology. sity is illustrated by concepts like ‘‘web time’’ (Karpf, Technological power can now be exercised in more 2012), that describes the increased pace of sociotechnical sophisticated, nimble and illusive ways than ever change brought by information technology. Or in the before, as the dematerialization of technology means terminology of Simon (1962), digitalization implies that that the ownership even of consumer products has the ‘‘short run’’, in which a system can be understood become possible to centralize. The artifacts that we con- formally, becomes increasingly short (Andersson et al., sume and surround ourselves with are increasingly 2014). As Sayer (1992: 122) points out, this in turn limits rented rather than owned, as apps, programs, and the usefulness of quantification, since the objects under technological platforms are increasingly located in the measure are not qualitatively invariant. cloud, and thus prone to constantly change without In our view, despite being largely neglected in the warning. The ‘‘zero-marginal-cost’’ of software has contemporary literature, this early characterization of resulted not in the end of capitalism, as some social digitalization remains in many ways accurate as a scientists rather optimistically theorized (Mason, description of the effects of digitalization on social 2015; Soderberg, 2015), but rather in a transition of life; implying a fluidity and instability of meanings business models from selling to renting. In other and structures constantly boiling under the surface of words, rather than undermining the private ownership the ostensible constancy of fixed numbers and symbols. regime of capital, this has had the effect of undermining However, digital technology has since developed in the already tenuous ownership of consumers (von some new and at the time unforeseen directions, Busch, 2008). Instead of workers gaining the ownership which have meant that the fluidity of meaning and of the means of production, they have increasingly lost structures have become channeled in unexpected ways. ownership even of their goods of consumption. While digitalization brings increasing centralization and sophistication in the expression of technological Fluid technology in the era of platforms power, technology’s function as a shaper of social First, at the time of these theories, the Internet was a behavior is in itself nothing new. Technology has highly fragmented environment of rapid and often always been in and for the power of its owners and informal experimentation. Today, the Internet has producers, as a force capable of shaping and directing infrastructurally instead become a place of extreme cen- social life in their interests. There is hardly an activity, tralization: information systems have turned out to be belief or form of interaction that is not mediated by ripe with natural monopolies, creating conditions for artifacts and thus affected by this hidden ideological large platform companies. The structures that typically face of technology (Feenberg, 1991), whether wedding follow from this have become increasingly similar to a rings and clothes, candles and incense, or money and art—these artifacts store and propagate societal struc- form of private governments, with power to control flows of information, and, as the ‘‘sharing economy’’ tures (Elder-Vass, 2017). Social life has always played (e.g. Uber and Airbnb) illustrates, at times even to tax out within technological platforms that shape and their user base. frame our interaction and provide context to it, grant- Second, it is not only the roll-out of new technology ing permanence to our symbols and our language that has changed with digitalization, but also the (Collins, 2014). The impact of the technological context To¨rnberg and To¨rnberg 7 is not merely incidental: churches, for instance, are seemingly natural and spontaneous outcome of expressions of power and authority, consciously human behavior. This form of distributed control fits designed to inspire awe toward the power of Gods, reli- into the individualization of power that is part and gious institution and holy men. They instill authority parcel of postmodernity; as Bauman (2000) notes, con- into the solemn priest behind the pulpit, and remind us trol has become part of individuality itself; no longer is of the larger story within which we are but minor the focus on producing homogeneity by whipping devi- players, thus shaping and giving meaning to our behav- ators into conformity, but rather on the emergence of a ior and interaction. Online digital platforms of today collective outcome in line with certain interests—direct- are not unlike such physical meeting places: they too ing the herd rather than the beast, through a shaping of provide the context within which rituals and social life context rather than through explicit command-and- take place. They condition our interactions, shape who control. The transition from technology being a has authority and who is heard. rather blunt tool for social control to a virtual social But the combination between centralization of scalpel thus implies that digitalization has brought an power and the dematerialization of technology implies era of platform power, in which technology provides a an important transition in the expression of techno- new level of herd control. logical power. While yesteryears churches were carved from stones, rocks and clay, the digital churches of The nature of digital data today constantly shift underneath our feet. While phys- ical churches were blunt tools for shaping our lives, What, then, are the implications of the condition of needing to be backed up by damnations and inquisi- postmodernity and the technological power of platform tions, the digital churches read and react to our every owners for digital data research? What limits do these gesture and expression. They are capable of customiz- observations imply for the computational study of digi- ing their expressions to individuals, or trying a hundred tal social life, and how do they clash with the tendency variations of the colors of the pulpit to see how its of Complexity Science to naturalize social life—seeing faithful are affected. What use culture emerges among social patterns not as the result of contingency and users is importantly controlled by what the system conflict, but as expressions of universal social laws? ‘‘affords’’ (Norman, 1999), and what can be done ‘‘fric- As we saw in the first section, the promise of the Big tionlessly’’ (Shaw, 2015): subtle design choices herd the Data revolution has described a world of previously users in certain directions, in ways related to the con- unimaginable data; a flood of coffee-table discussions cept of ‘‘nudging’’ (Thaler et al., 2013). revealing traces of the lives, dreams, and feelings of In other words, the increasing fragmentation and hundreds of millions of people. This has painted a pic- fluidity following from dematerialization has somewhat ture—which hangs centrally in the halls of e.g. paradoxically implied increased centralization of con- Computational Social Science—of the ‘‘true’’ relational trol, as it permits the owners of technology to express nature of social life being unveiled, showing a social life power by shaping meaning and structures through which is not only measurable but even predictable. gentle nudging of underlying technical rules. This con- While this painting shows a dreamy world for the trol does not congeal the constant boiling fluidity of social sciences, another reality appears when we lift our meaning, but rather dynamically directs its flow. gaze from the data feeders, and cast it upon the less Control moves to lower ontological strata, shaping out- than appetizing context in which the data is fed to us. comes through the underlying rules of interaction Rather than a spontaneous and natural production of rather than through explicit control. In this demateria- social traces, we see how the data is produced, selected lized modernity, the fluidity of meanings and structures and provided to us by platform owners pursuing their afford a form of control that seemingly paradoxically own interests. Many aspects are left out of this data. emerges from the bottom up. For instance, the platforms and their rules that shape This transition in the expression of power is remin- the online behavior are not readily visible: their inter- iscent of the transition described by Norbert Elias ests and incentives instead lie latent as hidden forces (2006) in The Court Society. Just as Luis XIV that guide individual behavior and the emergent embedded his control into the social rules of polite social practices of the platforms. Thus, at the same interaction rather than, as previous regimes, through time as the contextual aspects and the power of plat- violence and explicit control, power in the era of digital form owners are becoming increasingly central to platforms is expressed not top-down, but through invis- understanding social life, our focus as researchers is ible nudging and shaping of local behavior, molding of increasingly on the patterns of interaction, which, as social rules and practices, and thus, control is they lose their natural setting, become naturalized and embedded in the very rules of our interaction. The decontextualized, just in the way that complexity per- interests of platform owners thus appear to us as spectives have historically had a tendency of implying 8 Big Data & Society naturalization (Byrne and Callaghan, 2014; Uitermark, informal practice, to becoming encoded in a button, 2015). When Big Data is seen as merely an encoded, which ended up producing the macro-pattern of ‘‘vir- measurable version of social reality—possibly with ality’’ and ‘‘diffusion’’, appearing as repeated behavior some technological bias to be corrected for—the com- cascading through a network. plex social and technological forces that produced them The result of thinking of Big Data as providing a are flattened: the data is made to seem natural and form of privileged access to the social world is that inevitable rather than contingent and contested; they researchers flock to study relatively predictable and cor- are made subject of reification rather than critique. related social behavior on ostensibly disintermediated This idea of Big Data as an ‘‘encoding’’ of social life online platforms, while disregarding the sociotechnical disregards the complex interplay between the techno- conditions that lead to the formation of that behavior. logical and social aspects of human life that have pro- The digital platforms are developed using significantly duced the data. Rather than merely a one-way more sophisticated methods and larger data quantities encoding, the production of data takes place by digital than what is available to researchers, with even the platforms directing and limiting action by providing a most seemingly insignificant design decision being the ‘‘grammar of action’’ that make certain activities result of meticulous A/B-testing and data analysis. doable, and thus rendering social activities available From the basis of the Complexity Science metaphor for measurement, analysis, commodification, and of social behavior as the playing of a game, the data manipulation (Van Dijck, 2013). But at the same thus makes more visible the ‘‘playing of the game’’, time, users are not helpless puppets in this process: while obscuring the ‘‘rules of the game’’ and the inter- they are often aware of the ways that measures and ests that shaped them. The data thus becomes a perfect technologies play into their social lives, and reflexively fit for a naturalizing science that tends to see the rules take account of this in their use. They are not as universal and their outcomes as inevitable. ‘‘encoding’’ their behavior, but rather employing and The formal tools and mathematical models that we enacting the methods, performing through the meas- apply to study this world hinge on the stability of ures in front of an ‘‘imagined audience’’ (Litt, 2012): meaning and understanding that are exchanged. But ‘‘social actors produce methodical accounts of social such assumptions have not become less problematic life as part of social life’’ (Lynch, 1991). The platform through digitalization, but rather more so, as symbols and meaning are becoming more local in time and owners are in turn aware of these dynamics: the very creation of digital platforms tends to involve the imple- space. Interpretation has not become less central in mentation of sociological and social psychological the research process; its locus has merely moved, as ideas; their use ranging from the benign push (e.g. sug- interaction is simultaneously more quantified and its gesting friends through triadic closure) to what basic- meaning more fragmented and flickering. ally amounts to a weaponized social psychology (e.g. This does not mean that the observation of increas- ‘‘engagement maximization’’ through the application of ing complexity, native quantitativity, and the potential research on addiction). In short, measures not only for predictability are false. Big Data is seemingly para- describe, but are enacted and made part of social life, doxically associated to both these developments: it is in the type of continual process of reflexivity between simultaneously more liquid and more natively quanti- diverse actors and roles that is quintessential of the fied; it is simultaneously more open and more measur- vexatious nature of social life. able; it is simultaneously more bottom-up and more Online behavior and content are in other words a amenable to control. It is, in short, becoming easier consequence both of how digital technologies work to count, while at the same time harder to interpret and what people do with them, in ways that are exceed- what we are counting. ingly difficult to separate. Rather than thinking of The answer is not, as has been the case among some online social life through a separation between scholars, to simply reject computational methods or ‘‘human behavior’’ to be studied, and ‘‘technological suggest that the entire notion of ‘‘new data’’ is merely bias’’ to be in various ways ‘‘corrected for’’, content a red herring since many of its aspects have a long his- is perhaps better understood as the output of an tory (Marres, 2017; Uprichard et al., 2008)—the epis- entanglement between the two—a sociotechnical temological and methodological demands of system (Marres, 2017). This casts technology as a defin- complexity in general, and Big Data in particular, are ing feature of human society, rather than as something real and will have to be reckoned with. But neither is to be corrected for. The way that ‘‘virality’’ is employed the answer a methodological one: we will not find any to make claims about the new ‘‘instinctual’’ and ‘‘react- method to match and capture society in a single ana- ive’’ nature of digital social life is the case in point here; logy (Andersson et al., 2014; Archer, 2014). The solu- Halavais (2014) shows how the re-tweet emerged as a tion needs to concern the underlaborer on which the sociotechnical script on Twitter: beginning as an approach is founded. Instead of continuing to To¨rnberg and To¨rnberg 9 approach Big Data by extending ‘‘the tool found suc- digitalization thus serves as a reminder of aspects of cessful in one domain to decipher the other’’ (Khalil, the social world that the new computational view con- 1995: 414–415), we suggest following Perona’s (2007) tinues to leave out: as Andersson et al. (2014) argue, advice with regards to social complexity: to take ‘‘a society is neither a complex nor a complicated system, turn to ontology’’. Ontological perspectives are not but rather, it displays both these properties which only matters of philosophical curiosity, but have pro- makes it qualitatively different from both types of sys- found implications for how we can and should tems: thus, the reduction of social reality to these ‘‘ana- research, manage, and think about social phenomena. logical imaginations are simply misleading’’ (Archer, What is needed is a meta-theory capable of respecting 2013: 146). the openness and non-decomposability of social sys- The technological power of platform owners is to a tems in general, and digital social systems in particular, large part enabled by the same new tools for data ana- while at the same time admitting the methodological lysis as used by social scientists—indeed, the private and epistemological conditions of Big Data—large sector is often the driver for the development of these data sets characterized by relational complexity, emer- tools. These efforts have been immensely economically gence, and self-organization. Complex Realism pro- successful, as illustrated by companies like Google and vided such a response to the insights of Complexity Facebook. But we must not forget that the aims of Science in the 90s, bringing the patchy and partial these corporations are quite different from the aims of social ontology of complexity into dialogue with researchers: they seek prediction and control, while Critical Realism (e.g. Byrne and Callaghan, 2014). A researchers (at least should) seek explanation and under- potential way forward, to be pursued in future publica- standing. In trying to make a user click an ad, corpor- tions, could thus be to follow an analogous response to ations are less interested in the why than the how. These the insights of digital data. aims are implicitly built into the affordances of the tools, and just like the online platforms shape the actions of their users, these data analytical tools tend Conclusions to shape the behavior of their users, that is, our behav- The first section of this paper showed how the structure ior as social scientists. They thus nudge researchers of digital data is making trouble for the traditional toward pattern-finding and prediction, rather than in- social scientific variable-based approach, creating a depth understanding. push toward new social ontologies matching the struc- As noted by early scholars of digital technology, the ture of the data. This has sparked a renewed, complex flexibility and rapid change enabled by digitalization naturalism, within which social systems are increasingly are part of the larger processes of postmodernity. But approached through the formal methods of the natural the digital world is not well-described by the classical sciences—seeing social structures as patterns naturally understanding of postmodernity alone. It is also part of emerging from mass-interaction, which is taken to an increased centralization of technological power, and permit the leaving out of institutional, technological, a change in the role of technology in social life. The and contextual aspects of social life. In the second sec- postmodern aspects of digitalization do bring more tion, we revisited the discussions of the early days of ‘‘openness’’ (in the sense of e.g. Bhaskar, 1978) to soci- digitalization: these instead saw digitalization as part of ety, with social structures becoming more fragmented, the larger processes of postmodernity, implying liquid, and prone to qualitative change, which can argu- increased systemic openness as the transition from ably be seen in some cultures developing on the Internet atoms to bits brought the undermining of stabilizing (Nagle, 2017). This has furthermore rightly been under- forces of social systems. We extended this perspective stood to pose limits on quantification, since it implies by discussing how the liquidity of technology has that the measured objects are not qualitatively invari- turned into a means of control as online social life ant (Sayer, 1992, 2000). has become centralized into large platforms that work This is made more confounding by the fact that this to shape human behavior according to their interests. development is occurring in an increasingly natively Together, these developments reveal a clash between quantitative context, in which people are communicating the underlying assumptions of the computational through numbers and coded messages. While this approach to social data and the context in which the changes the locus of scientific interpretation, since researchers no longer need to transcribe conversations, data is produced. While the trouble-making of digital data may usefully help point to limits of the traditional it does not reduce the centrality of interpretation. While variable-based approach as well as the constructed transcription brings the researcher a direct experience of nature of scientific data, the new data brings new the inherent loss of nuances of meaning, making the limits, and are similarly constructed around certain local and contextual nature of meaning hard to ignore, methods and techniques. The early debate on the digital platforms conceal the ways that their 10 Big Data & Society that contributed greatly to this article: David Byrne, Brian quantitative data is produced by fluid and quickly chan- Castellani, Lasse Gerrits, Adrian MacKenzie, and Emma ging sociotechnical systems, whose signifiers flicker and Uprichard. vary over time and context. These changes are deter- mined in part by constantly evolving implicit social prac- Declaration of conflicting interests tices that are self-consciously hyper-ritualized, The author(s) declared no potential conflicts of interest with fragmented, and local to social context, and co-evolving respect to the research, authorship, and/or publication of this with technological change in the underlying platforms. article. Similarly problematic is the notion that human online behavior being more reactive should lead us to Funding disregard context and view social life through ana- logues like ‘‘flocks of birds and fish’’. While techno- The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of logical platforms do act on humans through their this article: the European Union’s Horizon 2020 research powerful social life, wielded and directed by platform and innovation programme under grant agreement No. owners, the fact that platforms are capable of herding 732942, as well as from Swedish Research Council project users through technological nudges and affordances grant 2016-03515_3. should not imply a reduced focus on context, but rather engage us to turn our gaze toward precisely the ORCID iD power of the platforms. But instead of moving towards Petter To¨ rnberg http://orcid.org/0000-0001-8722-8646 increasing focus of these contextual aspects and the role of platform owners, many scholars interested in digital References social life have been lured by the siren-call of new meth- ods and abundant digital data to lose their gaze from Alter A (2017) Irresistible: The Rise of Addictive Technology and the Business of keeping us Hooked. London: Penguin. precisely these factors. Thus, social patterns are natur- Anderson C (2008) The Petabyte age: Because more isn’t just alized: we take behavior on social platforms to be tell- more–more is different. Wired Magazine (16.07): 106–120. ing about the nature of social life, while it may in fact Anderson PW (1972) More is different. Science 177(4047): say more about the interests of the platform owners. 393–396. For us as researchers, this implies a need for not only Andersson C, To¨ rnberg A and To¨ rnberg P (2014) Societal studying processes of emergence, but for doing so while systems – Complex or worse? Futures 63: 145–157. keeping in mind that there is nothing natural about Andersson C and To¨ rnberg P (2017) Wickedness and the human behavior and that there is no such thing as anatomy of complexity. Futures 95(1): 118–138. ‘‘raw data’’. Social media should perhaps be thought Archer M (2014) Late Modernity: Trajectories towards off less like savannahs of free-running herds of humans, Morphogenic Society. Cham: Springer Science & Business than like zoos in which caged users are made to dance Media. Archer MS (2013) Social Morphogenesis. Dordrecht: Springer to the tune of capital; ‘‘no data, big or small, can be Science & Business Media. interpreted without an understanding of the process Arendt H (1958) The Human Condition. Chicago, IL: that generated them’’ (Shaw, 2015: 1), and these pro- University of Chicago Press. cesses are entangled in the interests of capital. Bajec IL and Heppner FH (2009) Organized flight in birds. This calls for a critical computational social science Animal Behaviour 78(4): 777–789. that does not sacrifice context, clarity, and critique for Ball P (2012) Why Society is a Complex Matter: Meeting the automatic identification of large-scale patterns, pre- Twenty-first Century Challenges with a New Kind of dicated in the notion that breadth could replace depth Science. Cham: Springer Science & Business Media. and context as basis for interpretation. If we are not to Barton AH (1968) Bringing society back in: Survey research be drawn by the siren-song of abundant data, sung by and macro-methodology. The American Behavioral the owners of technological platforms precisely to lure Scientist 12(2): 1–9. Baudrillard J (1994) Simulacra and Simulation. Ann Arbour: us into drowning in the data deluge, we must tie our- University of Michigan Press. selves to the mast of a critical and explicit metatheory: Bauman Z (2000) Liquid Modernity. Cambridge: Cambridge for only from a stable ontological position will we be Press. able to hear not only what the data has to sing to us Beck U (1992) Risk Society: Towards a New Modernity. about the social world, but also to listen for those London: Sage Publication Ltd. things about which they remain so curiously silent. Bell G, Hey T and Szalay A (2009) Beyond the data deluge. Science 323(5919): 1297–1298. Acknowledgments Bhaskar R (1978) A Realist Theory of Science. Hassocks: The authors would like to thank the participants at the Harvester Press. ODYCCEUS workshop On the Limits of Naturalism in a Boyd D and Crawford K (2012) Critical questions for big Digital World, in Venice, Italy, January 2018, for discussions data: Provocations for a cultural, technological, and To¨rnberg and To¨rnberg 11 scholarly phenomenon. Information, Communication & Hey T, Tansley S and Tolle Kristin M (2009) The Fourth Society 15(5): 662–679. Paradigm: Data-intensive Scientific Discovery. Vol. 1. Byrne D and Callaghan G (2014) Complexity Theory and the Redmond, WA: Microsoft research. Social Sciences: The State of the Art. London / New York, Jameson F (1991) Postmodernism, or, the Cultural Logic of NY: Routledge. Late Capitalism. London: Verso. Johnson S (2002) Emergence: The Connected Lives of Ants, Calhoun C (2002) Classical Sociological Theory. Oxford: Brains, Cities, and Software. New York, NY: Simon and Blackwell. Schuster. Candea M (2010) The Social after Gabriel Tarde: Debates and Jungherr A (2015) Analyzing Political Communication with Assessments. New York, NY: Routledge. Digital Trace Data. Cham: Springer International Castellani B (2014) Focus: Complexity and the failure of Publishing. quantitative social science. Discover Society. Available Karpf D (2012) Social science research methods in Internet at: https://discoversociety.org/2014/11/04/focus-complex- time. Information, Communication & Society 15(5): ity-and-the-failure-of-quantitative-social-science/ 639–661. (accessed September 2018). Khalil EL (1995) Organizations, naturalism, and complexity. Christakis NA (2012) A new kind of social science for the 21st Review of Social Economy 53(3): 393–419. century. Edge. Available at: http://edge.org/conversation/ Kitchin R (2014) Big Data, new epistemologies and paradigm a-21st-century-change-to-social-science (accessed shifts. Big Data & Society 1(1): 1–12. September 2018). Knorr-Cetina KD and Cicourel AV (1984) Advances in Social Collins R (2014) Interaction Ritual Chains. Princeton: Theory and Methodology: Toward an Integration of Micro- Princeton University Press. and Macro-sociologies. Boston, MA: Routledge & Kegan Conte R, Hegselmann R and Terna P (2013) Simulating Paul. Social Phenomena. Heidelberg: Springer Science & Lane DA (2016) Innovation cascades: Artefacts, organization Business Media. and attributions. Philosophical Transactions of the Royal Dahrendorf R (1968) Essays in the Theory of Society. Society B 371(1690): 1–9. Stanford, CA: Stanford University Press. Lane DA and Maxfield RR (2005) Ontological uncertainty Dorigo M and Stu¨ tzle T (2010) Ant colony optimization: and innovation. Journal of Evolutionary Economics 15(1): Overview and recent advances. In: Gendreau M and 3–50. Potvin JY (eds) Handbook of Metaheuristics. Latour B, Jensen P, Venturini T, et al. (2012) ‘The whole is International Series in Operations Research & always smaller than its parts’ – A digital test of Gabriel Management Science. Vol. 146. Boston, MA: Springer. Tardes’ monads. The British Journal of Sociology 63(4): Edwards A, Housley W, Williams M, et al. (2013) Digital 590–615. social research, social media and the sociological imagin- Lazer D, Pentland AS, Adamic L, et al. (2009) ation: Surrogacy, augmentation and re-orientation. Computational social science. Science 323(5915): International Journal of Social Research Methodology 721–723. 16(3): 245–260. Litt E (2012) Knock, knock. Who’s there? The imagined audi- Elder-Vass D (2017) Material parts in social structures. ence. Journal of Broadcasting & Electronic Media 56(3): Journal of Social Ontology 3(1): 89–105. 330–345. Elias N (2006) The Court Society. New York, NY: Pantheon Lynch M (1991) Method: measurement ordinary and scien- Books. tific measurement as ethnomethodological phenomena. Erdi P (2007) Complexity Explained. Berlin: Springer Science In: Button G (ed.) Ethnomethodology and the Human & Business Media. Sciences. New York, NY: Cambridge University Press, Feenberg A (1991) Critical Theory of Technology. New York, pp. 77–108. NY: Oxford University Press. Mackenzie A (2012) More parts than elements: How data- Galison P (1997) Image and Logic: A Material Culture of bases multiply. Environment and Planning D: Society and Microphysics. Chicago, IL: University of Chicago Press. Space 30(2): 335–350. Giddens A (2002) Runaway World: How Globalisation is Manovich L (2011) Trending: The promises and the chal- Reshaping Our Lives. London: Profile books. lenges of big social data. Debates in the Digital Halavais A (2014) Structure of Twitter: Social and technical. Humanities 460–475. Available at: http://manovich.net/ In: Weller K, Bruns A, Burgess J, et al. (eds) Twitter and index.php/projects/trending-the-promises-and-the-chal- Society. New York, NY: Peter Lang, pp. 29–42. lenges-of-big-social-data. Hayles NK (1999) How We Became Posthuman. Chicago, IL: Manovich L (2016) Online post. Facebook. Available at: University of Chicago Press. https://goo.gl/GdcJhm (accessed September 2018). Helbing D, Buzna L, Johansson A, et al. (2005) Self-orga- Manyika J, Chui M, Brown B, et al. (2011) Big Data: The nized pedestrian crowd dynamics: Experiments, simula- Next Frontier for Innovation, Competition, and tions, and design solutions. Transportation Science 39(1): Productivity. McKinsey Global Institute. Available at: 1–24. https://bigdatawg.nist.gov/pdf/MGI_big_data_full_ Helin J, Hernes T, Hjorth D, et al. (2014) The Oxford report.pdf (accessed September 2018). Handbook of Process Philosophy and Organization Marres N (2017) Digital Sociology: The Reinvention of Social Studies. Oxford: Oxford University Press. Research. Hoboken: John Wiley & Sons. 12 Big Data & Society Marres N and Weltevrede E (2013) Scraping the social? Issues So¨ derberg J (2015) Hacking Capitalism: The Free and Open in live social research. Journal of Cultural Economy 6(3): Source Software Movement. New York, NY / London: 313–335. Routledge. Mason P (2015) Postcapitalism: A Guide to Our Future. Thaler RH, Sunstein CR and Balz JP (2013) Choice architec- London: Penguin Random House. ture. In: Shafir Eldar (ed.) The Behavioral Foundations of Merton RK (1968) Social Theory and Social Structure. New Public Policy. Princeton, NJ: Princeton University Press. York, NY: Free press. Tornberg A (2017) The wicked nature of social systems – A Mitchell M (2009) Complexity: A Guided Tour. Oxford: complexity approach to sociology. Gothenburg University, Oxford University Press. Gothenburg. Mitchell WJ (1996) City of Bits: Space, Place, and the To¨ rnberg P (forthcoming) Complex realist economics: Infobahn. Cambridge: MIT press. Toward an ontology for an interested pluralism. Review Morin E (2007) Restricted complexity, general complexity. of Social Economy. In: Gershenson C, Aerts D and Edmonds B (eds) Uitermark J (2015) Longing for Wikitopia: The study and Worldviews, Science and Us: Philosophy and Complexity. politics of self-organisation. Urban Studies 52(13): Singapore: World Scientific. 2301–2312. Nagle A (2017) Kill All Normies: Online Culture Wars From Uprichard E (2013) Describing description (and keeping caus- 4Chan and Tumblr To Trump And The Alt-Right. ality): The case of academic articles on food and eating. Winchester: Zero Books. Sociology 47(2): 368–382. Negroponte N (1996) Being Digital. New York, NY: Knof. Uprichard E, Burrows R and Byrne D (2008) SPSS as an Norman DA (1999) Affordance, conventions, and design. ‘inscription device’: From causality to description? The Interactions 6(3): 38–43. Sociological Review 56(4): 606–622. Perona E (2007) The Confused State of Complexity Van Dijck J (2013) The Culture of Connectivity: A Critical Economics: An Ontological Explanation. Milano: Springer. History of Social Media. Oxford: Oxford University Press. Rosling H (2010) The Joy of Stats. London: BBC4. von Busch O (2008) Fashion-able. Hacktivism and engaged Savage M and Burrows R (2007) The coming crisis of empir- fashion design. University of Gothenburg. ical sociology. Sociology 41(5): 885–899. Watts DJ (2007) A twenty-first century science. Nature Sayer A (1992) Method in Social Science. 2nd ed. New York, 445(7127): 489–489. NY: Routledge. Watts DJ (2013) Computational social science: Exciting pro- Sayer A (2000) Realism and Social Science. London: Sage gress and future directions. The Bridge on Frontiers of Publication Ltd. Engineering 43(4): 5–10. Shaw R (2015) Big Data and reality. Big Data & Society 2(2): Wimsatt WC (2007) Re-engineering Philosophy for Limited 1–4. Beings: Piecewise Approximations to Reality. Cambridge: Simon HA (1962) The architecture of complexity. proceedings Harvard University Press. of the American Philosophical Society 106(6): 467–482. Wolfram S (2002) A New Kind of Science. Champaign, IL: Strogatz SH (2001) Exploring complex networks. Nature Wolfram media. 410(6825): 268–276.
Journal
Big Data & Society – SAGE
Published: Nov 15, 2018
Keywords: Big Data; data analytics; end of theory; computational social science; data-driven science; epistemology