【文章全文】Towards a sociology of computational and algorithmic journalism
This article advances a sociological approach to computational journalism. By“computational journalism” the article refers to the increasingly ubiquitous forms of algorithmic, social scientific, and mathematical forms of newswork adopted by many 21st-century newsrooms and touted by many educational institutions as “the future of news.” By “sociological approach,” the article endorses a research model that brackets, at least temporarily, many of the current industry concerns with the practical usability of newsroom analysis. The bulk of the article outlines a series of six lenses through which such an approach to computational journalism might be carried out. Four of these lenses are drawn from Schudson’s classic typology of the sociology of news— economic, political, cultural, and organizational approaches. In addition, the author adds Bordieuean field approaches and technological lenses to the mix. In each instance, the author discusses how particular approaches might need to be modified in order to study computational journalism in the digital age.
Keywords Computational journalism, culture, data, digital humanities, ethnography, journalism, news institutionalism, political economy, sociology of news, technology
Six approaches to a sociology of computational journalism
Michael Schudson’s “Sociology of News Production,” first published in Media, Culture, and Society in 1989 and revised several times since then, remains one of the key citations in any sociological analysis of journalism and news. The article has actually undergone several important mutations over the course of its 20-plus year history, with the key changes being outlined rather frankly in the latest incarnation of the piece (“Four Approaches to the Sociology of News,” published in the equally oft-revised Mass Media and Society [5th edition]). The different descriptions of each of the categories amount to more than simply revisions for the sake of revision, particularly in the latest piece, which signals its intentions by announcing that the “approaches” to studying news have now expanded from three to four. The relevant nuances of these category changes will be explored below. An equally important change in Schudson’s thinking—the argument that journalistic products are as real as they are “socially constructed”—lies outside the scope of this current paper, but would be valuable to analyze in its own right.
What, then, are the different lenses through which we can analyze the sociology of news production, and how can each of them be applied to the study of computational journalism?
Each of the six lenses discussed in this article—political, economic, field, organizational, cultural, and technological—are grounded in decades of serious journalism scholarship and yet can also be adapted to serve the needs of an academic agenda that focuses on a new research object: computational journalism. Analyzing computational journalism through a political lens would focus on public policies that encourage or discourage the adoption of computational processes and practices. An economic perspective would critically examine the stratification of the computational journalism field, and the manner by which differences in institutional resources create patterns of asymmetry in journalism’s ability to operate on behalf of the public. A field approach would expand this economic perspective, broadening the analytical lens to include other institutions in other fields—the field of computer programming and the foundation field, just to name two—and researching how particular field homologies create and recreate different distributions of financial, cultural, and symbolic power inside journalism. The organizational perspective would ground this research in real, daily work practices and routines, analyzing the manner in which potential technological advances succeeded, failed, or were modified due to institutional reticence, bureaucratic red tape, and institutional priorities. A cultural lens on computational journalism would not only examine the way human and material interactions affected the adoption of computational techniques, but also the way these interactions took place against the backdrop of larger symbolic systems and patterns of belief. Finally, a technological approach to studying the intersection of journalism with big data could discuss technology on its own terms, examining values in journalistic design, the hybrid nature of newsroom sense-making, and the changing status of journalistic evidence fostered by the exponential increase in types of digital evidence.
【文章来源】New Media Society 2013 15: 1005 originally published online 10 December 2012