1 the mode of information enacts a radical reconfiguration of language, one which constitutes subjects outside the pattern of the rational, autonomous individual. This familiar modern subject is displaced by the mode of information in favour of one that is multiplied, disseminated and decentred, continuously interpellated as an unstable identity. At the level of culture, this instability poses both dangers and challenges which, if they become part of a political movement, or are connected with the politics of feminism, ethnic/racial minorities, gay and lesbian positions, may lead to a fundamental challenge to modern social institutions and structures.”
Poster, M. 1995. The Second Media Age. Cambridge, England: Polity Press, p. 57
2 Thus, research on polarization and the Internet strongly suggests that the Internet facilitates homophily among at least one important subgroup of the US population—those who are politically aware and have strong partisan and ideological leanings. It does not, however, suggest that this leads to a breakdown of relations between those with different partisan identities. It is, furthermore, hard to know whether or not heightened clustering leads to political extremism. Among the broader population, the evidence seems to suggest that the Internet is associated with more exposure to alternative views rather than less. Although increased polarization and intragroup contact may lead to higher levels of collective action within each group, other sources of variation may be more important.”
Farrell, Henry. 2012. The Consequences of the Internet for Politics. Annual Review of Political Science15(1):35–52. DOI: https://doi.org/10.1146/annurev-polisci-030810-110815, p. 43
3 However, others argue that democratic public life benefits from a richness of voices and viewpoints, and that the journalism landscape should reflect social, organizational and epistemic diversity (Anderson, 2016; Benson, 2013; McQuail, 1992; Voakes, Kapfer, Kurpius, & Shanoyeon Chern, 1996). Should one hope, then, that algorithmic news distribution generates homogeneity, or at least a degree of similarity, in news consumption and agenda—or is the public best served if algorithms highlight varied information from a large number of sources that offer a potpourri of perspectives? Empirical research cannot truly answer these questions, as there is no agreed-upon standard for an “optimal” degree of personalization or pluralism in news.”
Nechushtai, Efrat, und Seth C Lewis. 2019. What kind of news gatekeepers do we want machines tobe? Filter bubbles, fragmentation, and the normative dimensions of algorithmic recommendations. Computers in Human Behavior 90:298–307. DOI: https://doi.org/10.1016/j.chb.2018.07.043, p. 303-4.