CCA Conference 2016 – I/O: Enforcing Idealized Newsmaking Practices Through Algorithm

During the Canadian Communication Association (CCA) Conference 2017 (Toronto, Canada), I presented the paper “ I/O: Enforcing Idealized Newsmaking Practices Through Algorithm”. Below you will find the abstract and my slides.


For more than two centuries the production of news has been dominated by mass media industry through processes of selecting, reporting, and distributing information. These operations are conducted by journalists through a generally agreed-upon collection of rules and protocols through which the goal is to represent reality and inform the public. Recent developments in digital technologies have directly impacted the way information, particularly the news, is produced and consumed. Blogs, social and mobile media, and search engines unlock and ‘democratize’ the means of news production and distribution, enabling virtually anyone to become a “journalist.” As a consequence, newspapers have had their role as gatekeepers questioned at the same time that the very concept of news has been challenged.

More recently, the media industry began to integrate computer algorithms in its practice as a strategy to reclaim its role as society’s gatekeeper. The adoption of new practices and the introduction of new actors (humans and non-humans) has an impact on the news form that cannot be ignored by scholars and practitioners. Considering this paradigm, this paper proposes that algorithms might now be understood as a new encoder/decoder of society, becoming a gatekeeper ‘organism’ that regulates flows of information as well as interprets and negotiates both public interests and the ‘value of the news.’ I discuss two interrelated modalities of algorithmic news: (1) economically efficient production, where news outlets utilize quantitative metrics, such as A/B testing, and automation to improve effectiveness and desirability of content; and (2) shared-gatekeeping, where visibility and distribution of information is based on user behaviour, location, and other non-disclosed attributes, as in the case of social media networks.

The assumption is that these modalities aim for more objective and efficient ways to deliver content to the public using a user-centric model: the news becomes contextualized in relation to the reader. However, the algorithmic news goes beyond its assumed role of aiding in media production and consumption: they serve as instruments of institutional power at the same time that they are shaped by the already-defined rules and protocols of the institutions that use them. The traditional values connected to journalism practices (informing the public with the most impartial view of the fact) were never disconnected from the more commercial aspects and market values (oriented around competitive and revenue-generating concerns). Now, a new set of values, attached to science and technology, are being incorporated into news practices, and already began to dominate conceptualizations of the “quality” of the information being produced and disseminated. These specific social values and norms are reflected in the design of algorithms, and, as a consequence, become central to how we represent reality and construct knowledge.