Core Post #3 - Mahnoor

As others have already noted, one topic addressed by this week’s readings is the notion that data collection and classification rely on algorithms and assumptions that reinforce and “repeat the most stereotypical and restricted social patterns, re-inscribing a normative vision of the human past and projecting it into the human future” (Crawford XIX). Wendy Chun also discussed the ways in which “Big Data” seeks to “repeat a highly selective and discriminatory past,” illustrating how data was used to “perpetuate angry microidentities” (49) during the 2016 election campaigns. This observation in particular provoked me to contemplate a dizzying number of questions. Apologies in advance for dragging you into my rabbit hole: Did the manipulative use of data shape these microidentities, or did they always exist in their current form, just in a different scale or in a kind of stasis, slumbering below the surface of civil discourse? That is to say, would hateful microidentities bide their time in private spaces and eventually erupt into mainstream politics without the aid of Big Data? Or would they gradually soften into extinction, hate expiring with a whimper instead of exploding with a bang? Are AI technologies, exploitative algorithms, and practices of data manipulation the authors or accelerants of political extremity? And does this distinction matter? 

Lippold too addresses the power of AI and data to create categories. In one example, Lippold described the experience of a friend who was categorized by Google as an ‘old’ ‘man’ despite being a young woman; as a scientific researcher, her online reading habits aligned with those of others in her field, namely old men. Lippold goes on to observe that because Google is only interested in users as consumers, if a user navigates the web like an old man does, then Google will treat that user according to the corresponding measurable type (106). Even if Lippold’s friend’s professional field gradually becomes more inclusive, the measurable type may not reflect the demographic reality, but rather the field’s historical inequities. In fact, in the digital realm, because these types are “reflecting” a false reality, they are in essence creating, or re-creating, a completely different one. I found this element of creation and its link to control compelling and unnerving: “measurable types [...] actively control; they modulate themselves to best fit the population they’re meant to describe” (Lippold 144). This sentiment reminded me of Edward Said’s Orientalism. The creation of measurable types in order to categorize individuals and groups echoes the ways in which narratives spun and histories written by Western writers and scholars about the Eastern “Other” enabled tangible Western domination in the pre-colonial and colonial period. However, I hesitate to take this comparison too far for reasons that I hope will become clear in the next paragraph.

I’d like to pivot now to the Couldry and Mejias reading. Like Chun, they look to past frameworks of exploitation to think about the intervention of data in our lives now. However, right after reading Couldry and Mejias’s piece about data colonialism, I felt that something rubbed me the wrong way, though I could not pinpoint exactly what it was. I agreed with some of the parallels they drew between colonialism and what they termed data colonialism, but upon reflection, I may have hesitated to fully accept their argument because of the scale and directionality they assigned to data colonialism: “The new data colonialism works both externally—on a global scale—and internally on its own home populations. The elites of data colonialism (think of Facebook) benefit from colonization in both dimensions, and North–South, East–West divisions no longer matter in the same way.” While these division-traversing scales may apply to the power structures responsible for the collection and control of data, more traditional modes of domination, expressed in the present as neo-colonial and neo-imperial modes, though at this point certainly imbricated in the weaponization of data, are still inextricably linked to cultural hierarchies. That is to say, colonialism, in a non-digital form, is still ongoing in some ways, and the “North–South, East–West divisions” do still matter. A specific point in the Couldry-Mejias text may further illustrate my desire to push back on their definitional argument, that is their desire to label the way digital power structures are operating as colonialism: 

“Consider the Spanish empire’s Requerimiento [...] The purpose of this proclamation, read in Spanish by conquistadors to a non-Spanish speaking audience, was to introduce the natives to the strange new world order they were about to be colonized under, and to demand their simple acceptance (or face extermination, which frequently arrived regardless of compliance). Today, in the era of data colonialism, we are accustomed to similarly incomprehensible documents called Terms of Service, which contain outlandish appropriative claims by corporations. The force of the Requerimiento depended on an effective monopoly of physical force. Today’s ‘muscle’ lies in various forms of economic concentration, one of which is the digital platform.”

I find the parallels drawn between the Spanish exploitation and extermination of indigenous people through official proclamation – here, a speech act described as a near-simultaneous announcement and enforcement of brutal policy – and the way in which users like me are constantly duped online to agree to conditions we do not understand to be quite a stretch. Similarly, though I agree that economic abuse can be and is often a form of violence, I still find it hard to accept that the force behind the Requerimiento as described can be equated to the force of a digital platform.