To what extent are data like or not like oil? It’s a common
metaphor, one that in the corporate world points to opportunities for profit,
while in the realm of political economic/Marxist critique (cf. Srinicek from
last week), it points to a new stage, relying on new forms of extraction and
accumulation, in capitalism’s speculative booms and inevitable crises. This
week’s readings help us take apart this metaphor, exposing the ways in which
data are always produced, cultured, and political – and perhaps shedding light
on the ways in which nature/culture binaries are reproduced in discourses of
algorithmic systems.
Couldry and Mejias do this explicitly, criticizing the ways
in which the language of oil naturalizes the processes of data colonialism by
suggesting that data exist for the taking: “This rests on the construction of
data as a ‘raw material’ with natural value” (p. 340). Their highlighting of
the language of “raw material” as applied to data brings to mind Gitelman’s
work in “Raw Data” Is an Oxymoron (2013), in which she critiques the
notion that data simply represent the real world and argues that data are,
rather, always culturally examined (here, she pulls from anthropological
notions of the raw and the cooked – the nature/culture binary in other words).
Cheney-Lippold similarly points to the ways in which data are
not “natural”: “Much like the social constructions of gender, race, sexuality,
and terrorist, the datafied world is not lying in wait to be discovered. Rather,
it’s epistemologically fabricated” (p. 45). Data collection and classification are
not processes applied to some kind of data “out there”; rather, they are how
data are produced. Crawford and Joler, in their discussion of extractivism in
algorithmic systems like the Echo, identify a tripartite structure: material
resources, labor, and data. In their mapping of the “anatomy” of an AI system,
they foreground the ways in which data are produced through various kinds of exploited
labor, including that done by supposed consumers through their online behavior,
as well as the work of Mechanical Turk pieceworkers and other low-skill,
low-paid laborers charged with annotating data to make it usable by machines.
Chun, meanwhile, tells the history of correlation, as it links
late 19th-early 20th century eugenics to the data
analytics and Big Data of the present moment. This, too, is in a way the story of
the relationship between nature and culture/the social as it is mediated
through the construction of data. Eugenics has its roots in an extension of
Darwin’s evolutionary theory, according to which humans are biological species –
like other animals – to whom the laws of natural and artificial selection apply.
Galton, and later Pearson, Fisher, and others took this as license to apply
these laws to the realm of the social (or, I guess, collapse the social into
the biological?) in their political projects to “improve the race” (the national
population). The production of data about social groups (and this was very much
the period of scientific racism, in which social groups were taken as
biological categories) was integral to these efforts.
This is a really interesting post. Oddly enough, your comment about the tripartite structure of material resources, labor, and data in the Crawford and Joler reading made me think about the film Nomadland. As Crawford and Joler discuss, Amazon not only closely surveils and regulates its workers, but it also owns patents for numerous devices that ensure “worker alienation” (XV). One of the more common criticisms of Nomadland is that it depicts the protagonist, Fern, working seasonally at Amazon without providing any real commentary on the company’s exploitative practices. Nevertheless, I find it striking that the film clearly wants to juxtapose Fern’s work at Amazon with her time as a nomad. Both conditions are fundamentally precarious, but the film’s depiction of Fern’s time as a nomad seems to imply that at least here she can find some semblance of freedom or peace of mind. Significantly, her nomadic journey is always associated with gorgeously shot scenery. By becoming a nomad, Fern turns to “nature” as a means of “escaping” the world of machines, technology, and data represented by Amazon.
ReplyDeleteNumerous scholars and critics have described Fern as akin to a contemporary cowboy and have noted the way that Nomadland draws narratively and aesthetically from westerns. And, just as westerns historically sought to justify the erasure of Indigenous peoples, Nomadland never acknowledges how Fern’s whiteness positions her in relation to the land she travels. She finds spiritual solace in “nature” without ever acknowledging she is occupying Indigenous land. In other words, Fern leaves a digitally mediated form of exploitation (Amazon) only to imbricate herself even more actively in another form of exploitation (settler-colonialism). To be frank, I’m not entirely sure what Nomadland has to say about the relationship between technology, nature, labor, and culture (or even what it thinks it has to say). But, as your post also addresses, it is telling that the logics of colonialism, capitalism, and white supremacy continue to emerge in an almost cyclical fashion as the film tries to navigate the aforementioned tripartite structure.