Read on Twitter
Excited to see this edited volume, *Anthropological Data in the Digital Age*, especially chapter by @lindsaypoirier Kim Fortun @profmikefortun & Brandon Costelloe-Kuehn, "Metadata, Digital Infrastructure, and the Data Ideologies of Cultural Anthropology": https://books.google.com/books?id=7pK7DwAAQBAJ&lpg=PA208&ots=xrSuefq4By&lr&pg=PA208#v=onepage&q&f=false
Some quick notes & comments on this chapter, which attempts 1st to compare cultural anthropologists’ & natural scientists’ assumptions about shareable data, & 2nd to push the anthros to rethink (from a perspective within their own epistemic culture) a reluctance to share data.
The driving concern is how anthropology could become more interpretively creative & responsive if data (not just finished writing) were available for re-interpretation. The goal is to enable more creative abductions, not reproducibility: context of discovery, not validation
Comment 1: the comparison, coming out of their participation in the @resdatall, is between what they see as the dominant data ideology of the undifferentiated positivistic sciences, & that of the interpretive humanities / social sciences. Further work needs a deeper dive here.
Reproducibility is a tangled mess across different research domains (see @SabinaLeonelli’s article: https://doi.org/10.1108/S0743-41542018000036B009): binary comparison is a missed opportunity for finding links, analogies, as well as distinctions btw anthropology & its other(s) (& their infrastructures).
Comment 2: The authors don’t use context of discovery vs. validation, but this is a useful distinction that makes sense of the affordances they wish to develop w/ their research infrastructure, PECE (Platform for Experimental Collaborative Ethnography: https://worldpece.org ).
Comment 3: The authors could do much more w/ the data-metadata relation: metadata doesn’t just *provide* context for data, it produces & naturalized the data / context divide (this is more Silverstein & Urban en- & co(n)-textualization than Derrida interability).
[I analyzed metadata from this semiotic-anthropological perspective awhile ago: https://doi.org/10.17615/yzbs-2j92. @npseaver does a nice job of contextualizing the context captured in different kinds of metadata: https://doi.org/10.1177/0163443715594102]
For me, one of the primary creative/abductive affordances of a platform like PECE is that it allows (re)interpreters to ask whether data is better thought as metadata to something else, or if metadata should be data. This gets at the process of anthropological theorizing.
Comment 4: The comparison btw researcher data ideologies suppresses comparisons w/ other knowledge workers’ data ideologies (& practices & infrastructures). I just finished reading Wilf’s Creativity on Demand ( https://www.press.uchicago.edu/ucp/books/book/chicago/C/bo34094110.html), so I’ve got a different comparison in mind.
Wilf is continuing his anthropological inquiry into routinized creativity, in this book looking at business innovation consultants who are also mainly concerned w/ methods of discovery. Through abductions & data reductions/decontextualizations, they are doing comparable work.
A likely worry about the open data anthropology envisioned by the authors is that it gets too close to the fast & loose methods of product innovation: does the method of (re)interpretation built into a PECE project veer too closely to the pragmatic flexibility of Post-It notes?
(For Wilf’s analysis of the Post-It note in business innovation, see https://doi.org/10.1086/688952 . He provides criticisms of the influence of design on some contemporary anthropologists in the book’s conclusion, but a fuller comparison of data ideologies remains to be done.)
Comment 5: while I appreciate that the comparison of data ideologies emerged from their observant participation in the @resdatall, the broad comparison also simplified the ways anthropologists have understood data. Anthros have reinterpreted others’ data before.
What if the authors had pointed to Gluckman’s seminars at Manchester, where ethnographers presented their field notes & analysis, & then the floor was open for reanalysis? What about historical anthros reinterpreting documents from the archive? What does digital/open add to this?
Or perhaps the comparison could be with sociological ethnographers, who have developed methods of sharing field notes as a validation defense. Their more robust methods literature has much to say on this. Convincing anthros to think about data reuse could start closer to home.
Comment 6: I think the authors need to do more than describe & compare data ideologies: we needed is some explanation of why ideologies cohere in their domains. I’ve been influenced by Jenny Fry’s domain-analytic research into information practices: https://doi.org/10.1016/j.ipm.2004.09.004
Using Richard Whitley’s categories, Fry shows how degrees of mutual dependence & task uncertainty w/in a research domain ties directly to the development of domain-specific info practices & research infrastructure. Anthro has low dependence & high uncertainty, thus little sharing
If something like PECE is going to work to promote data sharing, it’s because it is coupled w/ increased mutual dependencies in collaborative research projects, & also smuggled in just enough task certainty in anthro’s (mostly unspoken) working methods & interpretive strategies.
