AO: This orals document is framed around Fortun’s conceptualization of discursive risks and gaps (Fortun 2012). Fortun writes that the contemporary Late Industrial period is characterized by complex conditions for which there is no available idiom, no way of thinking that can grasp what is at hand. Fortun calls these “discursive gaps” and its opposite -- “discursive risks” -- emerge because of a tendency to rely on established idioms and ways of thinking nonetheless. Discursive risks or the established ways in which a particular phenomenon is discussed again and again can set ethnographers up to miss key aspects of the dynamics. Therefore, we borrowed a structure of ten analytic questions initially designed by Fortun for a project on air pollution in cities (unpublished manuscript) to develop ten analytic questions to help unsettle and unpack the nested problem space of tech development research data sharing in Nairobi in order to study its discursive gaps and risks.
The questions are employed to help tease out different scales of a phenomenon and to guide us in identifying discursive risks and gaps in approaches to an object of study. James and I (AO) have employed a modified version of these questions to query research collaborations.
On Developing Analytics:
AO: James and I had experience designing analytic questions with Kim and other UCI graduate students as part of the STS Across Borders initiative which we worked on from January - August 2018. As part of this process, James and I became familiar with PECE, PECE vocabulary and the style of analytic structures. In July 2018, James and I used this knowledge to draft a modified version of the analytic questions which we envisioned would help structure our joint querying. We reviewed this first draft of questions with Kim and based on her feedback, revised the questions to more closely follow the different levels of “sedimented questions” to help us address and identify the discursive risks of the work we would be querying. This first “final” list of questions was inputted into PECE by early August and then used to annotate. Once I began to read and annotate, new analytic questions continued to emerge (for example, this question to more explicitly capture any binaries and metaphors being used). When new questions arose, I added them (after touching base with James first). I did not delete any questions since the changes in questions themselves are an interesting and noteworthy part of the process.
On Identifying and Categorizing Artifacts:
AO: For this orals document, we leveraged both of the two distinct approaches used in my other two orals documents into one here, that is, to have a set of essays (the "querying" series) focused on the artifacts themselves as well as an accompanying set (the "analyzing" series) focused on our annotations and musings. We first identified key works that we wanted to include in our broader literature (see below for more details). Then, James and I created the artifacts and inputted them into the appropriate "querying" essay (which spans across the research life cycle). For those pieces that did not deal directly with one part of the cycle, we tried to file them as best we could, recognizing the limitations of the categorization strategy used. While annotating in more detail, it became clear that some of the works should be "filed" elsewhere so we moved the artifact appropriately. The intent behind our "research life cycle" heuristic was to help organize our thinking and identify where "collaboration" has been thought of the most within the context of research work.
Through discussion with each other and our shared advisor, we determined that it would be fruitful to try to extend the analytic structure to the wide range of disciplines publishing on “collaboration” across diverse discursive fields. Therefore, once we had a first draft of our bibliography, we leveraged the Google scholar search engine and library resources to identify works outside of our own discipline that we could apply our analytic structure to. We found that while there are several fields working on collaboration, there are significantly fewer thinking about collaboration explicitly in the context of research. After this online search, we added additional works to our bibliography from the following fields:
We focused especially on looking for work on collaboration from the disciplines of those we expect to encounter in our field sites (e.g. development economists, computer scientists, political scientists, policy makers, scholars of public administration, etc.).
In our keyword searches we used phrases and terms such as “economics + participatory methods,” “collaboration + computer science + users,” “big data + collaboration,” “crowdsourcing data,” to name a few.
Within each of the essays (one per aspect of the research life cycle), James and Angela each chose at least 2 per life cycle stage. We selected the artifacts to annotate based on the disciplines and topics we expected to most encounter in their field sites. For Angela, that was development-oriented fields and computer science/ informatics. For James, that was environmental and business / policy oriented fields.
In summary: James and I compiled a “final” bibliography (of over 90 references) and inputted them all in a Zotero folder. We put all references and analytic question probes into a google document. All of the references were inputted into PECE as artifacts. Find the style guide here. To upload them as artifacts, we both found the articles/books, reviewed the abstract and skimmed the piece. We wrote a critical commentary or the abstract and input the various meta data fields. Once all were inputted, we each then chose approximately 20 - 25 from the set to query in more depth. These were the ones that we each annotated using the shared structured analytic set of questions. We foregrounded the artifacts in the "Querying" essays and foregrounded our notes and analyses in the "Analyzing" essays. We used the annotations as well as the critical commentaries to draft the extended narratives which are our current understandings of the discursive gaps and risks of scholarship on collaboration across the research life cycle.
AO: After completing annotations on the subset of artifacts, I worked through each phase of the research life cycle beginning with "Research Design." To do this, rather than clicking the "analytic questions," I clicked the tag so that I could see all works that had been entered under that phase of the research cycle. I looked at all annotations I had done under each of the analytic questions, especially focusing on responses to "discursive risks" in order to develop a brief "abstract" for each life cycle that described briefly the patterns and trends. I used these abstracts per each phase as well as insights I jotted throughout the process to detail an "extended narrative" for "collaboration" across the research life cycle. In this way, not only did I leverage the analytic questions to look at collaboration but I also used the heuristic device of organizing the works across the research life cycle as a way to see what parts of "collaboration" analysts have looked at most frequently.