The article provides a number of philosophical insights on how we are anthropologists should engage with archives and the prevailing practices that continue to this day. Drawing from Derrida's Archive Fever, the authors push forward the post modern mission to take on archives, the archivists and the process of archiving itself as objects of study, complicating their understanding as being constitutive of the very 'past' and 'memory' that they seek to preserve. However, present day archiving of social scientific data continues to draw upon the belief of an ontological divide between data and context, instead of viewing data as being reflexively constituted through historical and cultural specific practices. Furthermore, there is even less of an inclination to consider data archiving and/ or the process of data collection, curation and analysis as being more than mere techniques of preserving data and context instead of being constitutive of the very data that they seek to preserve and represent.
Therefore, the need to take into account the performative and constitutive aspect of data archiving and curation practices is vital to the field of archive ethnography. Materials of archives need to be understood within their relation to the archiving practices that materially and discursively constitute them. The article draws upon examples of data cleaning, data anonymization and meta data preparation to explore the conceptual assumptions underpinning these practices of data curation i.e. as being neutral and technical. However, all three practices enact and embody normative conceptual assumptions ( sexual orientation, the personal vs meaningful , etc) that constitute the kind of data that is then preserved, categorized and discarded as being "unlikely", "impossible" or to be retained or discarded.
So therefore, what is required is not to reject categorization of any kind or archival practices in totality ( which is impossible as no archives would then exist) but rather to place them within the framework of analysis in order to engage with the assumptions underpinning them and to deconstruct the process in which they are naturalized and taken as granted. Furthermore, what is vital is to understand archived data as being "made and remade through multiple practices ... practices that are normally treated as neutral and innocent techniques ... (which) renders invisible their constitutive role in the ontological formation of data, records and artifacts" ( pg 166)
The article builds on the work of Derrida and Foucault to argue that data curation practices in social science quantitative research data are not merely neutral techniques but are perfomative as in these practices are constitutive of the very data that is being preserved. Therefore, data curation practices are historically and culturally constitutive practices that are based on specific cultural assumptions which constructs the notion of ontological divide between data and context and the assumption that representational practices such as data curation have no effect on the contruction of the material world or the context in which it is understood. The article argues that the approach towards understanding data curation practices should move away from representational lines towards engaging with them as being constitutive of their very objects of study and the knowledge that is produced. Therefore archival practices, in particular data curation ones, produce their own reality and knowledge, as priviliged topologies and by the process of naturalizaing and erasing the fact that these are privileged topologies.