Paper short abstract:

This paper reports on an attempt to re-count a large number: the approximately 29 million code repositories on the social media platform Github. Science studies might develop ways of re-counting large numbers as capital numbers.

Paper long abstract:

This paper reports on an attempt to re-count a large number, in this case, the approximately 29 million code repositories on the social media platform Github. Science studies and related fields have a decades-long interest in numbers and counting, both in understanding the power-politics of numbering and the diversity of numbering practices. Debates have recently centred on the accountability of numberings and stand against a background of sociological debates about challenges to empirical methods posed by increasingly abundant digital data. Drawing on these accounts of number, the paper will suggest that science studies might develop ways of re-counting large numbers as capital numbers. A capital number combines a sense of importance, the potential to be turned to profit and to name something. In the Github case, the genesis of capital numbers, such as '29 million' or '300 million,' can be re-counted in two different ways. We can analyse the sense-making processes that practically invest in and supplement those numbers with measurements, diagrams and statements. We can also re-count some of the acts of imitation, copying, cloning and duplication comprising mundane Github practice, and show how they give rise to capital numbers. The re-counting of capital numbers directly contests the value-laden measures associated with certain large numbers, showing that capital numbers hide other numericalities. It suggests instead that any ontological or epistemic transformation associated with large numbers -- for instance, in claims about the epistemic potentials of data -- pivots on fragile, weak and diffuse collectives.