What is Data?
Date
2023-04-14
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Philosophers of science researching data-intensive scientific practices have largely converged on the idea that data are relational artifacts, where data are defined through their relations to scientific practices. I use the relationality of data as a starting point to construct a schema that highlights three relata of data. My aim is to foreground relata that I believe have largely been pushed to the background in the current literature. My schema is: Communities use technologies to create data for a purpose. Beginning with a survey of philosophical literature, I show the development of data as a relational artifact. I first present the accounts of Patrick Suppes and of James Bogen and James Woodward. The relevant works of these philosophers focus on how data relates to theories, expanding it to proposing that data provides evidence toward claims for phenomena, which in turn provide evidence for claims of theories. Then, I present the relevant works of Ian Hacking and Sabina Leonelli. Here, data are considered in relation to laboratory practices and the content and form of data are investigated in more detail. I move on to constructing my schema of data. I argue that philosophers ought to foreground communities, technologies, and purposes when analyzing data practices. I do so by analyzing a historical case of thermometry before inspecting communities, technologies, and purposes in greater detail. Communities influence data practices through their complex interactions, done between individuals, groups, and groups of groups. Choices in technology affect the pace of data practices since tools directly influence the content of data that are created. Furthermore, technological choices are determined through community interactions, and often influenced by practical limitations of financial and resource limitations of a lab. Lastly, much of philosophical literature has focused on data's use as evidence, but I argue that there are multiple uses for data. In particular, data may be used as representation without abandoning relationality, and that data may be used to train algorithms. Evidence, representation, and training are distinctly important uses of data. My schema may be used a starting point for further philosophical investigation into data practices.
Description
Keywords
data, technology, philosophy of science
Citation
Chattoraj, A. (2023). What is data? (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.