Soft City Sensing: A turn to computational humanities in data-driven urbanism

Abstract

Data-driven urbanism is often entangled with the smart city and practiced in a way that prioritizes control over physical objects and downplays the human and political aspects of data. We label this approach ‘hard city sensing’ (HCS) and we argue that the rise of the ‘digital city’ offers the empirical foundation for more humanistic approaches. Driven by the ambition to untangle data-driven urbanism from HCS, this paper reviews two decades of scholarship that has used digital traces as an empirical ground for understanding urban phenomena. The review identifies four distinct ways of working with digital traces of which three pave the way for new ways of problematizing the city. Instead of abandoning the idea of data-driven urbanism, we propose the framework of 'soft city sensing' (SCS) as way to re-engage with it with inspiration from these pioneering works. However, this requires a willingness to revisit central epistemological commitments that currently serve as standards for how to “properly” do data projects. We therefore urge qualitative urban scholars to ponder the possibilities of furthering their urban interest by ‘thinking with algorithms’ while retaining their interpretative ambitions just as we identify a need for urban decion-makers to expand their criteria for what serves as valid data inputs to urban planning.