Abstract
Recent research argues that digital platforms contribute to the long-lasting erosion of genre boundaries and established cultural classifications in the “heads and habits” of art consumers (DiMaggio 1987). This paper draws on large volumes of YouTube data to illustrate how, on the contrary, cultural boundaries are techno-socially reproduced online, shedding light on the understudied role of algorithmic systems in the classification of cultural goods. Through a ground-up investigation of how both automated recommendations and the taste patterns of 202,509 platform users segment 14,865 music videos, this work provides empirical evidence of the persistent relevance of categorical boundaries in platformized cultural reception. The recommender system used by YouTube largely relates music videos of the same genre, thus reinforcing pre-existing artistic classifications in the digital circulation of culture. The study also documents how most users do not cross categorical borders in their interactions with musical content, and that boundary strength varies across music genres. I discuss how the interplay between the social patterning of taste and the algorithmic filtering of cultural content generates techno-social feedback loops, and conclude by drawing broad implications for the sociological study of culture and consumption.