Big Observation—Comparing Population Statistics and Recommender Systems as Modern Formats of Observation

Abstract

The article situates digital technologies in the history of modern formats of observation. The term “format of observation” helps to relate devices and systems that at first sight do not have much in common: official statistics, opinion and market research, monitoring instruments, and digital technologies such as search engines, social networks, and recommender systems. Despite their differences, they share one important characteristic: By collecting huge amounts of data and by systematically looking for patterns in these data, they are striking examples of the rise of quantification since the 18th century. This article draws on two examples—population statistics as the historically first modern format of observation and recommender systems as a prominent example of digital observation technologies—and compares them with regard to their technique of observation. How do statisticians produce the numbers that are represented in statistical tables? And how do recommender systems calculate personalized recommendations? Comparing these two cases not only sheds light on how official statistics and digital technologies observe but also provides insight into the practices and premises of social observation over the last two hundred years.