Emotion Twenty Questions: Toward a Crowd-Sourced Theory of Emotions Abe Kazemzadeh, Sungbok Lee, Panayiotis Georgiou, Shrikanth Narayanan ⋆ University of Southern California Abstract. This paper introduces a method for developing a socially- constructed theory of emotions that aims to reflect the aggregated judg- ments of ordinary people about emotion terms. Emotion Twenty Ques- tions (EMO20Q) is a dialog-based game that is similar to the familiar Twenty Questions game except that the object of guessing is the name for an emotion, rather than an arbitrary object. The game is implemented as a dyadic computer chat application using the Extensible Messaging and Presence Protocol (XMPP). We describe the idea of a theory that is socially-constructed by design, or crowd-sourced, as opposed to the de facto social construction of theories by the scientific community. This pa- per argues that such a subtle change in paradigm is useful when studying natural usage of emotion words, which can mean different things to dif- ferent people but still contain a shared, socially-defined meaning that can be arrived at through conversational dialogs. The game of EMO20Q pro- vides a framework for demonstrating this shared meaning and, moreover, provides a standardized way for collecting the judgments of ordinary peo- ple. The paper offers preliminary results of EMO20Q pilot experiments, showing that such a game is feasible and that it generates a range of questions that can be used to describe emotions. 1 Introduction There are some enterprises in which a careful disorderliness is the true method. –H. Melville, Moby Dick Science often begins with a good question. One may wonder whether the question a scientist asks is fundamentally different than the question of a non-scientist, or whether they are fundamentally the same. It is usual, in the field of affective computing, for scientists to treat people as experimental subjects when studying emotions. This paper takes a contrary view, where people who would formerly be considered as experimental subjects now function more like theory-generating scientists and the scientists play more of an editorial role. This change of roles ⋆ As we consider the players of Emotion Twenty Questions to be contributors to this theory, we would like to acknowledge the players for their contribution: Stephen Bodnar, Theodora Chaspari, Jimmy Gibson, Jangwon Kim, Michelle Koehn, An- geliki Metalinou, Emily Mower, Elly Setiadi, Kurt Weible, and Mary Weible. We also acknowledge the use of Ejabberd and iJab open source software. This work was supported by NSF and DARPA.