View article

Voice Assistants (VAs) like Amazon Alexa or Google Home assistant are getting more and more common in private homes. They provide various affordances such as controlling smart devices, telling jokes, playing music or even talking about feelings. Specifically interesting is their capability of natural speech. Interaction with VAs is supposed to be very easy and subject to anthropomorphism due to the human-like voice and “character” of the assistant (which, for instance, might tell jokes or express feelings (Purington et al 2017)). As speech interaction also has strong limitations in the current implementations, users nonetheless need to learn how to talk to the assistant in order to receive the desired response. Incorporating them into daily routines and practices requires a certain understanding about limitations as well as strategies for working around those. Studying such sociotechnical aspects of appropriation can be very informative and interesting for better understanding the role and potentials of technologies (Wulf, Rohde, Pipek, Stevens, 2011).

Recent research found that VAs suffer from lack of self-descriptiveness (Pyae and Johnson, 2018), especially regarding their available functions. From the perspective of CSCW, it is interesting to study how users discover and learn to interact with their VAs, how they share their experiences and knowledge with others, and what strategies they use to work around common breakdowns and