The topic of study of Human-Computer Interaction (HCI) is constantly changing as people develop new uses for new technologies. In this thesis I present three contributions to the field of HCI that address these ongoing changes. These contributions are around the themes of epistemology, experience, and evaluation.

I begin by importing from the field of Science & Technology Studies (STS) the notions of epistemology (the study of how we know what we know) and comparative epistemology (the comparison of different ways of knowing). STS researchers work from an intellectual position outside their field of study; I propose "epistemological reflection" as a way for HCI researchers to engage with questions of knowledge and validation while remaining within the field. I argue that epistemological orientations have impact throughout the research process, and that HCI currently lacks the vocabulary to discuss intellectual clashes on an epistemological level.

My second contribution is a study of the term "experience" in HCI, a discussion of its meanings in the field, and the identification of an emerging sub-field I call experience-focused HCI. Experience-focused HCI aims to design for the multiple, complex and situated experiences people have with technologies. This is not simply a shift from researching "tasks" to researching "experiences". Rather, it treats experiences as situated interactions formed in the course of a specific interaction, and recognizes any representation of an experience is inherently incomplete. Experience-focused HCI also implies engagement with themes of affect, aesthetics, the body, human practices, and the role of the artifact in knowledge production throughout an open-ended research process.

My third contribution is a set of methods for the evaluation of experience-focused HCI, based on a discussion of the epistemological foundations of evaluation: a successful evaluation not only validates the technology in question but also the topic of study and methods used to study it. Due to the open-ended nature of experiences, evaluation must shift from defining a priori metrics which can then be tested in laboratory situations to developing situated metrics through user experiences "on the ground".