About the Interview
When I decided to conduct an interview with media artist and theorist Brian House, I realized that the interview itself, as a genre, is a form of surveillance, albeit a highly naturalized one. An interview depends upon an interviewer's ability to not only ask and listen but also to collect and store information—in other words, to mine a person for something of value. But an interview is also a record of a particular interaction between two people, occurring at a particular time and place. And I wanted to play around with that idea.
To do so, I have mined multiple forms of "data exhaust" (House, 2013, para. 1) from the interview event itself. And I have presented these forms of data simultaneously, such that an audience can read the entire interview from beginning to end in multiple ways:
1) Video
A fly on the wall video captures the interviewee but not the interviewer, a method of data collection common in homemade, single-camera interviews.
2) Audio
A stream of audio data scrolls across the screen, enabling viewers to both hear the interview and see its changing volume through sound wave visualization.
3) Head-mounted point-of-view video, mimicking what each participant sees
Shaky-cam footage alludes to the ubiquity of POV perspective videos on platforms like YouTube and Google Maps while also bringing to mind the privacy and surveillance issues that have been raised by the use of technologies such as police body cameras and Google Glass.
4) Audio, mimicking what each participant hears
The participants' bodies are both microphoned. Through audio visualizations, each of these recordings is translated into a visual language that attempts to depict the rhythms of the conversation—its ebbs and flows—so that a viewer can hear what the conversation sounds like from each participant's body. The interviewer's audio is on the left side of the screen and the interviewee's audio is on the right side of the screen.
5) Keywords
The words most commonly used during the interview float across the screen when they are spoken (excluding stop words such as "a," "the," and "like"). This data exhaust alludes to natural language processing techniques used to mine data for marketing purposes while also highlighting the key concepts discussed in the interview.
6) Text
Transcription is a traditional means of presenting data mined from an interview event.
The following interview aims to capture, in some incomplete way, the particularities that result from the interplay of two personalities and bodies. By mining the experiences of two people—the vocal, bodily, gestural, and semiotic particularity of a singular social exchange—the interview creates a database in miniature, transforming the contingent particularities of everyday life into data that can be experienced in various forms, each with its own rhetorical implications. By presenting this data in multiple forms, the interview aims to spur reflection about the cultural logics underlying data collection and to encourage viewers to see data as inevitably rhetorical. In other words, we might ponder whether these streams of data convey the "relational and embodied and unstable" (Smith, 2013, para. 3) nature of my conversation with Brian House or whether they are nothing more than degraded representations of it. And we might, as House's work encourages us to do, be more mindful of the ways that we constantly "negotiate between algorithms and the rhythms of everyday life" (House).