Audio Narrative Project: Phone Survey Saga
by Issel Lim (MIT '05) [] & Jane Wu (MIT Grad. Student) []

Try the Project

Overall Concept

We began with the idea of communication, and focused on ways in which users can hear various people and their methods of communicating. Our project then evolved to involve one of the most frequently-used communication tools -- the telephone.

We came up with 5 questions:

  1. Who are you?
  2. Where are you now?
  3. Where are you most likely to meet people?
  4. What would be a random thing you'd say to a person beyond hello?
  5. Who was the last person you talked to?
  6. When was the last time you talked to someone you love?
  7. Why do you think we're doing this?


After devising the questions, we called up three groups of people: random numbers found by flipping through the phone book, friends, and the 21W.765 class. Their responses to these questions were recorded through speakerphone via digital video camera. The sound clips were then transferred onto a computer with a firewire card, edited with Premiere Effects to splice the files into separate answers and soundbytes, and converted from .WAV files to an .mp3 format.

After chopping up these linear conversations, we organized them in a database format by separating them into seven major categories, one for each of the seven questions. We further divided each category into 3 smaller categories, which represented the different groups of people we had called. We then created a Flash interface that would enable the user to interact with these sound clips in an investigative fashion. We first compiled the clips by using a random living room's objects as focal points for the sound. However, this beta version was too choppy, and the photograph had no real correlation to our creative process or our conversations. The sound clips were jumbled and didn't match the objects around the living room, which was too small overall.

Therefore, we decided to make it personal. After playing around with Adobe Photoshop, we created a new picture as a background for the project.

The image displays a unisex, though somewhat feminine, person, talking on the phone with her hand on the phone book. Her left shoe compiles the data for the question “Where do you meet the most people?” and her right shoe contains “Where are you now?” soundbytes. Her belt is a conglomeration of our miscellaneous remarks or sounds while conversing. The heart on her shirt displays “Who are you?” The thought bubble branching off consists of “Why do you think we're doing this?” while the highlight on her head plays several of our explanations. Her conversation bubble asks, “What would be the first thing you'd say to a random person beyond ‘hello'?” and the phone's conversation bubble questions, “Who was the last person you talked to?” Her watch compiles data for “When was the last time you talked to someone you loved?” We put the questions in big question marks, and the answers in small question marks. Random phrases from our conversation partners are listed in the yellow phonebook. Greetings and endings are arranged on the blue design below. The phone on the table contains miscellaneous telephonic noises, and the answering machine plays snippets from various samples.

The first version of the new image had labels detailing the questions for each mark. However, this had an overall messy and busy appearance, with too much labeling. There was no room for discovery, no sense of exploration. The users would already know and guess the responses behind each part of the drawing.

Therefore, the final image contains no word labels, so the user has to click on the questions to find out what they are. Similarly, we began each conversation by saying, “Hello, we're MIT students doing a project. Could you please help us?” and many of the people with whom we talked had no idea what we were talking about. Many of them would also never have thought about the answers to the questions (“When was the last time I talked to someone I loved?”) unless asked outright. We asked them to focus on and share a tidbit of their lives.

The simplicity of the interface, a seemingly childish stick figure, reflects on the innocence of the questions asked. These weren't deep questions; they were quick, and everyone had an answer for them. However, due to the layers and the possibilities of combination, the figure is interwoven with various question marks, accessories, and stop buttons. Due to the style of flash animation, the timing and customized conglomeration are different for each user , creating an unending play of words.

Envisioned User Experience

While observing outsiders interact with this piece, we saw an initial fascination with pressing as many buttons as possible. This usually results in a cacophony of sound. After the users grow accustomed to the digital aspect of the flash, they usually press buttons at regular intervals, which develops into a rich overlapping of sound. Often, one can hear one set of answers through the gaps in another. As in our first project, where we broke down language into key components, we divided these conversations into simplistic snippets. Users compile their own personal narratives, mentally grouping each answer into matching voices and overall personalities. Juxtaposition, synchronization, and organization contribute to various associative aspects of an overall narrative that's sometimes pleasing, sometimes divergent, but always interesting.

Questions of Narrative Addressed

When compiling the narrative, we wanted to show our own journey through the phone survey world. Hence, the girl on the phone with her finger leafing through the phone book. However, we added to this skeleton by breaking down people's narratives, then combining and enriching them with other people's answers. The narratives proceed categorically, rather than chronologically. The format allows the users to hear overlaying snippets and to combine these communications by associating juxtaposed snippets.

Our project is clearly divided into a database format, which illustrates Lev Manovich's essay. By studying how people interact with the project, we can determine whether the database format necessarily forms into a narrative. Through our own interactions with the system, it appeared that the project became more effective when multiple parts are played at the same time. Different answers would interweave with one another, providing a spectrum of our group as a whole. In essence, the user takes apart the database components, and assembles them in the most appealing manner at the time.

We have mixed questions answers in a blend of personalities that utilize different tones to provide categorical narratives. We offer the users building blocks. The formation of the narratives come only when the users interact with our system, building layers of answers that ultimately tell us who, what, when, where, and why. A phone survey saga.