Materializing Text Analytical Experiences:
the Bubblelines Hydra Project

There is a tradition in the digital humanities of interpreting texts through various forms of computer-generated text analysis and visualization. For example, the Text Analysis Portal for Research (TAPoR) and Voyant.org both provide online access to a broad-range of tools for work with a variety of data. However, these systems require a significant learning curve, both for students and senior humanities scholars. In addition, the standard computer screen and its interfaces can conceal the vastness and complexity of the material under study.

In this project, we experimented with creating new forms of interface that leverage physicality and kinesthetic intelligence (Ruecker, forthcoming). There were three primary objectives: (1) to provide a richer and more accessible introduction to the algorithmic study of texts; (2) to create new departure points for discussions of the value added, to all participating disciplines, in collaborations between the digital humanities and the fine arts; and (3) to demonstrate the depth and vastness of the digital artifacts that have been created as part of digital humanities scholarship.

Our proof of concept is the materialization of the Bubblelines interface. Bubblelines is an approach to the visualization of comparative search results across multiple documents (Giacometti et al. 2009). The concept, which originated with Carlos Fiorentino, has subsequently been implemented in two different ways: first, as a stand-alone web application and, second, as a component of Voyant.

For our physical substantiation, we originally considered automatically inflating helium balloons on various lengths of string. Another possibility, that also relies on gravity, is to drop a series of appropriately-sized balls on rods running from the ceiling to the floor, where each rod represents a different text, and a series of spacers are counted out to space the balls according to the distribution of search hits. Our third idea was to leverage the idea of bubbles. Returning to the horizontal form of display, we envisioned a course of parallel tracks, each filled with a viscous liquid, that could be interrupted by bubbles inserted through a series of valves controlled by Arduino.

Our team consisted of four undergraduate students in information design, 3D animation, and music, two colleagues from marketing and innovation and one from design, and three volunteer collaborators out of industry. We used the format of a week-long hack-fest, with two full days on site, examining the design problem, sharing our positions of expertise, evaluating our resources, and determining the design direction. Over the subsequent four days, we used smaller sub-teams who took the lead on different parts of the build. Everyone had an equal voice in all aspects of the project through an intentional subversion of the for-hire model of design and development. This hack-fest structure emerged out of my experience working with teams of designers, developers, and content experts since 2005. We chose an interesting, neutral location (in this case, TORCH Motorcycles), and we fed everyone, on and off site, throughout the fest. We presented The Milking Machine at the 2016 Congress of the Humanities, CSDH/SCHN Conference.