This paper looked at the implications of the third version of the Variable Media Questionnaire, an online database designed to help understand how an artwork may be best preserved when its initial form decays or expires. VMQ 3.0 introduced a shift from the previous two versions by viewing the fundamental unit of an artwork as not the artwork itself, but the pieces that go into its construction. In an attempt to capture as many impressions of the work as possible, questions about those components are posed to not just the artist whose name is on the wall next to the piece but also the curators, conservators, assistants, and even viewers who have experienced the work. But the modern reality of art is that it is not enough to treat an artwork as just a collection of physical parts, and so the VMQ also recognizes environments, user interactions, motivating ideas, and external references as aspects to be surveyed and considered when preserving or recreating the piece. Expanding the scope of the data collected by the questionnaire changes the focus of preservation from an artwork’s materials to its experiential characteristics, potentially exposing perspectives on the works it describes that had not been previously considered and pushing the VMQ into a new role as an epistemological prism through which art can be viewed.
Notes from the presentation, part of the “Still Accessible? Rethinking the Preservation of Media Art” panel, are below. Proceedings are available on the ISEA2010 site.
The Third Generation Variable Media Questionnaire, v1.1
June 18th, 2011
To start, I’d like to briefly reintroduce the second version of the Variable Media Questionnaire. The VMQ is a framework for understanding how to preserve works that are subject to decay and obsolescence.* It begins with the assumption that material works–remembering that digital code is ultimately stored on material somewhere–have a finite life span. So, while efforts to preserve an original work are useful, ultimately the question is going to become what should happen once to the work once that is no longer viable. That’s obviously a loaded question, and subject to many different answers from any number of people who are associated with the creation and display of the work. What the VMQ does is provide a model to begin asking the questions, and ultimately to record the answers to those questions when they’re posed to the various stakeholders involved.
In the second version, the VMQ’s framework was built around medium-independent behaviors and strategies. You can see the behavior categories in the tabs at the top of the screen. Each behavior that can be attributed to a work has a number of questions associated with it that describe the work as it exists, either now or in an “ideal” state. Once those questions have been answers it suggests another set of questions that are based on preservation strategies that are appropriate for the work given its behaviors. I’d like to show you a bit more about how it worked, but unfortunately the VMQ2 itself has fallen victim to what it’s trying to work around: it was made in a version of Filemaker Pro that no longer runs on my machine, so I’m limited to showing you documentation of the work instead of the work itself.
While the VMQ2 model is still useful, what we’ve found is that there are also some limitations. The biggest one is that there is limited flexibility in the model, which means that there is a limit to the specificity of the questions it suggests. Since each behavior has a fixed set of questions and options, it can be difficult to apply VMQ2 to a work that is outside the boundaries of its original conception. Even though the goal is to find medium-independent qualities in a work, each work is still going to be individual and will inevitably either have some pieces that don’t have relevant questions in the model or have questions in the model that don’t apply to the work. In addition, the implementation of the VMQ2 makes it difficult to answer questions about similar parts of a work that may require different strategies for preservation, like the television monitors and plants that make up Paik’s TV Garden.
The third versions of the VMQ addresses these issues by adjusting the underlying epistemological model of the questionnaire. The new focus for the questionnaire is flexibility, which it achieves by addressing an artwork less as a whole than as a set of parts* that are configured with specific relationships to one another. While the behaviors and strategies of VMQ2 are still part of VMQ3, they are now embedded at a lower level. Instead, the top hierarchical level for an artwork are the components from which it is built.
We didn’t just cut this schema from whole cloth. It is based on the way software is built from classes of objects in object-oriented programming. Each part of a work can be considered an instance of an abstract class of parts: a 42″ plasma screen is an instance of a media display, for example. Within an artwork, some parts may be modifiable within certain specifications: some artists may not care if you replace their 42″ plasma screen with a 50″ LCD. Interviews in the questionnaire are used to determine how much latitude can be applied to each part of a work. But as long as the updated version is within the same class of components, the questionnaire treats it the same as the original. As a result, these abstract parts are really functional logical components of the work that may or may not map directly to the physical components you see when the pieces is displayed or stored.
The best way to explain this is to show an example.*
This is Octris, a project I’m working on this summer that is going to be part of the Without Borders Festival in Maine when I get back. It is a 3 dimensional Tetris variant with a twist; over time the visuals fade away and the player has to position pieces only using audio cues that they learn before the graphics disappear. The director of the VEMI Lab where I’m developing this software researches functional equivalence, particularly non-visual perception of spatial information. Of course, my first instinct upon arriving there was to make a 3D Tetris game that could be played by people who are visually impaired.
If I were to break this project down into its physical components, it might look something like this. Octris uses a computer with some software I wrote, a monitor, a head mounted display, and a Wiimote:
Structure of an Artwork
Translating the work into VMQ3 terms and creating an interview requires a bit more analysis than physical components, however*. First I want to show you how data is structured in the VMQ. (Click on “VMQ Organization above.) A work has a stakeholder and a set of components. Each component has questions associated with it that are important to consider when attempting to preserve the piece. In turn, each of those questions has a set of pre-defined potential answers, usually corresponding to the preservation strategies that were central to previous versions of the VMQ. Once the piece has been put into this structure, an interview logs the interviewer, interviewee, and the selected responses to each question.
Functional components are not equivalent to physical ones, even if you add digital or virtual components like software. The idea is that each component fulfills a certain role in the overall scheme of the work. The variable media paradigm is always looking to the future, and in the future these roles may not be fulfilled by the same technology as they are now, but they will all need to be fulfilled somehow in order to recreate the experience of the original work.
- All of them, for a long enough time scale.
- I’m using the terms “component” and “part” interchangeably here. The VMQ calls them “components” right now, but for external data standardization reasons we’re transitioning to using the term “part”.
- This demo was done live at the ISEA panel. If you’d like to try it yourself, please go to the VMQ’s demo site.
- This explanatory Flash video by Jon Ippolito may help, at least until Flash disappears.