This text is from the publication Permanence Through Change: The Variable Media Approach, published in 2003 by the Solomon R. Guggenheim Museum and the Daniel Langlois Foundation for Art, Science, and Technology.
The following are excerpts from a conversation with Jon Ippolito, Caitlin Jones, Jeff Rothenberg, Carol Stringari, and Grahame Weinbren, which took place at Weinbren's studio, New York, on October 18, 2002. The introduction by Jeff Rothenberg outlines the challenges involved in emultating a work as complex as Grahame Weinbren and Roberta Friedman''s The Erl King.
Grahame Weinbren and Roberta Friedman's video piece The Erl King(1982-85) is in many ways a work ahead of its time. Running the minimal CP/M operating system on a primitive Zilog Z-80 based personal computer with a touch screen and a custom-built interface to multiple laser discs, this innovative piece provided a degree of interaction that is impressive even today. The authoring program that was created to enable the piece was far advanced in comparison to personal-computer software of the time.
The innovative interactivity of the piece alone is sufficient to make it worthy of preservation, but when combined with the technological prowess that achieved this degree of interactivity using the relatively crude and inexpensive equipment available, preservation becomes all the more desirable—and all the more problematic. In order to preserve both its behavior and the technological context in which it was created, it would be ideal to preserve the hardware— computer, videodisc drives, touch-sensitive display—utilized by the piece as well as the video content and the software, which created and controlled its behavior. No other approach would preserve The Erl King in its original form while retaining the ability to see how it was created and how it overcame the limitations of its hardware/software environment. Since preserving physical hardware indefinitely is not feasible, emulation of that hardware appears to be the most promising approach. This would entail preserving all of the work's original software (including its video content) in bitstream form and running that soft-ware on a virtual re-creation of the work's original hardware, by writing programs that emulate that hardware on a modern computer. Note that although the software for The Erl King was written in a version of the Pascal programming language (Pascal MT+), which originally needed to be compiled into machine code, the emulation approach would not require running the original compiler to recompile the original source code: instead, the executable object code would be run directly on the emulated version of the original computer.
Although the emulation approach to preserving this piece seems promising, it does involve some challenges. Emulating the Z-80 processor and the display are relatively straightforward, and it is unnecessary to emulate the details of the controllers for the original floppy-disk drives from which the program ran. The program can simply be transcribed from its original floppy disks onto a modern storage device and fed to the emulated processor as a logical bitstream (at a speed corresponding to the transfer rate of the original floppy disks). However, the videodisc drives and the special-purpose hardware that was built to control them and to allow directing their outputs to the display screen and audio amplifier would require specialized emulation. As with the floppy-disk drives, the best way to do this would probably be to transcribe the videodiscs' contents into bitstreams stored on modern storage devices (or in main memory, if speed turns out to be an issue) and to write software that emulates the logical behavior of the original special-purpose controller hardware. Assuming that these challenges can be overcome, emulation appears to offer the best chance of preserving the full richness of The Erl King.
The following are excerpts from a conversation with Jon Ippolito, Caitlin Jones, Jeff Rothenberg, Carol Stringari, and Grahame Weinbren, which took place at Weinbren's studio, New York, on October 18, 2002.
Grahame Weinbren: This piece started in about 1982. Roberta Friedman and I based the three-videodisc set up on technology that was developed for the 1982 World's Fair, which I had worked on in 1981 as a video editor and interactive designer. The idea was that it would be a kind of seamless, cinematic piece that the viewer could interrupt at any moment. It was based on a connection I made between two 19th-century texts, "Erlkonig" by Goethe and The Interpretation of Dreams by Freud. I became interested in the relationship between these two works and wanted to make a piece that would allow the viewer to similarly discover this connection.
A remarkable thing about Freud is not his analysis of dreams as much as his descriptions of the dreams themselves. He was an incredibly visual writer and I'm interested in the content of the dreams as Freud saw them. In one description, a sleeping father dreams that his child taps him on the shoulder and says, "Father, don't you see I'm burning?" And then he wakes up and runs into the other room to find that his child's body is on fire. It's a very intense, visual, cinematic moment that I was totally obsessed with for about ten years. I didn't really know what to do with it until I discovered Goethe's song "Erlkonig," which has this repeated chorus, "Father, don't you see the Erl King? Father, don't you hear the Erl King?" It's about this child who's riding with his father through the woods, and this supernatural creature keeps appearing and saying to the child, "Come with me."
Each time the child sees the creature, he turns to his father and says, "Father, don't you see the Erl King?" to which the Father replies, "No, no, it's just the shadows of the trees" or "a cloud of mist," etc. The discovery of parallels between this German, Romantic poem, which is put to music by Schubert, and the dream described by Freud is what really led to the impulse of making a piece in which the viewer could make the same kind of discovery by wandering through a mass of audio-visual material.
Jon Ippolito: When Grahame first showed me a glimpse of the programming behind The Erl King, I was struck with this program's inherent value as an historical artifact, apart from its merely functional role in the larger installation. The authoring system has elements of what I, in my loose sense, would call object-oriented programming. It's not a timeline, instead go-tos are built like objects that you're adding properties to, or removing properties from. But I have no idea what programming in Pascal was like in 1982; I don't know how this would compare to other applications written at the time.
Jeff Rothenberg: Well, Pascal was a fairly forward-thinking language and has most of the pre-object-oriented concepts of programming. It was actually a nice environment in which to be doing this sort of thing, but the interface is almost independent of the programming language it's
written in.
JI: One of the things you notice when you learn a little bit more about these menu items is the level of complexity of the "map"—the ability to program different tracks that cue up in sequence based on how many times someone has touched a particular target on the screen. That structure entrains the authoring system with the data output. You mentioned that you had been working before with a programmer who had tried to migrate the code. That was my first thought, "Oh, let's try this in Director, we'll just get all the cue points and all the video," but the cue points are intricately connected to this map that's been made of what someone has touched when. So it's not at all obvious how you'd even extricate data from code; they seem to be intertwined.
GW: Not only are the code and data inseparable, but the authoring system (the part of the software the artist uses to plot pathways through the data) and the run-time system (the part of the software that controls the hardware system during exhibition) are equally inseparable. For example, the use of global conditionals, which define a kind of map of the data, indicates which parts have already been accessed. Based on this map, the software closes off some passages and makes others available. Another set of issues comes out of the data structures, which depend on separating video from audio. The most significant of these is the "interactive cutaway": the viewer touches the screen, a new video image appears while the audio continues, and the screen eventually returns to the original video. This syntax is used throughout the piece and requires a lot of complex parameters which are defined in the authoring system and acted on by the run-time system.
Those are the issues that don't translate easily. And in fact, I used Director recently for my piece Frames—it's a lot simpler than The Erl King. We couldn't come close to the elaborate structures of The Erl King with Director. Nothing is missing from Frames, which uses a much simpler paradigm of interactivity, but the options just aren't there. [Another programmer] tried to migrate the Limosine system (this is what we Director, called the software that runs The Erl King) to the Visual Basic software language. But there were too many details.
JR: There are many reasons this piece would be a case for emulation. For the digital part of the work, it would be a shame to have to re-create a program that is still running. First of all it wouldn't be the same program when finished, and secondly, it takes a lot of work and is error prone. Who would decide whether it really was working the same way or not? So the idea is, there's no reason you can't keep running this program except that that machine becomes obsolete, but you can emulate that machine on any future computer and keep running that original program essentially forever.
That brings up another set of issues: Do we try to emulate analog devices such as laser disc players and the video monitor using digital technology. The most obvious solution, at this moment in technology, would be, "Let's take all this analog stuff and make it digital." So you copy essentially all the source material to digital form. Now you have a digital controller, which is talking to all these old analog devices, which you've now replaced. So you'd need to build a digital converter or translator, which takes all the output signals from the computer, which is saying, "OK, this meant go to frame such and such," to the videodiscs. I interpret that as doing something else on a DVD or a hard drive or some other way of accessing the material. The problem is that you now have access feed, size, and speed issues with digital video. As Grahame was saying, that would be pushing current digital technology. Digital is not quite there yet; it's getting there, but it can't quite do all of this analog stuff as fast and as seamlessly as the analog laser disc players.
JI: Beyond the technical complexities, what's exciting to me about this case study is how Grahame's authoring system prefigures today's software art. Programs like Auto-Illustrator and nato are more artist-made mediums than single pieces, and The Erl King offers an invaluable precedent for how to preserve works of this kind.