Descent to the Underworld:
Networked Creative Collaboration
One thousand years ago, stories were passed among cultures by troubadours and storytellers, who wandered from town to town. Like seamstresses, they reshaped and embroidered their tales so that today, we recognize the pattern of the Cinderella or Psyche stories, even when they’ve been done up in a different style. I wondered if the network could act as a high-tech Silk Road, serving as a transport for the transmission of stories in a similar fashion. That thought became the nucleus for first, “Story Streams” and then, “Descent to the Underworld”.
Descent is a collaborative project that employed 21st century technologies to re-tell a 5th century B.C. myth. Using the Access Grid, an open-source video collaboration program, and communicating across the Internet2 and partner international networks, sixty-five students from around the world formed an online classroom and studio, where they met over the course of a semester to discuss the story, storyboard their interpretation and then produce over sixty media assets, including sound track, sound effects and video footage. The media they created was embedded into a game-film, an online videogame whose game play generates short film clips, instead of points or health. At the end of game-play, the clips are compiled and streamed back to the player, unspooling a short filmic version of the game.
History
Descent grew out of “Story Streams”, an online live cinema production I had done in the winter of 2003; it is the first example of real-time, live cinema. “Story Streams” networked together directors in Paris, Montreal, Mexico City and Philadelphia, via the Real Broadband Network.
Each of the directors created a ten minute film based on the classic story of the Hero’s Journey. Then, the directors in Paris, Montreal and Mexico City uploaded their films, in real-time, to the director in Philadelphia. Working at the studios of the PBS affiliate in Philadelphia, the director mixed the three different narratives together in real time, and sent the finished stream back out over the Internet. In this case, while each film did follow a hero’s journey, each hero chose a totally different visual and narrative path.
Descent took the concept of networked storytelling to a new level. This time there were teams involved, instead of individuals, and the teams had to work together to produce one finished narrative.
Technology
Work on Descent began in the summer of 2004 when I was invited by Drexel University, in Philadelphia, to put together a project that would showcase Drexel’s participation in the Internet2 (I2) consortium. The Internet2 is a research and development consortium whose membership includes many universities and some corporations around the world. Because the capacity of the Internet2 net is so large, I looked for a video software program that would be able to leverage that that pipe. Many I2 projects use H.321, a videoconferencing platform, however H.321 is codecbased and produces smaller video windows. I ended up selecting Access Grid (AG), an open source program developed by the Futures Lab at Argonne National Laboratories at the University of Chicago. AG enables non-codec based, TV quality video and uncompressed audio (and has a window for text messaging). It can also scale to a large number of locations and project wall size video windows, providing an immersive environment that I felt would work well in the global classroom.
Access Grid is not widely deployed, therefore any universities that were interested in participating in the Descent project either had to have an existing AG node, or be willing to commit the time to set one up (no minor investment). MAGPI at the University of Pennsylvania agreed to serve as the AG node and Internet2 interface for the project. Apple Computer supplied two Apple X-serves, high powered servers that were necessary to power the AG communications.
Access Grid however, had not previously worked with the Apple server side and the engineers at Argonne worked rapidly to supply us with a beta version of the program so we could launch on time.
There were crashes, of course, and there were glitches. Running AG on the Apple servers for the first time meant that a lot of the applications were not available to us, for instance, recording the class sessions. We attempted to use Camtasia, an excellent program, but, like the Access Grid, it was a memory hog and crashed a number of systems. The audio would sometimes drop out completely, which is when we discovered why Skype is such a great product. There were times when several locations around the world all filtered in to MAGPI via AG, Skype and text messaging. In the middle of a live session like that, you need multiple redundant communications and we used all of them…and when the locations were all linked together and students and faculty were freely exchanging ideas, it really was magical.
There’s almost always a forest in the old stories. Thick and dark, deep and haunting, it signifies the unknown – and it must be conquered or there is no story. On this project, the technology was the forest. It produced anxiety, sleepless nights – and wonder. Without it, there would have been no story.
The Task
Eight universities signed on to the project: Drexel University in Philadelphia, which teamed with the University of Utah in Salt Lake City; Tsinghua University in Beijing, which teamed with the University of Washington in Seattle; Louisiana State University in Baton Rouge, which partnered with Northwestern University in Chicago; and the Universidade do Vale do Rio dos Sinos (Unisinos) in Sao Leopold, Brazil, which worked with the Faculty of Fine Arts in Prague (FAMU) - though FAMU withdrew halfway through the project.
The eight universities were split into four teams, three of which had to produce narrative films (twenty clips per team) and the fourth (LSU and Northwestern) which produced the sounds tracks for the narrative films, as well as the sound effects for the game itself.
The videogame was developed so that players could select which teams’ clips they wanted to see at the end of game play. It was also designed to allow players the option of having the computer randomly generate clips from different teams. Because of this database approach, each clip had to align with clips from other teams. This was done several different ways: first, we assigned the main narrative characters specific colors. The hero had to be or wear blue, the underworld king purple and the kidnapped person, white. Each clip had to open and close on a medium shot.
Each clip had to be precisely ten seconds, and there had to be twenty clips, each of which had to have a similar story arc (e.g., the hero arrives at a barrier to the underworld, and is barred from the entrance by a strong presence). Finally, each clip in a sequence received the same soundtrack.
Class Sessions
The classes went online in February, 2005 and worked together until the beginning of May, 2005. As a back-up to the weekly video sessions, bulletin boards were created for the teams to use for out-of-class communication. Ironically, they were more at ease with the boards - at least, in the beginning - then they were with the video screen. For two of the teams, this was due to a difference in languages. But culture, even in the US, played a very definite role. For instance, the Drexel team was comprised of students from a major urban area in the Northeast corridor, an area where the culture is more aggressive and “in your face”. Their partner team at the University of Utah was more reticent and tended not to challenge. They preferred to communicate through their professor during the online sessions, whereas the Drexel students by-passed their teacher and took charge themselves. While we worked with the Drexel team to bring their Utah teammates into the process, the Utah students never really came alive on screen. Off-screen was a different story however. Here, the Utah students dominated, posting their work and leaving comments on a regular basis.
The Beijing-Seattle team faced not only language issues, but also time zone challenges as well. The Seattle students logged on at 9 p.m. on Monday evenings – 10 a.m. Tuesday for the Chinese.
The Chinese students made the Utah team look positively verbose! They would ignore the camera and speak quietly amongst themselves quietly; when finished, one of the students would then face the camera and communicate with the Seattle team.
The Seattle team, happily laid-back, took their time and would wait out the discussion. Again, these two schools also used the bulletin board quite actively. Interestingly, the Chinese students were the only non-media majors involved in the project. They were all engineering students – and they produced the most beautiful illustrations for their film clips.
The Brazil and Prague team had a rough time getting off the ground. The Prague team came back from winter break and had trouble setting up their AG node. The Brazil team, meanwhile, was on summer break until mid-February. They had a hard time connecting, both to the technology and each other. In the end, Prague withdrew, leaving Brazil pretty much on their own, so we integrated some of the Brazil classes into the Drexel/Utah classes.
LSU and Northwestern had the least amount of challenges, culturally and technologically. Both faculty were conversant with the technology and had excellent IT support. Each school fielded a team of grad students, instead of undergrad, and that level of maturity helped when they were confronted with creating a soundtrack for film clips that had not yet been produced, meaning they relied on written narrative to get a feel for the story.
The Story
Despite the students’ initial on-screen reticence, they were all very conversant with the story of the Descent to the Underworld.
I chose the Descent story as the project backbone, because the story of a journey to the underworld appears in every culture, as does a story of creation. In the underworld stories, a lover or child dies, or is kidnapped by a ruler of the underworld. The bereaved person then goes in search of the loved one, wandering the world until he or she finds the entrance to the underworld, gets past the guardian and confronts/overcome the underworld king. Frequently there is one final challenge on the ascent, which many do not pass. The underworld stories most familiar to Western Culture are “Demeter and Persephone” and “Orpheus and Eurydice”. Other versions include the Nordic “Baldur”, the Egyptian “Isis”, the Indian “Savitri” and the American Indian, “Blue Jay”.
All of the students and faculty were sent copies of the different tales, and asked to work with their partner school in developing their own interpretation of the story. For the game itself, we chose the “Orpheus” motif, though we did not disclose that to the students. Ironically, they all chose the “Orpheus” storyline, right down to the gender roles (in the game we designed an option to rescue a man or a woman). While most of them had probably not seen Cocteau’s movie version, “Orphee”, it could be that this was the narrative most familiar to all of them, because it the Underworld narrative motif is found in a number of video games. Or maybe Jung was more right than we know.
Despite the similar narrative interpretations, there were many differences in their visual styles.
The Beijing-Seattle team set their story in 6th century China, illustrating the story with hand drawn water colors in very soft tones. The Drexel-Utah team used a combination of digital video, shot on the streets of West Philadelphia, and CGI, developed by Utah to represent the underworld scenes. Their colors were dark, often black, and heavily saturated. The Brazil team, using a few images from Prague, created a series of colorful, whimsical digital stills, which looked like collages.
Conclusion
Years ago, I was lucky enough to watch an old Turkish woman weave an oriental rug. Dressed all in black, kneeling before a wall-size loom, her hands worked the multicolored threads, selecting different colors as she wove the tapestry. Had she chosen a red where she took a blue, or a yellow where she used purple, she would have created a different image, even using the same threads. Watching her made me understand why the language of storytelling has appropriated the
language of weaving.
Descent reminds me of that rug. It brought together strands of stories, visuals, people and cultures and hung them from a fiber optic loom, where they were all woven together to form an arresting tapestry of tales. Unravel the skeins at www.gamefilm.tv .
