Fifteen years ago, I was sitting in a dimly lit room, leaning over a Steinbeck (an old film-editing machine) and cutting a strip of film. Aligning the film beneath the blade, I adjusted the blade once or twice and then cut the film in between two images on the strip. I released the filmstrip and did this once more, and then carefully pasted down clear plastic tape, matching the notches on the filmstrip with the notches on the tape. After several cuts and reattachments, an edited filmstrip developed. The process took two full weeks: three days of filming, five days for the film to develop, and six days of reviewing and cutting the film. The film ran for three minutes in front of a small classroom of students. It was a lot of work to put together, and it was satisfying to produce, but efficient it was not.
The filmstrip now lies in a box on my shelf. Though it will eventually erode, the images on the filmstrip and my arrangement of the cuts and the tape are still visible; they remind me of the time I invested as well as my experiences at design school. The object continues to hold value for the narrative it embodies as much as the narrative it represents.
Next to the filmstrip on my shelf sit several digital artifacts: Hi8 tapes, zip disks, floppy disks. Each contains an assignment from my film classes, and all rely on digital tools for access. I no longer have a Hi8 camera or a functional disk drive for accessing data on zip or floppy disks. What's more, I have no software to decode the obscure codec with which these files were encoded. All of the files look roughly the same: black, worn plastic artifacts with clever labeling, such as "Daniela II" and "Daniela III." Though they are still intact, the objects are relatively meaningless. I have no ability to understand and review the elements of this work.
Looking back at these objects, I can see that each one's form is significant. My filmstrip still reveals some of its imagery, but so, too, does it show physical evidence of my interactions: signs of creation as well as traces of wear and decay. Certain processes have affected the object, and the object reveals those effects to me. Yet this isn't true for my digital artifacts: I cannot see or understand what the disks and tapes represent, nor how the representations have changed over time. As a designer of new technology, I might wonder how to preserve a sense of valued historicity. How could I make visible the trajectory of ownership or the meaningful traces of interaction over time? How can I design for provenance ?
One place to look into questions of historicity, rather paradoxically, is amid a hotbed of technological change. In the San Francisco Bay Area, large festivals like Burning Man and Maker Faire, teaching facilities like the Crucible and Center for the Book, as well as independent studios and hacker spaces energize a surprising range of personal, "by hand" production. People engage in handwork as a response to, and an extension of, the technological culture in which they are embedded. Some turn to craftwork in order to counter the values embedded in a culture of planned obsolescence. Through my interactions and interviews with these craftspeople , I find a compelling argument for why traces of technique and provenance matter. Consider the story of a potter.
Mike has a background in chemistry, computer graphics, and engineering, and spent most of his professional career using software tools to design hardware. The result was a custom graphics chip that he could later view through a magnifying paperweight on his desk. Still, conventional craftwork was always close by. Mike's mother was a professional potter, and when she passed away, he and his wife decided to enroll in a local raku pottery class. Although Mike had little interest in the craft as a child, this "dabbling" with clay and glazes proved to be right up his alley. "It's the kind of challenge I like to take on, being an engineer," he explains. He began working with transparent glazes that contain soluble metals that become visible when heated in the kiln. Both the heat and the shape of the ceramic vase affect the patterns and colors that appear, "so you have to do a lot of testing and figure out how to pull it off." The pottery demands purposeful material exploration.
In this regard, the interactions involved in both Mike's pottery and his engineering work seem to have similar qualities: distance from the end product as well as dedicated attention. Through production, Mike accumulates specialized knowledge and balances strategy with surprise. Nevertheless, he views his engineering work as different from his pottery. Examining his completed graphics chip, Mike can just barely make out the sections he designed only two years ago. The engineering "just seems totally virtual. It all exists in your imagination," he says.
Out of the kiln, the vases allow Mike to formulate his understanding of how the chemicals and the shape of the clay play together with heat. Mike learns through experimentation with the numerous test tiles he creates before the full vase. The tiles materialize traces of handwork (dotting, patterning, layering) that enable Mike to assess the choices he made and to further develop his techniques for applying glaze. But though computer-chip design and pottery involve similar forms of problem solving, they suggest different trajectories of technique.
Unlike developing new color and form in a kiln, the craft of book repair entails "bringing something back into use," explains Thomas, a long-time binder. Thomas has no technical background. During college he majored in philosophy while working at an antiquarian bookstore. Once a month a customer would come into the bookstore to pick up damaged books and return them the next month, beautifully rebound. "It was just something that stuck out in my mind," he says. Years later, and after several stints in the restaurant industry, Thomas returned to books. He took a few binding lessons, promptly bought a small bindery, and continued his training. The business slowly grew into one of the largest (and only) trade binderies in the Bay Area. However, book repair is not its only preoccupation. Given the area's clientele, the bindery also produces numerous restaurant menus and e-reader covers. Thomas reflects on these changes:
If you brought your mother's cookbook to us, you want it to have the general feeling that it always did. But you really want to be able to open it up and handle it. And bringing it back so it's exactly the same condition or style that it was when it was first published is not really that important ... In a sense, that ties into the very books that I think are most important to us, and the ones that I think will ultimately be lost with the advent of the digital medium. Your first child's children's book may someday be a digital book with pictures and so forth. My girlfriend does a lot of cooking off of recipes she gets on the Internet. No granddaughter of hers is ever going to handle some book and say, "Grandma [Erica] had this." It's not going to happen.
Thomas anticipates that books will be displaced by the "digital medium" and, as that happens, they will lose their status as heirloom objects. An online recipe has no features to be repaired or sustained. It has little ability to invoke memories of the past. Regardless of its usefulness or emotional resonance, the online recipe will not endure over time.
The digital amalgam of online documents is therefore seen as different from the physical book. A cookbook might survive another generation of repair, but a collection of online recipes remains ephemeral. A particular book has meaning because of how it feels as well as how well it can be used. This imbued meaning has less to do with individual traces than with how the arrangement of material unfolds meaning through its creation and use. Though what it would mean to repair the digital is still unclear, in supporting digital assemblage, designers might consider to what degree such work might remain stable and accessible from one generation to the next.
In HCI research, provenance is typically discussed in relation to undo and back functionality or data-management and auditing software. We invoke undo when we make a mistake while typing a document. We press the back button when we want to revisit a Web page a few links prior. By enabling awareness, interaction histories help us more effectively navigate changes and engage in distributed collaboration. More broadly, we create interaction histories that support productivity as a utilitarian endeavor.
Yet as we have seen through stories of handwork, the life cycle of an artifact is often woven into additional concerns. Provenance is valued for its emotional resonance in addition to the pragmatic goals in which it is entrenched. Traces of use and ownership matter through the specific meanings they convey, the mistakes they unveil, and the educational opportunities they afford. Rather than relying on an undo stack, we might design for more evocative artifact histories. By repurposing the interaction histories for coordination, we reveal expressive traces of digital production.
There is also some irony in this analysis. Several digital document systems, such as Google Docs or Microsoft Word, maintain more useful histories than the tools designed to support creative work. For example, Photoshop and Illustrator create incomplete interaction histories. One might create multiple layers of interaction, but one can go back only 20 history states. Raising the number of records just decreases the program's performance. And each discrete state is itself incomplete: It reveals the result of an action (the mark) rather than how something was accomplished (the stroke of the brush) . Digital tools could more thoroughly represent the layered, idiosyncratic character of skilled mark making.
We find, too, that different artifacts entail different technical trajectories. Mike's glaze work was seen as more anchored to his pottery than his software engineering was to the computer chip. My film editing was infinitely more visible on the filmstrip than on the digital disks. Depending on the process of creation, distinctive material traces emerge. As such, designing mechanisms for interpreting interaction histories involves developing more sensitive and varied approaches to the codification of technique.
In HCI, several systems already engage affective responses to provenance. One genre of tools emphasizes the communication of techniques involved in production. Pierce and Paulos's Energy Mementos , for instance, play back patterns of energy generation through light and movement. These tools share production methods by conveying aspects of the process of creation: pace, rhythm, and skill. By identifying and communicating the key aspects of production, designers can expose provenance to enable reflective inquiry.
A second genre of tools focuses on attrition and decay In Zoran and Buechley's digital restoration , handmade pottery that has broken is then "restored" using a 3-D printer. The printer accentuates the broken edges by reassembling and transforming the object remains. Here digital production incorporates new materials and techniques while foregrounding the engagements (breakage and wear) that led to the object's current state. The object evolves through rearranging traces of use and, in turn, ideas of provenance.
Alongside delineations of use and skill, a third genre of tools involves sharing experiences of creation and ownership. Our Spyn system , for example, enables knitters to digitally annotate stitches while crafting, associating audio and visual recordings with locations on knit fabric. In trial deployments, knitters connected fabric with personal reflections at the time of production. By recalling these annotations, recipients of the knit articles became aware of the knitters' investments of effort and care. Such tools might enrich production work by capturing provenance through digital media, connecting intimate object interactions with processes of exchange and use.
Each genre of tools traces the use, technique, and ownership of artifacts in particular ways. They suggest paths for valuing processes of creation and trajectories of use that enable us to consider which elements of an artifact should remain stable over time and which materials, digital and non-digital, matter in the ongoing production of value. They provide a starting point for considering how we might design evocative traces of provenance.
Beyond handwork, this work suggests that designers can benefit from becoming sensitive to both the practical and affective dimensions of skill. What makes an artifact meaningful is bound up with how its traces are made, remade, and interpreted over time. In illustrating how materials figure into these handwork processes, I hope to encourage the design of technologies that meaningfully engage the expressivity of material histories .
3. Here the term provenance is taken to mean the itinerary that something follows in order to reach its current condition, beginning with its creation .
5. For a counter example, see Grossman, T., Matejka, J., and Fitzmaurice, G. Chronicle: Capture exploration, and playback of document workflow histories. Proc. of the 23rd Annual ACM Symposium on User Interface Software and Technology. ACM New York, 2010, 143152.
8. Rosner, D.K., and. Ryokai, K. Spyn: Augmenting the creative and communicative potential of craft. Proc. of the 28th International Conference on Human Factors in Computing Systems. ACM, New York, 2010, 24072416.
Daniela K. Rosner is a Ph.D. candidate at UC Berkeley's School of Information. Her research focuses on the interplay between computing, craft, and the creative communities that surround them.
©2011 ACM 1072-5220/11/0900 $10.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2011 ACM, Inc.