The streets of San Francisco: encounters with the Selle collection of street vendor photographs.

We frequently come to know archives through the weight of their sheer number in mass, which quantifies and qualifies their origins, ability, and options for future interpretation. These cumulative possibilities normally make up an archive's "story." Many archival reclamation projects are bent upon telling this story (such as the 2006 DVD collection and book, Electric Edwardians: The Story of the Mitchell & Kenyon Collection (1)), but the archive that is the focus of this essay involves a number of pictures that defy regular narratives. The Joseph Selle Collection at the Visual Studies Workshop, with over one million negatives of street vendor photographs, defies regular narration through its sheer magnitude of numbers and invites speculation about a different category of archive: the dual act of storytelling and counting.

This hybrid status may lie beyond the types of picture archives and their corresponding patterns evocatively proposed by the late historian and archivist Paul Vanderbilt. Responsible for modernizing major picture repositories such as the Picture Division of the Library of Congress, Vanderbilt opened the eyes of a new generation of historians and picture researchers to the interpretive possibilities of these resources. Vanderbilt listed four principal types of collections: 1) trade agencies; 2) working files of particular serial publications or promotional agencies; 3) critical collections of outstanding specimens (such as museum collections); and 4) repositories devoted to preservation as such. (2) This last category offered the most far-reaching possibilities for Vanderbilt as they were based on the contingencies of future development.

[ILLUSTRATION OMITTED]

These future contingencies invited the use of archives as an exploration rather than the routine selection of illustrations to accompany prescribed arguments. To encourage an open-ended, imaginative use of pictures, Vanderbilt worked out a long-term practice of forming combinations of images, usually in pairs, that were unrelated to each other by the usual archival categories of photographer, time period, geographic location, genre, and subject matter. Escaping the regulation of narrow control vocabularies, the pairings would reveal an unexpected line of interpretation and lead to larger associative patterns of imagery and ideas. Vanderbilt put his theories into practice over much of his career by posting combinations of unrelated pictures. These informal, reading-room "exhibitions" stimulated the imagination and encouraged conversations with like-minded visual researchers.

The key to these stimulating possibilities was the large number of pictures associated with this final type of image repository, posing unexpected and revealing juxtapositions and linkages. Another insightful historian and picture researcher, David Nye, drew on this potential with the photographic archive of General Electric. (3) There, Nye found a system of relationships between constituent elements of the corporation and its ideology that were only visible in the archive taken as a whole. Pictures directed toward consumers and management interrelated with those made for workers and the engineers. These various facets of the corporation only reveal themselves when the archive is apprehended systematically and within the current of many images rather than the single outstanding one.

Similarly, the key to unlocking the value of the Selle Collection lies in the many rather than the one, but this collection carries a multitude that sets it apart from Vanderbilt's institutional "preservation archive" or Nye's corporate "image world." The one million-plus images in the Selle collection are the product of a specialized small business in San Francisco, Fox Movie Flash, engaged in the bygone pictorial practice of street vendor photography. As such, the plethora of pictures belong to one overriding set, as diverse as they may be within that type. Since the pictures were all taken with a specially modified motion picture camera and stored on 100-foot rolls, the standard half-frame negatives are readily digitized using standard motion picture, post-production scanning equipment. This capacity for digitization animates much of the potential of this new type of archive and made possible the exploratory projects that ensued with the Selle Collection at the Visual Studies Workshop (VSW, publisher of Afterimage) in Rochester, New York.

My purpose with this article is to recount these projects--both in the sense of telling the story of and assessing the impact of numbers and the digitization of images--and to speculate on further project possibilities with special archives of this kind. The projects are Andy Eskind's groundbreaking initial work with the collection that established the basis for David Mount's video 17532 Pictures (2005) and Elisabeth Tonnard's artist's book, Two of Us: Encounters (2007). Each artist worked with the same set of images (about 18,000 digital images scanned as a pilot project). While each work bears its separate identity and provocative meanings, they share an overriding ambition "to verge on something else" that stems from the intractability of very large numbers. There is an instability and questioning that causes each artist/researcher to direct their project and seek meanings outside of regular boundaries. These experimental projects point out that archives of very large numbers operate as a kind of new math--an entirely different logic causing archival work to "verge" on boundary-testing genres and hybrid "artistic acts."

LAYING THE GROUNDWORK

As mentioned, if there is a story here to recount at all, it is thanks to Eskind's rediscovery of the collection in the spring and summer of 2003. Until then, there was little documentation of the collection: it scarcely had a name (the "Selle Collection" is still somewhat provisional). With over thirty years experience pioneering image databases at the George Eastman House in Rochester, New York, Eskind quickly saw the possibilities. He immediately set to work to recount the collection (in both senses) and to find out what it consisted of and where it came from. The numerical counting part was easy: just a matter of multiplying the 1,500 frames on each 100 foot roll by the number of rolls in each drawer, multiplied again by the number of drawers (1,500 images x 100 rolls x 7 drawers = 1,050,000 images total). If this calculation is not exact, even a casual survey of the collection would indicate that we are dealing with a hell of a lot of images here.

Learning the history of the collection was trickier and started from anecdotes about its origins provided to Eskind by the former coordinator of the VSW Research Center, William S. Johnson, and the founder and former director of VSW, Nathan Lyons. Though still somewhat sketchy, it appears that it was donated to VSW in the mid-1970s by a recent graduate of its MFA program, Brent Sikkema. He rescued the rolls of negatives after the collapse of Joseph Selle's company, Fox Movie Flash. As with other collections of images at VSW, this storehouse of negatives was not purchased according to a formal collections policy but "saved from the dumpster" because VSW possessed both the raw storage space and a keen receptiveness to the value of visual culture and vernacular images.

Eskind built on this account of the collection with clever detective work. He conducted background research into Fox Movie Flash and found it was in operation at 942 Market Street in San Francisco from the 1930s to the 1970s. He searched records about the proprietor and his next of kin. He located and interviewed the last living camera operator, Joe Reston, and learned some telling anecdotes of the life and craft of the street vendor photography business. Amazingly, Eskind discovered and purchased on eBay at least four of the surviving specimens of the modified DeVries motion picture cameras, the very cameras used by the Fox Movie Flash team. These cameras are a remarkable story of folksy photographic ingenuity. They appear bricolage-like as a whimsical assemblage of converted and merged equipment parts: a basic 35mm motion picture camera hacked with a modified shutter and flopped on its side. On top, a viewfinder sticks out and doubles as a platform for advertisement samples. Below protrudes a ticket dispenser. The street operators would bear these odd photographic rigs with harnesses that made them look like optical accordion players. Along with these finds came other accoutrements of the trade such as sample vintage prints and miscellaneous trade advertisements.

[ILLUSTRATION OMITTED]

Equally interesting was the information that was discovered within the rolls with a little more detective work. The rolls contain numerical markings, though an exact chronological sequence of the images and rolls is still elusive. Beyond the interpretation of frame numbers and markings, there were attempts to "crack the code" of the material frame itself by inspecting minute irregularities around the negative edge and the potentially matching frame of the film gate. Eskind invited in a retired engineer from Kodak, who had worked on the famous Zapruder film of the Kennedy assassination case, to consult on the investigation. Eskind also engaged other experts through listservs to find useful information within the pictures themselves. For example, film buffs found they could date particular images to the month by discerning in the background the title of movies on theater marquees. But as revealing as such embedded details might have been, the "collective view" enabled by the digitization of the images has generated even more avenues of interest.