Image and Artifact:
The Photograph as Evidence in the Digital Age

Martha A. Sandweiss

Journal of American History, 94 (June 2007) 193–202

The brief essays in this round table collectively explore how photographs can be used to understand the past. Their broad mix of voices—from the archivist and the historian, the photographer and the photographic subject—makes it clear that there is no one way to understand an image. Even a quick reading reveals interpretive tensions: the photographers’ intentions clash with the ambitions of the subjects, and both appear at odds with the needs of the viewers, who bring to the image their own experiences and interpretive concerns. Despite their differences, the essays taken together pose two critical questions. What does a historian need to know to interpret a photograph as a historical document? And how stable are images as records of the past?

The essays by the photographers Ted Engelmann, David Allen, and Jonathan Hyman indirectly raise the issue of whether one must understand the photographer’s intent in order to interpret an image as a historical document. Should it matter to the historian that Engelmann views his photographic work in Vietnam as therapy; that the photojournalist Allen finds his image of an exhausted Oklahoma City rescue worker “heartwarming”; or that Hyman seeks to create an archive of vernacular responses to the events of 9/11? Should it matter that, as Eric Sandweiss argues, the amateur photographer Charles Cushman had “no apparent intended audience” for his fourteen thousand color transparencies of the vernacular American landscape beyond a small circle of intimates? Can one make constructive use of these photographers’ images without considering their personal ambitions for their work?¹

These essays suggest that understanding a photographer’s motivations can help the historian perceive why an image maker makes certain views and not others, how one picture in a large body of work relates to another. But the essays simultaneously suggest how hard it can be to recover photographic intent and how unsteady it remains as a category of analysis. Sandweiss’s piece on Cushman, for example, conveys the difficulties of inferring intent from a large body of photographs when one has scant information about the pictures from the photographer. Conversely, Engelmann’s essay reveals so much about his personal use of photography “as a way to work through the wounds and scars from the American War in Viet Nam” that one is hard-pressed to interpret his photographs except in that light. But Engelmann’s is a rare case. Historians more often confront the difficulties of interpreting images without extensive biographical information on the photographer than those of interpreting pictures in light of the photographer’s own readings of them. In either case the historian must be mindful of photographic intent, not because it provides the only way of interpreting an image, but because it provides one possible starting point for a more complicated reading of a picture. The photographer’s intent may be fickle, unknowable, beyond the powers of the historian to ascertain. It does not necessarily adhere to the photograph itself in easily discernible ways. But to the extent that every photograph represents a point of view, in the literal as well as the interpretive sense, it is always worth inquiring what it is. One caution, however. The photographer’s understanding of his own pictures does not necessarily remain fixed. As David Allen’s essay suggests, time and public reaction altered his feeling about his photograph of Anthony “Skip” Fernandez III and his rescue dog, Aspen. The picture may now define Allen’s career, but it did not and could not do so when he snapped it. That it now seems so “powerful” to him is a reflection of other viewers’ responses over the past twelve years.²

If the photographer’s own reading of an image provides one way to begin thinking about it, the subject’s interpretation provides another. And, as Fernandez’s essay suggests, a subject’s interests can be at odds with those of the picture maker. The Oklahoma City rescue photograph that serves as the iconic and defining image of the photojournalist Allen’s career remains for Fernandez, its sleeping subject, a vague embarrassment that unfairly singles him out for recognition from his fellow rescue workers. The picture becomes richer as a historical document when the two interpretations play out in tandem, illuminating the ever-present tensions between photographer and subject. For its first fifty years, before the invention of the snapshot camera, photography remained an inherently collaborative medium. Slow exposure times and bulky equipment meant that most nineteenth- century photographic portraits represented a knowing collaboration between photographer and subject. But the invention of the snapshot camera in the late nineteenth century and the growing popularity of small 35 mm cameras, beginning in the 1930s, made it possible for photographers to make images without their subjects’ consent. Historians might thus keep in mind that the subject’s involvement in the creation of a photograph depends, among other things, on the technology used to make the image.³

The needs of the subject are no easier to ascertain than the needs of the photographer, indeed, they are often harder because the photographer (or a publisher) more often shapes the descriptive captions that adhere to the image in the archive or in print. As Barbara Orbach Natanson’s essay about the Library of Congress photo archives suggests, the subjects depicted in the library’s Farm Security Administration (fsa) and Office of War Information (owi) photographs may have perspectives on the images that challenge the pictures’ official captions in the archives. Consider Dorothea Lange’s “Migrant Mother,” the most widely recognized of all fsa images. Lange never called her “migrant mother.” Presenting the woman as an archetypal victim of the depression, she characterized her with broad strokes in a brief caption: “Nipomo, Calif. Mar. 1936. Migrant agricultural worker’s family. Seven hungry children. Mother aged 32, the father is a native Californian. Destitute in a pea pickers camp, because of the failure of the early pea crop. These people had just sold their tent in order to buy food. Most of the 2,500 people in this camp were destitute.” Lange used her subject’s anonymity to enhance the iconic power of the picture and identified the husband as native to the state in order to enlist the empathy of Californians grown weary of dust bowl refugees. Her migrant mother remained nameless by design. ⁴

But the woman’s grandson, Roger Sprague, now insists on giving fuller voice to the life of Florence Owens Thompson, the picture’s subject. On his own Web site, the “migrant grandson,” as he calls himself, identifies his grandmother as an Oklahoma-born Cherokee woman (elsewhere identified as Florence Leona Christie) who at the age of seventeen married Cleo Owens, the son of a Mississippi farmer. The couple moved to California in late 1924 with their three children. Owens died in 1931, leaving his wife with five children and pregnant with a sixth. In 1933 Florence became pregnant again, by a man she refused to name, and returned to Oklahoma with her children to have her baby there. She returned to California the following year with her six older children, leaving her baby with her mother, and joined the army of migrant workers in the San Joaquin Valley. In 1935, as her grandson puts it, a man named Jim Hill joined the family “and acted as husband and father to Florence and her children.”⁵ In reclaiming his grandmother’s story for his family, Sprague complicates Lange’s reading of the picture. Nothing suggests Lange knew anything about Florence Thompson’s complicated personal past. But it was in her interest not to know. No Indian woman who moved to California before the Great Depression and subsequently bore a child out of wedlock could embody, in an uncomplicated way, the image of the deserving poor that Lange sought to evoke with her carefully composed portrait. In the end, Lange’s ambitions for her photograph tell us much about Lange and the way she imagined her audience, but relatively little about Florence Leona Christie Owens Thompson.

The historian who would be content with understanding the photographer’s point of view alone would thus miss part of the richness of any photographic print. And yet, in considering the subject’s point of view, the historian must be mindful that it need not be any more fixed than the photographer’s. Florence Thompson cooperated with Lange, patiently posing again and again as the photographer rearranged her children and the material objects of her home. But over time Thompson came to feel “exploited” by the image, as it became increasingly recognizable and marketable. In the case of the “Times Square Kiss,” the iconic V-J Day image by the Life magazine photographer Alfred Eisen staedt discussed in Robert Hariman and John Louis Lucaites’s essay, the very identity of the subjects remains unclear. There is no shortage of people claiming to be the central figures in the picture. If, as Lucaites and Hariman claim, the photograph’s power comes from the larger cultural narratives it enacts, the subjects’ identities scarcely matter. Indeed, their very unrecognizability feeds into the iconic power of the image. As with “Migrant Mother,” the personal narratives of the picture’s subjects can disrupt the powerful cultural stories that adhere to the image in its multiple iterations.⁶

If photographers and subjects have their own needs for photographs, so too do the viewers. As the efforts of the “migrant grandson” suggest, a family member may have a different stake in a photograph than a stranger does. Likewise, a viewer who has experienced something similar to the event depicted in a photograph may respond to it differently from someone to whom the scene seems strange and unfamiliar. A member of one cultural group may respond to a picture differently from a member of another. A direct gaze into the camera lens may signal forthrightness and openness to one group, for example, but unspeakable rudeness to another. Meaning in photographs is rarely fixed or self-evident. Viewers construct meanings every bit as much as photographers or their subjects do. Moreover, readings not only change across cultural or social groups, they also shift across time. Photographs of the World Trade Center made during the 1970s began to look very different to us after September 11, 2001. Although no single event cuts us off from the world he depicted, Charles Cushman’s photographs of the vernacular landscape gain new meanings for us as that mid-twentieth-century landscape disappears. Pictures that once represented a culminating moment in a story become, with the passage of time, one more point on a historical continuum. Conversely, pictures of oncemundane subjects can become, in retrospect, documents of singularly important objects or events.

Scholars who work with photographs inevitably bring to their task the particular tools of their trade. As Colleen McDannell suggests in her essay on the use of fsa images as a source for the historian of religion, art historians and curators generally ask different questions than social or cultural historians do. Art historians, she argues, assume that “photographs engage viewers directly and thus should be mediated by only a minimum of explanatory text.” For them, “explaining who is in the photograph and what those subjects are doing misses the point entirely: a photograph is the mirror of the creativity and context of the photographer.” The historian, she suggests, can construct useful new narratives, however, “by moving back and forth between what is in the photograph, who took it, and why.” Moreover, while an art historian often values the singular image, for a historian the greatest value of an image may lie in its relation to a larger body of work, either within an archive or within what the essayist Michael Lesy calls the “sequence,” a selection of views carefully edited to create a visual narrative. McDannell may exaggerate the differences, since art historians have become increasingly concerned with matters of social history. But it remains a truism that in comparison to historians, art historians value form over content and give greater attention to the photographer’s training and stylistic antecedents. And the accidental detail that to the photographer (or later to the art historian) has little value, may be of immense value to the cultural historian curious about the details of historical costumes or architecture, land-use patterns, or any of the other myriad details of everyday life.⁷

Although art historians may pose different questions than do historians, they nonetheless have an abiding interest in something to which historians need to give greater attention: the physical form of the photograph. For photographs are not just images; they are physical artifacts. The physical form of the photographic image, prescribed by prevailing technology, determines what can be photographed, how it can be displayed or published, how it can be encountered by others, how it can circulate through public culture. Visual images may circulate freely now over the Internet, but original photographs (at least those not created with digital technologies) remain time-bound physical objects, and historians must be mindful of photographs as objects if they wish to make responsible use of them as primary-source documents.

In the beginning, photographs were singular images. Daguerreotypes, ambrotypes, and tintypes—all early forms of photograph popular in the United States—were unique objects produced without negatives. Historians do not need to understand the precise chemical processes that produced photographic images on silver-plated pieces of copper, glass plates, or cheap sheets of nonsilver metals. But they should understand the implications of the particular technologies. Because of their small size, early photographs more often served as personal items than as objects of public display. And because early photographers could not easily make multiple copies of a hard-won view, they most often made portraits, for which they could be assured paying buyers. But the content of photographic images, as well as their use and dissemination, changed enormously with the development of wet-plate-negative technologies in the late 1850s and the easier-to-use dry-plate-negative technologies in the early 1880s. Once photographers could make an unlimited number of paper prints from a single glass negative, they became bolder about making landscape views or documenting newsworthy events. But even then, photography did not truly become a form of mass communication. Not until the invention of halftone technology in the 1880s and its wider use in the 1890s could photographs be reproduced in the press without being engraved by hand on a wood block or metal printing plate. The wood engravings of the Civil War dead that appear in Harper’s Weekly only hint at the startling power of the original photographs.

The stories here about twentieth- and twenty-first-century photographs hint at a sea change at least as monumental as the changes that transformed photography from a technologically laborious medium to a simple point-and-shoot practice and supplanted unique Daguerrean images with mass-produced paper photographs. It is not simply that we see in these stories the fulfillment of photography’s promise to become a truly democratic art, accessible to anyone with an inexpensive camera (or even a camera phone). We see the impact of the digital revolution. In some of the essays presented here, the authors encounter historical photographs as digital images on a computer screen, rather than as materials in the archive. No longer actual physical objects, the photographs become freefloating images, dissociated from photographic chemicals or even printer’s ink. Writing about the digitization of the Library of Congress’s vast collection of photographs, easily read images, enormous numbers of pictures can be made accessible to researchers around the globe, text and image sources can be linked in exciting new ways, and research can become more collaborative. Both Colleen McDannell and Eric Sandweiss provide vivid evidence of the sort of historical research made possible by access to digitized collections. By moving back and forth between archival documents and the online images and captions available on the Library of Congress’s digitized Web site of fsa resources, McDannell crafts a new story about African American Catholicism in Chicago during the 1930s. Sandweiss relies on the digitized version of Charles Cushman’s enormous and difficult- to-view collection of 35mm slides to make visible a body of work virtually impossible to study in its original format. The issue of digital visibility goes beyond the simple uploading of digitized copies of negatives, prints, or transparencies. The Nebraska State Historical Society, for example, uses digital technology to enhance historical images, making visible details impossible to see with the human eye.⁸

Digital technologies thus make it possible for researchers to see images that the original photographers never saw. Unprinted negatives can be transformed into positive images; photographs can be cropped, enlarged, or enhanced to reveal details. What sort of historical archive does this become? With computers we can create an archive of images made in the past but that no one in the past ever saw. Should we take equally seriously as forms of visual evidence pictures photographers (or their employers) chose to print and those they repressed? Do we need to be mindful of the differences here? In digital archives we can encounter bodies of historical work for which there exists no original viewing context. Increasingly, we all encounter photographic images, not in archives, family albums, picture magazines, or books, but on computer screens. For the classroom teacher this can be a tremendous boon. With a Web-connected computer, we can introduce our students to millions of images from America’s past, affording them access to visual resources that previously required travel to an archive, patience, time, money. But let me voice a note of caution: encountering an image online is not the same as encountering the original photograph.

When we encounter a digitized photograph online, we need to ask whether we see the entire image or only a piece of it. We must still inquire about its original format, the original medium, the notes printed on the back or along the margins, the pictures that once lay next to it in an album. Rich historical stories can be constructed from such details. As we transform visual artifacts into disembodied digital images, we risk losing key information that allows us to understand how the image once circulated through a world of popular culture, how viewers once encountered it, how the physical object changed over time.

There is not, to my knowledge, any consistent standard by which different institutions transfer information about the original physical object to the online record associated with the digitized image.⁹ And that information about the photographic original can be hard to find, particularly in commercial archives such as Corbis, which make it easy to see a historical photograph of a specified event but difficult to learn the original photographer’s name or the source of the picture. Even the Library of Congress, the acknowledged leader in digitizing photographic archives, sometimes falls short in this regard. A quick survey of the Civil War images on the library’s extraordinary American Memory Web site, for example, reveals that one can determine whether a digitized image comes from a paper print or a glass-plate negative, but one cannot determine the size of the original. ¹⁰ Size matters, however. A stereograph’ for example—a double-image print that conveyed a three-dimensional effect when studied through a hand-held viewing device—would have been mass-produced, and that small card would have been less expensive and more widely circulated than a larger view. Marginalia and printed captions matter too. The Beinecke Library at Yale University remains a rare example of an institution that digitizes not only the complete face of a photographic object (including the edges of the mount and whatever notes might be included there) but the complete back as well. Any historian using the Beinecke’s Web site will be grateful for the seemingly mundane images of the backs of stereographs and snapshots. One finds there all sorts of useful information from the photographers, publishers, owners and users of the images. ¹¹ The ability to see the size, the edges, the marginal notes scrawled on the verso of the pictures, goes a long way toward alleviating the anxiety that one is missing something crucial by not holding the photograph in one’s hands.

It is critical to retain a sense of the photograph as a time-bound physical object, produced by a particular technology and circulated in a particular way. Consider war photographs. Unless we give some attention to technological issues, how can we expect our students to understand the enormous differences between photographs made of the Civil War, say, and the war in Vietnam or the ways our memories of those events have been shaped by the visual record? The almost total absence of actual Civil War battle photographs reflects the awkward and clumsy photographic technology of the times, not the absence of curiosity or ambition on the part of the photographers. The extraordinary My Lai photographs that Claude Cookman writes about here would have been impossible to make during the Civil War, not because civilians never felt frightened nor because soldiers never acted badly, but because nineteenth-century portraiture was a slower and more formal process that made it all but impossible to capture a fleeting facial expression or bodily gesture.¹² The evolving technologies of photography shape our records of war. We know World War II through visual records created with the new 35 mm cameras. We know the Vietnam War through color photography and television. And now we know the Iraq War through images captured on the cell phone cameras that enable every soldier to become a photojournalist and to send images around the world beneath the radar of would-be military censors. Digital reproductions of photographic images mute the important differences between the originals. But for all sorts of historical reasons, the physical form of the original continues to matter, because it reveals so much about the options available to the photographer and because it provides clues as to how viewers initially encountered and understood the image.

As we encounter photographic images online, it becomes increasingly difficult to understand the relationship between the digitized archive and the archive of original historical images. Few institutions digitize all their pictures, and the editorial or curatorial process that makes some images accessible online and others not generally remains invisible to the online user. On a visit to a photographic archive, one can feel some certainty that one has looked at the entire drawer of railroad photographs, for example, or all the prints by a particular photographer. But it is difficult to have that certainty when one consults an online archive. Might a curator have had a bias toward photographs of a particular neighborhood, a particular battle, a particular sort of farm work? Might she or he have mounted images online in response to institutional needs or the inquiries of a previous researcher? It remains very difficult to second-guess the relationship between most online archives and the physical archives from which they draw.

Moreover, when consulting digital images online, we cannot always be certain there was an original of any sort. The great documentary photographer Lewis Hine once said, “While photographs may not lie, liars may photograph.” ¹³ By this he meant that a photographer could exercise control over an image by deciding how to frame the view, how to arrange the subjects, when to snap the shutter. But he also meant to reiterate that what one sees in a photographic image once lay before the camera lens. That presumption about the physical existence of the scene preserved in the photographic view falls away in the era of Photoshop. The technologies that allow us to make accessible digital reproductions of historic photographs also allow us to invent digital images of scenes that never existed, except on a computer screen. This presents a serious problem for historians of the future. Increasingly, we will encounter all images in the same way—as digital images accessed on the computer. How will we know which is an invented digital image and which the digital copy of an image that also exists as a physical document, a source that provides corroborating evidence? Photographic manipulation is not new; photographers have long altered backgrounds, removed facial blemishes, combined two negatives to create a single print. What is new now, though, is the absence of any physical trail of evidence that lets the researcher see what was done.

Although the digital reproduction of photographic images raises questions about the value or meaning of the original photograph, older photomechanical means of reproduction have long raised similar issues. In photojournalism, for example, is the “original” photograph what the photographer made, or is it the photograph as it first appeared in print? During World War II, news photographers working overseas often shipped their unprocessed film back to the United States and saw their pictures only if and when they appeared in print.¹⁴ We might ask whether encountering a photograph in the pages of Life or Fortune is really any different from encountering a digitized image on the Internet; the magazine image likewise exists at a remove from the original and emphasizes visual information over the artful craftsmanship of an original print. Yet I would argue that the magazine reproduction of a picture differs from a digital reproduction in an online archive precisely because it appears in a particular historical context, given meaning by the words that accompany it, the layout of the page, the content of the publication, the audi ence targeted as buyers. As such, the magazine reproduction becomes a historical object in a way that an online reproduction—disembodied and severed from its historical rooting, made available for an infinite array of uses—is not. As Hariman and Lucaites argue in their essay on the “Times Square Kiss,”

A better understanding of icons, public arts, and public culture alike requires that one place the images in a history of production, circulation, and subsequent appropriation. The “last” stage, appropriation, is particularly important. As images are taken up, altered, or otherwise used persuasively by politicians, protesters, designers, artists, advertisers, journalists, and other people, and as they are displayed on billboards, T-shirts, murals, computer screens, and dozens of other surfaces, one can trace the making of public culture.¹⁵

In voicing my worries about digital picture archives, I do not mean to sound like a Luddite, unwilling to engage anything except the original photographic artifact. I simply mean to suggest that when we lose sight of the real thing—whether a daguerreotype, a magazine reproduction, or a T-shirt printed with a photograph—we lose an important opportunity to reconstruct this rich and vibrant “public culture.”

Ultimately, the question is whether the digitization of images and their new accessibility over the Internet fundamentally alter the meaning of the original pictures or whether they instead give us greater access to those original meanings. Should we approach digital archives simply as old photo archives made newly accessible or imagine that they represent a different form of information altogether? Either way, online images seem here to stay and historians need to think hard about whether old modes of inquiry will suffice or whether we must develop new interpretive skills to understand this new visual medium.

Anyone who teaches with photographs quickly discovers how provocative and engaging they can be for students, who inevitably discover something in the image they can connect with their own experiences. And yet we must ask whether all readings of historical photographs are equally useful, not simply as affirmations of personal experience but as insights into the world depicted by the photograph. Certainly, not all interpretations of a photograph can be verified by outside evidence, a basic tenet of any serious historical study. The responsible historian must treat a photograph as skeptically as any other sort of evidence, examining it as an artifact as well as an image (the very thing made harder by the digital turn) and considering it both in and through time, as an object whose multiple meanings mutate as the stories the photographers and subjects tell change and fall away, as new events occur, as the photograph itself appears in different contexts to new audiences.

As Michael Lesy argues in his essay, few historians are trained to be visually literate, and most feel unnerved by the possibility that any reading of an image is as good as any other. “Depending on who is looking and when, an image changes its meaning: Look now, see a rose. Look again, see a butterfly.”¹⁶ What must we know about a photograph to formulate an intelligent reading of it? Lesy proposes we become photographers. Eric Sandweiss indirectly argues for understanding the photographer’s life. The photographers writing here argue for the importance of their own intentions. Natanson, the archivist in the crew, makes a case for the importance of assembling related historical data. Each offers an important perspective. But collectively they make a case for the powerlessness of a photographic image (how can it possibly need so much research?) even as they argue for its extraordinary power.

To what end should historians engage photographs? Claude Cookman writes that “great photographs let us empathize with their subjects,” suggesting that photography sparks a human empathy between the viewer and the subject, between the present and the past. Certainly, such empathy is a key element of historical imagination, that quality of mind so necessary to any creative and understanding interpretation of the past. And yet the idea that a photograph should, as Cookman argues in his essay about the My Lai pictures, provoke a moral response that helps us “reflect on our responsibilities as humans and as citizens” also seems at odds with the conventional rules of historical inquiry, which demand that the historian reach beyond a personal response to a historical document to develop an interpretation that can be verified by others.¹⁷ A purely moral response to a picture remains contingent on each individual viewer’s own experiences, training, and cultural assumptions. In his essay about his work in Vietnam, Ted Engelmann reveals how intensely private one’s formative experiences can be and thus how very difficult it can be for one person to interpret a photograph precisely the same way as someone else. Indeed, photographs may be stable objects, but they have unstable meanings, shaped by their makers and their subjects, their immediate audiences and their subsequent viewers. In this sense, as in so many others, they resemble the literary documents with which historians generally feel more at ease. They answer the particular questions we ask of them, and as those questions change across time, the evidentiary value of the photographic document shifts as well.

For all the difficulties in interpreting photographs as historical documents, historians ought to find something deeply familiar about the task. Photographs are intrinsically historical; they necessarily capture a specific and fleeting moment of time. And making a photograph has much in common with the practice of crafting a history. One begins with a curiosity about the world, sifts through the evidence till an idea begins to emerge, and then carefully arranges the evidence for presentation to a broader audience. Photographer and historian alike are storytellers who must choose what to include and what to leave out, how close to stand to their subjects, how to frame their tales. Historians might sometimes feel confused about the photographer’s craft: What is objective science, what a more subjective art? But historians balance the same tensions. “History is not alone facts, not alone ideas, but facts in their relation to ideas,” the historian Hubert Howe Bancroft wrote.¹⁸ Photography is much the same thing. And in that, even the wariest historian can take comfort.