“About or Is?” Describing Archival Photographs

Institutions in the United States have long recognized the importance of acquiring and collecting photographic materials as documentation. One of the earliest and most widespread uses of photographs in an institutional context was the capture and recording of war activities. Writing in The American Archivist in 1958, Joe Thomas told the story of an 1860 U.S. expedition to Japan that used daguerreotypes to document governmental activities. He noted that the expedition was the first time that “an agency of the United States Government in connection with the transaction of public business had consciously created photographic records that were appropriate for preservation as evidence of its operations or as the embodiment of valuable information.”⁴ The U.S. National Archives began collecting photographs from federal agencies shortly after its inception in 1934 and even housed a separate division called Photographic Archives.⁵

Despite the long history of collecting and cataloging photographs, the development of archival descriptive standards specific to visual materials has been notably slow.⁶ Allen C. Benson pointed to the “general reluctance to recognize the unique and complex nature of archival photographs, collections, and image-constructed knowledge.”⁷ Joan M. Schwartz argued that by “embracing a textual model of recorded information and by adopting a bibliographic model of image classification, archives continue to fixate on the factual content rather than the functional origins of visual images.”⁸ Jane Greenberg agreed that current descriptive standards place constraints on access to visual materials. She suggested that metadata schema design should support functions across domains (such as discovery of images, regardless of disciplinary background) rather than focusing on designs to meet the needs of only one domain.⁹ According to Tim Schlack' s historical analysis, changing cultural conceptions about the value of archival photographs played a role in making access and preservation use cases more compelling for institutions.¹⁰

In terms of specific guidance, the literature suggests that archivists aim to develop visual literacy skills to improve description efforts. Though writing primarily about photograph appraisal, William H. Leary advised that archivists processing visual materials should be both “students of history” and “student[s] of the history of photography.”¹¹ Elisabeth Kaplan and Jeffrey Mifflin advocated for archivists to familiarize themselves with visual communication methods and borrow aesthetic approaches from photography, film, and video to formalize literacies.¹² In Photographs: Archival Care and Management, Helena Zinkham provided an in-depth discussion regarding how archivists might describe photographs. She instructed archivists to study both photographs and negatives carefully, noting any written information on the materials themselves or on their containers. Potential reference sources such as pictorial histories, histories of photography, and photography dictionaries, directories, and databases can help identify relevant details.¹³

The burden for archivists lies in adequately describing the archival photograph within changing temporalities for archival spaces that exist in perpetuity. As Schlack wrote, “Working with photographs then is a process of substantiating the most valid narrative that a photograph collection can evoke and transcribing it into the academic discourse of our time.”¹⁴ Given the relationship between archival description and access, the archival profession would benefit from user studies that explore aspects of interactions with visual materials. What do users need from the presentation and display of archival photographs in digital environments? What factors matter to them as they evaluate sources? Understanding the dynamics at play in the space of interactions is an important part of ensuring that descriptive models are an adequate fit for enabling access to visual materials.

How Does Library and Information Science Study Image Use?

Visual information (and visual methodologies) is an underexplored phenomenon in library and information science (LIS), though recent work indicates this may be changing.¹⁵According to Joan E. Beaudoin, developing a comprehensive understanding of scholarly image use is challenging for LIS simply because “aspects surrounding use are markedly different than those of textual materials.”¹⁶ Images are polysemic by nature: meaning depends on the viewer, and the same image may have multiple interpretations at the same time.¹⁷

Overwhelmingly, visual information behavior studies are quantitative in nature and tend to focus on visual search.¹⁸ A popular method used in studies is query analysis, which attempts to understand visual information needs by analyzing the kinds of search queries submitted to different corpuses. For example, Linda H. Armitage and Peter G. B. Enser collected, analyzed, and classified approximately 1,800 queries to anticipate user information needs for image-related materials at seven libraries.¹⁹ While acknowledging the value in creating image classification schemas, they nevertheless pointed to the difficulties in classifying queries with vague geographic locations (e.g., “London scenes”) or singular topics that may include many types of representational formats (e.g., paintings, photographs). Karen Collins also looked at user-submitted queries in two photographic archives to assess how well information systems were meeting the needs of users searching for visual materials.²⁰ Despite the perceived importance of provenance as an archival access point, all of the patrons in her study sought images individually rather than in the context of a collection. Noting this discrepancy, Collins concluded: “While maintaining the context in which images were created or used is necessary to preserve their evidential value, it is clear that few patrons are presently using images as primary source documents. A study of how patrons use images, and the implications for archives, would be interesting and useful.”²¹ Using the Library of Congress's American Memory Collection as their corpus, Youngok Choi and Edie M. Rasmussen analyzed the relationship between user-submitted search queries and participants' statements of information needs.²² One interesting finding in this study is that, given a set of images with accompanying metadata, study participants ranked “date,” “title,” and “subject descriptors” as the most relevant attributes—a noted difference from textual retrieval environments.²³

Visual information behavior studies also employ task-based scenarios to understand the context in which users submit queries for visual materials. For example, Raya Fidel analyzed and classified a hundred queries submitted to a stock photography agency, noting semantic differences between user queries motivated by informational purposes (e.g., an icon for a handicapped entrance) versus images serving as representations (e.g., a colorful picture).²⁴ These conceptual differences in query construction suggested to Fidel that the nature of the retrieval task might influence search behavior. Lori McCay-Peets and Elaine Toms also focused on task-based work contexts, interviewing thirty journalists and historians to understand motivations in using images.²⁵ Building on Fidel's work, McCay-Peets and Toms were interested in classifying whether participants used images as “illustration” or “information.” They suggested that visual information use studies should distinguish between what motivates image acquisition (physically acquiring the item) and image use (purposefully using the item). Marjo Markkula and Eero Sormunen also examined the practices of journalists working in a digital photo newspaper archives in Finland, identifying the different types of criteria used to determine relevancy.²⁶ However, they concluded that user queries from image searches revealed very little about user behavior; instead, they noted that observation sessions from the newsroom provided a more illuminating understanding of context for image needs and uses.²⁷

A small number of studies have employed qualitative approaches in their explorations of digital image collection use. Noting a gap in the literature on attempts to understand visual meaning-making practices in archival settings, Paul Conway and Ricardo Punzalan explored how expert users interacted with digitized photographs.²⁸ They developed a model called “Fields of Vision,” identifying three different modes of inquiry (storytelling, discovering, and landscaping) to help contextualize motivations for participants' use of digitized photographs. In another qualitative study, Kathleen Fear examined how users working with digital image collections perceived various Dublin Core metadata elements.²⁹ Although the stated purpose of the study was to examine search behavior, Fear's research offers contextual, nuanced descriptions of how users interact with digital image collections. In an effort aimed at understanding image use across disciplines, Joan E. Beaudoin investigated four professional groups (archaeologists, architects, art historians, and artists) to note similarities and differences in practices.³⁰ The most common motivating factor shared across all four groups was “knowledge development.”

Conceptualizing Information Use as Experience

Quantitative methods can be beneficial for understanding aspects of image use including useful baseline information about resource use. Yet such approaches can over-emphasize information access points—or how scholars locate particular objects using a specific information channel for object retrieval.³¹ Treating access and delivery mechanisms as proxies for use obscures important aspects of the information experience. As Laura Sheble and Barbara Wildemuth pointed out, transaction logs “cannot tell us anything about the users' cognitive or affective responses during the system interaction.”³² Citation analysis also presents an incomplete picture of use, as citation practices for digitized archival materials are continuously evolving. Databases like the Web of Science also do not currently provide interfaces that permit citation searches by material type, making citation analysis for visual materials like photographs particularly difficult.

User needs can also shift in the course of a single interaction. Quantitative approaches that count access to a resource do not capture dimensions that may in fact deter use. In a study of visual information seeking among video users, Dan Albertson noted, “Visual information can influence outcomes in an interactive process and presents variations in the types of needs, tasks, considerations, and decisions of users, as compared to information seeking in other contexts.”³³ In other words, ecological, emergent factors can mediate experiences significantly in ways that may not accurately reflect actual outcomes of use. If we only study cases where resource access signifies successful use, what can we learn about the characteristics that make environments usable and useful for scholars? The studies reviewed in the literature provide very little insight into critical dimensions related to scholarly use (or nonuse) of materials.

Investigating how scholars interact with visual materials can be helpful for revealing the emergent qualities and attributes that make that experience meaningful for participants. Ronald McIntyre and David Woodruff Smith explained:

Photographs, symbols, and words in themselves and apart from the meanings and interpretations given them by persons or other creatures possessing mentality, are only so many marks on paper. Their intentionality—their “representing,” or being “of” or “about” things other than themselves—is therefore not a character they have intrinsically, insofar as they are merely the physical objects that they are, but is derivative from their relation to intentional mental states.³⁴

Following this argument, visual representations become information through the relationships and meanings we inscribe onto them. Analyzing our experiences of things, then, can help us understand the contexts in which these things become meaningful.³⁵ In this research study, exploring how and why scholars used digitized archival photographs helped to reveal the factors and qualities that mattered to them. Underlying these experiences are important insights about how knowledge practices are changing in digital research environments.

Methodology

In the spring and summer of 2015, I conducted semi-structured, in-depth interviews with fifteen self-identified historians. I used three avenues to recruit potential research participants: (1) professional mailing lists, online forums and Twitter;³⁶ (2) individuals at five local institutions;³⁷ and (3) individuals recommended to me through colleagues and other recruited participants.³⁸ All recruitment invitations specified that to qualify for inclusion in this study, participants must have previous experience using digitized archival photographs in scholarly activities. Scholarly pursuits could include teaching, publications, presentations, or research projects. I posted initial recruitment invitations online or emailed them to potential participants, depending on the targeting strategy.³⁹ Recruitment strategies happened simultaneously, though the third strategy (snowball sampling) continued throughout data collection and analysis.

Interview sessions took place with recruited participants at settings of their choice, provided they were equipped with Internet access. Recruited participants came from a variety of academic departments, including history, English, African American Studies, American studies, classical studies, and musicology. Prior to interviews, each participant selected and submitted at least two examples of digitized archival photographs they had used in scholarly activities such as research and teaching. Material types included conference presentations, class presentations, course syllabi, dissertation chapters, and journal articles.

Table 1. Study Participants by Title and Primary Area of Research

View Fullscreen

The goal of this research was to understand and characterize historians' practices using digitized photographs as evidence in scholarly activities. In my methodological approach, I adopted a phenomenological stance to enable a focus on understanding “how persons construct meaning” by examining their particular experiences with certain phenomena.⁴⁰ The use of multiple types of data including examples of historians' specific uses helped to facilitate a complete picture of my phenomenon of interest (i.e., historians' experiences using digitized photographs as evidence).

I used Skype to conduct the majority of my interviews. I recorded each interview digitally and then transcribed them verbatim. On average, most sessions ran 1.5 hours. In the first part of the interview, I asked participants to provide demographic information and to answer questions about their image use practices.⁴¹ In the second part of the interview, I instructed participants to open a Web browser to the online location of their submitted photographs (see Figure 1).

View Figure

• Download as Powerpoint
• Download Figure

In my interviews, I drew on techniques from photo elicitation interviewing (PEI)⁴² and concurrent verbal probing⁴³ to ask participants about how they evaluated, selected, and used materials in scholarly activities. The PEI portion of the protocol contained fifteen interview questions with several embedded probes.⁴⁴ I used open coding and a hybrid approach to thematic analysis⁴⁵ to surface salient aspects of participant experiences. To strengthen and verify the analysis, I employed multiple types of triangulation. The goal was to assemble all participant materials and begin to formulate and approach the collection as the building of each case. During both data collection and analysis, I took memos as a strategy to “trace the thinking of the researcher and help guide a final conceptualization that answers research questions (or related ones) and offers a theory as an explanation for the answers.”⁴⁶

Findings

Overall, this research identified sixteen broad themes related to historians' experiences using digitized archival photographs. The findings shed light on both the practical concerns and the intellectual challenges that surfaced for historians working with visual materials in digital research environments. Each case demonstrated the extent to which material conditions in interactions (rather than just tools used to discover or access resources) influenced visual information use. Table 2 is a typology of themes that characterize historians' use, alongside examples of descriptive quotations from interview transcripts.

Table 2. Sixteen Themes of Experience

View Fullscreen

Table 2. Continued

View Fullscreen

The results of this study suggest that distinct practices remain critical to historians' image selection and use. Broad visual browsing of collections is an integral part of the research process, with participants initially approaching a body of images to cycle through and then locate potentially relevant material. P7, a historian of nineteenth-century American visual and material culture explained, “I wouldn't do keyword searches, because I know that keyword searches are going to limit what I find. And if I'm looking for everything, you know what I mean, if I don't know what I'm going to find yet, then I don't want to limit what I see.” Participants also used photographs to orient their research in two distinct ways. First, they used online image access initially to determine whether to make an in-person visit to a repository. Second, they downloaded interesting images “to get the lay of the land” at an early stage of research for later use as they may not “realize right away which images were going to stick out.” Participants were motivated to begin research projects after just being “fascinated with the images” and proceeding to see “what story would develop out of it.”

The affective power of photographs also motivated image selection and use.⁴⁷ Two participants described how their use of photographs “would foster empathy” or “make it more exciting for students.” P2, a historian of alternative green culture, explained, “I look for anything that catches my eye . . . and I don't always know what I'm looking for, but I'm looking for anything interesting and unusual.” Respondents also noted how affective dimensions of photographs assisted in translating historical concepts in pedagogical efforts. P11, a historian of music who has taught about the minstrel shows in the era of Jim Crow, described how photographs engage students in ways that listening to music often cannot:

You know, if we're listening to music from minstrel shows, they're very, very racist, horrifically racist, but college students are always really in tune to—they don't understand the racist images—or the racist language—they don't always understand that “Jim Crow” is a racist thing, or whatever. But you see the image of Jim Crow, that's really obvious. And then when you hear it, and so then I can say, “Well, this image, you can see how racist this is, this is easy to understand.” But in the music, they get sort of caught up in how can music be racist, and how can you have musical markers for racism, and how is this language racist?

Photographs also provided valuable historical reference for verification, documentation, or corroboration. They helped to contextualize a specific time and/or place of interest, for fleshing out “what things looked like then.” For example, P3 is a public historian who used archival photographs to re-create accurate details in restoring details to furnishings and layout in Orville Wright's mansion, the Hawthorn Hill house. Similarly, P4 used archival photographs to justify restoration recommendations for historic preservation work, noting how “a lot of times we need to look at these historic photos to actually know what it actually looked like and not just our idea of what it looked like.”

As a form of visual evidence, photographs prompted participants to make deeper historical connections. In several cases, the discovery of particular photographs helped to counter existing historical perspectives. P6, a public historian studying labor movements, described how the discovery of photographs showing African American women participating in the 1930s St. Louis nut processor strike led to a reexamination of the event:

So I knew that this strike had happened, but to find the images really drew me in deeper. And it was about the nut processor's strike in the thirties, where these African American women went on strike, and they were just literally unseen by the whole society. They were working longer hours than the Polish women who worked in the factory, and they worked in the basement, so some of the Polish women did not even know that there were African American women present in the factory . . . when this strike happened, it was really about them becoming visible for the first time. And this massive strike taking place in St. Louis, and really changing their conditions as they became unionized. Because the way African American women were able to go out on strike, is they aligned themselves with the unemployed councils, which were black women in the neighborhood. And white women in the neighborhood. So, they got white women to come to their picket lines. So when the white women saw the strike going on, they saw white women out on the picket line and joined the picket line. And it was images of that that really drew me into the fact that there was a whole, much more extraordinary history linked to the power of images, that hadn't been told before.

Several participants mentioned that addressing these absences in the historical record became an explicit motivational goal for further image use.

One significant change in contemporary historical practice is the amassing of personal collections of digital sources. Scholars can now use portable digital cameras to snap images in archival reading rooms, download digitized online sources, and/or use home scanning equipment—all factors that have undoubtedly contributed to the making of new forms of research collections.⁴⁸ Melissa Terras noted this trend in 2010: “Enthusiastic digitisation by amateurs, a phenomenon previously ignored by information professionals, is providing a rich source of online cultural heritage content which often documents areas not covered via traditional institutions.”⁴⁹ P1, whose dissertation work on African American musicians' self-expression in early twentieth-century Alabama, described a hybrid research process that included both print and digital sources as well as archival and nonarchival access points. For example, P1 used Google image search initially to browse images based on already-established criteria (in this case, pictures of musicians in certain poses), in turn facilitating a connection to source materials found on independent scholars' blogs. P1 mentioned tracking down a photograph that originally appeared on the cover of a folk album, having previously encountered it on the cover of a book on West Virginia fiddle music. P1 was thus able to “circumvent the traditional archive” by obtaining a digital scan of the photograph from the author of the book directly.

These new pathways can introduce additional challenges for scholars who may be used to relying on an archival repository for authentication and sourcing of materials. P11 discovered first-hand the complexities that can arise from locating an unsourced photograph online. In a PowerPoint Presentation for a class, P11 used what the participant thought was a digitized photograph of Anna Held, a performer from the early twentieth century. P11 initially found the photograph via Google Image Search, on a blog dedicated to vintage photographs of women at the turn of the twentieth century. The original photograph, digitized as part of the Macauley Theatre Collection at the University of Louisville Photographic Archives, tells a different story. According to its title, the woman pictured in the photograph is actually one of Anna Held's fencing girls.⁵⁰

The concept of trustworthiness played an important role in participants' decision-making. Factors such as the reputation of an institution or that an image originated from an archival institution helped to establish trust in using that image further. The notion of the archives as a harbinger of truth repeatedly surfaced in many participant descriptions, with P8 asserting, “I generally tend to take what the archive tells me as truth.” Additionally, personal relationships with archivists also helped to ensure trustworthiness. P15, a historian of the nineteenth-century American South, was more willing to use a photograph in personal scholarship because P15 knew the archivist who helped to digitize it, as well as the original donor's family.

In interactions with digital environments, functional aspects such as an ability to download images at different resolutions played a significant role in determining further use. Respondents wanted as much original descriptive information as possible (for example, the handwritten captions on the back of photographs) to be made available. P10, a historian who has studied the photographs of Lewis Hine, described the benefits of archival access and delivery thusly: “You know, having everything, having all the captions and having everything means you don't actually have to visit the repository. We can do it from a distance.” Indeed, the presentation and display of contextual information such as keywords, subject headings, original medium, size, and photographic captions were all important influences on selection and use, regardless of format preference.

Material aspects of interactions with photographs mattered for participants who used both digital surrogates and analog counterparts. Six participants stated an explicit preference for experiencing photographs in digital or analog formats (four preferred digital; two preferred analog). Digital surrogates provided possibilities for seeing intricate details, including those that might not be discernible in an analog counterpart. At the same time, examining photographs as analog objects prompted serendipitous discoveries. P2 discovered new details when viewing a photograph with a magnifying glass, altering the formulation of a research argument:

So, there are a number of things that I never would have seen, I think. . . . you know, the physical image has a whole affect to it that you don't get from a glowing screen, I think. Especially these nineteenth-century glass-plate collodion-albumen prints. And I just wouldn't have seen it—I've seen it, now that I know it's there, I can look at the digital image from the Beinecke, and zoom and zoom in to see it, but I never would have seen it if I was just looking at a 2-inch-by-3-inch digital version.

Regardless of analog or digital status, participants fluidly recalled interpretive choices and decision-making, in ways that cut across techno-social dimensions of their viewing experiences.⁵¹ For example, one participant noted that the ability to zoom in (a function of the viewing environment) was as useful and relevant as the photographic content itself. In other words, participants situated their descriptions within their specific viewing experiences, rather than in their interactions with the photograph itself. This embodied perspective suggests a form of digital visual literacy that places significance on the activity of the viewing experience itself.⁵²

Limitations and Future Work

This research study had numerous limitations. One limitation concerns the relatively small sample size (n = 15). There are no strict guidelines regarding sample size in qualitative case study research; instead, the emphasis is on the information-richness of cases. Toward that end, I used purposeful sampling to target information-rich cases and observed the principles of data saturation in my data collection. Relatedly, case study research is typically not generalizable; the goal of case studies is not to “infer findings from a sample to a population, but to engender patterns and linkages of theoretical importance.”⁵³

Another limitation concerns the interdisciplinary nature of my sample population. Despite using the umbrella term “historian,” the majority of my study participants were not teaching in academic history departments. However, prior to data collection, I made the decision to expand my sample population for a number of reasons. First, it was difficult initially to locate individual historians working in academic history departments who were using digitized photographs to make historical arguments. Through informal conversions conducted in the early stages of my research, I learned that institutions were not yet formally tracking scholarly use of their digital collections, preventing identification of potential recruits.⁵⁴ My particular focus on studying scholars using photographs from archival collections (as opposed to all visual materials) proved to be an additional challenge. Since I was most interested in how and why historians use digitized images as forms of evidence, expanding my sample population to encompass self-identified historians doing historical work allowed me to gather important aspects of those interactions while providing me with an adequate number of participants for case study research. I found great value in addressing the broad practices of individuals doing historical work with photographs, recognizing that my sample had unique qualifications and training. In fact, though most of my participants were not teaching in academic history departments, eight of fifteen had received their doctorates from “traditional” history departments.

The findings of this research paint a changing picture of the ways in which historians interact with archival sources. The formerly bounded physical experience of discovering, accessing, and using materials in an archival repository now has vastly different dimensions. Rather than designing for specific access points or information retrieval strategies, archival professionals should consider how use of digitized archival photographs is a multifaceted, multilayered experience in which meaning is constructed (or not) through a series of complex mediations. Archival research environments should aim to support fluid, generative interactions with materials rather than focus on accommodating specific access and use points. Future research should emphasize the information experiences of historians to holistically understand interpretive techniques in digital research environments and changes in critical scholarly practices. In particular, the role that affective factors such as empathy play in the selection and use of visual information is a promising area for further exploration.⁵⁵

This study identified potential shifts in where historians are locating photographs for use in their scholarly activities. Only one participant mentioned finding aids at all, while eight participants mentioned using Google image search. The emergence of different kinds of digital collections (e.g., amateur/hobbyist online collections and personal digital archives) might indicate that the institutional archives is not the starting point for image search. While trust in source materials remained a central concern for my participants, this study did not explore whether and how historians evaluated photographs located exclusively through Google image searches. Standardizing practices around the embedding of metadata into digitized photographs, as well as tracking use throughout the life cycle of images, remain two crucial areas of future research.

Although the study sample does not represent all historians, my findings suggest that historians are more comfortable using photographs as primary sources of evidence than previously anticipated. A significant number (thirteen) used digital surrogates as the basis for making historical arguments—in contrast to assertions made in both the historiographical and LIS literature regarding image use. This discrepancy points to the need for developing better metrics for understanding and analyzing use and users of digital collections.

Historians and archivists would benefit intellectually from sharing information about evolving work practices. The participants in this study wanted to know more about the appraisal and processing decisions archivists make and to see that information reflected in the archival record. At the same time, observing historical work in situ could help archivists learn more about historians' scholarly practices. Attending each other's professional conferences to serve on panels or at least perusing disciplinary journals would also help familiarize each constituency with emerging trends.

In this research, mediating factors influenced historians' selection and use of particular images, from how they first noticed and appraised materials to the obstacles that prevented them from using materials further. Regardless of digital or analog status, the process and practice of interpreting and using visual materials as evidence in scholarship is a complex endeavor that requires motivation and skill. In fact, it became clear that historians employ particular rhetorical strategies and perspectival orientations to construct their evidential case. Nearly all of the historians in this study named particular spectators in constructing their arguments. For example, four participants argued that photographs demonstrated specific evidence of public opinion from the time period. Exploring the strategies and tactics employed by historians in the making of historical evidence can prompt further development of discursive methods, helping to mature and broaden methodological scope in LIS research.

How can archivists and archival information systems best accommodate evolving information needs and practices amid such changes? A particular challenge lies with the difficulty in describing visual materials; photographs can have many interpretations, and this range in interpretive possibilities in turn has ramifications for description and use. What is to be done? Writing in 1988, Eric Margolis asserted, “In fact, the meaning of a particular photograph typically is generated as much by the context in which it is found—site, collection, date, photographer, caption—as by the image itself.”⁵⁶ Margolis underscored the importance of grounding study in information experiences to understand meaning making in digital environments. The value in studying these spaces is that they can reveal how information emerges rather than focusing on defining what information is. The latter is well-trodden territory, but orienting toward the former is critical for developing new understandings of, and literacies for, visual information use.