Methodological Approaches to Disclosing Historic Photographs

Eric Margolis and Jeremy Rowe

NOTES ON HISTORIC provenance: they were made by known or PHOTOGRAPHS important photographers; they are accompa- nied by collateral information, such as writ- Our research draws from the genre often ten documentation; or they are well-known called ‘historic’. In a real sense all photo- and have been studied by generations of graphs are historical; they are two-dimen- scholars. Other photographs are drawn from sional representation of scenes captured with the much larger category of ‘vernacular’ pho- lenses, and frozen in a fraction of a second. tographs. Such images come from more From the instant of exposure, the photograph obscure sources and usually offer little docu- recedes into the distance of time. However, mentary information to provide a ‘warranta- for the purpose of analysis and discussion, ble’ understanding of the photograph, other this chapter uses a conventional historians’ than that provided by the image itself. We definition of ‘historic’ photographs as being may be provided with only observable infor- 50 years or older. mation such as the size, format, and subjec- Like the photographs in Barrett’s descriptive description. Collateral information such tive and explanatory categories discussed as date, location, photographer’s name, sub- below, people tend to accept historical photo- ject, or reason for being made have been lost graphs at ‘face value’, as accurate, indexical or are unavailable to the researcher. reflections of reality. Historic images have The vernacular genre, as popularly been concentrated in libraries, museums and described, includes indigenous or ‘native’ archives, and have become increasingly pop- photographs, typically made by unknown or ular for illustrating books and used in ‘docu- amateur photographers that tend to depict mentary’ motion pictures. There are also less common subjects, objects, family, and events visible, but very active markets for sale and of daily life. There are literally millions of private collection of historic images. Some these vernacular historical photographs, historic photographs come with extensive which are becoming more accessible and

55632-Margolis-18.indd632-Margolis-18.indd 333737 33/14/2011/14/2011 4:04:144:04:14 PMPM 338 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

widely available to researchers. Access to photographs originally made for reasons vernacular historic photographs previously other than research. In either case, by them- required time-consuming and costly trips to selves photographs provide only ‘thin’ archives. Increasing numbers of large-scale descriptions, but this information can prove digitization projects, the resulting online helpful in constructing our own ‘thick’ access to these collections, and social media descriptions (Geertz, 1973). Erving Goffman and sharing sites like Flickr, Photobucket and (1976) emphasized the human ability to Picasa are making millions of vernacular make categorical inferences about the images readily available to researchers. glimpsed world in ‘real life’ and the glimpses Many sites also solicit and share input provided by photographs: about the images that can assist in identification and interpretation. Additionally, traffic To glimpse a world ... is to employ a set of categories more or less distinctive to glimpsing and often in buying and selling photos has moved from entirely adequate for the job they are designed to swap meets and estate sales into regional do (22).1 trade shows and conferences like the National Stereoscopic Association, the Daguerreian In the essays that introduce Goffman’s (1976) Society, and Papermania, and to large-scale study of the portrayal of gender in advertis- online auction sites like eBay and Delcampe. ing images, he makes the point that photo- These images, and where still extant, the graphs − posed or candid − are of the same associated data concerning their provenance, nature as a glimpse of ‘real life.’ constitute a significant resource for all the human and cultural sciences with tremen- We are all in our society trained to employ a some- dous untapped potential for researchers. In what common idiom of posture, position, and glances, wordlessly choreographing ourselves rela- what follows we explore ways to investigate tive to others in social situations with the effect the meanings of historic photographs. that interpretability of scenes is possible. ... however posed and ‘artificial’ a picture is, it is likely to contain elements that record instances of real things (21). INTRODUCTION: TO GLIMPSE A WORLD THAT WAS Visual ethnography has an academic interest in analyzing historical photographs to learn Clifford Geertz saw the strength of ethnogra- as much as possible about the way the world phy as its ability to ‘put us in touch with the was. The study of historic photographs builds lives of strangers’ and to ‘see things from the upon the same foundational issues as other other’s perspective’ (Geertz, 1983). In con- visual research techniques. There are also fronting our own history we similarly try to different analytic approaches, such as the fathom the lives of strangers. Traditionally, ‘postpositivist’ and ‘hermeneutic’ paradigm historic research primarily incorporated communities (Kuhn, 1970). Berger and Mohr document analysis and oral history. But as (1982) noted the connection of photography culture has become more visually literate, both to the modernist project of positivism photographs, graphics, and other images and to the discipline of sociology: offer the historian/ethnographer an addi- The camera was invented in 1839. August Comte tional window on the ‘webs of significance’ was just finishing his Cours De Philosophie Positive. that Geertz (following Max Weber) said Positivism and the camera and sociology grew up comprised culture (Geertz, 1973: 5). Some together. What sustained them all as practices was researchers have adopted and use cameras the belief that observable quantifiable facts, recorded by scientists and experts, would one day as recording mechanisms to provide their offer man such a total knowledge about nature own images for analysis. Others, the and society that he would be able to order them authors included, emphasize ‘found object’ both (99).

55632-Margolis-18.indd632-Margolis-18.indd 333838 33/14/2011/14/2011 4:04:144:04:14 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 339

This chapter presents two approaches to pho- ‘naturalistic’, or ‘superstitious and naïve’) that tographic research, each presented in the photographs have a special causal and structural relationship with the reality that they repre- voice of one co-author. In general we believe sent (282). that potential synergies exist in approaching photographic research from perspectives of We don’t necessarily agree with Mitchell’s both postpositivist evidence from indexical concern that positivist beliefs in images have and iconic sources, and interpretivist/herme- been killed off by postmodernists like Victor neutical approaches that draw on symbolic Burgin. As discussed later, Errol Morris, dimensions that are essential in the examina- among others, maintains a firm distinction tion of photographs. Each brings tools and between words and photographic images in lenses for viewing and analysis of photo- his analysis, and argues that photographs do graphs that are neither mutually exclusive have a ‘special’ relationship with what stood nor individually exhaustive. before the lens. While Margolis leans toward theoretical Currently, there are two vibrant meta- hermeneutic approaches, Rowe applies a physical conceptions of photographs. The postpositivist evidentiary process that he first we’ll term ‘postpositivist’ in that, while terms photographic forensics. In a previous confident in the indexical relationship co-authored piece about an album of Arizona between a photograph and the material world Indian school photographs, the co-authors in front of the lens, assertions about photo- found that each technique added value to the graphic meaning have two limitations: analysis: the whole was clearly more than the sum of the two parts (Margolis and Rowe, 1 These assertions are statistically probable, not 2002). This chapter continues the dialogue certain. between and about the two approaches to 2 They must be subject to rigorous testing and photographic analyses, providing general must be in principal falsifiable if they are to background to each approach, and brief be considered scientific (Philips and Burbules, examples that catalyze the discussion. 2000). Today, there is a growing awareness of the limitations of the strict positivist position that The photographic researcher/critic makes considers photographs as merely a ‘mirror ‘warranted’ statements backed by evidence − with a memory’ (as daguerreotypes were but the warrants can be extended or even labeled soon after their introduction in 1839). overturned at any time by new evidence. Similarly, few argue that cameras function The second conception emphasizes ‘inter- merely as ‘pencils of nature’, a popular pretivist’ views. In contrast to the postposi- implication in the nineteenth century. tivist perspective that considers photographs Englishman William Henry Fox Talbot pro- as documents that convey descriptive or duced the first book illustrated with photo- explanatory representations, merely reflect- graphs in 1844. Talbot laid the foundation for ing reality, or even representing ‘the thing in this interpretation with his introduction: ‘... itself’, John Tagg (1988), following Foucault, the plates of this work have been obtained by emphasized photography in terms of associ- the mere action of light upon sensitive paper. ated power relationships: ‘Photography as They have been formed by optical and chem- such has no identity. Its status as technology ical means alone, and without the aid of any varies with the power relations that invest it. one acquainted with the art of drawing....’ Its nature as a practice depends on the institu- (Fox Talbot, 1844 [1989] Book No. 1, np). tions and agents which define it and set it to W. J. T. Mitchell (1994) contended that: work’ (63). Hermeneutic perspectives emphasize photographs as texts, demanding seman- It is getting increasingly hard to find anyone who tic and semiotic interpretation to determine will defend the view (variously labeled ‘positivist’, meaning.

55632-Margolis-18.indd632-Margolis-18.indd 333939 33/14/2011/14/2011 4:04:144:04:14 PMPM 340 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

Postpositivist approaches to explanatory, interpretive, ethically evalua- researching historical photographs tive, aesthetically evaluative, and theoretical. Much of Barrett’s conceptualization can be Postpositivism emphasizes that photographs applied equally well within our framework represent ‘things in the world’. It recognizes for researching historical photographs. The that cameras are not simply mechanical tran- first two categories, for example, are useful scription devices, and acknowledges that pho- when discussing how historical photographs tographs result from photographers infusing emphasize a postpositivist approach to their own perspectives and interpretations of research. As Barrett explained, ‘The photo- subjects through decisions about framing and graphs are falsifiable in that potentially they composition, by manipulating depth of field could be empirically demonstrated to be true and exposure time, choosing when to release or false, accurate or inaccurate’ (Barrett, 1986: the shutter, etc. Oliver Wendell Holmes com- 56). Barrett defined these photographs as: mented in 1859: ‘The photograph has com- pleted the triumph, by making a sheet of • Descriptive: photographs such as crime scene paper reflect images like a mirror and hold photos, X-rays, portraits, and photographic them as a picture … the mirror with a reproductions of art. 2 memory’. An anonymous quotation plays on • Explanatory: includes photographs made by this by inserting the importance of the human visual sociologists and anthropologists and pho- element: ‘The camera is a mirror with a tographs like Andrew Davidhazy’s stop motion memory, but it can not think’. Nonetheless, shots.3 traces of the old positivist belief in ‘observable facts’, recorded mechanically, analyzed or In addition, Barrett’s typology is also useful quantified by ‘objective’ scientists and experts for considering photographs with more inter- lurks close to the surface of what Martin Jay pretive/hermeneutic characteristics. Our termed ‘the dominant scopic regime’ (Jay, intention, here, is to move away from having 1994). For postpositivists the most important two hard and fast approaches, as in the old quality of a photograph is its indexical con- paradigm wars. Our emphasis is rather on the nection to things in the world. As Edward interpenetration of postpositive and herme- Steichen described the relationship between neutic approaches to researching historical optical/chemical machinery and operator: photographs, without de-emphasizing their differences. As such, Barrett’s second two The camera is a witness of objects, places, and events. A photograph of an object is, in a sense, a categories work to emphasize interpretivist portrait. But the camera with its glass eye, the lens, views of research on historical photographs, and its memory, the film, can in itself produce little which regard photographs as texts to be more than mirrored verisimilitudes. A good photo- interpreted: graph requires more than that. .... The technical process simply serves as a vehicle of transcription and not as the art. • Ethically evaluative images include much of The photographer, unlike the painter and what has been termed ‘liberal documentary’ regardless of his subjective feelings, is forced by (Rosler, 1990). Barrett includes the works of Riis the very nature of his medium to concentrate on and Hine in the category that uses photography the object, on what Goethe referred to as ‘Das to force moral judgments in the viewer. Ding an sich’ in a portrait: on the person, or the • Aesthetically evaluative images seek a meadow, the mountain, the flowers, or the horse response from the viewer as well. They ‘function being photographed’ (Steichen, 1984: Chapter 10, as notifications that the photographic pres- facing plate 164). entation of people, places, objects, or events is worthy of aesthetic apprehension (Barrett, Art educator Terry Barrett (1986), created 1986: 57). a taxonomic system for art historians that • Interpretivist approaches to researching categorized photographs as descriptive, historical photographs.

55632-Margolis-18.indd632-Margolis-18.indd 334040 33/14/2011/14/2011 4:04:144:04:14 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 341

Hermeneutic perspectives emphasize photo- explanatory or derived information that can be graphs as texts demanding semantic and clearly verified to and by others (e.g. dating from semiotic interpretation to determine mean- format or image content, contextual comparisons ing. Allan Sekula termed assumptions that a with companion or other known images, verified photograph can ‘transmit truths’; ‘reflect period or more recent written identification, etc.); includes secondary source material. reality’; or be an ‘historical document’ as 3 Speculation − subjective inferences not fallacy. ‘The very term document entails a necessarily based on evidence provided by notion of legal or official truth, as well as a the photograph. It is the information that can notion of proximity to and verification of an augment the factual information that may range original event’ (Sekula, 1983: 195). Sekula from attribution by later generations based on also asked in regard to exploring and under- family legends, to hearsay or creative interpre- standing power relations: ‘How is historical tation to ‘leaps of faith’ based on desire to fit and social memory preserved, transformed, the image into a specific hypothesis or intellec- restricted, and obliterated by photographs?’ tual framework without sufficient factual basis (Sekula, 1983: 193). The iconic and symbolic (Rowe, 2002). life of photographs is the basis that the hermeneutic approach insists on for meaning − pho- The first two of Rowe’s strategies stand tographs are defined by their social and firmly in the postpositivist tradition. They cultural contexts and, as a result, can never be draw on what Roland Barthes termed: either ethically or aesthetically neutral. ‘studium’ − informed by education. Studium is the socially prescribed, statistically ‘aver- age’ meaning of the historian, the social scien- tist − interested in describing what the image POSTPOSITIVIST APPROACHES depicts: date, location, event, caption and so on. (Barthes, 1981: 22−26). Valuable research Postpositivist photographic researchers methods include standard historical research hypothesize that the image and its contextual into the photographer and the events depicted, information contain data such as where, how, content analysis of the image, which can be by whom, when and why the images were used for hypothesis testing and for a more produced; and use this information to attempt inductive process of hypothesis generation, to infer embedded representations within the and close analysis of the photograph itself. images. Ultimately the goal is to use this Many visual ethnographers recognize the information, and knowledge of photography, importance of understanding the relationship culture and context, to reach a warranted between the researcher and the subject, yet comprehension of the image. Jeremy Rowe these same researchers rarely address the devised a three-part strategy for photo importance of understanding the relationship research: between the photographer that created the image and the image, or exploring the con- 1 Evidence − objective, factual, documentary text in which the images were originally information provided by the photograph or created. The evolution from reliance on text its context. This is its provenance, or context alone, to incorporating images as sources for (e.g. format, content within the photograph, researchers, requires new methods and ana- attribution to photographic studio based on lytic techniques. Though many researchers imprint or printed identification from the period, etc.); the focus is on primary source information. reference the link between image and sub- 2 Interpretation − where we have little docu- ject, photographer and ethnographer, few mentation and must rely on deductions built have taken the next step to examine the infor- on circumstantial evidence and context that mation embedded in the image (format, can be clearly verified to and by others. media, imprint, notes, etc.) or of incorporat- Interpretation builds on the factual evidence with ing knowledge of the larger body of work

55632-Margolis-18.indd632-Margolis-18.indd 334141 33/14/2011/14/2011 4:04:144:04:14 PMPM 342 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

created by a photographer, or of comparisons the overall image. Even simple information with work by other photographers who about the original format of the photograph may have made additional images of an area can be potentially valuable, but is rarely or event. If such images can be located, noted in publications. they can provide a foundation for inferences Identification of the photographer can pro- such as the prominence of the subject at the vide valuable context about related work time, and permit triangulation of style and from which researchers can infer information context across the work by each individual about style, approach (as in documentary vs photographer. staged or posed), ability to interact with Recently, aided by copyright policy and the subject, use of props, etc. For example, emerging current practice, many library and studying photographer-produced stereo- archival collections, commercial sources, and graphs in the late nineteenth century can some collectors have been successful in provide insight regarding the entrepreneurial demanding source credits when images from nature of the photographer, interest in loca- their collections are published. However, tion work in addition to studio portraits, etc. providing credit to the photographer, when (Rowe, 1991, 2008). Knowledge that the known, in popular press publication captions photographer or photographic publisher pro- (other than to a few well-known fine art and duced primarily prints from original nega- journalistic photographers), has been rare. tives, as opposed to marketing copy When information about the photographer, photographs of the work of that of others, can format and other contextual information is offer additional insights. presented at all, it is usually separated Changes in mount colors or styles of from the image and appears in end credits or imprints, when combined with knowledge on a photo source page at the end of the about the photographer and changing styles, publication. can provide collateral information that can The interaction between the photographer provide a range of dates when the image may and viewer extends to the size, format, pres- have been produced (Rowe, 1997: 116−118; entation style, and labeling or captioning of 2007: 75−88). Imprints identifying the pho- the images that are produced. Changes in the tographer open the door to researching pho- presentation of the photograph in print or tographic business records and directories of online publication, e.g. cropping or extrac- photographers that can provide additional tion of detail within a photograph, can have data for analysis. significant implications on the viewer’s inter- The technology to scan, enlarge, crop and pretation and understanding of the image. enhance photographs during research has Such changes are rarely noted as photographs dramatically improved the ability of research- are reproduced. For example, images are ers to explore images for embedded clues. routinely cropped to fit layouts, and image Examination of information embedded within information such as format, mount type or the image, such as artwork, calendars, or style and photo credits are regularly removed. signage, can provide temporal cues about Historic images such as cartes de visite and when the image was originally made. cabinet cards are typically cropped and Software provides researchers with capabili- printed without mount information that can ties for comparison of images that were pre- often identify location, and provide assist- viously extremely difficult if not impossible ance in verifying date. Photographer’s using traditional magnification and photo- imprints are rarely included unless a notable graphic techniques. Included are techniques photographer produced the image. Stereo- such as: overlaying images; scaling; transfor- graphs are routinely reproduced without mations to match perspectives, contrast and noting the original format, and using only sharpness enhancement; and automated com- one of the stereo pairs instead of presenting parison tools that can identify and compare

55632-Margolis-18.indd632-Margolis-18.indd 334242 33/14/2011/14/2011 4:04:144:04:14 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 343

elements within the images.4 Rowe coined when copies were made of much earlier the term Photographic Forensics to describe images, such as the common practice of the full spectrum of photographic analysis making cartes de visite copies of daguerreo- from: examining image content using these types or ambrotypes, can similarly improve techniques to identify embedded clues; the accuracy of dating and attributions. mount and format data; contextual informa- Another potential source of contextual tion about the photographer (when known) information is knowledge about the photog- and related images by other photographers; rapher. For example, was she or he the sole and the contextual information associated operator in a small town, or one of a number with single photographs or collection of of studios in a city? Did she cater to upper or images (Rowe, 2002). working class clientele? Were the images Often, the date the photograph was taken known outside of the region and was the can add substantial information about the photographer recognized for artistic or docu- subject through an understanding of the style mentary styles? Is the photograph in question and symbolism of the era. For example, a typical of the photographer’s work, or is it woman wearing a black veil and clothing unusual in terms of pose, interaction with the may indicate mourning. Dark clothing with photographer, or inclusion of props? black gloves covering the palm but with bare Learning about the nature of the equip- fingers can symbolize a progression of the ment and photographic process frequently mourning process over time. Arrangement contributes useful information. Was the proc- and relationship of multiple individuals ess used common or more unique when the within the frame, or placement of hands or image was made? Did the process produce a contact between subjects; each had its own single copy of the image, like the daguerreo- Victorian symbolism. type, ambrotype, or tintype, or a negative Information on the approximate cost and capable of producing many prints? Was the resulting economic impact of the photo- camera mounted on a tripod or studio stand, graphs on the subject when it was made can or hand-held? Was lighting manipulated, or also be valuable. Knowing that a large was flash or auxiliary lighting added? Was daguerreotype was a significant expense in the equipment high-level professional or the 1840s, about $150, can be an indicator of lower- cost amateur gear? Does the image fit the wealth of the subject. Knowing that into the aesthetic style of the time, or is it smaller 1/6-plate-sized images cost about retrospective or groundbreaking? $30 in the 1840s, with the price dropping to In addition to examining clues within the $7.50 for two portraits by the late 1850s, image, researchers can explore available con- provides additional insight to the subjects of textual or documentary information that may these images (all prices adjusted to current be associated with the photograph. In com- dollars). prehending meaning, provenance and the The social context of cartes de visite, history of the photograph and the collection, which were exchanged in the 1860s, and archive, or source can provide additional clues knowledge that images of notable authors, about why and how the photograph was made performers, and politicians were collected and why it was valued and preserved. and included in family albums provides valu- Examining contextual information, such as able understanding of collections of family companion images in an album or collection, photographs of the era. The dates that tax proof sheets or other images made by the pho- stamps appear on the reverse, or changes in tographer, can reveal information about the mount style and borders occurred, also pro- photographer’s intent, as can journals, day- vide useful information in dating photo- books, assignments, or other collateral refer- graphs and discovering locations. Recognizing ences (cf. Margolis and Rowe, 2002). There photographic copies vs originals, particularly may also be captions and other descriptive

55632-Margolis-18.indd632-Margolis-18.indd 334343 33/14/2011/14/2011 4:04:144:04:14 PMPM 344 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

information about the image and subject. about the real subject of the image labeled If available, studying both the negative and ‘Kaloma’. The image of Kaloma took on a original print can be meaningful. life of its own through Boyer’s book cover As noted, images change their meaning as and the trail of auction catalog descriptions they are cropped or reprinted (cf. Margolis, that built upon his attribution. Since no prov- 1988). Some photographers use cropping to enance has yet surfaced definitively to iden- provide editorial control and to shape the tify this image, it provides an exemplar of the meaning that the image conveys. Christopher process of analyzing clues that are associated Lyman’s study of the Edward Curtis collec- with the image to demonstrate a postpositiv- tion, for example, revealed how Curtis him- ist approach to photographic research.5 self altered the images of Native Americans Photographic research is based on obtain- to eliminate traces of modern life as he ing as much information as possible about an composed shots and printed his negatives image, then building a logical context for (Lyman, 1982). a possible identification of the image. As Original owners can often provide sub- new information is located, it is compared, stantial information about the images in their and the interpretations checked for ‘fit’, personal collections. But most historical pho- given the new data. This embedded contex- tographs have been completely torn from tual information can be structured into three their original contexts. In most cases, it is categories of increasing confidence as only a matter of luck when the collateral noted above: evidence, interpretation, and information remains paired with the photo- speculation. graphs. Unfortunately, only a single gap in Each level of analysis can provide a valu- provenance can break the tenuous connection able source of information that can be evalu- to virtually all of this collateral information ated, verified, and weighted as part of the associated with an image. In the growing research process. For example, evidence such image marketplace, individual photographs as photographer’s imprints can usually pro- are often separated from collections or vide valuable insight about the image. But removed from albums with their associated for copy photographs the mount imprint may captions and context to sell the images indi- not accurately state the original photogra- vidually in an attempt to increase revenue. In pher. For example, the well-known images of other cases quirks of history and custody, Geronimo by Irwin, Randall and Wittick, and much of which is desirable documentation, is C. S. Fly were frequently copied, and today simply missing. But, as we shall see, this examples of these images exist with imprints lack of context does not make these images of dozens of other photographers. meaningless. Interpretation based on the format of the Confounding the impact of missing infor- photograph or information within the image, mation is the impact of dissemination of such as building signs, can help verify or incorrect information about a photograph. refute written identification that may have Unfortunately, once incorrect information been added to the mount. becomes widely available it takes on a life of All written information associated with an its own through references and citations, image should be confirmed, particularly if it making it extremely difficult to rein in misin- was added after the image was originally formation or correct errors. Rowe developed produced. Well-intentioned family members, a research strategy to determine the accuracy collectors and museum staff often add attri- of the attribution of an image published with butions to the photographs that pass through the label ‘Kaloma’, that was claimed to be their hands. Their impressions or knowledge, Josephine Earp by author Glenn Boyer (1998 and the accuracy of the written information, [1976]). Soon after the publication of Boyer’s should be verified before it is assumed to be book, I Married Wyatt Earp, debate arose correct.

55632-Margolis-18.indd632-Margolis-18.indd 334444 33/14/2011/14/2011 4:04:144:04:14 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 345

Speculation may be based on interpretation of available evidence, on emotional reaction to a photograph, or desire to ‘trim’ a piece of the puzzle of history to make it fit a particular research hypothesis or philosophy. Speculation can also be benign or uninten- tional, based on little knowledge or on incorrect information. Personal desire or a potentially escalating value for a given photograph can also drive speculative interpretations. For example, images of young Abraham Lincoln are rare, likely due to both his economic circumstances and frontier location. As a result, extremely high values are placed on any new image of Lincoln that might surface. Several well-intentioned individuals have found vintage photographs that they feel appear similar to known Lincoln portraits. Without specific evidence or provenance linking the images to Lincoln, attempts to generate such connections range from trying to convince experts to authenticate based on the apparent physical similarity, to computerized morphing of the image to the known Lincoln portraits. Each photographic identification is only as accurate as the weakest link in the chain of Figure 18.1 Kaloma, the vignetted image information about the image that is available of a beautiful young woman boldly posed at a given time. Anecdotes and speculation for the camera in a sheer gauze peignoir. can make great stories, but without evidence Collection of Jeremy Rowe vintage to support the assumptions, provide merely photography, vintagephoto.com weak links in the process of accurately identifying a photograph. Rowe examined the image labeled publication runs than actual photographs. ‘Kaloma’ and the attributions based on evi- Photogravures were often printed with title dence, interpretation of that evidence, and and publication data below the image and speculation about the image and its impact were commonly used to create many copies on the perception of the image and market of high-quality illustrations for books, post- value (Rowe, 2002). Based on evidence cards and art magazines. Though photogra- available to date, the trail of information vures had been used since the 1850s; their about the image of ‘Kaloma’ begins in 1914, surge in popularity was between 1890 and when the vignetted image of a beautiful 1920. young woman boldly posed for the camera in Many of the prints of Kaloma bear a credit a sheer gauze peignoir became popular to either the ABC Novelty Company in New (Figure 18.1). Titled ‘Kaloma’, it was origi- York or to the Pastime Novelty Company at nally produced as a photogravure. These 1313 Broadway, New York. Labels on the high-quality reproductions from photographs back of commercially framed prints indicate were produced from engraving plates on a that it was widely popular. Prints of ‘Kaloma’ printing press, and were much less costly for have surfaced with framing shop labels from

55632-Margolis-18.indd632-Margolis-18.indd 334545 33/14/2011/14/2011 4:04:144:04:14 PMPM 346 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

Hawaii and states throughout the USA and The following month, M.A.D. published a into Canada. response to the Ackerman letter by Bob Raynor The risqué image was popular and sold of H.C.A. Auctions. Mr. Raynor acknowl- well as both photogravure and silver print. edged that H.C.A. represented the Kaloma Also in 1914, the image appeared on the image as that of Josie Earp after researching cover of ‘Kaloma, Valse Hesitante (Hesitation the image, and argued ‘Both Sotheby’s and Waltz)’ composed by Gire Goulineaux, and Swann Galleries identified and sold the photo published by the Cosmopolitan Music image in 1996, both auctions prior to the Publishing Company, 1367−69 Broadway, December H.C.A auction’. Raynor stated: New York. Kaloma’s popularity continued as she became a pin-up during World War I, and Please note that the image was used as a dust appeared after the war on postcards. After cover of the book I Married Wyatt Earp, published discrete airbrushing darkened her peignoir, by University of Arizona Press, 1976. Additionally, the image was used in another book, Wyatt Earp’s Kaloma also appeared in other popular adver- Tombstone Vendetta, published by Talei, and also tising during that era. in Pioneer Jews, Houghton Mifflin, 1984.6 In all The relatively benign history of Kaloma instances the image was identified as Josephine changed significantly in 1976 when Glenn Earp (Raynor, 1997). Boyer used an airbrushed version of Kaloma as the cover illustration for I Married Wyatt Though this level of research is credible, it is Earp. Gradually, interest in the image began interesting to note that all of the references to shift from risqué nostalgia. Kaloma became hinge on, and post date, the attribution of an icon of the mania for Western collecting Boyer’s book cover. that grew through the 1980s and escalated As prices rose, so did the number of auction dramatically in the late 1990s. Almost entirely sales of the ‘Kaloma’ image. Sotheby’s 8 April as a result of the book cover attribution, 1998 sale included a photograph labeled an copies of the Kaloma image began to sell anonymous picture, taken in 1914 and titled for hundreds, then thousands of dollars as ‘Kaloma’, of a siren-like figure dressed in a portraits of Josephine Marcus Earp. sheer gown with a plunging neckline. Described Questions about the historical accuracy of in the catalog as a hand-tinted photograph of Boyer’s book and the attribution of the cover Josephine Marcus Earp, the one-time wife of photograph of Kaloma arose in the mid- lawman Wyatt Earp, the photograph was esti- 1990s. The debate about the cover image mated at $3000− 4000 and sold for $2,875. The escalated, reaching the popular press in the Sotheby’s catalog saw broad distribution and late 1990s. Donald Ackerman wrote to the afterward was frequently cited as the source Maine Antique Digest (M.A.D.) in June 1997, used to ‘identify’ Kaloma images as Josie Earp requesting assistance in verifying the attribu- in many subsequent auction and dealer sales. tion of the Kaloma image as Josie Earp. As the perceived value and notoriety of Ackerman noted the similarity to the early the image of ‘Kaloma’ rose, so did the stories silent film publicity stills that he was familiar that surrounded her: with and questioned the attribution to the 1880s and the strength of the purported link • Josephine Earp was born in 1861 and would have to Josie Earp. He further noted that recent been 53 in 1914. After this fact became an issue, Kaloma conveniently began to be described auction sale prices would likely draw more as a later print of an image of Josie taken in copies into the marketplace, and that addi- Tombstone, Arizona in 1881, when she would tional copies of the Kaloma image were have been 19 or 20 − roughly the same age as being offered by H.C.A. Auctions in their 27 the subject of Kaloma. April 1997 sale and an auction house in • At some point purported ties to C. S. Fly began to Kingston, New York on 28 May of that same surface as the original photographer of a drunken year (Ackerman, 1997). Josie coerced into posing for the portrait.

55632-Margolis-18.indd632-Margolis-18.indd 334646 33/14/2011/14/2011 4:04:154:04:15 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 347

• Legends of attribution prospered. Quotes from However, most images generally fit the sty- many sources have been touted as the definitive listic trends of their era. word on the history of the image. Unsupported The Kaloma image exhibits three strong tales of bar owners or those in attendance when stylistic elements that can be used to try to the image was supposedly made have been used assign a range of dates to the original photo- to rationalize the Kaloma image as a portrait of graphic image: Josie Earp. • Similarities with other, better-attributed images of Josie Earp have been cited but little 1 The sultry interaction between the subject in provenance has been given that could defini- Kaloma and the photographer is very direct. This tively connect any of the Kaloma images to Josie aesthetic style is more common and representa- Earp. tive of risqué images and nude studies from the • As an example of the market incentive, many postcard era (1905−1920) and appears rarely in copies of the ‘Kaloma’ image surfaced and cita- earlier nineteenth-century images. tions in auction catalogs and from dealer sales, 2 The full-figure vignetting of the image is also sty- all after the 1976 publication of I Married Wyatt listically more common during the postcard era. Earp were regularly used to ‘verify’ that Kaloma However, earlier images were reprinted in current is Josie Earp and establish a high potential formats years after they were originally taken. value. It is possible that the central image of Kaloma could have printed from an older negative and vignetted to update its appearance. Unfortunately, little concrete evidence has 3 The use of narrow depth of field (the range of been found to help settle the controversy. sharp focus in the photograph) was popularized Rowe (2002) published an analysis of the by art photographers in England and Europe evidence and speculation that revolved beginning in the late 1880s, and became popu- around the ‘Kaloma’ image that looked at a lar in America around the turn of the century. However, only a small fraction of commercial number of context and dating clues in the photographers regularly used this technique. photograph as a starting point for his analy- Aesthetically, the Kaloma image shares much sis. Several significant questions were posed. more with post-1900 images than it does with Why has no evidence of the tie between earlier images such as the cabinet cards that ‘Kaloma’ and Josephine Earp surfaced that were popular during the early 1880s, when the predates publication of Boyer’s book in image had to have been taken if Josephine Earp 1976? Also, where are the primary source was in fact the subject. citations from the period between Josie’s time in Tombstone, and the emergence of Risqué photographs like the Kaloma image Kaloma in 1914 that link the image to the have been made and sold since the 1840s. The personality? What do we actually know subjects of such ‘art’ photographs were not about the image titled ‘Kaloma’ from the usually identified. It is highly unlikely that available evidence? even if the subject of Kaloma had been identi- If one examines the image to see what fied at some point, such documentation by evidence is presented, certain inferences and photographer or publisher would still exist. interpretations can be derived. Photographic However, given the heated levels of discus- styles changed regularly every few years as sion about the current attributions, and possi- photographers sought to justify new portrait ble liability given Kaloma’s high sales prices business, and as lenses, formats and emul- based on its attribution to Josephine Earp, it is sions continually evolved. By looking at not likely that even if such information is large numbers of images it is possible to get available, that publishers or distributors would a feel for the photographic style of a given actively take sides in this matter. Obviously, era. Images that don’t fit the norm do exist, locating any original documentation of the and are often highly valued by collectors as sitter of the Kaloma image would be key to precursors of future styles and trends. unraveling the controversy about this image.

55632-Margolis-18.indd632-Margolis-18.indd 334747 33/14/2011/14/2011 4:04:154:04:15 PMPM 348 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

Copyright notifications have been printed To date, no copies of the Kaloma image have on photograph mounts and occasionally in been located on C. S. Fly studio mounts. the image area since the 1850s. Notices were Looking at the purported attribution of the occasionally printed or etched in the nega- subject as Josie Earp provides a bit of addi- tive, or later added to the mount or print with tional insight. During much of their lives, the a rubber stamp. Though copying and piracy Earps were popular, widely known, public were common, pirates rarely included previ- personalities. Though few commercial por- ous notices when illegally reproduced. The traits of the Earps exist today, if images of Kaloma images seen to date have all been the Earps had been available at the time it is associated with copyright notices dating from likely that they would have had a large and after 1914. The ‘Kaloma’ photograph on the ready market. Individual cabinet card por- sheet music is unattributed, though the music traits were relatively affordable, costing is copyrighted to Cosmopolitan Publishing about $1.25 per dozen, with group portraits Company. slightly more expensive at about $1.50 per The number and format of extant copies of dozen. It is also highly likely that period the image also provide some clues. If in fact documentation and references to such images the image were of Josephine Earp one would would have been produced, providing addi- assume that it would have been popular and tional evidence for such attributions. The many copies would have been made and mar- lack of such evidence is telling. keted soon after it had been taken. If it had Looking at the evidence provided by the been originally taken in the 1880s, most of image and trying to read the story that it tells the extant copies would likely be in the cabi- logically leads to it being an early twentieth- net card format. This would hold true even if century photograph of a beautiful young it were merely an unidentified nude image woman, likely made after about 1910, which made during that era. Virtually all of the was taken about the time that it first burst on copies of the Kaloma images identified to the scene in 1914. No clues clearly indicate date are printed paper photographs or post- this image was copied from an earlier image cards that date after the 1914 date indicated of Josie Earp or another as-yet unidentified in the copyright notice. Rowe is not aware of young woman. Given the broad exposure that any copies of this image that have surfaced in the image of Kaloma has had since the pub- original nineteenth-century cabinet card lication of the Boyer book, and strong inter- mounts. est in the legends of Tombstone and the West, The purported attribution to the C. S. Fly the search for compelling evidence to link studio in Tombstone can also be evaluated. In this image with Josie Earp will likely con- addition to his entrepreneurial personality, tinue. In short, at this point there is little Fly was known for the aggressive marketing evidence, some interpretation loosely based and promotion of his photographs, particu- on that evidence, and much speculation larly of his Geronimo series. Thousands of about the subject of the image known as copies of images made during the surrender Kaloma. of Geronimo in 1885 were printed and sold, each prominently identifying Fly as photographer and most carrying his copyright notice. Content analysis Similarly, portraits of personalities visiting Tombstone, and photographs of local events An important postpositivist approach is con- like the hanging of John Heath, were also tent analysis, a technique developed in the broadly distributed. If as speculated, Fly took field of media studies. It is a taxonomic and a salable image of Josie Earp, it is highly counting strategy for determining the relative unlikely that he would not have sought to frequency of certain representations within capitalize on the opportunity to sell copies. groups of images. Philip Bell (2001) gave an

55632-Margolis-18.indd632-Margolis-18.indd 334848 33/14/2011/14/2011 4:04:154:04:15 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 349

excellent review of the technical procedures procedures were used to determine if there but concluded: was a statistically significant difference in the points of comparison for the representa- It is also of limited value in many research contexts, tions of the observables between the early and might best be thought of as a necessary and later periods. Bell noted that not only but not sufficient methodology for answering could coders be trained to recognize easy questions about what the media depicts or repre- sents. Content analysis alone is seldom able to observables, e.g., blond or brunette, but also support statements about the significance, effects, they could accurately discriminate between or interpreted meaning of a domain of representa- more semiotic categories, for example ‘social tion (13). distance’, which he operationally defined as ‘how much of the (human) participant’s body The technical procedures, in Bell’s account, is represented in the frame’ (29). are a form of hypothesis testing. Content Similar procedures can be applied to analysis requires defining explicit categories sample populations of historic photographs. that are exhaustive and mutually exclusive, For example, one might examine representa- and employing checks like inter-coder relia- tive collections of studio portraits from two bility. These techniques are most applicable periods, the 1880s and the 1910s. Social dis- to comparisons between two discrete sample tance could define one axis of comparison populations that can be considered compre- and one might hypothesize that social dis- hensive representations of the sample popu- tance decreased as photographers and sub- lation. Content analysis is most applicable to jects became more comfortable with the research where a collection of images defines camera’s presence, or as exposure times the sample population than with less compre- decreased and head clamps became unneces- hensive collections that only are representa- sary, reducing the discomfort of posing for tive of the populations being analyzed. A portraits.7 Similarly, one might test the oft- hypothetical example to illustrate Bell’s noted increase in smiling faces over time. method involves an analysis of magazine The same sample of portraits could be coded covers. with a set of values for facial expression: One of Bell’s examples involved two sub- frown, deadpan, slightly upturned lips, teeth jects using explicitly defined categories to visible, mouth open in a toothy grin. Once code and compare the portrayal of subjects again, naive coders could be asked to dis- and social contexts expressed in images of criminate randomly presented images, then women appearing on the cover of Cleo their judgments compared until high values Magazine during two periods, 1972−4 and for agreement were achieved, and ultimately 1996−7. Points of comparison included the hypothesis that smiling increased could observables such as hair color (number of be tested. blonds and brunettes), age (older and younger There are two weaknesses of content anal- models), and social distance (involved or ysis as a tool for analyzing historic photo- distant). For example, values coded for social graphs. The accuracy of this technique distance were: ‘intimate, close personal, far depends on the ability of the sample to repre- personal, close social, far social, and public’ sent the total population, and is extremely (Bell, 2001: 29). sensitive to misinterpretation because of The two coders were naive in the sense sample bias due to the inability to identify or that they were blind to the hypotheses being have access to the entire population, and the tested. Each was trained to ‘classify images likelihood that some images are more likely according to the specified definitions’ (31). to be chosen than others. Although it is rela- The coding was then compared to establish tively easy to assemble collections of historic statistical reliability. The magazine covers photographs, in many cases it is difficult to were presented randomly, and technical create a representative sample that is realistic,

55632-Margolis-18.indd632-Margolis-18.indd 334949 33/14/2011/14/2011 4:04:154:04:15 PMPM 350 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

comprehensive, and unbiased, and that can them’ (Berger and Mohr, 1982: 96). Thus, the be used to provide accurate inferences about optical/chemical apparatus may record things the sample population. the photographer did not see, and much can Also, as Bell noted, even when representa- be learned by closely and systematically tive samples can be identified, the entire examining photographs. As noted above, one approach is under-theorized and cannot speak can frequently find clues: clocks recorded the to whether the hypotheses would have been time of day; calendars show month and year; ‘meaningful to those who habitually “read” and license plates, newspaper headlines and or ”use” the images’ (25). ‘In short, content flags similarly help date images. Postmarks, analysis cannot be used as though it reflects notations and dedications, provenance and un-problematically or a-theoretically the collateral contextual information can also social or ideological world...’ (24). This is expand the understanding of the image and especially true in the case of historic photo- its subjects. graphs. In most cases, it is virtually impos- In Margolis’ research on classrooms, what sible for modern coders and content analysts might be incidentally written on blackboards to know how the content was perceived at the provides clues to curriculum. Small details time the images were originally made. provide insights. The social and technologi- Still, content analysis has been used to test cal history of the medium provides useful hypotheses in historic photograph collec- information. tions. For example, in his 1993 book How Tintypes, daguerreotypes, panoramics, Teacher’s Taught, Larry Cuban employed, as Kodak snapshots, and digital files each had data, Frances Benjamin Johnston’s famous their day and produced certain social reac- set of more than 700 photographs made in tions to image-making technology. Each Washington DC schools in 1900 (Cuban, technique has its own representative vocabu- 1993). Cuban counted whether desks were lary of process evidence. Examples include: bolted down or movable and also compared quality of surface preparation and sharpness classroom activities to determine whether of focus and tonality of daguerreotypes; con- progressive or traditional classroom tech- sistency of collodion flow and contrast in niques were employed. In his analysis, he tintypes; size and evenness of exposure and found that traditional techniques of school- focus for panoramics; image size and processing were overwhelmingly depicted: ing or mounting artifacts for snapshots (such as the representative circular images pro- Out of almost 300 prints of elementary school duced by the first string-set Kodak cameras); classrooms, nearly 30 show groups of students working with large relief maps in geography, pre- and file format, bit depth and resolution for served rabbits and squirrels being used for a digital images. In addition, other contextual lesson, students watching a teacher carve into a information or comparisons with other simi- cow’s heart to show the parts of the organ, and lar images to infer studio quality or impor- classes taking a trip to the zoo. The remaining tance to the subject provide additional clues. 90% of the prints show students sitting in rows at their desks doing uniform tasks at the teacher’s Other information, such as estimating how direction (Cuban, 1993: 26).8 long an exposure might have taken, or how rapidly a series of subsequent images could be made using a given technology, may also Close inspection be useful (cf. Newhall, 1964; Leggat, 1995). Probably one of the most thorough and Where content analysis tries to tease mean- provocative postpositivist approaches to dis- ing from large collections of images, one can covering the meaning of photographs was also study single historical photographs in undertaken by filmmaker Errol Morris, who particular detail. ‘Photographs do not trans- writes the blog Zoom, for the New York late from appearances. They quote from Times. Morris wrote three long pieces

55632-Margolis-18.indd632-Margolis-18.indd 335050 33/14/2011/14/2011 4:04:154:04:15 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 351

comparing two Crimean War photographs Russians. One expert was ambivalent. Morris made by one of the first war photographers, (2007b) wondered: Roger Fenton, in 1855. Morris initially seemed ready to defend the ‘pencil of nature Would it be possible to order these photographs not based on anything that Fenton said (which view’, asserting ‘Photographs provide a might be unreliable) − but based on evidence in “window” into history ’. The implication was the photographs themselves? This idea appealed that the comment applied to the history of a to me because it did not require me to imagine specific moment, and specific place ‘as if we something about Fenton’s intentions, that is, have reached into the past and created a tiny about his internal mental state. peephole’ (Morris, 2007b). But Morris had The answer was provided by Morris’ friend no patience with interpretive or hermeneutic Dennis Purcell, who focused on small accounts. His obsession with Fenton’s two changes in the images. ‘We were scrutinizing images of the ‘Valley of the Shadow of individual rocks in the Fenton photographs: Death’ began in reaction to Susan Sontag, whom he cited from Regarding the Pain of When the rocks are uphill, and you look at the Others: road, you see that the balls are off the road... . Then, you look at the rocks after they have been Not surprisingly many of the canonical images of dislodged − rocks that were kicked and then tum- early war photography turn out to have been bled downhill − the balls are on the road... . In staged, or to have had their subjects tampered short, the first shot had to be taken when the balls with. … Fenton made two exposures from the were uphill (Morris, 2007c). same tripod position: in the first version of the celebrated photo he was to call ‘The Valley of the Morris concluded: ‘The one thing that we Shadow of Death ’… the cannonballs are thick on know about the rocks ... is that they were not the ground to the left of the road, but before posed. No one noticed them, let alone posed taking the second picture − the one that is always reproduced − he oversaw the scattering of the them. But together they helped unlock the cannonballs on the road itself (Sontag, 2003 secret of how to order the Fenton photo- quoted in Morris, 2007b). graphs.

The two photos were taken from the same I tried hard to prove that Keller and Sontag were position: one showed a valley and road cov- wrong.... I failed. I can’t deny it. But I did prove that they were right for the wrong reasons. It is ered with cannonballs, which Morris chris- not their assessment of Fenton’s character or lack tened ‘on’; the other showed the same view of character that establishes the order of the with no cannonballs present, which Morris pictures (Morris, 2007c).9 named ‘off’. Morris did not believe Sontag (and others discussed) could warrant their Of course, as Morris noted, both photographs conclusions about ‘faked’ photographs based were posed, if only by being framed by the merely on cursory views of the photographs, photographer. The real dispute between or on other primary source data about Morris and the ghost of Susan Sontag is actu- Fenton’s work in the Crimea. He consulted ally a paradigm difference between the post- five ‘esteemed curators’, experts on Fenton positivist focus on photographs as indexical familiar with his letters, photographs, and signs that are causally related to the objects documentary material on the Crimean War, that they reflect, and the hermeneutic view and received contrasting interpretations: two that the use and understanding of photo- supported Sontag’s argument, arguing that graphs is governed by socially established ‘off’ came first; two suggested the opposite, symbolic codes. While recognizing that pho- arguing that the balls were removed so the tographs cannot be ‘true’ or ‘false’, Morris road could be used or that the cannonballs insisted on photographic proof of which had been ‘harvested’ by British soldiers so image came first, while the semiotician/ they could be recycled and fired back at the hermeneutician insists that the meaning of

55632-Margolis-18.indd632-Margolis-18.indd 335151 33/14/2011/14/2011 4:04:154:04:15 PMPM 352 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

photographs, as perceived things in the world, differing understandings of the meanings of are created by interactions between photog- texts − including photographs.12 rapher, image, and viewer.10 Sontag’s point was not an attack on Fenton’s character, it Oral history, using photographs to solicit com- was the political argument that ‘early war ments and analysis about the image and its photography turn(s) out to have been staged, original context, can be very helpful but is lim- or to have had their subjects tampered with’ ited to the life span of potential interviewees. (Sontag, 2003; quoted in Morris, 2007b) and If human sources are not available, research- nothing in Morris’s dissertation on cannon- ers must seek to understand the meanings balls, especially the conclusion that she was of embedded information, such as Victorian right for the wrong reasons, disproves her hand gestures, dress, flowers, and group argument. Hermeneutic and postpositivist poses, by identifying and studying other paradigms rely on different approaches resources. and standards of ‘proof’ to warrant their assertions. A grounded theory approach to content analysis

INTERPRETIVE/HERMENEUTIC Content analysis is not restricted to postposi- ANALYSES tivist hypothesis testing; it can also be employed more iteratively and inductively. Systematic approaches associated with In keeping with the Sekula and Tagg view of grounded theory have been found to work photography as a technology of power and a effectively when applied to content analysis mechanism for creating ‘texts’ which must (Glaser, 1978; Glaser and Strauss, 1967; be ‘read’ to make sense, interpretative Strauss, 1987). Margolis collected and coded approaches seek photographic meaning in photographs of schools in the Farm Security quite different ways from the postpositivists. Administration − Office of War Information This is not to argue that hermeneutic Archives (FSA−OWI) into emergent catego- approaches may not take advantage of the ries using a three-phase process (Margolis, techniques discussed above, just that herme- 2005). neutic approaches may go further and in dif- As a first step, Margolis counted the ferent directions from the post-positivist number of school photos in these collections approaches. In essence, there are two semi- by year and discovered that they increased otic approaches to analyzing photographs, steadily (as shown in Table 18.1): both discussed in other chapters in this The second phase applied ‘open’ coding volume: categories that were neither mutually exclu- 1 Structural semiotics, which assumes there are sive nor exhaustive. Using a ‘constant com- certain signs that can be read and understood by parative’ approach, the photographs were everyone or nearly everyone in a culture.11 examined sequentially and provisional 2 Social semiotics/iconology, which argues that judgments were made about appropriate cat- different people and different social groups have egory coding for each image. For example,

Table 18.1 School photos in collections, by year Date 1935 1936 1937 1938 1939 1940 1941 1942 1943 Total School photos 42 79 74 177 292 436 351 1305 1711 4467

55632-Margolis-18.indd632-Margolis-18.indd 335252 33/14/2011/14/2011 4:04:154:04:15 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 353

FSA−OWI images were coded using catego- While grounded theory avoids some issues ries such as ‘connotation’ date, location, related to the conclusion that postpositivist photographer, race of teachers and students, content analysis is un- or under-theorized, as well as ‘denotative’ or semiotic categories this approach cannot solve problems related describing the image: social distance, per- to adequacy of the research sample in repre- ceived age of children, and whether the senting overall population, nor does it address schools appeared to be ‘healthy happy places’ Bell’s question: Does the analysis yield state- or impoverished and unhealthy. The catego- ments that are meaningful to those who ries Margolis used were not mutually exclu- habitually ‘read’ or ‘use’ the images (Bell, sive; each photograph could be coded into 2001: 25)? Both forms of content analysis more than one category if appropriate. remain at the epistemological level of struc- The third phase examined similarities and turalist semiotics; the investigators define the differences, and each category was reconsid- categories as if these ‘resources’ were avail- ered for applicability using axial coding. As able to everyone (Jewitt and Omori, 2001). new codes were identified, they were added. The categories were collapsed or expanded when appropriate as the process continued. The process was both inductive in reaching INTENTIONALLY CREATED for hypotheses, and deductive in testing theo- PHOTOGRAPHIC SYMBOLS ries as they emerged (Charmaz and Mitchell, 2001). In the FSA−OWI project Margolis In Camera Lucida, Barthes coined the term concluded that the number of images coded ‘punctum’ to characterize the personal/ ‘schools as social problems’ declined, and emotional effect some photographs have on the number of schools depicted as ‘healthy the viewer (Barthes, 1981: 22−26). Many American schools’ increased. Moreover, photographers seek to create images that schools in the social problem category disap- function at this level, and such images virtu- peared entirely when the sponsorship of the ally cry out for semiotic/symbolic examina- documentary process shifted from the Farm tion.15 As Rowe noted previously, above and Security Administration’s mission of the beyond the ‘truth’ of the image there is a 1930s to the Office of War Information ‘sultry interaction’ between the photograph project of the 1940s. A theoretical hunch led of Kaloma and viewers. The ‘risqué’ symbol- to adding the code ‘patriotic’. As a result he ism is an integral component of the image found only a single image of saluting the flag and cannot simply be ignored. Similarly, the before 1942, while in 1942−3 alone, 18 shots cabinet card (Figure 18.2) is a complex of pledging allegiance to the flag were taken image with carefully constructed symbolic (Margolis, 2005: 111).13 elements on many levels. The tableau is care- Theoretical coding offers another potential fully created and posed in a photographer’s benefit: the ability to identify and recode for studio to show ‘Professor Lutz’ being teased categories that were noted as absent, or null. by his students.16 Analysis can begin with an Margolis repeated analysis of the collection, inventory of iconic qualities. The professor is adding the codes ‘Black schools’ and White male; all the future teachers are women. The schools’. He discovered that these categories image blatantly parodies the ‘normal school’ were mutually exclusive within the collec- class for future teachers. tion. Additional research of photographs Each woman poses, or was posed, to rep- from other sources from the 1930s and 1940s resent specific types of resistant student found a number of Black and White children behavior. Two appear to engage, and poten- in the same class; however, such images were tially distract, the instructor while the rest of not found the FSA−OWI collection.14 the class acts out a range of misbehaviors.

55632-Margolis-18.indd632-Margolis-18.indd 335353 33/14/2011/14/2011 4:04:154:04:15 PMPM 354 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

Figure 18.2 An original cabinet card group photograph, c. late 1890s or early 1900s, by a portrait photographer at Blodgett’s Studio in Hicksville, Ohio. Collection of Eric Margolis

None of the other students attend to the sexuality, and repressed desire that saturate this lesson. Two pass a note; one pokes her class- image. Sex, of course, is an often unspoken mate with a pencil, while another feigns text of schooling − and photography. sleep. The woman on the right appears to be Semiotic/iconographical approaches, though, tickling the student with her hands raised. cannot ultimately tell us the meaning the While the professor concentrates on his image had to its subjects and participants. lesson with the two students, another reaches All the women signed the back of the card, so to take the pencil from behind his ear. perhaps it was made as an end of term gift. Other meanings about the image and its Alternately, the image could be a play or intent can be deduced. The body is of course tableau, or even a photo to be used as a lesson the site of desire, as are classrooms full of about teaching. On the other hand, it is hard bodies. This image can be seen to speak to to miss Barthes’ ‘punctum’ as it was con- intentionally or desublimated sexuality. structed by both photographer and subjects. Despite a death grip on his pointer, the male body of Professor Lutz (or Lust?) struggles to maintain rigidity and decorum while surrounded by teasing young women. Underlying CONCLUDING REMARKS sexual tensions are revealed in images of school classrooms composed of all female The role of historic photographs in research, students with a male professor. The arch gaze and their use by historians, sociologists, of the spectators contributes to the power, ethnographers, geographers, and others is

55632-Margolis-18.indd632-Margolis-18.indd 335454 33/14/2011/14/2011 4:04:154:04:15 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 355

still being defined. When Margolis began intricate texts that must be responded to studying coal miners in the mid-1970s his- with the emotions, as well as with analytical torical societies and libraries gave essentially reason. unrestricted access to the copying and use of This chapter does not propose a single their photographs. Rowe used to peruse methodology for the studying and analyzing boxes of unaccessioned material while curat- of historic photographs. Images are complex, ing exhibits and researching early photo- and as the value of the information that graphers of the American Southwest.17 they contain is recognized, the palette of Discoveries of new materials and relation- approaches to analyze and explore them will ships between images didn’t depend on cata- also be expanded and refined. Our intent is to loging and ‘finding aids’ alone; luck and share approaches that have proven valuable, proximity were valuable collaborators. The to catalyze discussion, and to exchange ideas rules of discovery have changed, though, as for continuing the evolution of approaches to archives have begun to realize the value of photographic research. their collections and the fragility of photographic materials and have emphasized pres- ervation over convenience. As we noted NOTES above, the digitization of vast collections and their availability on the Internet has made 1 McDermott and Raley (Chapter 20) begin from images available as never before. ‘Accidents a similar epistemological presupposition in their chapter, ‘Looking closely: toward a natural history of of discovery’ that used to only come from human ingenuity’, tracing all visual social science to scanning Hollinger boxes now also come the rapid processing of symbolic interactions done by online. Increasing access to enormous online all people. public collections like the California Digital 2 Holmes, Oliver Wendell (1859) ‘The stereoscope Library and ‘American Memory’, which and the stereograph’, Atlantic Monthly, 3(20): 738−748. incorporated hundreds of smaller collections, 3 http://people.rit.edu/andpph/exhibit-3.html or commercial archives like Alinari, Corbis, 4 See Mark Klett’s contribution (Chapter 6), and Getty continue to grow. Online sharing ‘Repeat photography in landscape research’, for an of images through Facebook and FLICKR excellent discussion of the digital technologies he and other sites permits individuals, and uses to make rephotographs of historic images. 5 Some of the content of the present essay is increasingly museums, to post images, creat- excerpted, condensed, or elaborated from my previ- ing huge visual resources with research ously published work in the study of historic photo- potential. These formal and informal graphs and especially the Kaloma image. For example, resources and online access appear to see: http://vintagephoto.com/reference/kaloma/ approach Oliver Wendell Holmes’ dream, 1-02JosieKaloma%20article.htm 6 See: Rochlin, H. (1984) Pioneer Jews – A New in 1859, of an ‘Imperial, National, or City Life in the Far West. Boston, MA: Houghton Mifflin. Stereographic Library’ where people could 7 Clamps and stands were used to hold subject’s visit to see ‘any object, natural or artificial’ heads still during long exposures, such as in a studio. (Trachtenberg, 1989:16). Improvements in the plates in use by the mid 1880s Historians and social scientists are just allowed for shorter exposures, and the clamps were eventually unnecessary. beginning to come to grips with the fact that 8 Many of Frances Benjamin Johnston’s photo- photographs are much more than illustra- graphs can be examined online. Several hundred of tions. Photographs are finally being recog- images of school classes can be found in her collec- nized as valuable data sources that should be tion at the Library of Congress (keyword: ‘school’). examined, using a variety of rigorous Available from: http://lcweb2.loc.gov/pp/fbjquery. html approaches and techniques, to explore the 9 Morris’s postpositivism is most visible in his depth of information that they contain. Also, Popperian assertion of falsifiability (in the Postscript we agree with Barthes that photographs are to his three-part series):

55632-Margolis-18.indd632-Margolis-18.indd 335555 33/14/2011/14/2011 4:04:164:04:16 PMPM 356 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

I spoke with Dennis Purcell recently and asked, ‘Do 15 In an earlier article, Margolis and Rowe con- you think these essays will put this issue – the issue ducted a detailed hermeneutic analysis of a photo- of which came first – finally to rest.’ Dennis replied, graph from the Phoenix Indian School with the title ‘No. I don’t think so. There could be some guy who ‘Good Night’ (see Margolis and Rowe, 2002). reads your essays, writes in, and says: ‘You know, 16 This is an original cabinet card group photo- there aren’t just two photographs. I found another. graph, c. late 1890s or early 1900s, by a portrait There are actually three’ (Morris, 2007b). photographer at Blodgett’s Studio in Hicksville, 10 In an earlier New York Times blog, Morris also Ohio. recognized that social knowledge was essential: 17 Archives frequently went unguarded. Researchers brought briefcases gear including cam- Without a caption, without a context, without eras, lights, and copy stands and were left alone. On some idea about what the picture is a picture a more positive note, while reviewing a collection for of, I can’t answer. I simply cannot talk about an exhibition, Rowe was able to identify and reas- the photograph as being true or false inde- semble a missing album (Number 2 in a three-album pendently of beliefs about the picture. A series documenting construction of the Phoenix, caption less photograph, stripped of all con- Prescott and Santa Fe railroad, c. 1891 by J. C. text, is virtually meaningless. I need to know Burge), by identifying mount and captions informa- more (Morris, 2007a). tion that appeared on photographs that had been (In this he follows Gombrich (1961), who argued disbound and distributed throughout the collection back in the 1960s that only captions or labels [as by individual image topic. statements or propositions] can be true or false, not pictures.) But unless one is discussing formal logical propositions, this conclusion seems to defy common sense since if I look at a painting, say Jackson REFERENCES Pollock’s ‘Number 5’, I don’t need a caption to tell me the meaning of what I am seeing, I just make it up. As John Berger wrote, ‘In every act of looking there is Ackerman, D. (1997) Mrs. Earp? [Letters to the editor: an expectation of meaning’ (Berger and Mohr, 1982: June]. [Online]. Maine Antiques Digest. Available 117). Morris is right in that ‘truth’ is not a quality of from: http://maineantiquedigest.com/articles_ photographs, but as Berger makes clear, meaning is. archive/articles/lett0697.htm [Accessed: 2001]. In the case of vernacular photographs torn from their Barrett, T. (1986) ‘A theoretical construct for interpret- context, we also create meaning in viewing them, ing photographs’, Studies in Art Education, 27(2): but, as I will argue, we make the meanings using 52−60. socially established and learned clues. 11 See Winfried Nöth’s contribution (Chapter 16), Barthes, R. (1981) Camera Lucida: Reflections on ‘Visual semiotics’, for a detailed discussion. Photography. Tr. R. Howard. New York: Hill and 12 Marion Müller’s contribution (Chapter 15) in Wang. this volume provides an excellent introduction to a Bell, P. (2001) ‘Content analysis of visual images’, in long German literature on iconography and iconol- T. van Leeuwen and C. Jewitt (eds), Handbook of ogy, which has been little noticed outside Germany. Visual Analysis. London: Sage Publications. Social semiotics/iconology lends itself to use of tech- pp. 10−34. niques, like audience studies, in communication and Berger, J. and Mohr, J. (1982) Another Way of Telling. photo-elicitation research. These techniques are also New York: Pantheon. addressed in Francesco Lapenta’s contribution Boyer, G. G. (1998 [1976]) I Married Wyatt Earp. The (Chapter 11) in this volume. 13 It might be possible to discover in the FSA−OWI Recollections of Josephine Sarah Marcus Earp. written archives whether the decision to make ‘patri- Tucson, AZ: University of Arizona Press. otic’ photographs was made by Roy Stryker who Charmaz, K. and Mitchell, R. (2001) ‘Grounded theory directed the project, or by individual photographers. in ethnography’, in P. Atkinson, A. Coffey, S. The images and their sparse captions do not provide Delamont, J. Lofland and L. Lofland (eds), Handbook that information. of Ethnography. London: Sage. pp. 160−174. 14 Margolis speculated that integrated class- Cuban, L. (1993) How Teachers Taught: Consistency rooms would have angered the Southern ‘Dixiecrats’ and Change in American Classrooms 1890−1990, who were an important part of the Roosevelt admin- 2nd edn. New York and London: Teachers College istration. Testing this hypothesis, as with the ‘patri- Press. otic’ images, however, would require analysis of many additional sources beyond the sample collec- Fox Talbot, W. H. (1844 [1989]) The Pencil of Nature tion assembled by Margolis for his research. (Facsimile ed). New York: Hans P. Kraus, Jr.

55632-Margolis-18.indd632-Margolis-18.indd 335656 33/14/2011/14/2011 4:04:174:04:17 PMPM DISCLOSING HISTORIC PHOTOGRAPHS 357

Geertz, C. (1973) ‘Thick description: Toward an inter- com/2007/07/10/pictures-are-supposed-to-be- pretative theory of culture’, in C. Geertz (ed.), The worth-a-thousand-words/ [Accessed: 10 July 2007]. Interpretation of Cultures; Selected Essays. New Morris, E. (2007b) Which Came First, the Chicken or York: Basic Books. pp. 3−30. the Egg? (Part One). [Online]. Available from: http:// Geertz, C. (1983) ‘From the “Native’s Point of View”: opinionator.blogs.nytimes.com/2007/09/25/which- On the nature of anthropological understanding’, came-first-the-chicken-or-the-egg-part-one/ In Local Knowledge: Further Essays in Interpretative [Accessed: 25 September 2007]. Anthropology. New York: Basic Books. pp. 55−70. Morris, E. (2007c) Which Came First? (Part Three): Can Glaser, B. (1978) Advances in the Methodology of George, Lionel and Marmaduke Help Us Order the Grounded Theory: Theoretical Sensitivity. Mill Valley, Fenton Photographs? [Online]. Available from: CA: The Sociology Press. http://opinionator.blogs.nytimes.com/2007/10/23/ Glaser, B. and Strauss, A. (1967) The Discovery of which-came-first-part-three-can-george-lionel-and- Grounded Theory: Strategies for Qualitative Research. marmaduke-help-us-order-the-fenton-photographs/ Chicago, IL: Aldine. [Accessed: 23 October 2007]. Goffman, E. (1976) Gender Advertisements. New York: Newhall, B. (1964) The History of Photography: Harper Colophon Books. 1839 to the Present Day. New York: Museum of Gombrich, E. H. (1961) Art and Illusion. New York: Modern Art. Bollingen Foundation. Philips, D. C. and Burbules, N. (2000) ‘What is post- Jay, M. (1994) Downcast Eyes: The Denigration positivism’, in D. C. Philips and N. Burbules (eds), of Vision in Twentieth-century French Thought. Postpositivism and Educational Research. Lanham, Berkeley and Los Angeles, CA: University of California MD: Rowman and Littlefield Publishers. Press. Raynor, B. (1997) Re: Mr. Ackerman’s Letter as Jewitt, C. and Omori, R. (2001) ‘Visual meaning: Recently Published by Maine Antique Digest [Letters A social semiotic approach’, in T. van Leeuwen to the editor: July]. [Online]. Maine Antiques Digest. and C. Jewitt (eds), Handbook of Visual Analysis. Available from: http://maineantiquedigest.com/ London: Sage Publications. pp. 134−156. articles_archive/articles/lett797.htm [Accessed: Kuhn, T. (1970) The Structure of Scientific Revolutions, 2001]. 2nd edn. Chicago, IL: University of Chicago Press. Rosler, M. (1990) ‘In, around, and afterthoughts (on Leggat, R. (1995) A History of Photography from documentary photography)’, in R. Bolton (ed), Its Beginnings Till the 1920s. [Online]. Available The Contest of Meaning: Critical Histories of from: http://www.rleggat.com/photohistory/ Photography. Cambridge, MA: MIT Press. Lyman, C. M. (1982) The Vanishing Race and Other pp. 304−341 Illusions: Photographs of Indians by Edward Rowe, J. (1991), ‘Dudley P. Flanders stereoscopic S. Curtis. Washington, DC: Smithsonian Institution views of a trip to Arizona’, Stereoworld, National Press. Stereoscopic Association, 18(5): 28−33. Margolis, E. (1988) ‘Mining photographs: Unearthing Rowe, J. (1997) Photographers in Arizona 1850−1920: the meaning of historical photos’ Radical History A History and Directory. Nevada City, CA: Carl Review, 40(January): 32−48. Mautz Publishing. Margolis, E. (2005) ‘Liberal documentary goes to Rowe, J. (2002) Evidence, Interpretation, and school: Farm Security Administration photographs of Speculation: Thoughts on Kaloma, the Purported students, teachers and schools’, in D. Holloway and Photograph of Josie Earp. [Online]. Available from: J. Beck (eds), American Visual Cultures. London and http://www.maineantiquedigest.com/articles/oct02/ New York: Continuum. josi1002.htm [Accessed: 2 March 2003] Margolis, E. and Rowe, J. (2002) Manufacturing Rowe, J. (2007) Real Photographic Postcards: A History Assimilation: Photographs of Indian Schools in and Portfolio. Nevada City, CA: Carl Mautz Arizona. [Online]. Available from: http://courses. Publishing. ed.asu.edu/margolis/paper/paper.htm [Accessed: 1 Rowe, J. (2008) ‘Arizona pioneer photographer George October 2002]. H. Rothrock’, Journal of Arizona History, 49(Winter): Mitchell, W. J. T. (1994) Picture Theory: Essays on 355−392. Verbal and Visual Representation. Chicago, IL: Sekula, A. (1983) ‘Photography between labor and University of Chicago Press. capital’, In B. H. D. Buchloh and R. Wilkie (eds), Morris, E. (2007a) Liar, Liar, Pants on Fire. [Online]. Mining Photographs and Other Pictures, 1848−1968. Available from: http://opinionator.blogs.nytimes. Halifax, NS: Nova Scotia Press.

55632-Margolis-18.indd632-Margolis-18.indd 335757 33/14/2011/14/2011 4:04:174:04:17 PMPM 358 THE SAGE HANDBOOK OF VISUAL RESEARCH METHODS

Steichen, E. (1984) A Life in Photography. New York: Tagg, J. (1988) The Burden of Representation. Amherst, Bonanza Books and the Museum of Modern Art. MA: The University of Massachusetts Press. Strauss, A. (1987) Qualitative Analysis for Social Trachtenberg, A. (1989) Reading American Scientists. New York: The Press Syndicate of the Photographs: Images as History. New York: Hill and University of Cambridge. Wang.

55632-Margolis-18.indd632-Margolis-18.indd 335858 33/14/2011/14/2011 4:04:174:04:17 PMPM