We Are What We Eat: Preservation, Discovery, and Research in Historical Scholarship

Alea Henle

The late historian Rhys Isaac devoted several paragraphs to describing the diaries at the heart of his analysis of eighteenth-century Virginia patriarch Landon Carter in the 2004 award-winning monograph Landon Carter’s Uneasy Kingdom. Isaac noted that Carter left his diaries where they might be read by family members and visitors during his lifetime, and after his death generations of the family retained possession. Family members thus may well have made decisions about what contents would survive before they deposited the incomplete set of volumes with the University of Virginia’s Alderman Library in 1960. Unfortunately, but not unsurprisingly, Isaac did not integrate the diaries’ histories’ further in his analysis.
Too often historians draw on primary sources for content and overlook the conditions which made them available for use. Postcolonial scholars and those working on the histories of women, the poor, the illiterate, and disadvantaged ethnicities and sexualities highlight the ways archives are culturally constructed. Digital humanists also address biases in digital initiatives, but these have yet to make a sufficient impact on other bodies of historical research. On the whole, people are more interested in what they may extract from sources than in reflecting on sources’ availability.
We run the risk of re-inscribing centuries of power inequities when we focus only upon provenance and lament how few sources have detailed provenance or fail to consider how their sources survived for discovery and use. Until recently, successive generations of historians largely considered archives as reservoirs of materials and ignored the power of behind-the-scenes decision-making. Yet the very people interested in preserving materials for the writing of history (i.e. archivists, librarians, museum curators, collectors, historians) embraced practices (including selection, de-selection, arrangement, editing, and publication) which affected what sources survive for public access. Likewise, digital developments over the past decades, unequal access to resources, and idiosyncratic research practices limit scholars’ exposure to various bodies of primary and secondary literature.
Sources have histories which encompass creation, use, preservation, and discovery. Historical research, writing, and instruction should analyze how and why sources were preserved, thus placing resources in broader contexts and highlighting the actions and inactions contributing to persistence and discovery. Of equal import, historians need to consider the role of their research practices in shaping the pool of sources with which they work.
Chance plays a part in source survival, but historical scholarship and instruction depends upon materials made available due to successful preservation efforts, benign neglect, and/or failed destruction. Employees of archives, libraries, museums, and special collections actively make decisions to add materials to their collections, as do individual collectors. Within these collections, items are accessioned, processed, arranged, described, repaired, and otherwise made more or less available. The older the collection, the more opportunities for individuals and groups to affect what survives and in what form.
Provenance is generally defined in terms of ownership. The plural, provenances, denotes lines of ownership of more than one object, such as different copies of a volume. In the case of art history, material culture, and rare books, the application of provenance centers on tracing the ownership history of items. Provenance is also a foundational archival principle used to facilitate organizing institutional and personal records while maintaining distinctions between records produced by different creators.
Inasmuch as historians do address issues of preservation and discovery, this involves analyzing the provenance of central documents. Scholars of material culture incorporate attention to the ownership and use history of the objects they study. Rare book enthusiasts explore the provenance of specific volumes. Similarly, scholars of the history of the book study the production, authorship, and use of written items. Historians documenting disadvantaged groups have produced intriguing works exploring source production and survival, but these have had limited impact on other historical fields.
Ironically, historians’ focus on provenance contributes to the general absence of its use in scholarship. Use of provenance as successive possession relies on the availability of information about ownership and control. Historical repositories include information about previous owners in finding aids to manuscript and archival collections, but this usually involves three types of information: date(s) of acquisition, whether materials were donated or purchased, and name(s) of donor(s) or seller(s). Finding aids and collection guides sometimes supplement this with further information, but report the knowledge of whoever processed collections based on their willingness, time, and ability to conduct research on the collection, which can lead to missing documentation, error, and misinterpretation.
Older materials, those documenting disadvantaged groups, and even papers of professional men of middling station are prone to gaps in knowledge about provenance. For example, Concord, New Hampshire antiquarian John Farmer was a prime mover behind the New Hampshire Historical Society, established in Concord in 1823. Farmer’s papers include as much or more information about the Society’s early years as its institution records. The finding aid indicates the provenance of the bulk of Farmer’s papers as two purchases in 1928, one of approximately 400 letters and a second of about 100. Over the intervening decades, the letters were arranged chronologically without specification as to which belonged to which lot.
Repositories often keep authority files, separate from collections, which may contain information about a given item’s original acquisition and subsequent treatment. These in turn draw on and/or duplicate information in institutional records, which may include donation or accession records and, of equal or more import, correspondence between the donor or seller and the institution. Librarians and archivists may be willing to open their records to researchers, but without guarantee of discovery of beneficial information.
Unfortunately, in too many cases provenance records are incomplete, particularly for manuscript collections. Broken links make establishing even speculative chains of ownership and possession difficult or even impossible. Whereas bound volumes may carry distinctive ownership marks to facilitate provenance investigations, most loose-leaf manuscripts lack equivalent markings denoting successive ownership. These factors combine to restrict the use of provenance in most historical scholarship to a few selected primary sources.
Provenance or successive ownership constitutes only one element with respect to placing sources in context. Historical repositories’ processes of preserving and making materials discoverable include extensive points at which people may consciously and unconsciously affect what items, manuscripts in particular, receive enhanced access, replication in other formats, or a host of other interventions.
The custodial history of records matters, inasmuch as shifting ideas of what constituted history affected what materials a given institution accepted. In the mid-twentieth century prominent historical repositories including the Newberry Library, New York Public Library, and Library of Congress turned down the Mary Earhart Dillon collection, which included papers of prominent suffragists such as Carrie Chapman Catt and Anna Howard Shaw. Radcliffe’s Schlesinger Library eventually acquired and preserved the collection.
Further, over the centuries historical repositories have used a wide array of approaches to sorting and arranging materials for use. Shortly after establishment in 1804, the New-York Historical Society acquired the papers of General Horatio Gates, who fought in the American Revolution. A society officer divided the materials into two categories; one set was bound for consultation and the other tied up in string and marked as unimportant. A few decades later, a new generation of society officers reviewed the unbound papers and identified three volumes’ worth to bind and add to the materials made available to researchers. The officers also repackaged and labelled other documents they considered “really are private and of no value.” The “really are private” materials likely included Gates’ correspondence with his first and second wives and financial records. The collection was re-processed under twentieth-century archival principles and now incorporates these materials.[1]
Even recent collections processed under modern archival principles may require subsequent review. The University of South Alabama Archives twice appraised the records of Blacksher, Menefee & Stein, a biracial Mobile law firm whose partners participated in important Alabama civil rights cases. The university archives’ original interest centered on the files for these cases. The first assay separated records into items for retention and non-relevant materials for disposal but ran into sufficient issues of determination that staff decided to conduct a re-appraisal. The second review uncovered non-legal materials of value such as records documenting the Social Justice Commission of the Archdiocese of Mobile and local clergy and laity active in the fight for human and civil rights.[2]
Decisions to re-process older collections under modern principles can cause other collections to be left unprocessed or partially processed. Archival processing is time-intensive, contributing to backlogs of unprocessed and thus largely inaccessible materials. After all, if a researcher cannot ascertain a particular collection resides in a given archive, how can they access it? In 2005, two archivists challenged fellow professionals to rethink and reduce interventions with the goal of providing “more product, less process.” Less intensive processing, for example at the folder level instead of the item level, reduces the level of detail associated with any given collection but speeds up the process of making collections available.[3]
Today, format discussions tend to focus on the equivalency of digital copies to physical originals, but for centuries the emphasis was on print equivalency. Many historians endorsed print as a method of preservation. Printing permitted dissemination of copies far and wide, thus preserving a document’s contents against the loss of the original to fire, flood, war, negligence, or other disasters. A quintessential example of the acceptance of printed equivalency is the United States’ Declaration of Independence. The federal government’s copy was subjected to abuse for decades, which did not concern historians and historical societies since the text was widely available.
Yet acceptance of printed materials as equivalent to manuscripts, or conversely manuscript as equivalent to print, was and is problematic, particularly with respect to publications of historical materials dating before the development of modern historical editing ethics. Earlier historical editors considered it their job to do only what the authors would have done had they lived to prepare the collections. Their actions, however, contrast sharply with modern practices. New England historian and editor Jared Sparks is a quintessential example. When preparing a documentary edition of George Washington’s writings, Sparks omitted materials considered inappropriate for publication, substituted terms (changing Washington’s references to “Old Put” to the more respectful “General Putnam”), and altered grammar and spelling to conform to nineteenth-century standards. Sparks’ editorial practices proved controversial during his own lifetime. In the 1850s, he waged a pamphlet war with Lord Mahon, an English historian, who eventually conceded some of Sparks’ supposed alterations arose from his use of Washington’s letter books rather than the versions of letters actually sent. The matter of which version to include in edited works (drafts, sent letters, copies of drafts selected by authors) continues to perplex editors into the twenty-first century.
Sparks’ contemporary, statesman and historian Charles Francis Adams, described Sparks in private as “a mere white-washer.” Ironically, the same might be said of Adams, who published several editions of the writings of his grandparents, John and Abigail. He liberally omitted material, not least Abigail’s account of a stillborn pregnancy, graphic descriptions of diseases, discussions of family finances, sensual endearments, and references to domestic matters. He also altered grammar and punctuation and modernized Abigail’s phonetic spellings. He did exhibit some restraint and made only one outright substitution in his edition of Abigail’s letters: where she wrote “Spanking” he substituted “personal chastisement.”[4]
This matter warrants attention because digital initiatives mingle materials processed at different times and under different rules. Although an excellent resource that makes valuable materials more widely accessible, Alexander Street Press’ North American Women’s Letters and Diaries emphasizes letters and diaries’ original authors and de-emphasizes their sources. Most of the letters and diaries come from printed editions, some from the latter half of the twentieth-century and prepared using modern concepts of historical editing. But many derive from editions published in the nineteenth-century when most editors followed practices which have fallen into disfavor. Thus letters from Abigail Adams to Thomas Jefferson edited by the staff of the Library of Congress more recently under modern historical editing ethics appear in the same format as the version of Abigail’s letter to her husband where Charles Francis Adams substituted “personal chastisement.”
Such false equivalencies risk misleading or distorting users’ perceptions. The lack of contextualization of sources distorts digital developments, impacting historical interpretation. The existence of well-funded databases or digital resources may inhibit the development of competitors regardless of problematic design or interface. Thus problematic database design decisions cast long shadows. In the United States, the earliest database and digitization projects were largely built on pre-existing twentieth-century microfilm initiatives, replicating earlier investments in sources that over-represented the papers of prominent deceased white men of station, newspapers, and printed works. Numerous recent digital initiatives have focused on historically disempowered groups, but much remediation remains to approach equity.
Other aspects of digital impacts on discoverability also warrant consideration, for instance perceptions of completeness. A number of high-profile digital resources of archival material are developed in phases. Both commercial and non-profit entities often charge fees for access to specific collections, either yearly subscriptions or outright purchases with lesser maintenance fees. Institutions purchase and offer some but not all of the collections, dependent upon shifts in funding and local needs. JSTOR, America’s Historical Newspapers, and other resources were developed and offered for purchase as distinct phases or collections. Newspaper digitization illustrates an area in which researchers risk assuming the extent of available content. Chronicling America, a free, searchable database of historic newspapers published in the United States, seeks to provide access to a selective number of newspapers by state. Ethnic press publications, particularly those in non-Western European languages, are still being added. Each state determines which newspapers to include using their own criteria. The Illinois Digital Newspaper Project consortium went through several steps before determining which titles to add. The project selected twenty-four titles over consecutive awards which “represents just a fraction of a wealth of significant newspaper titles.” Researchers should not assume that just because a title is included in Chronicling America that all issues of the newspaper were found and digitized.[5]
Placing sources in broader contexts requires researchers to interrogate how their own skills and interactions with physical and digital repositories affect what materials they have to work with. Unfortunately, research tendencies on the part of many historical scholars effectively reduce their exposure to important bodies of scholarship that discuss the impact of library and archival labors on availability and accessibility of historical resources. The 2012 Ithaka report listed a “typical search strategy” for historians as involving Google Books, Interlibrary Loan, JSTOR, WorldCat, Archive Grid, and Achive Finder. While all of these are solid tools, cumulatively they emphasize books, primary sources, and the specific array of resources in JSTOR. Historians have a reputation with American academic librarians for using JSTOR and Project Muse as primary means for identifying secondary literature. Both JSTOR and Project Muse have many strengths, including deep archival holdings and full-text searching, but they are artificially-created bodies of work missing most current scholarship due to publisher embargoes restricting access to recent issues. Reliance on JSTOR or Project Muse as major sources of secondary literature leads to missing important contemporary materials. Historians would be better advised to use these as supplements to broader discovery options such as the indices America: History & Life, Historical Abstracts, MLA International Bibliography, or Google Scholar and equivalent web search engines.
Sources’ histories or contexts—how and why they are preserved for use, made available for discovery, and enter into any given scholar’s body of research—warrant attention in historical scholarship. From the date of creation, all sources accumulate histories that incorporate decisions and accidents of preservation, processing, digitization, discovery, and more. In turn, scholars’ research practices effectively include and exclude items. These cumulatively shape what materials scholars have to work with.
The archaeological concepts of assemblages and deposition offer a means towards incorporating the context of sources into historical scholarship. Assemblages are collections of material sharing “contextual proximity,” while deposition is the process during which the materials were laid down or became part of an associational layer of sediment. In archaeological practices, an assemblage could be all artifacts associated with a particular site or section of a site (tomb, house). An equivalent application with respect to physical and digital historical research might be all items in a scholar’s body of research obtained from a particular archive or database.
Alternatively, consider all sources as having multiple collective histories based on any criteria by which they can be grouped. Abigail Adams’ letters may be combined, individually or as groups, with other sources based on content, creator characteristics, or other criteria. For instance, her letters to John written during the American Revolution form collectives based on marriage, physical separation, wartime conditions, and more.
Such an approach addresses some of the problems inherent in considering the matters of preservation, discovery, and research outlined above. Even items with little to no documented provenance may be placed in conversation with other items sharing particular commonalities. While these artifacts may not share provenances, accumulating available information about any such collection’s known ownership and preservation history could help flesh out awareness of how and why these materials as a whole survived. In particular, this may help parse out the complex factors contributing to the persistence of particular sources and groups of sources documenting the lives and experiences of historically marginalized groups.
Placing sources within the broader contexts of their creation to discovery offers the potential for a more holistic view of the bodies of materials available for historical analysis. I do not mean to say historians need to be all things to all people, search all existing databases, or consider sources and their histories from all angles. Rather, consider how analysis of sources and their histories relate to your particular interest, subject, or argument might contribute to your work. Sources, primary and secondary, are the raw material which feeds historical scholarship. What we find and how we find it shapes our analyses. We are what we eat.


Alea Henle has worked in academic and special research libraries for over a decade, and is currently the Head of Access & Borrow at Miami University Libraries (Ohio). She has a masters in Library Science from Simmons College and a Ph.D. in U.S. history from the University of Connecticut. Her research interests include how efforts to preserve historical materials in the early United States shaped resources available to modern historians.



[1]”Stated Meeting, May 4, 1847,” Proceedings of the New York Historical Society for the Year 1847, 60-61; New-York Historical Society, “Guide to the Horatio Gates Papers, 1726-1828.” 

[2]Carol Ellis and Russell D. James, “‘If at First You Don’t Succeed’: Blacksheer, Menefee & Stein, A Second Appraisal,” Provenance, Journal of the Society of Georgia Archivists, 20 (no.1, 2002), 21-31.

[3]Mark A. Greene and Dennis Meissner, “More Product, Less Process: Revamping Traditional Archival Processing,” American Archivist, 68 (no. 2, 2005), 208-63.

[4]Charles Francis Adams, as quoted in Lyman H. Butterfield, “Editing American Historical Documents,” Proceedings of the Massachusetts Historical Society, 3rd Series, 78 (1996), 88; Lyman H. Butterfield, “Introduction,” Diary and Autobiography of John Adams, (1962), I:R26-R30, and Lyman H. Butterfield, “Introduction,” Adams Family Correspondence (1963), I:R34-R38; Founding Families: Digital Editions of the Papers of the Winthrops and the Adamses, ed. C. James Taylor, (2007). 

[5]Marek Sroka and Tracy Nectoux, “‘The Dwindling Legacy that Is Food for Mice and Flames’: Discovery and Preservation of Illinois Historic Newspapers Through the Illinois Digital Newspaper Project, 2009-2015,” Journal of the Illinois State Historical Society, 110  (no.1, 2017): 87-107, esp. 96-97.

Further Reading

Terry Cook, “The Archive(s) Is a Foreign Country: Historians, Archivists, and the Changing Archival Landscape.” The American Archivist, 74, (no.2, 2011), 600-62.
Katharina Hering, Michael Kramer, Joshua Sternfeld, and Kate Theimer. “Digital Historiography and the Archives.” AHA Today. Jan. 21, 2014. Cross-posted at Issues in Digital History and the Journal of Digital Humanities.
Tim Hithcock, “Confronting the Digital or How Academic History Writing Lost the Plot.” Cultural and Social History, 10, (no.1, 2013), 9-23.
Laura Millar, “The Death of the Fonds and the Resurrection of Provenance: Archival Context in Time and Space.” Archivaria, 53 (2002), 1-15.
Andrew Prescott, “Digitizing the Historical Record.” Genres of Scholarly Knowledge Production, Jan. 1, 2015
Lara Putnam “The Transnational and the Text-Searchable: Digitized Sources and the Shadows They Cast.” American Historical Review, 121, (no. 2, 2016), 377-402.
Lena Roland and David Bawden. “The Future of History: Investigating the Preservation of Information in the Digital Age,” Library & Information History, 28 (no. 3, 2012), 220-36.
Janine Solberg, “Googling the Archive: Digital Tools and the Practice of History,” Advances in the History of Rhetoric 15 (no. 1, 2012), 53-76.
Michel-Rolph Trouillot, Silencing the Past: Power and the Production of History, (1995)