Archivists are commonly asked by researchers to produce everything available about a particular topic. While understandable from a researcher standpoint, fulfilling the request is a challenge. Unlike library holdings, archival material is rarely described to the item-level. This makes it difficult for archivists to do more than point researchers to where everything about a particular topic could be. The result is a persistent disconnect between researcher expectations and archival practice. It’s also an underlying cause of the increasingly prevalent, though by no means new, “lost in the archives” narrative in which archival material is deemed lost because it was not readily described in desired terms or, perhaps more accurately, widely recognized to exist.
In actuality, most records that are deemed to be “found”, or “discovered,” have been available for use by way of archival finding aids and lost thanks only to the failure of anyone to read them. A recent example is media coverage regarding the “discovery” of an unproduced Edith Wharton play that, as pointed out by Eric Colleary, Curator of Theatre & Performing Arts at the Ransom Center where the work was housed, had been listed in print finding aids since the 1980s and in electronic finding aids since 2006.
The disconnect between researcher expectations and archival practice was also evident in Dennis Molinaro’s piece regarding the discovery of Canada’s Secret Archives. Molinaro rightly draws attention to the substantial number of government records that have yet to be transferred to Library and Archives Canada for use and access by the Canadian public. I argued that this issue is symptomatic of decades of chronic underfunding and non-existent political will, rather than a concerted effort to suppress the public record. After unsuccessfully requesting files pertaining to wiretapping during the Cold War or obtaining finding aids for untransferred records, Molinaro concluded that the Canadian government is maintaining a secret archives where “no one in the general public is permitted to know the contents, and there’s a separate system that has been developed for storing and sorting this information.”
While Molinaro’s framing of the Canadian government’s legacy of undervaluing and failing to prioritize recordkeeping as a secret archives is one that merits further consideration, it will serve here as an entry point for examining what researchers expect and what archivists can provide. Specifically, why it isn’t possible to ever obtain everything about X held in an archives or, more importantly, to bypass the sometimes daunting and unglamourous work of archival research.
Library versus archival description
To fully unpack why seeing everything about X in an archives is so challenging, it helps to consider the distinction between library and archival description. When you find something of interest in a library catalogue and click on the corresponding title, a catalogue record is displayed. Contained within that record is pertinent information about the book such as the author, date of publication and related subjects that speak to the aboutness – topics or themes – of the book. Because catalogue records reflect a mass-produced object, they can be reused and repurposed by any number of libraries because the author and aboutness of a book doesn’t change. Once Kazuo Ishiguro’s Remains of the Day has been catalogued, edits or descriptive improvements that take place at one library can be put to use by others. Archival description is decidedly less utilitarian.
As a repository for corporate or personal records of enduring value, an archives predominantly holds material that differs from library holdings by being one of a kind. Copies, electronic or otherwise, may exist, but a fundamental aspect of archival practice is preserving original records in a manner that documents the purpose, intent and use of the record as a way of upholding its authenticity and evidential value. While archivists may collaborate on high-level biographical or organizational histories, particularly when related or complementary holdings are held by multiple institutions, descriptive work for the records themselves can’t be shared because only one archives holds those records.
Being unable to transfer or reuse archival descriptive records is further complicated by the purpose of archival description. It differs from library descriptive practice by prioritizing not the aboutness of a standalone record, but the original use, order and purpose of a group of records. In the Canadian context these records are referred to as a fonds. Instead of describing each record, archival description provides a high-level overview of the records within a fonds, focusing on information about the person or entity that created the records and the manner in which those records were used. While approaches to archival description may vary, the end-goal is always the same: provide entry points by which researchers can read, assess and make sense of the records at hand. Within this practice, and in part due to record volume, item-level descriptions are rare. In their place are file-level descriptions or lists, accompanied by dates of creation, formats and general notes, which serve as the primary means of accessing the records. The work of drawing intellectual conclusions, listening for conspicuous absences and reading between the lines is left to those who consult the records during the research process.
File lists and file descriptions versus researcher assumptions
While positing the existence of a secret archives, Molinaro referenced requests from various departments that he narrow his search request. He also expressed frustration with departments that requested more time to generate a list of files pertaining to wiretapping during the Cold War. This type of request is important because it suggests an underlying assumption that files pertaining to that topic are or should be readily identifiable. In actuality, archival file names are rarely subject-based. Rather, they reflect the categorization of their creator or the contents of the file as determined by the archivist. Each title simultaneously represents an entry point into the records and a requirement for further investigation. Part of that process includes understanding the environment in which the records, and the files, were created. For example, government and corporate records tend to align with specific organizational functions that can obfuscate the contents of a file unless you are readily familiar with the operations of the organization as a whole. This means that department files labelled with organizationally meaningful references, such as core functions, would be significantly more likely than those labelled “Cold War wiretapping.”
If available, file-level descriptions can provide more insight about the contents of a file, but, much like file lists, they present their own set of challenges. First, they vary from fonds to fonds. Some may provide a robust paragraph identifying people, document titles or themes captured by the records in a file, while others may simply identify for-your-information type notes such as “includes list of committee members” or “marginalia appears throughout.”
Second, archival description is prone, like most anything, to subjectivity. Whether information is deemed worthy of mention, and to what extent, is ultimately influenced by the cultural or social lens of the archivist. And, as Jarrett Drake has outlined, there are countless examples that have served to establish “memorials and monuments to wealthy, white, cisgendered and heterosexual men.” This has, in turn, helped to create a legacy of archival descriptions that have perpetuated the systematic marginalization and underrepresentation of Black, Indigenous and people of colour.
Finally, due to the labour intensive nature of descriptive work, fonds are often described only once, leaving the language and technical considerations of the era in which they were created to resonate indefinitely. This means antiquated terminology, the assumption of on-site research and, rather rudely, zero consideration for web-based keyword searching.
Given the limitations and stylistic quirks of archival description, pulling everything for manual review in order to find everything about X is one way forward, though one that can only be achieved with a small number of records. When a CN Tower is involved, it can be, frankly, impossible. One must also examine who would be doing the work of manually reviewing the records, keeping in mind that while archives staff can and do assist you with your research, they aren’t research assistants. The role of an archivist isn’t to perform close readings of records in order to identify hints of curious or compelling narratives. That’s the job of a researcher.
When you ask an archivist to pull together every file pertaining to a particular topic, you are effectively requesting a keyword search. Anything more amounts to asking that they do original research on your behalf. Whether broad searches of this nature are conducted manually or electronically – in Molinaro’s case the age of the records means they would all need to be reviewed manually – a failure to find desired search terms would result in a negative response about the availability of information on a particular topic. Therefore one can and should view requests from archivists to broaden or narrow a request not as attempts to suppress records, but as a limitation of working with available descriptive information and a desire to provide the best possible options for more detailed review.
Legitimate archival discoveries
To conclude, I’ll offer some counterpoints to the “lost in the archives” narrative, which, in addition to erasing the labour of archivists, perpetuates the Googlization of the research process and masks the extensive intellectual labour and time investment of original research. In May of this year it was announced that the mother of Leonardo da Vinci had been identified by Professor Martin Kemp as Caterina di Meo Lippi. He concluded her identity after thoroughly reviewing 15th-century property tax records. Another centuries-old question was answered in 2015 by Professor Frans Grijzenhout, who identified the location of a 17th-century Delft house depicted by Johannes Vermeer after reviewing a ledger documenting the dredging of canals in Delft.
The research work involved in both these findings is perhaps best captured by historian Marie Hicks who recently outlined, via Twitter thread, the innumerable number of hours she spent reviewing seemingly inconsequential archival records only to hit on proof that women didn’t leave computing, but were instead pushed out.
In each of these examples the question “give me everything on X” would have failed to yield files clearly labelled with the newly identified evidence, yet the answers were there for the finding, and an archivist most definitely ensured there were clues available to lead the way. While it may be exciting to imagine a secret archives where teams of employees are hoarding “the good stuff” and working to keep records from the public, the reality is significantly less diabolical.
Danielle Robichaud is Digital Archivist in Special Collections & Archives at the University of Waterloo. Her work focuses on improving the discoverability of archival holdings and increasing access to digitized primary resources in a manner that diversifies the narratives surfaced online through outreach and content development.