Dissociation refers to the separation of items (usually specimens, objects, and material culture) from their context, metadata, and related sources of information (including labels, cataloguing data, and records). The term first emerged as part of conservators’ increasing focus on inhibiting damage and loss before it occurs, known as “preventative conservation” (Caple, 2011; Rose, 1991; Staniforth, 2013). From the late 1980s, conservators like Stefan Michalski from the Canadian Conservation Institute (CCI) began elaborating on how to monitor and control for nine physical threats, including fire, water, pests, contaminants, temperature, and humidity (Michalski, 1990). R. Robert Waller then added a tenth, non-physical threat he termed “custodial neglect,” later renamed “dissociation” to avoid apportioning blame (Michalski, personal communication, 1 May 2017; R. Waller & Cato, 2005). Dissociation encompassed: “all the ways that data are dissociated from objects and collections and that objects are dissociated from collections and institutions that are not the result of criminal activity” (R. Waller & Cato, 2005), including “Loss of collection-related documents such as field notes” (R. Waller, 2015, n.p.).

Museums have long been aware of the importance of data associated with items. As Alfred Whiting, Curator of Anthropology at Dartmouth College Museum, wrote: “The essence of a museum collection is the data associated with the specimens. Specimens without data are at best curios … A collection so badly catalogued that the data cannot be readily found is more frustrating than no collection at all. Catalogues can be that bad” (1966, p. 85). Archivists often refer to the importance of context in similar ways. Renowned Australian archivist Peter J. Scott wrote with his colleague G. Finlay:

Records and archives derive much of their meaning and value from the administrative (or other) context in which they were originally created; furthermore we maintain that preservation of the association between archives and their original historic context is vital to a full and proper understanding of the evidence and information which they contain (1978, p. 115).

Dissociation in museums (and elsewhere) cuts both ways, impacting as much on the value and utility of documentary heritage as on other materials to which it relates.

An example of the deleterious effects of these splits is found in the Donald Thomson Collection, widely considered one of the most significant collections of Australian Aboriginal cultural heritage in the world. Thomson’s life work is now categorised into several parts. The Donald Thomson Ethnohistory Collection, also referred to as Thomson’s “literary estate” (Allen, 2008, p. 390), is on the Australian Register of the UNESCO Memory of the World Program. In keeping with that designation, it is comprised of documentary heritage, including photographs, sound recordings, maps, drawings, notebooks and field notes, language lists, correspondence, and illustrations (UNESCO Australian Memory of the World Committee, 2008). There are also many hundreds of ethnographic objects, donated to the University of Melbourne on Thomson’s death in 1970; and natural history specimens, including mammals and fishes (Thomson et al., 1985) and botanical specimens. The significance of the collection and its meaning for Indigenous and non-Indigenous peoples is found in the relationships within and between these parts. Anthropologist Nicholas Peterson wrote: “It is in the context of the collection that the minute particulars of his hundreds of pages of fieldnotes come into their own through their close relationship with the objects and images” (Peterson, 2005, p. 43). Thomson’s assistant Judith Wiseman spent more than a decade after his death transcribing thousands of pages of handwritten notes, and capturing as many cross-references between parts of the collection as possible with little regard for physical format or disciplinary divisions, reporting on her progress in 1978: “Cross referencing—to continue more or less for ever” (Wiseman, 1978).

And yet, the subsequent management and documentation of the collections has produced numerous problematic separations. Until recently the ‘literary estate’ remained the property of Thomson’s family, the ethnographic artefacts belonged to the University of Melbourne (donated by Thomson’s widow), and the natural history specimens were sent to what was then called the National Museum of Victoria (NMV), later Museums Victoria (MV). From 1973 the collections were co-located at MV due to a joint loan agreement between the museum, the university, and the Thomson family, though the natural history specimens were housed in a separate department due to disciplinary splits in the institution. Despite Wiseman’s work in the 1970s and 1980s (see Jones, 2018), in more recent years other splits have emerged. For example, though a published microfiche catalogue of the artefacts contained many cross-references (Ramsay, 1987), now that they are documented in the MV CMS these references are no longer publicly visible, and are only found in an internal notes field. A print catalogue of the mammal and fish specimens (drawn almost entirely from Thomson’s registration cards, notes, and diaries) fails to mention any of the material culture collection; and the Memory of the World entry for the Ethnohistory Collection does not even mention artefacts or natural history specimens. To take a single example, there are photographs and field notes about dugong hunting on Cape York Peninsula in the ethnohistory collection, including descriptions and images of the making of a dugong hunting rope. The rope (and others like it) are in the material culture collection, and there are dugong skulls and associated specimens in the natural history collection, all collected by the same person in the same context. For Indigenous community members or other users interested in the culture and practice of dugong hunting on the Cape, divisions between documentary heritage, natural science, and ethnographic collections and their different ways of preserving and cataloguing collections are not just irrelevant, they complicate the generation of meaning and understanding. The umbrella provenance of ‘Donald Thomson’ does little to surface the many significant relationships between collection items so prized by Peterson and Wiseman.

Whether considering objects, specimens, cataloguing data, or archives, there are different types and degrees of dissociation. When the Flinders’ chronometer was registered in 1937 it was already dissociated from its history, and from any documentation or metadata that might confirm its provenance. There are also many artefacts in museums that were apparently not registered on acquisition, like the ‘Gweagal shield’ in the British Museum. Some argue it is the shield collected by Captain James Cook and Joseph Banks on the east coast of Australia during their voyage to the South Pacific in 1770, while others remain unconvinced. The shield was finally registered in 1978; but, with no definitive evidence and no paper trail to follow, debate continues (Nugent & Sculthorpe, 2018; Shield, n.d.; Thomas, 2018). By comparison, the Jazz Messengers album and photograph discussed earlier are registered, documented, and managed in their own right; but the separate institutional and technological silos used for the collection and archives mean every subsequent user has to rediscover the association between material that sits on either side of the divide. This work becomes even harder when archives are not documented and searchable, particularly where institutional processes have resulted in artefacts and their associated archives being split across different departments. This is exemplified in Jones’ (2021) introductory discussion of Aboriginal stone tools sent to the British Museum in 1901 by Tasmanian amateur collector Reverend Christopher George Wilkinson. Due to institutional restructuring in the twentieth century, the tools are now held in a different department from the bound correspondence files containing Wilkinson’s extensive and evocative letters (Jones, 2021, pp. 1–2). Similarly, in more recent decades, developments such as the move away from shared central files for digital records can silo and fragment organisational documentation in ways that hinder discovery, access, and use.

Even more challenging is where archives and associated collection materials end up in different institutions. There are several factors that contribute to this kind of fragmentation. Many pre-registrar documenters (collectors, curators, source community members) moved between institutions as they pursued careers in the sector, while others relied on non-salaried incomes (or worked without pay, as in the case of women who were assumed to be financially supported by male family members). Curator-collectors moved from institution to institution, much as their modern counterparts, to secure jobs and advance their careers, often taking their field notes and research with them, and moving further away from source community relationships over time. Records tended to end up in the archives of whichever institution was home at career’s end, or wherever family members felt the curator had the strongest ties. Meanwhile, national and international networks of amateur and professional collectors sold objects to multiple institutions that could not or would not purchase whole collections, resulting in their (often arbitrary) break up along with documentation. Federated cross-institutional catalogues and linked data have improved this situation to some degree, but it remains a problem, not just for researchers, curators, collections managers, registrars, and archivists but for those source communities whose tangible and intangible heritage was collected and traded by institutions around the world.

Locating layers of information pertaining to collections is dependent on the implementation of common institutional practices. While most museums maintain a catalogue in one form or another, effective use requires attention to the histories of its production. William Sturtevant (1973, p. 45), Smithsonian anthropologist and curator, called for the need to interview museum staff about “specific systems of record-keeping” in order to understand “what lies behind” typical descriptive records such as catalogues—just one example of the over-reliance on implicit knowledge in managing collections (Jones, 2021, p. 102). There is a need to understand how catalogues have been arranged and indexed, whether the information has been sourced from surrogate catalogues in other formats, what additional information has been added, and where this information has come from—also known as data provenance: “where a piece of data came from and the process by which it arrived in the database” (Buneman et al., 2001, p. 316). The process of locating information beyond the catalogue is akin to an organizational ‘scavenger hunt’ as relevant documents (papers, photographs) may be kept, “in more than one location: near the catalogs, or in the library, archives, department office, registrar’s office, old safes and closets” (Sturtevant, 1973, p. 45).

The impact of dissociation and the effort required to locate relevant information is significant. At a broad level they result in a loss of meaning (Greene, 2002). Archival and object files function simultaneously as evidence and memory. They are evidence of the most basic function of the museum (to collect and interpret objects) and of the legal ownership of the collection, and as such remain active from a records management perspective. At Museums Victoria, for example, the archives program was developed to address the requirements of State Records and Freedom of Information legislation (McNulty, 2000). But, as with all archives, there are a multitude of other possible uses, only some of which relate to why the records were created and kept. As recognized by UNESCO, documents support engagement with and understanding of the past, providing: “the means for understanding social, political, collective as well as personal history” (2015).

In addition to UNESCO’s focus on state memory and history, dissociation also impacts the discoverability and use of museum collections and archives for Indigenous communities who see their ancestors, cultural heritage, and intangible knowledge collected, traded, bought, plundered, dug up, and stolen by generations of collectors. Fragmented documentation is one reason institutional inventories prepared under the Native American Graves Protection and Repatriation Act (NAGPRA) can take so long to compile, slowing the return process for ancestral remains. Writing about the process at the Peabody Museum, Barbara Isaac lists some of the reasons why an extension was sought, including the dissociation of human remains and related archaeological material, the need to navigate multiple (often outdated) cataloguing systems, and:

The documents that contain information relevant to identifying and collating these assemblages are scattered not only through the Peabody Museum but also across other Harvard archives, and at times are actually to be found in the archives of other entirely autonomous institutions that at one time had shared expeditions and interests (Isaac, 2002, p. 161).

Similar issues face those engaged in other repatriation and reparation work, including in response to growing calls for broader data and archival sovereignty and the return of associated archival records, photographs, audiovisual records, and other documentary heritage (Booker, 2023; Evans et al., 2018; Thorpe, 2021; Thorpe et al., 2020). Diana Marsh’s study of Indigenous community members’ barriers to archival use, discovery, and access calls attention to considerations of archives’ organization, contents, and users (Marsh, 2023, pp. 90–92). Using archives is a labor-intensive process that depends on knowledge and stamina as well as qualities of the archival material, which may be challenging due to the contents (such as handwriting or discipline-specific language), organization, and idiosyncratic documentation conventions. Source communities also face the problem of navigating dispersed records, or “archival diasporas” (Punzalan, 2014), reporting difficulties locating which archival repositories contain records of interest and understanding the complexities of diverse search interfaces.

As documentation surrounding the context of museum objects fragments, so does our understanding of the relationships among the people associated with objects. This makes it increasingly difficult to surface the many relationships with various actors over time, resulting in incomplete or inaccurate finding aids—often at the expense of the participation of source communities, marginalized individuals, and counter-narratives. This disassociated provenance knowledge has a narrowing effect on opportunities for deeper research and creates barriers to reconstructing the broad impact of exchange practices and networks of actors or, in extreme cases, even provenance itself.

A pertinent case study here is the relationship between Harvard University’s Peabody Museum of Anthropology and Ethnology (PM) and the Harvard University Archives (HUA). In comparison with many other coeval collections, the PM has an extraordinarily complete run of accession files, dating almost all the way back to its founding in 1866. The key actors of the PM’s earliest collections (1866 – 1909) were its curator/directors, responsible not only for the administration of the organization, but also for building its collections through donations and purchase. Jeffries Wyman and Frederic W. Putnam also published on the collections and taught courses in Harvard’s nascent Anthropology program while helping to establish the field as an academic and professional discipline in North America. Both also were active in archaeological expeditions (Wyman in Maine/Massachusetts/Florida and Putnam in Ohio) that added to the museum’s collections (Wyman, 1868; Putnam, 1884). Putnam’s collection development resulted in voluminous correspondence with commercial collector/scavengers, his assistants and bursary-supported associates (such as Zelia Nuttall, and siblings Francis and Susette La Flesche) as well as many amateur archaeologists. As the museum profession developed and matured, and registration became more standardized, staff had to contend with how best to organize the mass of provenances contained in ledgers, correspondence, (extraordinarily detailed) annual reports, trustee meeting minutes, and field notes. The principles of museum registration (which focuses on transfer of legal title) and archival original order are often at odds when it comes to these permanently active records (Demb, 2011).

Over the past twenty-five years a series of projects have been undertaken to arrange, describe, and make accessible accession files and other documentation. This included digitizing the PM’s earliest catalogue ledgers and attaching them to its CMS (Demb, 1998); and digitizing much of the accession file materials and linking them to the CMS (Simmons University, 2019), though the completeness of each original file and its digital proxy is not yet (and may never be) complete, due to privacy and confidentiality requirements pertaining to content of the files, the sheer volume of related dissociated documentation, and the complexities of its organizational and actor histories. It is unclear when the Peabody’s accession files were organized as a central file series, although they have certainly existed as such for many decades. Up until the late 1990s, they were held in file cabinets and some still included collections materials such as glass plate negatives and prints. It was common to find field notes and field notebooks in the files along with deeds of gift, correspondence, print-outs of catalogue records from earlier collections management systems, and registrarial forms. They were re-housed with funding from an Institute of Museum and Library Services (ILMS) grant. Materials documenting the larger context of the acquisition, such as field notes and field notebooks, were removed to the archives, documented by ‘separation sheets’ tracking their location; and materials considered museum collections ‘proper’ were removed to object storage. The first database inventory of the accession files was created, as the CMS’ registration module did not include this information. Prior to this exercise, the registrar had relied on cover sheets in each file. The preservation benefits to the project were significant, as were efforts to retain an audit trail of any removals to other locations within the museum.

However, dissociation of accession file materials had already occurred at the University, as the relationship of the Museum to central administration strengthened and became more formalized over time. Some years prior to the re-housing of the accession files, a sub-set of early director records were transferred to HUA with the best of intentions, to document the institution’s administrative relationship to the larger organization (Harvard University Archives, 1866-1942; 1851-1968). While laudatory in its approach to ensuring that its ‘associated’ or ‘affiliated’ organizations are documented centrally as part of the HUA mandate, this called into question whether the archival principle of original order was upheld; but perhaps even more saliently, the nature of the early curator/directors’ work, unlike their more recent successors, meant that the correspondence documented not only the founding and ongoing management of the museum, but also its collecting activities. The accession correspondence that had not previously made its way into the accession files is part of the director records now held at HUA. The volume of the correspondence and the archival practice of group level cataloging means that this correspondence remains divorced from the objects it documents, continuing the dissociation that already existed before its transfer. The only way to reconstruct the provenience of the objects documented in this correspondence is to review each page, noting where an object is mentioned, and then cross-check its descriptive elements, collector and donor name, and letter date against the PM’s online CMS and the digitized web versions of early Peabody Museum annual reports (which list many acquisitions at the item level with their accession numbers)—a labor intensive process made slightly less complicated by the recent digitization and cataloging of the PM’s accession files. It does not address the ongoing complexity of access to early fieldnotes, which are often in bound volumes too fragile to be consulted on a regular basis or digitized in a cost-effective manner.

While reassemblage may be a worthy undertaking when researching objects at their holding institutions, the separations of object and information is further complicated when viewed from the vantage point of objects in transit. The exchange of museum specimens was a regular part of administrative and curatorial operations between most natural history museums, scientists, and collectors in the nineteenth and twentieth centuries. While specimens (objects) were deemed mobile based on their status as duplicates (Nichols, 2021), relatively limited descriptive (catalogue) information tended to accompany objects (Nichols, 2022). At the Smithsonian Institution, a ‘Distribution of Specimens’ form was the primary document for recording basic descriptive information for specimens sent in exchange. While a collector name and Smithsonian catalogue number was commonly listed, there was no systematic transfer of the collector’s associated papers or field notes. The result of this practice led to widespread redistribution of objects, but extraordinary dissociation from their documentary contexts.

In an exchange from the Smithsonian to the PM in 1889, the Smithsonian sent a pair of moccasins described by their Distribution of Specimens form simply as “[Nature of Object] Moccasins, [Locality] Zuni, Ariz., [Collector] Col. J. Stevenson, [Catalogue Number] 41723, [No. of Specimens] 1 pr” (Woman’s Moccasins, 2023). While these moccasins remain in the PM collection, and the PM online CMS lists James Stevenson as collector and Smithsonian Institution as exchange participant (or donor), there is no reference to the extensive archives or publications of the Bureau of American Ethnology, which organized Stevenson’s collecting and research efforts, and whose records are held at the NAA in Washington, D.C.

Dissociation has many impacts. These include: breaking the relationship between items and evidence, potentially affecting the ability to establish provenance and legal ownership; disrupting the generation of meaning and understanding from distributed collections and archives; increasing the labour required to work with collections for curators, collections staff, and researchers (including an increased requirement for duplicated work); and complicating collections repatriation and reparation work to improve accessibility and usability by source communities.