THE NATURE OF ARCHAEOLOGICAL COLLECTIONS

Archaeological collections consist of material remains and their associated records. These material remains, generally including artifacts and “ecofacts” (or animal bone, shells, and soil samples), have typically been recovered from the ground. The number of artifacts found in any one collection can range from zero to the thousands, although most collections have more rather than fewer objects. Decisions on what is included in any collection are typically made by the archaeologist/principal investigator in consultation with stakeholders, including Native American tribes, and the repository. In some cases, human remains can also be found in an archaeological collection. Records can include paper, film, and digital descriptions or representations of the fieldwork as it was carried out and as the artifacts were removed from their archaeological context. Laboratory records and reports in both paper and digital format are also part of a collection. In sum, archaeological collections consist of artifacts; ecofacts; and paper, film, and digital materials, and each material type has its own long-term preservation needs.

Archaeologists and archaeological collections managers often distinguish between legacy collections and incoming collections. Legacy collections are defined as those inherited by a museum or archaeological repository. They are typically older, larger, and more poorly packaged or housed than incoming collections. In legacy collections, context or provenience information—the heart of the archaeological enterprise—is often missing (if it was collected at all) or is difficult to reconstruct. Artifacts can be missing too. The archaeologist who generated the collection is usually no longer active or available, and the collections themselves have been forgotten or overlooked, languishing in storage. Incoming collections—collections recently generated in the field and typically subject to modern standards for collection, processing, and transfer—are usually in better shape and are presumably of more relevant and immediate use to researchers. Although these distinctions are critical for collections managers, the differences for researchers can be less meaningful. Incoming collections recovered using incompatible methods and strategies, for example, can be difficult if not impossible to compare, making them as problematic as legacy collections for comparative research projects (King Reference King2016:5–6, 10–13).

Although Maryland is a small state, its holdings include some nine million artifacts recovered from more than 5,800 archaeological sites. Multiply these numbers by the numbers of repositories in other states and territories, and a rough sense of the magnitude of archaeological holdings in the United States alone begins to emerge. How likely is it that these materials will be used for research? The MAC Lab, which actively promotes collections accessibility, receives an average of 53 requests per year for access to its collections, but it may be the outlier. Compare the MAC Lab's experience with that found in a survey conducted by the Council of Virginia Archaeologists (COVA) in 2011 of facilities holding archaeological collections in that state. COVA found that, although the majority of responding organizations reported that collections in their custody were available for study, usually by appointment, more than 40% reported that their collections had either never or only rarely been used (White and Breen Reference White and Breen2012:9).

Curating archaeological collections is an expensive endeavor and more so if the collections are only rarely or never used. What, then, can repositories do to make collections in their custody more accessible for research?

IDENTIFY AND PRIORITIZE THE ARCHAEOLOGICAL COLLECTIONS IN YOUR CUSTODY

To make collections attractive for research, a repository's first step must be to understand the collections in its custody. Before the MAC Lab came online in 1998, archaeological collections in state custody had never been fully inventoried. This situation is not unusual. The 2011 COVA survey found that “many [repositories in Virginia] could not provide even a list of collection names stored in their repository” and two-thirds lacked a collections policy (White and Breen Reference White and Breen2012:10). The move into the MAC Lab in 1998 provided an opportunity to begin the process of inventorying the Maryland collections. Developing such an inventory is a basic first step for any repository and enhances preservation and security, as well as the ability to meet research needs.

The inventorying process begins with the creation of a spreadsheet or database that, for each collection, lists the site number, project, site function, cultural affiliation, principal investigator, level of investigation, number of boxes of artifacts, site date and context, presence of field and laboratory records, and current storage location. In preparing the inventory, each box, even if labeled on the outside, should be opened and its contents checked. Boxes can be mislabeled or contain artifacts from more than one site, and only a visual inspection can confirm their contents. One should identify any collections that contain human remains or associated funerary objects or both. In addition to requirements of the Native American Graves Protection and Repatriation Act (NAGPRA) for institutions receiving federal support, states and other jurisdictions often have laws defining the treatment and disposition of these materials.

The process of inventorying may also be a good time to list any potential preservation problems encountered, such as disintegrating bags, conservation needs, or evidence of insect activity. The inventory's development should be supervised by a curatorial or collections professional, but much of the work can be undertaken by trained volunteers. Engage the assistance of researchers who have requested access to a collection and other volunteers in this step to enhance the collection's usefulness.

The move of our collections to the MAC Lab, along with the development of the inventory, provided an opportunity to imagine how these collections could be used for research. In 1998, MAC Lab staff invited archaeologists from across the state to participate in a one-day workshop focused on identifying, ranking, and prioritizing collections most likely to be of research value. Participants in the workshop were selected to ensure a broad range of expertise with different site types. Many participants had worked in the state for years and were well aware of the sites that had been investigated and the resulting collections. At the end of the workshop, participants identified 30 archaeological collections with potentially high research value. The 30 collections covered all time periods and regions of the state and included towns and domestic, military, and industrial sites. Native American, European, and African occupations were well represented among them.

This priority list identified those collections that should be immediately rehoused in archival packaging meeting the state's newly adopted curation standards. Doing so addressed the preservation aspect of the lab's responsibility. At the same time, the list formed the basis for soliciting grant funds to assess the collections, recatalog them as necessary, and digitize the artifact catalogs, site reports, and lab and field records; this represented the MAC Lab's accessibility responsibility. Two grants from the National Endowment for the Humanities (NEH) allowed the MAC Lab to hire staff to undertake the project. Now that finding aids for the collections, along with digitized records, can be found online at http://www.jefpat.org/NEHWeb, researchers have the opportunity to review materials and determine whether a collection will be valuable to their research, all without traveling to the MAC Lab. The list also served to demonstrate the national significance of the collections to the NEH.

Identifying and prioritizing the 30 collections for attention transformed an impossible job into a manageable one. If identifying 30 collections is too much for an institution, begin by identifying the top 5 or 10. The point is to develop a feasible plan to make progress. In hindsight, our focus on archaeologists as the only stake-holders for prioritizing archaeological collections was short sighted and limiting. More than 20 years later, it is clear that Native Americans and nontribal community stakeholders are an important audience for archaeological collections. The MAC Lab has developed and nurtured these relationships with positive outcomes for all participants. Our recommendation is that tribes and nonprofessional stakeholders have a place at the table as their advice and support are critical to twenty-first-century archaeological practice.

ENSURE ACCURATE PROVENIENCE INFORMATION

Although archaeologists value objects, they value provenience even more. They spend a considerable amount of time in the field carefully recording provenience—an artifact's location in space and time—and maintaining associations of artifacts as they are removed from their original contexts. Ixchel Faniel and colleagues (Reference Faniel, Kansa, Kansa, Barrera-Gómez and Yakel2013) found that most “data reusers,” as the authors described collections-based researchers, want two things: (1) basic information about objects (when and where they were found and by whom) and (2) the chronology of a site, including stratigraphic information for all objects and relationships between strata. Collections with what Lyman (Reference Lyman2012) describes as “fine-grained” provenience are prized for their value in addressing temporal and spatial variation at a microscale. Almost all repositories, then, can anticipate that one of the first orders of business for collections researchers will be an assessment of whether provenience can be reconstructed for any given object or collection.

Theoretically, maintaining contextual information should be part and parcel of a repository's system of archaeological collections management. In reality, even incoming collections can arrive with poor provenience documentation. Rebecca Frank and her colleagues (Reference Frank, Yakel and Faniel2015:151–152) found that archaeologists, who appreciate the destructive nature of archaeological excavation, still resist responsibility for the long-term preservation of archaeological data. One archaeologist in their study lamented that a lot of contextual information “only seems to exist in the heads of people,” whereas others worried about the long-term preservation of digitally collected provenience information.

To assist archaeologists before they transfer collections to a repository, S. Terry Childs and Danielle Benden (Reference Childs and Benden2017) developed a checklist that includes protocols for tracking artifacts from the field to the lab to minimize the loss of context or provenience. Childs and Benden describe this checklist as essential for the “Collections Management Cycle,” or the link that connects the field and lab work with the collections work. If your repository regularly accepts collections, this checklist may be useful to distribute to the archaeologists using your facility. To this checklist might be added specific questions about how provenience or context was maintained and where that information might be found in the collection.

Once a collection is transferred to a repository, managers may determine that preservation needs require that like materials be curated together. Metals and organics, for example, may be pulled for storage in specially created microenvironments. In addition, objects may be retrieved for display or further study without proper tracking documentation. Through these types of activities, provenience information can be (and has been) inadvertently severed from objects. Protocols for tracking artifacts as they move through and within a facility are critical tools for preserving provenience information (see Buck and Gilmore Reference Buck and Gilmore2010; Kipp Reference Kipp2016; Reibel Reference Reibel2018).

For collections that have been prioritized for research, repositories can get ahead of the curve by assembling provenience information in a usable form. This can be a challenging task. Archaeological records are the first place to look for mining provenience information. Maps, especially site plans, are critical: if they do not exist, they should be reconstructed from the records if at all possible. The maps/site plans and any other records can be used to create a table or spreadsheet listing every provenience or soil layer that appears to have been excavated or recorded, including proveniences yielding no artifacts. The table should show any information about that provenience, including name, description, coordinates, elevations, and any other data that can be extracted from the records. Be aware that provenience systems vary widely, can be cumbersome, often have missing information, and can include errors, so avoid making too many assumptions until all the information has been collated. If the site has many components and the records are sufficiently detailed, use them to assign designations to the different periods or components. Perhaps most importantly, for future researchers and users, memorialize this work in a record that describes how the provenience system was reconstructed, including any assumptions that were made and other project-specific instructions.

Profile or cross-sectional drawings can be tackled next. Profile drawings are critical for establishing stratigraphic relationships between soil levels and the artifacts contained within them. These drawings, however, are often missing from collections because excavators may have instead relied on the recording of elevation data to maintain stratigraphic control. Labels found on artifacts and artifact bags can be very useful because they may contain information linking the artifacts to soil levels, sometimes in a code that has been forgotten long ago but that may be cracked using the information assembled in the provenience table. If principal investigators are still available, contact them; it is sometimes surprising what is and is not turned over to a repository. Any provenience information, regardless of where it may be found, is preferable to none at all (although Silberman [Reference Silberman2015] argues that materials that have lost all context have no “inherent” value requiring their retention, an important topic for consideration but beyond the scope of this article). Finally, if a site map does not survive, assess if the information, such as coordinates, exists to create one.

Software systems can be very useful for managing provenience systems. That said, there is a difference between collections management software and software developed for the analysis of archaeological data. Collections management software privileges objects and their current provenience—that is, where they might be found in a repository. That information is, of course, important. Nonetheless, although archaeological provenience might be included as part of a collection management record, it is typically not recorded in the most usable ways for researchers. Most researchers opt instead to use commercially available database products for archaeological analyses. Geographic information system (GIS) and computer-aided design (CAD) technology also have a role to play in the reconstruction and representation of provenience information. This technology is increasingly available and used by archaeologists to represent spatial relationships in digital form.

Finally, recognize that the physical arrangement of archaeological collections within a repository is always meaningful. Space, and the spaces occupied by collections, can reveal assumptions about what we value in collections and, by inference, what we value about the past. The collections deemed suitable to accession, how these collections are organized and stored, and who can and cannot have access to them are determined by many factors, including disciplinary practices, ethics, funds, interests of the museum, and contemporary understandings of social realities. Archaeologist David Palmer (Reference Palmer and Barnes2011) makes this point clearly and elegantly in his work focused on the study of Jim Crow practices in early twentieth-century Louisiana. Biases in the field toward larger sites yielding more artifacts, he observes, mean that sites with a smaller archaeological signature (and typically associated with poorer people and people of color) are often missing in collections. If a repository has an archaeological collections policy or collections statement, it may pay to prioritize these smaller collections or those types missing from your repository. Repositories without a collections policy should make it a priority to develop and adopt one (see Malaro and DeAngelis Reference Malaro and DeAngelis2012).

CREATE FINDING AIDS FOR THE COLLECTIONS

Once the inventory of collections is complete and one has an understanding of the archaeological context, a basic finding aid for each site or collection can be created. It should include a summary of the site's history and context, details about the archaeological investigation, and the size and nature of the collection. A description of the types of field and laboratory records associated with the excavation should be prepared, with quantities of each type of record given if possible. Any type of specialized analysis, including radiocarbon dating, ceramic vessel counts, or faunal or paleobotanical studies, should also be listed in the finding aid. Do final reports or report drafts exist for the excavation? Have there been publications, dissertations, or master's theses written on this work? Once they are assembled, making finding aids available online will make it easier for researchers to review your holdings. Finding aids prepared for the 30 prioritized MAC Lab collections can be found at http://www.jefpat.org/NEHWeb: they suggest the kinds of basic information users are seeking. Examples of similar finding aids can also be found at http://colonialencounters.org/index.aspx.

PROVIDE DIGITAL ACCESS TO COLLECTIONS

Traveling to a repository, especially one located far from a researcher's home base, can be time consuming and expensive. For this reason, a growing number of repositories are experimenting with the digital delivery of collection information. Providing information about a collection digitally can minimize the handling of artifact and records, improve accessibility, and enhance preservation. The digital delivery of information, however, comes with its own costs. Many institutions are turning to online data repositories for making their data available, including Digital Antiquity's Digital Archaeological Record (tDAR) and the Alexandria Archive Institute's Open Context (Sheehan Reference Sheehan2015). Others choose to manage their own data. The MAC Lab does both.

Beginning in 2000, the MAC Lab experimented with developing online access to its collections through the creation of web-based type collections, today known as “Diagnostic Artifacts in Maryland” (http://www.jefpat.org/diagnostic/index.htm). Supported by a grant from the National Center for Preservation Technology and Training, this online type collection began with Native American and colonial ceramics. It has since grown to include projectile points, postcolonial ceramics, porcelain, small finds, and table glass. This type collection tool provides detailed essays about each artifact type, as well as descriptions and high-resolution photographs of selected artifacts from the facility's collections. The purpose of the online type collection is to help other researchers identify, date, and interpret artifacts from their sites or collections.

In 2003, the MAC Lab added a new website focused on the 30 collections earlier identified as the state's most important ones for research. The website, Archaeological Collections in Maryland (http://www.jefpat.org/NEHWeb/), is very basic, with a finding aid, a site map rendered as a PDF, and scans of field and laboratory records, including paper and film records and site reports. Although they are not available online, artifact catalogs in Microsoft Excel are available on request from the curators. The website is easily navigable, although using the materials online can be cumbersome and usually requires researchers to download files and print them. Nonetheless, the website provides material that a collections researcher would encounter in hard copy if visiting the lab. This can help a researcher determine whether a visit to the MAC Lab is in order.

Maryland Unearthed (http://jefpat.org/mdunearth/), a newer website benefiting from two decades of managing the state's collections, lists features and other sealed contexts (such as middens or living floors) for almost 300 archaeological collections generated through excavation. A site summary is included for each collection, along with a searchable database for features and other contexts. The online search feature contains more than 4,000 records from sites ranging in date from 9000 BC to the mid-twentieth century. Although artifact catalogs recovered from these contexts are not included on Maryland Unearthed, a researcher can quickly assess what deposits might be useful to a particular research question and contact curatorial staff for additional information or to schedule a visit. The website's primary audience is archaeologists interested in learning more about the archaeological collections housed at the lab; the site is also easily accessible for students, educators, museum curators, and others interested in Maryland's past.

Additional websites include Curators Choice (http://www.jefpat.org/curatorschoice.html), a monthly series highlighting a significant or unusual artifact chosen from the collections by staff, and the blog, Maryland History by the Object (https://jeffersonpatterson.wordpress.com). This blog uses an artifact from the collection as a springboard for exploring larger topics in Maryland's past, such as trade, industry, politics, sports, or religion. Maryland Archeobotany (http://www.jefpat.org/archeobotany/Home.aspx) aggregates paleobotanical data from the lab's collections and is one of the few paleobotanical websites. Maryland Archeobotany is complemented by the Wood and Charcoal Identification website (http://www.jefpat.org/WoodAndCharcoalID/index-WoodAndCharcoalID.htm).

PROVIDE INCENTIVES TO OFFSET COSTS FOR ON-SITE RESEARCHERS

The MAC Lab, located in a rural part of the state at least an hour's drive from major population centers, has created a competitive fellowship program to offset the costs of on-site research requiring five days or longer. The fellowship program is funded annually by a donor, and fellows are required to conduct research using materials curated by the lab. If a fellowship program is not feasible, provision of a smaller stipend, meals, wi-fi, and a dedicated workspace, as well as available accommodations nearby, can serve as incentives to researchers.

TRANSFERRING COLLECTIONS AS A RESPONSIBLE OPTION

As the foregoing sections indicate, managing archaeological collections and making them accessible for research, exhibition, or education require investments of time and money. Anticipating what researchers might be looking for requires an understanding of the kinds of research questions that materials from a particular collection may generate. Not surprisingly, many archaeological collections are housed in facilities that may not have these resources or staff with the needed expertise. For example, the survey of repositories in Virginia revealed that “many of the survey respondents were historical societies and other smaller institutions, often with no professional archaeologist on staff” (White and Breen Reference White and Breen2012:13).

If a repository lacks the most basic requirements for managing archaeological collections—at least one curator or collections manager with a background in archaeology, a collections policy, and the ability to make collections accessible to a broad and diverse audience—the repository's staff should carefully consider whether retaining archaeological collections is in the best interest of the collection, the institution, and the potential audience for it (Sullivan Reference Sullivan1993). Transferring custody of a collection to an appropriate facility can be fraught with challenges, including practical and emotional ones, and requires owners to place the collection's welfare above the institution's interests. This can be hard to do and often spurs concerns about a community's control of its heritage. Transferring a collection is not simply a matter of driving boxes to a new repository but a methodical process that itself requires resources. In addition, there may be no available receiving institution. These concerns must all be considered in any effort to transfer collections, but they should not dissuade a repository from making a decision that enhances a collection's preservation and accessibility.