II. Operationalizing the art research collective collection

In response to these needs, OCLC Research launched the Operationalizing the Art Research Collective Collection (OpArt) project.Footnote ¹ The goal was to help art libraries address their sustainability challenges through collaborative approaches, by:

• • Demonstrating the value of data-driven analysis to identify collaborative opportunities and partners, through detailed analysis of collection and resource sharing data.

• • Exploring models of art library collaboration, through case studies of real-world partnerships to understand how art libraries create value within collaborations.

In 2023, OCLC Research published two reports documenting our findings: Sustaining art research collections: Using data to explore collaboration,Footnote ² and Sustaining art research collections: Case studies in collaboration.Footnote ³ This article focuses on findings from the first report, which covers the quantitative analysis, and is based on a talk presented at the 2023 IFLA Art Libraries Section Satellite Conference.Footnote ⁴

III. The art research collective collection

A collective collection is ‘the combined collections of two or more institutions, viewed as a single, distinct resource, usually through aggregation and analysis of metadata about the collections.’Footnote ⁵ OCLC Research has done a great deal of investigation into collective collections, usually with the focus on the collections of a pre-defined group of libraries. But the OpArt project is our first study of a collective collection focused on a specific subject area: art. A disciplinary approach to collective collection analysis offers insights into how the scholarly record in a specific discipline is defined and described, and what those characteristics might suggest for partnerships and innovation.

For this project, we defined the art research scholarly record broadly, as materials that are of interest to art research scholars, that support and document art research and are distributed across collections of art libraries and other institutions around the world. The materials in these collections form the global art research collective collection, which in turn is an example of a disciplinary collective collection.Footnote ⁶

The disciplinary collective collection approximates the disciplinary scholarly record, so stewardship of the art research collective collection is, in practice, stewardship of the art research scholarly record. But the art research collective collection is still an abstract concept. For our project, we used data from WorldCat, the world's most comprehensive database of information about library collections.Footnote ⁷ WorldCat data allowed us to transform the concept of an art research collective collection into a tangible set of library holdings through the creation of a proxy art research collective collection – the combined holdings of 85 institutions in the US and Canada that explicitly support art research as part of their mission. While the resulting collection is only a fraction of the global art research collective collection, it nevertheless offered a tractable data set suitable for exploring patterns of analysis to help identify collaborative opportunities in art research collections.

In addition to the bibliographic data used to construct the proxy collective collection, we also explored data from WorldShare ILL, OCLC's interlibrary loan network which includes more than 10,000 libraries and is the largest ILL network in the world.Footnote ⁸ The data covers the period 2017–2021, and involves transactions where a proxy group member was either a borrower or lender. Taken together, the collection and resource sharing data equipped us with a quantitative resource with which to explore collaborative opportunities in an art research collection context. The data-driven approach we adopt in this study is made possible by the unique aggregation of bibliographic and holdings data in WorldCat, as well as WorldShare ILL's extensive network of resource sharing partners.

IV. Analysis

The proxy art research collective collection featured the combined holdings of 85 institutions, accounting for 8.6 million distinct publications (roughly equivalent to FRBR manifestationsFootnote ⁹), based on 16 million total holdings across all members of the group. Even though the proxy collection is only a fraction of the global art research collective collection, it is nevertheless a collection of significant size. We examined the characteristics of this collection across a wide variety of dimensions, such as publication date ranges, languages, countries of origin and others, all of which are documented in our report. In this section, we will focus on findings pertaining to holdings overlap and material types.

With 16 million total holdings but only 8.6 million distinct publications, the collective collection clearly exhibits some degree of overlap in group-wide holdings. Figure 1 illustrates the overlap pattern present in the collection. The finding that stands out is that about three quarters of the publications are held by only one member within the group, so a great deal of uniqueness is present across the 85 individual collections. But there is duplication as well: for example, about two percent of the publications are held by more than ten institutions within the group, accounting for nearly 150,000 distinct publications.

Fig. 1. Holdings Overlap: Percent of publications, by number of proxy group member holdings.

Our US- and Canada-focused proxy group is just one grouping of institutions relevant to art research. As a point of comparison, we also examined the holdings of the internationally distributed Art Discovery Group Catalog (ADGC). The ADGC is an art-focused research subset of WorldCat, based on the holdings of nearly 60 institutions worldwide, as well as several networks and union catalogs. Before calculating the size of this collective collection, we removed the holdings of members that are in both the ADGC and the proxy group. With this adjustment, the ADGC collective collection accounts for 12.7 million distinct publications, based on 15.5 million total holdings.

We found that the overlap in publications between the ADGC and proxy art research collective collections is about 20 percent - in other words, only one in five publications in the proxy collective collection were also held by at least one member of the ADGC group (proxy group members excluded). The key takeaway is that given two art research-focused collective collections, constructed from two separate groups of institutions, the overlap across the collective collections is quite low. This reinforces the point that there is distinctiveness across art research collections, reflecting different collecting strengths and priorities.

Returning to the US- and Canada-focused proxy collection, we also examined the distribution of material types. The salient finding is that the vast majority of the publications in the collection are books or book-like objects. But as Figure 2 indicates, there is a wide range of other material types present too, including serials, musical recordings, and images.

Fig. 2. Distribution of material types in the proxy collective collection.

Drilling down deeper into the content of the proxy collection, we examined material types frequently found in art research collections – the material types that art libraries prioritize and specialize in collecting. As Figure 3 shows, exhibition catalogs, auction catalogs and artist files are particularly prevalent in art research collections; exhibition catalogs alone account for nearly 700,000 distinct publications across the holdings of the proxy group.

Fig. 3. Special categories of art research materials in the proxy collection.

We calculated the holdings overlap within the proxy group for these special material types. Figure 4 shows the holding overlap patterns for the five most frequently held types: exhibition catalogs, auction catalogs, artist files, artist books and trade catalogs. These materials exhibit a diverse range of holding overlap patterns, with some materials more widely held than others. For reference, the figure includes the holdings overlap for the overall proxy collection, where we found that about three quarters of the materials accounted for only a single holding in the group. Using that result as a benchmark, it is clear that some of the special types – artist files and trade catalogs – have an even larger percentage with only a single holding in the group. In short, within the proxy art research collective collection, there are some categories of specialized materials that are very rare or even held within the collection of a single art library.

Fig. 4. Holdings overlap (special categories of materials): Percent of publications, by number of proxy group member holdings.

An implication of this data analysis is that art libraries can bring unique value to collaborations, based on the specialized nature of their holdings. Figure 5 shows the special material types identified in Figure 3. The last column shows the number of ILL transactions in our resource sharing data set involving those material types; as mentioned earlier, these are transactions in which one of the proxy group members is either a borrower or lender. For example, exhibition catalogs were loaned almost 25,000 times in the five years covered by our data. Detailed examination of this data showed these materials were loaned almost exclusively by the art libraries in our proxy group, and the most frequent borrowers include a number of large academic libraries.

Fig. 5. Special material categories collected vs. shared by art research libraries.

This is an interesting indicator of partnership opportunities for specialized libraries like art libraries: the fact that they are the primary holders of these special materials and are willing to share them makes them valuable partners for other institutions seeking access for their patrons. Therefore, these special material types are especially interesting as a starting point for thinking about opportunities for innovative collaborative partnerships.

V. Findings and implications

Two important findings emerge from our study of the proxy art research collective collection. First, the art research collective collection is a specialized collection, characterized by many rare and unique materials, as well as specialized material types, that are of particular interest to art research scholars.

Second, this is a decentralized collection, spread over the local collections of many institutions, with many materials held by only a few or a single institution. The art research collective collection is built on the wide range of collecting strengths and specializations of many local collections.

Because the art research collective collection is both specialized and decentralized, no single local collection can offer a representative view of the scale and scope of the art research scholarly record. This scale and scope is instead manifested through the collective collection of institutions located all over the world. The art research scholarly record is therefore a networked collective collection, and its stewardship must necessarily be a collective responsibility.

This suggests that to maximize access to and use of the art research collective collection, and to ensure its sustainability, art libraries need innovative stewardship models. Since stewardship of the art research collective collection is a collective responsibility, collaboration is a key pathway for achieving this goal. So innovative and collaborative stewardship models are needed. Where can art libraries look for these models? Based on our work, we believe there are at least three channels through which institutions can develop innovative, collaborative stewardship models: partners, collections and resource sharing:

Partnerships: The first channel for innovation is the partners that art libraries choose to collaborate with. Look beyond peers – that is, other art libraries – and consider partnering with other types of institutions that are stakeholders in art research. The distinctive collecting priorities, skills and expertise of art libraries can be a valuable complement to partners of many types.

For example, in our second OpArt report, we highlighted case studies of art libraries successfully partnering with academic libraries. These models are potentially replicable in other circumstances. Moreover, potential partners of this kind may be both geographically nearby and numerous. In our work, we identified academic libraries in the US and Canada whose collections significantly overlapped with the proxy art research collective collection. This suggests a shared interest in collecting art research materials. Figure 6 shows the locations of these academic libraries, indicated by the green dots, that are in the Northeast region of the US. The orange dots are the locations of art libraries from the proxy group that are also in the Northeast US. As the data shows, there are several academic libraries near each of the proxy group art libraries, each representing a potentially valuable partner.

Fig. 6. Regional view of proxy group members and academic library co-collectors.

Collections: Another channel for innovative stewardship collaborations is through leveraging complementarities across collections. Understand how your collection aligns with what others have, and how this can be leveraged to extend access and use. Because the art research collective collection is diffused over many institutions, and is rich with distinctive and rare materials, any collaborative partner will likely bring a unique strength within their collection.

This point is illustrated by the case study in our second report that examines the membership of OCAD University, a small art and design university in Toronto, with the Ontario Council of University Libraries (OCUL), a regional consortium of university libraries of various sizes and types. OCAD University's specialization in art and design makes it a welcome partner with a unique collection strength and expertise within the OCUL membership.

Figure 7 provides further evidence of the value of specialized collecting strengths, in the form of an excerpt from a table listing subject headings in the proxy art research collective collection that have about 5,000 holdings across the proxy group – i.e. subject areas not heavily collected compared to other topics. This data shows the richness and variation in subjects touched on in art research-focused collections and suggests opportunity for collaboration. For example, not everyone has the resources to prioritize collecting in, say, Outdoor Sculpture. A key implication of this data is that there is scope to think more deeply about how to bring collecting specializations together in a collaborative setting.

Fig. 7. Sample of subjects with about 5,000 proxy group member holdings.

Resource sharing: Finally, a third channel for innovative, collaborative stewardship is resource sharing, both in person and via interlibrary loan. High degrees of specialization, combined with many rare or unique materials across local collections, suggests the need for robust resource sharing mechanisms to maximize access and use. This may include embracing greater openness to sharing, such as expanded sharing of special collections, or extending borrowing privileges to researchers at institutions outside the art library community.

An example is the partnership between the Museum of Fine Arts, Houston (MFAH) and Rice University, detailed in a case study in our second report. In exchange for sharing Rice's offsite storage facility for monographs, MFAH extended borrowing privileges to Rice faculty and students – a benefit of special interest to Rice, which had recently established a new doctoral program in Art History.

Art libraries are already heavily engaged in resource sharing, both within the art library community and beyond. Figure 8 shows an image generated from interlibrary loan data from the Getty Research Institute, one of the institutions in the proxy group. This data, for 2019, shows the geographic pattern of lending transactions originating at Getty that year. This is just one institution, but it illustrates that both the need for art research materials, and the ability to share them, is truly global.

Fig. 8. Getty Research Institute: Geographical pattern of interlibrary loan lending transactions, 2019.

VI. Conclusion

Looking at the art research scholarly record through the lens of a collective collection, and conceiving of it as a networked collection with shared stewardship responsibility, is an opportunity to reflect and move forward on some key questions:

• • What are the opportunities to create value through collaborative stewardship of the art research collective collection?

Although the art research collective collection, as it is represented in WorldCat, is still largely an analog collection, art libraries will continue investing in digital and online access to their collections. Limited resources are available for digital stewardship work, and the environmental impacts of digital collections are a looming issue. This underscores the need to reduce duplicative work across institutions, and to work together to amplify what is rare and uniquely valuable to scholarship. So we can also ask:

• • What innovations in stewardship are needed to support shared digital evolution and sustainability?

And lastly, as we work to operationalize innovative partnerships, we need to reflect expansively on:

• • Who are the potential partners in these collaborations?

The findings presented here illustrate the role data-driven analysis can play in supporting efforts to shed light on these questions, taking into account different groupings of institutions, collection contexts, and institutional priorities.

The focus in this article has been on the findings of our first report, which examines how collection and resource sharing data can be used to identify valuable collaborative opportunities and prospective partners. But this is not the full story of collaboration for art libraries. There is also an art to collaborating effectively: in organizing, operating and sustaining successful partnerships. This art of collaboration is practiced in many existing partnerships involving art libraries. In our second report, Sustaining art research collections: Case studies in collaboration, we document the on-the-ground experiences of participants in three such collaborations: the Museum of Fine Arts, Houston and Rice University; OCAD University and the Ontario Council of University Libraries; and the Worcester Art Museum and the College of the Holy Cross.Footnote ¹⁰ These are all examples of libraries affiliated with institutions with an art research focus partnering with academic libraries.

We constructed a detailed case study for each partnership, based on interviews of people directly involved in building and maintaining the collaboration. In addition to the details of the collaborations, we also gathered practical advice and lessons learned from our interviewees on the art of successful collaboration. We synthesized the findings from the interviews into a set of general recommendations for effective collaboration, which, although derived from experiences in the art library community, nevertheless apply much more broadly. We encourage readers to explore this report to learn more about the practical art of collaboration.