For generations, most research libraries have had employees with deep subject expertise. Once known as bibliographers, these scholars and librarians originally focused their efforts on selection for collection building. Today, there is real anxiety about the role of subject expertise and academic liaisons in research libraries. We argue that evidence about scholars’ practices and needs should be a key input into reorganizing library subject expertise.[1]

Librarian subject expertise and liaison roles

At many research libraries, the role of subject expertise is less needed for selection of general materials in key humanities disciplines. Demand-driven acquisitions have pushed selection to users, while content bundles, consortial purchasing, and approval plans have aggregated decision-making in another direction. To be sure, there remain material types, world regions, and even specific fields where these patterns have been slower to emerge if they will emerge at all. But broadly speaking, there has been a decline in the relative need for subject expertise in support of collection development for general collections.[2]

This is not to say there is no value to subject expertise. To the contrary, there are enormous opportunities to utilize subject expertise in other ways. Specifically, as we emphasize in this issue brief, there are opportunities for subject expert librarians to partner with scholars and students to support their research, teaching, and learning.

Over time, there have been substantial efforts to reshape subject bibliographer roles into subject liaisons. Many research libraries in doing so have shifted or expanded the responsibilities of these professionals albeit while retaining the structure of these positions as aligned on a disciplinary or departmental basis with the faculty. That said, even the most well-resourced libraries have never achieved comprehensiveness in providing subject expertise, with unevenness of opportunities for service provision and engagement as a result.

Today’s library leaders want subject specialists and liaisons to provide strong engagement and valued services.

Today’s library leaders want subject specialists and liaisons to provide strong engagement and valued services. While many libraries have turned once again to hiring PhD’s who can bring deep subject expertise to these roles, others are seeing that subject expertise may not be the only, or in some cases most important, ingredient in formulating a valuable liaison program.[3]

At a basic level, in evaluating the efficacy of existing liaison programs at specific libraries, several analyses have found that key benefits and improvements to the model hinge on increased outreach, awareness, and engagement.[4]

One prominent direction has been to consider how to pair subject expertise with functional expertise. A number of libraries have integrated subject liaison positions with geospatial, statistical and data, digital humanities, and other forms of expertise, including undergraduate instruction and information literacy. In some cases, these have yielded hybrid or matrix models for responsibility. Implicit in these models is the question whether among liaisons with departmental affiliations, subject expertise is giving way to becoming, at least in part, “account managers” with responsibility for connecting “clients” with resources.[5]

Others have focused on a new vision for the role that is yet further radically decoupled from collections responsibility. Anne Kenney’s issue brief on liaison work emphasized the importance of aligning liaison roles and defining their success in conjunction with the university’s mission. In her framing, academic productivity was a key factor.[6]

And in recent years, the Association of Research Libraries has sponsored a large-scale effort to provide the basis for reformulating liaison roles and liaison work.[7]

A full literature review would include many other publications and initiatives over the past decade. There is no shortage of interest in redefining liaison roles and reframing subject expertise.

Disciplinary structure

Notwithstanding this interest in redefining the roles, few observers have advocated for a change in the disciplinary, subject, or departmental structure for organizing subject expertise that has been with us since the bibliographer days. Where a disciplinary approach is not taken, it is most typically because libraries cannot afford the number of subject specialists that would be required and instead group disciplines together.

Most research libraries organize liaisons on a disciplinary basis, built on the reality or legacy of subject expertise.

To be sure, there are more fundamentally different approaches. Some STEM-focused institutions have never had a model organized primarily on librarian subject expertise. And, in terms of recent efforts to rethink this approach, the University of Kansas reorganized its engagement roles to focus on user communities such as undergraduate and graduate students, along with faculty members, rather than on an ultimately disciplinary basis.

But these alternatives are rare. Most research libraries organize liaisons on a disciplinary basis, built on the reality or legacy of subject expertise.

Scholar practices

We developed the Ithaka S+R series of studies of the research practices of scholars to help libraries and other academic information organizations better support scholars. Having completed five such projects, with four more in progress at the time of this writing, our work reflects a key underlying theme in liaison discourse, the call to “focus on what users do” as opposed to “what librarians do.”[8] Of the five projects completed, three of them focus on fields that are largely humanistic in their methods, history, art history, and religious studies.[9] For further information about the methodologies for the three humanistic projects, see Appendix 1.  In subsequent sections of this paper, we examine what we have learned from these projects that bear on the question of how libraries may wish to organize and deliver subject-related expertise to scholars.

Our research builds on ongoing efforts within library and information science to recognize and understand the unique information needs of scholars by discipline and in the humanities more widely.[10] Each of our projects is “scholar centered,” which is to say we do not ask scholars how they engage with library services, collections, or employees but rather in a more open-ended way what their experience and practices are like as researchers. Our intention is not to evaluate the “efficacy” of liaison services as they are currently organized but rather to determine what services if any may be needed.

The findings from these projects offer a number of opportunities for thinking through common challenges associated with discipline-centric models of library liaison services including ensuring staff capacity for meaningful patron engagement, providing sufficient expertise across subject areas, and recognizing the evolution within and fluidity between disciplinary boundaries.

Common practices and needs: Who do humanists rely on?

Looking across the humanities, our research has found that scholars share commonalities in terms of who they identify as the main actors involved in their research processes. Our findings point to four kinds of actors humanities scholars generally describe engaging with throughout their research: the scholar themselves, their colleagues, graduate students, and information professionals.  Crucially, the majority of humanists continue to position themselves as the central actor in their research activities.

Humanists amass large information collections over the course of their careers. Whenever possible they create copies of primary content when visiting archives and special collections to conduct to analysis off-site and reduce their amount of research travel. These personal collections fill voids where archives and special collections are unable to digitize and make collections available. The advent of highly portable and cheap scanning and photographing technology has made the prevalence of personal collections of primary source copies that much higher. Humanists also continue to maintain personal collections of scholarly books. They will purchase books if they are unavailable through their library and/or they are planning to use the works for sustained analysis.

Scholars amass large information collections in both paper and digital formats[11]

Humanists report only minimally sharing primary or secondary information they collect with other scholars or institutions. And, in contrast to many scholars in the sciences, collaborative research activities in the humanities remains low.[12] Humanists do report working with students; however, students are commonly positioned as having a support role, such as through research assistantships, rather than as full partners or collaborators. Examples of the kinds of research support activities assistantships involve include: finding and retrieving secondary content provided by the scholar, conducting literature reviews, and transcribing oral history interviews. Some humanists also report that they benefit from working with students because the students introduce new research techniques or tools, for example, for literature searching or citation management.

Humanists not only position themselves as the main actors in their research because their collaborative research activities remain low but also because they perceive their advanced research skills as central to their identities as advanced scholars. Over the course of their careers scholars develop skills and knowledge in research that are finely attuned to their specialized research areas. Their ability to conduct research in their subject area is often considered essential to their positionality as experts in the field. Examples of how this deep knowledge manifests includes awareness of the content of specific archives, museums or other special collections, and close relationships with the stewards of those collections. Humanists also perceive their expertise as extending to secondary material discovery because of their responsibility to keep up with relevant literature and ability to discern through sources. The networks humanists build with other scholars in their field is also reflective of this culture of specialization. Humanists report the value of their collegial relationships as a mechanism for keeping up through conference presentations and informal interactions and resource recommendations.

Humanists exhibit clear patterns in their engagement with and reliance on information professionals. Crucially, humanists are more likely to rely on information professionals for support seeking primary rather than secondary information for their research. These professionals are typically located in archives, museums and other special collections not located at their home institution and they are more likely to be archivists and curators rather than librarians. Their support can be crucial to finding information, particularly when working with collections that have not yet been processed and/or have not been made fully available online. This points to the need to think beyond the secondary literature and to a variety of primary sources—including those in general and digital collections such as newspapers and images—rather than just special collections.

In contrast, humanists are much less likely to consult the subject specialist librarians at their home institutions because they are not perceived as experts. Even if, as appears to be increasingly the case, subject specialists have advanced degrees in the relevant subject area, subject expertise at a disciplinary level is not what is being sought. Rather, for research support humanists are looking for engagement at the level of their own sub-discipline, which is rarely available through the library.

Rather, humanists value librarians and libraries at their home institutions more for their teaching and learning support than as partners in research.

This is not to say that humanists do not value librarians as partners. Rather, humanists value librarians and libraries at their home institutions more for their teaching and learning support than as partners in research. Many are concerned that students, particularly graduate students, have gaps in their research skills. They do not see themselves as having full responsibility over ameliorating those gaps and they are open to libraries providing support in this area.[13]

Common practices and needs: What tools do humanists rely on?

Humanists have commonalities in the discovery and information management tools they rely on for their research. Platforms offering comprehensive discovery and, increasingly, digital access to materials, are generally improving scholarly research experiences for humanists.

Whenever possible, humanists prefer a single easy to use, comprehensive digital discovery platform, but in practice they often still need to check across multiple platforms. They desire their discovery platforms to have more predictive qualities and approximate this by using Amazon’s recommendations function. They increasingly are using social media, particularly academic social networking sites such as, but they rely on these tools more for self-promotion than focused information discovery for specific research projects.

Humanists generally experience greater challenges seeking primary information than secondary information because there are relatively less comprehensive digital discovery options available. They do not generally report using systematic methodologies towards digital discovery.[14] When seeking secondary content, Google Scholar is increasingly preferred over other discovery platforms. When seeking primary information they consult primary-source specific discovery platforms, Google, colleagues, collection websites and stewards of specific collections.

Working with primary and secondary information produced and held beyond the West is also a major challenge. Online discovery for primary collections is uneven based on locale, and the conditions at and regulations for using collections varies widely. For some scholars, travelling and using information on site remains central to their research activities. Humanists also experience challenges because their home institutions often do not subscribe to databases of secondary content from beyond the West due to cost, which necessitates developing networks and travelling abroad to acquire content.

Humanists amass collections of primary and secondary information over the course of their careers that they manage on their own with minimal intervention. These collections include information in both analogue and evolving digital formats. As discussed earlier, their collection of primary materials is increasingly facilitated through advent of cheap copying and scanning tools, such as portable scanners and scanning apps that utilize cell phone cameras. This information is generally organized and stored idiosyncratically with minimal intervention or support, such as through software or institutional storage programs. Idiosyncratic approaches to information storage and organize reflect that the collections are mainly for humanists’ personal use and rarely shared with others. They encounter little to no requirements for how this content is stored and preserved because the information collection rarely falls under the purview of external review, such as a requirement for grant applications or through institutional review of human subjects research.

Variance by discipline and sub-discipline

Our in-depth work in art history, history, and religious studies is helpful for teasing apart where the research support needs of humanists converge and diverge. Research support needs can diverge depending on the primary content that comprises the bases of study. For example, while all humanists amass large personal collections of information and these collections are generally not systematically organized or stored, art historians are more likely to rely on visual and material cultural material for their analysis. When considering supporting the information management needs of art historians, therefore, the ways in which they work with visual content must be taken into account. Their focus on working with digital images also means that many art historians are especially attuned to considerations of image quality, provenance, and copyright. In contrast, historians generally place a greater emphasis on working with documents, which leads to needs associated with archives, such as improved mechanisms for digital discovery and better archival training for PhD students.

Considering that scholars are highly specialized within their disciplines, however, researcher activities and their attendant support needs can also be understood as more granular than that of their general discipline. For example, in art history, architectural historians and archaeologists rely on three dimensional modelling. In history, oral historians collect and work with recorded testimonies and must contend with the complexity of preserving and making those materials available. In religious studies, scholars in biblical studies rely on software which enables them to compare and annotate canonical texts across various languages.

While the majority of scholars across the three fields we studied did not express strong needs for digital humanities research support, it is important to recognize that engagement with digital humanities, while relatively low overall, varies in degree of uptake across the humanities. As evidenced by some work with 3D modelling, some approaches to digital art history are at the forefront of the field and differ widely from other methods that have emerged through the digital humanities. Specific service needs appear to be held at a sub-disciplinary (or perhaps trans-disciplinary) level.

Sub-disciplinary needs are rarely met today by the liaison model. The most well-resourced research library at a university with strong emphasis on the humanities might have two liaisons for history, one for American history and one for European or global history. But most historians, for example, organize their work around a focused geographical area and time period. Art historians organize their work in clear subfields, such as Greek and Roman, African, Modern, etc. Religious studies scholars not only often align their work with disciplines in either the humanities or the social sciences, but also further distinguish their work by religion, time period, and geographic area under study. Scholars often find themselves relying on archivists who know a given relevant primary source collection extremely well and can provide contacts for and advice about other archival holdings, rather than on a library subject specialist at their home institution who is mostly focused on the published literature. But it is unusual that there is someone in the library they can turn to for help in keeping their workflows and toolsets up to date, when these vary based on sub-disciplinary practices or even individual scholar needs.

A further tension is that humanities scholarship can be inter-disciplinary. Religious studies is particularly instructive in this regard because it is a traditionally an interdisciplinary pursuit. Scholars in religious studies may be aligned with the humanities or the social sciences and their research support needs reflect this alignment accordingly. Furthermore, religious studies scholars’ work aligned with the humanities may draw on such disciplines as American studies, history, literature, linguistics, art history, classics, and various area studies. Many receive at least some of their academic training in another humanities discipline, and they continue to attend conferences and publish alongside others in that discipline while concurrently identifying as religious studies scholars.[15]

Redefining the role

These findings suggest a few different ways that the role of liaison or subject specialist could be rethought. In this section, we will suggest three fundamentally different models. Our objective is not to suggest that these are the only or the best models available but rather to show how we can link evidence about scholarly practices into planning the right liaison model.

The first model is for an integrated macrodisciplinary team. This model recognizes the extensive commonalities in the practices and needs across these three humanities fields. This model ends the alignment of individual subject specialists with a given field of study. Rather, subject specialists are reorganized into an integrated team to support the needs of a macrodiscipline like the humanities. Members of the team would remain responsible for regular engagement with a given group of scholars, not necessarily on a departmental basis, in order to stress the importance of individual rather than collective engagement. Every team member would also provide secondary coverage for another colleague, to ensure that gaps as a result of leave or vacancies did not yield a lack of coverage by the library. Team members would refer researchers to other team members and other colleagues with the appropriate expertise rather than feeling the need to solve every problem themselves.

Ultimately, this model is a shift away from attempting to provide subject expertise at the disciplinary level. In restaffing for this model over time, an integrated macrodisciplinary team would incorporate individuals with complementary skills who together as a group can support the information needs of the humanities. The team itself would emphasize less disciplinary subject specialization, so in staffing this model libraries would reduce job requirements for subject-specific background and expertise while increasing the importance of broader humanistic expertise. In parallel, they would likely increase job requirements for functional and methodological expertise that can become part of the ongoing suite of services offered in support of the humanities. This model would facilitate peer learning among team members.

But if such a model makes sense for organizing the common needs across the humanities fields, it provides little basis for specialized needs, on either a sub-disciplinary or methodological basis, that we have heard about from scholars again and again. Perhaps a collaborative approach can be considered.

Think about a library that is part of a research university consortium or system, such as the University of California system, the Big Ten Academic Alliance, the Association of Southeast Research Libraries, the Greater Western Library Association, or the Ivy Plus. While the scale of these collaborations varies in terms of the number of institutional members, each of them has an extraordinary array of library collections, expertise, and services.

To this point, the vast majority of the collaboration across these initiatives is in the area of collections, whether it be collaborative licensing, resource sharing, shared print, or digital initiatives. Where initiatives have been pursued to share expertise, they have sometimes focused on less well-resourced area studies priorities (as 2CUL pursued).[16] But there is no inherent reason that such collaborations, or others like them, could not move more deeply in the direction of shared subdisciplinary expertise.

For example, scholars express a need for subdisciplinary expertise–pre-Columbian Andean art, post-colonial Caribbean history, and medieval Chinese Buddhism. It is impossible to provide this level of granularity through a locally organized liaison program, but expertise could be coordinated and shared across institutions. This would push liaison programs much further towards support for primary source discovery, key research questions, and similar research support, as opposed to supporting teaching and learning, which could have benefits but also tradeoffs.

In this model, one can imagine that there could be opportunities to deepen collaboration with scholarly societies. Even today, certain societies include programming for and by subject librarians at their annual meetings. Perhaps societies could be a vehicle or clearinghouse for such collaborations, rather than or in supplement to more traditional library collaboration vehicles.

In another collaborative model, libraries could provide shared methodological expertise, focusing on both methodologies and the tools needed to support them. In our projects, we encountered needs for expertise and support on topics such as oral history, three dimensional modeling, and biblical software and tools. Again, none of these can likely find dedicated support at an institutional level, but looking across a group of peer or related institutions it is possible to imagine justifying resources dedicated to these methodologies and their tools.

In this model, there might also be third parties that could play a coordinating or clearinghouse role. For example, ICPSR already provides training sessions for data librarians, and the HathiTrust Research Center provides training and community focused on text mining. There may not be appropriate vehicles for all specific methodologies and tools, but it is worth thinking about whether there would be collaborative vehicles more appropriate than, or at least able to work in complement with, traditional library collaboration vehicles.

In the shared models, it would be vital to clearly specify how reporting relationships, work prioritization, and performance appraisal, would work. And, it is possible to imagine a final model in which one of the shared models is implemented in conjunction with the team-based approach discussed first.

Each of these conceptual models could be well-suited to the common research practices and associated information needs that cut across humanities fields, and several of them to the more granular functional and subdisciplinary needs that emerge. Could they also be well-suited to remaining selection and collections-related responsibilities? Could they even extend to a team-based model for supporting instruction, research skills development, and other learning support? These are questions we cannot answer with the evidence from our projects, but they are important to consider in evaluating these models.

Looking ahead

Many libraries are in various stages of rethinking their subject specialist or liaison programs. Whether the specific models we introduced above are right for any given institution or group of institutions, evidence about scholarly practices is a necessary input into planning the right liaison program for your institution. It should feature centrally in library planning for these programs.

In this paper, we have focused only on the humanities, where our evidence basis is richest. Over time, we will conduct similar types of analysis for other macrodisciplinary groups as well. But we make no assumption that each discipline or macrodisciplinary group has the same needs or should receive the same services. Indeed, the mindless pursuit of service equity has too often led to least common denominator service offerings rather than a more scholar-centric array of offerings such as the ones we have proposed for consideration.

Appendix 1

The three projects from which we are drawing findings were conducted using slightly different methodologies.

  • The project on history included interviews with 39 historians and graduate students, many by telephone, conducted by Ithaka S+R research staff with support from the National Endowment for the Humanities.
  • The project on art history included interviews with 76 academic art historians, curators, graduate students, and librarians in and for the field of art history, mostly through a series of site visits, conducted by Ithaka S+R research staff with support from the Getty and Kress foundations.
  • The project on religious studies included interviews with 102 religious studies scholars, conducted by research teams from each of 18 higher education institutions that were trained by Ithaka S+R staff and utilized a common semi-structured interview script. For further information on this project’s collaborative methodology, see Cooper and Schonfeld, 7-13.


  1. This paper is based in part on a talk that we gave at CNI in Spring 2017 ( Thanks to Jessica Clemons, Kimberly Lutz, Holly Mercer, and Barbara Rockenbach for reviewing a draft of the paper and providing extremely helpful feedback.
  2. Much has been written on changes in how materials are selected into general collections and the relative merits of varying approaches. For two examples, see  David W. Lewis, “From Stacks to the Web: The Transformation of Academic Library Collecting,” College & Research Libraries 74:2 (March 2013),, and Luke Swindler, “New Consortial Model for E-Books Acquisitions,” College & Research Libraries 77, no. 3 (2016),
  3. One consideration is that even some research support functions requiring subject expertise are seeing new competition. See for example, Roger C. Schonfeld, “Defining a New Content Type: The Exploratory Resource,” The Scholarly Kitchen, July 18, 2017,
  4. See for example, Julie Arendt and Megan Lotts, “What Liaisons Say about Themselves and What Faculty Say about Their Liaisons, a US Survey,” portal: Libraries and the Academy 12.2 (2012): 155-177, and Louise Cooke, et al, “Evaluating the Impact of Academic Liaison Librarians on Their User Community: A Review and Case Study,” New Review of Academic Librarianship 17.1 (2011): 5-30.
  5. One excellent resource is Janice Jaguszewski and Karen Williams, New Roles for New Times: Transforming Liaison Roles in Research Libraries (Association of Research Libraries, 2013),
  6. Anne R. Kenney, “Leveraging the Liaison Model: From Defining 21st Century Research Libraries to Implementing 21st Century Research Universities,” Ithaka S+R, last modified 25 March 2014,
  7. Barbara Rockenbach, Judy Ruttenberg, Kornelia Tancheva, and Rita Vine, “Association of Research Libraries/Columbia University/Cornell University/University of Toronto Pilot Library Liaison Institute,”
  8. Jaguszewski and Williams, 4.
  9. Jennifer Rutner and Roger C. Schonfeld, “Supporting the Changing Research Practices of Historians,” Ithaka S+R, December 10, 2012,;  Matthew P. Long and Roger C. Schonfeld, “Supporting the Changing Research Practices of Art Historians,” Ithaka S+R, April 30, 2014,; Danielle Cooper and Roger C. Schonfeld, “Supporting the Changing Research Practices of Religious Studies Scholars,” Ithaka S+R, February 8, 2017,
  10. See, for example, Ellen Collins, Monica E. Bulger, and Eric T. Meyer, “Discipline Matters: Technology Use in the Humanities,” Arts and Humanities in Higher Education 11.1-2 (2012): 76-92; Ellen Collins and Michael Jubb, “How Do Researchers in the Humanities Use Information Resources?” Liber Quarterly 21.2 (2012); Max Kemman, Martijn Kleppe, and Stef Scagliola, “Just Google It—Digital Research Practices of Humanities Scholars.” arXiv preprint arXiv: 1309.2434 (2013); Hieke Huistra and Bram Mellink, “Phrasing History: Selecting Sources in Digital Repositories,” Historical Methods: A Journal of Quantitative and Interdisciplinary History 49.4 (2016): 220-229; Kim Martin et al, “The Role of Agency in Historians’ Experiences of Serendipity in Physical and Digital Information Environments,” Journal of Documentation 72.6 (2016): 1008-1026.
  11. Top photo from page 10 of Princeton Theological Seminary’s report as part of  the Religious Studies project, see Virginia Dearborn, Jenifer Gundry, and Kate Skrebutenas, “The Research Practices and Support Needs of Advanced Scholars in Religion and Theology: A Local Report by Princeton Theological Seminary Library,” Princeton Theological Seminary, 2016,; bottom photo from page 6 of Luther Seminary’s report as part of the Religious Studies project, see Trisha Burr and Andrew Keck, “Faculty Research Practices at Luther Seminary.” Luther Seminary, 2016,
  12. A major exception is digital humanities research, which is often characterized by scholars as a highly collaborative endeavor involving team-based structures with a diverse set of partners. Only a minority of humanities scholars, however, describe digital humanities work as their primary research activity or output. Examples of the kinds of actors involved in collaborative digital humanities work include: graduate students, postdoctoral candidates, scholars in different fields and at different institutions, information professionals, and computer programmers, among others. For further information on the collaborative nature of digital humanities work, see, for example: Lynne Siemens, “‘It’s a Team if You Use “Reply All”’: An Exploration of Research Teams in Digital Humanities Environments.” Literary and Linguistic Computing (2009): 225-233.
  13. In our project on research support needs in history, we interviewed graduate students in addition to professors. These students also reported concerns about gaps in their training. See Rutner and Schonfeld, 23.
  14. One reviewer of this issue brief observed that humanists’ lack of systematic methodologies towards digital discovery may be related to earlier practices of browsing books in physical stacks and preferences to serendipity.
  15. The complexities of categorizing the religious studies scholar is further explored in Cooper and Schonfeld, 10-14.
  16. See for example Gwen Glazer, “Cornell, Columbia Libraries to Share Latin American, Iberian Studies Collections,” May 9, 2012,