Since the digitisation of primary sources, digital humanities (DH) has become an increasingly relevant topic for libraries as the (large) digital collections offer novel ways of conducting research for humanities scholars. This asks new questions, activities and services of libraries as providers of this data and their role as research partners in digital research. By expanding their activities beyond the traditional library activities (such as the offering of physical collections and support and training in information literacy), libraries are required to act more and more as publishers of their own digital collections and provide the expertise that is needed alongside the publication of that data. This not only changes the inner workings of the library as an organisation, but also the relationship with their research community.

Although much has been published on digital humanities and/in libraries, many publications are focused on libraries in the United States (e.g. Christian-Lamb et al., 2016; Hartsell-Gundy et al., 2015; Digital humanities in libraries, 2013; Padilla et al., 2019; White & Gilbert, 2016; Zhang et al., 2015). Little research has been done in Europe on activities, organisation and relationships between digital humanities researchers and the library. In the spring of 2019, the LIBER Working Group Digital Humanities and Digital Cultural Heritage1 therefore conducted a survey among members to gain knowledge on current practices of digital humanities in Europe’s research libraries and to close this gap in research on DH and libraries in Europe. This article describes the development of the LIBER survey and places the outcomes in the context of other surveys on digital humanities/scholarship2 in libraries, examining the new roles libraries take within digital research and therefore aims to answer the question “What role do European academic libraries take within the context of digital scholarship and how does that compare to libraries in other parts of the world?”

This paper first describes the development of the LIBER survey, then introduces four similar surveys done in other library communities and subsequently compares the output of those surveys within four themes.

In the spring of 2018, the Working Group conducted a mini survey of 20 questions amongst members to test questions, answers and question types. This was then used to produce case studies around digital humanities in libraries (Wilms, 2018). The Working Group organised a workshop at the 2018 LIBER conference collecting input from members on topics and specific questions. After the workshop and the publication of the mini survey report, all feedback was collated and discussed during a meeting of the Working Group’s core team together with several members of the digital scholarship team of the British Library in September 2018. The questions were then aligned to the survey and subsequent report by Rikk Mulligan (2016) on Digital Scholarship published by the Association of Research Libraries (ARL), in order to facilitate comparison of results. This resulted in the first draft of the survey covering the following themes:

The draft was tested by three members of the group and examined by two social scientists on survey ethics and practices. The final version consisted of 83 questions which were coded into the online survey tool SurveyMonkey. Responses were accepted from February 1 to March 15, 2019. All questions and collated answers of the survey are published in Wilms et al. (2020).

A total of 56 colleagues from 54 institutions in 20 countries completed the survey, mainly from Western Europe (71% of respondents). This represents 12% of all LIBER members and 50% of LIBER member countries (LIBER Network, n.d.). The respondents work in different types of libraries, ranging from small museum libraries to large national libraries, as can be seen in Figure 1. A more elaborate description of the respondents and all detailed findings can be found in Wilms et al. (2019).

Before presenting further results, I trace the history of similar surveys conducted in other parts of the world.

Over the past decade, multiple surveys have been conducted in library communities. In November 2011 the Association of Research Libraries (ARL) in the United States published a report titled ‘Digital Humanities’ (Bryson et al., 2011) Following this research, a survey was done amongst eight libraries in Hong Kong (Wong & Li, 2015). This 20 question-survey conducted in 2014 asked similar questions as the 2011 ARL survey related to – the then relatively new practice of – digital humanities in libraries.

Following the 2011 publication of ARL, in 2016 the Association funded new research into digital scholarship in libraries. Rikk Mulligan conducted a survey under 73 ARL members in February 2016 resulting in the report ‘Supporting Digital Scholarship’ (Mulligan, 2016). The survey consisted of 31 questions and also collected job descriptions, project websites, organisation charts, project planning documents and service descriptions. A response rate of 59% was achieved among ARL members. This report and survey outcomes are used in this article as it is the most up-to-date research conducted in the United States on digital scholarship in libraries.

In June 2017 Clare McKenzie and Kerry Ross conducted research among the member libraries of CAUL and CONZUL, the university library associations in Australia and New Zealand. They too were inspired by the 2011 report of the ARL, thus focusing on digital humanities. The survey consisted of 18 questions and was filled out by 17 libraries; a response rate of 36% of all CAUL and CONZUL associated libraries. The report and data were published in February 2018 (McKenzie & Ross, 2018).

The Irish Consortium of National and University Libraries (CONUL) also sought to investigate the activities their libraries undertook within digital scholarship. Modelling their survey on the 2016 ARL-report of Mulligan, they focused on digital scholarship, using the activities Mulligan used in his research as a basis. In February 2018 the 15-question survey was filled out by 85% of CONUL members, which are 11 libraries. The report by Joy et al. (2019) was published in February 2019, along with the survey data.

The association of Research Libraries UK (RLUK) was again inspired by Mulligan’s report and following the establishment of their Digital Scholarship Network conducted a survey under their members between January and April 2019. Again, this survey was closely based on the 2016 ARL survey, working from the same activities and adding RLUK-specific questions. The 42-question survey consisted of multiple sections, which could be filled out separately, but in total they received 38 responses from 30 member libraries, representing 81% of all members. The report compares the outcomes of their survey to those of CONUL and ARL (Greenhall, 2019).

Four of the Irish libraries that took part in the CONUL survey have also taken the LIBER survey, as have six of the RLUK libraries. Where needed, these libraries have been excluded from the LIBER dataset for the comparison in the remainder of this article in order to avoid any overlap.

The various surveys described above and discussed in this article were conducted at different times (ranging from 2015 to 2019) and focus on different regions, ranging from a community of eight libraries to a national network of 73 libraries. These differences should be taken into account when reading the comparisons in this article. Each survey posed questions and answer options slightly differently. Nonetheless it is possible to identify four themes which were covered in all surveys. These are: Activities, Funding and organisation, Staff and skills, and Future. Each theme is discussed below.

Digital scholarship encompasses a wide range of activities, which makes it difficult to create a complete list of all possibilities within libraries. The LIBER survey uses the TaDiRAH research taxonomy of digital research (Borek et al., 2016). This taxonomy lists eight activity nodes: ‘Capture,’ ‘Creation,’ ‘Enrichment,’ ‘Analysis,’ ‘Interpretation,’ ‘Storage,’ ‘Dissemination’ and ‘Meta-activities’. These are subdivided into more specific activities, such as imaging, visualisation and project management. Each surveyed library was asked whether they do a sub-activity as a regular activity, an ad-hoc activity or not at all. For the purposes of comparison in this article the LIBER results4 have been compressed into the eight main activity nodes with all absolute answers averaged per activity node and displayed in a 100% stacked bar, as can be seen in Figure 2.

What becomes clear is that the more traditional activities of libraries appear most often as a regular or ad-hoc activity. More than 60% of the surveyed libraries mention ‘Capture’ (27), ‘Enrichment’ (32), ‘Storage’ (38) and ‘Meta-activities’ (28) as a regular or ad-hoc activity. ‘Capture’ includes activities such as the digitisation of text and images; ‘Enrichment’ includes producing metadata and data cleaning; and ‘Storage’ includes metadata creation and preservation. We would expect to see these activities appear regularly in the questionnaire responses as this is what libraries are most familiar with. Activities related to the analysis and interpretation of data, such as modelling, visualisations or network analysis — which are core elements of digital humanities — are mentioned less often as being part of regular or ad-hoc library activities in LIBER libraries.

The surveys from the UK, Ireland and the US have not opted for the same taxonomy, but chose to use a list of activities, created by the author of the ARL survey, Rikk Mulligan. This list consists of 19 activities and an ‘Other’ option. Where the RLUK use the whole list, the CONUL survey chose to bar the activity ‘Technical upkeep’. The CAUL/CONZUL survey defines activities by referencing specific tools, hardware and platforms, while the Hong Kong survey summarises activities within seven themes and an ‘Other’ option.

To facilitate the comparison of the surveys, all activities and themes are categorised using the top nodes of TaDiRAH with the mapping that can be seen in Table 1.

The ARL survey asked if an activity was done in the library, outside of the library but in the same institution, or outside of the institution. CONUL added the option ‘Not currently supported’ to this list and RLUK again added ‘Unknown’. All three surveys also had the option ‘Other’. It has to be noted here that the options about activities existing outside of the library, but in the same institution are most likely not applicable to the national libraries, that usually operate as independent organisations.

For the ARL, RLUK and CONUL surveys, the absolute answers to the types and locations of activities are categorised using the table above and then averages are calculated per category. This is plotted into a 100% stacked bar graph per category as can be seen in Figure 3.

When looking at activities that happen in the library (the light blue bars in Figure 3), RLUK and CONUL show similar results to those of the LIBER survey. The more traditional library activities included in ‘Capture,’ ‘Enrichment’ and ‘Storage’ occur most often. The ARL survey shows a different view of library activities around digital collections, as they also have a large number of respondents describing activities around ‘Creation’ – which in this case means software creation and/or interface design. This corresponds to the finding that the ARL libraries are also quite active in the analysis and interpretation of data, and meta-activities such as project organisation.

The CAUL/CONZUL survey is rather difficult to compare to the other surveys as it revolves not around activities but asked input on software and hardware. However, when using the mapping in Table 1, a similar image to that of the European libraries appears. The more traditional categories mentioned earlier of ‘Storage’ and ‘Capture’ are mentioned more often than the more DH-specific roles around analysis. Over half of the surveyed libraries offer an institutional repository (8), server space (5) and access to scanning (8) and 3D modelling equipment (2). A remarkable effort is also seen in the ‘Dissemination’ category since bibliographic management tools are supported by 13 libraries and web publishing tools by 5 libraries.

The Hong Kong-survey does not map conclusively to all categories used within TaDiRAH. However, it is interesting to note that almost all libraries (86%) offer digital collections and almost half (46%) provide or build tools for the use of those collections, again showing the categories ‘Capture’ and ‘Creation’ to be more natural to the library than ‘Analysis’ (14%).

In conclusion, the various types of activities listed by the surveyed libraries show that US libraries play a different role in digital scholarship at the time of the surveys than the other surveyed libraries, who seem to be more active in research support, while the ARL libraries appear to be engaging more in the research itself.

The surveyed libraries range from having fully implemented digital humanities centres to those that had “not developed an awareness of the trend of digital humanities” (Wong & Li, 2015). It is therefore to be expected that the funding of the various libraries differs a lot. However, when combining the figures around funding from the different surveys5 many similarities are seen between the surveys. As with the previous questions, not all surveys had the same answer options, but similar categories can be seen. Figure 4 depicts the percentage of libraries that use the various types of funding mentioned in the surveys, collated into categories.

Staff and skill-building is a common theme throughout the surveys, as what is possible in terms of activities within the organisation relies heavily on the people working in the libraries and the position they have. When looking at the roles and skills for DH/DS in the library two main categories are found in the surveys. First, questions around roles and position within the library, and second, questions around skills and skill-building. These will therefore be discussed separately.

LIBER asked libraries to indicate how much of a priority digital humanities currently is in their library on a scale from 0 to 100 and what they thought the priority would be in 2022 (Figure 6).

Figure 6 shows that most libraries do not see a major rise in prioritisation in the next four years after the survey with an average increase of 5,45 points. Just eight libraries see a potential rise of 20 or more points and 10 libraries indicate the priority to decrease (slightly), generally with around 10 points. However, most libraries who indicate the priority to be high (80 or more) see that being similar in four years.

The survey conducted by LIBER’s Working Group paints a clear picture of the status of the use of digital collections in digital humanities research. Libraries6 are currently most comfortable with activities close to their traditions, such as providing (digital) storage, digitisation and collection building. They are, however, interested in expanding those into the digital realm. Developing the technical skills of staff is needed for this, but in many libraries the plans and budgets seem to be in place to do this. Their role currently revolves around research support, but the intention is there to develop that into research engagement.

The survey conducted within the LIBER community does have its limitations. As LIBER membership is not evenly distributed across Europe it does not show a balanced view of Europe’s research libraries. This is again amplified by the fact that respondents from already overrepresented member countries participated a lot in the survey. In future work this might be overcome by using different data collection techniques to ensure libraries that are not yet represented well are included in the results, possibly also by engaging with libraries that are not members of LIBER. Also, a next iteration of the survey should be shortened. The survey (83 questions) was very long, and many respondents dropped out before completion7.

The research done in the various other surveys across Europe, Hong Kong, the US, Australia and New Zealand also shows the digital humanities/scholarship library landscape as one of growth. Whereas the US-based libraries are already engaging in the whole research lifecycle, they still see more opportunities to grow their role and responsibilities. The other surveyed libraries are also increasingly getting involved in digital research beyond the offering of digital collections.

All organisations see their limitations and work on improving them. They position their staff with crucial mandates in management teams and incorporate digital scholarship into the workings of the library by using the library’s general budget as a funding resource. Although differences exist between activities in the various parts of the world, libraries feel comfortable to choose what works best for them and their research community, but also experiment with new techniques and their implementation in their organisations.

The future is bright for digital scholarship in libraries as many of the themes discussed for the future closely relate to the work libraries have been doing in the past decades. They open collections, facilitate research, never stop improving themselves and work to share research as widely and openly as possible.