{"id":262,"date":"2019-04-22T21:39:34","date_gmt":"2019-04-22T19:39:34","guid":{"rendered":"https:\/\/dhmethods.hypotheses.org\/?p=262"},"modified":"2019-04-25T16:02:47","modified_gmt":"2019-04-25T14:02:47","slug":"here-be-dragons-open-access-to-research-data-in-the-humanities","status":"publish","type":"post","link":"https:\/\/dhmethods.hypotheses.org\/262","title":{"rendered":"\u201cHere be dragons\u201d: Open Access to Research Data in the Humanities"},"content":{"rendered":"\n

Guest post by our blog competition<\/a> winner, Ulrike Wuttke.<\/p>\n

Introduction<\/h4>\n

#1 It has been estimated by Stefan Winkler Nees (from the Deutsche Forschungsgemeinschaft-DFG) in 2011 that 90% of all digital research data is lost.[1]<\/a> We don\u2019t know how many of this data belonged to the Humanities and hopefully, these numbers are better today, but we can assume that still a lot of Humanities data (and other data) is lost, because of missing infrastructures, or because no one has taken care of the long-term availability of this data in time.[2]<\/a><\/p>\n

\"\"<\/p>\n

\n
Are Your Data FAIR?<\/figcaption>\n<\/figure>\n

But, even if data is not lost: Does the available data sparkle joy, to borrow a term from the ubiquitous Marie Kondo? Are these datasets accessible for research, well documented using standards, available in interoperable formats etc., in short: is it FAIR[3]<\/a> data, too?<\/p>\n

<\/p>\n

#2 The digital transformation of research has dramatically increased the creation, gathering, and use of research data in all disciplines. It also led to the development of the new paradigm of Open Science which not only promotes that published results should be open access, but also that the underlying data should be open (for example in the H2020 Open Research Data Pilot).[4]<\/a> Together, these developments resulted in a heightened demand for the publication of research data, promoted by national and international funding agencies[5]<\/a>, and research institutions, often directly reflected in (funding) guidelines, which can be more or less binding.[6] Among the many reasons to underpin the demand for Open Research Data, efficiency (reuse), reproducibility (transparency), impact, and public trust are prominent.[7]<\/a> Slowly, the responsible handling of research data and FAIR publication of research data is becoming an integral part of Good scientific practice.[8]<\/a><\/p>\n

The vision goes beyond that: we are talking here about the grand vision of the European Open Science Cloud (EOSC)[9]<\/a> and an even bigger vision for a global Open Science Ecosystem. To realise this vision and for humanities data playing a considerable role in it, our \u201cefforts should be guided by the twin aims of ensuring that data meets the FAIR principles, and that it is effectively preserved in trusted, certified repositories\u201d[10]<\/a>. The FAIR principles have gained since their publication in 2016[11]<\/a> great acceptance as guiding principle, because they take not only into account that not all data can be open[12]<\/a>, leading, for example, the European Research Council to formulate the principle \u201cas open as possible, as closed as necessary\u201d[13]<\/a>, but because they also put great emphasis on \u201cthe attributes data need to have to enable and enhance reuse, by humans and machines\u201d, in a nutshell metadata.[14]<\/a><\/p>\n

\"\"<\/p>\n

\n
Vision for Humanities Data<\/figcaption>\n<\/figure>\n

In the following, I will first line out key challenges related to Open Access to Research Data in the humanities and then discuss some perspectives to improve the current situation. I will focus on educational aspects, and tools for research data management as keystones for Open & FAIR Research Data. As the German situation is best known to me, examples are mainly drawn from a German context.<\/p>\n

\u201cHere be dragons\u201d<\/strong><\/h2>\n

#4 The phrase \u201cHic sunt dracones\u201d (transl. \u201cHere be Dragons\u201d), is used on some old maps of the world to describe an area that was unknown to the cartographer. I found it quite appropriate to summarize the ambivalence of humanists towards data and all these \u201cfancy\u201d concepts discussed by \u201cinfrastructure people\u201d like FAIR Data, the EOSC, or Research Data Management. First of all, it must be said that humanities researchers tend to be ambivalent about the concept of \u2018data\u2019[15]<\/a> and that \u201c[t]here are issues surrounding [\u2026] the acceptance of the \u2018research data concept\u2019\u201d[16]<\/a>. In short, they just don\u2019t use the word \u201cdata\u201d, but talk about \u201csources\u201d, \u201cresearch materials\u201d etc., which leads to the fact that the whole \u201cdata talk\u201d doesn\u2019t appeal to them. Additionally, an expeditionary survey conducted by PARTHENOS in 2017 among researchers in the domain of digital humanities, language studies, and cultural heritage showed that the FAIR Principles and the EOSC, concepts and recommendations, thriving among \u201cinfrastructure folks\u201d, are relatively little known in the research communities themselves.[17]<\/a> Often, the publication of research data only comes as an afterthought (if at all).[18]<\/a> However, at the end of a project, it is often too late to publish the data in a meaningful way because of the lack of documentation and the lack of resources to prepare the data properly for publishing.<\/p>\n

#5 Let me discuss some reasons that currently seem to form barriers for establishing a culture of open sharing in the humanities. In general, \u201cissues surrounding incentivisation\u201d[19]<\/a> can be observed. Given the strong competition and the traditional humanities reputation system based on traditional \u2018long story scientific publication formats\u2019 (monographs, book chapters, or articles as significant scientific publications) as opposed to \u2018data\u2019, there is low motivation to publish research data.[20]<\/a> This lack of incentives goes hand in hand with researchers\u2019 (perceived) fear of being scooped or that someone else will be the first to publish something based on their data, e.g. an exciting manuscript or object they have found.[21]<\/a> A research tradition that has been based on (and rewarded) secretiveness is not easy to change with nothing fundamental opposite as a reward. Other prejudices often brought up are: nobody will understand the data, nobody will need the data, someone will sell the data and last but not least, (perceived) lack of technical skills.[22]<\/a><\/p>\n

\n

\u201cThe inherent controversy in the meaning of \u201cdata\u201d and the importance of personal interpretation on data for humanities researchers is not conducive to sharing.\u201d[23]<\/a><\/p>\n<\/blockquote>\n

#6 Even those who are willing to publish data, as it must also be duly acknowledged that there is a long and thriving tradition of humanities corpora collection and publication continued in the digital, e.g. at the academies, are facing obstacles. Especially legal issues or doubts about possible legal issues are often mentioned in this context: The legal regulations concerning research data are complicated and not internationally aligned and there are many actors involved in the production of humanities research data, not only humanists. This leads to the fact that humanities research is often based on data under copyright restriction (from cultural heritage institutions or other actors) which makes it difficult to publish them as ORD.[24]<\/a><\/p>\n

#7 Additionally, we are dealing with is issues around the availability and sustainability of specialist support structures for humanities research data support as well as the lack of practical guidance.[25]<\/a> Humanities data centers and other data services for the humanities are often dependent on third-party funding (project based).[26]<\/a> This leads to issues of trust on the side of the researchers which may result in these services not being in high demand[27]<\/a> (and may even result in an unwillingness to \u201cGo Digital\u201d at all), it also leads to the problem of sustaining \u201cliving systems\u201d[28]<\/a> (which need to be frequently updated, migrated, and curated). However, current efforts, especially from the Digital Humanities community, also have led to positive developments:<\/p>\n

\n

\u201cFor example, the emergence of linked open data over the past decade has been supported both by the establishment of effective standards for modeling and disseminating such data, and the growth of practices and social expectations supporting its creation. These developments have meant that expertly modeled data from specific domains can be accessed and combined flexibly, rather than remaining in isolation or striving for self-sufficiency.\u201d[29]<\/a><\/p>\n<\/blockquote>\n

\"\"<\/p>\n

\n
Fellowship of the Data<\/figcaption>\n<\/figure>\n

#8 To sum up my observations so far: Humanities research data, in general, is rather heterogeneous, idiosyncratic, and complex[30]<\/a> and humanists are ambivalent about the term \u201cdata\u201d. Digital practices are already part of the research activities of many humanists, especially in the Digital Humanities, but they are not equally fully developed. This leads to the fact that the potential of digital research data and methods is not fully exploited, because the digital research process is not carefully planned, with other words many research data already exist in digital form, but they are not findable, quality controlled, and reusable.[31]<\/a> All in all, the land of FAIR Research Data is still unknown territory for many humanists, or at least scary as if dragons would indeed live there. In the next part, therefore, I will argue for increased efforts for awareness raising and skills building and a \u201cfellowship of the data\u201d, a support system to facilitate the quest for FAIR data in the humanities.<\/p>\n

Perpectives<\/h2>\n

#9 Naturally, I cannot offer immediate solutions, but I would like to point out some paths that need to be pursued with increased intensity in the near future to facilitate FAIR data in the humanities (and beyond). These paths can focus on different stakeholders in the FAIR ecosystem such as research institutions, funding bodies, or publishers, or individual researchers and research communities. In my opinion, special attention has to be paid to the researchers and research communities themselves, so that recommendations, policies, services, etc. are aligned and known to disciplinary practices and cultures.[32]<\/a> If the researchers are not on our side in this quest, we are prone to lose the battle or at least experience a delay in realising the goals.<\/p>\n

#10 The most urgent points in my opinion are the following. We need to work on incentives for the FAIR publication of research data, e.g. wider adoption of DORA (Declaration on Research Assessment).[33]<\/a> We also need to Invest in the development of beneficial environments for aggregation (think EOSC, German NFDI):<\/p>\n

\n

\u201cThe interdisciplinary bundling of humanities data repositories and the development of adequate research tools and services for linked data represents a great opportunity for humanities research.\u201d[34]<\/a><\/em><\/p>\n<\/blockquote>\n

Another highly important task is educating the next (and this) generation of (digital) humanities researchers[35]<\/a> to deal with the datafication[36]<\/a> of research and education practices, but also \u201cinfrastructure people\u201d in discipline-specific contexts. Humanists need to be aware of limits of publication tools and how they expose (or not) the underlying data model. We need to strive to offer more idiosyncratic options and take time for critical consideration of how early choices have an impact on how data is published and can be used (curatorial perspective). It is our task to prevent research data from becoming trapped (in specific formats or hardware). In an ideal world we \u201cdesign our data without tool dependencies\u201d = \u201ctool-agnostic approach\u201d vs. \u201ctool-dependent approach\u201d[37]<\/a>. The progressing digital transformation of humanities research along with the increasing importance of digital research infrastructures calls not only for a certain level of \u201cdata literacy\u201d[38]<\/a> but even for an expansion of this concept to a certain level of \u201cdata infrastructure literacy\u201d, a term recently coined by Gray et al. 2018[39]<\/a>.<\/p>\n

Much discussed in the context I have just outlined is Research Data Management:<\/p>\n

\n

\u201cResearch Data Management describes the process to curate (or manage) research data along the research data lifecycle and includes various activities such as planning, producing, selection, analysis, archiving, and preparation for reuse. Because data are very heterogeneous, discipline and data specific solutions can be required.\u201d[40]<\/a><\/em><\/p>\n<\/blockquote>\n

Given this, admittedly not very appealing sounding definition, it may not come as a surprise that researchers in general still consider research data management [41]<\/a>as an extra, tedious, time-consuming task that diverts them from \u201creal research\u201d and humanists especially consider it as almost opposed to the hermeneutic humanistic research practice[42]<\/a>. However, acceptance by the researchers is the key success factor for establishing standards in Research Data Management.[43]<\/a> Therefore, we need to show humanists the added value of Research Data Management for the planning process of digital projects. Lemaire (2018) has recently convincingly argued that RDM is a process already inherent in the research process itself (although at the moment rather implicit) and that it can be an instrument for reviewing the research concept.[44]<\/a><\/p>\n

#12 The digital turn of humanities research and the requirement for FAIR research data, that is sustainable and quality controlled handling of research data, requires thorough planning of the digital research process before<\/em> the start, a process that is guided and documented in a Research Data Management Plan.[45]<\/a> Given all the advantages of Research Data Management planning, efforts should be increased regarding awareness raising and skills building[46]<\/a> already as part of university curricula[47]<\/a>[, and tools for consequent data management planning (and handling)[48]<\/a> with the end in mind, that is (if possible) the publication of FAIR research data.<\/p>\n

RDMO[49]<\/a><\/h2>\n

#13 Given that the management of research data is increasingly regarded as a process of active support and care during the whole research process (and not only of producing a mere static document), tools are needed that support active Research Data Management Planning[50]<\/a>, e.g. by providing different and up-to-date status information to different participants of the research process. I am currently part of a project that is developing such a tool, the Research Data Management Organiser (RDMO).[51]<\/a><\/p>\n

#14 In a nutshell, with RDMO the research data management process can be organised as a collaborative effort encompassing its different stakeholders, besides researchers, especially infrastructure partners such as libraries or computing centres. One of the use scenarios of RDMO is library staff using RDMO\u2019s question catalogue to work out the data management strategy for projects with researchers and other partners\/experts.[52]<\/a> RDMO can be adapted to the requirements of communities or organisations (e.g. institutional or discipline-specific guidelines) and has multilingual capabilities. At the moment we are working with a range of very different institutions and communities in Germany to further improve this tool.<\/p>\n

Conclusion<\/h2>\n

#15 On the one hand, researchers need to be aware of the issues at hand and take their responsibility (take the issues of digital research practices seriously!).[53]<\/a> On the other hand, we need to work on institutional availability and sustainability of research data (management) and support (e.g. via data centres and local experts)[54]<\/a> and clever and efficient connection between initiatives on different levels. We need to provide adequate RDM tools that support researchers to prepare their data for publication from the very beginning. Libraries already have an important role in this ecosystem and I dare to say that they are capable and designated to take a leading role in this field in the future if they invest in it: in heads and infrastructure(!).[55]<\/a> Especially I would like to refer here to their role in the creation of vast digital corpora of open research data, which cannot be left as a task to researchers alone.[56]<\/a> Last but not least, we also need to make research data management less scary: it\u2019s not a scientific revolution and doesn\u2019t mean that all skills learned so far in a typical humanities curriculum are to be thrown overboard, quite the opposite is true.[57]<\/a><\/p>\n

#16 There is a lot at stake for the humanities, maybe the very question what we want the future of the humanities to be. When it comes to Open and FAIR research data in the humanities, I can only say it with Queen: \u201cI want it all, and I want it now!\u201d<\/p>\n

\"\"<\/figure>\n
\n

\u201cI want it all, I want all, I want it all, and I want it now.\u201d Queen (1989).<\/p>\n<\/blockquote>\n

To create this broad culture of FAIR data sharing in the humanities we have to roll up our sleeves, team up, and distribute hats:<\/p>\n

    \n
  1. Embrace Open principles,<\/li>\n
  2. bridge the gap between the digital and the humanities and look what we can learn from the Digital Humanities and other more data-savvy disciplines.[58]<\/a><\/li>\n<\/ol>\n

    What are your thoughts and suggestions on this topic? Do you agree? Do you have additional hints for me that lead to more discipline specific information and insight about the handling of research data, data sharing, and the FAIR principles in the humanities? Please, leave your thought below or contact me via Twitter or e-mail. Looking forward to discuss this topic further with you!<\/p>\n

    You can find the original post<\/a> on Ulrike’s blog at: https:\/\/ulrikewuttke.wordpress.com\/<\/a><\/p>\n

    Notes<\/h2>\n

    [1] See Stefan Winkler-Nees, Vorwort, In: B\u00fcttner, Stephan; Hobohm, Hans-Christoph; M\u00fcller, Lars (ed.): Handbuch Forschungsdatenmanagement, Bad Honnef 2011, p. 5.<\/p>\n

    [2] See Geisteswissenschaftliche Datenzentren im deutschsprachigen Raum, Grundsatzpapier zur Sicherung der langfristigen Verf\u00fcgbarkeit von Forschungsdaten, Version 1.0, DHd AG Datenzentren, 2018, p. 9. Accessed from http:\/\/doi.org\/<\/a>10.5281<\/a>\/zenodo.1134760<\/a>, 25.03.2019.<\/p>\n

    [3] See Wilkinson, M. D., et al. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data<\/em>, 3<\/em>, 160018. Accessed from https:\/\/doi.org\/10.1038\/sdata.2016.18<\/a>, 25.03.2019. Webseite FORCE11: https:\/\/www.force11.org\/group\/fairgroup\/fairprinciples<\/a>.<\/span><\/p>\n

    [4] See European Research Council (ERC), Guidelines on Implementation of Open Access to Scientific Publications and Research Data in projects supported by the European Council under Horizon 2020, Version 1.1., 21. April 2017. Accessed from http:\/\/ec.europa.eu\/research\/participants\/data\/ref\/h2020\/other\/hi\/oa-pilot\/h2020-hi-erc-oa-guide_en.pdf<\/a>, 25.03.2019. See also from the UK document Concordat on Open Research Data, 2016. Accessed from https:\/\/www.ukri.org\/files\/legacy\/documents\/concordatonopenresearchdata-pdf\/<\/a>, 25.03.2019 or the recommendation from the Steuerungsgremium der Schwerpunktinitiative \u201eDigitale Information\u201c der Allianz der deutschen Wissenschaftsorganisationen (2017), Den digitalen Wandel in der Wissenschaft gestalten: Die Schwerpunktinitiative \u201eDigitale Information\u201c der Allianz der deutschen Wissenschaftsorganisationen, Leitbild 2018 \u2013 2022, 2017, Accessed from: http:\/\/doi.org\/10.2312\/allianzoa.015<\/a>.<\/p>\n

    [5] See for example EC Directorate-General for Research & Innovation, H2020 Programme: Guidelines on FAIR Data Management in Horizon 2020, Version 3.0, 26. July 2016, Accessed from: http:\/\/ec.europa.eu\/research\/participants\/data\/ref\/h2020\/grants_manual\/hi\/oa_pilot\/h2020-hi-oa-data-mgt_en.pdf<\/a>, 25.03.2019, Deutsche Forschungsgemeinschaft \/DFG), Leitlinien zum Umgang mit Forschungsdaten, 30.09.2015. Accessed from: http:\/\/www.dfg.de\/download\/pdf\/foerderung\/antragstellung\/forschungsdaten\/richtlinien_forschungsdaten.pdf<\/a>, 25.03.2019.<\/p>\n

    [6] While German funders already give recommendations, other funders already issued more binding guidelines that include the demand for a DMP, such as the Schweizerische Nationalfonds (SNF). See Schweizerische Nationalfonds, Open Research Data, (o.J.). Accessed from: http:\/\/www.snf.ch\/en\/theSNSF\/research-policies\/open_research_data\/Pages\/default.aspx#FAIR%20Data%20Principles%20for%20Research%20Data%20Management<\/a>, 25.03.2019.<\/p>\n

    [7] \u201cOpen research data (ORD) have the potential not only to deliver greater efficiencies in research, but to improve its rigour and reproducibility, to enhance its impact, and to increase public trust in its results.\u201d Open Research Data Task Force, Realising the potential: Final Report of the Open Research Data Task Force. N.P., 2018, p. 3. Accessed from https:\/\/assets.publishing.service.gov.uk\/government\/uploads\/system\/uploads\/attachment_data\/file\/775006\/Realising-the-potential-ORDTF-July-2018.pdf<\/a>, 22.03.2019.<\/p>\n

    [8] See recommendation 7 \u201cSicherung und Aufbewahrung von Prim\u00e4rdaten\u201d of the DFG-Denkschrift zur Sicherung der guten wissenschaftlichen Praxis, in: Deutsche Forschungsgemeinschaft (DFG), Sicherung guter wissenschaftlicher Praxis: Empfehlungen der Kommission \u201eSelbstkontrolle in der Wissenschaft\u201c, Weinheim 2013, p. 21-22. Accessed from: http:\/\/doi.org\/10.1002\/9783527679188.oth1<\/a>, 25.03.2019.<\/p>\n

    [9] https:\/\/ec.europa.eu\/research\/openscience\/index.cfm?pg=open-science-cloud<\/a>.<\/p>\n

    [10] Open Research Data Task Force, Realising the potential: Final Report of the Open Research Data Task Force. N.P., 2018, p. 8. Accessed from https:\/\/assets.publishing.service.gov.uk\/government\/uploads\/system\/uploads\/attachment_data\/file\/775006\/Realising-the-potential-ORDTF-July-2018.pdf<\/a>, 25.03.2019.<\/p>\n

    [11] The term FAIR was launched in 2014. See https:\/\/www.force11.org\/group\/fairgroup\/fairprinciples<\/a>.<\/p>\n

    [12] Principles for Open Data in Science have been formulated in the Panton Principles, demanding that data should be placed in public domain. See Panton Principles, Principles for open data in science. Murray-Rust, Peter; Neylon, Cameron; Pollock, Rufus; Wilbanks, John; (19 Feb 2010). Retrieved 31.03.2019 from https:\/\/pantonprinciples.org\/<\/a>. This demand has been criticized as difficult to realize because of two main reasons 1) the legal system of some countries, including Germany, does not really allow complete renunciation of rights by the right holder (i.e. public domain) 2) it removes all obligations to quote, which remove an important incentive, see Ulrich Herb, Open Science in der Soziologie. Eine interdisziplin\u00e4re Bestandsaufnahme zur offenen Wissenschaft und eine Untersuchung ihrer Verbreitung in der Soziologie, Gl\u00fcckstadt 2015, p. 126, accessed from: https:\/\/doi.org\/10.5281\/zenodo.31234<\/a>, 25.03.2019.<\/p>\n

    [13] European Research Council (ERC), Guidelines on Implementation of Open Access to Scientific Publications and Research Data in projects supported by the European Council under Horizon 2020, Version 1.1., 21. April 2017, p. 6. Accessed from http:\/\/ec.europa.eu\/research\/participants\/data\/ref\/h2020\/other\/hi\/oa-pilot\/h2020-hi-erc-oa-guide_en.pdf<\/a>, 25.03.2019. Not all data needs to be published. A selection of what data is relevant and interesting for scientific reuse, like archives have always done, is necessary. For guidelines see for example \u00a0Angus Whyte & Andrew Wilson, How to appraise and select research data for curation, Digital Curation Centre How-to Guides, Edinburgh 2010. Accessed from: http:\/\/www.dcc.ac.uk\/resources\/how-guides\/appraise-select-data<\/a>, 25.03.2019.<\/p>\n

    [14] European Commission Expert Group on FAIR Data, Turning FAIR into reality: Final Report and Action Plan from the European Commission Expert Group on FAIR Data, Brussels 2018, p. 18, accessed from: https:\/\/ec.europa.eu\/info\/publications\/turning-fair-reality_en<\/a>, 25.03.2019. The FAIR principles provide guidance on \u201chow to facilitate knowledge discovery by assisting humans and machines in their discovery of, access to, integration and analysis of, task-appropriate scientific data and their associated algorithms and workflows\u201d Quote from: https:\/\/www.force11.org\/group\/fairgroup\/fairprinciples<\/a>, accessed: 25.03.2019.<\/p>\n