10 practical tips to fight against the culture of non-citation in the humanities

I still remember well my first encounter with the complexities of digital citation practices. Coming from a traditional humanities background, it was during my master studies when I first came across a large-scale digital research tool, the Hungarian National Corpus [1]. It brought an instant revelation to my student life and changed the way I think about linguistics and our capabilities of exploring languages for good. To record this experience and say thank you for the helpfulness of the corpus creators, I had left a lengthy paragraph of acknowledgement in my masters thesis and sent a hard copy to them. Well, it did not fully serve its purpose as their immediate response suggested:  ‘This is all great but where is the proper citation?’ 

Only years later did I recognize that I was not the only person who got lost in the emerging (written or unwritten) norms of digital citation culture. In fact, several studies report a great deal of uncertainty that many humanities scholars feel when confronted with the issue of how to cite their online sources. This is especially true for content types that are not rooted in the print culture. In the study [2] that inspired the title of this post, Jonathan Blaney and Jusith Siefring even apostrophe the citation of digital resources as “the unloved son of the unloved son.” (paragraph 7) .

The honest reflection of research realities in citations is key

Today it’s clear to me that citation ethics is a real game changer in scholarly transparency that goes well beyond the basic matters around plagiarism. By giving credit where it’s due in standardized formats, we as authors can practice our power to grow scholarly transparency and increase resource creators’ symbolic capital and credit in academia at the same time. Even though we have the sense that the information systems measuring the (re)use and impact of digital tools and scholarly data are still in their infancy and institutional research evaluation criteria are only slowly adapting to the prevalence of scholarly objects that cannot be placed on a bookshelf, with small changes in our citation routines we can make a big contribution to accelerate them. 

What follows is a collection of useful tips that I learned since my aforementioned citation fiasco. They translate the principle of generous citation culture into practicalities to make all kinds of scholarly contributions equally visible and traceable. Although the practice of citing and to be cited are as inseparable as breathing in and breathing out, in this post, I separated these perspectives for practical reasons and collected good practices to both. In a second post, we will dive into specificities of the different content types (software, blog posts, images etc.) and see practicalities around their citation through a set of examples. 

As a (re)user… 

1. Think in terms of the research process not only in terms of publications  

A governing principle: make your knowledge provenance clear in your citations. This allows your readers to follow you on your provenance trail and recreate the same discovery experience that lead you to your conclusions. That said, cite and acknowledge everything that you worked with during your research: blog posts that gave you the basic idea, web resources, the corpora you consulted, the annotation schema you borrowed, the vocabulary along which you structured your data etc. If you used the digital edition of a work, or a preprint version of an article, cite that as well. This will inform your peers about the existence of these easily accessible resources (plus in this case you don’t need to worry about page numbers). You will find a rich set of examples in the next blog post and in the references section of this post. 

Pro tip: If you are concerned about citing something that may change, by submitting them to the Internet Archive, you can create digital citations that never break. Like this one [3], where you can find a brief tutorial provided by Internet Archive. 

2. Give credit to the intermediaries 

You may want to acknowledge the archive (digital or physical) or the aggregation platform where you found your sources/data to recognize their vital role in curation and discovery. The following solution shared by Andrew Newmann in his blog ‘Digital sources, analog citations’ [4] I find simple and easy: 

Figure 2. Screenshot from Thomas Ridout, Ten Years in Upper Canada in Peace and War, 1805-1815 (1890) p. 346. The Internet Archive.Figure 3. Screenshot from Philippo Alegambe, ed.  Mortes illustres et gesta eorum de Societate Jesu (1657). Google Books.

One can adopt this schema for Europeana or Wikimedia resources as well, for instance: 

Gertrude Käsebier, 1905. Happy Days. [Photograph.] Museum für Kunst und Gewerbe Hamburg, Sammlung Juhl. Found in Europeana Collection 2048429_Ag_DE_DDB_MKG. CC0, Source: https://www.europeana.eu/portal/en/record/2048429/item_36FQJ2YT6YSX4LYPEPBEBVEFT6P6G23G.html?q=%22Gertrude+K%C3%A4sebier%22#dcId=1583140820017&p=2 Wilhelm Busch, Lehrer Lämpel (aus Max und Moritz), marked as Public Domain (CCO 1.0), via Wikimedia Commons, Source: https://commons.wikimedia.org/wiki/File:L%C3%A4mpel.jpg  [5]

3. Move everything you can to the reference list from in-text mentions and acknowledgement sections

This is a piece of advice that is especially important to highlight in the era of automatic citation extraction systems and indexing services. By doing so, you can make sure that your acknowledgement enters the bibliometric information systems as a standard citation that can be used to demonstrate the impact of the content/tool/service that was instrumental in your work.  

 4. Credit the creators in the ways they want to be credited

It’s very easy if you find a ‘Cite as’ note associated with the resource you wish to cite. If not, and you are not sure about the creators’ citation preferences, it is well-worth to check with the creators/curators, even if it means extra work. It not only has to do with keeping a good academic karma. By crediting creators in ways that are meaningful and beneficial to them, you encourage them to keep sharing their work, you can help their work to be recognized by their institutions (yes, it’s about making them visible in those slowly-changing tenure and promotion criteria again) or you can support funding decisions about maintaining tools that are important for your work. When it comes to data, tools, or other, non bibliographic content types, some might prefer to cite an associated paper over the dataset. Although this is clearly not optimal and raises transparency concerns, these preferences are, temporarily, understandable as long as metrics and information systems are limited to the recognition of research articles only. 

5. Be as precise as you can be

Being specific about e.g. the exact version of software, tool or the pipeline you used is an obvious point for research transparency. The Grobid family is a nice case in point here: it is worth to clearly indicate which specific instance you are using.

• Grobid: https://github.com/kermitt2/grobid
• Grobid-Dictionaries: https://github.com/MedKhem/grobid-dictionaries
• Grobid-Cat: https://github.com/MedKhem/grobid-cat
• Grobid-NER: https://github.com/kermitt2/grobid-ner
• Grobid-quantities: https://github.com/kermitt2/grobid-quantities
• Grobid-astro: https://github.com/kermitt2/grobid-astro
• Grobid-SMECTA: https://github.com/Vi-dot/grobid-smecta

6. Include Persistent Identifiers (DOIs, handles, ARKs etc.) when possible  

PIDs are powerful citation instruments that fulfil many vital functions [6] in scholarly communication. For instance, they clearly identify the resource in question and disambiguate them from related ones or ones with similar names/titles (think about all those typos people make in reference sections!). But in the context of citations, impact and recognition, their real power lies in their ability to make the citation in question machine-readable and actionable. This opens up a range of possibilities to make the resource visible in indexing platforms and databases and track their (citation) life on the web. For instance, if you include a DOI to your citation, the Altmetrics information associated with it will show the creators how you interacted with their resources. Since there is usually a set of metadata associated with PIDs, when you have only limited space e.g. on slides and posters, you can use them as citations. 

Pro tip: Manually adding DOIs in your citations can be a laborious task. With the help of this nifty little tool called CrossRef Simple Text Query, you can simply copy-paste your reference list and (most) DOIs will be automatically added. (Thanks for the tip, @Olmo van den Akker (@DenOlmo).

7. Explore the full potentials of citation management systems such as Zenodo

Zenodo is flexible enough to accommodate a wide range of content types that you may want to cite later. By keeping record with your sources in a well-organized Zotero library, you do a big favour for your future self looking for that one missing reference (or the full name of that third author) right before the submission deadline. 

Pro tip: In this resource [7] published by the Harvard Library, you can find a handy tutorial on how to expand the use of Zotero even to primary resources such as archival records. 

As a creator….

1. Create a ‘Cite as’ snippet

By doing so, you don’t only make sure that people will credit you for your contribution the way you prefer but you also contribute to a better citation culture in general. Note that many repositories and discovery platforms (such as Zotero, HAL and NAKALA) already have ‘Cite as’ sections as in-built features. 

Pro tip: If you run a blog on Hypotheses, you can automatically generate ‘Cite as’ snippets by using the Cite plugin [8]. 

2. Make the citation of your resources easier (and easily trackable!) by assigning a PID to them

See point 6) above. The easiest way to get a PID to your resources is to share them in a repository. You can find easy to follow, step-by-step guidelines for sharing your publications here, sharing your data here and sharing your software here (more details are to come about each in the next post). When it comes to choosing a venue for your publications, prefer those publishers/platforms who provide PIDs. 

3. Link it or lose it 

Having Persistent Identifiers on your resources enables you to meaningfully link different types of related entities together (e.g. your author ID (such as ORCID) with your papers, the datasets you used with the resulting publications etc.). This way you can open a much broader window on your research and of course implicitly guide others to use and acknowledge you for all the work you did, not only for the canonical publication. 

Pro tip: Thanks to a recently established collaboration between CCSD (the technical infrastructure behind the French HAL repository) and Software Heritage, a software preservation service, you can now deposit your source code in the HAL repository. Once the deposit has been validated, and if you want it transferred, it gets imported to Software Heritage which generates a unique identifier to the code deposit allowing its traceability. As a second step, this identifier appears on your HAL record.  You can find more information here [9]. 

In the next post, we will focus on resource types that are important for humanities research and look at practical citation guides pertaining to them. In the meantime, tell us about your experience with citations and citation ethics here in the comments or as part of the twitter discussion! What is the resource type that is the least visible in terms of citations in your field? Is there a preference in your field to cite hard-copy over the digital and publications over novel types of scholarly objects? 

Many thanks to Alba Irollo, Mohamed Khemakhem, Frances Madden, Naomi Thruan and Ulrike Wuttke for their useful inputs for/comments on this post.

REFERENCES 

1. Csaba Oravecz, Tamás Váradi, Bálint Sass: The Hungarian Gigaword Corpus. In: Proceedings of LREC 2014, 2014. 
2. Blaney, J., & Siefring, J. (2016). A Culture of non-citation: Assessing the digital impact of British History Online and the Early English Books Online Text Creation Partnership. Digital Humanities Quarterly, 011(1).
3. Rossi, A. (2013). Fixing Broken Links on the Internet | Internet Archive Blogs. https://blog.archive.org/2013/10/25/fixing-broken-links/ [Blog post.]
4. Newmann, A. (2019, February 20). Digital Sources, Analog Citations. Omohundro Institute of Early American History and Culture https://blog.oieahc.wm.edu/digital-sources-analog-citations/ [Blog post.]
5. The latter example is from: Wuttke, U. (2019, June 16). Is Your Research Future Proof? Data Management Techniques & Tools for Digital Historians. https://doi.org/10.5281/zenodo.3247014 [Presentation slides.]
6. Madden, F. (2019, November 8). Why Use Persistent Identifiers? The PID Forum. https://www.pidforum.org/t/why-use-persistent-identifiers/714/4 [Blog post.]
7. Fauxsmith, J. (2019). Research Guides: Zotero for Archival Research: Use Zotero for Archival Research. https://guides.library.harvard.edu/c.php?g=531255&p=3634253 [Website.]
8.  Pacaud, F. (2019). Make it easy for readers to cite your posts using the Cite plugin. The House of Blogs. Retrieved 2 March 2020, from https://houseofblogs.hypotheses.org/152 [Blog post.]
9. Depositing scientific software into Software Heritage. (2018). Retrieved 2 March 2020, from https://www.softwareheritage.org/2018/09/28/depositing-scientific-software-into-software-heritage/ [Website.]

Cover image: Citation needed. Source: http://www.futureatlas.com, CC-BY 2.0. Found on: https://commons.wikimedia.org/wiki/File:%22Citation_needed%22.jpg.

---

---