Sjoerd Levelt is assistant professor in the Program for Cultures, Civilizations and Ideas of Bilkent University, Ankara. He studied Dutch and English Medieval Studies in Amsterdam, Berkeley and Oxford, received his PhD in Combined Historical Studies at the Warburg Institute (2010), and in 2012 was elected to the Fellowship of the Royal Historical Society. He won the Society for Renaissance Studies Book Prize 2012 for his book Jan van Naaldwijk’s Chronicles of Holland: Continuity and Transformation in the Historical Tradition of Holland during the Early Sixteenth Century, a study of the late medieval historical traditions of Holland and their continuations in the early modern period. He previously taught at the Universities of Exeter and Sussex, and was a research fellow on The Poly-Olbion Project, where he studied John Selden’s prose commentary to Michael Drayton’s seventeenth-century chorographical poem about Britain, Poly-Olbion, in relation to Selden’s annotations in his own books, now in the Bodleian Library, Oxford. Levelt’s areas of interest include the medieval and early modern historiographical traditions of the Netherlands and Britain, book history, and manuscript culture in the first centuries of printing. He tweets at @slevelt. In this blog, Sjoerd talks about best practices surrounding the publication of images of archival sources on social media.
As a medievalist and earlymodernist on twitter, I regularly tweet images from manuscripts and early printed books, which I either photographed myself during library research, or which I found browsing digital collections. This puts me in the company of many historical twitter accounts tweeting images of historical books, from good ones such as repositories like @BLMedieval, projects like @SexyCodicology, and individuals like @PiersatPenn, and bad ones which I won’t link to here, for reasons which I shall explain below. The images I tweet are usually classified as either public domain, or shared under Creative Commons licenses, allowing reuse. The libraries whose books I tweet images from usually request, rather than demand, attribution; this means that considerations about how exactly to share these images are usually not primarily of a legal, but of an ethical nature.
“We are the notes who say ‘ni!'”https://t.co/N1D6WuNIsO pic.twitter.com/Luc7xEpaVx
— Sjoerd Levelt (@SLevelt) February 11, 2017
Besides the picture and my caption, there are types of information I include in my tweets of medieval manuscript images; in particular, there are repository references, and there is descriptive information. Here are some thoughts on which information I include in my picture tweets, and why.
Repository and shelfmark references
Repository references are crucial, and for so many reasons. First of all, it is a matter of courtesy: libraries and museums care for their materials, and digitize them, at great expense. They then give them away for free for me and you to reuse as we like. A little hat tip, I always feel, is the least I owe them. Secondly, it is a way for people to follow up where you got the picture from; if I can entice one person to browse through a manuscript online, my job is done. Thirdly, it is a way for myself to be able to find things back when I want to, sometimes at a much later date. Fourthly, there is this: it’s history, not a viral feed!As Sarah Werner points out, not only repository, but also shelfmark references serve an important purpose: they identify the unique object (in my case, usually, the individual book) from which the image is taken, making it possible for others to follow up on details that may be particularly interesting to them. Finally, I don’t begrudge angels their slice of pizza.
whenever you retweet a manuscript image w/ proper identification and a link to the source, an angel eats a slice of delicious pizza for you
— Karl Steel (@KarlSteel) December 1, 2015
Repository references can come in many forms, and can contain a range of types of information. Over time, I have developed the following practice – this is my current personal practice, but I don’t suggest it is best practice, and I am open to suggestions for what I could do better. An alternative suggestion, using hashtags for identifying manuscripts, is proposed here. Some of the choices come down to a balance between being thorough and making appealing picture tweets within the character limit. That said, because of all of the above, I personally feel there is never a reason not to include a repository reference which outweighs all the reasons there are to include it. In principle, I apply the same minimum standard to retweets, too: I will not retweet image tweets which do not include attribution to the repository. What I really appreciate from other picture tweeters, however, is answering queries. It is the hallmark of a caring tweeter, versus the medievalreacts of this world.
So, at a minimum, I include, on the picture itself (i.e., added with a picture editor), a repository reference and shelfmark (i.e., the unique identifier of the book within the repository), often abbreviated, but enough to make it clear and searchable. So, for example, ‘Bibliothèque nationale de France, Département des manuscrits, Français 2664’ can become ‘BnF Français 2664’, or anywhere in between the two – enough to identify the manuscript, even if not a proper full shelfmark. I prioritise this reference on the image over all other forms of reference because it has been my experience that people take images (and tweets) and recirculate them, either on twitter or on other platforms, stripping them of any reference information you have included in the tweet. So, you may have included a link and repository and shelfmark reference, but the image starts circulating without the information (but with the joke you made!) regardless. Including the shelfmark on the image makes it a little less likely for the repository reference to get lost. People may run away with the interesting detail you found in a digitized manuscript, and perhaps “steal” one of your jokes, but at least the source of the image remains identifiable, in whatever context it turns up, unless someone makes the effort to crop the image to delete the information – which they’re very unlikely to do. I use a simple image editor (Mac Preview) to add the reference to the image.
Additionally, if nothing else, I tag the repository’s twitter @-handle to the picture, if they have one, using twitter’s “Who’s in this photo?” functionality. That way people can follow up, and the institution can see what is being done with their material.
In many cases, I also include the repository’s @-handle, the shelfmark reference, and regularly also a link to the digitized version, in my tweet. If I do the latter, however, I don’t do it specifically so people can admire the detail I have tweeted – I have, after all, already tweeted that; I do it in the hope that someone will take the time to browse through the manuscript. So, to make that a little more likely, I link to the whole work, not to the specific page view. This is also a reason why I don’t include a folio number in my references: except when a tweet has specific scholarly purposes that require one, I feel the collection/shelfmark suffices.
Sometimes, however, tweets can start to feel clunky when they include too much information; a simple repository/shelfmark reference, in my experience, is rarely ugly, but also including a link can sometimes make the additional information in the tweet overbearing. For that reason, I regularly choose to either link or provide the reference; although I still feel best practice is to do both. Also, I feel bad for the angels, whom I’m depriving of pizza.
Besides repository references, twitter allows to include a further type of data, which however is not visible to most users of the platform: descriptive information. This is a functionality twitter has introduced not so long ago, which you can use to make your picture tweets more accessible for people who are visual impaired. It is only readable via dedicated assistive technology. You can learn here how to turn on this functionality. I hope it is useful to people who access my tweets through assistive technology, and have once done a twitter poll which suggested it was indeed useful to some. I do not know, however, whether the descriptions that I give are any good, and would be very grateful for any feedback any users could give me.
So, this is my personal practice at the moment – it is very much subject to change, as twitter slackens its character allowance, and as I meet new issues, or am persuaded to improve. What is your practice, and why? And, as a reader of my tweets, what more (or less) would you appreciate to see me include?
[This is a revised version of a post on my blog, levelt.wordpress.com; it was greatly improved by responses I received to that post from several readers, in particular Olivia Thompson.]