The Letters of 1916 is a project involving the transcription and compilation of letters written by or to Irish residents between November 1915 and October 1916. Originally begun at Trinity College Dublin, the Letters of 1916 project has recently been transferred to Maynooth University. My Digital Scholarly Editing class has had the privilege to assist with the upload and transcription of some of the letters on the website. But one of the things that has fascinated me most about this project is the use of crowdsourcing to transcribe the letters and the methods of promotion that have been utilised to garner public attention.
Crowdsourcing the Transcriptions
At initial glance, one wouldn’t think there would be many letters to transcribe for such a short period of time. But once you step out of a modern mindset and realise that, during this time period, letter writing was really the only way to communicate, you begin to understand the sheer breadth of what this project is attempting to undertake. Factor in that during this time, Ireland is firmly enmeshed in World War I, and the Easter Uprising, a prominent event in Ireland’s fight for independence, occurred in April of 1916, there is quite a bit of activity going on for the citizenry to discuss. So how does a small research team transcribe all of these letters?
Enter the concept of crowdsourcing. Wikipedia defines crowdsourcing as “the process of obtaining needed services, ideas, or content by soliciting contributions from a large group of people, and especially from an online community, rather than from traditional employees or suppliers.” The idea is to leverage your audience, the group of people most invested in your product, to assist with the collection of your content. While some have maligned the concept of crowdsourcing as nothing but free labour (see Crowdsourcing: Sabotaging our Value), crowdsourcing has become an important tool in the content collection space, especially among non-profit endeavours.
So how does crowdsourcing work with the Letters of 1916? It’s fairly simple. Anyone can upload a letter (although many of the letters are contributed from agencies such as the National Archives of Ireland, the Military Archives of Ireland, University College of Dublin Archives, etc.) and once the letter is uploaded, any user can then transcribe the letter using a standard set of transcription tools provided by the website (Letters of 1916 utilises Omeka and Scripto to assist with the transcription efforts).
By the Numbers
The success of the crowdsourced transcription effort has been great. As of 31 October 2014, more than 1,000 letters have been transcribed or were in the process of being transcribed (approximately 71% of all currently uploaded letters), and October saw the addition of more than 30 new members to the transcription effort. For more information on the numbers, please refer to the October 2014 Progress Report.
These numbers show a positive trend in the use of crowdsourcing to leverage the audience for the Letters of 1916 in the creation of content for the website. Unfortunately, Omeka tracks only character counts and doesn’t really provide a solid look into the demographics of the transcribers or the extent of their contribution beyond the actual number of characters transcribed. Therefore, it is difficult to see the contributions of those transcribers who are proofing other people’s contributions but aren’t contributing to the overall character count, as they may only be changing a word here or there. This is one area where Omeka really falls short in terms of attempting to understand the scope of contribution of your user base. But the overall sense is that the crowdsourcing effort is highly successful.
Modes of Engagement
So what contributes to the crowdsourcing effort being as successful as it is? It is difficult to tease out any one particular item, but I would posit that the use of social media by the team has led to a strong engagement. The team utilises Twitter heavily to promote not only the site but also items related to the site, such as news articles and other events occurring within the Digital Humanities space. In addition, the team holds a monthly twitter chat utilising the hashtag #AskLetters1916 (check out Storify for the latest #AskLetters1916 chat). The team also leverages Facebook in addition to a blog to advertise the site to those interested in the history of the time period or with a general interest in Irish History or Digital Humanities.
Room for Improvement
While the site has been relatively successful in its efforts to leverage crowdsourcing, that doesn’t imply there isn’t room for improvement. While analysing the site for this article, I came across a couple of items that I thought could be improved from an interaction and usability standpoint.
First, there is A LOT of information on the site. So much, in fact, that it is very easy to get lost in the weeds and forget why you came – and that’s before you even get to the transcription area of the website. There are 7 top-level menu options, many of which have multiple sub-menus. There are a number of really interesting and helpful resources related to education and current news as well as the obligatory “About Us” and Sponsorship pages. These are well and good, but if the main purpose of the site is to contribute a letter or transcribe a letter, I wonder why there isn’t a persistent top-level menu item just for that. Yes, there is a “Contribute” top-level menu, but to do the actual transcription or contribution, one has to navigate to a sub-menu and follow links to login or signup. In addition, the menu item is easily lost among the other items.
As an alternative, I would suggest adding a persistent option for Contribution in the form of a button, that is coloured so as to stand out. I would also place it in the upper-right hand corner where the login / signup metaphor typically exists. This would draw the eye to the primary purpose of the site as well as facilitate a faster workflow for those who are returning to the site and simply wish to login to submit a new transcription or upload a new letter.
My second suggestion comes more from the transcription workflow itself. When attempting to transcribe a letter, the only view available is by category. As a user, I have no way of knowing which items are already transcribed and awaiting review, which are completed, and which have not been started. I have no options to sort or even filter. I’m simply presented with a long, scrollable list by category of the items loaded into the system. Once an item is selected, I can then see the status of the item (Not Started, Needs Review, Completed, etc.), but it requires an additional click to do the actual transcription (a click I deem largely unnecessary). Finally, the transcription tool itself is a little clunky. There is a toolbar provided to assist the user with standard TEI encodings, but as the average user may have no knowledge of TEI and the transcription page provides no explanation for how encoding should be handled, a number of transcriptions require a lot of clean up in order to conform to standards. Much of these complaints are, however, limited by the Omeka and Scripto software, so they are a criticism aimed more at those particular implementations than at the Letters of 1916 project itself.
Criticisms aside, the Letters of 1916 project has done a great job of garnering attention and drawing in its audience in order to facilitate the creation of content. The next step in the process is to migrate from the transcription desk to a site that is searchable and discoverable. With the implementation of a strong search mechanic and a few visualisations of the data to add a little spice, I think the Letters of 1916 will set itself up to be a rousing success.