Disclaimer 1: many of the ideas have been inspired by the work of Alex Gil (@elotroalex) and Denis Tenen (@dennistenen)1 and discussions with the two of them and others at DHI Beirut, THATCamp Beirut, and DHSI.
Disclaimer 2: I am teaching short workshops on some of the ideas outlined in this post at Digital Humanities Institute - Beirut on 10 March 2017 and at DH Abu Dhabi on 10 April 2017. Basic slides are available here and here.
In the world of (academic) publishing, large aggregators and indexers have turned into and acquired publishing presses and generate obscene profits by charging the public (every tax-payer worldwide) multiple times over. First by charging the predominantly publicly-funded academic for publishing the results of her publicly-funded research and by enforcing a culture of pro bono labour among academic reviewers and editors; second by selling this content to equally predominantly publicly-funded libraries, which then increasingly demand access fees from members of the public, who want to access their collections; and third by offloading the cost of long-term preservation to, again, publicly-funded institutions. This system not only created a hierarchy of academics and institutions in the relatively well-off “West”—two classes divided by their ability to pay for being published and accessing publications (their own and others). It also increasingly prevents anybody outside western academia from accessing cutting-edge research and participating in intellectual discourse.
One of the opportunities afforded by the digital humanities and the stated goal of this endeavour is to remove the middlemen—be they technical or entrepreneurial—between authors, readers, and the library-cum-archive. There are two main obstacles to this aim:
We will not be able to change copyright legislation and the vested business interests in sustaining and expanding the regime of profit-generating copy and distribution rights in the foreseeable future, but we can all provide our knowledge under a creative commons licence. In order to do so, we need to re-claim the means of production. We argue that by doing so the main argument for restrictive copyright—namely, the provision of allegedly expensive services such as quality control and metadata curation—collapses. By the time of writing, printing costs and the global distribution of heavy and voluminous books are already negligible as the main avenue of scholarly publication, the journal, has already moved to digital online publication.
The main principles in our effort to (re)claim the means of production are: accessibility, simplicity, sustainability, and credibility. They shall pertain both to the intellectual endeavour and to the tools employed.
Looking at the most broadly-employed software in academic contexts and beyond, Microsoft’s Word, a piece of bloated and expensive proprietary software, what are the functions we need to replace?
To avoid overtly complex software and proprietary formats, form and content must be separated. Structural / semantic and representational information as well as metadata must be embedded in the text itself in order to be inseparable. We suggest using plain text with rudimentary markup following the conventions of MarkDown and a short block of metadata written in Yaml as the format of choice.
TXT), this string of letters happens to be human readable. Plain text has been with us since the early days of computing and TXT files could be viewed and edited with 1980s hard- and software. We can therefore assume that this basic format will remain accessible for the years to come.Writing is a process and subject to change. We need to be able to try out different structures and formulations and, more often than one would like to, we discover that yesterday’s deletions would have been worth keeping. Not to mention an external editor or collaborators that quickly make any approach involving ever-longer file names futile (we all have our folders full of text.docx, text-new-version.docx, text-new-version-2004-01-01.docx, text-new-version-2004-01-01-comments-by-tg-2.docx etc.).
{»We need additional collaborative tools«}
While it would be absolutely sufficient to publish / distribute the plain-text files by means of a USB key, a graphical user interface (GUI) that translates the structural information into a formatted and aesthetically pleasing layout is often {==advisable==}{»better wording?«}—be they a printed paper copy or a website:
HTML, DOCX, and PDF. Pandoc, which is under active development by John MacFarlane, a professor of philosophy at UC Berkeley, also supports the formatting of references using BibTeX (another plain text format for storing structured bibliographic information) and CSL (citation style language).HTML and CSS: HTML and CSS are well-established standards to separate form and content maintained by the World Wide Web Consortium (W3C). While the hyper text markup language (HTML) carries the content of our text as well as all the structural and formating information and the metadata, cascading stylesheets (CSS) provide the actual layout. This combination allows to provide different layouts for different contexts and devices using the same content.HTML5 supports semantic tags and machine-readable metadata following various standards, such as those provided by Dublin Core (DC) or schema.org. If one includes Dublin Core metadata in the head of HTML files, aggregators, search engines, and reference managers can find and extract the structured information on author, title, publication date, keywords, etc.{»mention copyright«}
A licence is formal agreement that specifies the rights and duties of both the licensor (e.g. us as authors) and the licencee (e.g. us as readers). Its most important purpose within our discussion is to assure the readers of our texts of their rights to read, copy, and cite them. An open licence might for instance allow reproduction of the text but might prohibit charging for accessing the reproduction.
Formulating one’s own licence text is a challenge and one might not be familiar enough with the necessary “legalese” to write a text readers can rely on. In consequence we suggest looking at established licences and having made a case for open access {»open science, open knowledge etc.«}, we suggest to start with creative commons licences.5
Websites of more than a single page of text tend to be technically complex and require an infrastructure of data storage, internet connections, web addresses, some content management system (cms), and databases that rarely come for free and without the need of maintenance. In addition, contemporary dynamic websites are almost impossible to archive or download in their entirety. To reduce complexity and the number of technologies, we suggest using static, self-contained websites without a content management system and no database.
HTML and CSS and nested files and folders. These can either be hosted online or distributed on a USB key.How to deal with sensitive data / material, that should not be publicly accessible, such as ethnographic field notes?
{»add relevant literature«}
Tenen, Dennis and Grant Wythoff. “Sustainable Authorship in Plain Text Using Pandoc and Markdown.” ↩
MarkDown really is only a convention and John Gruber’s (and Adam Swartz’ [yes, the Adam Swartz]) canonical description of the Markdown syntax is at least partially ambiguous and and lacks some core functionality for academic writing, such as support for tables and footnotes. In consequence, a plethora of formats (MultiMarkdown, GitHub flavored markdown, etc.) and software implementations have proliferated inspired by and based on Markdown that make the actual rendering of Markdown in HTML rather unpredictable beyond the core functionality. In recent years, a group of people involving John MacFarlane, professor of philosophy at UC Berkeley and author of Pandoc, proposed and developed are more rigid standard which they call CommonMark. ↩
There is a great list of open source licences, including links to their full texts, at opensource.org ↩
As always, I have made the slides available on GitHub.
]]>