Can open-access online publishing be interdisciplinary in both merit and means?

Not long before we set out to build an internal prototype of the Listening Experience Database and bootstrap it with an initial dataset, another research project had kicked off, which involved the authors of this article in its team. It was a short project aimed at assessing and advancing the state of maturity of open education.

The term “open education” alone is loaded with significance and prone to multiple interpretations, even within the project itself. Open was the learning material that we went hunting for (you might have heard of massive online courses, or MOOCs), and open were the data that described these resources and the educational context where they lived; and finally, open was the documentation produced by the project itself as a result of analysing the other two “declinations”, as it were, of open education. This last product was an Open Education handbook, an ever-evolving and crowd-sourced publication available in pretty much all the existing online publishing formats such as Booktype, Wikibooks, epub and of course a downloadable, DRM-free PDF.

When that strand of research was over and its material published, the ineluctable scholarly curiosity got us to sit around once more and pose the unfailing question “Where to go next?”. As it often comes natural, the search for possible parallels with listening experiences arose. What are the parameters that define such a thing as “open music listening”? We might as well try to provide this notion bottom-up, starting from the contributions that the Listening Experience Database set out to provide and envisaging the point where any loose ends can be tied.

We committed to deliver open data right on the outset. Everything that our community contributes to the database is re-delivered back to the public after going through a review process. Not even the metadata on the approval history are held back. More so, we deliver linked open data, that is, a machine-readable form of what can be browsed and searched on the project website. Linked open data are rendered using open standards and, to abide by its recommended set of practices, they reuse as much as they can. This means that our data capture system makes every effort to make sure, whenever possible, that the things described in the database are not isolated concepts that are being re-defined over and over again, despite the Web showing that they are of common knowledge. So therefore, we should see to it that Ray Charles as described here is the same as the one on MusicBrainz; that “The Letters of Charles Burney” is the same edition as one on the British National Bibliography, and so on.

As for the range of possibilities that doing so opens, the sky is the limit. For a low-hanging fruit, this approach makes it much easier for our institutions to build recommendations of learning material on the things described in this – let us now overhaul it terminologically – knowledge base. And what is this, other than implicitly making the learning material itself better “data-wise connected” to the rest of the world? These are the basic mechanisms of making sure the objects described (such as a piece of music, an author, a location, or an instrument) is represented in the same way as external material – think lecture notes, a study book or a whole online course – is annotated with it.

That takes us back to our pursuit for an open ecosystem where the listening experience lives. In LED, the documented accounts of the experience in text form are exposed alongside the references used for describing the context around it, thus making the material and the data describing it both open as one single resource. That makes two of the aspects covered from our experience with the educational domain. Disseminating the results of this research – to be often taken quite literally as the activities that led us to publish some unique literary sources for the first time – through a comprehensive form of documentation is the next aspect to be addressed. Publishing an online book as an epub or a downloadable PDF is all well and good, yet this time around we have the advantage of controlling both the material and the data ourselves – and in a combined form, at that.

From a purely academic perspective, this is a rare opportunity for a research endeavour to materialise its production and documentation as a continuously integrated living system, which naturally evolves because so does the knowledge around the entities it describes and formally incorporates. Taken broadly, the database website itself is but one form of documentation of the dataset it is built upon. We picked a set of dimensions according to which we expected users to be incline to explore this dataset, but this is by no means intended to be exhaustive of the potential behind the data themselves.

Having adopted the Linked Data paradigm allowed us to conceptually reiterate that the LED website is not the database, merely a (re-)presentation of its dataset, and that essentially the database is what you make it. That could very well include its documentation. Now we need to take it one step further, and one of the ways to do so is an academic publication that, along with incorporating original research, reflects the same potency for growth that the resources it is based upon have. In that sense, it is time to implement the lessons learned from the field of semantic publishing and its software offspring. We can therefore only welcome an effort such as the Academic Book of the Future, and envisage it as a potential driver for promoting an evolved open-access publishing paradigm; one through we can implicitly teach the merits of the Semantic Web through the means by which we deliver our contribution to musical studies.

– Alessandro Adamou and Mathieu d’Aquin