Dave Beer makes some interesting points in this short article which frames current debates about open access in terms of trends within music culture, which have been driven by broadly similar structural processes and have been playing themselves out for much longer. Some people argue that rendering academic publishing more ‘open’ could prove hugely problematic, as unrestricted access to the means of ‘making public’ could lead to an even more confusing mass of literature than exists at present:
Kirby points out that the ‘footprint’ of academic publishing has grown ‘exponentially’ since the 1960s. His point appears to be that even in the current system we are already faced with an unwieldy mass of journals and journal articles. This, as the statistics indicate, already represents an unfathomable amount of published resource. Anecdotally we can probably reflect here on how hard it is to keep up with the articles published in our own specialist areas let alone across entire disciplines and cognate disciplines. The point, it would seem, is that this situation could very well get even worse if we adopt an unlimited open model of publishing. The implicit observation is that academic publishers limit and classify the output of academics and thus filter and order the content. Kirby’s suggestion is that with open access we will be opening the floodgates whilst also losing some of these ordering mechanisms. As such we might well end up, Kirby (2012, p. 259) claims, with even more material and with ‘no obvious means to make any sense of it’.
There’s a certain plausibility to this argument though I think it, as well as Dave’s response, ignores the significance of network filtering (as happens through social media) and correspondingly overestimates the significance of bureaucratic filtering (as happens through publishing corporations). Journals are becoming less important to discovery and perhaps less important as cyphers of quality, at least in terms of the reading decisions individuals make (clearly this is untrue in terms of institutional use of prestige as a cypher for quality and associated decisions about the distribution of scarce resources). Nonetheless, it’s unrealistic to imagine that a socially networked academia could, as an unintended emergent effect of sustained online communication, “limit and classify the output of academics and thus filter and order the content”. Which is why the mechanisms being deployed within the music industry to filter content and aid discoverability are potentially so relevant:
Against this backdrop of cultural cacophony, music cultures have found ways to organise content so that it might find an audience. The broadcast model, despite the recent changes, remains relatively powerful in music. And so we still have large organisations shaping consumption at the top end of the market through TV shows like The X Factor. Amongst the chaos in music we also have systems that enable people to locate new music that they might like to listen to. Here predictive analytics are increasingly beginning to rule. Music fans are often told what music they are likely to want to listen to by machines. Software applications built in to iTunes, Last.fm and the like are predictive of tastes. These predictive systems provide a means for managing the chaos as the music automatically ‘finds’ its audience. This is an era in which, as Scott Lash (2006) has put it, the data ‘find us’. The practices of tagging music with metadata and the accumulation of personal profiles combine to enable this form of ‘knowing capitalism’ (Thrift, 2005) to operate. If this were to happen in academic research then one of the solutions to the intellectual din of open access might be apps that enable research articles to ‘find’ their desired audiences. This app might well predict what type of thing you might want to read and recommend it to you. An early version of this type of recommendation system, although based just on links with individual articles rather than one’s comprehensive browsing history, is provided by Elsevier’s ScienceDirect platform. It sounds nice and convenient, but as with music we are forced to question the underlying algorithmic infrastructures that will then be so powerful in shaping the formation of knowledge (Beer, 2009; Hayles, 2006). These algorithms will shape the knowledge we encounter and will in turn then impact upon our ideas.