|The era of machine learning is upon us.|
Sunday, 6 May 2018
Academic boycott of new journal over open access
Many months ago, this blogger wrote about ‘Open Science and Open Culture’. I’d like to continue this line of discourse if I may, (having recently emerged from a writing malaise borne out of a seemingly unending job search) with a discussion on academic publishing and open access. Of particular interest is the boycotting of a brand new academic journal entitled “Nature Machine Intelligence”, which is to be launched in January 2019 by publishing giant, Springer Nature.
Nature Machine Intelligence will be an “online-only journal for research and perspectives from the fast-moving fields of artificial intelligence, machine learning and robotics” (quote source here). The fledgling commercial subscription journal has, however, been met with a certain amount of hostility by the academic community. Indeed, 2,816 individuals (at time of writing) have refused to “submit to, review, or edit for this new journal” in a Statement on Nature Machine Intelligence. This move is a signifier of a larger cultural shift that has been growing since the inception of the internet, which seeks to move away from ‘paid-for’ access to research knowledge to an open access model. The objective of this open access model is that anyone with an internet connection might furnish themselves with the latest scientific and academic knowledge and insights.
The traditional system of academic publishing
In order to have a piece of work published by a respected publisher (a must for many academics – ‘publish or die’), academics must usually assign the copyright in their work to the publisher. The publisher then locks the work behind paywalls, charging money to gain access to the research output via subscription payments. This maximalist use of copyright – to limit access to academic knowledge to only those who pay the toll – is what proponents of openness argue hinders development in society. Subscriptions to academic journals will usually only ever be paid by universities and other large institutions or companies and even at that, these large organisations face problems in being able to afford publisher subscription charges: see this famous memo from Harvard, one of the worlds richest and most prestigious universities, for an example of the budget strain that besets many institutions as a result of prohibitive publishing charges. Probably the most galling aspect of the situation is that – despite academic research being funded mostly from the public purse – large swathes of the general public will never be able to access research output and are therefore unable to gain any of the benefits brought about from these tax payer funded investments.
The push for openness
The driving rationale behind the push for openness is to create a counterpoint to an ‘all rights reserved’ culture. Bestowing property rights in intellectual creations is intended to “promote the progress of science and the useful arts” (Article I, § 8.01.8 of the U.S. Constitution), yet there is a growing pushback within society which posits that the propertisation of all intellectual creations actually hinders the development of science and the arts. Innovation is always built upon that which came before it – everyone is always somehow “standing on the shoulder of giants” after all. However, if standing on the shoulders of giants entails insurmountable difficulties – such as exorbitant licence fees, expensive copyright infringement lawsuits or even a complete inability to gain access to knowledge and research – innovation will likely be stilted. Academic publishing is something that many openness advocates would describe as the archetypal example of innovation being hindered because intellectual creations have been commodified and propertised. For publishers to use copyright as a tool to generate profit from research output that was funded largely by the tax payer presents legitimacy issues for the existence of copyright, as well as the publishing industry.
The internet has been the fundamental driver behind open values. Various international legal instruments have been crafted to push forward the openness agenda, with the objective of making academic research available to anyone with an internet connection. These include the Budapest Open Access Initiative, the Montreal Declaration, the Bethesda Statement, the Berlin Declaration, the Durham Statement of Open Access to Legal Scholarship. The opening of the Berlin Declaration is particularly powerful:
“The Internet has fundamentally changed the practical and economic realities of distributing scientific knowledge and cultural heritage. For the first time ever, the Internet now offers the chance to constitute a global and interactive representation of human knowledge, including cultural heritage and the guarantee of worldwide access.”
This paragraph emphasises the key importance that the sharing of knowledge has to our species and how the internet has made the sharing of knowledge easier than it has ever been. Indeed, the dramatically decreased costs of sharing academic outputs that has been engendered by the internet makes spiralling publisher subscription costs somewhat confounding. Making knowledge available to everyone opens up possibilities of collaboration and follow-on innovation; it helps us develop our understanding of the world, as well as our understanding of ourselves.
Statement on Nature Machine Intelligence – the boycott of Nature Machine Intelligence
Many of the major academic journals that cover machine learning (which covers the plethora of science relating to artificial intelligence) are already open source. These include the Journal of Machine Learning Research, NIPS, ICML and others. Importantly, these journals do not charge for access, nor do they charge authors for publication, which is in direct contrast to the subscription based Nature Machine Intelligence. The Statement on Nature Machine Intelligence explaining the boycott emphatically declares that signatories
“see no role for closed access or author-fee publication in the future of machine learning and believe the adoption of this new journal as an outlet of record for the machine learning community would be a retrodgrade step. In contrast, we would welcome new zero-cost open access journals and conferences in artificial and machine learning.”
The boycott of this new subscription journal by nearly 3,000 experts within its specific scientific field signals that the academic community are pushing back against the closed-access model employed by the majority of academic publishers. Open access is moving into the mainstream culture within academia. With it comes new and exciting opportunities of follow-on innovation, collaboration and new possibilities for the advancement of human knowledge.