What is open research?

Next month I am attending the first conference of the Open Scholarship Initiative. The format of the conference is a series of working groups examining key questions for open research, and I will be taking part in a workshop that will address the question ‘What is open?’. I have written this post partly to organise my own thoughts, and also to share with the rest of the working group, and more widely.

What is open research? It is a fairly fundamental question, superficially simple, but I am sure one that will generate much debate. For me, there a number of lenses through which to view this question.

The first is to frame as different degrees of openness, and I would argue that there are three different qualities to consider. Can content be:

  • discovered?
  • accessed?
  • reused?

These characteristics have a sequential relationship, in that content can’t be reused if it can’t be accessed, and access depends on discovery. In addition, all three steps have some public value attached to them. If content can be discovered, at least a potential user can take steps to access or reuse. And similarly, there is value, for example, in being able to just read a scholarly article, even if more value might accrue if full reuse permissions were attached. The relative values of the steps will vary depending on content. While there is a value in access to a scholarly article without the ability to reuse, this is less true of a dataset.

As well as there being value at all three levels, and despite the focus of debates on access and reuse, there is potentially a lot to be gained from improvements in discovery. For scholarly articles discovery is in good shape. Using general purpose search engines, Google Scholar, Scopus or Web of Knowledge (if you have access to the latter two…) finding scholarly articles in English is reasonably straightforward. It is harder for non-English content, but still possible, given that many non-English articles contain titles and abstracts in English.

But there are areas where even improving discovery might make huge advances for open research. For research data, it is not even possible to know whether or not data addressing a particular research question even exists. Similarly, it is hard to find out about ‘negative results’, or experimental approaches that didn’t work. It is also not always straightforward to find out who is researching a particular topic or question, although funder databases, like Gateway to Research in the UK, are starting to address this.

A second way of framing the question of openness in research is to consider the barriers_ _to discovery, access or reuse. There are many potential barriers, and they will vary in importance depending on the content, and along the discovery-access-reuse continuum. One that we don’t talk about very often is internet access itself, but in places that don’t have reliable electricity supply or fixed or mobile internet access, debates about the minutiae of open research must look quite esoteric. There are also barriers created by expertise. These are real, but can be overplayed, particularly within the academy. Despite what some researchers believe, there are plenty of people outside of universities who are perfectly capable of making use of research-related content. But I still think it is true that a commitment to open research should also mean a commitment to making content useful for people with expertise that is different from your own.

Another barrier can be know-how, which is subtly different from expertise. Take text-mining as an example of reuse of scholarly articles. It isn’t necessary to have deep expertise in the process of text-mining to carry out this activity, but straightforward and easy to use software tools are required for those without the deep expertise.

The barrier which receives a lot of attention is the financial one; the much-maligned pay-wall. I wouldn’t downplay the importance of this barrier, but I do think that sometimes the debate becomes overly simplistic. The size of the financial barrier is important as well as its existence. I wouldn’t disagree that £40 for a 10-page article represents a real barrier, but it would be possible to identify lower costs that were more reasonable. And there is a separate issue about the ease of access. At least part of the problem with the pay-wall is the need to stop and go through the process of payment. At least part of the reason that Sci-Hub is popular seems to be related to its ease of use. I know it’s a poor analogy and not without its own problems, but Spotify’s £10 per month subscription that buys easy and friction-free access to almost all musical content is an interesting alternative way of thinking about paid-for content.

There are plenty of other barriers (issues of licensing, for example). I think it’s important to think through the barriers carefully. Some are more important than others, and the solutions are not always obvious.

So those are my two lenses on the question ‘what is open research?’. The degrees of openness and the barriers to openness. There will be plenty of other ways of approaching the question. I haven’t even mentioned the issue of what to make discoverable, accessible or reusable, for example. But hopefully there is some food for thought here, and a start to the debate.