Octopus is a new platform for scholarly research: somewhat like a preprint server, with instant publication, but Instead of a full article, a researcher posts just a component of an article: a problem statement, or a hypothesis.
I attended the Octopus launch meeting (29 June 2022), and from the calibre of the sponsors of the free service (including funding and support from JISC and from Research England, with David Sweeney, executive chair, enthusiastically endorsing it), it sounded at times like Octopus was the biggest research innovation since the scholarly paper was created in the 17th century.
What is Octopus? It simply divides the research process into eight components (hence the name). Instead of publishing a complete research paper, the academic can simply publish something in any one of these eight components, which Alex Freeman, the creator of Octopus, called “publication types”. They comprise:
- Real-world application
Octopus looks like this (Alex Freeman pointed out this is sample data only, but it gives a good impression of how the system will work):
So, if you have a contribution to make or a comment on the methodology, you can post a note about it and it will appear under the “protocol” section. The rules for posting a very simple:
- Publication is immediate. Any review takes place after publication.
- Every post has its own DOI (so potentially eight DOIs for one full paper)
- You can only post in relation to what is already on the platform.
On the crucial question of accreditation for contributions, Steven Hill, director of research at Research England, seemed to suggest that Octopus would be valid for researchers to obtain credit (presumably towards their regular appraisal).
When asked if this supplements or replaces traditional publishing, Alex Freeman stated clearly it would take place alongside published articles, but would not replace them.
The great appeal of this model is that every academic paper can be subdivided into components, such as methodology, so it makes sense to provide access to individual components of the research workflow.
Are there any problems with this model? I can imagine quite a few (not that I am against innovation, but I’ve spent a long time observing the motivation behind the scholarly publishing process)
- Researchers publish papers to get promotion. They are interested in the topic, but I think it is an open question the extent to which they are prepared to participate in a debate about details of their research. They may be very reluctant to provide interpretations of analysis that can then be modified by other researchers. At present, this happens by someone publishing a whole article.
- One physicist immediately commented that the eight-stages set up within Octopus worked better for life sciences than for physics. More than that, this structure is unlikely to work for the humanities, although it might for social science.
- More fundamentally, we were all taught that there are two ways of doing scientific thinking: inductive and deductive. The Octopus method is only deductive. Where is a framework for research based on observation, with no hypothesis behind it?
- Reviews were mentioned in the launch event, and Alex Freeman stated they would be quality checked like other peer review. I would love to see reviews given a score, but I don’t know how feasible this will be. I find reviews a fascinating part of the scholarly ecosystem, but that is a subject for another post.
- The Octopus system, if used widely would generate many thousands more articles than we currently have. It seems odd to fix a problem of too many articles by creating a system that will generate a multiple of those articles, by several times.
- This structure does not work for many kinds of academic article, such as the literature review, of the medical case study, or the discovery of a new species of plant.
- There was no mention of any restriction from posting new content to Octopus. The academic publishing system makes it difficult to post new content, or to comment on published content, next to the original content. There could be many people adding “this is a cure for cancer” type of comments, with no curation.
- A neutral platform such as Octopus eliminates all commercial publishing, so you can expect the scholarly publishers to respond with modifications to this system if it ever looks like becoming widely used.
More than one of the panellists commented negatively on the idea of a published academic article as a narrative, which seems odd, given that other initiatives are trying to persuade researchers to structure their article more like a narrative, to make it more accessible. After all, we all love a story.
So I wish Octopus well, but it is very early days before there is any indication that it might be a success. I think it deserves all credit for challenging the existing, and very imperfect, scholarly publishing model. I just hope it gets enough usage, and has a good enough feedback loop, to evolve and start to be a genuine research tool.