Over the last month I’ve found myself in a strange vortex of image permission issues. So I thought I’d have a little rant, but also wanted to play around with some possible implications for debates about open access and ownership in academic publishing.
The disappearing figure
I have a paper close to publication that includes (or rather, was going to include) a composite figure made of eight images that had been published in popular articles about neuroscience. I attributed and referenced them all fully, naively assuming that my duty had been dispatched. But I soon learned that as a journal is a commercial enterprise I would have to obtain, and likely pay for, permission to reprint the photographs. Fair enough, and I’ll be on top of this from the outset next time, but the quixotic nature of my quest for permissions suggests that the system isn’t exactly a well-oiled, rational machine…
My first attempt, to navigate a US news website’s byzantine tangle of press contacts and customer feedback forms, led exactly nowhere. Contacting the University where the reported research study had been conducted led me to the press release containing the images, but my hopes were dashed when it turned out that the images in the press release weren’t actually from the study. Indeed, no-one knew where they were from. And this wasn’t the only instance of an image featured on a website, and republished by countless news feeds and blogs, whose origin was effectively unknown.
An indication of the scale of this problem is that Getty Images offers an Image GuaranteeTM service, indemnifying you against claims should the owners or subjects of orphaned images come out of the woodwork. The problem for academic research is that it’s the publisher who gets to make the insurance gamble, and understandably enough they prefer not to place too many chips on the table.
Brownie points for New Scientist, who pointed me to the image libraries they’d used, and Discover magazine and The Daily Mail were also helpful, digging out details of the photographers who’d taken the featured snaps. The photographers both graciously (and freely) gave me permission to reprint. But unfortunately, both featured people. Or should I say, models. Even permission from the publication and photographer, to reprint images already in the public domain, was not enough. If images feature people and are to be reprinted in a commercial publication that is not deemed public-interest journalism, they must be directly approached to sign a ‘model release’ form, and this hasn’t been possible. Final tally for my figure? Two out of eight + a bunch of hyperlinks that might expire.
We’ve encountered a similar problem with using photographs from a Museion event in an academic thesis, despite participants giving permission to be featured on our website, and I recently heard an anecdote about a book that had to be withdrawn because retrospective model permission could not be obtained for the museum visitor featured in a photograph on its cover.
In the public interest?
Setting aside my sci-commer’s twitching at the notion of a single thing called a public interest, I appreciate the intended distinction between commercial use and public interest journalism. However, as a distinction it’s an extremely creaky one – it only operates at the extremes, and under an outdated model of information flow. Why do we still agree that a news company holding a video camera, whatever their funding structure, motivations, and ethical practices might be, is excused from an issue that dogs small-scale academic publications and of course the cadre of bloggers who would fiercely defend their ‘public interest’ goals?
It seems to me that we should be able to argue that academic publication is in the public interest, but a lack of clarity over who pays for it and how makes it too onorous or risky for journals to argue the case. Without wanting to get into the open access debate here, one possible consequence of free access to publications could be a more sensible position on reprinting images.
This is not an Author
When you consider scientific figures, the rabbit hole gets even more psychedelic. The idea of an image creator is controversial within many scientific disciplines, which wrestle with the thin digital line between data representation and artistic manipulation. Again, open access provides some interesting thought experiments here – for instance, if data itself is freely available, would remaking a scientific figure from that data require permission of the ‘creator’ of the original image?
And to throw Alice down the rabbit hole as well, when looking at biomedical images a model struts back onto the scene. Difficult questions about the ownership of biomedical data and confidentiality have been largely kept at bay by the difficulty of identifying someone from medical images, which are often group averages or cropped for facial and other strongly identifying characteristics. Future developments in biomarkers and digital image processing could shift this criterion in interesting ways, and in useful juxtaposition to the questions of what ‘identifiable’ means in a photograph, and what we implicitly sign up to by placing ourselves in various image-able situations. As a final thought experiment, if people whose medical data was presented in a scientific figure could be identified from it (and setting aside confidentiality issues), could they claim that a model release must be signed, even though the figure doesn’t ‘look like them’ in the classic sense?
In browsing around these issues, I came across a workshop on The Lives of Property in Oxford in September, asking amongst other interesting questions ‘How do competing understandings of ‘property’ affect the circulation of scientific knowledge and artifacts?’. I’d love to hear from anyone who attends, and from readers better versed than me in copyright issues, or anyone with ideas for solutions…
In the meantime, don’t worry – all the images in this post are creative commons or copyright-expired, and I gave myself permission to use my (pseudo) anonymised photo as the featured image.