> Based on Journal X’s practices, my photographs would be
> isolated from the paper, uploaded to Wikimedia Commons,
> and available for corporations who normally pay for my
> images to get them as freebies.
As noted in the article, the journal doesn't have the right to change the license on someone else's work. This is an ongoing problem with liberal copyright licenses in general; I often receive emails from people who ask me to release my software under MIT or BSD3 rather than GPL so they can "relicense" it, and sometimes even several back-and-forth emails are insufficient to convince them that copyright licenses aren't mutable by anyone but the owner.
For the author of this piece, I think the solution is relatively straightforward:
1. Point out to the journal that having a photo in one of their articles does not grant automatic permission to put that photo on Wikimedia. If the journal's software is unable to handle these cases separately, then the software should be corrected.
2. Ask Wikimedia kindly to remove the author's photos from their collection, or at least correct the license metadata. I'm sure the Wikimedia editors would be willing to do this, though they might become unhappy with the journal editor who uploaded photos without permission.
3. If someone uses the author's photos for commercial purposes, contact them and let them know that such use require a commercial license. They will likely be uncooperative (c.f. the various newspapers who like to source uncredited photos from Twitter), but some gentle reminders about copyright infringement's RIAA-engorged penalties should bring them around. If nothing else, they will likely become much stricter about validating ownership before using a photo.
> that copyright licenses aren't mutable by anyone but the owner
Well, sort of. Certainly if you've released something under CC0 or WTFPL, any derivative work can be "relicensed" however they want. MIT and BSD and CC-BY have some fairly trivial conditions that must be respected and fulfilled, but otherwise a derivative work can still use a different license.
I don't think many people care about the particulars of open source software licenses except in the context of producing derivative works.
At most journals, the authors of the article must certify that they have the rights to the whole manuscript and agree to relicense it. And I imagine that many biomedical journals wouldn't see the benefit of making an exception for this unusual edge case, or in spending the money to write new software to do so.
> I often receive emails from people who ask me to release my software under MIT or BSD3 rather than GPL so they can "relicense" it,
That makes no sense. People are asking GPL software to be relicensed under BSD/MIT to make them usable in situations where the GPL is a legal hassle or impossible to use. I doubt that people even consider relicensing BSD as something else (besides people practically downgrading BSD to GPL).
1. User wants to use my GPL'd library in their application. Their application is released as source code under the 3-clause BSD license.
2. User sends me an email asking me to change my library's license to MIT or 3-clause BSD so their open-source app can use it.
3. I reply with the longform equivalent of "u wot m8?"
4. They reply that if I change my license to MIT, they will be able to "relicense" it to 3-clause BSD and use it as a dependency in their application's build script.
5. I explain that they don't need permission to just depend on a library, since they're not distributing any of it. And besides, if they did include some of my code in their tarballs, they still can't change the license because they don't own the copyright.
I suspect this stems from the same sort of pseudo-religious copyright knowledge that leads to "no copyright intended" disclaimers on YouTube. Lots of people grew up in the era of Napster and copy-pasted MySpace javascript, and never really gave much thought to how copyright actually works.
Presumably, User wants to include your library in their open-source BSD release. If that is the case, what they're asking is correct since a work extending / modifying a GPL licensed code needs to be released as GPL as well. That is the main difference between GPL and non-copyleft licenses.
They could distribute their code without your library, as you say, but that would require more hassle from the end-user, especially if there's some configuration / integration process that is required to make both work together.
My libraries are mostly in Haskell, so depending on them is a matter of adding a line to a standardized build file. I have not yet received mail from a user asking me to change a license because they wanted to actually include code in their works -- it's almost always a misunderstanding of copyright.
Presumably what they want to do is mix the BSD code with their own code and release the whole under a restrictive licence. That's permissible under the BSD licence, but doesn't involve relicencing the BSD code, since that's still available under the BSD licence.
The journal in this case is presumably licensed under a CC licence that permits commercial reuse and derivatives. In that case, they do need the same (or less restrictive) licence on each image that they want to include. Fair use won't be sufficient.
While I doubt they would, in a case like that, they could have included a link to the image instead of the image itself. This is less than ideal in a lot of ways, but it would appear to solve most of the issues.
The journal wants to ensure that anyone can read the articles and all supporting information so they insist the author of the piece allow them to publish it under a Creative Commons license. The author would like to use one of this photographer's pictures and requests the permission. The photographer doesn't want to allow that. He (the photographer) suggests "I'd be happy to let you could use it in the article as long as you won't let anyone else copy it", but that's not OK with the journal. So the picture doesn't get used.
All sounds right to me. That's how it's supposed to work.
This doesn't seem like a collision of fair use and creative commons so much as a simple refusal to use fair use in the first place. The publications have decided that fair use is not sufficient for their needs, requiring instead a more explicit licensing agreement that fits their needs better.
The author even brings this up "This is not a problem with Creative Commons per se, but a structural issue with publishers that are inflexible in how they handle content." Then IMMEDIATELY continues with "I thought I’d share it here, though, as a counter-intuitive example of how Creative Commons can suppress the distribution of information."
EXCEPT this is NOT an example (counter intuitive, or otherwise) of how CC can suppress distribution. Rather this is an example of publishers STEALING images.
As the author pointed out, if the paper uses copyrighted photos with permission, or by fair use, the publisher of the paper CAN NOT just apply CC licensing to those photos. This is not a problem with CC this is a problem with the publisher.
From what I read in the post I don't think the publisher is stealing the images so much as rejecting them if they don't meet the license requirements.
While this paper illustrates an unfortunate side-effect, it's really no different to the Linux kernel rejecting code that isn't under a GPL2-compatible license.
This is almost identical to the issue of open-source license conflicts, where everyone is OK with their code being freely downloaded / compiled / modified by users, but not necessarily with the terms under which it can then be redistributed by those users.
The author says this is for logistical reasons, but I can imagine some principled arguments for the journals having such a policy too. The journal _is_ more useful for everyone if it can also serve as a source of reusable scientific images. In the ideal case, the authors would notice that they could not use a copyrighted photo of an ant, so they would go out and take their own photo, or pay a photographer to do so. If they just used ant photos under fair use, then nobody would have an incentive to create new CC-licensed ones.
author wants to "have his cake and eat it too", and argues from a position of untenable entitlement. As a private photographer, I cannot afford it. But you, sir, are not entitled to have someone distribute your work on your behalf.
Perhaps the photographer should consider changing his or her business model wherein he or she gets paid upfront on a commissioned basis to take interesting photographs, and release everything freely (with the concomitant distribution benefits).
For me, I'd like people were compensated fairly for their time and investment, but only once. Then, we will not have to think about this stupid dilemmas. I wish copyright did not exist, or for it to be really short. I prefer the former, as in selling a service.
> Printing images as natural history data in a scholarly publication should be considered Fair Use.
The author's entire argument rests on this single sentence. And, so far as I know, it is incorrect.
[I think that by "should be" the author is really saying "is" in a flowery way: otherwise his entire article is a nonsense hypothetical rather than, as he says, something which happened in reality.]
Publication is not straightforward. For example, if I publish an article in a Springer journal, Springer owns the copyright on it. They, not I (the scholar), have published it and are making money off of it. It is my experience in publishing scholarly work that inclusion of other figures is often considered to be copyright violation and not fair use.
And let's say that he's self-published, or used some service where he owns the copyright. Even then, he can't change the license on something just because of "fair use". He is free to publish his article under Creative Commons, or license the figure, but he can't do both.
This phenomenon of a "permissions mismatch" is common enough with various open-source software licenses that it has its own name: license incompatibility [1].
the op must clearly state fair use and forbid re-licensing to the magazine and the magazine should not relicense (scenario 1). But i suspect, op doesn't do it because of the fear the magazine will simply skip media with such restrictions (scenario 2). So op allows fair use, and the magazine re-licenses wrongly (scenario 3) [edit, or the op may license her work in cc- scenario 4)].
imho, it makes more sense for OP to use this fair use permission as a tool for generating potential interest for his work, and watch the magazines re-license them wrongly (scenario 3). if sufficient interest shows up, at least some new users may accept to pay for this media at the end. it all boils down to a conversion funnel analysis to decide which scenario makes more sense. ranting about it in a blog may also increase conversions.
There is no "collision" here whatsoever. Some people might like to distribute a photographers work under a Creative Commons licence. The photographer does not want to permit this. That's the whole thing. The fact that fair use exists has nothing to do with this.
For the author of this piece, I think the solution is relatively straightforward:
1. Point out to the journal that having a photo in one of their articles does not grant automatic permission to put that photo on Wikimedia. If the journal's software is unable to handle these cases separately, then the software should be corrected.
2. Ask Wikimedia kindly to remove the author's photos from their collection, or at least correct the license metadata. I'm sure the Wikimedia editors would be willing to do this, though they might become unhappy with the journal editor who uploaded photos without permission.
3. If someone uses the author's photos for commercial purposes, contact them and let them know that such use require a commercial license. They will likely be uncooperative (c.f. the various newspapers who like to source uncredited photos from Twitter), but some gentle reminders about copyright infringement's RIAA-engorged penalties should bring them around. If nothing else, they will likely become much stricter about validating ownership before using a photo.