On ReadCube, and Nature‘s give-away

December 9, 2014

It’s been a week since Nature announced what they are now calling “read-only sharing by subscribers” — a much more accurate title than the one they originally used on that piece, “Nature makes all articles free to view” [old link, which now redirects]. I didn’t want to leap straight in with a comment at the time, because this is a complex issue and I felt it better to give my thoughts time to percolate.

Meanwhile, other commentators have weighed in, and have mostly been pretty negative. John Wilbanks described it as “canonization of a system that says a small number of companies not only do control the world’s knowledge, but should control all the world’s knowledge”; Ross Mounce characterised it as “beggar access”; Peter Murray-Rust says “Nature’s fauxpen access leaves me very sad and very angry”. Perhaps surprisingly, Michael Eisen is more temperate, asking whether Nature’s policy is “a magnanimous gesture or a cynical ploy”, and concluding only “At the end of the day, this is a pretty cynical move”.

I am a bit more optimistic (although as you will see, still not really happy).

First of all, let’s say clearly that this is a step in a good direction. Nature‘s papers are now at least somewhat easier for regular people to get hold of, and that is to be applauded. Even if Mike Eisen’s cynical reading is correct, it’s still a net good.

But — and it’s a big but — I have a huge problem with the use of ReadCube, or any equivalent, to provide a crippled form of access. Rather than Ross’s term “beggar access”, which focusses on the need to get a subscriber to share a link, I think the best term to describe what Nature is offering here is “broken access”. Broken by deliberately locking the content into the ReadCube jail, to prevent printing, downloading, copy-pasting, etc.

My issue with this is two-fold: both practical and philosophical. Practically, PDFs are very far from perfect, but there’s a lot we can do with them (including printing, downloading, copy-pasting, etc.) Most crucially, when I download a PDF, I have it forever. I can refer back to it whenever I need it, without depending on a third party. It becomes part of my research toolkit. I know it’s not going to vanish when my back is turned.

By contrast, we never know when we’re going to be able to read these Nature papers. Certainly not when we’re offline. Maybe not when there’s a service outage. Probably not after the end of the one-year pilot. And you can’t build research on something that you can’t rely on existing. It’s not real.

But the philosophical issue is really burns is that ReadCube exists precisely in order to take away functionality. Its purpose is to make access limited, ephemeral, unreliable and less useful. And I find that offensive. The idea of doing work to remove functionality hurts me. The idea of all those clever people doing all that hard work to take functionality away. It’s wrong. It’s burning value.

So I end up feeling conflicted about the new Nature policy. It is a forward step; but one that I literally don’t ever see myself taking advantage of. A much more useful policy (to me anyway) would be to keep new articles under lock and key, but make them truly open after, say, a year. Because for a scientist, usefulness trumps timeliness.

Finally, Matt makes this point:

Nature papers are short, typically 5 pages or fewer. With big, modern monitors, you can usually get away with screen-shotting a whole page in one go, or in two takes and the world’s easiest GIMP stitch at worst. So by not allowing people to download the PDFs, all they’ve done is ensure that the people who really need their own offline copy will have to waste maybe 15 minutes assembling one. So the ‘barrier’ they’ve put up is low and crossable, it’s just annoying. Is that what Nature wants to be known for, annoying their users to death?
It’s a real issue. For many people, their main impression of Nature is now going to be “that journal where you have the read the papers in that weird UI where you can’t do anything”.
Advertisements

22 Responses to “On ReadCube, and Nature‘s give-away”

  1. David Marjanović Says:

    Here’s a thread about this. I’m linking directly to the first comment (number 9) that comments on how easy the “protection” is to crack; further such comments follow till number 43.


  2. That’s right, Nature’s “read-only sharing by subscribers” is not OA, but this is not the point. I see no discussion about the fact that it is compliant with the new Open Definition v2.0. I find this extremely puzzling. In my opinion this shows a big weakness in this open definition. The weak point seems to lie here:

    “2.2.6 Technical Restriction Prohibition. The license may prohibit distribution of the work in a manner where technical measures impose restrictions on the exercise of otherwise allowed rights.”

    “May”, not “must”!
    All over the place when it is about the author of data there are
    plenty of “must”.

    All this Nature story revolves around DRM. There has been a discussion here in the comments, where this weakness has been pointed and apparently dismissed with no argument.

    So I ask here publicly: is the material released through Nature’s Readcube scheme Open Data or Open Content?

    Would be glad to be corrected.

  3. Mike Taylor Says:

    Well, that is ridiculous. Then the Open Definition v2.0 is useless. Completely useless.

    Was v1.0 the same?


  4. Compare with the very clear CC 4.0 licence: in Section 2/Scope/a. Licence grant/5.2
    “No downstream restrictions. You may not offer or impose any additional or different terms or conditions on, or apply any Effective Technological Measures to, the Licensed Material if doing so restricts exercise of the Licensed Rights by any recipient of the Licensed Material.”
    That is, I believe, the motivation for the middle part of the article Content sharing is *not* open access and why NPG is committed to both:
    “To many, “open access” means licensed under a Creative Commons Attribution CC-BY license. We offer CC-BY on all the open access journals and hybrid journals we own, and the majority of titles we publish on behalf of societies.
    […]
    Authors can opt for a Creative Commons Attribution license on *all* our fully open access journals, and all NPG-owned hybrid journals
    […]
    Content sharing is quite different – the 49 journals participating in this initiative are mostly available only to subscribers. We are offering something additional for subscribers at no additional cost, as an experiment, to learn how to meet the needs of researchers better.”

    There seems to be a tension between Open Definition and CC, I don’t know why, nor I take any part. But my initial question was why not include in the Open definition a text as clear as the one quoted from the CC licenses?

    If anybody cares, see also Beggar Access: use DRM for academics, enjoy wathcing them fight for the copyright.

  5. Aaron Wolf Says:

    This is a complete misunderstanding of the Open Definition 2.0. The Nature release does not meet the definition, period.

    “An open work **must** satisfy the following requirements in its distribution:”

    and “The work must be provided in a convenient and modifiable form such that there are no unnecessary technological obstacles to the performance of the licensed rights.”

    “The license must allow free use of the licensed work. … The license must allow redistribution of the licensed work …The license must allow the creation of derivatives of the licensed work and allow the distribution of such derivatives … The license must allow any part of the work to be freely used, distributed, or modified separately from any other part of the work … The license must not discriminate against any person or group. … The license must allow use, redistribution, modification, and compilation for any purpose.”

    GO READ THE DEFINITION. http://opendefinition.org/od/

    chorasimilarity is completely wrong. And Mike Taylor, you should not just take his word for it.

    2.2.6 does NOT say that a work can have DRM restrictions and still be Open. It does not say that. You are completely misunderstanding. Section 2.2.6 says that the license may or may not require that *derivatives* are DRM-free. In other words, one license can say “this is Open but you can make non-Open derivatives with DRM” while another can say “you must keep this Open, no DRM for any derivatives allowed”. But in BOTH cases, the work is only Open when there’s no DRM. It’s just a matter of whether DRM is further prohibited for derivate works. The “may” part refers only to whether the Openness is required to be maintained for all derivatives. If a license *allows* derivatives to have added DRM, that still means any such derivatives will no longer themselves be Open by the definition.


  6. OD 2.0 section 2 pertains to the license. A compliant license may prohibit DRM. A must here would mean an unconditional license or public domain would be non-open!

    OD 2.0 section 1 pertains to the work. 1.3 says “The work must be provided in a convenient and modifiable form such that there are no unnecessary technological obstacles to the performance of the licensed rights.”

    A work with DRM applied to it is not an OD compliant work whether the work’s license prohibits DRM or not.

    I can’t tell where “here in the comments” is intended to link to above, but if similar misreadings of the Open Definition are occurring elsewhere, I’d love to see them corrected, and to get suggestions as to make the OD easier to read and understand correctly.

    See http://opendefinition.org/od/ for the current version and feel free to file issues at https://github.com/okfn/opendefinition/issues

  7. the bear Says:

    I am on the OD committee and have raised this on the public od-dscuss list https://lists.okfn.org/pipermail/od-discuss/2014-December/001182.html. Read the following discussion. There has been a complete misunderstanding and – the simple sense is that applying DRM does not stop something being Open. The committee have agreed they will look into the cause of the misunderstanding and possibly reword (but not change the meaning) in a future revision.

    Please rest assured that OD 2.0 is still vitally valuable

  8. Mike Taylor Says:

    Thanks for chipping in, Aaron — it’s very good to know that the intention of the Open Definition is not what chorasimilarity read it is. That said:

    2.2.6 does NOT say that a work can have DRM restrictions and still be Open. It does not say that. You are completely misunderstanding.

    I’m afraid it does say that. Again, great to know this was not the intention, but reading the actual words on the screen, that’s what it says.

    Section 2.2.6 says that the license may or may not require that *derivatives* are DRM-free. In other words, one license can say “this is Open but you can make non-Open derivatives with DRM” while another can say “you must keep this Open, no DRM for any derivatives allowed”.

    OK, so the goal is to allow both copyleft licences, such as CC By-SA, and non-copyleft licences, such as CC By. That goal is good and sensible (although I’d have thought clause 2.2.6 is unnecessary given that 2.2.3 covers it); but again, it’s not what the text actually says, because it doesn’t mention derivatives.

    Looks to me like a small wording change will fix this.

  9. Mike Taylor Says:

    And thanks to Mike and Peter for further clarifications. Sorry that your comments, and Aaron’s, were held in moderation — it’s probably because of the links, which the spam-filter is sensitive about. Now that you’ve had comments OKd on here once, subsequent comments should go straight through.

  10. Mike Taylor Says:

    Peter, what is the intended distinction between “must”, “must” and “shall” in the Open Definition? e.g. sections 1, 1.1 and 1.2 begin “An open work must …” (no italics), “The work must …” (in italics) and “the work shall”.

  11. Mike Taylor Says:

    I just read the comment thread that Peter linked to. Very clear and helpful: the fixes proposed there all look good to me.

    One other suggestion for any Open Definition people who read this: Mike Linksvayer asks “does whatever ReadCube does constitute an “effective technical measure”?” In practice, due to laws like the DMCA with its anti-circumvention clause, any technological measure is “effective” — not technically but legally. That could even apply to something as trivial as the ROT13 cipher. So I think there is no need to make a distinction between effective and ineffective technological measures.

  12. Aaron Wolf Says:

    Mike, I see you have recognized a parsing problem, but you have insisted on your own reading as the only one. But it is indeed a serious source of error and confusion.

    Section 2.2.6 verbatim: “The license may prohibit distribution of the work in a manner where technical measures impose restrictions on the exercise of otherwise allowed rights.”

    This was written as:

    The license may prohibit ( distribution of the work in a manner where technical measures impose restrictions…)

    in other words: What may the license prohibit? Distribution with DRM is something it may prohibit.

    You read the same text as: The license may (prohibit distribution of the work…)

    Which is confusing and contradictory to everything else stated in the definition. Any reasoned debate would recognize your reading as an error, but it is fair to blame it on bad wording that is open to this parsing error. Thus, we must indeed fix the wording so your reading is not possible.

  13. Mike Taylor Says:

    That all sounds right to me, Aaron. Thanks for your gracious response.


  14. Thank you for the replies, especially for those non emotional ones.

    I understand the intentions behind the point 2.2.6 of the OD: there are licences for works which don’t have any DRM restrictions, my example is the one used by arXiv.

    A “must” in the point 2.2.6 of the OD would imply that articles from arXiv.org are not open. Mind that the licence used by arXiv is also non-exclusive, which solves the theoretical danger related to any conceivable future DRM clauses added on top.

    On the other side, consider the case of a work with a licence which satisfies the all the points of the OD as it is now, but which is under a DRM lock.

    The example of Nature shows that is possible. Btw, somebody from the OD would care to explain to Mike Taylor which paragraph from the OD is not satisfied by works available through Readcube. (irony here)

    “Thus, we must indeed fix the wording so your reading is not possible.”

    Yes, no need to blame readers for that.

  15. Mike Taylor Says:

    chorasimilarity,

    The NPG articles on ReadCube fail to satisfy the Open Definition because of this clause:

    1.3 Open Format

    The work must be provided in a convenient and modifiable form such that there are no unnecessary technological obstacles to the performance of the licensed rights. Specifically, data should be machine-readable, available in bulk, and provided in an open format (i.e., a format with a freely available published specification which places no restrictions, monetary or otherwise, upon its use) or, at the very least, can be processed with at least one free/libre/open-source software tool.


  16. Thank you, I know that. In that paragraph the catch is “unnecessary”, and also, at the end, ” at the very least, can be processed with at least one free/libre/open-source software tool”. How would a lawyer read “processed”? If I look at Readcube using tools from a linux distro then is OK, right?

    The real problem behind all that is the following: it is not sufficient to classify as open or non-open as a function(licence you use) solely (although there are ones better than others). DRM is a lock which is independent on that (unless the licence prohibits DRM).

    Maybe is puzzling why I don’t have anything to say against Nature. Why should I? They are a business and they have their ways. They are perhaps ones of the fey publishing businesses which deliver something to the authors and readers, through their editorial process.

    On the other side, I don’t see why the academia should facilitate the life of legacy publishers.

    I wish not for one OD, but for many. Let the author bargain with the publisher to pick one OD which fits (and add that OD label to it) and let the readers decide if the work is open based on the publisher point of view and also based on the definition of open provided.

    This way there will always be a pressure on the OD providers to improve their work. There will always be a pressure on the publisher to not try PR stunts.

  17. Mike Taylor Says:

    “I wish not for one OD, but for many. Let the author bargain with the publisher to pick one OD which fits.”

    Surely you mean you wish for many open licences? Having multiple definitions is a recipe for confusion.


  18. No, the licence is not the whole picture. I’m OK with CC-BY 4.0 and I’m OK also with arXiv non-exclusive licence, because both provide (different) ways to avoid DRM locks.

    Having multiple definitions is not confusing as long as the work comes with a link to the definition.

    What can happen? Suppose you have 100 definitions of what is open. Some of them consider only licences, some other consider DRM, and some other consider the race of the author, color of hair, or the English accent of the author. Who cares? People will have opinions about these, and time will filter better definitions, in ways we can’t forecast. Publishers will always try to convince readers they are good (because this increases their $ income) and being open is more and more important. So they will always try to tweak the meaning of open. Having multiple definitions is an adaptability trait, against these predators.

  19. Mike Taylor Says:

    Having multiple definitions is confusing, always. I am surprised that this assertion is controversial. When you have multiple definitions, you have no definition.


  20. So many open definitions coming by,
    That’s so confusing!
    Next time I better use a non-exclusive
    Or even better I should ask for CC-BY.


  21. […] I wrote last week that I can’t support Nature’s new broken-access initiative for two reasons: practically, I can’t rely on it; and philosophically I can’t abide work being done to reduce utility. […]

  22. McBlawg Says:

    […] On ReadCube, and Nature‘s give-away – Mike Taylor […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: