I am currently working with a startup that has to do with prescription drug refilling. Obviously, interactions between drugs and such can have fatal consequences if not understood, so we responsibly are providing that information to our users.
Elsevier provides an API for that data, but wanted to charge around $40k/year for access. Our business model is of course free to users (we charge the pharmacies a modest monthly fee)
$40k/yr simply too damn high...so potentially life-and-health saving information is kept from the people whom need it the most.
Yes we did eventually find a lower cost provider, but as far as shedding tears for Elsevier?
It's amazing the logic hoops their supporters go through. I can't find the site (it was linked from another journal article on HN) again right now, but I went to a journal trade site once and started reading the comments. The article was about open access, scihub, and how journals needs to be ready, but the comments I could only assume were paid shills.
There was one guy whose position was that journals needed to limit access because the layperson would not understand what they were reading. He was questioning a person employed in home care who admitted to use SciHub to read topics she was interested in because according to him she had no need for that knowledge. This is what open access to knowledge is up against.
I agree, some of the opinions tossed around by incumbents in the industry are mind-blowing. My favorites: "A PDF is a weapons-grade tool for piracy." ...And on parents of sick kids looking up articles: "Do you really think a search and downloading of technical medical literature means anything?"
And they wonder why NaturalNews and other anti-vaccination sites get so many hits. The arguments they are using are utterly indefensible.
They now have a real problem. They were able to be gatekeepers in the U.S. and other Western nations, but the instant one nation doesn't play ball and publishes all the material as it doesn't recognise international copyright... well, this has parallels to when India refused to recognise pharmaceutical patents. They had to rapidly change their business model to more favourable terms, because it suddenly (!!!) became clear to the pharmaceutical industry that this could destroy the value of their patents.
That's pretty much identical with the position of the Catholic Church for a very long time: laymen aren't able to understand the Bible so it should only be accessible to professionals.
I don't know if the Catholics changed their opinion, but I think it would be important for science to be more accessible :)
Copyright laws are an affront to the free market, since they don't allow selling what one wants or what the customert wants. Instead one can only sell what the copyright holder allows.
Copyright laws are a finite duration social and legal grant to the creators of intellectual property with the intention of encouraging creation of work.
It is artificial relative to a particular vision of a "free" market, but the same could be said about any of the mechanisms that structure the market, thereby allowing it to function. In order to have a market you must have some rules. So what we do is try to figure out what rules work well and which ones impede things.
Copyright/Patent/Trademark isn't a priori bad. It has funded many musicians, writers, inventors, painters, etc. It has often been referred to as one of the great innovations of Enlightenment era free-market practice. It allowed generations of thinkers to have compensation when previously the only form of wealth was goods and land backed up by physical violence.
This is a nuanced topic, and I think it makes sense to treat it as such.
> Copyright/Patent/Trademark isn't a priori bad. It has funded many musicians, writers, inventors, painters, etc.
There is no reason why one can't pay money even when copyright laws wouldn't exist. I often do, for exampl for open source projects or for free ebooks where the author still asks for a donation if one can afford. The only difference is that the creator cannot force you, which I consider as a good thing.
If you're trying to get paid for creative work, donations simply don't work as a model.
There's a fundamental problem for both the arts and sciences, which is that in general terms, knowledge and creativity do not have a market value.
You can't place a market value on a Mozart opera or the Theory of Relativity. You can't even place a market value on last week's most downloaded SoundCloud track. Or on the Linux kernel.
Copyright is an attempt to butt-join creativity with a market economy. As such, it almost works, and it's the least bad solution for the context.
But it's still the wrong answer. I'm not sure what a better answer would be, but it could be something like giving a person with unusual verifiable talent extra resources and time to pursue their interests.
Academia used to work like that, but academia is always highly politicised, both internally and externally, which adds friction and error bars.
If there was a way to create a similar system with less politics, that might work better.
Or not. It's a hard problem to solve. But the current marketisation of everything - and FOSS is still marketised in its own way, IMO - is making things worse, not better.
That's the patronage or tip jar model. It has also funded many people. But not as many as the copyright model.
To flip your force model around, you cannot force people to create for you. But the hope of future money does provide incentive to many of them. Robert Heinlein, for example, started writing in order to pay of his mortgage.
> To flip your force model around, you cannot force people to create for you. But the hope of future money does provide incentive to many of them.
But who says that this future money has to be enforced by copyright laws? It is also possible by either asking for donations (from which for example Tarn and Zach Adams, the creators of Dwarf Fortress, live) or something similar to Kickstarter/Indiegogo (the product won't be created/released until some money is donated).
So in other words: Even this incentive wouldn't be killed if there were no copyright laws.
No one is forcing you to read their works in the first place.
Nor is your example correct. The first sale doctrine still applies. If you own a legal copy then you do not need to give money to the copyright holder to sell the copy. If someone else has a copy, then you can buy it without sending any money to the copyright holder.
Granted, we appear to have collectively and foolishly decided to reflexively click on "I agree", and switch to on-line rentals instead of actual purchases, making this harder. But it is only illegal if you make a copy. So don't do that. Then there's no (data) highway robbery. Problem solved!
> But it is only illegal if you make a copy. So don't do that.
So in other words you argue against free speech (creating a copy) and pro censorship. Writing texts that obey the censorship laws (or in western countries copyright laws) is illegal about nowhere.
Wow. Really? Be sure to tell Stallman that the Free Software Foundation, based as it is on strong copyright, is really against free speech and is pro censorship.
First: The FSF just uses copyright as a means to an end. They would rather wish that their four freedoms were part of the law.
Second: I'm no advocate of the FSF way. For some reasons I think that the idea that one has to share the source code is dubious (among these reasons: what, for example, if I share the source code but obfuscate it; or if I change the source code so that it from now on is generated from some metacode: Do I have to share source code that is on the same level as the original one or do I have to share the metacode?). Instead in my opinion the community should develop much better methods of reverse engineering so that sharing the source code will not make much of a difference anymore. So in this sense the way to go is some kind of BSD-like license which also allows universal reverse engineering of built binaries. Such a license would also work if there were no/very weak copyright laws (since one can always do this, it just might be that the laws disallow it) in opposite the the GPL/AGPL.
Why do you continue to talk to me, a known censor and oppressor of free speech?
As for the FSF way, you aren't forced to share the code. That's one of the four freedoms you mentioned. It's only when you distribute something based on the code where you are forced (by the horrid anti-free speech Stallman) to also include the code through one of several mechanism.
If you do decide to distribute it, and your software incorporates someone else's work under the GPL, then there's already a clause which prohibits you from obfuscating the code. ("The source code for a work means the preferred form of the work for making modifications to it".) It's in v2 and v3.
Disassemblers are a thing. They are quite good these days. Let's assume perfect reverse engineering. Even with source code it's hard to reverse-engineer how a large project works.
Though not perfect. No one had yet reverse engineered Skype.
> universal reverse engineering of built binaries
Your complaints seems to be with contract law, not copyright law. I believe US and EU copyright already allows you to reverse engineer built binaries, with some restriction.
The problem is that people voluntarily agree to licenses which restrict that right.
40k/year sounds quite high for life-saving information. Life-saving information should be free (or, rather, whatever necessary costs it has should be socialized).
I think that's a great point--socialization is a solution that should be considered a realistic option. It seems like publishing would have a comparable bureaucratic load to that faced by science funding agencies.
It's a complex topic though. The papers are distinct from the research (and public money), and the authors willfully give publishers the rights to their writing. This is also not "free" but in exchange for prestige for promotions, etc. I don't think this is a good system though, but it's legal. NIH does require open access after one year for publicly funded work, at least, but I think a better solution is to require summary reports submitted directly to NIH for every project.
Even better would be to do away with reports and journal articles that only report a tiny fraction experimental data, and focus more on "selling" the science rather than reporting it. I suspect there is a very large overlap in research, thousands of animals are being killed needlessly, simply because the data that is not published in articles is discarded, even though they could be useful for other studies.
Suppose you needed 40k/yr of AWS computing time and couldn't afford it. And suppose someone hacked into AWS and redistributed it for free.
As far as I can tell, if the argument you made is correct, you should support the same argument that says the AWS hacker is right.
(To address the two obvious rebuttals, viz 1. Marginal cost and 2. Sci hub is getting data from those who have a subscription, merely in violation of ToS:
1. Marginal cost for AWS is likely low, if there's any unused CPU time.
2. Change the scenario slightly to someone signing up for a bunch of free AWS accounts and handing those out. )
What? no. Elsevier does not make any science, they just put a stamp on it and sell it. They would be more likely to make you pay $40/y for using their linux in someone else's server. Because you have to have "elsevier linux" to get grants these days.
If author wants his paper to be open access (something that is often required by their funding grants), they have to pay a high publication fee. Other journals charge paper submission (for consideration) fees as well.
But the error in your argument is at the start. You assume that the market is free and scientists have an alternative to publishing in elsevier. If scientists want their work public, there are tons of free options, preprint archives etc. The problem is that funding sources require high impact (i.e. elsevier) publications to give you grants. There is a self-perpetuating cycle between funding sources <-> (elsevier, NPG etc.; talking about biology). This is a racket.
PS. How it is different from buying servers? It would be comparable if , say, the US army had a requirement that all soldiers must do a yearly internship in amazon (paid by the army), thus amazon didnt need to pay any salaries, yet could charge as much as the competitors. It would all be legal, but grossly immoral.
We're not talking about open access here. I don't see why you brought it in.
From what I understand, most journals that charge for access offer free publication. Even if fees are charged, they don't cover costs.
>This is a racket.
Maybe if they were bribing grant committees to require Elsevier publication. If those committees decide that of their own free will, then they think the benefits of Elsevier are worth the costs.
If they really were just extracting rent, I'd expect to see far higher margins, at least above 50%.
> If those committees decide that of their own free will, then they think the benefits of Elsevier are worth the costs.
We don't know what's happening here, but when elsevier holds the keys to funding scientists, they are not going to turn against them. Elsevier can "bribe" without money in this case. Counter example: the mathematicians' boycott of elsevier. They could pull it off because their career does not depend on it. A biologists' boycott? never happened.
Elsevier can ask for all the money in the world and i don't care, the problem is that it's an abuse of copyright law to give them exclusive access without an alternative method for the public to see the results of publicly funded work. Why this kind of provision does not exist is not clear to me, but i suspect lobbying from publishers has played a role.
While you can dismiss it as a 'greedy business', elsevier plays an institutional role here, and it holds that role purely by inertia.
p.s. Yes , 30% is exorbitant profit, academia has different standards, and most scientists would prefer a non-profit.
>Elsevier can ask for all the money in the world and i don't care, the problem is that it's an abuse of copyright law to give them exclusive access without an alternative method for the public to see the results of publicly funded work.
As I've mentioned before, complaining about publically funded work misses the point. The funding only covered research, not publication. If you want it to be published, you need to pay the costs of that. CLaiming that you have rights to someone else's work (publication costs) because you paid for something else (research costs) is nonsense.
You can require the government to pay for publication costs as well (i.e. publish in open access), and then the natural consequence is that the cost of each study goes up, and therefore less get funded.
> most scientists would prefer a non-profit.
Then set one up. If they can outcompete Elsevier, more power to them. But that doesn't justify illegally stealing from them.
> The funding only covered research, not publication
Not true, many grants require publication. Any grant giving money to do research that doesnt require any results is fraud.
> Then set one up.
PLoS, frontiers, eLife, there are plenty and they are good, but until we can drag the critical mass of high-impact scientists into these open publishers, they will keep being underdogs. You need a systemic change here that will at once pull the good reviewers and good scientists to them, in other respects they are much better than elsevier in some technical respects. The problem is that academia is not the free market. The currency is 'prestige' and its proven very very hard to change attitudes without a) a generation dying or b) regulatory action or c) doing something illegal. It's regulations (specifically requiring high impact publications) that brought elsevier to become the arbiter of scientific importance after all. We either have to regulate paywalls out of academia completely, or find a way to replace the institutional role of publishing with something else.
In any case, your 'this is free market' approach is wrong here i think. Academia is not a free market.
I didn't say they don't require publication, only that they aren't paying for it.
Plenty of Elsevier titles offer open access for a fee now. And PLOS and co are putting pressure to reduce margins.
I don't think it's bad enough that something illegal needs to be done to force it to fix itself.
And I've asked before: if Elsevier slashed all their prices by 30%, would people be happy with it? If the answer is no, then they aren't only complaining about profit.
Grants typically include publishing costs, it's very rare the scientist will pay out of pocket. And because grant agencies encourage high impact publications it's not small amounts. Ironically, grants that require open access while encouraging high-impact publication end up paying more, as high-impact journals charge more for open access.
I agree that money is not the major issue, but the entrapment of useful information in closed systems and its decentralization. We have machine learning systems that could consume the vast amounts of data and come up with potentially life-saving results, but we don't have access to the data, because for 50+ years its been more and more locked in.
My understanding is that the publisher subsidizes publishing costs and makes it up on access charges. So if the grant didn't pay the cost to publish, the publisher will.
I agree that it's a nice ideal to have everything open, but people need to realize that there are costs to doing so, and merely saying "I paid for it, therefore I have a right to read it" doesn't do much.
>I agree that it's a nice ideal to have everything open, but people need to realize that there are costs to doing so, and merely saying "I paid for it, therefore I have a right to read it" doesn't do much.
You're right, that's why places like sci-hub exist, and Elsevier et al dont seem to be winning this fight…
> Grants typically include publishing costs, it's very rare the scientist will pay out of pocket.
Though for small labs just starting out, two or three open access publications written into a grant definitely could have gone to salary for someone. Even if it's written into a grant, it's not a trivial expense.
As a minor objection to your analysis, some of the papers they distribute are by employees of the US government. There is no copyright for those papers, therefore Elsevier could not get anything signed over to time.
Elsevier and all other journals charge the same price for access to these papers as they do for papers where the publisher controls the copyright. To a first approximation, that means the value of the copyright is not significant.
It's a minor objection because there are few papers which are in the public domain. The publishers use contracts to prevent or limit redistribution.
(I have wondered if a SciHub variant with only papers which are out of copyright or not under copyright would be viable, even on a US server. I invite someone else with money and lawyers to try it out.)
Anyway, to get back to your AWS server scenerio, the parallelism doesn't work that way. In the print publication era, multiple organizations functioned as an archive for the material, including the institution library, various university libraries, and state libraries. I have gone to the local(ish) chemistry library and read publications from the 1960s without paying extra costs to read the article or make copies. I can just walk in as a member of the public and do that.
However, it's more cost effective to centralize things at Elsevier (that savings to the library get passed along as profit to the publisher), Elsevier moved towards an electronic-only publication system mediated through user agreements instead of copyright, so the library cannot act as an archive, and that user agreement must extend to all users, including members of the public. (I have signed a paper to that effect, which lets me as a member of the public to access online publications but without the right to print them out or make copies.)
In your AWS analogy, Elsevier/Amazon once distributed their software to anyone, and there were dozens of places willing to provide $5K/year to host those services for users of the local cloud, including anonymous hackers. Elsevier/Amazon decided to centralize, by only hosting new services on their machines rather than distribute the software. They also decided to require verified logins. So now our anonymous hacker is forced to either give up doing science, pay potentially $1,000s for legit access, or use subterfuge to do what was once inexpensive or even free, as part of a library's mission to support the public.
But in real life AWS never distributed all of its code, so that others could provide AWS services, which is why your analogy doesn't really work as a metaphor for this issue.
"some of the papers they distribute are by employees of the US government. There is no copyright for those papers, therefore Elsevier could not get anything signed over to time.
Elsevier and all other journals charge the same price for access to these papers as they do for papers where the publisher controls the copyright. To a first approximation, that means the value of the copyright is not significant."
It also could mean that the market is broken. In a 'normal' free market, somebody would take the set of copy-right free publications, start selling access to it for half what the big players ask, and grab that part of the market. Part of the reason that doesn't happen is that the big publishers only sell bundles. You have to go there for some stuff, and to get that stuff, you get access to the rest, too. That decreases the portential value customers put on having access to only the copyright-free stuff.
Nevertheless, you should still be able to get those copyright-free texts cheaper elsewhere, for example from their authors, if they wish to provide that service.
There are four barriers to entry: 1) acquire the possible publications, 2) determine if it's in the public domain, and 3) face likely legal action design to suck away time and money, 4) come up with a good revenue model.
1) isn't easy because access to the the newest publications require a user agreement. I think it can be done by starting with SciHub since SciHub hasn't signed an agreement in the first place. For older publications, paper scans are possible, but also expensive.
2) is easiest in those rare cases where the paper says "this paper was authored by an employee of the United States and is not subject to copyright", or something similar. Otherwise, it might be inferred from the author's institutional affiliation ("William J. Wiswesser, US. Department of Agriculture, Frederick, Maryland 21701" to pull an example from one paper in my collection). This is labor intensive and therefore expensive.
3) While the US doesn't recognize a "sweat of the brow" doctrine, other place might. Even if a public domain archive were legal, it could be sued for, say, knowingly trying to subvert the policies in #1. (Also, will these documents have the same DOI as what the publisher uses? Might they sue for copying that in violation of some user agreement on their web site?)
4) Yeah, I've no clue how to make money off of this. Especially if SciHub is part of the competition.
"for example from their authors"
Wiswesser, above, died in 1989. I'm going through a lot of historical papers these days.
You can still go to a university and access papers for free. In fact, you can log on to their wifi anonymously, and often get access to those papers.
I'm not sure why the past actions of a company as you describe would matter for current ethical questions? If Amazon had previously allowed anyone to use it for free as you describe, then upgraded and stopped allowing it, would it then be morally permissible to use hacked accounts or similar to do your computing with?
"You can still go to a university and access papers for free"
Yes. I don't understand why you mentioned that. I even pointed out that I could do that.
"In fact, you can log on to their wifi anonymously"
No, I cannot. The local college and the 1-hour-away university do not have anonymous wifi. I have signed up for a library card for non-anonymous access to the local college. It does not carry postgraduate journals in chemistry. The university does, but I am only allowed to read them from dedicated terminals. I am not allowed to copy the articles and I do not have wifi access.
(Why do you think you know more about my local academic library policies than I do?)
BTW, I can make InterLibrary Loan requests if I really want a copy. I believe it's about US$10/copy.
"... would matter for current ethical questions"
I was attempting to demonstrate how your scenario was not a good metaphor for describing the situation with scientific publications, so could not be used to clarify the ethical issues. You can stretch the metaphor to try to compensate, as you did, but there are several other factor which are also not captured:
1) people get jobs and tenure based in part on the significance of their work. Part of that is often based on where one publishes, eg, "Cell", which is an Elsevier journal, is considered to be a prestigious place to publish. Thus, short-term personal goals of advertising one's work may be at odds with long-term community goals of easy access to scientific data. 2) peer review is almost entirely done on a volunteer basis.
Thus, to be useful, we would also have to posit that people's jobs depend in part on running their code on the Amazon cloud and no other cloud provider, and that most people are expected to volunteer their time to support AWS. Perhaps this is something that Bezos dreams about, but it is not the current reality.
This, your AWS metaphor concerns only a part of one facet of the issue. Ethics is very clear cut when there is tunnel vision.
It sounded like you were complaining that there's less access than there used to be. I seem to have misunderstood you, then.
Re your other points: AWS relies in part on open source software, which arguably corresponds to free peer review.
There's no direct analogue to prestige, but that's something the publisher arguably deserves payment for. Prestige doesn't come out of nowhere, it takes a long time to build up a reputation, etc.
No, it doesn't. They are remarkably different. If I distribute free package X, which Amazon uses, then there's nothing to prevent anyone else from using X. Both can benefit the same way.
If I spend 6 hours to do a peer review on paper X, which is then accepted, then Elsevier benefits financially, because they have a new paper to sell, but no other archive can benefit that way. There is also a gain in prestige to Elsevier from having published a higher quality paper.
If the paper is not accepted, then the reviewer gets feedback from my review, but no one else does. (I am supposed to benefit in the future from a favor-in-kind by peer reviewers.)
While it's true that much open source software is at no cost, I sell free software, exactly as described at https://www.gnu.org/philosophy/selling.html . While AWS might purchase said software from me, Elsevier/ACS/etc. will never purchase my peer review.
You can still go to a university and access papers for free.
Maybe you can, I certainly can't. Neither of the two universities near me allow the general public access to their journal database. Nor are the IP numbers used by for the public anonymous wifi white-listed for accessing journal databases.
Are either of those ones you'd have gone to back when they had physical libraries?
My point was only that if a university funded public access in the past, there's nothing preventing them from allowing it now as well, as long as you go there.
Sure they both have physical libraries and both let me go there and read books. But their licencing agreement with the journal database owners forbid them from giving out access to anyone who isn't a student or faculty.
My "argument", if I was even making one, was that the cost to Elsevier to provide an API, is a tiny fraction of $40k/yr, and thus their business practices and model is all about rent-seeking and not about saving peoples lives, which I personally find disgusting.
You realize that an API fronting service like Mashery alone would cost them several hundred thousand dollars ?
It's not something you hacked up in a weekend in your basement. It's something that requires project managers, business analysts, engineers, sales people, etc. These people require salaries, pensions, healthcare, etc.
You can of course build a competitor if you think you can build something for a fraction of the cost.
Fundamentally Elsevier's content is not funded or created by them. But they have a monopoly on it because most studies cannot be published in multiple places.
In the case of AWS market forces could undercut their margins.
That sort of copyright monopoly was granted by our legal system to encourage people to create intellectual property, but it is being twisted in this situation, as the creators are being funded by a mechanism completely irrelevant to Elsevier's costs.
In this case what they are doing is more similar in a sense to patent trolls.
>Fundamentally Elsevier's content is not funded or created by them.
They pay publication and editing costs, and in return get copyright signed over to them.
>market forces could undercut their margins.
And market forces can open new open access journals as well. To date, this hasn't solved it, perhaps because their margins weren't so high to begin with.
>patent trolls
Generally this refers to patents that should not have been issued, or that don't apply to the defending companies' product. It's not applicable here.
It is part of the fabric of scientific research that findings are shared. Science is overtly collaborative. Every scientists specifically builds on the knowledge of others, and shares what they discover with the world.
Governments and a variety of organizations fund research with the express purpose of discovering new knowledge that ultimately benefits everyone.
That's the WHOLE POINT of publishing a paper. At a fundamental level, scientific publishing is about trying to share the results of your research with the scientific community.
The funding is tied to the act of publishing, having nothing to do with how people consume the papers. Charging people above cost to read the published research is antithetical to the whole idea of how the world does science.
The example I gave above was where you needed the processing power to run your medical startup.
>There is obviously certain types of data that should be provided as a public service to all of society
"Should" what's meant by that? If you want data to be open, you need to fund that, or develop your own and release that freely. Statements of the kind "everyone deserves scarce resource X" really mean "we should tax rich people in order to subsidize X for poor people". Which is a fine thing to argue for, but don't present it as a universal right.
I am currently working with a startup that has to do with prescription drug refilling. Obviously, interactions between drugs and such can have fatal consequences if not understood, so we responsibly are providing that information to our users.
Elsevier provides an API for that data, but wanted to charge around $40k/year for access. Our business model is of course free to users (we charge the pharmacies a modest monthly fee)
$40k/yr simply too damn high...so potentially life-and-health saving information is kept from the people whom need it the most.
Yes we did eventually find a lower cost provider, but as far as shedding tears for Elsevier?
Hell fucking no.