Hacker News new | past | comments | ask | show | jobs | submit login
Science’s Pirate Queen (theverge.com)
496 points by IntronExon on Feb 8, 2018 | hide | past | web | favorite | 251 comments



To me, Alexandra Elbakyan is kind of like a Martin Luther of science. That is, she enables the lay public to read and interpret science rather than awaiting the interpretation and approval of academics from high-status institutions. That may be a ‘sacrilege’ view here, given the audience of HN. And, in a way, I do feel somewhat nervous of people making decisions based on papers they read without understanding of the philosophy and practice of science. Yet, it’s unavoidable and already being done through popular journalism.

With sci-hub (and with some reference books obtained via libgen), science becomes accessible to everyone with basic scientific literacy - a surprising large number of people. I see these two tools together as enablers of a new generation of scientists. For instance, recently I have seen many of my collegues and friends sending me direct links to research papers. That’s crazy and would have never happened just a few years ago. These tools will drive the democratization of science. Just as I now see the religious kids at the coffee shop discussing their scripture, I hope to soon see science kids there discussing classic science papers - doing it not for economic or career reasons, but simply to understand and discover.

To me, that’s not an economic issue but one of morality.


>she enables the lay public to read and interpret science rather than awaiting the interpretation and approval of academics from high-status institutions

>science becomes accessible to everyone with basic scientific literacy

The thing is most research papers are simply not accessible to neophytes; you often need to dedicate a significant portion of your life to the study of a given field in order to understand and critically evaluate the latest research in that field. It's not that academics are sitting in a high throne, dictating what is truth and what is not, its that you need to be an academic with many years of formation in order to understand, let alone improve, the research that is being undertaken. A "basic science literacy" isn't enough.

Still, it's very positive that, in principle, anyone can grab a book or three of libgen, study hard, become knowledgeable in the field, then grab as many papers as they want off libgen and read and interpret them. That's undoubtedly very positive.


I would argue that most people who studied an empirical discipline in college and did ok (good school and/or good grades) can pick up a paper and, using the internet, understand and critique it creatively within three weeks. The basic idea of science (minus a little nuance) is "I'll believe it when I see it, and I can show that it can't be explained otherwise." This isn't too much to think about.

If a paper doesn't link back to its foundations directly (citations) and clarify its assumptions, that isn't the fault of the reader.

One really dumb thing that many academics do is, in an effort to be respected in their field or accepted to a solid paper or a good conference, they write obscurely. With very few exceptions, this results in bad science writing because the point of science writing is to express clearly and precisely the steps and formulations that lead to an ideological or project accomplishment. But that bad writing doesn't make the paper less intelligible for normal people. It simply makes it take longer to read, for a typical person.

One of the reasons it's easier to understand research than to do it, is that it's easier to verify a solution than to create it. This is the famous Columbus' Egg, and it is a principle that should be well known to computer scientists who understand the time complexity of verification versus solution.

So, I argue that basic science literacy is enough to understand and critique. Verification of an idea is different than creating it. Papers that don't reference how they fit into their field are flawed in a way that can't be blamed on the reader. Obscure, bad writing doesn't change the meaning of a paper, even if it hides it. And the basic tool of science is a pretty accessible idea.


Would you care to put that argument to a test?

Because it appear to me that papers like (picking one from the front page of cell.com) http://www.cell.com/molecular-cell/fulltext/S1097-2765(18)30... "A Metabolic Basis for Endothelial-to-Mesenchymal Transition" require a lot of specialized knowledge to understand.

The same for https://arxiv.org/abs/1802.02935 "Suppression and revival of long-range ferromagnetic order in the multiorbital Fermi-Hubbard model", picked from a recent arxiv.org preprint in condensed matter physics.

Or "High-Content Surface and Total Expression siRNA Kinase Library Screen with VX-809 Treatment Reveals Kinase Targets that Enhance F508del-CFTR Rescue" from https://pubs.acs.org/doi/10.1021/acs.molpharmaceut.7b00928 .

What does "critique it creatively" mean?

I remember how long it took me, as a young smithling in graduate school, to read a scientific paper, in the field I was entering, where I had several years of education. It took a while to understand the literature, and even longer to be able to critique the literature in any more than superficial way.

I don't follow your comment "If a paper doesn't link back to its foundations directly". Papers are often like a skyscraper. The 50th floor links back to the 49th, which links back to the 48th. There are support structures which might span a few dozen floors, and (to stretch the analogy), express elevators connecting it to the ground floor.

But few papers really reference the foundations, because the audience is expected to know the foundations.


Are these characteristic papers, or the densest ones you could find? I submit that the vast majority of papers can be understood by people who have had about 4 years of university, and another 3-4 years of post-graduate education. Do you really believe that cobbling some portion of this study together over a lifetime is not even possible, or maybe even something that would be increasingly common, with access to the literature?

People have endless time to become experts on quackery and pop science; wouldn't it be nice to start seeing people who have taken up actual science as a hobby, and take it seriously enough to hold themselves to the same standards as professionals - or as seriously as the 20-somethings that are often writing these things?

Also, plenty of people get professional training, and end up working outside of the field. What if they want to continue learning? Is it somehow physically impossible for them to understand journal articles if they are not connected to an institution with journal subscriptions, even with training?


They were not the densest ones I could find. Not by far. I'm surprised you even thought of it. How many science articles not in your field have you read?

Take a look at the other articles in the current issue of Cell, at http://www.cell.com/cell/current . (This is one of the journals with the highest impact factors, so is less likely to consider the deliberately opaque language you mention.) You'll see the one I gave was not even the densest in that issue. Here are the titles to some of the others:

"Fast-Spiking Interneurons Supply Feedforward Control of Bursting, Calcium, and Plasticity for Efficient Learning", "In Situ Structure of Neuronal C9orf72 Poly-GA Aggregates Reveals Proteasome Recruitment", "5-HT2C Receptor Structures Reveal the Structural Basis of GPCR Polypharmacology", "Abstract Image Multiscale Structuring of the E. coli Chromosome by Nucleoid-Associated and Condensin Proteins", or "KRAS Dimerization Impacts MEK Inhibitor Sensitivity and Oncogenic Activity of Mutant KRAS".

I submit that your original statement, that "most people who studied an empirical discipline in college and did ok (good school and/or good grades) can pick up a paper and, using the internet, understand and critique it creatively within three weeks." is woefully undervaluing the expertise needed for each field.

I have 3 years of post-graduate education in physics. I can't read a chemistry paper for the life of me. I never even took organic chemistry. I find the idea that I can critique a paper on, say, some new reaction, with only a few weeks of internet-based research, to be ludicrous.

I note that you appear to have changed your statement. Originally it was just college, and now it's "another 3-4 years of post-graduate education." Originally it was "creatively within three weeks" and now it's "understand journal articles" with no time limit.

It feels like you think my objection to your original claim is representative of some absolute stance of mine against amateurism. As an amateur historian, I do not see how you have made that shift, nor do hold that opinion.

I don't understand the relevancy of your second paragraph at all. People get into quackery and pop science because it seems easy, and it comes with the built-in belief that experts are wrong and can be ignored.

If anything, it is the quacks and fringe science people who believe that three weeks of research is enough for them to be able to critique the standard views on evolution, global warming, chemtrails, pyramid construction, shape of the earth, etc., and critique to such a level that they should be listened to.

Nor do I understand the relevancy of your third paragraph. Could you clarify how you drew that inference from what I wrote? I am not connected to an institution with journal subscriptions yet I read and understand journal articles (acquired through ILL, links from Google Scholar, and yes, even Sci-Hub), so obviously I think it's physically possible for that to happen.

Are you sure you haven't just misunderstood my previous comment?


You aren't wrong, but I am not sure where you are going with this?


I was expressing a disagreement with the argument originaly made by totalZero and expanded upon by pessimizer.

I'm pretty much going where most HN comment threads go - nowhere.

Where are you going by making a metacomment?


> What does "critique it creatively" mean?

I assume it means to critique it using modern interpretive dance.



I suggest a key component of being able to "critique it creatively" is understanding the paper in the context of other work in the immediate sub-discipline, and the wider discipline.

I suspect there are insufficient focused waking hours in a three week period for any old university graduate to achieve that.

If I branch out into a nearby sub-discipline I end up spending a lot of time reading a lot of papers just to get sufficient context to make sense of a single paper, and that is starting from a base of knowledge of the broader discipline accumulated by someone several years post-PhD.

Even for practicing scientists, non-superficial critique of anything outside of ones immediate area of expertise takes work.

To work efficiently in your own sub-discipline, you have to be at the point where you can see references to Smith & Jones (1992) and implicitly know what that paper is about, and why it is referenced. That takes experience (you have to have read a lot of papers so you are already familiar with a reasonable portion of the references in a given paper) and you have to be practicing, (or at least I have to), otherwise you start to forget.


>> The thing is most research papers are simply not accessible to neophytes

To most neophytes that is. We only need a few of those with recently gained access to turn out to be a dormant Newton or Einstein. Society as a whole gets a better bargain.


Indeed! And the more neophytes that take the plunge and pull themselves up by their own bootstraps, the more neophytes there will be to encourage and enable other neophytes to do the same (network effects and whatnot).


I agree with your sentiment 100%, but "pull themselves up by their own bootstraps" is a phrase that came into existence to MOCK the idea that you can do it all on your own. It's obviously absurd, until it got so common that we use it without thinking. If you need a fresh quip to get the point try: "that poor mother can get by drinking her own milk!"


>The thing is most research papers are simply not accessible to neophytes

That very much depends on how you define "neophyte" and what field you're talking about. You don't need to be a doctor to make sense of most clinical trials, you just need a reasonable education in statistics. A mechanical engineer can usefully learn from a lot of materials science or fluid dynamics research. Our society is increasingly well-educated, which creates a huge number of people who can make use of sci-hub in one way or another.

If you're writing papers that are of no use whatsoever to anyone outside of a tiny coterie of specialists in your field, then either you're doing absolutely cutting-edge theoretical work or you're just really bad at writing papers.


That is a good point, but it's also true that you don't know what you don't know.

I've learned a lot about things I never thought I would be interested in by searching for terms and maths that I saw in papers and didn't understand.

It's not like I've become an expert in anything new, but it gets me thinking about all kinds of interesting things. It makes me want to go back to school, and reinforces the idea that I could create more if I did.

So I know that not everyone sees it this way, but I think that gating the means to easily access that sort of aspirational hope and inspiration is reprehensible.


> The thing is most research papers are simply not accessible to neophytes

The other thing is that it's not just neophytes without access. Heck, even in-academia researchers often have insufficient access (I recently spoke to someone at Nancy, France, which is wealthy enough), let alone the droves of former academics.


The argument that people should be prevented from reading information that they may not understand bugs me..


It's not just about laymen. There are quite a few PhD-holders working outside academia. Many of them cannot even access their own papers without paying ridiculous fees.


Even within; between funding sources or position switches, you might lose access to your institution's library, or your institution might be small with significant gaps in their subscriptions.

Ultimately, Sci-hub is often an easier and more pleasant way to get papers you have legitimate access to. At this point, I don't think there's anyone in my department that doesn't use Sci-hub from time to time. It's basically expected that anyone on filing fees for dissertation writing will use sci-hub rather than go through the hassle of requesting a special library access code (since they're no longer 'paying/earning' tuition).

There's no going back.


And even within academia. Big universities can afford to subscribe to thousands of journals; small universities and research institutes can't. When I worked at a research institute I would routinely encounter paywalls trying to access papers.


"Basic scientific literacy" is not enough to read most papers (in part because most papers are badly written and need familiarity with the genre to be understood). Almost everyone that I know that uses sci-hub does so because it's faster than finding the correct institutional proxy, repository, login credentials, etc. that they would need otherwise (the sci-hub UX is vastly superior).


The ability of a non-scientist to understand a paper at a useful level varies greatly based on the area of study and the reader's knowledge of the topic. 'Basic scientific literacy' may not be enough, but there are many more individuals with an amateur level of advanced understanding or interest in a given area who benefit greatly. I personally have no institutional access but I read a lot of papers that I can understand with an undergrad level of science background and a keen interest in a specific subject. I'm sure there are many others like this.


And let us not forget those who ended up working in wildly different fields from their degree or degrees. Such "cross-pollination" between different fields can be very productive.


"Basic scientific Literacy" here probably should include basic understanding of statistics, and knowledge on how to read a journal paper in a foreign scientific domain. You don't have to know the lay of the land in a particular domain to understand the specific claim or methodology outlined, knowing how much it may mean in a domain. . . well that is context determined.


> (in part because most papers are badly written and need familiarity with the genre to be understood)

Considering science is a community/social enterprise, it's surprising that such papers are published at all.


> To me, Alexandra Elbakyan is kind of like a Martin Luther of science. That is, she enables the lay public to read and interpret science rather than awaiting the interpretation and approval of academics from high-status institutions.

I don't think the analogy works. Luthor took a single text and translated it, making it accessible. The fact that it was accessible didn't affect the creation of the text.

By contrast "the interpretation and approval of academics from high-status institutions." is an intrinsic part of the creation of the document. It's part of the peer-review essential to the creation of good science.

Experts disagreeing on the fundamental nature of the holy trinity is not the same as experts disagreeing on the experimental design behind testing a new category of drugs.


Still - I think the analogy is suiting. Martin Luther started a movement which granted common people access to religious texts (by translating them from Latin). Sci-Hub removes other obstacles to access knowledge, but the goal is somewhat similar.

This is not to say that peer review and expert validation doesn't have it's place in the process - but it could (should) be done in open, with as many eyes watching as possible. So IMHO anything that denies general public access to scientific work should be removed.


I agree. This is clearly how science should be done for the benefit of humanity. Closed access journals are the "property is theft" of scholarship. Being able to read papers and engage directly with the material is how innovation will happen. If not SciHub then someone else. It was and is just too important of a service not to be the new status quou.


Completely agree. Science is not a published paper about an experiment. It is a community activity, the results of which only improve with increased participation and experimental replication. Anything enabling a greater number of people to participate makes science more science-y.


Why not create an alternative platform where papers can be published for free, by the authors, instead of stealing them and publishing them like warez?


See https://news.ycombinator.com/item?id=16341265 - basically, many top-tier journals won't accept papers that are also published in open access journals.


> Martin Luther of science

Martin Luther was an anti-semite and called for synagogues to be set on fire. Not a good comparison.


We scientists WANT (desperately want!) anyone and everyone, experts and neophytes alike, to read and pay attention to our work! The unavailability of scientific research papers is in no way due to snooty scientists restricting the availability of papers to others in the ivory tower. This is on the publishers, who charge massive rent-seeking mark-ups for actually doing very little work and providing very little value in this digital age. It is also on the system of sticks and carrots that prevents many scientists (notably those who do not yet have tenure) from publishing in free and open journals by tying our livelihood to publication in more traditional (non-free, non-open) journals that have high Impact Factors.


That same year, the AAP and Elsevier also supported and lobbied in favor of a bill that would have prevented the government from requiring agencies to make research published through a journal Open Access at any point

I realize I shouldn't be surprised, but that shocked me. God forbid taxpayers should have access to the research they paid for. How these organizations expect any public sympathy is beyond me.


In comparison several European countries are increasingly mandating that publicly funded research should be open access. Why this simple concept would elude the US government is beyond me.

However Open Access is not panacea. Most of the times it just moves the cost from the subscriber to the author. The open access costs are equally extortionate for several journals in the ballpark of £2000 for a single article. As a result publicly funded grants must also include a quite significant budget for publishing costs further inflating the already bloated administrative overhead of research programmes. So you either pay the journals to open up your article or pay for a subscription. It does increase visibility and accessibility of research but, still, we are not quite there yet.


> However Open Access is not panacea. Most of the times it just moves the cost from the subscriber to the author.

I feel lucky to work in a field where journals are OA, but authors do not have to pay to publish. Our major journals are published by learned societies that were founded with endowments large enough to cover publication costs in perpetuity. In fact, that’s how a lot of journals used to be published before big publishers, like Elsevier and Springer, convinced those societies to entrust them with the journals, claiming it would make things easier for everyone.


>>but authors do not have to pay to publish

My company has to pay to publish in our preferred journal of open access (PeerJ). I gladly pay for lifetime memberships for all of our authors, which is pittance and affords lifetime publication rights with no additional fees.

But we do pay. Most OA journals do have fee waivers and are very liberal with them, though.


PeerJ is different, though, in that you pay through lifetime memberships rather than a new fee every time you want to publish an article. Their fees are also substantially lower than those of the traditional publishers.


No question. I support PeerJ to the fullest extent and the price is more than fair. (It's actually unfair, honestly, it's far too low.)

But there is a fee. And it's important for supporters to be honest so it doesn't appear one-sided, so, here I am.


Out of curiosity, what's the field? In physics nothing is particularly open-access but that doesn't matter because everything with any significance at all gets put up on arxiv six months before publication.


Well, it does matter somewhat: authors still publish in journals at the traditional publishers, and universities still subscribe to it, wasting research funds in developed countries and making the barriers to participation higher for researchers in emerging economies.


For its conferences, ACM SIGPLAN asks the authors to pay for the open-access fee if possible, but will cover it if the author cannot afford it.


Don't know about OP. In my experience a lot/most Computational Biology papers are Open Access.


I can second this. I've had a much better experience in Computational Biology compared to other fields wrt paper access.


> Why this simple concept would elude the US government is beyond me.

The NIH, the single biggest funder of research in the US, has an open access mandate.


The single biggest funder of research in the world.


In reality the authors and editors do 90% of the work and are not paid. If the articles are electronic then the "publishing" costs are minimal. No real reason for everything to be open access (and I mean real open access, not charging authors).


True, those who do the real work doesn't get paid, but the ones who publish it gets paid.


Even simpler concepts like "passing a budget" or "vet your staff before hiring" elude the U.S. government these days so I really can't expect them to even comprehend the concept of open access journals.


Intentional or not, pay-to-publish also acts as a barrier-to-entry for less well-funded research groups.


This is a very real thing - a open access paper or two per year is, for example, a non-trivial amount of my startup funding.


Many OA journals have fee waivers.


Said fee waivers are often for fairly specific requirements, and "I'd rather spend the money on something else" isn't among them.


> For instance, in 2013, the Obama administration mandated that copies of research conducted through federal agencies must be uploaded to free repositories within 12 months of publishing.

All federally funded research in the US is mandated to be open access.


> Why this simple concept would elude the US government is beyond me.

At this point the US government pretty much entirely works for monied interests. It has waxed and waned over the years but for the past thirty years or so, it has strongly waxed.


> Why this simple concept would elude the US government is beyond me.

Money.


Does money not exist in Europe?


I suspect politicians in Europe are less dependent on private contributions.


Europe has wider variation in political corruption (less corrupt in Germany, more corrupt in Romania). Whereas US politicians always seem to be moderately corrupt, wherever they come from.


Kind of you to put it that way. 'Always' is too blanket, but those who are not are clearly diminishing in number. (They ... 'retire'.)


I'm pretty sure we have that thing here too.


There's no need to pay a publisher £2000. The authors can self-publish.


Self-publishing is not without its problems.

Peer review, while not perfect, guarantees that the text has been reviewed by quality researchers. Journals have an interest in filtering bad publications, if only to keep their reputation.

In contrast, a self-published paper has no guarantee of quality, simply because everyone can do it. An author's reputation is often a proxy to assess a publication's quality, but this is clearly bad for new researchers.

I think there are good alternatives - my last publication costed around $300, which were used to pay for the conference where I presented it.


Ah, so we simply create a protocol that transparantly breaks down the peer review process into a rainbow of gradients thereby allowing the desired process of review by what you call quality reseachers. Such protocol could allow a review-for-free mechanism that could be time consuming and it could build a fund for the review process. Arguably the review process needs not to be finished at any time. It is probably big fun to revisit the process in the licht of later findings. Lastly but not least the protocol could help establish a name for reviewing scientists who just happen to do a lot of reading. A (large) bit like github does for programmers.


Self-publications and 'gray literature' are generally deemed worthless and often even considered to contribute a large amount of negative value to a CV. To put things simple, you'll loose your job.


Not if I want to keep my job come review time.


Seriously - everyone, if you can afford it, boycott Elsevier. Don't review for them, don't publish with them (of course, for junior academics that need to collect impact points that's difficult...), don't buy them. The "criticism and scandals" section on Elsevier's Wikipedia entry is huge; their behaviour is shameful.


Publicy funded institutions pay for the scientists that write the papers and review their colleagues' papers. And then the public has to pay Elsevier again for the right to read the papers. It is an absurd system. The value added by the publisher does not justify the enormous sums the publishers charge.


We are all interacting with this HN website over technology whose very underpinnings was the assumption its freely available -The protocol stack, the application-to-protocol bindings, the definitions which a program source code represents, the compiler tools..

Am I alone in seeing the contradictions, of people defending for-profit science publishing, using a vehicle whose very existence is predicated on government funded science and technology coming with 'its free' as a requirement?


Do you believe the tech industry is so different with its walled gardens, proprietary software, and closed protocols built off of decades of publicly funded research and infrastructure? Why?

Is it contradictory to support an industry which does that while supporting this? Would you be for the liberation of Google's property?


It's categorically different. The arrow is in the opposite direction. The tech industry's private siloed stacks are built on top of that public infrastructure and are an elaboration of it. Whereas the publishing industry's walled gardens are the foundation upon which public research is built.


It's the same.

Publishers are the "gateway" to science, and they have cornered that market, using publicly funded science.

Google is the "gateway" to search, built largely using publicly funded science. Facebook is the gateway to online social interaction, and is again built largely using publicly funded science. Et cetera.


The content in Google and much of it in Facebook is discoverable via those services, but exist outside of it.

Publishers are not mere discovery services - they are gatekeepers, not portals.


"built using publicly funded..." Is not the relevant direction. You'll come to absurd conclusions. A building in San Francisco is literally built using publicly funded safety science, and materials science, etc. Should we then force it to be free to access?


A building is not information, in the sense that you cannot freely copy it.


Well, things are a bit different in the US where your universities are privately funded but in many countries the publishing industry publishes research from researchers employed by state funded institutions and sells the research back to libraries in state funded institutions. So, really they're being doubly financed by the state.


There are moments in what google do where it's explicitly in the public domain. QUIC is now in IETF process. Substantial parts of IPv6 are done in IETF. Google public DNS is in IETF. I believe Google docs and sheets uses open doc standards.

There are moments which are not. I think their smarts behind ad placement logic and the ML derived stuff may be trade secret not patent. I believe their use of patents acquired through Motorola was defensive not offensive but I disliked that move.

Google exploit first mover advantage and category killer scale as much as IPR. If they used IPR as much as apple do, to defend the indefensible (slide to open for goodness sake) I'd repudiate that.

I don't entirely like Google. In this space is see them as a lesser evil than say SCO or Oracle


I think the argument is that the science building the underlying infrastructure is public and free to share. And that is a fundamental part of the reason for it reaching the size/scale that it is right now.


Publicly funded research should without reservation be free and available to all (if that's what the author chooses). I say that as someone who hopes to publish within the next 18 months. Not sure it's going to be any good, but it will be free.

The big publishers are public parasites that make money off the work of other people and through controlling information and progress. The quicker they go into the dustbin of history the better. My issue is less with the argument and more with the general sentiment.

I don't understand how we can get outraged that companies like Elsevier operate this way when it is fundamentally the same closed model that is so prevalent in tech. I don't imagine as an industry we are so self-unaware, contradictory, or (to use a less nice word) hypocritical, but the sentiment doesn't match with the actions of this industry. where does the disconnect come from?

We celebrate piracy in this thread, but if these same tactics were turned on almost any company represented on HN, it would not be seen as a virtue.


>fundamentally the same closed model prevalent in tech

They're totally different. If tech were like academic publishing, software would be developed in publicly-subsidized studios and then given, for free, to for-profit distributors. These distributors would play no significant part in the development of the software, but would happily charge the public an arm and a leg for any usage of it.


Yes. I don't buy "information wants to be free" but I do believe "the internet worked better than the others because it was open spec and free" (in no sense without cost)


That's true. ITU-T standards (eg for ATM and OSI protocols) used to be really expensive. Buy one 8 page document for 50$, realize that this is mostly a collection of references to other documents, keep buying until you have all the information. This takes a lot of time, as the documents aren't available online. Or you go to a university library with a notebook ("making a copy is strictly forbidden"). It was a system that shut out small companies.

Or you implement IETF RfCs. There's an FTP server with all standards documents. If you have just email but no Internet yet (think uucp mail or bbs), there's a service where you can request documents to be sent to you.


Nobody ever gives the full version of that quote:

"On the one hand information wants to be expensive, because it's so valuable. The right information in the right place just changes your life. On the other hand, information wants to be free, because the cost of getting it out is getting lower and lower all the time. So you have these two fighting against each other."

Brand wasn't a naive utopian; he acutely understood the conflicts that the internet would bring about.


To save others from panicked searching, Stewart Brand still is not a naive utopian; he has not died.


Without defending the current system of publication, many things that are government funded are not free. For instance, our ironically named freeways. And nearly all government offered services from postal to patents come with costs and fees, which can become quite obscene.

---

@pwinnski - This is likely either a dialect or national language issue. In the US, at least the section where I am from, freeways are what you might call highways. They have no traffic lights, are high speed, and have toll booths for entry and spread along them at intervals. Hence the reason I stated their naming quite ironic! They are built by government, but are not freely available. Responding here since hacker news seems to stop you from replying in a thread once something you post has been mass downvoted.


I'm confused. Freeways are, in fact, free. Both in movement and in cost to use. If a highway charges tolls, it is not a freeway. If a highway has traffic lights, it is not a freeway.

You may be conflating price with cost, but the two are not the same. The government funding of the cost of freeways means that the price is free to all users.


In many countries there are taxes you have to pay to be allowed to use the roads.


Yes, but that's because use of a road causes wear. Homeowners didn't want to foot the bill for the road outside their house, and potholes are a real problem to safety. People who use the roads are the ones creating potholes, so why not charge them for that? It's a public service, like the fire service or the national health service.


The fact there’s a reason for it doesn’t change the fact.

However yes, public highways cost money to maintain. They can’t be truly free unless the construction and maintenance were free - donated land and volunteer labour? Right. Otherwise somebody has to pay for it.

Frankly it’s potentially the same way with published scientific papers, we do need a system of professional peer review. That could be voluntary, but would probably be better if it were independently funded in some way. ArXiv is one approach but I’m not sure it’s a complete solution.


Now imagine your state funded public highways were vetted by some private company, let's call them "Elsewhere, Inc", who wrote a 3-line review of each highway and then charged taxpaying members of the public $30-$50 per journey to use those highways. And then Elsewhere, Inc managed to convince most civil engineering companies to gauge employees' performance solely by whether or not Elsewhere, Inc has reviewed the roads they built.


Actually, imagine that rather than writing the reviews themselves, Elsewhere simply asked other road builders to evaluate the roads of their peers for free. That's the situation in publishing -- the publishers don't even do the peer review -- they get content for free from authors, get peer review for free from reviewers, and then charge readers to access the final papers!


You know it's bad when I'm being snarky and I still don't push it as far as they do with a straight face.


Peer review is already being done and it's free. Distribution is "too cheap to meter". Editing is pretty much not done at all at the publisher level.

So to fit the road tax model - we already pay for research. So let's add 0.0001$ extra tax for publishing, and kill off the current publishers.


Peer reviewers are not usually compensated.


Peer review is voluntary and done by other researchers in the field. Publishers mostly take care of editing the final text (basically typos and formatting) and uploading them to their servers.


We pay taxes whether we use the roads or not. Those taxes are decoupled from any given road. Further, when I visit a state in which I do not live, I use those freeways without paying a cent, directly or indirectly.

If the complaint is that roads require maintenance and are therefore not free, that's a torture of the word.


You are paying taxes that fund research as well.


@Sci_Hub twitter:

"Unfortunately, the Verge article about Sci-Hub founder Alexandra Elbakyan has many inaccuracies"

"I'm going to publish a separate blog post explaining the inaccuracies later"

https://twitter.com/Sci_Hub/status/961829490803449856

https://twitter.com/Sci_Hub/status/961836113731072003


"Publishing powerhouses like Proceedings of the National Academy of Sciences have estimated its internal cost per-article to be around $3,700. Nature, meanwhile, says that each article sets it back around $30,000 to $40,000"

I work in book publishing, not journals, so I may be missing something in Nature's internal structure, but that seems like a lot to spend on one article. As in I think we spend less than that to do a whole book.


Those numbers can't be right, honestly.

Bjoern Brembs has been writing a lot about open access publishing and the costs associated, here he cites a few numbers:

>There is ample material on how much the publication of a scholarly open-access article costs. SciELO publishes for under US$100 per article. SciELO, however, publishes largely in regions of the world where labor and other associated costs are comparatively low. How much costs do other organizations cite? The Open Library of Humanities is on the record with about US$500, Ubiquity, Hindawi and PeerJ are also on record within the US$100-500 range and Scholastica, RIO Journal, Science Open, F1000Research or arpha mention similar figures. Thus, there are now about ten different companies or organizations which all agree that open access publishing costs about US$100-500 and Independent analysts agree: it is established that the actual costs of publishing a scholarly article are in the low hundreds of US$.

Note that he focuses on open access articles, not closed access + open access articles, but I can't think of a reason why closed access would cost more.

http://bjoern.brembs.net/2017/11/is-a-cost-neutral-transitio...


> Those numbers can't be right, honestly.

Just a very rough calculation: If Nature has a staff of around 150 [1], each paid $100,000 year, 50 issues per year, 20 research articles per issue. This gives

150 times 100,000/(50 times 20) = $15,000 per article.

My numbers are probably off in more than one way.

[1] https://www.nature.com/nature/about/journal-staff


They have a staff of 150 to publish 20 articles each month. That seems quite top-heavy. Remember that authors and reviewers of the articles are not paid.


Good point made by 'IanCal here:

https://news.ycombinator.com/item?id=16339193

Nature drops about 95% of submitted papers somewhere in the pipeline, so part of the costs applies to 20x more papers than you see in the end.


All right, but then why are most open-access papers under $3000?

If an editor can survive with these fees even when the paper will be publicly available, how can Nature need $30,000 for a paper people will pay to read?


> why are most open-access papers under $3000?

Ordinary scientific journals don't have reviewers in paid staff. An editor sends an article (without reading it) to 2-4 non-paid reviewers, and makes the publish / don't publish decision based on the reviews. Also the editor is not paid, either.


Not Nature, reviewers.


Nature has in house staff editors and reviewers that vet papers before sending them out to per-review. Most papers sent to Nature get rejected long before the per-review process starts.


Perhaps they have to manually apply Latex templates with a Morse key.


But Nature sells subscriptions and advertising...


Publishing isn't that generous, in general, salary-wise. Only the folks at the top are getting that kind of money.

Also, vetting submissions should be a relatively small part of your time. When you're reading a lot of submissions, you learn how to weed out the weak fairly quickly.


[disclaimer, work for Digital Science, whose parent company owns the parent (or just merged?) company that produces Nature, my own thoughts etc.]

Nature have paid editors who also review articles before they go out for peer review, which means they have fairly obvious costs per article submitted (on top of other things like organising peer review). However, you don't pay to submit, you only pay if you're accepted. Nature accepts, I think, about 5% of the articles submitted to them. That means that for every fixed cost per submission the cost per paper published is 20x higher.


I'd just like to clarify this:

> However, you don't pay to submit, you only pay if you're accepted.

I should have been clearer that the main Nature journal does not, I believe, charge for accepted papers as it charges readers instead. However the general setup is that with open access papers people pay only if accepted.


Is paying editors typical for a journal? Id guess Nature is comparatively well off financially, so would smaller journals be able to afford to pay editors in the same way?


In general (as far as I'm aware), no, it's not usually the case.


Nature almost certainly took their top-line costs and divided by number of articles to get those values, which is obviously a ridiculous way to look at article costs.


When I published in Nature and Science (haven't published in PNAS but I have in other journals) it seemed to be a bit more organized with more actual employed staff vs. volunteer academic editors and so on. Not sure if that could explain this large a discrepancy.


This has been my experience at PNAS too. Professional editors, much more responsive interaction, stronger page layout, etc.


Poor ones. Well in that case they really should simply go out of business. If your model is that broken you have no right to exist in the first place. But something - such as their profits posted year-after-year-after-year - makes me suspect that their story isn't complete.


Nature almost certainly didn’t round down that number, as a higher figure suits them better, but I do believe that they have significant costs. Reason? For every published article, Nature rejects about 12 (https://www.nature.com/nature/for-authors/editorial-criteria...).

That page also says ”On submission, the manuscript is assigned to an editor covering the subject area, who seeks informal advice from scientific advisors and editorial colleagues, and who makes this initial decision”

⇒ multiple persons, some of them employed by Nature, read each paper. Nature’s editors are highly educated (https://www.nature.com/nature/about/editors/), so their time won’t be cheap.

That can add up. For example, _if_ it costs a manweek to decide whether to do peer review on a single paper, finding that one accepted paper out and would be a quarter of a man year of work. I guess the actual effort is lower, but still significant. People have to read those papers, and that takes time.

And that is just step 1. Peer reviewers likely work for free, but managing volunteers costs money, too.


thats because they are bullshiting. elsevier charges $5000 for open access which is supposed to cover its internal cost and then some. Nature is probably accounting here its various promotional / PR events and conferences as part of the cost of the article, which is unethical anyway, since the science should be popularized based on merit, not on the amount spent to advertise it.


There's a few reasons for that. One is that they take their existing profit margins and incorporate them as "cost" - and these margins are really very comfortable.

Another is that it's not the cost of one article, but also of every article they have had to reject instead of that one. That number is pretty high for Nature (the journal).


> But if America’s access were further restricted, it would be a blow to the site, and to many of the “capitalists” that use it.

The Verge are definitely misunderstanding Elbakyan when she says this:

>"the capitalists have started blocking Sci-Hub domains, so the site may not be accessible at the regular addresses."

Elbakyan is referring to Elsevier and the ACS and the other major publishes putting pressure on SciHub and making it inaccessible. No capitalists are using the service in the Marxist sense of the word "capitalist" which is definitely what Elbakyan was going for.

The Verge's writing here misrepresents what she's saying, and truly does change the message of what she has said. I think they need to edit and fix that.


Elbakyan says so herself as well: https://twitter.com/Sci_Hub/status/961829490803449856

And she'll go into more detail later: https://twitter.com/Sci_Hub/status/961836113731072003


Yeah, she seems to be making a point there that Verge's usage seems to be (unintentionally) obfuscating.


First it was music, then it was movies, Google took books off the table by accident, now journal articles. You can see where this is going: small subscriptions for millions of subscribers world-wide. Amazon Prime, YouTube Red, or Spotify for scientists. My questions: 1) can I as an author get 30% like musicians on iTunes?, 2) will this make science more attractive as a career?


1) Probably not. People want to produce papers more than people want to consume papers.

2) Also probably not. Because most papers aren't read widely. The folks I know who have things that generate a small drip of royalties (usually licenses from patents) treat it as a nice little bit of fun-money, but it's nowhere near enough to either fund a lab or really change the calculus of "Is this job worth doing" from a monetary perspective.

Especially when compared to "Fuck it, I'll go work for industry."


Based on bibliographic notes alone we can say that for each produced paper, a lot more are consumed.


Most of the references were also in the last paper of the authors, and will almost surreally be in the next paper. (And there are also self citations ...)

I'd probably divide the number of citation in math by 3 and in physics by 5 to discount repetitions.


As for your second question: please no. Of every 200 PhD students, only one ends up becoming a professor. Science is attractive enough as it is, despite pretty much every aspect of it (pay, overtime, job security, etc.) is abysmal.


Professor isn't the only career possible with a PhD.


It depends on the field. It basically is that way for some fields. Those fields are almost like pyramid schemes (not that the topics are worthless, but the economics behind the level of investment simply do not work). Clearly fields like computer science have all sorts of ways to directly use advanced skills outside of academia, but that's not the case always.


Musicians get 70% of retail is on iTunes downloads. (Or rather, the owners of the copyrights do.)


The article (and the discussion here) seems very focused on the legalities of Sci-Hub. But this misses the real question:

Has there been a noticeable impact on general scientific progress as a result?

I feel like there must at least be anecdotal stories of people reading, exploring and making connections they otherwise would not have been able to because free > $35/paper.


Yeah, me. My company is publishing six papers this year and a large mover forward is the ability to read all the papers in our industry fee-free. Elbakyan's work moved us so dramatically that I instituted a no-fee publication policy; we will only publish in fee-free, open access, and open peer review journals - providing all data and supplementary material openly to everyone either through the portal's site (if supported) or a public repository.

EDIT: In my opinion Elbakyan should be considered one of the most important people of her generation. It is my belief that academia will see fit to write her out.


Apart from the personal anecdotes by people, it also allows university negotiators to take a tougher stance when negotiating prices with the traditional publishers, as their threat of cancelling subscriptions has become more credible. This has been most noticeable in Germany, where universities have come together in Projekt DEAL through which many have, this year, cancelled their Elsevier subscriptions [1]. The Dutch are likely to follow [2], and several rumours of other countries have surfaced as well.

[1] https://www.projekt-deal.de/about-deal/ [2] https://www.scienceguide.nl/2018/01/we-are-prepared-for-a-no...


For any student not part of a university with a journal access program, Sci-Hub is a massive boon. I've not personally had this issue, but there are quite a few amazing post graduates that I know of who rely on Sci-Hub a fair amount.


> I feel like there must at least be anecdotal stories of people reading, exploring and making connections they otherwise would not have been able to because free > $35/paper.

And I wonder how many of those people don't know about DeepDyve.com? It's essentially Spotify for journals. The $49/month ($30/month if you buy a yearly subscription) plan gives unlimited online access to most papers from a huge list of journals [1] from a huge list of publishers [2]. (There are also group plans available, but they don't list the prices for those on their website, as far as I can see).

You can also buy 5 packs of rental tokens for $20. A rental token gives you access to an article for 30 days. People on the free plan get 5-minute full text previews, so you can at least skim an article to get a better idea if you want to consume a token to rent it.

Open access is much nicer, of course, but most cases I've seen where someone says that the cost of non-open access, such as the $35/paper so many publishers charge if you buy directly through them, was a serious issue to their research would have been able to get those papers through DeepDyve at a cost well within their budget, but for some reason very few people in the wild seem to have heard of it even though it has been around for quite a while.

[1] https://www.deepdyve.com/browse/journals?page=all

[2] https://www.deepdyve.com/browse/publishers


Hadn't heard of DeepDyve and I'm always willing to look into paying for something using that type of model.

Out of five major journals in my discipline, four are "Preview Only."

What a surprise. "Huge list," indeed.


> Out of five major journals in my discipline, four are "Preview Only."

What discipline?

> What a surprise. "Huge list," indeed.

The journal list there lists ~13100 journals, of which ~9200 are available and ~3900 are preview only. Seems reasonable huge to me.


Just as spotify for music has almost everything, "spotify for journals" needs to have 80-90%+ coverage or it is an annoying waste of time.


It seems to be missing all of the ACS journals (a big chunk of chemistry) and IEEE (a big chunk of electrical engineering and computer science).

Also, I find the idea of merely 'renting' a paper objectionable. Deepdyve charges individually for PDF downloads, and otherwise only allows you to print 20 pages a month.


Most of the stuff I mention in comments that I forward to developers came from several years of free access to ACM and IEEE where I could follow-up on any line of research I wanted. I had to read hundreds of papers, esp experience reports, to get an idea of what would or wouldn't work in a variety of contexts for INFOSEC, high reliability, productivity, and maintainability. There have been countless people stunned by some of the prior works wishing they'd known about them before for their more recent explorations. Supports idea that having them available with something like SciHub for those without free access would've boosted researcher or programmer effectiveness for those using them.

EDIT: The CompSci people themselves that are pushing progress often cite paywalled works. That may also support that access to them would accelerate progress outside CompSci if its a prerequisite to developer certain ideas evolving from older ones.


I'm a medical doctor and frequently use sci-hub to review papers relevant to care of my patients. Prior to sci-hub, I would often have a clinical question, look it up, find a paywalled article, attempt to access it via the hospital's access, find it usually wasn't, give up, and then have spent 10 minutes of precious time doing this, feel like an idiot for wasting time, and just make a decision without the best information.

Now I get the DOI/pubmed #, go directly to scihub (or at least one of the scihub domains that isn't blocked by the hospital for being "unethical") and have the paper in 20 seconds.


Finally an article that dips into Elbakyan's worldview and motivations. I _hate_ the typical lazy coverage which projects American "civil rights hero" stereotype on her. She isn't, she is borderline crazy person (that is, self-proclaimed "communist" supporting Putin's crony capitalism) with no fear of consequences and with incentives aligned with the majority of scientific community. It can be made into a _much_ richer and personal story than your typical "fight the man" trope, but most journalists can't be bothered.


I wish your comment was higher. You describe her personality quite accurately, based on what I know. That being said, "borderline crazy person with no fear of consequences" is the exactly the type of people that can push humanity forward.

Richard Stallman is another example of such a person, I think. You have to be an extremist to make other players in the field to yield.


Yes, I guess you can do that comparison with Richard Stallman if you are willing to disregard what they stand for ("free as in freedom" vs "glory to Motherland").


What about "¡Viva la Revolución!"? Stallman was an ardent supporter of the Chávez regime in Venezuela. It took him some years to realize that Chávez was, to put it mildly, kind of a dick.


Sci-Hub's Twitter says there are "many inaccuracies" regarding the Verge article.

https://twitter.com/Sci_Hub/status/961829490803449856


In good old days, there used to be DC++. Currently I use what is touted as the best client for it AirDC++.

Why can't peer to peer file sharing like Direct Connect be used for such purposes? People have successfully used i2p, torrents, emule,etc.?

Can they be tracked and shut down?


LibGen's contents can be donwloaded/distributed over torrents. Participate!


Knew from the title it's going to be about SciHub.

As someone who is just out of a grad programme and no longer has access to the university network, I'm incredibly thankful to her.

I wish putting papers on ArXiV were the norm in all disciplines, not just math/physics/cs, but until that happens, this is the answer (and when it does, SciHub will essentially become a mirror). The licensing deals that the publishers have made with countries like Germany are still extortion, and aren't a solution.

You can learn more about SciHub, and LibGin (a similar project for science books which mirrors SciHub articles) from Wikipedia: https://en.wikipedia.org/wiki/Sci-Hub

The Wiki page also lists mirrors, in case any of the domains are blocked.

If you have some bitcoin to spare, SciHub is donation-supported, and you can help the fight to make science available to the people that do it - and everyone else.

----------------------------

For those out of the loop: currently, access to papers done by scientists (paid for by government/their home institution), typeset by the same scientists, reviewed by their peers (for free) and distributed electronically costs bug bucks to access.

The publishers still cite hosting/maintenance costs as a justification, and yet when someone manages to do the same for free (e.g. SciHub), the publishers go to war.

SciHub demonstrates that hosting costs are not a justification for atrocious access fees (to the tune of $30 to see one paper!), and yet the racketeering scheme (a relic from the days of Guttenberg's press) still, somehow, persists.

To address this problem, people in certain disciplines (math, physics) currently publish their pre-prints online on sites like ArXiV and their own university pages. However, these self-publication methods are not peer-reviewed, and so are not a direct solution to the problem. Some big-name scientists have started open-access online journals, but it's hard to get everyone on board, as academic performance evaluations are based on publications in well-established journals -- so only established academics can afford to be published in new journals when more established alternatives are present, but have little motivation to do so, since the university pays the access fee anyway.

SciHub is the interim solution to this chicken-and-egg problem. The long-term solution is still SciHub, but legal. Alexandra Elbakyan has done the footwork on the implementation side, the rest of the battle is going to be social (getting the academics on board) and legal (preventing the publishers like Elsevier from crushing the initiative, and, in the long run, destroying their monopoly on providing access to government-funded research altogether -- which means, effectively, wiping them out altogether).

All of this is talked about in the article in more detail.


LibGen is a great concept, and I have uploaded a lot of my own scans of scholarly material there, but I fear it is sort of collapsing under its own weight now. Its contents will continue to remain available thanks to their smart decision to decentralize it (torrents are available to download and seed the material). However, there is absolutely no curating of LibGen’s holdings. That means that a lot of metadata is wrong or incomplete, and some books are uploaded in a dozen different versions and you don’t know which ones are high-quality and which ones are low-quality scans or outright corrupt files.

LibGen is just the latest in a long series of pirate ebook libraries that have popped up and then been either shut down by copyright holders or abandoned by admins. It’ll eventually give way to another, and hopefully the next generation will maintain its strengths while having higher-quality metadata.


>That means that a lot of metadata is wrong or incomplete, and some books are uploaded in a dozen different versions and you don’t know which ones are high-quality and which ones are low-quality scans or outright corrupt files.

The rule of the thumb is to ignore the metadata and just look for the highest id number.


It's LibGen, not LibGin.


In another era, LibGin is a library about illicit alcoholic beverage production.


This would be ginious!


Thanks, it's too late to edit the typo now :)

It's still библиотека колхоз in my mind (a predecessor whose contents was merged into LibGen).


Sci-hub, and the availability of research, have been and is, something to be grateful for.

It seems like fear is the motivating function for restriction of knowledge, and that is understandable.

"Science's...Queen"


What's the reason this entire thing isn't already entirely over torrent?


It is. LibGen is available via torrent.


But no search function, is that it?



The academic publishing model, and academic research in general, will be dramatically different in 15 years, maybe fewer

In california, venture capital already funds more research than the NIH ($5B vs $1.5-2B). Much of this is one step later stage than academic research but increasingly they overlap directly. VC funding for life science research is skyrocketing, and NIH funding is declining. In January 2018 alone there was $1B invested in seed or series a biotech deals which is more than Stanford gets from the NIH in two years

The fact is that the publication bias is ruining scientific quality. Anywhere from 40-75% of academic science isn't reproducible, either because of chance, poor quality control / documentation, cherry picking data or outright fraud. The list of researchers and universities that VCs and pharma will work with is quite small, and many VCs are just funding the basic science themselves rather than invest in stuff that just doesn't work half the time https://lifescivc.com/2011/03/academic-bias-biotech-failures...

Young scientists are treated horribly by the academic world, with brilliant PhDs and postdocs with 15 years of research experience in top labs making barely minimum wage with no hope of tenure. Many universities make it very hard for students to explore non academic jobs because they realize they have the most highly educated captive workforce the world has ever seen

If a good alternative to working in academia opens up, things will change fast


I see this situation as resulting from governmental decisions that reduce funding for research and make doing it harder. There are also bonuses to make companies that extract value from the public research sector. All this quantitative easing seems to give the capitalists money but not the scientists, unless they are happy working for the capitalists.


NIH puts over $30B into research annually. (https://www.nih.gov/about-nih/what-we-do/budget)


I was referring to CA funding not total budget. VC funds $15B / year; top 15 pharma funds $75B. VC and NIh funding are closer when you consider overhead and non-life sci research


You're off on the NIH extramural budget by a factor of >10X.


I was referring to spending in CA although the I think I misquoted the figure as $1.5-2B; think it's $3.7B in ca

Total NIH budget is like $32B. Half of that goes to basic research and some goes to non-life sci research. Life sci VC funding in 2017 was $15B. Some of that goes to admin costs obviously but not nearly as much as NIH if you look at headcount at companies

https://www.hhs.gov/about/budget/budget-in-brief/nih/index.h...


To be honest though, CA is the best case for VC funding, and I'm willing to wager the number of states with significant VC funding of science can be counted on your fingers.

In comparison, the NIH funding is nation wide, and specifically has mechanisms for underfunded states. What do you think the ratio of funding between the NIH and life sci VC funding is in say...Idaho?


I agree. Massachusetts is maybe the only other state like that. I admittedly used an outlier example as it illustrates a trend, and I know so many scientists who dont realize 1) how hard it is these days to get tenure and 2) how much opportunity there is to do cool science outside of academia, so I use this example to show them that there are other options

For what it's worth, I know there are movements to attract more VC funding to places like Pittsburgh, NYC, San Diego, Texas, North Carolina etc. still early stages, but I think in 15 years things could be very different


Considering that scientific publishing is a racket[1], I find it very hard to have any sympathy whatsoever for the likes of Elsevier

1. https://news.ycombinator.com/item?id=15993603


What is an "open" paper ? I mean, the paper is a very succinct description of the research done. For example, if one demonstrates how one builds a CNN but at the same time, doesn't provide the data set used to build it, then it's worthless. In the same vein, if one provides explanations on how to detect motion in a picture but doesn't provide the full source code of its detection pipeline, then it may be super tough to rebuild that code (because it assumes the paper actually gives all the necessary information, not the "general idea"))

So, how open is open ?


This is why about 90% of the papers Google publishes are useless.

They describe deep learning successes, with datasets that aren't public, on hardware that isn't public, with software that isn't public.

They describe technologies, be it spanner, borg, flumejava or others, relying on closed hard- and software.

If an average university professor can't replicate research 1:1 based solely on existing, open technology and the paper, the paper should not be accepted by any journal. In fact, it shouldn't even be accepted on conferences, or be able to get a doi code. That's not worthy of being called science.


I don't think they're quite useless (they tend to contain some interesting ideas, which presumably work as google have generally implemented them) but as you say they're definitely not 'science' as they can't be replicated.


While I obviously agree that fully reproducible results is the ideal, just knowing that something is possible and having some idea where to start is often valuable in itself.


First of all, if it’s not reproducible, it’s not science. The entire point of science is getting results that give insight to others.

Second, often the papers set prerequisites that are not available to anyone else, because these, too, are undocumented.

This is cool if you want an inspiration, but large parts of these papers are as useful as watching Star Trek for actual science. They can serve as inspiration, but they can’t tell you how to replicate something, or give you possible leads to expand on it.

And this issue gets even worse – if I’m writing an undergrad thesis on a tech that Google has several years lead on, my thesis won’t be able to produce anything new. I can’t expand on the state of the art, because the state of the art isn’t even available to me.


The Open Access movement (making articles openly available) is a smaller part of the larger Open Science movement, which encompasses among others making code and data open, registering research plans in advance, publishing null results, etc.


I wonder what would happen if Apple bought Elsevier, and made all of its journals available on any Apple device for $100/yr? A reasonable thing to do with its cash hoard, no? Amazon, or Bezos, could probably raise the money to swing a hostile takeover.


> I wonder what would happen if Apple bought Elsevier

I'd much rather see them go down than be bought out.


That wouldn't solve any problem, just shift the paywall to someone else.


I wonder if she needs financial support for lawyers and stuff. I would gladly make a donation.


Dumb question:

How make a free high impact journal?


Rather than answer this outright (as I don't have a complete answer), I'd encourage you to imagine what some of the things that would make a high impact journal, one that you'd find useful. Even if you don't have all of the answers, that would move the ball forward a bit, and put some skin in the game/conversation, as it were. What would make the journal high impact? Would there be costs associated with that? If so, how would you fund it?


Modern heros


I like to see this for what it is, piracy. If I want access to a paper, I have sci-hub and that's beneficial to me. I'm saying its a good thing.

Sidenote: those pictures are creepy.


It's interesting how Elbakyan generally receives only praise in discussions like this. Do no supporters of capitalism want to denounce this self-styled communist and complain about the harm done to the profits of law-abiding businesses like Elsevier?


Law-abiding is by no means synonymous with good, or even neutral in a moral sense. Obligatory reposting of old HN content (especially under the section "Is Elsevier really so evil?"): "Why I still won’t review for or publish with Elsevier–and think you shouldn’t either" http://www.talyarkoni.org/blog/2016/12/12/why-i-still-wont-r...

I'm also discovering that law-abiding is as much a reputation as a technicality. Companies get away with breaking the law all the time. They just aren't universally or publicly punished for it.


I'm not sure that morality would be an issue. The duty of companies in a capitalist system is simply to make money for their owners. If profit is maximised by selling fewer copies for higher fees, that's what they are supposed to do.


Please delete this irrelevant and incorrect widely debunk myth.


Very few of the misdeeds that Elsevier are accused of seem to be illegal. There's a claim that they publish copyright violations in some cases.

It seems like people are saying that there's some sort of capitalist code of conduct and Elsevier isn't living up to it, so it's not the fault of capitalism if the current situation sucks.


If you’re committed to ignoring substance in favor of rhetoric, you need to work on your rhetoric. A lot. I doubt that many here will be interested in shifty replies which fail to touch the substance of previous posts.



Even if that's what they're "supposed to do" (and I do agree to some extent that that holds for Elsevier), that doesn't make them a Good thing. Doesn't make them evil, either, but when any alternative is likely to be better, I won't shed a tear if they go away.


>>The duty of companies in a capitalist system is simply to make money for their owners

Shit. I am an owner of a company that makes millions of dollars per year and that is not how I run my business.

Hmm. I must have missed the ownership memo.


I would hope that supporters of capitalism aren’t deluded enough to believe that this system of cronyism and protectionism isn’t capitalism. Having called out the false premise underlying your challenge, they’d go about their lives.


The only protectionism that I can see is the copyright system itself, which is the essential element that makes scientific journal articles "capital" and subject to capitalism. Cronyism is just a normal part of business deal-making.


A capitalist would only complain if a government was the pirate. An individual pirate (group) is just a disrupting competitor to one type of business.


You are conflating "capitalist" with "anarchocapitalist"


Not really, since the system has been set up with the journal articles designated as intellectual property, unauthorised access to that property is treated as theft, regardless who is doing it.


That's orthogonal to capitalism, which is simply the economic activity of extracting rent from a resource ("capital") the capitalist controls.


If articles weren't treated as capital, there'd be no place for capitalism as far as I can see. The articles would just be hosted on an archive.org / Wikimedia / arXiv - style website, and Sci-Hub already shows it can be done on a shoestring budget.

But that would be communism (I suppose) and I was asking if anyone was defending capitalism in this instance.


Styles of economic structure have nothing to do with how research is disseminated.


I think capitalism is often pernicious but I'm opposed to "stealing" to a fault. You could consider this theft (and the big journals obviously do). Whether that theft is justifiable or the moral equivalent of stealing more concrete stuff is up for debate.

I'd imagine there are a number of scholars who don't agree with their papers being used this way, so there's certainly another side to it.


I would never consider it theft, because I think the conversion of information into property and the criminalisation of copying is fundamentally misguided.

It does of course enable certain capitalist business models which involve controlling a scarce resource, even though in this case the scarcity is artificial and threatened by the likes of Sci-Hub.


"Information" is simply the name for products with 0 marginal cost. If marginal cost is 0.001$, is stealing it OK? $1? $100?

Marginal cost is not the same as average cost. If it's OK to steal something that has high fixed cost but 0 marginal cost, then surely it's OK to steal groceries, or houses, which have low fixed costs and high marginal costs, yes?


I don't consider it possible to steal information, because when I make my copy it doesn't deprive anybody else of their own copies. That would be the case even if I had a ridiculously expensive Internet connection and it cost me $1 to download it. Physical goods can't be copied in that way.


> I don't consider it possible to steal information, because when I make my copy it doesn't deprive anybody else of their own copies.

According to that definition, it is possible to steal information: just delete their copies after making yours.


I'd imagine there are a number of scholars who don't agree with their papers being used this way

Then shouldn't they choose another journal to publish their paper in? Maybe open access?

Why don't they do that?


Because their industry's employees hire people based on publishing records in certain "reputable" journals.


Hmmm... it sounds like these journals do offer something that other journals don't - high impact, reputation for high quality content.

Is that why the journals charge money for access?


Who's the thief? and who are they hurting?

I have a response but that comment reply will cost 0.000000001BTC


Can somebody explain the exact process for free access that people would propose? Most options seem to come with significant and foreseeable 'unforeseen consequences.'

For instance if the idea is that research is submitted to a private journal, as typical, but then the journal is not allowed to charge a fee then that is going to rather lower their prioritization of submissions from public research.

A far worse idea would be government covering the journal fees which would do the exact opposite and overly incentivize public research as companies could send their publication fees through the roof and still have them paid due to government price insensitivity.

Maybe another idea would to have a public government research journal where all research that received public funding is freely available. But this also runs into many problems. One would be that the fundamental point of a journal is to work as a filtering mechanism. We might argue that a lot of mediocre science gets published today, even in more reputable journals. And that's after some odd 80%+ of papers, for those more reputable journals, is rejected. A government clearing house would lose its purpose as a quality filter. And it would also run into the same problems as #1 if we then have the authors submit it in the private industry, where publishing rights/exclusivity are typically part of the model.

So what's the idea?


The problem is solved, save for entrenched incumbents.

Open Access: PLoS https://www.plos.org/

Closed Access: (P)PNAS https://www.google.com/search?q=ppnas http://andrewgelman.com/2017/10/04/breaking-pnas-changes-slo...

Closed Access journals DO NOT serve as a quality filter. The unpaid peer-reviewers serve as a quality filter (and often a poot one anyway).


Despite publishing in PLoS journals a fair amount, I'm not a huge fan of the actual customer experience of publishing on their platform. I once seriously gave up first authorship on a paper in exchange for not having to deal with PLoS Currents' submission system.

Similarly, the BMC-system appears to be having serious issues with turning around papers swiftly enough.


Of course you're right, but the review process is connected to the journal. PLoS journals are very accepting and so they end up being seen as a repository for publications that could not get published elsewhere. That's not to say nothing of value is published there, but rather that if you take a random paper from one of their journals, it's going to be of a different quality than a random paper from the Annual Review of Astronomy and Astrophysics.

And like others have mentioned, a supplement is a great idea because of publication bias against things like negative results. However, I would again see this as supplementing and supplanting when I think most people here are gunning for the latter.


I wonder if university libraries around the world could federate their efforts and take on the burden of publishing and hosting articles. The money they currently spend on subscriptions should more than cover the expenses.


Here's one thing I don't understand about the call for open access.

Academic papers are not read by the general public because they are dense and usually very technical, accessible to researchers in that field but not really to many others.

Given this, the subtext to arguments for open access that journal fees are stifling scientific awareness or potential research by depriving the public seems incorrect.

Because a relatively small specialist community actually consumes scientific literature, I don't see paid journals as some egregious societal inefficiency.

Universities make so much money nowadays that it's certainly a travesty if researchers have to pay for access to these journals themselves, but the point remains that it just doesn't seem like that big of a deal that the public doesn't have free access to scientific journals.


> Academic papers are not read by the general public because they are dense and usually very technical, accessible to researchers in that field but not really to many others.

You underestimate how quickly a motivated person with a solid general education can start to understand specialized texts. In my case, I have recently been diagnosed with a chronic disease that changed my life, and that most of the doctors in my area are surprisingly ignorant about

Thanks to sci-hub, I've been able to access the latest research about my condition, and so have helped my doctors understand and offer me some of the latest therapies (in theory, that should be their job, but most doctors seem fairly uninterested, or else just too busy to care). I've been unable to work much since getting sick, so there's no way in hell I would be able to afford $30-40 a pop for journal articles. Besides, I'm already paying for the research in the form of 20+ years of tax payments.

And there are many people like me out there.


Because 1) I paid for it, and denying it to me is as stealing an unused item from my house and 2) laypeople DO use scientific papers. I read the paper about 40Hz light being applied to mice with Alzheimer’s and used it to built a light-flickering headset to try on my mother. Open access is just basic morals for taxpayer funded research.


You paid for the F-35 too. Do you deserve a ride in one?

Are you concerned that your unsupervised human trials might harm your mother? Your anecdote makes a point in favor of restricting access.


I was not concerned about harming my mother as she was dying of Alzheimer’s and used to be a scientist — she wanted to try it herself.


How come only people at universities should get access to knowledge?

$30-40+ per article is, effectively, no access. It's ridiculous. Well, it's access for the wealthy and academia I guess.

I read about a lot of stuff, many different subjects, and a lot of the papers by the people I want to read, and articles mentioned in other books and papers, are of this $30-to-view kind. It sounds like that never happened to you? I think this black-and-white elitist 'researchers vs public' idea is a silly model of reality. Even someone who's never read any papers in their life, may one day have some medical, legal, historical, artistic etc need or desire to become better educated in some area, which is prevented. Or was, before SciHub. But that's not an 'egregious inefficiency' so never mind?! That's a long way from As We May Think.


the journal industry (NPG, elsevier, etc...) are the very definition of rent seeking. They add basically no value on top of the work.

Also the very idea that it's ok to gatekeeper off knowledge that was paid with public funds because you think the general public won't be interested in it is ridiculous and wrong.

LOTS of people regularly read scientific papers - many people in technical fields, people who studied an area of research but are not active researchers, and so on.


I don't think its right to say they add no value. They definitely don't add value anywhere near what they charge (clearly based on their profit margins) but they do provide a service (albeit often minor). They coordinate reviewers (who they don't compensate), certain journals (Nature and Science in my experience) provide typesetting and editorial service beyond just basic copyediting, they provide a website to download the content.

I'm in no way saying these services couldn't easily be replicated at a much lower/negligible cost but they do provide a service.


> They ... provide typesetting and editorial service beyond just basic copyediting.

Not necessarily. For-profit publishers today – perhaps more often than not – expect the unpaid volunteer editors of the journal to do any typesetting or copy-editing, and then just send the publisher a camera-ready PDF to print.


Minor correction: many publishers don't even coordinate the reviewers. That's usually left to editors. Editors are generally unpaid, or are paid some trivial insignificant amount.


Additionally, I am a member of several societies who have outsourced their journal publication to a commercial publisher, and who use the money from those contracts to pay for society activities like travel scholarships, etc.


While I am no longer working in Academia and don't have access to journal subscriptions I'm still able to easily read papers in my discipline (Materials Science) and many others across a wide range of related physical science fields. Granted, PhDs are a minority of the general public, I'm still a member of the public and I'd like to read journal articles that come about as the result of my tax-funded research grants.

Other smart people, with some dedication and effort could learn enough to read scientific journals. People learn to read music or learn other languages. If you wanted to learn science you should have access to the works you helped fund.


There are plenty of researchers who use sci-hub, and I personally know several doctors who were beyond ecstatic when I told them about sci-hub.

So it's not just pedestrians who have problems accessing the most recent research information.


Am a researcher, and I love sci-hub. My institution doesn't have access to Elsevier journals (bless them!), so I happily sci-hub all of their articles.


> Universities make so much money nowadays

It might be true in some countries but that's certainly not the case everywhere.


And even if it were the case everywhere just because they make money doesn't mean some other private company should exist to siphon some of it away we should figure out how to cut back the universities so people can afford it.


My taxes go to funding a lot of this research. I am paying for it already, so why do I need to pay for it again?


In an abstract sense (meaning, not coupled to the current economic situation), the journals provide archival services.

This may include scanning old publications, OCR'ing them, and making them available for search and retrieval; a service which wasn't expected when the research was done.

This may include continuing content curation. I have sent corrections to a journal because their system contained incorrect information, such as in incomplete title. In one case they combined two letters to the editor into a single DOI entry, and needed to break it into two. They have staff which can handle those corrections, and those staff need to be paid.

They may also provide addition search capabilities beyond the free ones available from, say, Google Scholar, and they may update the interfaces for improved usability and for better integration with tools like citation managers.

While I don't see how those activities justify the current high prices, I disagree with the idea that public funding for the original research means there should be no additional funding for continued archiving and access.


There is a realist argument and the argument that wins support.

Realistically, most people won't touch scientific papers. You might get a few people who can wrangle up incredible things from them, but those people would have succeeded in any environment. Most research will continue to be done at major institutions like companies and universities.

But that appeal to the heart is a winning argument, and it's not wrong. Open access in the end is a good thing, even if the arguments for it are stretched, and they are stretched. I don't trust a lot of my own colleagues to bulletproof their papers (get in a graduate course, you'll see how many holes are in most research), the general public with no scientific training is at best limited and at worst dangerous.

It is a natural step, though. Your's won't be a popular question (do we really have to downvote the guy for asking a question, even if you don't like that question?), but it's a good one.


I'll never read my car's technical manual, but I'm damn glad my mechanic is legally permitted to. No stretch needed.

https://www.gnu.org/philosophy/free-sw.en.html


Thanks for the understanding. I just wonder sometimes, if open everything is always good. I get that complete openness is part of the zeitgeist on this site and within the dominant liberal western social and political philosophy, but openness can sometimes lead to unintended negative consequences.

I believe that specialists in a field are the ones who will actually make use papers. Specialists should have access, that's something I can get behind, but the general public as you say won't make use of research.

I understand what you're saying about appealing to emotion though.


Closedness usually leads to unintended negative consequences, as well-evidenced throughout history. Openness sometimes leads to unintended negative consequences.


Open access isn’t just about making papers available to the general public, it is also about making papers available to students and scholars in the developing world, whose institutions definitely don’t “make so much money nowadays” and often do not subscribe to a wide range of publications.


https://whoneedsaccess.org/

Also consider that of every 200 PhD students, only one ends up with tenure. That means there are a lot of former academics out there.


I wonder if academic papers are usually dense and non-accessible because they are not opened up to the public. Perhaps more effort should be made on the part of the academic to make their findings more accessible without compromising on the science. And this is coming from an ex-academic who have published dense and non-accessible papers.


Will this happen? in a paper about path dependence or international relations, you don't have space to make your argument understandable to people who don't know the basics of the field. As it is, a lot of papers take serious trimming to get into publication.

Many arguments that have wide appeal do end up published as books. no disagreement on your goal, but is it something realistic for much academic work?


Doubtful, you often have very little space to work with and have to get your point across succinctly. You don't have space to summarize the field to get a lay person up to speed adequately to understand the findings of the paper. Some journals, like PNAS, now require a paragraph describing the significance of the work in addition to an abstract that summarizes the work. The significance paragraph is supposed to be aimed towards understandability by a broader audience.


> you often have very little space to work with and have to get your point across succinctly.

Why are century-old obsolete technical limitations taking priority over clear and effective communication?


I don't think eliminating length restrictions would be beneficial even if they are no longer necessary. You can already basically have as much room in a supplemental/supporting section in many journals to include extra data/methods if needed. But if you extend the main article people will ramble on and in my opinion it will make it harder to read the articles for the intended audience (other scientists) with probably only marginal benefit for a broader audience.


Maybe a larger portion of the public would read academic papers if they had access to them.


"Universities make so much money nowadays" - There are a huge number of public universities in the U.S. running on razor thin budgets.


That sounds like something medieval clergy would say. Prepare for Renaissance.


I'm a guy starting my own deep learning startup in my living room. I'm not associated with a university, and I barely make enough money to live off of. There's numerous papers that are relevant to what I'm working on, and at $30 a pop, that adds up quickly. Furthermore, the abstract for a paper is sometimes so vague that you don't know if it is relevant until you read half way through it.

Also, I am diabetic. Pretty much everything in pop science says to eat complex carbs, and avoid fats and simple sugars. There are a number of papers that talk about the benefits of a low carb and high fat diet, but they're behind a paywall.

So it's more than just a small specialist community that consumes scientific literature.


These are weak examples. AI papers are published on arxiv.org long before they are published in journals (by then they are obsolete)

Diabetic "pop science" says plenty about low-carb high-fat diets, the most famous of which may be Gary Taubes's New York Times article from 2002.

http://www.nytimes.com/2002/07/07/magazine/what-if-it-s-all-...




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: