Michael Eisen's goal is to change the way scientific findings are disseminated. Most research papers today are locked behind paywalls, and access can cost hundreds of dollars per article. The general public, and most scientists, don't have comprehensive access to the most up-to-date research, even though much of it is funded by U.S. taxpayers.
"It's a completely ridiculous system," says Eisen, an acclaimed biologist at UC Berkeley, an independent candidate for Senate in California running against Sen. Dianne Feinstein (D), and a co-founder of the Public Library of Science, or PLOS, which publishes some of the largest and most prestigious academic journals in the world. These publications stand out for another reason: They're open access, meaning that anyone with an internet connection can read them for free.
PLOS seeks to break up the academic publishing cartel, and it's a leading force in the so-called open science movement, which aims to give the public access to cutting-edge research and democratize scientific progress. This movement became widely publicized after famed hacker and Reddit co-founder Aaron Swartz sought to upend the publishing system by uploading millions of articles for free; he was prosecuted relentlessly, and ultimately committed suicide in 2013.
Eisen first thought he could simply convince his fellow scientists to start uploading their work, but that didn't work because universities and funding agencies use journals as a proxy for quality. They base tenure and award decisions in large part on how many articles a researcher publishes, and on the reputations of the publishers.
To encourage a switch in researchers' thinking, PLOS's first journal, PLOS Biology, attempted to emulate what Eisen describes as the "snooty" journals such as Science and Nature, which generate prestige in part by rejecting most submitted papers. PLOS Biology became well regarded and provided a proof of concept for PLOS's model, in which funding agencies or universities pay a flat fee up front (typically $1,500, but adjusted based on ability to pay) that's then made accessible for free.
The multidisciplinary journal PLOS ONE, created in 2006, used this same model to become the largest academic publication in the world, though it's been surpassed by other open access sources. PLOS ONE puts papers through a fairly typical peer review process, but it doesn't ask editors to determine a paper's importance; the journal will publish any study that follows sound science and reports its data. According to Eisen, this model encourages more thorough experiments, rather than flashy results that aren't reproducible, and allows readers to determine whether a particular study is important and valid.
Reason spoke with Eisen at the BioHack the Planet Conference in Oakland, a gathering for DIY scientists known as biohackers who eschew traditional research institutions. They often carry out experiments in garage labs and share their raw findings on the internet in real time, a publishing model to which Eisen believes all scientists should aspire.
Eisen also discussed why scientists and universities continue to prop up the academic publishing monopoly, how scientific progress suffers from the current regime, why he's running for senate as an independent, why he beleives political parties are obsolete, and the way forward for the open science movement.
This is a rush transcript. Check all quotes against the audio for accuracy.
Michael Eisen: Labs get hundreds of thousands or millions of dollars that come in from the public. After they write up what they've discovered, if you're a member of the public and you want to get access to it, you quite often have to pay again to get access to it.
So, it's a completely ridiculous system. When the internet was invented, that should have allowed scientists to share their work completely freely with the world.
Monticello: What are the consequences for it being so expensive for the average person to access this research?
Michael Eisen: They can't get access to it if they need it for their health, they can't get access to it if it's important for their job. If you're a student at Harvard you can get access to everything, but if you're a student at a community college or a smaller university, or you're in high school, or you're just curious about something that you read about, you can't get access to the primary literature on that topic.
And that's the most visible problem, but an even bigger problem is that it makes science work less effectively. The vast majority of scientists in the world, including in the United States, don't have comprehensive access to the scientific literature. Even though it's all online.
It happens all the time that somebody is doing something in their lab, and they don't know about something that could make their work better, and push it to the next level because they don't have access to the journal in which that work was published.
And finally, it's bad for the public because it's just really expensive.
Monticello: Can you tell us about how much money it costs for a large institution like UC Berkeley, or like Harvard?
Michael Eisen: Berkeley probably spends like $30 million a year to get access. Science journals take in about $10 billion in revenue a year. That's on the order of $5000 per article. Even though all the knowledge that those articles contain was generated, not by the journal, but by the scientists who send the paper to them.
All that actual important work that's done by the journal … I mean, the most important thing they do is get those papers peer-reviewed. That's done by scientists who are being paid by universities as well. So, the journals, they're basically being paid $5000 per paper to lock publicly generated knowledge behind pay walls.
You know, they have that monopoly control because scientists give it to them, because the currency of the success in academia is where you publish your papers. It's very, very difficult to get scientists to abandon established, high profile journals in favor of what they should do, which is just posted their stuff on the Internet for free.
We give them monopoly control over the literature, and they do what anybody with monopoly control over something does, which is they charge as much as they possibly can to a captive audience.
Monticello: You began several open source journals, including what was, for several years, the largest scientific journal in the world, PLoS One.
Michael Eisen: Yep.
Monticello: But this has been an entrenched system for decades, right? So, how did you kind of approach changing it?
Michael Eisen: In 1995, 1996, you saw this big movement for science journals to go online. We first became sort of aware of its problems when another big dramatic change in science, which was the growth of the genome sciences. We were doing experiments on this massive scale where we could study the behavior of every gene in the human genome in one experiment, and we went to journals and said, "Hey, can we have the papers that you now have in electronic form, so we can do this new kind of science?" and they all said no.
We all of a sudden realized how ridiculous it was that publishers, and not the scientific community, actually owned the knowledge we generated. So, we originally thought that it would be as simple as just saying, "Hey, everybody, this system's really dumb. Let's do something different." But that didn't work.
What we arrived at was a model in which publishers still exist, they take papers, they peer-reviewed them, they accept them or reject them from journals, but instead of charging for access to those papers, when you publish a paper in the journal, you just pay a fee to the publisher to cover the publisher's expenses.
The organization's first journal, PLOS Biology, started in 2003 and was an attempt to emulate what Eisen describes as the snooty journals such as Science and Nature, which generate prestige in part by rejecting most papers submitted to them. PLOS Biology became well regarded and provided a proof of concept for PLOS's pay-up-front, open access model. The multidisciplinary journal PLOS ONE, created in 2006, used this same model to become the largest academic publication in the world.
Michael Eisen: It's now no longer the biggest scientific journal in the world, because other people saw how successful we were with this, and they've copied the model, which is fantastic.
It's not grown as rapidly or as much as we'd like, it's still only maybe 20 or 30 percent of papers are published in this way, and that's still because of this sort of legacy power of publishers, but I do think things are changing.
Monticello: So like you're saying, journals have sort of become a proxy for the quality of a researcher and their research, and so that was sort of the Catch-22 that was at the heart of efforts to change, right? That everybody wants to publish in a prestigious journal, because it shows that there are a good researcher, but the fact that everybody wants to publish in these journals means that overall research is a lower quality, because they're looking for big results, they're looking for the flashy things.
Michael Eisen: Yeah, so there's another problem … Forget about how journals are paid for, and whether their contents are freely available. It encourages people to do work that will be perceived as sexy and important when it's conducted. Often the most exciting results are the most fragile results. They're the ones that need the most kind of validation and repetition. But, if the only thing that matters is you got your paper into Nature, and it doesn't matter if six months from now everybody thinks that it's false, then you have no incentive to do the kind of work that's necessary to prove that something's important.
So, those people are penalized in the system, because they don't get their work into high-profile journals, and they don't get grants, they don't get students.
It's a terrible, terrible system, and it's perpetuated primarily by the institutions who should know better, by universities and funding agencies who are the ones who give people jobs, and tenure, and grants based on their publications.
And what's more, there is abundant evidence that this system doesn't even do what it's trying to do. It's not very good at picking out the best science. Many of the most important things that, you know, the most important discoveries in science were not published in high-profile journals, and most of the stuff that's published in high-profile journals is forgotten in a year or two, it's not significant.
The presence in these journals is not an indicator of importance, is not an indicator of quality, it's not an indicator of durability. If anything, it's actually kind of inversely correlated. And yet, we still use it.
Monticello: And there's been sort of a crisis in a lot of scientific fields in the past few years, where a lot of the seminal studies are not able to be reproduced in other, you know …
Michael Eisen: Yeah. I mean, I should say that I think to some extent this reproducibility crisis is a bit overblown. There is a problem, but the problem is not that sort of science is dysfunctional, and that most research that people do isn't reproducible. It's particularly the stuff that's published in these high profile journals. This system in which it sort of encourages people to get a flashy result, very rarely do they get results that actually transfer from whatever particular system they're studying to the things we really care about.
So, they might've made an observation that's valid in a very, very narrow set of circumstances, but it doesn't translate into being important and meaningful. So, the real crisis is that the stuff that gets the most attention and the most money, and the most sort of rewards for the people who do it is precisely the stuff that is least likely to replicate in a meaningful sense. That is, it's least likely to actually turn out to be useful.
Monticello: And another part of that is that researchers really have no incentive to publish things when they don't produce great results, right?
Michael Eisen: Absolutely. There's a counter incentive.
Monticello: So, nobody knows if you did this study and it didn't work out.
Michael Eisen: Right. The same experiments are done hundreds of times in some cases before somebody actually publishes that it didn't work.
Monticello: So, how did the PLoS journals overcome that? Because you are making researchers pay upfront, right? So, why would I, as a researcher, want to pay you to tell the world that I didn't find anything?
Michael Eisen: So, first of all, most journals will only publish things that they deem to be new and important to people in their field. PLoS One does not impose that criteria. The only thing we ask is that the work be a legitimate work of science. So, we will publish and have published many, many papers that are negative results.
And it turns out that, actually, people do want to publish those things, because it may seem like a negative result when you look from the outside, but from the inside, that's a year of some graduate student's time and if you did good experiments, and you did it well, that's something to be proud of. It turns out that actually, in a world where it isn't a huge pain to get your work published, where you don't have to convince some editor that this is important, or it's going to cure cancer or whatever, people are actually happy to publish.
Monticello: So, could you just briefly take me through the process? Like, let's say that I as an individual have a breakthrough in my garage. I want to publish this paper, I want to publish it in PLoS. How do I do that? How much does it cost? And how do you ensure quality, so how is it peer-reviewed?
Michael Eisen: So, you send the paper to the journal. PLoS One has thousands of editors. So, they'll find competent reviewers, and the questions for PLoS One are very simple. Do they explain their methods well? Are the data clear and present in the paper? Is it well enough written, clearly enough written that I can understand what they did? And most importantly, the things that they claim at the end of the paper, that these are the conclusions, do the data support those conclusions?
Monticello: But so it's a pretty typical peer review process, but you don't …
Michael Eisen: Yeah, but what's different is, unlike in other journals, we do not ask the reviewers, "Is this paper important?" They're much better at recognizing flaws in the reasoning, or the data in a paper, but the reviewers are just bad at figuring out whether something is important. We just take that decision out of their hands and let readers decide the papers they're reading are actually interesting and important for themselves.
Michael Eisen: So, you were asking about the economics. In PLoS One, there is a publication fee of about $1500. If you come from countries where research funding is much lower, we either charge a discounted fee, or no fee. Authors who don't have, you know, like if you're doing this experiment in your garage and you come to us and say, "I don't have any support for my work," we don't charge you for that.
Our goal is to try to never stand in the way of, like the finances stand in the way of someone publishing things. I should also point out, I think $1500 is too much.
Monticello: So, even at $1500, have you noticed that that has led to, or have you seen any sort of major research, or you know, things that are of really high quality coming from individuals, or small groups, or lesser-known institutions versus these big institutions?
Michael Eisen: Yeah. I mean, you know, most science isn't done at Harvard. We publish many things from independent scientists, oh we've published stuff from student groups. It's our goal to not just, you know, not put a criteria on importance, but not to decide who's a legitimate scientist. We judge the work.
I spoke with Eisen at a conference in Oakland for DIY scientists known as biohackers, who eschew traditional research institutions, often carrying out experiments in garage labs and sharing their raw findings on the internet in real time.
Michael Eisen: The biohackers are not trying to build careers, by definition, they're not trying to build careers in academia that require Nature papers, or any papers. They are much more likely to just have websites where they share ideas and information, and I think that's the way that it should go.
Monticello: Right, as long as you can get the information.
Michael Eisen: Yeah. This whole infrastructure for peer review, and editorial oversight, that all exists to serve universities, and granting agencies, and other people. It has very little utility to me as a consumer of scientific knowledge.
Mostly when I'm looking for something, I Google it, and I decide on my own if it's interesting. I don't really care where it was published. So, I actually think if we were doing things right, we would take our cue from biohackers and others, who have just basically said, "The system is dumb, let's just use the Internet for what it's good for, which is the rapid and nodeless exchange of information."
Monticello: So, shifting gears a little bit here, last few questions. So, you have announced that you are running for Senate in 2018 in California.
Michael Eisen: Yeah.
Monticello: Why did you decide to run as an independent?
Michael Eisen: I think everybody should be an independent. I think political parties are bad for the country. People should, you know, form coalitions around ideas. They're so obsessed with successes as an institution, and not at all concerned with doing the right thing.
Monticello: Do you identify with any particular political philosophy, or what are the kinds of issues that you would want to focus on?
Michael Eisen: My political philosophy is, there are big problems we face, I want to go at those problems without preconceptions of what the right answer is. I think we're so locked into this idea, you know, different parties, they know in advance pretty much how they're going to respond to every issue, and that doesn't work out.
Monticello: Right. So, what major reforms would you put in place?
Michael Eisen: So, the biggest one is, you know, we have to deal with our changing climate. My primary objective is to get into Congress, and tried to change the way we think about this problem. To recognize that, you know, I'm not a climate pessimist for example. There's so much pessimism. People think the world's all collapsing. We have a big problem, but I really believe that we can solve this problem by the right application of, you know, science, technology, engineering, agriculture, if we came at this from the point of view that this is a real problem, but it's also a problem to make people abandon all the things that make them enjoy their lives today.
Figuring out ways to capture carbon out of the atmosphere, or two, you know, capture it before we put it into the atmosphere. We know how to do some of those things, but they're too expensive. We should be investing heavily in better and more efficient forms of transportation, we should be changing the diet of the population in ways that work.
I think the most immediate area for us to think about is agriculture. Government heavily controls our agricultural system, by subsidies, and all sorts of other things. We basically have a socialists agricultural system in this country. But, we do it really, really poorly. I think we could free up the way that, you know, free up farmers to do a lot more interesting things, that give them a lot more autonomy in the decisions they may, that aren't really forced into things by the government, but we give them incentives to do things that are actually good for the environment.
Monticello: So, you've kind of identified two problems there, right? One is that we have interventions that aren't good, and the other is that we should be investing in certain technologies, and progressing things. My question is, so since you're skeptical that any peer reviewer can know the significance of something, and we've seen what happens when the government gets involved in these decisions, so why do you have confidence that you can change that?
Michael Eisen: It's a big challenge, it's a personal challenge I have. I believe in principle, that government can do a lot of good things, but almost every time I've interacted with the government, they don't do a good job at what they're doing. I think part of the problem there is that we've structured so many of these programs around sort of ideological, with ideological objectives, and not practical, empirical objectives.
You know, I'm a big believer in removing government subsidies from, you know, that tell people, "You have to grow this," and try to say to people, "Well, we need an agricultural system that captures carbon, and puts it back into the air. We will pay you if you figure out ways of doing that."
Monticello: So, it's sort of higher-level targets, like we need to reduce …
Michael Eisen: Yeah, like by far, the most effective and subversive force we have in our current universe is the free market. We have to let it work on these things. The one thing the government can do in all of these things, it just has a lot of money, and sometimes you have to spend a lot of money to do kind of far-reaching things.
I think it's one thing the government actually has done a historically good job in, is investing in new technologies.
Monticello: And how would you like to see the intellectual property regime changed?
Michael Eisen: We just have to have government-wide policy that, if the government's spending money on something, it belongs to the public. That's the simplest piece of legislation you could ever write. It be one paragraph.
More broadly, we need, you know, our patent system is completely broken. It does not work to incentivize people to develop new ideas. It's become just kind of a game for lawyers. It's just one of those things where people are so afraid of touching it at all that we end up with a system that everybody knows is broken.
Monticello: So, what's next for you, and the Public Library of Science, and just the open access science movement in general?
Michael Eisen: The next thing for me is to change the way we do peer review. The Internet makes it possible for everybody who ever interacts with a work of science to review it. We should not have a system in which all the important judgments about the value of a work are made at one fixed point in time by two or three people.
So, what I want to do over the next several years in publishing is create a system in which everything gets published quickly and freely. What we need is good ways for people to find papers that are relevant to them, and technology can solve that problem way better than journals can.
Scientists, if we make it easy for them when they're reading something to say, "Hey, I think this is a great idea, this particular part of this paper. Hey, they were really onto something, " or, "They did this really poorly," we should harness the power of the whole scientific community, who are constantly interacting with the literature in different ways, to be the new form of peer review.
PLoS was all about doing this from within the journal publishing system, but I think now the journal publishing system should go completely. We're 20 years into the Internet, and we're still acting as if the Internet doesn't exist in the way that science is published, and that's just totally nuts.
Monticello: Alright, well great. Thank you. We've been talking with Michael Eisen. He is a co-founder of the Public Library of Science, and a professor of molecular and cell biology at UC Berkeley. Thank you so much for talking to us today. For Reason, I'm Justin Monticello.
Michael Eisen: My pleasure.
Monticello: For Reason, I'm Justin Monticello.