The first Western science journals in the world — Journal des Sçavans in France and England’s Philosophical Transactions of the Royal Society — will turn 350 years old next year.
The science journal has, as a communication system that owes thanks to the printing press, brought about some quite peculiar situations. According to Jan Nolin, professor of information sciences at the University of Gothenburg, many habits of today’s science can be defined as ‘paper constructs’. Although scientists have adopted digital technologies when it comes to publishing the results of their studies, they still cling to paper – or electronic forms imitating the medium. The limitations inherent to paper as a medium have limited and shaped the philosophy of how science could and should be practiced, Nolin writes. Now, by redefining the medium, we define again what ‘science’ actually means.
Australian quantum physicist and publicist Michael Nielsen implies in his essay from five years ago that the most open system for trading information in the 17th century is now holding back more efficient solutions, as paying for publications in journals, not the new media, is still the norm.
He also says that the Internet has significantly increased the part of useful knowledge created by scientists that happens to be publishable. Nielsen says the greatest change in the creative process after the invention of writing is going on right now, with scientific publishing houses turning from service providers into technology companies.
According to Nielsen, the ideal criterion for science would be extreme openness: making as much information of different kinds public, as well as readable, not only for humans, but machines as well. I studied one such experiment a bit. It might be a little too raw but it proves interesting nonetheless, and it’s called ‘open notebook science’. Time for some notes I made.
Open Notebook Science. Image credit: Matt @kwinkunks, Flickr Creative Commons
Publishing negative data
The philosophical core of open notebook science (ONS) is simple: The scientist discloses the entire process of the study in real time. He or she places all initial data on the web, both successful and failed experiments, as well as all notes and musings.
The method was probably most discussed on the Internet during the second half of the noughties. It was implied that ONS simplifies publishing so-called negative data (The experiment was carried out but it didn’t work out because…). Seeing such information could help other researchers to save time and grant money, as past mistakes would not be repeated.
In addition, there was hope that ONS would render the current model obsolete. As the success of a scientific career is quite directly related to publishing articles, there’s almost always bound to be someone somewhere who is basically doing the same thing at the same time. You don’t know the details (and don’t know the person), but you know that you must publish first, before the hypothetical competitor. That’s where luck comes in.
“Competition is useful, but it shouldn’t waste years of research time”, writes bio-informatician Pedro Beltrao on the topic.
Bradley and his open notebook projects
The term ‘open notebook science’ was coined in 2006 by Jean-Claude Bradley, a chemist at Drexel University in the United States. In 2005, he started a wiki-based project called UsefulChem. Currently, new compounds to fight malaria and HI virus, as well as cleaning up water contaminated with arsenic, are being pursued using open notebooks. As a side project, databases on the solubility and melting temperatures of organic substances have been created, with the latter currently including information about 27,000 compounds.
Many researchers possess highly organised chemistry-related information that is not public, only because no one has asked, according to Bradley.
“It is true that keeping an open notebook has not gone mainstream”, he muses while answering an e-mail from me. “But it only takes a few contributors in a given field to gain most of the benefit. For me, one of the most important indicators that our use of open notebook science is having impact is that we are approaching 1000 queries a day for solubility and melting point data. And by looking at the nature of the queries I believe that most of these searches are providing useful answers to those looking for specific information”.
In some cases, Bradley researches the background of experimenters (using public information) and tries to contact them.
“We often see searches related to environmental issues like toxicology or how chemicals are transported through soil”, he says. “The applications are always interesting and revealing about scientific information needs, and being able to provide this information in a completely open way is in my opinion by far the most efficient way to facilitate scientific progress”.
In his commentary published May of last year in the journal Chemistry World, Bradley reminisced that in the more closed scientific space of the past, trust of fellow scientists who had conducted the experiment and managed the data, but could still be mistaken, was really important. “Today, it is a choice. Optimally, trust should have no place in science”, he wrote.
The new Chemical Rediscovery Survey project was launched this summer, in cooperation with Bradley, and aims to collect as much raw data in an open format as possible about the consequences of blending of two compounds.
Open notebook science in Estonia: PlutoF
Some tinkering is going on in Estonia as well. PlutoF, the information system managed by NATARC, the joint network of local science universities, the Estonian Museum of Natural History, and the Ministry of the Environment makes it possible for researchers to manage, analyse, and publish data related to living nature on the web.
“When talking about open-access data, management is essential”, Urmas Kõljalg, academician and head of the NATARC board, opined at the Open-Access Week conference that took place at the University of Tartu Library last autumn. “Most of the times, the data management needs of the scientist are ignored. Excel is being used, as well as text files – but later, when uploading items in repository, he has to start from scratch. We don’t write our own software for using e-mail or working with spreadsheets; we use what’s already there instead. I don’t believe that the conception of open-access data will enter the picture, as long as the science does not possess common software solutions”.
In 2012, the PlutoF information system was used by over 19,000 researchers (including hobbyists) from more than 35 countries. The average time spent in the system during a day was half an hour. Estonia-centered PlutoF databases can be found in the eElurikkus project, for example.
Displaying uncertainty in science
Still, the Estonian presence on the web does not seem to possess such an ONS enthusiast as Jean Claude Bradley.
“I have been thinking about it, but as I have not been in the position of being able to decide it myself, I haven’t gone too far with it”, Jaan Aru, researcher of consciousness at UT, as well as long-time science blogger, reflects. Still, he has sometimes briefly described his half-finished projects on the web.
Careless publishing of scientific findings can bring about a severe fine by those funding the projects, Vice-Rector for Development at UT Erik Puura says. But what if the funder has no objections?
“One should ponder if the invention might be something that could be used in business, thus nourishing the Estonian economy”, Puura says. “Then again, when going for publishing, it must have a point. To me, the main point would be educating people”.
Thus, the openness of science is necessary for the public as whole. Aru agrees that persistent, on-going web-based description of projects would be a nice way to communicate with society. “In layman’s terms, science is something that’s clear, solid and somewhat out of reach. When reading these open descriptions, everyone understands that scientists go wrong too; there’s a lot of uncertainty in science as the process goes along the path, falling and stumbling”, he muses. “Scientists would become more human in the eyes of society”.
It could also add some impetus to the scientist’s own work. “At least in my experience, writing stuff down constantly can lead to discoveries about your own project that you wouldn’t have noticed otherwise”, Aru says.
Idea theft, copyright, sensitive data
According to Jaan Aru, one of the major obstacles to the spread of ONS philosophy is the fear of idea theft, shared by many scientists (something that should actually be quite quickly detected in a system such as PlutoF). Bradley has talked about ’embarrassment’ as well.
Then again, researchers met various legal obstacles, specific to their fields. Chemical substances don’t think or have opinions, but when it comes to medical and human sciences, there’s a lot of sensitive data involved. Linguists have to take into account that a talking person can be identified by the voice. The objects of humanitarian sciences – texts – could be copyrighted. These are only a couple of examples.
The Internet is full of discussions about different forms of open science. The European Commission is struggling with legal matters related to the topic, and the Estonian Science Foundation has a special work group to deal with local legal acts as well.
There has already been an opinion expressed at the Open Access Week conference at UT Library that an Estonian citizen’s right to secrecy is greater than the public’s right to get to know things. This hinders practising scientific studies not conducted by the universities as well.
The cost of open notebook science
I couldn’t find out the cost of ONS. Jean-Claude is using a variety of different (freeware) web-based tools. The academician Kõljalg was not able to estimate the costs of maintaining PlutoF at the conference.
John Wilbanks, the one-time vice president of the Science Commons project has written that in the end, the core of all discussion about open access – innovation – tends always to get buried under arguments over profits, business models, and legislation.
