Imagine logging on to your computer in the morning and being presented with a list of 10 requests for your help. The requests have been filtered for you by specialized software from millions of applications filed overnight by scientists around the globe. There is no obligation to reply, but one particular question from a materials scientist in Hungary catches your eye. That researcher is seeking to develop a new crystal but is facing an unexpected hitch concerning the way particles diffuse on a particular lattice structure.

This just happens to be a subject that you – a condensed-matter theorist in California – know like the back of your hand, so you reply with the outline of a possible solution. In breaking the materials scientist’s log-jam, you feel gratified in helping to move science forward, having tackled a problem that would have taken them weeks to solve. Meanwhile, there is also the enticing prospect of developing a long-term collaboration with your new-found colleague and perhaps even writing a paper together.

This not-too-unlikely vision of the future is painted in Reinventing Discovery by Michael Nielsen as one of the exciting possibilities of “open science”. Nielsen, a physicist by training, originally worked in quantum computing and information before quitting research to develop new tools for scientific collaboration. He and other advocates of open science believe that researchers could benefit enormously if only they used the power of modern communications technology to share ideas, data, papers, results – everything, in fact. By collaborating more often and more creatively, scientists could crack problems faster and gain unexpected new insights through what the author dubs “designed serendipity”.

Nielsen’s idea of parcelling out requests for help to others with more time or expertise has already proved its worth as the mechanism underpinning a number of successful “citizen-science” projects. These include the massively popular Galaxy Zoo, in which members of the public are invited to view images of galaxies obtained by the Sloan Digital Sky Survey (SDSS) and classify them as either elliptical or spiral – a task at which computers are notoriously bad. So far more than 250,000 people have contributed to the mammoth task of analysing the SDSS’s 930,000 images, and their efforts have spawned some 25 genuine scientific papers.

Unfortunately, collaborative projects such as Galaxy Zoo are very much the exception in science, rather than the rule, because scientists are a notoriously conservative bunch when it comes to adopting new communications technologies. Nielsen is aware of this problem. He describes, for example, how one “well-known physicist” told him that Paul Ginsparg – the physicist who set up the popular arXiv preprint server used in many branches of physics – had “wasted his talent”. What Ginsparg was doing, the anonymous physicist complained, was “like garbage collecting” and beneath someone of Ginsparg’s abilities.

But as Nielsen points out, the even bigger problem with scientists ever fully adopting the concepts of open science is that there is often no motivation to do so. Why bother sharing your data in an online forum if it is only going to hand your rival the answers? Why spill the beans on your hard-thought-out ideas in a blog if it will merely let your competitors write those key papers that should rightly be yours? Why, in other words, should you sacrifice your career towards some nebulous goal of openness?

Nielsen is honest enough to admit all this, pointing out that “networked science” will not make much progress until scientific papers stop being the currency by which all scientific careers are judged. Publishing papers is, after all, what counts in science – it gets you grants, wins you research professorships and earns you kudos among your peers. As long as scientific papers continue to be the decisive output in science, there can be little hope of any true revolution.

Transforming science, Nielsen suggests, will be a 50-year enterprise and will only properly happen once we as a community learn to value openness and data sharing as much as we value research papers. What we need, for example, is some form of standardized way of allowing experimental data to be cited, just as we have an accepted way of allowing papers to be cited. Journal citations are a valuable and widely used way of judging research quality in quantitative terms – and a similar scheme for data is appealing and intriguing.

It is a shame, however, that the author fails to really develop this concept of a “data-citation-tracking service”. The same goes for the requesting-help concept mentioned above – it is an intriguing idea, but one that is not fully fleshed out. Nielsen’s get-out clause is that, well, we are just at the start of a long process, and how can anyone possibly know how things will develop or predict what different communities of researchers will want? But given that he has spent several years writing the book, Nielsen ought to be as qualified as anyone to make a few predictions – and he could have been more clear-cut in saying exactly what scientists should do next.

Indeed, on several occasions, just as the book appears to be coming to the boil and some great insight seems imminent, Nielsen pulls his punches. No, scientific blogging is probably not going to transform the world. Citizen-science projects such as Galaxy Zoo are important, but “it’s not obvious whether they’re curiosities or harbingers of a broader change”. Creating an open scientific culture seems to require “an impossible change” in how scientists work.

Nielsen also misses an opportunity by not saying more about how online tools can change the very nature of scientific explanation and the ways in which knowledge is constructed. Some areas of science, particularly astronomy and biology, are amassing so much data that computers can extract information that would be impossible for humans to deduce. If we can obtain purely statistical models of complex phenomena, does one still need traditional theories and hypotheses? Can computers reveal deeper truths than humans can? Nielsen raises these questions but does not really offer answers.

One thing that Nielsen is clear about, however, is that open science will never fully progress until funding agencies insist on it. After all, they control the cash that researchers need and pretty much have scientists eating out of their hands. “If the grant agencies decided that as part of the granting process, grant applicants would have to dance a jig downtown,” Nielsen jokes, “the world’s streets would soon be filled with dancing professors.”

Funding agencies have already made some progress in forcing researchers to deposit preprints and new papers in open-access databases, but Nielsen complains that further efforts at opening up the scientific literature are being hampered by dastardly journals publishers. However, in criticizing publishers for charging subscription fees, he (like many scientists) largely ignores the fact that the filtering, archiving and peer-reviewing that publishers perform or manage is not cheap. In particular, much time, energy and money goes into dealing with papers that do not meet a particular quality threshold and so will never appear as a “product”. Any open system that seeks to replace traditional subscription models will have to address this challenge. Nielsen, by and large, does not.

The author concludes by saying that he wrote this book to light “an almighty fire under the scientific community”. I am not quite sure, though, that it will really set the community ablaze. Indeed, one could argue that things have not moved on significantly since the author discussed open science in an article he wrote for Physics World in 2009 (May pp30–35), and which forms the basis for chapters eight and nine of the current book. So although this is a well-written book – Nielsen’s descriptions of citizen-science projects are particularly lucid – it could have been stronger on telling scientists what to do next. Nielsen’s dream of open science is likely to remain just that for some time.