Last year I wrote a column in response to a letter from a group of eminent academics bemoaning the current state of peer review. “Support for research that might lead to major new scientific discoveries is virtually forbidden nowadays,” the letter claimed, “and science is in serious danger of stagnating.”
I explored the problems of the system and what could be done to fix it. I looked forward but didn’t think to look back. I’d assumed, like many others, that the current system had always been so. I hadn’t realised that the first use of the term peer review was in 1969.
This took me by surprise. Peer review has such a timeless and noble aura that I had assumed the original instructions for it had arrived in the Ark of the Covenant, tucked down neatly beside the Ten Commandments. Fortunately Dr Aileen Fyfe of the University of St Andrews set me right on this. In September she gave a wonderful overview of the history of academic publishing at the Open Access Scholarly Publications Association conference. If you have a spare 54 minutes, do watch the recording. It will not be time wasted. Essentially, her contention is that the history of peer review is inextricably entwined with the business challenges of academic publishing and is not the long-established and hallowed tradition that many assume it to be.
Most assume that the starting point of peer review was the foundation of the world’s oldest academic journal, Philosophical Transactions, in 1665. In fact Henry Oldenburg, its founder and first editor, was more of a speculator and magpie than a robust gatekeeper of knowledge. His journal, and indeed most of the journals that followed in the next two hundred years, acted as a noticeboard for discovery and development. They made no pretence to judge the scientific accuracy of what they presented or even to seek out original research. Instead they were recording debate and discussion for the benefit of those who had not been able to make it to the actual meetings.
When the Royal Society took over Philosophical Transactions in 1752 it set up an editorial committee for the journal. The committee’s role wasn’t so much to provide peer review as to represent the society’s interests and to ensure that nothing published would damage its reputation. Votes were taken on whether to publish a piece, but these were based only on the paper’s abstract. What they were voting on was not the intellectual truth of a proposal but whether or not it would embarrass the society.
The situation was a little different across the English Channel. The French Academy of Sciences assessed the merits of discoveries through sub-committees who would look at the science and even try to replicate the experiments and the findings. This did not last; it was abandoned in the early nineteenth century for being too laborious and time-consuming.
Both these models opened up research to external assessment but it was only in 1832 that the Royal Society sought independent reports, signalling the start of what we would later call peer review. However, the practice was not widespread; most independent journals relied on the all-powerful editor to decide whether to publish, without recourse to any objective validation of the science. At the time, as printing became cheaper and more widespread, there were actually more academic outlets than there was research to supply it, and the editor’s role was to drum up trade rather than stem a flood of papers.
Thus, peer review was not seen as an essential part of academic publishing early on, persisting well into the 1950s. In her talk Fyfe mentioned that throughout this period the journal Nature was happy to publish research so long as it was “not outright wrong”. It only made refereeing compulsory in 1973, partly due to the sheer weight of research produced. After the war changes in scientific methods and increases in the numbers of researchers were huge. Between 1950 and 1970 the number of PhDs awarded in the States more than quadrupled, and there were similar rises elsewhere across the West.
This increase in and of itself was not enough. Peer review cannot realistically happen without copies of the paper to review. Until the twentieth century, making copies of papers was not easy. Ray Spier wrote in 2002 that “replicate manuscripts were tedious to create and it was not until the 1890s, when the typewriter became available, that carbon papers could be used to make replicate copies, three to five at the most. These could then be circulated to a committee for examination.”
The limits of carbon paper meant that the number of copies was limited and typewriters could only cope with the most common symbols and letters. These limitations were overcome by the commercial availability of the Xerox photocopier from 1959 onwards. Suddenly, any number of copies of the original manuscript could be printed and sent to any number of reviewers.
It was a perfect storm: the need for more rigorous academic filtering of submissions coincided with the widespread availability of the tools to make it possible. Within a decade of the Xerox machine becoming available the term peer review had been coined, and the modern peer review industry was born.
This article first appeared in Funding Insight in October 2015 and is reproduced with kind permission of Research Professional. For more articles like this, visit www.researchprofessional.com
The last 2 paragraphs start with the same sentence. 😉
Ah! Thanks Shaun. I wish I could say that it was entirely intentional to replicate the photocopiers I reference. But in all honesty I can't: it was just an earlier draft not being properly updated. All rectified now.
Peer review is now an essential quality hallmark for academic journals – I'm guessing this stems from a time when it was very difficult and expensive to do it. However, now we have the means to distribute an infinite number of electronic copies of a manuscript easily, every journal can do peer review. If they can find the academics to review.
With paper publication rates doubling every nine years (http://blogs.nature.com/news/2014/05/global-scientific-output-doubles-every-nine-years.html) we seem to have created a bit of a monster and it is no wonder peer review is creaking and high profile failures are noted.
However, now we have the means of holding truly global discourse that is open to anyone with an internet connection, is it time to move on from the flawed idea that the published paper is the final word? Some areas, such as theoretical physics have well established ways of making the initial draft the start of an open conversation (arxiv.org) on the discovery detailed within and not (bar the odd correction) the final word until someone else does work that adds to or discredits the original paper (and gets it published). This might get rid of anomalies such as high profile incorrect papers getting widely cited by others as they highlight that this work is wrong (thus fooling the metrics) as these papers would be weeded out in discussion rather than potentially waved through by peer reviewers short on time.