We always like to get a historical perspective on how scientists have tried to correct the record, such as this attempt in 1756 to retract a published opinion about some of the work of Benjamin Franklin. Although that 18th century note used the word “retract,” it wasn’t a retraction like what we see today, in which an entire piece of writing is pulled from the record. These modern-day retractions are a relatively recent phenomenon, which only took off within the last few decades, according to science historian Alex Csiszar at Harvard University. He spoke to us about the history of retractions – and why an organization like Retraction Watch couldn’t have existed 100 years ago.
Retraction Watch: First of all, let’s start with something you found that appears to break our previous record for the earliest retraction – a “retractation” by William Molyneux of some assertions about the properties of a stone, published in 1684. Could this be the earliest English-language retraction?
Alex Csiszar: No, for many reasons. First, I found this in a few minutes by searching under a variation of the word retraction. I’m certain that if I spent more time, I could find even earlier examples of similar notes, going much further back.
But this history is so much more fascinating than what any keyword search or Google Ngram will tell us. There’s an important distinction to be made between what this 1684 note is doing and modern-day retractions, and this difference tells us a lot about how our ideas about research ethics have changed. For a very long time, authors published similar notes that retracted an opinion or claim they made. The notes basically read: ‘This particular claim I made is wrong, so I’m correcting that.’ But there was rarely a thought to cancelling or pulling an entire publication. Keep in mind that we are talking about a time when writers often had little control over how their works found their way into print, so much so that our modern idea of an “author” solely responsible for a publication barely existed.
A very early and important forerunner is a book called Retractationes written by the fifth-century theologian Saint Augustine near the end of his life. He’d long wanted to write a book that collected together everything “which most justly displease me in my books.” The right modern word to translate the title is probably Revisions or Reconsiderations.
RW: When did this notion of “retraction” start to take hold?
AC: While I think the phenomenon is very old, in the 18th century we do see a change with a big increase in the number of periodicals focused on philosophy, literature, and even science. Most of these periodicals were filled with reviews and critiques of books, and they were part of a much bigger conversation that was also happening through oral discussions in salons or coffeehouses, and even in letter-writing. Participants called it the Republic of Letters. Criticism was an especially cherished activity, and people saw all these new periodicals as the ideal venue in which to practice it. Just as it was becoming routine to critique or revise the claims of another writer, it would be relatively common (well, maybe not quite as common) to admit when one’s own earlier opinion was in need of revision or even “retraction” – it was all a contribution to the evolving debates. Out of this emerged a big Enlightenment idea that “knowledge” progressed and grew through critique and revision.
RW: But even if those notes sometimes included the word “retract”, they were still closer to what we’d call corrections today, yes? When did authors start trying to retract entire publications?
AC: That’s true. I don’t think you’ll find very many cases of entire scientific publications getting retracted until the 20th century. It was not long before the beginning of the 20th century that scientists began to start believing in a body of works called “the scientific literature,” and that this was so important that it should be safeguarded in some way. Some started to think that scientific knowledge belonged in designated journals that were different from other periodicals, such as newspapers, or even popular science journals. But retracting something was relatively difficult to do – everything was obviously in print, which makes it much harder to remove from circulation or mark as invalid in an effective way. Once something was printed, the author and editor basically lost control of it, and it became a fool’s mission to try to remove it. There have occasionally been examples of authors disowning a book and even physically trying to take one out of circulation – including one infamous case of an author of a book on fossils who found out he had been hoaxed – but they’re rare. States and large institutions have always had more luck with this – the many indexes of prohibited books compiled by censors in past centuries bear an important resemblance to modern-day lists of retracted articles.
RW: Can you think of one early example of such a scientific retraction, where authors or publishers try to remove an entire publication?
AC: Sure. One occurred in 1889, when a new journal, Acta Mathematica, advertised a prize, sponsored by the King of Sweden, to someone making a significant discovery in mathematical analysis. It was won by Henri Poincaré, an up-and-coming French mathematician, for a paper on the famous three-body problem. But just after the journal began distributing his winning paper, a sub-editor of the journal contacted Poincaré to inquire about a confusing point. Poincaré very quickly realized that the small problem was actually a huge mistake. He sent a letter to the editor, who was really upset – the prize had been a way for the journal to promote itself throughout Europe, and publishing an incorrect essay would have been terrible publicity. Not very many copies had been sent out, and so the editor tried to collect them all back, writing to mathematicians begging them to return the paper (and to keep quiet about it). The editor was successful enough to save the journal’s reputation, and in fact a radically revised paper published a year later is what established Poincaré’s European fame. It’s still famous today as a forerunner of chaos theory.
But this is really the exception that proves the rule – like I said, publishing a flagship paper with a major mistake looked really bad for a newer journal trying to get noticed. For your everyday articles in established publications, while you may find savage critiques, rebuttals, and some sheepish corrections, you don’t find many instances where authors or editors tried to retract them in our modern sense.
RW: When did retractions – as we know them – start to become more common?
AC: For that we need to look further ahead in time to the decades following World War II, and especially the 1960s and 1970s. This is also about the time that people began to see editorial peer review as an essential requirement of scientific journals, and I think this is no coincidence. In the United States, the perceived successes of physics in the war brought about a big change in the relationship of scientists to the state. They argued that the war showed that large-scale public funding of science was crucial to the welfare of the nation, but also that control over such funding should be in the hands of the scientific community.
Spokespeople for science wanted to show that they were trustworthy and accountable to the public while maintaining their autonomy. The scientific literature became a potent symbol of the public character of science, and formal procedures such as peer review and retraction have become public badges of accountability. In this climate, retractions eventually spread like wildfire.
Before then, retractable science wasn’t often publicized – specialists in their field likely knew which papers were no longer valid, but didn’t feel the need to make that information public. The relationship between scientific authorship, career advancement, and (often public) funding sources has changed that. Today, the shift to digital publishing in science has brought renewed attention to the “public” in scientific publication. I think that’s where Retraction Watch and other public watchdog organizations come in. But it’s not simply the technology that’s changed. Today, the specter of fraud in science has become a common motif in public engagement with contemporary science. This might be because there really is more scientific fraud now, but what is certain is that what counts as fraud has changed a great deal.
RW: In what other ways were retractions different historically than they are now?
AC: For one, the reasons for retraction have changed dramatically. Until the 20th century, for example, authors self-plagiarized constantly – i.e., allowed duplicates of their work to be published, just to get the word out. This was not usually seen as a bad thing. The idea that there should be just one copy of something out there is relatively new. And of course, today, duplicating your work is considered a retractable offense!
We might imagine the scientific literature as a partial record of the ideas, claims, and even the wrong turns that scientists have made over time. Another way to see it is as a record of well-tested discoveries that are, by and large, correct – a storehouse of facts, the canon law of nature. For very long the first idea was more prevalent, whereas recently the second idea is probably more dominant. The literature has just become such a powerful public emblem of scientific truth and integrity. Earlier I brought up the indexes of prohibited books of past centuries as a point of comparison. I think that is quite apt. If you want a real precursor to the reverence with which many now regard the scientific literature, don’t look to earlier science – look to theology.
Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy.