Friday, June 26, 2020

When the mob decides truth



The Film Mercury, 1926 (Wikicommons) – When the mob decides truth.

Until recently, it was almost impossible to remove an article from the published scientific literature. You would have to ask each university library for permission to go to the stacks and tear it out from a bound volume. Your request would almost certainly be denied.

All of that has changed with online publishing. Now, you only need permission from the publishing company, and removal is just a click away. The ease of online removal can lead to abuse, as noted back in 2005:

Before the advent of electronic journals, it was very hard for publishers to purge articles from their journals. At best, they could publish a later retraction. [...]

Now, however, with publishers controlling their own digital archives, and print copies no longer being produced, it has proven to be entirely too easy for some publishers to purge these archives of unwanted articles, much to the dismay of those who, like me, fear for the long-term integrity and trustworthiness of the published record of science and our intellectual heritage. In addition, if such materials can be removed, it often means they can be modified after publication as well.

Elsevier, for example, has removed about 30 articles so far from its ScienceDirect journal article archive, just since the year 2000, for various reasons. [...] The fear that many of us have is that individuals, corporate entities, and even governments, including ours, will begin to use such techniques to control the published record for political purposes or in order to cover up embarrassing information. (Davidson 2005)

That fear has come true with the removal of a paper by J. Phillippe Rushton and Donald Templer from the psychology journal Personality and Individual Differences. Rushton is known for his belief that cognitive ability varies not only between individuals but also between human populations. That was not, however, the subject of the removed paper. The subject was body coloration, specifically the fact that darker animals tend to be larger, more polygynous, and more aggressive. This correlation seems to hold true not only between species but also within species.

I believe such a correlation exists, but it’s not a simple one of cause and effect (see my last post). In any case, my opinion doesn’t matter. What matters is the right of all researchers to present their findings and interpretations in the scientific literature. If errors are made, others will point them out. That’s how the system works. 

Unfortunately, that’s not how some people want the system to work. Rushton had enemies, and they now see an opportunity to destroy his legacy, much of it being papers he published in Elsevier journals. I suspect they identified the above paper as the easiest target for removal, a kind of “test case.” It’s not about human cognition and is viewed with skepticism even by Rushton’s defenders, who seem to have fallen back to a defense line around his IQ work. Pauvres naïfs.

Demands for removal began a year ago, but it was really the events of the last month that made the journal give in.


My email exchange

Initially, I wasn't sure who authorized the removal. Was it Elsevier, i.e., the publisher? Or was it the current editor of Personality and Individual Differences? I emailed the latter, Don Saklofske, partly to protest this decision and partly to confirm he had been responsible. The following is my email exchange with him and with Elsevier:


Dear Dr. Saklofske:

I am writing with regard to your decision to remove the 2012 article by J. Philippe Rushton and Donald Templer from your journal.  This is an unusual move and breaks with longstanding practice. Once an article has passed peer review and been published, it remains in the scientific literature even if subsequently proven wrong. There have been a few cases of articles being withdrawn shortly after publication, but there have been no cases, until now, of an article being removed eight years later.

My personal judgment of this article is like that of many articles I read. I agree with parts of it and disagree with others. It is true that darker-colored animals tend to be larger and more aggressive, this being true not only between species but also within species. We can disagree about the causes, but the correlation is real and has been confirmed by other researchers.

I could argue this point at greater length, but I shouldn't have to. None of us has the right to sit in judgment on an article that is already established in the scientific literature. If one disagrees with an article, one is always free to write down one's criticisms and submit them for publication to the journal in question, but no one has the right to "unpublish" an existing article, however much one disagrees with it.

I urge you to reconsider your decision. You have created a dangerous precedent.

Yours sincerely,

Peter Frost

----------------------------------------------------------------------

Hello Peter... thank you for your email.  Indeed this was a difficult and challenging investigation and resulting decision that began last year but for which the controversy had been ongoing even before I became editor of PAID.  I am forwarding your letter to Catriona Fennell, Director of Publishing Services at Elsevier, who would have a much greater knowledge of the timelines on retracted articles following publication.

Sincerely

don

D.H. Saklofske, Ph.D

Editor: PAID

----------------------------------------------------------------------

Don,

Perhaps I am mistaken. Was this your decision or was it Elsevier's? In other words, who actually made the decision and who will take responsibility for it?

Sincerely,

Peter Frost

-----------------------------------------------------------------------

Hello Peter...  decisions related to corrigendums, letters of concern/warning, and retractions 'rests with the editor'!   I along with a panel of PAID Sr. Associate Editors comprised the signatories who reviewed the 'evidence' resulting in the decision to retract the Rushton and Templer article.

This was NOT Elsevier's decision; their office was consulted and advised of our investigation and actions only because they are the owners and publishers of the journal and it was important that I then understand their position on such matters re. legal and ethical guidelines. However I also thought you were also raising the point of 'time between publication to retraction' and this might be better known by the publisher of PAID and many other journals across varying disciplines. Should I have misunderstood, I apologize and withdraw my previous request to Elsevier.

Lastly,  retraction of journal articles  is not so uncommon (e.g. see Brainard and You; www.sciencemag.org › news › 2018/10 ›) and while the time from publication to retraction is usually less than 8 years, we began our examination of this paper last year (2019) following increased concerns from the scientific community, and two years after my appointment as editor.

Thank you for sharing your comments and viewpoint.

don

D.H. Saklofske, Ph.D

Editor: PAID

cc.  Elsevier: Catriona Fennell and Gail Rodney

-----------------------------------------------------------------------

Dear Dr Frost,

Thank you for your comments, we appreciate that there are a variety of views on how the literature should be corrected.

Since 2009, the Committee on Publication Ethics (COPE) guidelines (updated in 2019) have recommended retraction for cases where misconduct has taken place, but also in cases of error:
"Journal editors should consider retracting a publication if:
• they have clear evidence that the findings are unreliable, either as a result of misconduct (e.g. data fabrication) or honest error (e.g. miscalculation or experimental error)"
https://publicationethics.org/files/u661/Retractions_COPE_gline_final_3_Sept_09__2_.pdf

Elsevier journals endorse these guidelines from COPE and put them into practice, as do most major publishing houses.  Analysis by Retraction Watch, who have compiled a database of >18,000 retractions, found that at least 40% of retractions were due to error rather than to fraud:
https://www.sciencemag.org/news/2018/10/what-massive-database-retracted-papers-reveals-about-science-publishing-s-death-penalty
However, it is likely that retractions due to misconduct receive more amount of attention in the media and community.

It is not particularly unusual for older papers to be retracted, please see below some examples of retractions from Elsevier journals several years after publication, in one case a 1985 paper being retracted in 2013. More data is available, also from other publishing houses, from the Retraction Watch database: http://retractiondatabase.org/

Sincerely yours,

Catriona Fennell

Director Publishing Services
STM Journals
Elsevier
Radarweg 29, 1043NX Amsterdam, The Netherlands.

-------------------------------------------------------------------------

Dear Catriona Fennell,

I looked through the examples of retractions you provided. All of them concern papers in engineering or the medical sciences. Most of them were retracted because the same material had been published elsewhere, either by the same author (duplication) or by another author (plagiarism). There were a few other reasons:

- Paper retracted at author's request
- Fabrication or falsification of data
- Inability to confirm authorship of the paper and inability to interrogate the data presented in the paper

None of these examples resembles the retraction of the paper by J. Philippe Rushton and Donald Templer. That paper was in the social sciences, and there was no duplication or plagiarism involved. Nor do any of the other reasons apply. The reason seems to be more ideological. Am I right?

Sincerely,

Peter Frost



Conclusion

There were no further replies from Catriona Fennell or Don Saklofske. Perhaps they consider the case closed. They did prove me wrong on one point: several longstanding articles have already been removed from the scientific literature. The record is a paper published in 1999 and removed in 2019. Removal was justified on the following grounds:

Despite contact with Futase Hospital and Kurume University in place of the co-authors, who could not be located, the Journal was unable to confirm whether ethical approval had been granted for this study and has been unable to confirm the authorship of this paper. The Journal was also unable to interrogate the data presented in this paper as no records have remained of this study. This constitutes a violation of our publishing policies and publishing ethics standards.https://www.sciencedirect.com/science/article/pii/S8756328298001859?via%3Dihub

After twenty years it’s often difficult to locate the authors of a paper, especially if they are grad students. Their academic affiliation has changed or they may have left academia entirely. Even if they can be located, they may no longer have the raw data to support their findings. My PhD data files are on floppy disks. How can I read them today? And would they still be readable?

So if you dislike a scientific paper, and if its authors are no longer available, you can get rid of it by making a plausible accusation. Who is going to prove you wrong? This is another kind of abuse alongside the political and ideological one. "Science" increasingly belongs to established researchers with secure positions and access to legal assistance. Yet, historically, most innovative research has been done by individuals working alone with little institutional support. Charles Darwin was a country squire with no academic affiliation. Albert Einstein published major papers while working at a patent office. Intellectual breakthroughs tend to be made by outsiders.

Outsiders are losing their place in the academic community, especially ideological outsiders. This may be one reason why scientific and technological progress is slowing down. Indeed, such progress may sow the seeds of its destruction by creating better ways to manage information. And people.

But there’s another reason why outsiders are being squeezed out of academia. During the late 20th century, Christianity could no longer control what people said and believed, but it was still strong enough to keep other belief systems from taking over and imposing their controls. That happy interregnum is over. We’re moving into an intellectual environment where insiders are no longer interested in finding truth. They want to decide truth. To that end, they want to decide who gets published and who remains published. If you fall out of favor, they may delete all of your publications, and you will cease to exist as an intellectual entity. You’ll be unpersoned.


A few words to the journal editor

Don Saklofske,

You have created a precedent, and we’ll see more of these “removals.” I suspect you realize the gravity of your decision but feel you had no choice. Such a decision must be especially difficult for you, an evolutionary psychologist who has worked on genetic determination of cognition, impulsiveness, and empathy. Your research interests, however, have to be weighed against the treatment you’ve seen meted out to certain academics, including some at your university. Why share their fate?

So you had no choice. Anyway, someone else would have done the same thing sooner or later.
And, anyway, J. Philippe Rushton was a racist, like those Confederate generals whose statues have been torn down and taken away.

Apparently, Rushton is like a lot of people nowadays, such as Christopher Columbus, George Washington, Thomas Jefferson, and Sir John A. Macdonald. Of course, you’re not like those people either. Your name is further down the list, and it’s not a statue that will disappear when your time comes.

So remember: the more you give in now, the more you’ll have to give in later. At best, you’re buying yourself time, and not as much as you think.


References

Davidson, L.A. (2005). The End of Print: Digitization and Its Consequence-Revolutionary Changes in Scholarly and Social Communication and in Scientific Research. International Journal of Toxicology 24(1): 25-34

Rushton, J. P., and D.I. Templer. (2012). Do pigmentation and the melanocortin system modulate aggression and sexuality in humans as they do in other animals? Personality and Individual Differences 53(1): 4-8

Friday, June 19, 2020

Cleansing the scientific literature ... again



J. Philippe Rushton (Wikicommons). Eight years later, one of his studies is being “removed” from the scientific literature.



Seven years ago I wrote about Danish psychologist Helmuth Nyborg and the attempts to "unpublish" a study he had already published. At first thought, the idea seemed strange to me:

I was initially stumped by the ruling that Dr. Nyborg must withdraw his study from the scientific literature. How can one withdraw an already published study? Then the penny dropped. Most journals are now published online, and cash-strapped university libraries have been phasing out their paper subscriptions. (Frost 2013)

I and many others complained to the Danish minister responsible for that decision, and it was reversed.

Now, the same thing is happening again, in the heart of the Free World:

An article claiming that skin pigmentation is related to aggression and sexuality in humans will be retracted, Elsevier announced today. The study, "Do pigmentation and the melanocortin system modulate aggression and sexuality in humans as they do in other animals?" was published online in Personality and Individual Differences, an Elsevier journal, on March 15, 2012. The study's authors, John Rushton and Donald Templer, both deceased, hypothesized that skin color was related to aggression and sexuality in humans. It has been cited just nine times in eight years, according to Clarivate Analytics' Web of Science. (Retraction Watch 2020)

According to Google Scholar it has been cited fifteen times. Anyway, a study’s worthiness isn't decided by a show of hands, either now or in the future. Opinions change, and fringe science may eventually become mainstream. Or vice versa.

That brings me to another point. For eight years that study was legitimate. Now, it no longer is. Thanks to George Floyd.

Actually, his death was only a pretext. The mob feels no differently today than it did before May 25. What has changed is its ability to get what it wants ... with no pushback. Let’s not kid ourselves. This is a power grab by people who already have much power over the most important aspect of human culture—the flow of information. If you control a few chokepoints, you can get most people to believe almost anything.

“Power” may not be the right word. The aim is not simply to control institutions but rather to control how we perceive and understand reality. These are people who believe that ideas matter, and they want control over ideas, even in the scientific literature (!).

I don't wish to judge whether Rushton and Templer were "right." Once a study has passed peer review and been published, that judgement belongs solely to the reader. Personally, I feel they were right in some respects and wrong in others. They were right to argue that darker-colored animals tend to be larger and more aggressive, this being true not only between species but also within species:

Ducrest et al. (2008) reviewed data on over 40 wild vertebrate species showing that within each species, darker pigmented individuals averaged higher levels of aggression and sexual activity than lighter pigmented individuals, with a larger body mass, more resistance to stress, and greater physical activity when grooming. The relationship between coloring and behavioral dominance was robust across three species of mammal (African lion, soay sheep, and white-tailed deer), four species of fish (mosquito fish, guppy, green swordtail, and Arctic charr), four species of reptile (asp viper, adder, fence lizard, and spiny lizard), one amphibian species (spadefoot toad) and 36 species of bird.

In captive Hermann's tortoises (Eurotestudo boettgeri), another reptile species, Mafli, Wakamatsu, and Roulin (2011) found darker shell coloration predicted greater aggressiveness and boldness. Darker individuals were more aggressive in male-male confrontations and bolder towards humans, independent of body size and ambient temperature. (Melanin based color traits are a criterion in mate choice.)

Validation of the pigmentation system as causal to the naturalistic observations was demonstrated by experimentally manipulating pharmacological dosages and by studies of cross-fostering (Ducrest et al., 2008). Thus, melanocortin hormone levels predicted the amount of testosterone and other sexual steroids along with concomitant increases (or decreases) in aggression and sexual behavior. Placing darker versus lighter pigmented individuals with adoptive parents of the opposite pigmentation did not modify offspring behavior. (Rushton and Templer 2012)

Yes, melanin does correlate with aggressiveness, especially male aggressiveness. Unlike Rushton and Templer, however, I don't believe the correlation is causal, at least not wholly. It probably began as an accidental association: newborns are generally less pigmented, and this has resulted in a mental association of lightness with weakness and immaturity. Conversely, darkness is associated with strength and maturity. The age difference in pigmentation has been amplified by sexual selection in many species, particularly polygynous species where males have to compete against each other for access to females

We humans make the same mental association, particularly darker-skinned humans. Among them, the contrast between infant and adult pigmentation is striking:

There is a rather widespread concept in Black Africa, according to which human beings, before "coming" into this world, dwell in heaven, where they are white. For, heaven itself is white and all the beings dwelling there are also white. Therefore the whiter a child is at birth, the more splendid it is. In other words, at that particular moment in a person's life, special importance is attached to the whiteness of his colour, which is endowed with exceptional qualities.

According to the same concept, it is also claimed that a newborn baby is not only white but also a soft being during the time between his birth and his acceptance into the society. Furthermore, during this entire period, he is not considered a real person, and this may go so far that parents and society may do away with him at will for reasons that are peculiar to each social group. Having been done away with, these beings are considered to return automatically to the place where they came from, that is, to heaven. (Zahan 1974, pp. 385-387)

This mental association may have become a factor in the struggle by men for mates. In highly polygynous societies, such as those of sub-Saharan Africa, darker-skinned men would be seen as more masculine, and threatening, by other men ... and by women.  Among human populations, darkness of skin correlates significantly with the polygyny rate, even after adjusting for latitudinal variation in skin color (Manning et al. 2004). It looks like selection has favored a darker color in adult males, particularly in a context of intra-male rivalry for mates, and this selection has probably occurred in many species.

I may be wrong. Perhaps Rushton and Templer were wrong. Perhaps nobody knows the truth on this point. That's why we don't unpublish scientific studies. No one has the last word in intellectual debate, and that's how things should be.

Elsevier is aware of the taboo it's treading on:

It is a general principle of scholarly communication that the editor of a learned journal is solely and independently responsible for deciding which articles submitted to the journal shall be published. In making this decision the editor is guided by policies of the journal's editorial board and constrained by such legal requirements in force regarding libel, copyright infringement and plagiarism.  An outcome of this principle is the importance of the scholarly archive as a permanent, historic record of the transactions of scholarship. Articles that have been published shall remain extant, exact and unaltered as far as is possible. However, very occasionally circumstances may arise where an article is published that must later be retracted or even removed. (Elsevier 2020)

Until now, unpublishing wasn’t even “very occasional.” I know of a few cases where a paper was retracted shortly after publication. But eight years after? That just wasn’t done. Now it’s been done. The taboo has been broken, and we're going to see more and more "removals."


References

Elsevier (2020). Article withdrawal. https://www.elsevier.com/about/policies/article-withdrawal

Frost, P. (2012). Dark coloration and male aggressiveness: Is there a link? Evo and Proud, March 31
https://evoandproud.blogspot.com/2012/03/dark-coloration-and-male-aggressiveness.html

Frost, P. (2013). Cleansing the scientific literature. Evo and Proud, November 23
http://evoandproud.blogspot.com/2013/11/cleansing-scientific-literature.html

Manning, J.T., P.E. Bundred, and F.M. Mather. (2004). Second to fourth digit ratio, sexual selection, and skin colour. Evolution and Human Behavior 25(1): 38-50.
https://doi.org/10.1016/s1090-5138(03)00082-5

Retraction Watch. (2020). Elsevier journal to retract 2012 paper widely derided as racist. June 17 https://retractionwatch.com/2020/06/17/elsevier-journal-to-retract-2012-paper-widely-derided-as-racist/

Rushton, J. P., and D.I. Templer. (2012). Do pigmentation and the melanocortin system modulate aggression and sexuality in humans as they do in other animals? Personality and Individual Differences 53(1): 4-8
https://www.sciencedirect.com/science/article/pii/S0191886912000840 

Zahan, D. (1974). White, Red and Black: Colour Symbolism in Black Africa. In A. Portmann and R. Ritsema (Eds.) The Realms of Colour, Eranos 41 (1972): 365-395, Leiden: Eranos.
https://doi.org/10.2307/1573594