Re: Vanishing Act

[Ann Okerson <ann.okerson@yale.edu>]

Liblicense-l readers:  The following message, though long, is of
considerable interest as a discussion of the "purged" articles of which
there has been recent mention.  The e-mail format coming to liblicense had
gotten somewhat garbled and in particular a few of the proper names of
authors of articles have lost accented vowels and so may seem a bit
peculiar.  We apologize for that gap, but the item is otherwise forwarded
as received from M. Lapelerie <lapeleri@voltaire.timone.univ-mrs.fr>

************************************

In the the Chronicle issue of January 10, 2003, Andrea L. Foster published
a paper about the way the Editor Eslevier is retracting some articles
already published: "Elsevier's Vanishing Act. To the dismay of scholars,
the publishing giant quietly purges articles from its database". A case
study raises several issues.

A research, performed in Science Direct on November 15, 2002, shows 38
"missing" papers in the database. And generaly, in place of the paper, we
can see the ritual text:" This paper has been removed for legal reasons."
When it was possible, that is, when the author names and the title of the
articles were mentionned in the electronic notice, I asked the authors why
their papers have been removed. In some cases it was not possible to get
in touch with the authors, because neither author name nor article title
was mentionned in the online edition: they seemed to be definitively
"missing in action".

Some of authors never answered my email. For instance, Prof.Nikitas
Assimakopoulos. Some of them answered, and the answers are very
interesting: Some papers are "retracted" just for material reasons.

For instance, the following paper was first published incorrectly.: Sophia
E. Daire, John M. C. Plane, Stuart D. Gamblin, Pavel Soldn, Edmond P. F.
Lee and Timothy G. Wright: A theoretical study of the ligand-exchange
reactions of Na+X complexes (X=O,O2,N2,CO2 and H2O): implications for the
upper atmosphere, Journal of Atmospheric and Solar-Terrestrial Physics,
Volume 64, Issue 4, Pages 443-450 (March 2002).

Timothy Wright wrote me (email of 15/11/2002): "The original paper that
was published was not correct since the Journal did not add in the
corrections we noted at the proof stage. For some reason, they were unable
to do anything except publish the "wrong" paper and then publish the
correct paper - all very strange. Thus, there are two versions, with the
later one being correct". As Tim adds:"Nothing sinister I am afraid!".

Another case of material errors. Gary N. Greenberg : Internet resources
for occupational and environmental health professionals, Toxicology,
Volume 173, Issues 1-2, Pages 145-152 (25 April 2002). The paper is
republished 5 months later."

The publisher actually goofed in font translation", writes Greenberg
(email 15/11/2002). "Every italic word was introduced with an erroneous
character and ended with another. In apology, they re-published the
article, and even allowed me to release the whole text to the web."

Another case is much more strange. It's: Undurti N. Das : Abrupt and
complete occlusion of tumor-feeding vessels by [gamma]-linolenic acid,
Nutrition, Volume 18, Issue 9, Pages 767-769 (September 2002). The author
answers me (email 6/11/2002): "Thanks for your letter about my paper. By
mistake this paper was published twice in the journal: Nutrition, August
and september. Hence, the paper that was published in the September 2002
issue was removed."

Those 3 "retractions" have nothing to do with any "legal reason". As Gary
Greenberg wrote it down: "I have no clue why they declare it to be a legal
reason." This all-purpose expression may unduly induce suspicion. In fact
why do librarians accept to pay so high subscriptions to a so careless
editor who is able to publish twice the same paper!

Some "retractions" seem to follow real misconducts.

In one case, the authors declines to answer. Klaus Bellmann and Anshuman
Khare published : A systems dynamic perspective on the development of
recycling strategy for end-of-life vehicles, Technovation, Volume 21,
Issue 8, Pages 489-499 (August 2001). The paper is withdrawn. Both authors
send me exactly the same answer: " I am sorry: it's a closed chapter and
we would not like to re-visit it". Anshuman Khare adds few words in an
answer to a second email: "I am unable to discuss this any further. It was
a delicate matter and has been handled well by the editor" (email
18/11/2002).

Another scientist is much more straightforward. Dr. Siu Yuen Chan and
Richard W. Wong have published a short paper, with the same or a very
similar title, at the same time (November/December 2001) in 3 different
Elsevier journals of the Trends series.

1) EGFR as a transcription factor?,Trends in Endocrinology and Metabolism,
Volume 12, Issue 10, Page 431 (1 December 2001).

2) EGFR as a transcription factor?, Trends in Genetics, Volume 17, Issue
11, Pages 625-626 (1 November 2001).

3) Epidermal growth factor receptor: a transcription factor?, Trends in
Biochemical Sciences, Volume 26, Issue 11, Pages 645-646 (1 November
2001).

The 3 papers are "retracted". Dr. S.Y. Chan acknowledges that the
"missing" papers have been removed for hard reasons (email 23/11/2002):
"In short, the paper was a commentary submitted by my previous student who
has left the laboratory. He put my name and I did not realise that the
commentary was extracted from someone's work until the original author
contacted me by e-mail. There had been negotiations between the editors of
the journals concerned, and the matter has been very unpleasant to the
editors as well." This case is a clear misconduct.

The misconduct and its consequences seem to leave a very bitter taste to
the authors, Bellman, Khare or Chan. And Chan adds: "It was a very painful
and dreadful experience and actually I do not want to mention it again."

But some papers have been retracted for very curious reasons.

First awkward "retraction": A. S. Argon, R. E. Cohen and A. C. Patel, A
mechanistic model of case-II diffusion of a diluent into a glassy polymer,
Computational and Theoretical Polymer Science, Volume 9, Issues 3-4, Pages
339-352 (December 1999). On November 15, 2002, this paper appears as
retracted. Asked why his paper has been removed, Ali S. Argon, Professor
at the MIT,answers (email 21/11/2002): "Thank you for this observation. I
have no idea why this has been done, and have raised the problem with the
editor of the Journal." Very strange: the authors doesn't know his paper
has been withdrawn! Three weeks later, I got the information from Argon
(email 13/12/2002): "Dear Francois Lapelerie, This is belated reply to
your questions of November 20, on the strange listing of our paper in
Computational and Theoretical Polymer Science. When I looked into the
matter through the editor of the journal and the publisher (Elsevier), it
turned out that the publisher had not realized that around the same time
we had published two fairly different papers with the same title on the
same subject in two different journals. Without looking into the matter
carefully to see if these papers were actually the same they deleted the
one in C&TPS. Thanks to your alert observation I pointed out to Elsevier
that the two papers were different in nature and requested that they
restore the C&TPS paper to the web. They have now done so. Thank you for
alerting me to this problem." This raises a very big issue.

Second awkward "retraction": Joseph B. Kadane, Elias Moreno, Maria Eglee
Perez and Luis Raul Pericchi, Applying non-parametric robust Bayesian
analysis to non-opinionated judicial neutrality, Journal of Statistical
Planning and Inference, Volume 102, Issue 2, 1 April 2002, Pages 425-439.
The answer of Joseph Kadane is really interesting for librarians. In his 2
emails (15/11/2002), he says: " Thank you for your inquiry. In submitting
my paper to a journal, implicitly I give them permission to publish it in
their journal, but not necessarily to sell the electronic version.  All
that happened is that I declined to give Elsevier the electronic rights to
my article. I did not sign their copyright form, and hence retained the
copyright. They published it anyway, and did not agree to my request not
to charge for electronic copies. Rather than distribute it without charge
to anyone that requests it,they have chosen to remove it from their
electronic system. I stand behind the paper" This is another big or even
bigger issue.

Third awkward and murky "retraction": Antonio Arnaiz-Villena, Nagah
Elaiwa, Carlos Silvera, Ahmed Rostom, Juan Moscoso, Eduardo Gmez-Casado,
Luis Allende, Pilar Varela, Jorge Martnez-Laso: The origins of the
Palestinians and their genetic relatedness with other Mediterranean
populations, Human Immunology, vol. 62, n 9, september 2001, pp. 889-900.
The case is public and seems to be much more known. But it useful to go
back to the begining because all is not as clear as it seems to be and
some arguments and expressions are misleading.

In this case, the Publisher Elsevier made the first move, when it sent his
(in)famous letter on Octobre 3, 2001, to all subscribers of Human
Immunology. I received this letter of Paul Taylor, Senior Publishing
Editor, which deserve to be cited entirely because it has not been read by
enough people:

"Regrettably it has come to our attention that an article in a recent
issue of Human Immunolgy included certain statements that the American
Society of Histocompatibility and Immunogenetics (ASHI, the owner of the
journal), the editor-in-chief, and we as Publisher found were entirely
inappropriate for articles published in this journal. Together ASHI, the
editor and we made the decision to withdraw the article as it appears in
that issue of the journal . All electronic versions of the article are no
longer available, therefore, and we have informed Current Contents of
these actions. We would like to advise you to either ignore the article in
question (including the mention of the title in the Contents listing for
this issue) or preferably, to physically remove the relevent pages.
Elsevier Science aims for the highest standards in scientific publishing
and we deeply regret any upset that has been caused by the original
publication of this artile and its subsequent withdrawal ". <sic> This
letter is very confused about the real reasons of the "retraction" and
it's difficult to have an opinion when we read it, and much more difficult
when it's impossible to read the original paper itself : because it has
been already withdrawn! The case is so sensitive that many articles were
published in The Chronicle, in Nature, and in newspapers: The Observer,
Jerusalem Post, Al-Ahram weekly on-line, etc. I got in touch with Antonio
Arnaiz-Villena and we exchanged many emails. Antonio Arnaiz-Villena was
guest editor of a special issue of Human Immunology, and all the papers
were peer reviewed: he wrote himself 2 papers for this issue. He was the
lead author of this withdrawn paper. He made no scientific misconduct. He
only added some historical background of the Palestinian history,
commentaries and maps on the actual situation of Palestinians. Some are
not true (for instance the origins of the 1st israelo-arab war), some are
true, but are not useful to understand a very specialised paper about the
genetic relatedness between Palestinians and Jews! They are true but they
are not in the right place...

Some of Arnaiz-Villena's commentaries have been later cited out of their
context and misunderstood. For instance, Arnaiz-Villena uses the words
"colonists" and "concentration camps". He agrees that the expressions are
not very well chosen. "I should have said settlers not colonists and
refugee, not concentration, camps", he adds. But who knows now where the
Palestinians are supposed to be in "concentration camps"? Many people, who
no longer have the text itself, may think that Arnaiz-Villena says
Israelians are "concentrating" Palestinians in camps. That's wrong.
Arnaiz-Villena says many Palestinians "have a refugee status and live
either in concentration camps or are scattered in Jordan, Syria and
Lebanon" (page 892 of the article). The "concentration camps" are in ..
Arab countries!

As in other cases, the authors were not aware of the withdrawing. And in
the on-line edition of Human Immunology, you just find this few words:
"Article has been withdrawn by the American Society for Histocompatibility
and Immunogenetics (the copyright owner), the Editor and the Publisher,
and will not be available in electronic format, Pages 889-900", without
any mention of the authors or of the title.

Conclusions?

1.  As Foster writes it down, "retracted" articles are not very high
figure: a very small percentage of articles published by Elsevier. This
fact seems to show that scientific information goes its way rather
honestly. But at an extraordinary cost.

2. The figure being low, nevertheless papers are withdrawn. And this calls
into question the process of scientific publishing. Referees (or
reviewers) are the essential part in this process. When a paper is
published, it has been peer-reviewed, or at least we hope it's routine for
Elsevier journals. If the Editor and/or the Publisher of a journal
withdraw a published paper, what does that means? Simply that they
disclaim the referees, or that they have chosen very bad referees.
Something is wrong with the editor or the publisher. Arnaiz-Villena's
paper has been peer-reviewed; Argon's paper has been peer-reviewed; Chan's
paper has been peer-reviewed: Editors or Publishers have to do some
inhouse cleaning , or soon, spring cleaning, before deleting papers.

The way some cases are handled by Elsevier may raise very strong concerns.
They may be isolated cases, or they may be just the begining of a
dangerous process.

3. Words are very confusing, on purpose. Elsevier uses "retract",
"withdraw", "remove", some use "purge" or "expung". The editor of Human
Immunology, Dr. Nicole Suciu-Foca (who never answered my email), uses the
word "delete" about Antonio's paper: "This paper has been deleted from the
scientific literature" (Human Immunology, vil. 62, n 10, october 2001, p.
1063). In fact, when we had only hard copy of scientific journals, papers
were "retracted" in an intellectual sense only. The original paper
remained. Another paper, or errata, or corrigenda, or answers were
published, telling what was wrong with the original paper: plagiarism,
misconduct, factual error, miscalculation. In some cases a short strip of
paper had to be pasted (glue and scissors were in the librarian toolbox,
last century). We can refer to the paper published by Sheldon Krimsky, in
Nature Genetics (For the record, vol. 10, n2, february 2002,p. 139). But
never an editor asked its subscriber to "physically remove" the original
paper. When Elsevier uses "retract" or "retracted" today, it means much
more than intellectually "retracting". It means to really "remove",
"withdraw", "erase", or "delete" as courteously Suciu-Foca says. Because
today, with electronic edition, it's very simple to "retract" a paper: as
simple as a clic. When you retract, you delete, definitively (or
temporarily definitively: See Argon).

When I told Paul Taylor I was upset by his "advise", he answered (email
10/19/2002): "Regarding your questions on my advise to "physically remove
the relevant pages", I do see that this could be regarded as an
inappropriate practical suggestion, and apologise for that. My letter had
the sole purpose of informing subscribers that a paper had been withdrawn
and should be ignored- my more practical comment about removing the pages
was indeed perhaps not very sensible, but it was only a suggestion that
you can totally ignore if you wish.  I do hope this is all a little
clearer now: our intention was only to inform you and everyone else who
reads Human Immunology (including Current Contents, who themselves
suggested placing a retraction notice in their databases) that the article
has been withdrawn and should be regarded as not having been published."

I was pleased to see that Paul Taylor was "retracting" himself his advise
(In fact, when you buy a journal, you are the owner of this journal and
nobody is allowed to ask you to tear off pages). But I was amazed to
discover a new category of scientific paper, the
"elsevirian-never-published-published-paper".

4. Who decide to delete? We don't know. Authors don't know, because their
papers are retracted and they even don't know that their papers are being
retracted (Arnaiz-Villena, Argon, for instance). The editor or the
publisher retracts. But Daviess Menefee tells (6/11/2002) us that is in
place a new " strict internal policy of review before any article is
removed and that review must include senior publishing management and the
legal department". The author has not his place in the new system.
Elsevier is prosecutor, bailiff, judge and executioner. The author should
be implicated in a new system and at least be his own lawyer. And the
Argon case demontrates that procedures are very careless. This is not the
expression of "the highest standards in scientific publishing", as Paul
Taylor writes it, in his letter.

5. Elsevier asked databases vendors to erase Arnaiz-Villena's paper,
namely from Current Contents and Science Citation Index. And it has been
done at least by ISI. Villena's paper is no longer in the WOK (Web of
Science). I asked the European Bureau of ISI what was their opinion in
that matter. Sophie Panagi (email 12/19/2002) gave an official answer from
ISI: "Generally, it is not the role of ISI to exclude or edit a journal
article--as their content is the property of the publisher and author. In
this case, it was the editor of the journal that issued the request to
remove the published article. Therefore, ISI complied by removing the
article from the ISI Web of Science database. The article was also indexed
in ISI Current Contents and the CD edition of the ISI Science Citation
Index. These products were publicly released before the removal request
was received from the publisher". Is an editor allowed to ask databases
vendors to erase the notice of a so-called "retracted" paper, even if the
article remains in hard copy in several libraries? That's dubious, but
anyway, librarians must know this fact.

6. Who owns what? Law is different in different countries. But. Papers
have been erased very simply. Why? Because, as says Kadane, "in submitting
my paper to a journal, implicitly I give them permission to publish it in
their journal". The editor or the publisher is the owner of the copyright,
and allowed to erase the paper, if he wants to.

And what happened to Kadane is very interesting. Kadane declined "to give
Elsevier the electronic rights to (his) article because he is very
generous and willing to distribute it free of charge "to any one that
requests it". And Elsevier "have chosen to remove it from their electronic
system". As Kadane says, "The only ethical issue here is the way that
Elsevier treats its authors."

I do hope that one day, all authors will do the same thing as Kadane did:
decline to give Elsevier and other editors the electronic rights to their
articles. An occasion to find out a more balanced agreement between
scientists and publishers (because the actual situation is not the result
of negociations and is very unbalanced).

7. Is this just the begining of a new era? If all scientific information
goes digital, what can happen? Let's have a look at what Sheldom Krimsky
experienced when he wanted to wright a paper about Arnaiz-Villena. He told
me (email 11/28/2002): "I would like to write a letter or commentary to
Nature. I have not be able to get a copy of the original "retracted"
paper. Do you have one? Can you FAX it to me? There is not a single
library in the Boston area with a hard copy of the journal." Is this the
future, "selon Elsevier" ?


8 . The actual system currently at work in scientific information is
awful. Science is a kind of black box:

1- You put (huge amounts of) money into one side of the black box and
2- You get scientific articles from the other side.
3- Scientific editors receive, free of charge (and sometimes ask for a fee
from scientists), all those articles (divide the huge amounts of money by
the number of papers: you will get the cost of a single paper) .
4- Scientific editors ask scientists to peer-review, free of charge (again
and of course) papers they have themselves submited.
5- Scientific editors publish scientific peer-reviewed journals.
6- The same scientists pay the same editors simply to read a hard copy of
papers they have given away gratis to editors and they have peer-reviewed
for free (A part of the huge amounts of money you put into the black box
goes straight to the editors and is a part of the cost of every paper).
7- If the same scientists want to access electronicaly to their papers,
they have to give more money to the publishers.
8- Pretty soon (5 to 10 years), when the cost of huge databases of seldom
used older references will be higher than the profit they get from them,
scientific publishers will be willing to sell the ownership ( not only
access) of those databases to scientists (i.e. university libraries). And
scientists (i.e. libraries) will pay some more money for the third time
for the same information. This money will go into the black box (see 1).
Scientific editors have invented the perpetual movement.
9-The cost of the electronic subscription of scientific journals: this
cost is always a % (more or less) of the hard copy. Why? Because it 's
apparently the best way to ransack libraries. The cost of an electronic
edition has no relation with the cost of a paper edition.

It's time for scientists and librarians to erase, purge, delete outdated
commercial editors.

To go back to the basics, librarians and scientists don't need digital
autodafe and Digital Terminatrix, who like Nicole Suciu-Foca, "deletes"
papers "from the scientific literature".

By the way, missing, erased, deleted, removed, withdrawn, expunged, purged
Antonio's paper is still available, in its original PDF version, on the
Internet, at the following address:

http://www.idesign.fl.net.au/camel_/tidbits/Being_Jewish_Doesn't_Mean_You're_Different/
(And this is extremely unlawful!!!)