The following exchange took place on the EvolDir listserv in October 2005.
To view the original exchange and related comments from others on the list,
go to the EvolDir listserv. I reproduce the unedited exchange
here solely for convenience (e-mail headers, addresses and phone numbers removed).

One additional response by me is appended at the end. I will happily add any others
I receive, as they become available.

Additional questions have also been raised about questionable data from, or conduct by, Anders Moller.

ORIGINAL POSTING BY MOUSSEAU TO EVOLDIR 4 Oct 2005

Dear Colleagues,

Below are links to recent developments in the case against Anders Moller.
The first link is a recent essay by Moller concerning his view of the
ordeal. The second is a link to the French CNRS (Moller's employer) report
conducted by an independent board of inquiry that clears Moller. In
essence, the conclusion of the impartial committee was that there was no
evidence to support accusations of intellectual dishonesty.

Moller ISBE Newsletter piece (pdf)

CNRS Decision regarding Moller case (pdf)

Best wishes,

Tim

Dr. Timothy A. Mousseau
Professor of Biological Sciences
Dept of Biological Sciences
University of South Carolina
Columbia, SC 29208 USA

RESPONSE BY PALMER 6 Oct 2005

Further to the recent developments in the case regarding data
fabrication by A.P. Moller posted to this list by Tim Mousseau, I
offer the following.

A commentary I prepared on another -- more compelling -- example
of questionable data published by A.P. Moller was rejected
editorially (without review) by Nature (Mar. 24, 2004), TREE (May 12,
2004) and Evolution (Oct. 24, 2004). It was also submitted (Mar. 30,
2004) to the CNRS committee that evaluated Moller's 'conviction' by
the Danish Committee on Scientific Dishonesty, but the CNRS committee
ignored it and chose to focus solely on the data at issue in the now
infamous and withdrawn Oikos paper.

A careful reading of the decision by the CNRS "committee of wise
men" reveals that no clear conclusion was possible -- either guilty
or innocent -- regarding scientific misconduct in the Oikos incident:
"Lacking the material evidence necessary to establish innocence, the
committee was equally unable to reach this conclusion . . . [so] the
presumption of innocence must be applied".

As peer-reviewed journals and CNRS seem unwilling to consider
additional evidence regarding suspicious irregularities in data
published by A.P. Moller, I make these data publicly available. In a
nutshell, scatterplots of data from two early papers by Moller,
reporting an association between size and asymmetry in barn swallow
outer tail feathers, appear strongly inconsistent with similar data
from a later paper that wished to draw a different conclusion. A
plausible explanation for the patterns exhibited in these
scatterplots would be welcome.

A pdf copy of this commentary, including scatterplots of the
suspect data, may be downloaded from:

Moller Data Irregularity (pdf)

Judge for yourself whether these published data are questionable or not.

Regards, Rich Palmer

A. Richard Palmer
Systematics and Evolution Group
Department of Biological Sciences
University of Alberta
Edmonton, Alberta T6G 2E9 CANADA

RESPONSE BY MOUSSEAU AND MOLLER 8 Oct 2005

Dear Colleagues,

For those of you who are not interested in this PUBLIC discussion, I
apologize for continuing it longer. I can assure you that I have no interest
in perpetuating this debate, or mediating it in any way. However, in the
face of ongoing and unjustified PUBLIC criticisms of my colleague, Anders
Moller, whose behavior I have found to be above reproach and certainly not
worthy of unsubstantiated attacks, I offer the following response to Rich
Palmer's recent posting:

Rich has insinuated that changes in tail length FA distributions among three
Moller papers (Moller 1990, 1992, 1999) reflects selective use of data (or
worse).

See
http://www2.biology.ualberta.ca/palmer/pubs/05MollerComm/MollerDataIrreg.pdf
for details.

As far as I know, Rich has not asked Moller directly for clarification
concerning possible biological explanations. Thus, for my own benefit, and
that of this group, I have asked Moller myself. His response:

+++++++++++++
Dear Tim,

I have recorded asymmetry in barn swallows annually since the 1980's.
Concomitant with a dramatic increase in tail length in males (Moller & Szep
2005) there has been a dramatic decrease in mean and variance in asymmetry.
Furthermore, the relationship between unsigned asymmetry and character size
has changed from significantly negative to completely flat, with this
temporal effect being present in males, but not in females. All this
information is contained in a manuscript that was submitted to Evolution,
but rejected because it was not "novel". I would be happy to share this
manuscript with Palmer or anybody else. Perhaps Palmer's emotional attitude
prevented him from looking for a biological explanation and from contacting
a colleague before making public accusations.

All the best,

Anders

Moller, A. P. and Szep, T. 2005. Rapid evolutionary change in a secondary sexual
character linked to climatic change. J. Evol. Biol. 18:481-495.

++++++++++++++

A copy of Moller's rejected paper showing changes in FA through time can be viewed at:

http://cricket.biol.sc.edu/dedication/moller/moller-FA-year.pdf

Another paper that may interest those following this discussion:

Direct and indirect tests for publication bias: Asymmetry and sexual
selection. - Anim. Behav. 70:497-506, 2005 (authors A. P. Moller, R.
Thornhill and S. W. Gangestad).

Sincerely,

Tim Mousseau

Dr. Timothy A. Mousseau
Professor of Biological Sciences
Dept of Biological Sciences
University of South Carolina

FOLLOWUP BY PALMER 10 Oct 2005

In response to Mousseau and Moller, and perhaps others who might
think that Moller's reply and unpublished MS provide a satisfactory
explanation, I repeat my question regarding Moller's published data.
It has yet to be answered, so I hope the question is clearer this
time.

Why, are symmetrical (or nearly symmetrical) individuals ABSENT
among shorter-tailed male barn swallows (less than 100mm tail length;
shaded areas in the graphs) in two early studies by Moller (1990,
1992), yet PRESENT in two subsequent graphs in a later paper by
Moller (1999) for the same population? For actual graphs, see:

http://www2.biology.ualberta.ca/palmer/pubs/05MollerComm/MollerDataIrreg.pdf

In the former two studies, the absence of symmetrical (or nearly
symmetrical) individuals among shorter-tailed individuals yielded
graphs showing a tidy decrease of asymmetry with increasing tail
size, which was a 'predicted' outcome in both papers. In the latter
paper, symmetrical (or nearly symmetrical) individuals were expected
to be most common at all tail lengths (not just longer ones) and were
found to be so, in sharp contrast to the earlier studies.

This crucial point is not addressed in Moller's reply to Mousseau,
nor Moller's unpublished MS submitted to Evolution:

http://cricket.biol.sc.edu/dedication/moller/moller-FA-year.pdf

both of which deal with change in average FA , and overall
size-dependence of FA, over time, not with the apparently 'missing'
symmetrical individuals at small tail lengths in the 1990 and 1992
studies.

One possible resolution to this debate would be for Moller to post
publicly (for example, on Mousseau's web site) his raw data on the
lengths of the right and left outer tail feathers of male barn
swallows from his well-studied Kraghede population, for each year,
for the years 1988 - 2004. Perhaps the patterns exhibited in the
1988 and 1990 sampling years were simply flukes. This possibility
cannot be judged merely by examining summary statistics like those in
Moller's unpublished MS.

As for public discussion, a public clarification will go further
towards putting the community's concerns to rest than private
re-assurances.

Regards, Rich Palmer

PS. To my chagrin, when double checking Moller's original 1992
paper, I discovered a graphing error in the figure I posted earlier
(two extraneous points in panel 1b -- the uppermost two). I have now
corrected the posted file. If anything, the discrepancy between the
early and later Moller data sets is now even stronger. I apologize
for this careless oversight.

o--------------------------------------------------o
A. Richard Palmer
Systematics and Evolution Group
Department of Biological Sciences
University of Alberta
Edmonton, Alberta T6G 2E9 CANADA

RESPONSE BY MOLLER POSTED ON EVOLDIR WEB SITE 17 Oct 2005
BUT NOT SENT TO THE LIST

A. R. Palmer - a divine guardian of good scientific practice, or just another human being?

Palmer has in his recent posting on evoldir stated that my responses to his
questions do not address the issues about why there was an absence of
symmetric males among short-tailed males in early samples, but many such
symmetric males in recent samples of barn swallows from my study population
in Denmark. However, I will reiterate once more for Palmer in an attempt to
convey the message. I have show in the manuscript posted on the website of
Prof. T. A. Mousseau by honestly listing all available information that (i)
tail length of males has increased by over 1.2 standard deviations during
the last 20 years; (ii) asymmetry in tail and wing length has decreased
dramatically during the same period; and (iii) the relationship between
asymmetry and tail length in males, but not in females has changed from
being significantly negative to being flat during that period. This
addresses exactly Palmer's question because it shows that since short-tailed
males (at least by the standards of the 1980's) are absent or rare in recent
years, this fraction of the population does no longer contribute to the
distribution of asymmetry. Therefore, there is no deficiency of symmetric
short-tailed males any more, because there are hardly any short-tailed males
in the population any more. Finally, any differences in individual data
points among publications, as identified by the detective work of Palmer,
are due to differences in selection criteria for inclusion of observations
in specific analyses.

Having read the recent postings by A. P. Palmer on EvolDir between my other
required activities, I have started to wonder what is the problem at stake.
Since when has a professor at a small university in western Canada been
appointed the universal oracle of all matters related to asymmetry?
Recently, A. R. Palmer has listed on the web a number of different findings
from his analyses of some of my papers. He has in addition posted a number
of questions that he apparently imagined that I had to answer in public on
the internet until he some future day felt satisfied with the replies. This
entire procedure seems utterly absurd to me because as far as I know Palmer
has never been elected over-seer in these or any other matters. He has no
legal or moral obligation or right to pursue what he is doing. At worst, his
behavior reminds me more of prosecutions in certain medieval societies, or
in certain parts of North America a couple of hundred years ago when the
same person was acting as sheriff, judge and guard of the local prison.

If A. R. Palmer had a saintly or godly behavior that made him a superior of
others, I might be able to understand his ways. However, that appears not to
be the case. When Palmer wrote a piece for the newsletter of the
International Society for Behavioral Ecology a couple of years ago, the
president of the society, Prof. N. B. Davies, wrote in the subsequent issue
that all future contributions to the newsletter would be required to reach
certain minimum standards in terms of normal collegial behavior.
Furthermore, Palmer has as the senior author of a paper published in
Evolution in 1997 reproduced four figures from four other papers without
requesting permission to reproduce these from the copyright holders. That
was done against the advice of at least one referee who recommended that he
not proceed with this illegal behavior. Hence, Palmer clearly has no rights
based on a higher moral standing to pursue what he is doing.

I would be happy to share data sets with others, as I have done hundreds of
times including with Palmer himself. However, I do not subscribe to the idea
that a single person should be allowed to conduct personal vendettas under
the disguise of promoting a fair society of scientists devoid of dishonesty
and fraud. When such activities are left in the hands of single individuals
they often become tools to meet the desires of that individual rather than a
fair and open-minded inquiry, as I have already explained in detail in my
recent article in the newsletter of the International Society for Behavioral
Ecology. We are all best left without such unsolicited actions by parties
directly or indirectly involved in a case.

Perhaps A. R. Palmer should focus on his teaching, research or some other
activity and stop harassing people by posting his thoughts and beliefs in
public on the internet. Then I and other persons being scrutinized in public
could be left on their own to pursue their everyday life at work and at home
together with colleagues and families.

Finally, I would like to invite Palmer or anybody else interested to
personally inspect all data sheets in my office at their own expense to
erase any doubt about the existence of the authentic data. I have all my
notes since October 1969 neatly ordered so they are readily accessible. In
addition, I would like to invite anybody at their own expense to join me in
the field to see how asymmetry measurements are made, to check for feather
breakage or molt, and to observe variation in asymmetry among individuals. A
visit of less than a week would suffice for most people. And that would be a
memorable week for most people. Please address enquiries and bookings to
amoller@snv.jussieu.fr

Anders Pape Moller

http://www2.biology.ualberta.ca/palmer/pubs/05MollerComm/MollerDataIrreg.pdf

This is not because birds with tail lengths less than 100mm were absent in 1999. So the
discrepancy between the patterns in the 1990/1992 papers and the 1999 paper
remains unexplained. Furthermore, as explained in detail in Rowe et al. 1997
the patterns exhibited by barn swallow tail feathers in the Moller's 1990 and 1992
papers depart significantly from what would be expected for size-dependent FA
and his 1999 rebuttal provides no satisfactory explanation for this discrepancy.

OBSERVATION 2: Moller's defense that "any differences in
individual data points among publications, as identified by the detective work of
Palmer, are due to differences in selection criteria for inclusion of observations in
specific analyses" certainly raises some interesting questions regarding what criteria
Moller used to include/exclude data, given that all of these papers simply describe
the association between tail length and tail asymmetry where, one would presume,
the criteria for inclusion would be the same.

But as "just another human being", perhaps I am missing something important here.

OBSERVATION 3: Enrollments at the University of Alberta in 2004/2005 were:

Undergraduate full-time: 27,562
Graduate full-time: 4,064
Undergraduate part-time: 2,224
Graduate part-time: 1,816
Total: 35,666

This compares to an enrollment of "30,000 students including 8000 in postgraduate studies"
at the Pierre & Marie Curie University where Moller works.
So, if I work at a "small university in western Canada", by his own criterion,
Moller works at an even smaller university in Paris, which, I hope, does not trouble
him too deeply.

Facts like these, of course, should never confound a good argument.

OBSERVATION 4: The tongue-in-cheek piece I published in the
International Society for Behavioral Ecology newsletter:

Palmer, A. R. & L.M. Hammond. 2000.
The emperor's codpiece: A postmodern perspective on biological asymmetries.
International Society for Behavioral Ecology, Newsletter 12:13-20.
(download full-text pdf file: 104K)

was reviewed prior to publication, is thoroughly referenced, and includes many
quotations from the original publications to avoid misrepresenting any authors'
statements. Even though the characters in the piece were fictional, Moller (and
Randy Thornhill) felt unfairly and unjustly criticized. Significantly, Moller and
Thornhill were invited to publish a response where they could rebut any unjust
claims or misrepresentations regarding their data or conclusions, but they elected
not to.

OBSERVATION 5: Regarding copyright infringement, the Rowe et al. 1997 paper
does not directly reproduce any figures from Moller's or anyone else's
papers. As stated clearly in the methods, the data were digitized from published
figures and replotted on comparable axes. The source of the data
was fully acknowledged for each figure.

Furthermore, by waiting eight years to grumble about supposed copyright infringement
interesting questions arise regarding why such concerns should be aired now.

OBSERVATION 6: Although trips to Paris -- enjoyable as they might be -- to
inspect Moller's original data may ease the minds of individual scientists,
they will hardly address concerns of the larger community. A public posting or
publication of tabulated raw data (or scatterplots) on the lengths of the right and
left outer tail feathers of male barn swallows from Moller's well-studied Kraghede
population, for each year, for the years 1988 - 2004, would resolve the debate more
effectively than any other single action.

Moller may do so here, if he wishes.

Curiously, since CNRS is based in Paris, and presumably had access to
"all [Moller's] notes since October 1969 [which are] neatly ordered
so they are readily accessible", one might wonder why they were

"Lacking the material evidence necessary to establish innocence"

in their investigation of the disputed Oikos paper.

A.R. Palmer Home Page, FA references
All rights reserved.
(revised March 26, 2009)