The journal Geology has retracted a paper that, when it was published in May 2010, was hailed as a major step forward in understanding what happened when the Indian and Asian land masses collided millions of years ago. As The Hindu reported when the paper was first published:
Dr. [Anju] Pandey and her colleagues used sophisticated analytical techniques to demonstrate the occurrence of relict majorite, a variety of mineral garnet, in rocks collected from the Himalayas. Majorite is stable only under ultra-high pressure conditions, meaning that it must have been formed very deep down in the Earth’s crust, before surfacing millions of years later.
“Our findings are significant because researchers have disagreed about the depth of subduction of the Indian plate beneath Asia,” said Dr. Pandey.
In fact, the previous depth estimates conflicted with estimates based on computer models. The new results suggest that the leading edge of the Indian plate sank to a depth around double that of previous estimates.
“Our results are backed up by computer modelling and will radically improve our understanding of the subduction of the Indian continental crust beneath the Himalayas,” said Pandey, according to an NOC release.
It turns out, however, that as best as anyone can tell, the key data are from Norway, not the Himalayas, and were published in 1998 by another group. According to the retraction notice, which appears in the May 2011 issue (link to 1998 paper added):
This article describes the first reported record of relict majorite in the Himalaya. Subsequent to publication of the article, Geology received queries from several leading metamorphic petrologists from around the world, indicating that the key photograph (Figure 2C) in the Pandey et al. paper appears to be the same as Figure 3A from a paper by van Roermund, H.L.M., and Drury, M.R. (1998; “Ultra-high pressure (P > 6 GPa) garnet peridotites in Western Norway: exhumation of mantle rocks from >185 km depth”, Terra Nova, v. 10, p. 295–301, doi:10.1046/j.1365-3121.1998.00213.x). The image also appeared on the cover of the December 1998 issue of Terra Nova.
Geology has learned that the photographed thin section cannot be located. Until analytical results can be replicated from a Himalayan rock, we must conclude that relict majorite has not been demonstrated to exist in Himalayan eclogites.
Accordingly, Geology retracts the article by Pandey et al.
Author Contributions: A. Pandey obtained thin sections from P.K. Verma, and photographed and analyzed the minerals shown in the relevant figures. M. Leech worked with Pandey on the tectonic implications of the reported majorite, based solely on the published photomicrographs. A. Milton managed the laser ICP-MS laboratory used by Pandey, and set up the instrument for analysis of Pandey’s thin sections. P. Singh originally studied the relevant samples as part of his Ph.D. dissertation, but did not find any evidence for majoritic garnet through probe data. Verma supplied the thin sections for Pandey (Professor Verma is now deceased).
The Geology paper has yet to be cited, according to Thomson Scientific’s Web of Knowledge.
So where did the rock actually come from?
Verma’s death — shortly after the paper was published, according to co-author Leech, of San Francisco State University — makes it difficult to know. Leech, who was left trying to figure out what had happened once other researchers began asking questions, tells Retraction Watch:
Pandey cannot or will not produce the thin section in question or provide additional information to prove the texture in Fig. 2c came from a Himalayan rock. She says that she sent the thin section back to PK Verma in early 2010 for further study, and that the thin section is now lost.
As for the other co-authors: Preeti Singh was one of Verma’s graduate students, and collected the samples in the field with him in 1999 and 2004, says Leech. As the retraction notice says, he “did not find any evidence for majoritic garnet through probe data.” And Andy Milton manages the plasma mass spectrometry lab at the National Oceanography Centre in Southampton, UK, and his only involvement was to help Pandey to use that method to collect her elemental data, says Leech:
He did not have any knowledge of the manuscript, nor that he was included as a co-author, until after the paper was published in Geology.
Apparently, it was the fact that everyone worked on different parts of the paper that allowed confusion to reign:
I worked with Pandey only on the tectonic implications of majorite in the Himalaya based solely on the photomicrographs she provided to me and that were subsequently published by Geology. I have never personally seen the thin section(s) from which the photomicrographs were taken. I worked with Pandey on earlier drafts of the Geology manuscript and the earlier text and figures were similar except for the published photomicrographs. At some point prior to final submission to Geology, I suggested to Pandey that she take better photomicrographs to show the exsolution textures she was describing in the paper, particularly higher magnification images. When she provided the images of exsolved pyroxenes that were eventually published, I thought she had much stronger evidence of majorite than in earlier drafts of the manuscript. Unfortunately I, nor two expert reviewers of the manuscript, did not recognize Fig. 2c as an image (rotated 90°) from van Roermund and Drury’s 1998 paper.
To the best of my knowledge, Pandey collected all data and produced all the figures and tables herself. Because Pandey clearly used a photomicrograph of van Roermund and Drury’s Norwegian garnet peridotite, it is clear the paper is fraudulent and I now do not trust any other data published within that paper. Until and unless someone independently finds evidence of majoritic garnet in the Tso Morari Complex, we must assume that there is no evidence at all.
In other words, if this were a police procedural, the case for majoritic garnet would have to be thrown out, because at best there was a failure in the chain of custody. And in this case, the detective seems to have borrowed evidence from another case — which just happened to land on the front page of the New York Post.
We’ve tried to contact Pandey, and will update with anything we hear back.
Hat tip: John Timmer