This is an edited version of a letter to which Eugenie Reich, who has been covering the Taleyarkhan affair for Nature, has yet to respond. The letter exposes serious deficiencies in the stance that the journal has insisted on taking.

Date: 23 May 2007 11:52:53 +0100
From: Brian Josephson
To: Eugenie Reich
Subject: Purdue secrets leaked to Lafayette Journal and Courier

Dear Eugenie,


    Are you are aware that details of the confidential Purdue investigation have now been leaked to a local newspaper?  In case not, the link is at http://www.jconline.com/apps/pbcs.dll/article?AID=200770518040

I don't think Taleyarkhan's attackers will be very pleased at this leak because among other things:

(i) it shows that Purdue had good reason to clear Taleyarkhan in its first investigation; (ii) it backs up Taleyarkhan's allegation (I believe made in a letter to the NY Times that was not published) that the report was 'cherry-picked' to make him look bad.

In more detail, the investigation noted that the accusers had provided no evidence that Taleyarkhan had participated in the disputed experiment (except for advising on the paper), and in this connection Yiban Xu has stated explicitly that he had not been involved on the experimental side and that "the comments Taleyarkhan made didn't change the data or his analysis".  This I don't think appeared in the congressional submission.

This also has ramifications regarding the fraud allegation since it would require Xu and Taleyarkhan to be plotting together. That is different from cases such as those of Schon and Ninov where the fraud was the work of individuals.

And the part that was presented to Congress can reasonably be characterised as having been 'sexed up' (as they put these days).  It is hardly a crime to have some involvement with a colleague's attempt at a replication of one's experiment.  And while, in retrospect, it was unwise for Taleyarkhan to have had any involvement at all, it may have been hard to foresee this at the time.  And I should imagine no student was forced to work in this area of research; it was for them to decide pros and cons and, given my own experience, I'm sure they had many people advising them of the cons.

Now the interesting questions, two in particular.  First, how did it come about that such a wildly distorted version of the report got to Congress? I doubt if Miller himself was responsible.  Rather, he was the pawn of someone else.  Who was that; what happened there?

Secondly there is the curious matter of Putterman's claim (which you reported) to have found proof that Taleyarkhan had used DARPA funds for an experiment not covered by the grant, and furthermore not acknowledged the use of these funds in the published paper.  But his case turns out to be seriously flawed.  For example:

1. Coblenz was shown the PRL experiment not because of a claimed funding connection but because he had indicated an interest in seeing it.  A special demonstration was put on for the benefit of a number of visitors. Should this be a problem?

2. Putterman doesn't say what the titles of the other grant applications were, but the fact that the word 'sonofusion' doesn't appear in them proves very little; there must be many different ways to characterise the same research.  Low marks for Putterman there.

3. To quote your article, the paper was submitted to PRL in September 2005, and

Taleyarkhan's postdoc, Yiban Xu, was a co-author on the PRL paper. The DARPA grant paid all of his salary for March and April 2005, and at least half of it from May until December.

The editor of Nature, in the course of discussion, told me he considered this the strongest point.  But it is not.  Presumably we are supposed to infer that from March to September 2005 Xu was being paid by DARPA, while working on the PRL expt. that did not acknowledge DAPRA funding.  But as Taleyarkhan has noted, there was a period, which the analysis fails to take into account, between when the experimental work was completed and when the paper was submitted.  During this time Xu had shifted to working on the new experiment.

I have an email where Taleyarkhan lists all the things that happen in between taking the readings and submitting a paper.  I don't think I need hunt this up: clearly there would be discussions about what the data mean and how to analyse them, drafts would be circulated for comment, etc. etc.  One does not simply switch off the apparatus and straight away write a paper and submit it the next week.

A similar analysis applies in the case of Cho, presumably.

4. Again, a similar point applies in the case of Taleyarkhan's funding, but here the red herring is brought up about people who are 'partly funded by DARPA' being treated as DARPA-funded.  I contacted Mark Frankel about this and he said you had not discussed the details of the specific case with him, so we don't know what he would have said in this case knowing the details (he declined the option of taking time to get into it in detail). If there is a rule that if a person has their salary partly paid by DARPA then they must mention DARPA in everything they do, it must be broken by people thousands of times a day.

Anyway, that is not the point.  It is only some distance on in the story that DARPA's rules get a look in.  Prior to that, the article says the following:

Taleyarkhan insists no DARPA money was used for that work, but after checking accounts at Purdue University, where Taleyarkhan is based, Putterman believes otherwise.

It is agreed that only 1/3 of Taleyarkhan's salary came from DARPA, so (according to most people, I suggest) he would be entitled to spend 2/3 of his time in other work without it being considered use of DARPA funds. That conflicts with the gloss that Putterman tried to put on the situation.

What it is said sometimes happens in such situations is that a group of people whose funding might be threatened by certain developments get together and say 'this must be stopped'.  That seems, to me at least, the most plausible explanation for the disreputable activity I have described in the above, involving people clearly ganging up on Taleyarkhan.

I have been critical of your contributions to Nature in the past on account of what appeared to me to be serious bias.  However, I do believe it is your intention to tell the truth, and I hope that now the facts are out in the open you will be able to write about this other side also.


Yours sincerely,

    Brian Josephson