Errare humanum est sed perseverare diabolicum

With a past PhD student of Francesco Stellacci’s, Predrag Djuranovic, weighing in to the debate, I think there are some important questions and some interesting observations that should be made.
One reaction to the recent posting by Predrag Djuranovic on Raphael’s blog is that we simply have a disgruntled student’s word against that of his advisor. However, we also have a link to part of the student’s work, which was performed in Francesco Stellacci’s lab at MIT. Note that the student is still at MIT, albeit with a different supervisor/advisor and thesis topic.
My reaction is that Francesco Stellacci had a smart graduate student in his lab, Predrag Djuranovic, capable of an analysis close to that of a leading practitioner of STM, Philip Moriarty. Result: Lots more papers on striped nanoparticles and some with STM images, graduate student is shifted sideways to a PhD in another lab at MIT. This paints a rather disturbing picture.

So at this point the STM data do indeed look to be an artefact and, without further analysis the case for stripes is weak.
How can this happen with peer-review? After all, peer-review articles were held up as a paragon of science and blogs as the antithesis of science by “Someone has to say it”, a commenter on Raphael’s blog. Well, peer-review is far from perfect and artefacts will at times escape the notice of the reviewers. Why? Even the most assiduous of reviewers is limited by their expertise. The assumption is that someone else is dealing with those parts one cannot review. This is also true of the parts of a paper based on previous work outside the reviewer’s remit – one assumes it is right. One might also remember that not all reviewers can be bothered to do their job in the first place. There is an editorial duty to seek reviews from a suitable breadth of expertise. However, this doesn’t always happen. The 100 posts at science fraud amply illustrate the failures of peer review.
Science should be self-righting. So I do not worry too much about the failures of peer-review. We should all remember Seneca’s wise words “errare humanum est”, though we would be wise to heed the rest of the phrase “sed perseverare diabolicum”, which roughly translates as “but to persist is diabolical”. It is the persistence (individual and institutional) that prevents science self-righting.

Updated 3 November 2013