Guest post: A seismologist from Japan looks at the STAP cells mess

By Robert J. Geller

I’m a seismologist at the University of Tokyo, where I’ve been since 1984. I was the first tenured foreign faculty member in the history of our university. If you look at my publication list on the Researcher ID site you can see that my main research recent interests are modeling seismic wave propagation and analyzing observed seismic waveform data to determine the seismic velocity structure of the Earth’s interior. I also have a longstanding interest in science policy in Japan. After the 2011 Tohoku earthquake I wrote a commentary article in Nature criticizing the Japanese government’s earthquake prediction program. You can read my recent article on the issues involved in restarting Japan’s nuclear plants, and you can follow me on Twitter (@rjgeller), where I mostly tweet in Japanese. The STAP cells mess is having a big effect on the global stem cell research community, but is also impacting all scientists in Japan. Paul Knoepfler kindly invited me to guest-post here.

Like other non-specialists, the first time I heard of either STAP cells or Haruko Obokata was at the end of January, when her two papers were published in Nature to the accompaniment of a gale-force blast of publicity in the Japanese media orchestrated by Riken. Her press conference, where she was flanked by her co-author Teruhiko Wakayama, was featured on every network’s news broadcast. She was lionized by the media as the epitome of a “rikejo” (a woman in science fields), who decorated her lab in pink, and was featured not only on news programs but also on infotainment shows. The publicity emphasized her personality more than her research itself, treating her more like a pop star than a unit leader at Riken. Still this was a feel-good story—a young scientist who was reported to have made a major breakthrough—and she was invited to attend a meeting of the Prime Minister’s Council for Science and Technology Policy on Feb. 14.

As everyone now knows, reports of problems with the figures in the Nature papers began appearing on the internet in early February. A tweeter with the user name @JuuichiJigen played a major role in presenting clear and extensive lists of problems, posting mostly in Japanese. Problems were also reported on the pubpeer site. Obokata skipped the Prime Minister’s Science Council meeting on Feb. 14, and on Feb. 15 the Mainichi newspaper reported that Riken had opened an investigation into possible irregularities in the Obokata et al. papers. Nature is also investigating.

It’s March 13 in Tokyo as I type this, and as we all know the problems have snowballed, and now also include apparently serious problems related to Obokata’s 2011 Ph.D. thesis at Waseda University, which has launched its own investigation. Riken has scheduled a press conference (to be webcast) at 1400JST on March 14.

I’m an outsider to the stem cell field, but this mess impacts public trust and support for every field of science in Japan. The longer it’s allowed to drag on the worse the ultimate impact will be. So it’s in the interest of every scientist in Japan that Riken, Nature, Waseda, and other institutions fully and transparently identify the problems and take appropriate actions. Let me make a few observations.

Nature:  First, to take a random example, the Obokata et al. article refers to “pasture pipettes” rather than “Pasteur pipettes.” The existence of such an obvious error suggests that the nine authors, as well as the referees, editors, and copy editors were all, to put it bluntly, slackers. Some of the actions taken by the authors in apparently inappropriately editing and recycling figures might arguably have been uncatchable by the referees and editors, but surely this simple typo should have been caught by someone. Maybe the referees and editors who missed this should be put out to pasture.

More seriously, Nature has some heavy-duty explaining to do. An earlier version of the Obokata et al. paper was rejected, but then the problem-beset version published in late January was finally accepted. I’d like to see Nature, while protecting the identity of the referees, disclose (on their website) all of the editorial correspondence (redacted appropriately) and every version of the submitted papers, so we can all judge whether or not the decision by Nature to accept the papers was appropriate based on the information available to them at the time.

Nature Publishing Group not only publishes journals. Other subsidiaries provide editing and web/print publishing services, sell advertorial supplements in Nature’s journals, and so on. Riken is reported to be one of their customers. Without implying that the commercial side in any way intervened in the editorial side (something Nature has strongly denied in response to a question from a Japanese newspaper), the potential for possible conflict of interest here is obvious. Just as Nature asks authors to disclose their conflicts of interest (something Obokata et al. failed to do in the case of their patent application), Nature’s parent company should also disclose its own conflicts of interests more clearly.

Waseda: Two co-authors of the Nature article, Masayuki Yamato (Tokyo Women’s Medical University) and Charles A. Vacanti (Harvard Medical School) were also external examiners for Obokata’s Ph.D. thesis at Waseda University in 2011. It is hard to believe that neither they nor the two internal examiners failed to notice the complete absence of cited references in the introductory chapter to the thesis, which could and should have led to the discovery that it was plagiarized from the National Institutes of Health home page, and should have in turn led to an investigation that might well have uncovered some of the other major problems. In view of these problems it is hard to see how this thesis can, on reexamination, be judged as appropriate justification for award of the Ph.D. degree. One hopes that this is an isolated case rather than the tip of a giant iceberg, but probably every university in Japan, not just Waseda, should conduct a systematic investigation of all of their Ph.D. theses, taking appropriate actions in all cases of large-scale plagiarism or data manipulation. This is obviously something most university administrators would rather not face up to, and I personally hope that a large number of problems are not uncovered, but the health of the scientific enterprise demands integrity.

Riken: Riken obviously has big governance problems. These basically stem from the fact that while nominally being independent it is actually heavily controlled by government bureaucrats on secondment from the Ministry of Education and other ministries. (Nominally this is not the case.) The Japanese civil service is divided into “career” and “non-career” entrants. The former are fast-tracked from entrance, and are gradually winnowed out as they are rotated every 18 months or so. The bureaucrats who rise to the top do so by getting along with their superiors and not being tagged with failures. They don’t actually have to do something positive; in fact doing so would probably offend someone senior who would push them off the fast track. Their interests in resolving the STAP cell mess would best be served by throwing Obokata under the bus and nominally admonishing the other Riken-affiliated co-authors, thereby providing pseudo-closure. They could then just go on with business as usual, telling the public “nothing to see here; move along.” On the other hand, the interests of the scientific community in Japan, not to mention the interests of the public, would best be served by unearthing and correcting the various structural problems that exist.

Government policy decisions in Japan are made based on recommendations by advisory panels. This sounds good, but the problem is that the advisory committees are hand-picked by the government, and the academic members consist solely of researchers cynically called “go-yo gakusha,” who are selected because they will agree with whatever the government says. (In case you’re wondering, the government has never asked me to serve on one of these committees. That’s why even though I’m 62 now I still have time for research.) I’m afraid that until this system is changed, which seems unlikely to happen any time soon, the structural problems that led to the STAP cell mess won’t be corrected, and business as usual will continue.

© Robert J. Geller, 2014

17 thoughts on “Guest post: A seismologist from Japan looks at the STAP cells mess


  1. I appreciated Dr. Geller’s well balanced comments. This scandal will need to be interpreted from a social context and many people will surely do that. One will be Japan’s scientific career system, in which there used to be no postdocs until relatively recently. Modernization of career passages has been causing a lot of pains in Japan and this scandal probably reflects deeply rooted and structural problems of the science society. Unfortunately, Japan’s unique bureaucratic sytem will not make it easy to address such problems efficiently.


  2. i have nothing to add to your post except that i feel you’re being modest by restricting your concern to Japan.

    i think this entire fiasco embodies the problems with crony capitalism, and i believe it has gotten the public’s attention unlike any other controversy i can remember.

    attention for science is a good thing, especially when it’s through things like this–the scientific process.


  3. Thank you for a great post. I would like to add a component to what is missing from this entire discussion so far- the responsibility of the journal Nature in catching plagiarism in the first place.

    There are several sophisticated computer tools for detection of plagiarism. It is very surprising the world’s top science journal does not use this. Had they used this, it would have caught the simple methods plagiarism in the STAP paper, thereby triggering more widespread investigation, preventing these papers from ever being published, and sparing all of us this sad saga.

    Therefore, I believe Nature bears much if not all the responsibility for this situation. A journal of the caliber of Nature should *expect* people to submit plagiarized papers all the time. So it has a responsibility to *detect* this. The journal did not take even the most elementary steps to try and detect plagiarism (or has very under-performing tools and process for this). So it is neglecting its responsibility as the world’s premier journal to safeguard the community from overt plagiarism, in which case it is nothing more than a high-profile platform for people to make big grandiose and often false claims.

    The lesson to me in all this has nothing to do with the authors, nor Obokata, nor Japanese science. Plagiarizers abound and there is no shortage of them. This is nothing new. What has been exposed, rather, is that the world’s top science journal does little to detect plagiarism, has a broken and damaged process of peer review, and all too eager to promote its brand even at the cost of people’s livelihood (time and money spent replicating false work).

    After all, if Nature and its peer review process cannot even catch word-for-word plagiarism, then what hope do we have that it is catching much more sophisticated, subtle and hard to detect statistical errors, methodological errors, etc?

    This situation also highlighted the immense power of post-publication crowd-sourced peer review centralized by Paul Knoepfler. After all, how can 3 reviewers and overworked editors compare to thousands of people. This is a watershed moment in how we will disseminate science going forward.


    • There was no plagiarism in the Nature methods section. The introduction to her PhD thesis was allegedly plagiarized.

      This post like the OP jumps from one incident to very broad conclusions. There are fraudulent papers all over the world in diverse scientific cultures. This is not specific to the japanese situation.

      And most importantly: Its still to early to say that her work is not reproducible.


      • There was indeed overt plagiarism of text from a paper of a different lab into the methods section of the published paper in Nature. Please review the long list of issues with the paper previously posted on this blog prior to defending it.

        The broad conclusion that Nature has a responsibility to catch simple copy-paste plagiarism is therefore highly relevant, and a central point few people are making. In my opinion this is the key illustration as to why Nature and traditional peer review are out of date and on an inevitable decline towards irrelevance.


  4. I think Robert is completely missing a point. The problem stems from XXXXX lack of training to conduct a scientific research. XXXXX This is not a unique problem. Rather, Japan’s higher education system lacks training on rigorous scientific writing. I got my Ph.D. in Japan, too. Was never taught how to write a report (or thesis for that matter). Everybody is pretty much on his/her own. I bet none of Japanese members of Obokata’s dissertation committee cared about writing. They didn’t get any formal training, either, after all. Robert should have noticed this. Students (even in Tokyo University, the best university in Japan!) do plagiarism in their reports all the time. (I know this because I did TA.) I was very surprised to learn that training on rigorous writing starts very early in the US. Elementary kids learn how to differentiate opinions from facts in writing. High schools kids are asked to use citation and reference in their reports. (They are even asked to use a specific style, such as APA!) Thesis requires very strict formatting, like margins, spacing, paragraph style, and of course, citation and reference. These never happen in Japan. Japan needs to adopt this kind of systematic training and plagiarism checker.

    (XXXXX edited for content)


    • I don’t fully agree. This isn’t a problem restricted to Japan, but now all universities try to push as many PhD’s out the door as quickly (in 3 years) as possible. They are strongly incentivized to do so by the Ministry of Education, Culture, Sports, Science and Technology, as certain types of funding scale with this statistic; also the PhD students pay the university valuable tuition fees instead of getting paid a salary. On top of all this, in Japan nearly all the PhD’s are submitted at the end of December, with the candidates expecting to finish by March. In a big university, a professor might typically supervise a few of his/her own Master and PhD candidates, and sit in the evaluation committees of 5 more, all in the space of 1 month. A PhD student might send in a first draft a few weeks before official submission, making it almost impossible to make all the changes.

      This means there will be a higher percentage of candidates not meeting international standards. The committee can only oblige by pushing the standards lower and lower. It logically follows that at some point the bar will be so low that it actually gets embarrassing.

      I would argue that Japan took a conscious decision to go down this path, as the universities wanted more free student manpower to do research, and more money from the government.


      • that’s so pathetic that they’re trying to cover up something that’s been published.

        of course any institution can say “oh it’s not final” when they’ve been caught red-handed.

        these people have no shame. they know that once they admit one fault, it’s definitely “baby WITH the bathwater” (especially when it comes to academia).


    • Hi, Ken
      I took a M.Sc. in Univ. of Tokyo, though I’ve got bachelor’s in another univ. in Japan. I don’t mean to deny all of your univ. experiences. However I want to say two points in your claim.

      1. I think students learning in famous univ. may learn their major or related subjects using textbooks written in English. (In my case, half of my textbooks were written in English.) So, they must have many chances to know how to express specific terms, expressions, styles and so on, if they study hard. However they don’t realized such things cause they are always busy to try to translate all or search for translated ones.

      2. Students are recommended to read many papers as possible in any labs to know current accademic trends and rivals’ doings. And also they may have trainings for accademic meeting. (About this trainings I am not sure every univ. does) At least, I’ve got several lessons from my senior.
      I think there are many chances to know how to write/understand papers. Students can learn by self-training with the help of univ.’s guidance. That’s why univ. are called high-leveled educational organization, not a school.

      To editor
      It would be helpful to correct spells and grammers in my opinion. There may lots of mistakes. It have passed long time since I used English in my work. thank you. :-)


  5. Great comment. I am doing currently a PhD at a top three university in Japan, although in field of social sciences. All PhD theses, I have read, in my field contain a plagiarized parts without correct attributions – not a single one has been published as a book in English, some in Japanese though. Last year, I was at a conference, where I spotted a clear usage of someone else’s work during a paper presentation of a, coincidentally, young Japanese female researcher. No action was taken and I believe the paper was published in its original form.

    I am afraid that if Japanese universities start checking each and every PhD work for plagiarism, then they will be only handful of degree holders left.

    In my case (my department and faculty):
    – no proper supervising – “do whatever you like”
    – no teaching or training in science ethics
    – no teaching or training in science methodology and research design, among others, – “just read this book, you would learn it”
    – no punishment for plagiarized work after I have discovered some wrongdoings.
    – pressure to graduate as soon as possible without any support for future career or academic position
    – etc.


  6. I am deeply, deeply disappointed by how RIKEN is seeing the situation. That is all I can say now.


  7. I quite agree with Prof. Robert J. Geller.
    As the professor has already mentioned, only Obokata would be thrown under the bus, so the other co-authors and those involved in this incident would never shamelessly take responsibility for their fault, more’s pity. (XXXX edited for content)
    Incidentally, I’m only a Japanese unknown researcher on math.


    • As a Japanese taxpayer, I hope that Riken is broken up on this occasion, so it will lead to tax saving. The reason I’m unable to accept the existence of Riken is that Riken doesn’t seem to have a sense of crisis about so-called Obokata’s incident. In my view, the investigation by Riken is so lenient with Obokata and her colleagues in this incident that the fact is not in the least uncovered, and I cannot help saying that that’s out of the question.
      Be that as it may, I’d like to advocate that Riken be broken up by The Japanese Government’s condemnation.


  8. I’ve worked at several labs in the US, but similar fraud is going on. Most if not all papers that were published contained some level of fraud…


  9. Do you think STAP cells are the same as spore-like stem cells like Vacanti insists?

    They said,
    “Obokata worked with top stem cell scientists to show rigorously that these cells were triggered to return to a stem cell-like state by an environmental stress, such as being placed for 30 minutes in a mild acid bath.”
    -from The Boston Globe-

    It seems to me this is not what the Nature paper said. To me, if Vacanti is right, the Nature paper shouldn’t show TCR rearrangement. But they did. His claim is inconsistent. Are spore-like cells are also CD45+ cells? How do you think?

Comments are closed.