方舟子的Blog

中国流行指控学术造假，有人担心文/革式迫害

记者：David Cyranoski
英国《自然》2006年5月24日

（方舟子译）

　　中国科学有着被一把双刃剑切割的危险：一方面是猖獗的科学不端行为，另

一方面是基于虚假指控的迫害。

　　由于对官方恰当地处理造假事件一事缺乏信心，这导致人们越来越依赖于那

些质疑中国科学家的履历和著作的网站。但是许多人担心这种未经试验的指控会

导致的伤害。100多名在美国的华人科学家向中国政府递交了一封公开信，要求

它设置确保公正地调查科学不端指控的机制。

　　中国承认它面临着严重的科学不端的问题，包括剽窃以及捏造和窜改数据。

还不清楚这个问题严重到什么程度，但是最近一大批指控已引起了对这个话题的

关注。

　　在今年3月，北京清华大学医学院副院长【译按：应是助理院长】刘辉在被

指控冒用另一个H. Liu的论文为其论文表注水后被解雇（见《自然》440,
728;
2006）。据报道刘辉否认这些指控，归咎于办事员失误导致的混淆。在今年4月，

成都四川大学因为澄清它的一名教授没有伪造论文，遭到了中国媒体的批评；这

篇论文自从它在2000年发表后就遭到攻击。曾在2003年宣布开发出中国第一个数

字信号处理芯片并引发爱国热情的上海西安交通大学【译按：原文如此】的陈进

2周前被其大学谴责伪造研究并从一家外国公司剽窃设计。

　　在所有这三起案例中，一个有众多读者的中文网站新语丝
（http://www.xys.org）对加强公众舆论发挥了关键作用，该网站在揭露中国科

学造假方面享有声誉。

　　在前两个案例中，张贴在新语丝上的指控导致中国媒体进行跟踪报道。该网

站的拥有人、住在加利福尼亚圣地亚哥的生物化学学者方是民（方舟子）声称他

首先公布了据信把外国芯片改换标签的陈进公司的名称【方舟子按：我和记者说

的是“我们最先确定了那家为陈进打磨芯片并改换标志的建筑设计公司”（We
were
the first one to identify the architecture company which polished
and
re-labeled Chen’s chip.
）。“120人公开信”的签署者中有人污蔑我曾经为
陈进辩护，记者来问我是否有此事，我于是向他解释了一下新语丝网站在这个事

件中的所作所为：“我从未为陈进辩护。有关陈进的指控首先被张贴在几个读者

众多的互联网论坛，包括我们的论坛（新语丝有一个读者可以自由张贴其评论的

论坛）。在有记者确认并与揭发者联系之前，我没有在新语丝上发表这个指控，

因为我们的政策是不发表匿名指控。事实上，我帮助某些记者调查并报道这个案

件，提供了某些关键信息。我在我们的网站上发表了许多文章支持这一调查并披

露了一些信息。我们最先确定了那家为陈进打磨芯片并改换标志的建筑设计公司。我

写过一篇文章要求对这一案件进行深入的调查，让更多人承担责任并起诉陈进。”

(I never defended Jin Chen. The allegation about Jin Chen was
first
posted to several popular Internet bulletin boards, including
ours
(New Threads have a bulletin board that our readers can freely
post
their comments). I didn’t publish this allegation on New Threads
until
some journalists have identified and contacted the
whistleblower,
because it is our policy not to publish anonymous allegations. In
fact,
I help some journalists to investigate and report this case
by
providing some critical information. I published many articles
to
support the investigations and disclose some information on our
web
site. We were the first one to identify the architecture company
which
polished and re-labeled Chen’s chip. And I wrote an article to ask
for
further investigating this case, finding more people accountable
and
prosecuting Chen.)】。

　　在缺乏足够的正式调查机制的情况下，该网站所拥有的质疑科学家的力量已

让它成为指控不端行为的关注中心。

　　印第安纳波利斯的印第安纳大学的免疫学家傅新元说，四川大学事件促使他

写了一封致中国科技部部长和中国科学院院长等科技高官的信，要求他们采取行

动。这封信在其同行中引起了共鸣——它在美国华人生物学家圈子中传了5天，

收集到120个签名，包括两名在中国的科研人员。“我深受感动。”傅新元说。

　　在表明揭露所有不端行为的必要性之后，该信集中在无根据的指控的问题上，

特别是那些没有给出实验程序有错的证据就进行学术攻击的。它以谴责“在缺乏

适当调查的情况下……匿名公开进行人身攻击”的现象结束。

　　傅新元说四川大学事件就是一个相关的例子。四川大学副校长魏于全在2000

年的《自然·医学》上发表了一篇论文，描述使用外源内皮细胞做为抑制肿瘤生

长的疫苗。该论文声称在小鼠中获得成功，并提出该技术在人类中也能有效

（《自然·医学》6, 1160–1166; 2000）。

　　西安交通大学的免疫病理学家司履生在2001年审核魏于全的一份资金申请报

告时，首次看到这篇论文，他怀疑它含有捏造的数据。今天3月26日，在听说魏

于全在用这篇论文进一步申请大批资金后，司履生在新语丝上攻击这篇论文。

　　司履生的信引发了中国媒体的愤慨和魏于全所在大学的调查。四川大学的调

查结论说魏于全没有犯错误，这一关于魏于全的研究的争端不过是一场普通的学

术争论。中国媒体继续批评魏于全和四川大学，但是许多科学家认为司履生的攻

击是不负责任的，是由于对科学概念和步骤做了不合理的解释。

　　例如，司履生认为小鼠免疫系统应该对外源细胞中的所有蛋白都做出反应，

而魏于全的论文表明免疫小鼠只对少数几种抗原做出反应。“这违背了免疫学的

基本原理，”司履生说。

　　但是马里兰巴尔的摩约翰斯·霍普金斯大学医学院的免疫学家、傅新元信件

的签署者陈列平不同意司履生的看法。【方舟子按：陈列平还是中国科学院“海

外百人计划”获得者，在生物物理研究所感染与免疫中心兼职。】陈列平说有选

择性地只对一种或少数几中外来抗原做出免疫反应是一种众所周知的现象，称为

免疫优势。

　　司履生也对魏于全所用的小鼠数量提出疑问，估计要用到大约4万只。“这

个数量大得难以置信，”他说。魏于全说司履生算错了数量，实际上用的小鼠少

于5000只；陈列平支持魏于全的说法。

　　但是即使是那些为魏于全辩护的人也承认魏于全的反应于事无补。例如，司

履生声称魏于全至今拒绝出示其原始数据，大多数人同意公布原始数据将会平息

这个话题。魏于全告诉《自然》，“我没有说我不能出示原始数据供调查”，但

是他没有澄清他是否将会公开他的数据。他否认所有的不端指控。

　　四川大学的对此事的调查未能使许多人相信真相已大白，主要是因为它缺乏

透明度。“四川大学最近对指控造假自己做的调查完全是个笑话，”加州大学伯

克利分校的神经生物学家、上海神经科学研究所所长蒲慕明说。《自然》希望了

解四川大学调查步骤的细节和调查委员成员组成情况，该要求被交给了魏于全；

到《自然》印刷时他还没有提供有关该调查的任何信息。

　　蒲慕明相信这一事件表明大多数中国大学没有能力调查自己的成员。“调查

结果很可能受到大学自身利益的影响，比如为了保护它的名声，”他说。

　　傅新元在5月8日发出的信呼吁更高层次的资金提供机构，例如科技部、中国

科学院和国家自然科学基金的更大参与。

　　这些机构已经设有调查部门。中国科学院在1997年建立了其道德委员会，并

在2001年制定了准则。在1998年成立的国家自然科学基金说它在最初的5年调查

了445起不端行为的指控（在那段时间它资助了大约3万个项目）。对那些最严重

的案例，委员会无限期地禁止肇事者申请基金。

　　但是许多科学家感到这些委员会是无效的，对他们解决问题的能力缺乏信心，

这导致那些不满的人在互联网上公布其控告。例如，司履生说他考虑过把他的举

报交给中国科学院或科技部，但是他没能找到详细的联系方式。因此他把其指控

贴到了新语丝。《自然》试图与中国科学院和国家自然科学基金的委员会联系，

也没有成功。

　　“正是由于（实际上）缺乏这种正规的机制才使新语丝变得重要，”傅新元

说。但是傅新元做为一名人权提倡者，担心媒体根据网上不负责任的指控，特别

是那些没有表明其身份的指控进行的炒作，让人回想起中国的“大字报”。

　　在1970年代【译按：原文如此】文化/大革命期间，这些贴在墙上的手写张贴

被用于迫害那些被认为是政府敌人的人。“任何人都能写任何事情，人们读了它

就假定它是对的，”陈列平说。“在学术界再发生这种事会是非常可怕的。”

　　自从方是民在2001年【译按：应是2000年】建立其网站用于揭露坏科学并试

图提高中国科研道德水平后，已受到了广泛赞扬。他为他的做法进行了辩护。他

说在收到的来稿中，他只发表其中的大约10％，而且他只发表那些向他表明真实

身份的来稿者的指控。他补充说，他会做一些初步的调查，有时也向外部专家征

求意见。

　　但是有几位科学家给《自然》写信，对方是民的网站变得如此有威力表示担

忧，说他们害怕姓名被公开，因为他们害怕成为他的敌人。

　　傅新元说，他希望能够见到中国建立一个由在科学不端行为方面受过训练的

专家组成的新的机构，对造假的指控进行调查，类似于美国的科研诚实办公室。

《自然·医学》的主编Juan-Carlos
Lopez说，那对解决司履生指控魏于全一案
肯定是必要的。“已经有够多的‘他说，她说’的废话，”Lopez说。“是能胜

任的权威机构介入的时候了。”

　　人们并不清楚这将会如何发生。傅新元和他的共同签名者也还未收到中国权

威机构的任何答复。

http://www.nature.com/news/2006/060522/full/441392a.html

Nature

Published online: 24 May 2006; | doi:10.1038/441392a

Named and shamed
As accusations of scientific misconduct in China become rife,
some
fear persecution reminiscent of that used in the Cultural
Revolution.
David Cyranoski

Chinese science risks being sliced up by a double-edged sword:
rampant
scientific misconduct on the one hand, and persecution based on
false
accusations on the other.

The lack of confidence in official mechanisms for properly
investigating fraud has led to increased reliance on websites
that
challenge the records and publications of Chinese scientists. But
many
are concerned about the damage such untested allegations can
cause;
more than 100 Chinese scientists based in the United States have
sent
an open letter to the Chinese government, asking it to set up
mechanisms to ensure that claims of scientific misconduct are
investigated fairly.

China admits it faces a serious problem with scientific
misconduct,
including plagiarism, and the fabrication and falsification of
data.
The scale of the problem is unknown, but a recent spate of
allegations
has drawn attention to the issue.

In March, Hui Liu, the vice-dean of Tsinghua University medical
school
in Beijing, was fired, following claims that he had boosted
his
publication list with papers by another H. Liu (see Nature 440,
728;
2006). Liu has reportedly denied the charges and blamed the mix-up
on
a clerical error. In April, Sichuan University in Chengdu was
criticized by the Chinese media for finding one of its
professors
innocent of fabricating a paper; the paper has been under attack
since
its publication in 2000. And two weeks ago, Jin Chen of
Shanghai’s
Xi’an Jiaotong University, whose announcements of one of China’s
first
digital signal-processing chips in 2003 stoked patriotic fervour,
was
condemned by his university for faking research and stealing
designs
from a foreign company.

In all three cases, a popular Chinese-language website known as
New
Threads (http://www.xys.org), which has a reputation for
disclosing
scientific fraud in China, played a key role in fuelling public
outcry.

In the first two cases, postings of the accusations on New Threads
led
to the Chinese media picking up on the stories. And the website’s
owner,
Shi-min Fang, a biochemist based in San Diego, California, claims
he
was the first to post the name of Chen’s company which
supposedly
re-labelled foreign chips.

The power of the website to implicate scientists in the absence
of
adequate formal mechanisms of investigation has put it at the
centre
of concerns over claims of misconduct.

Xin-Yuan Fu, an immunologist at Indiana University in
Indianapolis,
says it was the Sichuan University case that drove him to write
a
letter to key science-policy officials, including China’s science
and
technology minister and the head of the Chinese Academy of
Sciences,
asking them to take action. The letter struck a chord among his
peers
— within five days of circulating it to other Chinese
biologists
based in the United States, Fu’s letter had collected 120
signatures,
including those of two researchers in China. “I was overwhelmed,”
says
Fu.

After noting the need to expose all types of misconduct, the
letter
focuses on the problem of unfounded allegations, particularly
those
that attack scientific claims without giving evidence of
faulty
laboratory procedures. It ends by condemning the tendency to
make
“personal attacks anonymously in public… in the absence of
proper
investigation”.

Fu says the Sichuan University incident is a case in point. Yuquan
Wei,
vice-president of the university, published a paper in Nature
Medicine
in 2000 detailing the use of foreign endothelial cells as a vaccine
to
prevent tumour growth. The paper claimed success in mice and
suggested
the technique could work in humans (Nature Med. 6, 1160–1166;
2000).

But Lusheng Si, an immunopathologist at Xi’an Jiaotong University
who
first came across the paper when reviewing a grant proposal by Wei
in
2001, suspected that it contained fabricated data. On 26 March
this
year, after hearing that Wei was using the paper to request a
further
large grant, Si attacked the paper on New Threads.

The letter led to a media fury in China and an investigation by
Wei’s
university. Sichuan concluded that Wei had committed no offence,
and
that the dispute over Wei’s research was simply a
run-of-the-mill
academic disagreement. The media in China has continued to
criticize
Wei and Sichuan University, but many scientists think Si’s attack
was
irresponsible and based on unsound interpretation of
scientific
concepts and procedures.

Si contends, for example, that the mouse immune system should
respond
to all proteins in foreign cells, whereas Wei’s paper suggests
that
immunized mice selectively respond to a few antigens. “This
violates a
fundamental law of immunology,” Si says.

But Lieping Chen, an immunologist at Johns Hopkins University
School
of Medicine in Baltimore, Maryland, and a signatory to Fu’s
letter,
disagrees with Si. Chen says that a selective immune response to
one
or a few foreign proteins is an aspect of well-known
phenom–enon
known as immunodominance.

Si also questions the number of mice Wei used, estimating this to
be
around 40,000. “This is too big to believe,” he says. Wei, backed
by
Chen, says Si has miscalculated the number, and that less than
5,000
mice were actually used.

But even those who defend Wei admit that his response hasn’t
helped.
For example, Si claims that Wei has so far refused to release his
raw
data, which most agree would settle the issue. Wei told Nature, “I
did
not say I cannot release raw data for inspection”, but he has
not
clarified whether he will make his data available. He has denied
all
misconduct.

The university’s investigation into the matter has failed to
convince
many that the truth won out, mainly because it lacked
transparency.
“The recent self-investigation into alleged fraud at Sichuan
University is a total joke,” says Mu-ming Poo, a neurobiologist at
the
University of California, Berkeley, and head of the Institute
of
Neurosciences in Shanghai. Nature’s request for details on
the
university procedure and an introduction to members of the
investigation committee was referred to Wei; as Nature went to
press
he had not provided any information about the investigation.

The recent self-investigation into alleged fraud at Sichuan
University
is a total joke.

Poo believes the incident is indicative of the fact that most
Chinese
universities lack the capacity to investigate one of their own.
“The
outcome is likely to be influenced by the university’s own
interests,
such as protecting its reputation,” he says.

Fu’s letter, sent on 8 May, calls for greater involvement of
higher-level funding bodies such as the science ministry, the
Chinese
Academy of Sciences (CAS) and the Natural Science Foundation of
China
(NSFC).

These institutions already have investigatory bodies. The CAS
established its ethics committee in 1997 and drafted guidelines in
2001.
The NSFC committee, established in 1998, says it investigated
445
allegations of misconduct in its first five years (out of an
estimated
30,000 projects that it funded during that time). In the most
severe
cases, the committee indefinitely blocks perpetrators from
applying
for funds.

But many scientists feel these committees are ineffective, and a
lack
of confidence in their ability to settle matters is driving those
with
grievances to publish them on the Internet. For example, Si says
he
considered sending his complaint to the CAS or to the science
ministry,
but he was unable to find contact details for either. So he posted
his
accusation on New Threads instead. Nature’s attempts to contact
the
committees of the CAS and the NSFC were also unsuccessful.

“It is the [effective] absence of such formal mechanisms that
makes
New Threads important,” says Fu. But Fu, a human-rights advocate,
is
worried that the media frenzy following irresponsible
web-based
accusations, particularly by those who don’t identify
themselves,
hearkens back to China’s ‘big letter’ posters or ‘dazibao’.

These wall-mounted handwritten posters were used to persecute
those
considered enemies of the government during the Cultural Revolution
in
the 1970s. “Anyone could write anything, and people would read it
and
assume it was right,” says Chen. “It would be a terrible thing to
go
through again, in academia.”

Fang, who has been widely praised since setting up his website in
2001
for exposing bad science and trying to raise the profile of
research
ethics in China, defends his postings. He says he only accepts
about
10% of submitted letters, and that he only publishes allegations
from
correspondents who identify themselves to him. He adds that he
does
some preliminary investigation and sometimes asks outside experts
for
their opinions.

But several scientists have written to Nature to express concern
over
how powerful Fang’s website has become, saying they are afraid to
be
named for fear of becoming his enemy.

Ideally, Fu says he would like to see China establish a new
agency
staffed by experts trained in scientific misconduct that
could
investigate claims of fraud, akin to the US Office of
Research
Integrity. That would certainly be necessary to resolve the case of
Si
versus Wei, says Nature Medicine’s editor-in-chief Juan-Carlos
Lopez.
“There’s been enough of this ‘he said, she said’ nonsense,” says
Lopez.
“It’s time for the competent authorities to get involved.”

How likely that is to happen is unclear. Fu and his
co-signatories
have yet to receive any response from the Chinese
authorities.

(XYS20060525)

◇◇新语丝(www.xys.org)(xys.dxiong.com)(xys.3322.org)(xys.xlogit.com)◇◇