![[Critical Mass]](/archives/cmlogo.gif)
![[dot]](/archives/cmdot.gif){kind=small}
December 16, 2009
![[feather]](/archives/feather.gif){kind=small}
You don't need a weatherman...
Neil Hamilton has rightly pointed to one of the uglier elephants in academe's living room, observing that amid all the faculty-led ruckus about threats to academic freedom, the decline of tenure, and so on, there is a huge, unpleasant, salient fact that professors do not generally acknowledge: professors aren't keeping their end of the bargain. All the freedom and security professors declare they want and need to do their jobs is predicated on the expectation that faculty will self-police, within disciplines and also within institutions--but they just don't do it.
As Hamilton notes,
... members of a peer-review profession cannot aggressively justify and defend their control over professional work when they do not both understand the profession's social contract and internalize their responsibilities under the social contract. The social contract of each peer-review profession is the tacit agreement between society and members of a profession that regulates their relationship with each other, in particular the profession's control over professional work. Essentially, in order for the public to grant a peer-review profession more autonomy and control over the work different from the control that society and employers exercise over other occupations, the public must trust that the profession and its members will use the autonomy at least to some degree to benefit the public in the area of the profession's responsibility, not abuse occupational control over the work merely to serve self-interest.The simple fact is that all the data available indicate that a substantial proportion of graduate students and faculty members do not clearly understand the profession's social contract, academic freedom, shared governance, and each professor's and the faculty's specific duties that justify the profession's claims to autonomy. Osmosis-like diffusion of these concepts and duties does not work.
Hamilton notes that a lot of the feelings of endangerment that plague faculty culture are self-created and self-fulfilling: "The predictable result of an anemic defense of a profession's social contract during a time of market change is that the society and employers will restructure control of the profession's work toward the regulatory and employer control typical for other occupations -- essentially the default employment arrangements in a market economy."
Hamilton has good general ideas about how to remedy the problem. Education in professional norms, standards, and ethics should be a standard part of grad school, he notes. Continuing education for faculty is needed, too. Ditto for strong leadership at the administrative and governance level. But what about the specifics?
Thinking about ClimateGate and what it may tell us how academia should define professional ethics and enforce those definitions, Northwestern law professor Jim Lindgren argues against the traditional academic model of internal investigation as a mechanism of accountability:
According to press reports, Penn State University is conducting an inquiry to determine whether it should institute a formal ethical investigation of Michael Mann, the Penn State professor who was the lead author of the paper that invented the "Hockey Stick." At issue are CRU emails and his role in ClimateGate.Frankly, I am not a big fan of academic investigations.
First, academic investigations are not how science – or social science – is supposed to operate. They are a hard type of official coercion, which ought to be reserved for only the most egregious cases.
Second, sometimes the investigations are half-hearted, conducted by colleagues who understandably would much rather see no evil.
Third, even when the investigators are diligent and unbiased, academic investigations are often conducted in secret, which makes it easy for the researcher to mislead the investigators with specious arguments that would be unlikely to hold up in the light of day.
For one or more of these reasons, I fully expect Penn State not to bring formal charges against Professor Mann – and if it does, I expect him to be cleared by his colleagues. Though I have read only a few dozen CRU emails, in my opinion Mann's errors should be corrected in the usual way, not by organized groups telling people what to think.
But if I were Professor Mann's dean at Penn State, I would try to determine whether he has fully shared his data, metadata, and computer code. To the extent that he hasn't already, I would try to make him do so – at least for his most important or most controversial articles in recent years. And, for reason #3 above, I wouldn't take Mann's word for it. I'd call his critics and ask them to name the few most important Mann papers for which the data and computer code are needed for replication.
If Mann is still withholding the data and code necessary for replication, I'd ask him to replicate his most important or most controversial recent work (certainly not everything) and to release the data and code so that others might do so. If Mann couldn't replicate his own work, I would ask him to announce that fact to the scientific community, so that serious scientists would know whether his work is replicable.
Thus, if I were Professor Mann's dean, probably the only power I'd use would be to further the scientific enterprise. And even that would not be necessary if ethical standards were higher in the subfield of paleoclimatology.
(For those who might be wondering, I did not call for a formal investigation of Michael Bellesiles back in 2000–2002. It was Bellesiles’s supporters who most frequently called for an investigation, though some of his critics did as well. Emory's investigation was triggered by prominent members of its faculty pushing privately for a formal inquiry. Apparently, Bellesiles's public supporters, being too lazy or too biased to bother checking the evidence that could be found in an hour or two in any major academic research library, miscalculated that Bellesiles would be vindicated. He wasn't.)
Fascinating points in themselves -- and even more so when one considers what a costly debacle Colorado's investigation of Ward Churchill turned out to be. Lindgren makes good points about why "investigating" concerns about research misconduct don't work--but he also begs the question of what ought to happen when / if someone like Mann doesn't jump through all the hoops he thinks deans should make him jump through. I can see the argument right now: "I have a research program to keep on track. I have grants to keep on schedule. I don't have the time or the resources to just sit around and replicate work that I assert is just fine as is. You have no call to require me to jump through these hoops. You are targeting me because you have been influenced by the right-wing machine that is hostile to my research results. My academic freedom is at stake--and by implication, that of my colleagues!" That's stonewalling--but it's also stonewalling with a point.
I'd love to hear readers' thoughts on what academia can do to make integrity and accountability a stronger professional value--and to enforce that value effectively.
Trackback Pings:
TrackBack URL for this entry:
http://www.erinoconnor.org/movabletype/mt-tb.cgi/1770
Comments:
One thing that might help is a revolution in academic publishing. Especially when research is publicly-funded, it is ridiculous to lock it away in very expensive, subscription-only journals.
I often see a media snippet on some possibly-interesting research finding..but nothing is available on-line except a link to an expensive journal. (Evidently, the journalist wrote the item based entirely on the university's press release.) Whereas once scholarly journals were a key enabler of information interchange, they now seem to be a major inhibitor of the same.
Also...in trying to understand the forces that drive conformity, in academia and elsewhere, C S Lewis's analysis of the Inner Ring is valuable. Excerpts linked at my blog (Nov 28 post)
There's really no such thing as professional integrity.
There is only the personal integrity of the various members of the profession, and their willingness (or unwillingness) to associate with those who prove themselves to be of dubious character.
Start fixing things there: raise better people.
Suppose that the investigation turns up positive proof that Mann falsified data in some way? Can/ would there still not be something done? What if it turned out that he didn't so much falsify as disregard accepted practice among scientists - not throwing out data without a reason, that kind of thing?
As to enforcing integrity and accountability, I'd ask whether there are written rules agreed upon by all parties regarding transparency and sharing of data. If not, there should be. And it might be worthwhile to have a policy, on which everyone must sign off, stating that we'll go where the data take us and not let ideology of any sort get in the way. That might seem to be a feel-good do-nothing act, but it would be something to point to when people start to go off the rails.
What is really needed is a good foundation to start with, one that lays things out in advance.
Part of the problem is that for the most part the "enforcement" occurs randomly as the result of attention rather than science.
also stonewalling with a point -- well said.
I agree with the comment on personal ethics as being what is in question.
The most apparent fix for the issue in the paleo-climate community (of which I am a peripheral member) lies in addressing the disconnect between a very outdated form of communication (traditional, peer-reviewed journal publications) and the nature of the modern research enterprise, which is highly data generative and complex model directed. Traditional journal articles are not as useful to the scientific enterprise as are media-rich reporting mechanisms which (to be useful to the end user) make possible and in some cases require the publication of data, computer codes, etc. along with a summary of the analysis of such data. The culture of peer-reviewed journal publications is less effective than it use to be (in the sciences, at any rate - I don't have any ken for the nature of communication outside of my broad field) because the nature of what it is we are communicating to one another has changed significantly. Movement away from traditional publishing towards more open formats (arxiv.org, for example in the sciences) is steady though traditional communication vehicles continue to carry the weight of prestige with them. The situation is fluid, however, and (I believe) will tend towards a format that is most conducive to the way science is really done. That is, after all, why we communicate in the first place.
There is a cultural aspect to all of this, of course. Though much of this material is produced at tax-payer expense (via public granting agencies), the raw material is still viewed as proprietary to the researcher or research group that generates it. The environment of professional academic science is competition driven to some degree and this engenders the type of defensive behavior on display in the CRU emails. This is the broader issue in academia that "ClimateGate" highlights - how do we develop a culture that places the academic (or scientific, in this case) enterprise above the personal interests of its participants? Can this even be done? Can any amount of talking to graduate students about ethics and social contracts really alter their human nature? Or should we be actively trying to design the culture (of publication and peer-review, for example) to take advantage of our base motivations to shepherd our efforts towards a productive social end?
For Lindgren's program to be implementable, Prof. Mann would have to be a terminable employee. I do not think you are going to improve academic integrity much without formal sanctions and without oversight by people not infected by a guild mentality.
1. Replace tenure with renewable multi-year contracts.
2. Replace unwieldy self-replicating boards of trustees with boards of modest size elected by resident alumni; require that every trustee file an oath of office with the county clerk which states in explicit terms that he is responsible for oversight of the fulfillment of the school's academic programs.
3. Modify labor law to incorporate greater transparency for prospective employers. The rubrics of job interviews are maddening, as is the paranoia some people have about giving honest references.
4. Alter the consequences of dismissal from academic institutions in such a way that departments and administrators might be more willing to do it. A less sclerotic and privilege-laden labor market might be of help (then again it might not):
a. Break down the barriers between higher education and the rest of the academic world through a thorough reconstitution of the practice of secondary education - most especially including the closure of extant teacher training faculties, the replacement of public agencies with philanthropic enterprise, and the abolition of collective bargaining.
b. Attempt to weaken the barriers which have appeared between working professionals and their academic counterparts. Robert Bork joined the faculty of Yale Law School in 1961 after seven years working in a law firm, something quite contrary to latter-day hiring practices. Trustees should not be shy about calling in the chair of the economics department and asking him why corporation economists are untouchable.
c. Reduce the number of years of study required to train for a position as a college professor by abolishing the baccalaureate degree with its distribution requirements and replacing it with a brief and specialized substitute.
d. Dignify ABD status with a formally conferred degree.
e. Consider a federal law which incorporates the moderator's never-pay-for-a-PhD. principle. Institutions can provide full scholarships for a doctoral candidate or they can purchase a franchise to do otherwise. The franchises would be fixed in number and distributed via multiple-price auctions.
