Você está na página 1de 2

commentary

Expressing a consensus on candour


If governments are to define deception as research misconduct, science
deserves clarity and rigour in the definition.
acceptance in Scientia. As act IV opens, the Groping for norms of candour
Louis M. Guenin
fraudians have scanned OSTP’s entry and “Integrity”, despite early institutionalization
Is there a problem about misconduct in awarded it an impressively low legalism score. in “Office of Research Integrity” (ORI), is not
research? When this question first received But soon counter-examples appear. In the the root notion here. Integrity is variously
widespread attention in the United States a definition derby, there has yet to be a winner. understood as a set of virtuous dispositions
decade ago, many would have understood it This drama accentuates the influence of a or, after integer or “whole”, as the attribute of
as a query about the frequency of misdeeds. perception of science and law at loggerheads. so adhering to a set of commitments as to
As university and government investigations To be sure, the law often mangles the very keep oneself intact. The pivotal concept here
have since become routine, some of the pub- notion of science. As we encounter the US is candour, the attribute on a given occasion
lic’s fears about the incidence of infractions Supreme Court defining science, we can of not uttering anything that one believes
have been eased. But it happens that govern- only despair to read one account (Sir Karl false or misleading. We describe breaches
ments have not settled upon a rigorous defi- Popper’s) unwittingly adopted as if there of candour as deception. The government
nition of wrongdoing. were no rival1. But in respect of misconduct might reasonably assert that one who stoops
What is “misconduct”? A definition cases, two observations are telling. First, the to deception in quest of distinction betrays
recently published for comment by the White seldom-read opinions of the government’s such lack of distinction that further support
House Office of Science and Technology Poli- reviewing tribunal reveal earnest study would waste public funds3. An incident of
cy (OSTP) has returned this question to cen- of pertinent science, cogent reasoning self-promoting deception might also be
tre stage. On the premise that science deserves and painstaking appraisal of conflicting viewed as a selective disadvantage reliably
the best of scientific, moral and legal thinking evidence. Opinions in the most publicized predicting failure in grant competition. Since
combined, I propose here a different defini- cases — like Thereza Imanishi-Kari’s, almost 1987 the US government, followed recently
tion — one that not only avoids counter- all acquittals — encourage the belief that by European bodies, has prohibited “fabrica-
examples to which OSTP’s well-intentioned this panel can capably adjudicate charges of tion, falsification, and plagiarism”. Despite
proposal succumbs, but that sounds, in the wrongdoing2. Many more cases have ended the non-moral senses of “fabrication” (as
most fundamental notions of right and in admissions of guilt. synonym of “construction”) and “falsifi-
wrong, pertinent to the experimental setting. Second, in the throes of scruples about cation” (as antonym of “verification”),
language, it has escaped notice that a consen- the regulations toss off the foregoing
The imagined drama sus obtains among scientists about what offences without definition. They then add
As some would have it, OSTP opens a new act conduct is misconduct. Consider how rarely another, “serious deviation” from “accepted
in a drama featuring science as agonist, law as scientists disagree, in respect of untruthful- practices”.
antagonist. The plot begins about ten years ness, about a statement of the form “Act a by S A National Academy of Sciences panel
ago as aliens invade Scientia. The aliens are was wrong”. Fraudians have been unable to observed early on that “absence of a clear,
legislators, lawyers and bureaucrats whose find any category of conduct that one thinks explicit definition ... has impeded develop-
avowed mission is to expel dissemblers. wrong and another innocent. But by dint of ment of effective institutional oversight
Scientists immediately protest that prevari- antixenophonia — avoidance of a foreign and government policies and procedures”4.
cators are rare and that Scientia deserves tongue — the definition derby has yielded Although we inherit a storehouse of philo-
home rule. To allay native fears, the aliens only imprecise sketches of what constitutes sophical and legal machinery bearing on
announce a “collegial” investigatory process. misconduct. breaches of candour, two reasons have been
But the natives soon learn that under this By unshackling the linguistic tools with urged for shunning it. First, it is suggested
scheme, the alien prosecutor is also the which we have learnt to express norms of that such machinery will not work for the lab-
judge, and the judge conducts no trials. conduct, we might capture this scientific con- oratory and that scientists would not readily
In act II, some convicted scientists obtain sensus. I take as my task to bag that quarry. A comprehend it anyway. In fact, scientists have
hearings before a higher alien authority. To search for norms can be understood as an criticized the sketchy misconduct texts of the
everyone’s surprise, the reviewing tribunal exercise in hypothesis formation. The data past decade as vague and overreaching; mis-
acquits one accused after another. Few read consist of our judgements about what is right representation of material facts, to take the
the tribunal’s lengthy opinions, but reports and wrong. We test a hypothesis to see if it most nuanced of norms about candour, is
put down the acquittals to prosecutorial explains and correctly predicts those judge- understood even in the rough-and-tumble
incompetence. Taking a longer view, native ments. If a draft norm condemns an action world of finance. Everyone prefers clarity to
observers whimsically dubbed “fraudians” that a scientific consensus absolves, or vice opacity, brevity to verbosity. Beyond that,
doubt whether an alien tribunal can under- versa, we reject the norm, draft another, and one can say little about our storehouse’s tools
stand Scientia. In act III, frustrated alien test the new one. until we experiment with them.
prosecutors attempt to dragoon the natives The second contention is that moral and

A
into fewer hearings. Meanwhile alien leaders legal phraseology may impede ethics educa-
decide to solicit native views about miscon- bsence of a clear tion. If this is not self-contradictory, it is a
duct. Fraudians organize a competition to canard. Few of us learn that to burgle we must
select that acceptable definition of miscon- definition has break and enter a domicile at night with
duct containing the fewest alien legalisms. intent to commit a felony; we all learn that we
Many committees submit brief definitions, impeded effective should never invade another’s property. We
but each entry is either too vague or suc- usually do not learn morals from law books.
cumbs to counter-examples, and none gains
oversight and policies. When seeking to stimulate thinking by young
NATURE | VOL 402 | 9 DECEMBER 1999 | www.nature.com © 1999 Macmillan Magazines Ltd 577
commentary
scientists, instructors select provocative dis-
cussions, not rules. What is in play is clarity. To capture a consensus
Litigants would seem no better served by Research Misconduct Defined. It shall constitute research misconduct for an investigator in the conduct of
norms eschewing legal distinctions than research to commit misrepresentation, plagiarism, or misuse of another’s work.
would patients by physicians’ orders eschew-
ing medical distinctions. ● The conduct of research ● A fact is material if it or its believing the utterance false, or
consists in (i) any inquiry funded contrary affects the plausibility, (ii) while believing the effect
Nuance unbound by federal funds, or (ii) the replicability, execution, or effect misleading, omits a material
OSTP proposes to prohibit and define fabri- preparation or communication on safety of a result, hypothesis, fact without which the utterance
cation, falsification and plagiarism, and sur- related to any such inquiry of or procedure, or is such that it is misleading.
prisingly enough to convert departure from (a) a document published or would be rational for others to
“accepted practices” from a sufficient to a nec- provided to another, including alter their course of conduct in ● Plagiarism is the intentional
essary condition of misconduct5. This seems any government agency, reliance upon it. presentation of the words of
aimed at consensus wrongs, but antixeno- institution of higher education, another as the presenter’s own.
phonia has exacted a toll in precision. research institution, or ● Misrepresentation occurs if
OSTP renders fabrication as “making up professional organization, or and only if an investigator, ● Misuse of another’s work is
results and recording or reporting them”. This (b) a formal oral presentation. intentionally or recklessly the intentional presentation as
metaphor leaves the shop sense of fabrication indicating belief in the the presenter’s own, without
to differ from the regulatory only insofar as ● An investigator is a person investigator’s utterance, attribution appropriate for the
“making” differs from “making up”. That puts who participates in the conduct (i) delivers a false utterance medium, of the ideas or work of
a big load on a little adverb. OSTP goes on to of research. concerning a material fact while another.
define falsification as “manipulating research
materials, equipment, or processes, or chang- council of the National Academy of Sciences in transcription of words), this text exonerates
ing or omitting data or results such that the has opposed offences predicated on reckless- plagiarists who copy without appropriating
research is not accurately represented in the ness7. OSTP’s phrase may reflect a consensus and convicts non-deceptive patent infringers
research record”. Suppose that Jeeves deliber- to sanction some recklessness, but it encom- who appropriate without copying. OSTP
ately misstates an experimental step. (A prin- passes too much. For any first author, a paper’s would absolve copying that is not a “serious
cipal charge in one publicized case was fail- every sentence is doubtless intentional. OSTP departure”, but our considered understand-
ure to disclose pooling of patient samples.) would exclude “honest error”, but that contra- ing of plagiarism is unqualified (we do not
According to OSTP’s text, because Jeeves has dicts the condemnation of recklessness and condone “a little plagiarism”). OSTP anom-
not misstated or omitted either “data” or begs for a definition of “honest”. That returns alously condemns “reckless plagiarism”. And
“results”, no misconduct occurs. The same us to the key tests for culpability — what the its mention of work confidentially reviewed
result follows if Jeeves deliberately misstates investigator intentionally or recklessly indi- may suggest freedom to copy work otherwise
the value of a physical constant or signs a cates and what he believes. reviewed10. Review ought not to be men-
patent application asserting a falsehood. The Deceptive omissions have long been tioned: copying is the wrong, regardless of
cure for under-inclusiveness here does not lie recognized as wrongful. To show that an source. We need precisely to define plagiarism
in expanding the notion of “the research omission renders “the research” not “accu- and misuse of another’s work (see box). If
record”, for OSTP the crux. For any academic rately represented”, as OSTP imagines, one Jeeves copies by unattributed paraphrase
discipline, it is mistaken to suppose that must show the omission misleading. The rather than verbatim transcription, Jeeves
deception relates peculiarly to the record, word “material”, a threshold already observed may commit misuse.
esteemed though it be. OSTP, for example, by ORI8, forestalls prosecution of peccadil- In the norms of misrepresentation,
would appropriately capture deception in loes. In the Imanishi-Kari case, it was held plagiarism, and misuse of another’s work, we
oral presentations. What is fundamentally that alleged wrongdoing about the “January would appear to have the simplest hypothesis
wrong about deception is harm to others, not fusion” data concerned a “trivial” matter9. We explaining our considered moral judgements.
to a medium. We need to specify the activity in may define “material” for the experimental In time their etymologies may seem no more
the course of which harm by self-promoting setting (see box). OSTP would instead say alien than those of “science” and our many
deception should be forbidden. That activity that “a significant departure from accepted other borrowings from the ancients. ■
we familiarly call “the conduct of research”. practices of the scientific community for Louis M. Guenin is in the Department of
Rigour here is also simplicity. “Fabrica- maintaining the integrity of the research Microbiology and Molecular Genetics, Harvard
tion” and “falsification” are indirect ways of record” is a necessary condition for finding Medical School, 200 Longwood Ave, Boston,
expressing misrepresentation. As it happens, misconduct. This recrudescence of the Massachusetts 02115, USA.
our richest source of experiential guidance “deviation” phrase is vague and — like a track e-mail: guenin@hms.harvard.edu
devolves from disclosure requirements of the umpire ranking pole-vaulters by technique Further commentary on the author’s and OSTP’s
federal securities laws6, thousands of judicial instead of height vaulted — adverts to method proposals may be found at
encounters with which have exposed many when results control. When Amelie’s paper http://www.hms.harvard.edu/dms/cos/guenin.html
nuances of reliance on written communica- omits the data from a botched experiment, 1. Daubert v. Merrell Dow Pharmaceuticals, 113 S. Ct. 2786 (1993).
tions. An investigator induces and betrays a we do not vindicate Amelie by conjuring 2. Guenin, L. M. & Davis, B. D. Sci. Eng. Ethics 2, 47–54 (1996).
listener’s trust by signalling “I believe it” while “accepted practices” for “maintaining integri- 3. Guenin, L. M. Acad. Med. 71, 595–603 (1996).
4. National Academy of Sciences. Responsible Science (National
believing a false utterance false or a misleading ty of the research record”; we observe that the Academy Press, Washington DC, 1992).
omission misleading. This understanding discarded data are immaterial in the context. 5. 64 Fed. Reg. 55723 (14 Oct. 1999).
defines misrepresentation in the conduct of OSTP renders “plagiarism” as “the appro- 6. Rule 10b-5 under the Securities Exchange Act of 1934, 17 C.F.R.
§ 240.10b-5.
research (see box above). Here we imagine no priation of another person’s ideas, processes, 7. Council of the National Academy of Sciences. Letter to William
access to private mental states, only inferences results, or words without giving appropriate Raub, 15 March 1996.
from behaviour. credit, including those obtained through 8. 60 Fed. Reg. 10588 (27 Feb. 1995).
9. In re Imanishi-Kari, Department of Health and Human Services,
OSTP would instead capture conduct confidential review of others’ research pro-
Departmental Appeals Board, Research Integrity Adjudications
committed “intentionally, or knowingly, or in posals and manuscripts”. Apart from doing Panel, Decision No. 1582 (1996).
reckless disregard of accepted practices”. The violence to “plagiarism” (which consists only 10. Guenin, L. M. Science 271, 1790 (1996).

578 © 1999 Macmillan Magazines Ltd NATURE | VOL 402 | 9 DECEMBER 1999 | www.nature.com

Você também pode gostar