"She's like the poster child for collaboration". Jon Miller was talking last weekend at the AAAS session about scientists' virtues. The "she" was Jennifer Doudna, who had given a talk the evening before.
She's also one of the scientists at the center of an intensely bitter public dispute over credit for discovering what CRISPR can do.
With a patent showdown coming up in March, and "who gets the Nobel prize" stakes as well, it all exploded in January. One of the most powerful science journals (Cell) published a partisan CRISPR history without so much as a conflict of interest statement.
(Some places to catch up on what happened and the strength of feeling: Sharon Begley in STAT, Nathaniel Comfort on Genotopia, Michael Eisen on it is NOT Junk, Carolyn Johnson in The Washington Post, Joanna Rothkopf in Jezebel, Sarah Zhang in WIRED.)
"From the outside, it's kind of comically petty", Sarah Zhang wrote. "…As the conversation expanded though, its importance came into focus: The episode heralds not only how science gets done today, but how the narrative of scientific discovery gets written".
It's not petty. And it's not new. "Creeping propertization" of science, especially after World War II, is making it more acute, and the sheer growth in the number of scientists makes multiple discovery more likely. But sociologist Robert Merton long ago argued that struggle over credit and priority has never really been aberrant: it's typical.
Consider this from Galileo about a rival: he "attempted to rob me of that glory which was mine". He used "a sly way of attempting to establish his priority".
From Halley about another scientist: he is "a lazy and malicious thief", "as lazy and slothful as he is corrupt".
Descartes: "tell him as little as possible about…my unpublished opinions, for if I'm not greatly mistaken, he is a man who is seeking to acquire a reputation at my expense and through shady practices".
The thing is, scientific results require validation. We can be sure we're onto something, or have committed ourselves to something really worthwhile, but it's not for us to determine the value of what we've done. We're too biased and prone to error. So the attention and esteem of other scientists matters. And it matters a lot. Merton argued:
Rather than being mutually exclusive, joy in discovery and eagerness for recognition by scientific peers are stamped out of the same psychological coin. They both express a basic commitment to the value of advancing knowledge.
Aspirations and ideals can fall a long way from behavior, in science as in anything else. A 2002 survey of over 3,400 NIH-funded scientists by Melissa Anderson and colleagues found that they encountered more behavior that contravened scientific norms than embodied it. In the same group plus another group of more than 4,100 early career researchers, Anderson and colleagues found a lot of agreement about what good scientific norms are – but some differences between the two groups as well.
Anderson and her colleagues work from the basis of Merton's norms – aspirations and ideals for scientific work. He condensed a range of issues around scientific practice and integrity into 4 norms at first (in 1942), then added another years later. Physicist John Ziman gave them an acronym that stuck: CUDOS. (The O was the later add-on). In the Mertonian norms, advance of scientific knowledge needs scientists to be:
- Communalist – common ownership and openness of knowledge and scientific discoveries
- Universal – validity depends on scientists transcending race, gender, religion, and nationality
- Disinterested – not striving for personal gain
- Original – aiming to expand the body of knowledge
- Skeptical – detached logical scrutiny, based on empiricism
Merton wrote about humility a lot. And Ragnvald Kalleberg [PDF] has written that it should be norm 6: turning the acronym into CUDOSH.
This slide of 10 virtues were the top-ranked ones by hundreds of respected scientists Robert Pennock and Jon Miller studied. Honesty and curiosity were way ahead of the pack – being collaborative came last. (Only 7% counted it one of the 3 most important attributes of a scientist).
Only one has a rider on it, and it's revealing. Humility is limited towards evidence only. Miller said, "At first we had humility as a virtue, but that didn't resonate very well". Once restricted, it managed to make the list – but even then only 20% ranked it in their top 3.
One thing this new list of virtues has in common with the Mertonian norms is that they're both based on the worldview of U.S. academic science (and thus the society in which its embedded).
Pennock and Miller's first analysis comes from a group that's close to 90% men, with few people of color. (They're in the process of studying scientists earlier in their careers.) Ouch.
Being a scientist doesn't make you immune to societal values and circumstances. How critical you can be of senior people, how much courage it takes for independence of thought, how much collaboration is valued – these things aren't the same globally. Even though much of the dominant western science culture spreads, that doesn't mean that values, and the attributes you need to be a good scientist, are universal.
Robert Merges is a lawyer. What he writes about science in relation to its attitudes to sharing science's results and products, is an interesting perspective more generally. Science is not so much a commons, as a "more or less well-identified circle of similarly situated scientists" who share "a largely implicit code of conduct".
What's more, the CRISPR fight is a reminder that where we're headed with patents could lead us to a "tragedy of the anticommons" too, where too many people have the right to exclude too many others from advancing knowledge.
Science has a status bias problem of people associated with prestigious institutions, gender, race, and being a native speaker of English. And the U.S. as a country forms another elite (see for example, in getting an article accepted). There are serious questions to ask about what weight the values of elites and insiders should have on matters like humility about science, and humility in science and scientists. And because science culture can spread, and could change, the conversation about values is worth working on.
Humility and generosity, and the other opposites of arrogance and pride, seem to me to be central to advancing knowledge. Marguerite La Caze goes back to Descartes, on the philosophy of generosity [PDF]: "If we have generosity, we will not prefer ourselves to others". If we're serious about humility and generosity, I think it leads straight to valuing and enabling collaboration, openness, and diversity.
"We just have to change the incentives". I hear that a lot, when it comes to scientists' behavior. Especially in debates about journal impact factors, or post-publication review – and this week, about preprints, at the ASAPBio meeting (#ASAPBio on Twitter). "Academic incentives" tends to be code for the journal citations it takes to get grants, jobs, promotions, and career opportunities like fellowships. Change that, and inexorably, carried by that weight, the problem will be solved.
Maybe. I don't want to downplay that kind of incentive at all. But the social and psychological power of credit to scientists is powerful, too. That kind of less obvious factor can block change, or just express itself by re-forming into the same patterns in the new setting – right back where we started.
Our behavior – and any failure and unintended negative consequences of interventions intended to change it – will be inexplicable if social and psychological motivations are discounted. It's not going to be easy to run counter to something coded into science's DNA. We might have to work on editing that, too.
Explore further: What values are important to scientists?