A Daniel Sarewitz wrote an opinion piece in Nature awhile back to argue that the pressure to publish regularly has driven down the quality of science. Moreover, he claims to have identified

…a destructive feedback between the production of poor-quality science, the responsibility to cite previous work and the compulsion to publish.

Sarewitz ends with an exhortation of sorts. To publish much less.

Current trajectories threaten science with drowning in the noise of its own rising productivity, a future that Price described as “senility”. Avoiding this destiny will, in part, require much more selective publication. Rising quality can thus emerge from declining scientific efficiency and productivity. We can start by publishing less, and less often, whatever the promotional e-mails promise us.

Interestingly, this “Price” he seemingly follows in thought wrote in 1963, long before modern search engines were remotely conceivable and was, as Sarewitz himself observes “an elitist”.

Within a couple of generations, [Derek de Solla Price] said, it would lead to a world in which “we should have two scientists for every man, woman, child, and dog in the population”. Price was also an elitist who believed that quality could not be maintained amid such growth. He showed that scientific eminence was concentrated in a very small percentage of researchers, and that the number of leading scientists would therefore grow much more slowly than the number of merely good ones, and that would yield “an even greater preponderance of manpower able to write scientific papers, but not able to write distinguished ones”.

Price was worried about “distinguished”, but Sarewitz has adopted this elitism to claim that pressure to publish is actually causing the promotion of mistaken or bad science. And so we should all, I surmise, slow down and publish less. It is unclear what Sarewitz thinks about the “merely good” scientists identified by Price and whether they should be driven away or not. If not explicitly stated, this piece does have a whiff of ol’ Steve McKnight’s complaints about the riff-raff to it.

Gary S. McDowell and Jessica K. Polka wrote in to observe that slowing the pace of publication is likely to hurt younger scientists who are trying to establish themselves.

In today’s competitive arena, asking this of scientists — particularly junior ones — is to ask them to fall on their swords.

Investing more effort in fewer but ‘more complete’ publications could hold back early-career researchers, who already face fierce competition. To generate a first-author publication, graduate students on average take more than a year longer than they did in the 1980s (R. D. Vale Proc. Natl Acad. Sci. USA 112, 1343913446; 2015). Introducing further delays for junior scientists is not an option as long as performance is rated by publication metrics.

One Richard Ebright commented thusly:

Wrong. All publications, by all researchers, at all career stages, should be complete stories. No-one benefits from publication of “minimum publishable units.”

This is as wrong as wrong can be.

LPU: A Case Study

Let’s take the case of W-18. It hit the mainstream media following identification in a few human drug overdose cases as “This new street drug is 10,000 times more potent than morphine” [WaPo version].

Obviously, this is a case for the pharmacological and substance abuse sciences to leap into action and provide some straight dope, er information, on the situation.

In the delusional world of the “complete story” tellers, this should be accomplished by single labs or groups, beavering away in isolation, not to report their findings on W-18 until they have it all. That might incorporate wide ranging in vitro pharmacology to describe activity or inactivity at major suspected sites of action. Pharmacokinetic data in one small and at least one large experimental species, maybe some human if possible. Behavioral pharmacology on a host of the usual assays for dose-effects, toxicity, behavioral domains, dependence, withdrawal, reward or drug liking, liability for compulsive use patterns, cognitive impact with chronic use. This list goes on. And for each in vivo measure, we may need to parse the contribution of several signalling systems that might be identified by the in vitro work.

That is a whole lot of time, effort and money.

In the world of the complete-story tellers, these might be going on in parallel in multiple lab groups who are duplicating each others’ work in whole or in part.

Choices or assumptions made that lead to blind alleys will waste everyone’s time equally.

Did I mention the funding yet?

Ah yes, the funding. Of course a full bore effort on this requires a modern research lab to have the cash to conduct the work. Sometimes it can be squeezed alongside existing projects or initial efforts excused in the pursuit of Preliminary Data. But at some point, people are going to have to propose grants. Which are are going to take fire for a lack of evidence that:

1) There is any such thing as a W-18 problem. Media? pfah, everyone knows they overblow everything. [This is where even the tiniest least publishable unit from epidemiologists, drug toxicologists or even Case Reports from Emergency Department physicians goes a loooooong way. And not just for grant skeptics. Any PI should consider whether a putative new human health risk is worth pouring effort and lab resources into. LPU can help that PI to judge and, if warranted, to defend a research proposal.]

2) There isn’t anything new here. This is just a potent synthetic opiate, right? That’s what the media headlines claim. [Except it is based on a the patent description of a mouse writhing task. We have NO idea if it is even active at endogenous opiate receptors from the media and the patent. And hey! Guess what? The UNC drug evaluation core run by Bryan Roth found no freaking mu opioid receptor or delta or kappa opioid receptor activity for W-18!!! Twitter is a pretty LPU venue. And yet think of how much work this saves. It will potentially head off a lot of mistaken assaying looking for opioid activity across all kinds of lab types. After all, the above listed logical progression is not what happens. People don’t necessarily wait for comprehensive in vitro pharmacology to be available before trying out their favored behavioral assay.]

3) Whoa, totally unexpected turn for W-18 already. So what next? [Well, it would be nice if there were Case Reports of toxic effects eh? To point us in the right direction- are their hints of the systems that are affected in medical emergency cases? And if some investigators had launched some pilot experiments in their own favored domains before finding out the results from Roth, wouldn’t it be useful to know what they have found? Why IS it that W-18 is active in writhing…or can’t this patent claim be replicated? Is there an active metabolite formed? This obviously wouldn’t have come up in the UNC assays as they just focus on the parent compound in the in vitro assays.]


Science is iterative and collaborative.

It generates knowledge best and with the most efficiency when people are aware of what their peers are finding out as quickly as possible.

Waiting while several groups pursue a supposed “complete story” in parallel only for one to “win” and be able to publish while the other ones shake their scooped heads in shame and fail to publish such mediocrity is BAD SCIENCE.