The enterprise of Science for a long while now has been a good trade, and it’s important to keep it that way. Professionals need standards, and for Science the most important standard of all is Peer Review.
Peer Review, for the laity, is the practice by which research is evaluated and refereed by other scholars in the field before work is allowed to be published. In academia, this is the basic mechanism for quality control, research grants, and career advancement. Although Peer Review has a long history, it did not become a major force in Science until after World War II.
There have recently been criticisms of Peer Review, and of the institutions of Science more generally, some of which are quite lowbrow. These criticisms are dangerous and come off as rather histrionic. Since 1950, the number of PhDs in the United States has increased by a factor of almost 30, and roughly 55,000 new PhDs now graduate every year. The only way to manage so many professional academics and make their work legible is through control and standardization.
Peer Review gives us a barrier to entry; it helps filter out anything that’s too reckless, unconventional, plainspoken, or, most worrisome of all, threatening to our good trade. We need gatekeepers. How else can we ensure that scientists are serving Science?
Yes, maybe someone like Einstein couldn’t get published today. His paper on general relativity, as Frank Wilczek points out, muses on the philosophy of space and time, recapitulates tensor calculus at length, includes no charts or colorful graphics, and, most damningly, cites the work of other physicists exactly zero times. No filter is perfect—without Peer Review, how much nonsense would we have to suffer to get one Einstein paper?
And yes, Peer Review can occasionally feed popular biases and permit established names to domineer their fields, by keeping out papers that present threatening ideas or fail to sufficiently cite the conventional wisdom. For example, Alzheimer’s research languished for 30 years because work deviating from the favored hypothesis—that beta-amyloid was a key cause of the disease—was rejected, censored, or willfully ignored by scholars. A tiny thimbleful of other embarrassing examples exist, but I’ve yet to see such cases examined in Peer-Reviewed papers, so we should look on this phenomenon with a measure of skepticism.
And yes, I’m willing to concede that Peer Review is slow, expensive, inconsistent, and regularly fails to catch fraud. And yes, because career success is built on Peer Review, it may incentivize researchers to massage statistics, self-promote, and do incremental, uncreative work that can be reliably published. And yes, most research findings turn out to be false. But none of these are arguments against Peer Review. If anything they suggest that we need more Peer Review, not less, to combat all of these creeping problems!
Let’s be honest: most people aren’t credentialed enough to sort out good research from bad. Peer Review gives us an authoritative voice. What matters is that the great mass of people can point to Nature or The Lancet and know that something is True, because it has been subjected to Peer Review and is therefore proper Science.