Sunday, February 23, 2014

Against "consensus" science

I often cringe when I hear the media use the word consensus in reference to some area of science. The concept of consensus is antithetical to the scientific method. I have no problem talking about "mainstream" or "conventional" science, or saying that "most scholars accept" some claim. But "consensus" is not right.

Much like the equally bad phrase "settled science," consensus implies that the debate is entirely over. That indisputable proof exists. That anyone who disagrees is simply wrong as a matter of objective fact. Worst of all, it implies that truth can be determined by a majority vote.

Who believes in gravity?

In science, nothing is ever final. Everything is constantly subject to re-evaluation and re-testing. In fact, if a statement is not falsifiable, then it's not considered a scientific claim in the first place. And it takes only one person--not a majority--to overturn an existing finding.

"But wait a minute, Jason," every reader is surely now saying. "Didn't you use the term consensus to describe the foundations of IQ research?" Guilty as charged, but that was actually a calculated appeal to the journalists who formed my target audience. The reference is in "Why can't we talk about IQ?" published by Politico. (Of all my responses to the dissertation controversy from eight months ago, that's the one I am most proud of.) To show that the media had denounced scientific findings about which there is little technical dispute, I wrote:
What scholars of mental ability know, but have never successfully gotten the media to understand, is that a scientific consensus, based on an extensive and consistent literature, has long been reached on many of the questions that still seem controversial to journalists. [emphasis added]
I went back and forth on whether to use the word consensus to describe that set of views shared by nearly all cognitive psychologists. In the end I went with it, because the word carries a certain resonance with my intended audience. It was an appeal to those on the Left who are accustomed to hearing consensus thrown around in the context of global warming, second-hand smoke, same-sex parenting, and so on.

I wanted Politico readers to consider that they might be just as wrong about IQ as they believe their opponents are about those other issues. So I deliberately used "consensus" rather than "mainstream" or "rarely disputed."

Anyway, this whole post was motivated by the alleged consensus on government preschool, which I discussed in National Review recently.


  1. I went back and forth on whether to use the word consensus to describe that set of views shared by nearly all cognitive psychologists.

    That's not true. Nearly all differential psychologists may share those views, but not cognitive psychologists. Lots of cognitive psychologists know nothing about the g factor or individual differences in intelligence, because that's not what they study.

  2. If differences in IQ are heritable and we accept evolution exists, we must then allow that differences in IQ may not be permanent. Rather, they are based on evolutionary context. If there are differences in IQ, they can not definitely claimed to be racial in nature, but merely contextually racial. Thus, they may one day come to look very different. It's a good thing that science is not settled.

  3. And it takes only one person--not a majority--to overturn an existing finding.

    With respect, Jason, this is nuts. One person may, like Wegener, have a contrarian idea that eventually becomes accepted; but he first has to convince a majority of researchers in his field that he's right. He has to, in other words, persuade a ... consensus.

    And "consensus" is not a synonym for "unanimity." Look it up.

    1. There is a difference between overturning an existing finding for one person, and overturning an existing finding for the majority.

    2. My Theory on Why East Asians and Ashkenazi Jews Have Highest Average IQ Scores in the World

  4. This comment has been removed by the author.

  5. Hi Jason, I wrote something..

    The American Dream is Already Dead: How Jason Richwine is Actually Right in His US Immigration Policy that Disfavours Races with Lower Average IQ

  6. The problem is that one can't be an expert in all fields. Thus, the safest route is to rely first on consensus view when outside your own area of expertise.