Self-correction in science

A common claim about the superiority of science over other ways of knowing is that science is self-correcting; science may take wrong turns from time to time, but it eventually finds its way back on the right road. As a supporter of science, I believe in the power of the scientific method; and generally speaking, it’s true that science self-corrects. However, it’s important to understand how human limitations—scientists are human, after all—sometimes undermine the process of self-correction.

Science will never give full understanding of anything. All that we can hope for are useful approximations of the objective reality we hope is out there. Under ideal circumstances, science is certainly self-correcting in the sense that it provides a process for arriving at consistently closer approximations. But, in too many instances the self-correcting potential of ideal science cannot overcome common human frailties. The most famous example from the field of astronomy will illustrate this.

Physics, like all of the sciences, started out as ‘natural philosophy,’ which functioned as an integrated branch of the whole philosophy/religion of the ancient Greeks. Science in its rudimentary form was thus shackled to the Greek worldview that placed humans forever at the center of the universe and effectively limited scientific thought to what would become Ptolemaic theory. This geocentric view of mankind’s place in the universe also prevailed because it conformed nicely to what the ancients observed with their limited senses, and because it had a powerful appeal to human emotions that subsequent theories could never have.

It is testimony to the power of the human mind and the potential of science that at least one individual was able to overcome all of this and figure out a closer approximation of the truth. An ancient Greek astronomer named Aristarchus proposed a heliocentric universe in the 3rd century B.C. Unfortunately his hypothesis was quickly squashed by contemporaries who condemned his idea as impious and foolish—in other words, it didn’t conform to the dominant philosophy/religion of the day. The Copernican revolution did finally take place, 1,800 years later, but those who have faith that science is the best way (or only way) to know things shouldn’t take much comfort from this example. Yes, the scientific method was eventually successful, but the self-correction was at best tragically slow.

This example has some scary implications, because the weak sister of modern science, the study of human behavior, is currently at a stage comparable to physics 2,000 years ago and shows no signs of correcting itself. With all of the social and behavioral problems facing an increasingly complex and technological world, it is possible that modern society cannot survive another 2,000 years without viable theories of individual and group behavior. So, it is important that all of us who depend on science to solve (or at least mitigate) the world’s problems understand how the three major things that prevented physics from correcting itself for about 2,000 years—the debilitating effects of ideology, the limiting nature of human perspective, and the immense power of emotions to mislead—are still at work today preventing the newer branches of science from correcting themselves.

To appreciate the ways in which science’s ability to self-correct can be thwarted, one has to be very clear about the basics of the scientific method. They can be outlined in rudimentary form in the following manner:

  • Preliminary observations of some natural phenomenon are made
  • A scientist brainstorms possible explanations of what is observed
  • A workable hypothesis is formed
  • An organized plan for additional observations and experiments is made and carried out
  • If additional evidence for the hypothesis is found, it advances to a level of confidence higher than that of a hypothesis but lower than that of a theory (we can call it a conjecture)
  • The newly elevated conjecture is then presented for the peer review process, in which some scientists will find evidence to support the conjecture and others will try to tear it apart
  • If the conjecture survives the peer review process and gains additional evidence and the support of a large number of scientists it will eventually become accepted as a viable theory

The two parts of this process that make self-correction possible are the brainstorming and peer review stages. Unfettered brainstorming makes it possible for scientists to consider all possibilities—that’s how we got Einstein and Georges Lemaître, the father of the big bang1. If religion or philosophy makes some ideas unthinkable, the brainstorming stage will be inhibited, and ‘unpopular’ possibilities will be missed. The humanist philosophy that dominates the behavior and social sciences departments today is making the self-correction in those fields impossible just as much as the philosophy of the ancient Greeks made physics impossible.

The peer review process makes it possible to challenge popular but false notions. In modern times this stage has become highly susceptible to the negative influences of politics and government funding. The controversy over global warming / climate change is a good example. Regardless of a person’s views on climate change, it should be deeply disturbing that one side of what should be a scientific debate has been corrupted by government funding into political advocacy. When any scientist becomes an advocate of policy, he is no longer a scientist, because science can only serve one master—the search for truth. It is even more disturbing that those on the other side of the scientific debate have been tagged with the vicious label of ‘deniers.’ Those who use this label in such a pejorative manner are trying to preemptively shut down the peer review process and mandate scientific orthodoxy.

It is dangerous, therefore, to assume that science has an inherent ability to overcome human failings to the point that we can depend on it to be self-correcting. That it can effectively reach that goal is demonstrated by the fairly rapid acceptance of big bang theory in the mid-20th century over the strong objections of those who were philosophically opposed to it.  But what has only recently become true of a branch of science that is over 450 years old, is not true of the newer sciences. Biology, psychology, and the social sciences are nowhere near the stage where self-correction is automatic.

Now, there is one assumption I’ve made in this discussion, that self-correction means we’re making better and better approximations of reality. But there is another issue: what if there isn’t always an objective truth that we can get closer to by self-correction? That’s an altogether different topic, but let me say that relativity and quantum mechanics do suggest that this might be true. This is a topic for another article.

[1] “Father” in more than one sense: Lemaître was also a priest.

Recommended reading:

  • The Structure of Scientific Revolutions by Thomas Kuhn