Motivated Reasoning
Why Did We Evolve the Ability to Reason? To Argue!
The final step in human evolution is—the lawyer?
Posted March 31, 2011
In a fascinating new paper in the journal Behavioral and Brain Sciences titled "Why Do Humans Reason? Arguments for an Argumentative Theory," Hugo Mercier and Dan Sperber offer the hypothesis that human beings' capacities for reasoning evolved to enable them to craft and evaluate arguments better and therefore enhance communication (and survival).
Typically, we assume that the benefit we get from reasoning is the ability to make better decisions. And if better decision-making enhances our chances of survival, then the process of natural selection would favor those with better faculties of reasoning. However, the ability of reason to generate better decisions has come under heavy criticism lately, due to the discovery of widespread, systematic, and predictable cognitive biases and dysfunctions, such as the well-known confirmation bias (by which new information that confirms our beliefs is overemphasized, and information contrary to them is dismissed).
Why, then, did we evolve as reasoning creatures, if our reason does not help us make better decisions? Mercier and Sperber make the case that human reason did not evolve because it allows us to make better decisions, but instead because it allows us to make better arguments, as well as evaluate the arguments of others. As they say in the conclusion to the paper:
Reasoning contributes to the effectiveness and reliability of communication by enabling communicators to argue for their claim and by enabling addressees to assess these arguments. It thus increases both in quantity and in epistemic quality the information humans are able to share. (pp. 71-72)
Focusing on the value of communication to human survival and evolution, Mercier and Sperber argue that, in order for communication to be reliable, there must be some way for people to convey the truth of what they say, and also to judge the truth of what others say to them ("epistemic vigilence"). Certainly, one mechanism that evolved to allow us to do this is the "tell," subtle bodily signals that reveal one is lying. But judging others' communication isn't necessarily as simple as distinguishing between truth and lies; sometimes several people are sharing their opinions, and the listener has to decide which of them (if any) to believe. And how does the listener do that? Ideally, by evaluating the arguments they make in support of their positions—and this, according to Mercier and Sperber, is why we developed the capacity to reason.
This is also why, they claim, our reasoning does not seem to produce good decisions—because that's not what it was selected to do! If reason developed to allow us to make better arguments, to support our positions and defend our actions, then we should be able to point out particular aspects of it that serve this purpose better than the purpose of good decision-making. And Mercier and Sperber spend much of the paper doing just this: for instance, the confirmation bias does not help make good decisions, certainly, but it does help a person to craft good arguments by allowing him or her to filter out counterproductive information and focus on just the facts (and other arguments) that will support his or her case. So rather than being a flaw when judged against the standard of good decision-making, the confirmation bias may be an evolutionarily selected trait that promotes better argumentation, communication, and therefore survival.
Thinking more strategically, reason also allows us to anticipate disagreement with our beliefs, opinions, and actions, which may lead us to craft counterarguments against potential disagreement, engaging in what scholars call "motivated reasoning." In cases of motivated reasoning, the motivation is taken, not to be truth, which seems most noble and in line with traditional understanding of the purpose of reason, but rather to be success, namely in whatever you are arguing for.
This seems to describe the behavior of the participants in structured debate in schools, or lawyers in a courtroom, in which the goal is to win the argument or case. We would like to think, of course, that "victory" in either case is correlated with finding the truth, but this is more likely to characterize the outcome of the debate or trail as a whole, rather than the intention or behavior of any particular party. A debate competitor is trying to win the argument, and an attorney in a trial is trying to win the case for his or her client (or the state if he or she is a prosecutor). Rather than bemoaning their motivated reasoning, we count on it to drive debaters and attorneys to make the best arguments possible, which then contributes to the overall goal of the process: truth. Much like sports, in which we want all the players to be driven to win so we spectators can enjoy a good game or match, adversarial systems such as debates and trials depend on the participants to be goal-oriented, to be motivated by winning, in order to get the best performances (arguments) out of them. Only then will we get the best result from the overall process: a resolution to an argument or dispute that incorporates all the available facts and arguments that all parties could provide, therefore providing the best possible approximation of truth.
Of course, the bête noire of argument in the current day is political argument, in which talking heads in politics and the media seem to talk past each other at best, and bark at each other at worst. But is this all bad? Only if people listen to such "debate" and think that all the participants are trying to convey objective truth out of the goodness of their hearts, rather than pursuing a goal. But if we take into account what their actual goal—winning an argument, promoting a candidate or platform, influencing opinion, or assassinating character—and we take what they say in that context, it can still provide us with a lot of useful information (and theater!).
If we really want to improve the quality of political debate, we should focus on improving the structure of poltical debate, not the incentives of the participants. If we force them to make good arguments, we might actually get good arguments from them—and, if Mercier and Sperber are right, they would be fulfilling their evolutionary destiny as well!
-----
If you're interested, see Mercier's website devoted to the ideas explored in his and Sperber's article, featuring other foundational work and background.
You can follow me on Twitter and also at the following blogs: Economics and Ethics, The Comics Professor, and The Literary Table.