Thou shall not take sides - preprint available

Does taking sides make us dumb? Our very fallible human reasoning

We all believe in something. And among these beliefs, we do not limit ourselves to moral choices. We also make statements about how we think the real world is. And we feel that these beliefs, this taking sides is not only not problematic, but it is also the right thing to do. A new preprint from André Martins at aArXiv.com defends that the opposite is true. By combining results from Logic and the results of several known experiments on how we human reasons, his paper shows that the very action of taking sides and believing should be avoided because it can prevent us from learning. Taking sides, it seems, turns even the very smart among us dumb. Simulations presented in the paper also show that simply wanting to have one option to believe might be at the heart of the appearance and strengthening of extremist points of view. Some serious problems that arise from our taking sides also identified as generating errors in scientific inquiries. The author presents a discussion of the problem, how it affects society and research in general and suggests ways to make the problem of believing less serious. When Socrates claimed he knew nothing, he had almost completely understood the problem. But we forgot it and we kept choosing descriptions of the world we choose to defend. The problem, according to the paper, is not what we believe; the problem is that we believe.

The Results

We have been aware for a few decades now that, when we humans reason naturally, we commit a staggering amount of mistakes. We make very basic logical errors, we are influenced by others in ways that can make us fail even trivial tasks where we would not have failed at all. And we fail at those tasks just to agree with our group. Our reasoning does allow us to navigate the problems of the daily life in a reasonably competent way but, without training, it does not go further than that. We are simply much dumber than anyone among us would want to believe. New trivially simple logical problems are often beyond our natural skills. More recently, researchers have also observed in their experiments that we the main function of our reasoning is not as a tool for solving problems by finding the best available answer. Instead of looking for truth or best answers, our reasoning seems to have, as its primary function, a simple argumentative cause. We reason to establish arguments to convince other people and, as long as our arguments work, they do not need to be right. Quite the contrary, they could even be quite incompetent and wrong, as long as they get the work done. As in any heated discussion, our brains seem to work to get us ahead in the fight with our opponents. And, in this fight, being right or wrong is just an incident. What matters is that we sound convincing. At the very least, we should sound convincing to the social group we belong, to those people from where we draw validation. And when our social group is attacked by outside ideas, we use our brain power not to examine the situation and try to find who is right. We use it to defend the group we feel where we belong. We hold sets of beliefs that show clearly that we do not reason in an independent, correct way. Instead, we look for justifications for the conclusions we want to be true. And, in an almost unexpected twist, the smarter we are, the better we are at the defending our conclusions. And that can make smarter people less capable of changing their opinions, of learning.

In a recent preprint “Thou shall not take sides: Cognition, Logic and the need for changing how we believe”, currently at the ArXiv preprint server at http://arxiv.org/abs/1508.05169, André Martins, from Universidade de São Paulo has combined these results from Cognitive Psychology with our current knowledge on Logic about accepting propositions as true. By observing that our logical methods can not provide any certainty about statements about the real world and that the best we can hope for is to find ways to estimate probabilities for those statements, it becomes clear that our side-taking has no logical origin. While there are statistical methods that try to accept or reject ideas, they are actually a desperate attempt to justify our desire to have just one idea to believe. As it has always been clear that they never ruled out any ideas. While there are circumstances where one might decide to act as if one idea is true, in logical terms, only an estimate of probability makes any sense (and even this might be beyond our current reach for many problems, given our actual state of knowledge). The conclusion: we should never take sides and adopt beliefs as if those ideas were true. For there is no logical reason to do that and, when we embrace these identity-defining beliefs, our instincts make us unable to learn, they make our minds work to defend those beliefs instead of inspecting them. Taking sides and believing make us stupid; sometimes the more stupid, the more intelligent we are. Unless you actually just care about defending your side, regardless of who is right, you should not hold beliefs. You can and actually should have probabilistic beliefs, but those are just uncertainties where some ideas are more uncertain and others less uncertain.

Simulations in the preprint also strongly suggest that our desire to have one single idea to believe and defend is a fundamental key in understanding the origin and spread of unjustified extreme beliefs. Indeed, a simple change in the mental model of the agents, where they start looking for answer that combine the available ideas instead of looking for one to discard, has the amazing effect of avoiding extreme beliefs in the problems studied in the preprint.

Finally, why we actually can trust Science is discussed. And we learn that, despite being the competent and correct way to learn about the world, scientific work still has much room for improvement. As scientists are also humans, the preprint shows the consequences of the desire to have one single idea in the beliefs of different areas. And, while some scientific fields are lucky enough that the mistakes the researcher do there on a daily basis are of no consequence to the reliability of their ideas, in other areas the problems do cast much larger doubts on the reliability of our knowledge. The author shows that taking sides causes us problems in all aspects of our lives. We become less capable of learning, groups become extremists with no reason, and even our scientific knowledge has suffered greatly from this natural instinct. The conclusion is that we should start learning as soon as possible to never take sides on issues that are not just moral issues, but that include statements about how the real world is. It is not that our emotional side can drive us to wrong decisions, our ability to reason does exactly the same. We have tools to avoid that (or, at least, to minimize the problem), it is fundamental that we learn to use them. One of them is to never trust our own beliefs. They may be probable. But we simply never know if they are right.