All reasoning is in service of winning arguments? I knew it all along! It's hard for me to express any skepticism of the study cited in this New York Times article without going all meta, so I'll just let the article speak for itself:
Now some researchers are suggesting that reason evolved for a completely different purpose: to win arguments. Rationality. . . is nothing more or less than a servant of the hard-wired compulsion to triumph in the debating arena. According to this view, bias, lack of logic and other supposed flaws that pollute the stream of reason are instead social adaptations that enable one group to persuade (and defeat) another. Certitude works, however sharply it may depart from the truth.
The idea, labeled the argumentative theory of reasoning, is the brainchild of French cognitive social scientists, and it has stirred excited discussion (and appalled dissent) among philosophers, political scientists, educators and psychologists, some of whom say it offers profound insight into the way people think and behave.
The Times ran a follow up blog post from one of the researchers cited in the piece, Hugo Mercier, who defends his thesis here:
We do not claim that reasoning has nothing to do with the truth. We claim that reasoning did not evolve to allow the lone reasoner to find the truth. We think it evolved to argue. But arguing is not only about trying to convince other people; it’s also about listening to their arguments. So reasoning is two-sided. On the one hand, it is used to produce arguments. Here its goal is to convince people. Accordingly, it displays a strong confirmation bias — what people see as the “rhetoric” side of reasoning. On the other hand, reasoning is also used to evaluate arguments. Here its goal is to tease out good arguments from bad ones so as to accept warranted conclusions and, if things go well, get better beliefs and make better decisions in the end.
If you'd rather, you can click here to read the study itself, or see a summary located at Hugo Mercier's own website, or listen to Edge.org's interview with him. Viewing the video below is optional; in fact, you probably shouldn't.
Wes Alwan says
Ha, you beat me to this one — although I probably would never have gotten to it; it infuriated me too much.
First, a genetic account of the development of some trait is not an account of its “function,” as the author seems to think. It doesn’t matter if reasoning developed in the service of talking Woolly Mammoths into voluntarily becoming food: its function depends on how it is fact used today in today’s environment — and the cultural environment is especially important here. Likewise, I might say the real function of computers is doing math — not email — because the function of performing mathematical calculations was highly implicated in their original development. Biological functions frequently change function during evolution, and in fact evolution wouldn’t be possible without such function-hopping. And the result might be multiple functions: knives may be for cutting, but they butter my bread very well. So this study would have to show not only that a) reasoning developed because the function of “winning arguments” was selected for in some ancestral environment and b) reasoning still has this function in today’s environment, but c) reasoning does not have other functions in today’s environment. And the author of this study is, in fact, making this idiotic claim: “Reasoning doesn’t have this function of helping us to get better beliefs and make better decisions.” Mercier simply doesn’t understand the concept of “function.”
Second, this study is itself unverifiable schlock. It is complete garbage. It cannot possible even meet the burden of showing (a). There’s a nice brief podcast (http://www.rationallyspeakingpodcast.org/show/rs18-evolutionary-psychology.html) by Massimo Pigliucci that goes over why much evolutionary psychology is, like this study, so bad. He points out that “the main problem is that evolutionary psychologists typically … tend to focus on precisely the worst examples from the standpoint of testability”; they focus on uniquely human traits and speculate about ancestral environments about which we know virtually nothing (especially given the fact that the social/cultural environment itself applies its own evolutionary pressures — and fossils and tree rings simply don’t tell us what this environment was). If we’re talking about traits that are common to other species, these are testable (e.g. fear responses to danger); if we’re talking about, e.g., why human beings have an exceptionally large brain among primates, there are plenty of competing theories among which we can’t choose; we can’t devise a test to say what selective pressure led to a larger brain. And not all traits are the results of natural selection: conceivably they might not have been subject to selective pressures at all; or are a by-product of other selected traits.
The study itself doesn’t even meet the standards of the philosophical speculation it is meant to improve upon (the ridiculously named “Cartesian” view that the function of reasoning is to create more reliable beliefs). The list of “predictions” that are meant to serve as confirmations of the theory could be accounted for by any number of other theories, including the theory the authors oppose; “confirmation bias” itself could be the result of some other faculty with some other function. Whatever the case, the theory is completely un-testable. And the idea that it has any bearing upon a philosophical theory of reasoning is a conceit: to claim that reasoning can be used to improve beliefs is not a genetic theory, and is not incompatible with any genetic theory you choose, even the theory that reasoning developed for the purpose of creating less reliable beliefs! (Some tools can be used for opposite purposes).
This infuriating garbage is not science.
So this is yet another example of the influence of scientism: in this case, pseudo-scientific nonsense that gets top billing in the New York Times as a scientific discovery overturning of some classical philosophical hypothesis. And smart people will take this dumb idea — which ultimately is just a dressing up of shallow, emotive schaudenfreude over the limits of reason — and repeat it to each other; it will be another block in the “it’s all relative, man” pyramid. Over cocktails: “But I read in the New York Times the other day that reasoning doesn’t help lead us to the truth! It’s just to win arguments.” And this all fits with my claim that scientism is in fact not scientific or rational in spirit, but bottom a form of anti-intellectualism.
Wes Alwan says
By the way, I’m not arguing against the idea that reasoning doesn’t have social functions (that we can’t detect by looking at how we in fact use it socially) or the idea that it developed in part because of social pressures. I’m saying the latter is un-verifiable (although general enough to be plausible) and the former does not preclude other functions and does not tell us what the primary function of reasoning is, or what it’s most useful for. Further, the specific claims about social function — that reasoning errors are themselves adaptive for the purposes of winning over others (as opposed to the unfortunate side-effects of other functions, e.g. the appetites) — are rank, unverifiable speculation.
And I should make clear that I’m responding to Mercier’s summary of his study, especially the absurd list of predictions, not the study itself (which conceivably
Daniel Horne says
Great response — more rants, please! But are you thoroughly opposed to all aspects of evolutionary psychology? I have no dog in the fight – I simply haven’t take the time to take the theory (“theory”?) seriously, citing to myself dumbed-down versions of your objections above. But I have a lot of respect for Robert Wright, and I know his “The Moral Animal” was very well received.
FWIW, Wright recorded a smart discussion on the subject here, with Stanford biology prof Joan Roughgarden:
Also a good exchange on E.P. between Robert Wright and John Horgan here:
Wes Alwan says
Thanks — I’ll check out the links; I always feel vaguely guilty after these rants, and I’m not opposed to all aspects (see the Pigliucci podcast for a good discussion — it’s not all low quality; a lot of it happens to be).
It seems wise to maintain the dated distinction between hard and soft science, following Karl Popper. Behavioral science is seldom falsifiable, hence it ain’t science.
If the definition of scientism is faulty observations arising from misapplication of the scientific method yet accepted owing to valid credentials, or application of the method to subject matter that is beyond falsifiability, then I condemn scientism wholeheartedly.
Isn’t there reasoning that has nothing to do with arguments, though? Like reasoning how to get home if you got lost, reasoning certain causes to certain effects, building tools and machines, and so forth and so on? Or am I confusing multiple definitions for “reasoning” here? I just think it’s strange that they said “all reasoning”.
Very interesting rant by Wes, that was a good read.
Another thing, is how do you win arguments prior to language? Isn’t reason required to invent a system of symbols for communication?
Or does the study consider war and violence to fall under “arguments and debate”?