When It Comes To Thinking, 2 Fish Heads Are Better Than 1

Transcript

STEVE INSKEEP, HOST:

Once again on today's program, we're pretty much covering everything, including this report on how fish think.

DAVID GREENE, HOST:

NPR social science correspondence, Shankar Vedantam, is here to talk about this. And OK, Shankar, this is taking us to a new direction - fish researching fish as part of social science.

SHANKAR VEDANTAM, BYLINE: I'm going to argue it is, David, and here's why.

GREENE: OK.

VEDANTAM: Social sciences have long known that when you aggregate answers from people you aggregate their estimates, you often get...

GREENE: What do you mean, aggregate estimates - what do you mean here?

VEDANTAM: Well, think about, for example, polls that are conducted during elections. If you take a bunch of polls and add up all the answers and get the average of the polls, the average of the polls is likely to give you a better estimate of which way the election is going than any individual poll.

GREENE: I see.

VEDANTAM: It turns out this principal is not just true for human beings and polls. It's also true for fish.

GREENE: Are pollsters just getting bored with humans? Are they deciding to survey fish now? What's happening?

VEDANTAM: Well, Christian Agrillo, at the University of Padova, Italy found that when two fish swim together, they make better decisions than when the fish are swimming alone. So when you place them in a situation where they have to decide which way to swim, the fish end up making better decisions when they're swimming in pairs than when they are swimming individually.

GREENE: And why do we think that is?

VEDANTAM: We're not entirely sure. Among humans, aggregation seems to work because the overly positive estimates cancel out the overly negative estimates. With fish, it seems to be more of a meritocracy David, which is the fish somehow figure out which fish is the smarter fish, and they both end up doing whatever the smarter fish decides.

GREENE: Is this a model for we humans?

VEDANTAM: You know, I can imagine presidential debates being completely revolutionized by this insight, David.

GREENE: Oh, you mean the two candidates, instead of fighting with each other, one would eventually decide, oh you're smarter than me so I'll just go with what you're saying.

VEDANTAM: Exactly. Would solve all the problems.

GREENE: That will never happen.

VEDANTAM: Of course not.

GREENE: And maybe we're going to leave the studio now and I'll follow you. You tell me which way to go.

VEDANTAM: The window looks good, David.

GREENE: That's awesome. Shankar Vedantam, thanks for being here as always.

VEDANTAM: Sure, David.

GREENE: I think we've now heard it all, Steve. That was NPR's Shankar Vedantam, who joins us each week to explain interesting social science research. And he's on twitter@hiddenbrain and Steve and I are tweeting throughout the day as well.

INSKEEP: Yeah, you can find links to stories from the program, other things we find interesting. The program is @morningedition. This guy to my left is @nprgreene.

GREENE: And you're @nprinskeep.

INSKEEP: This sounds good to me. You're listening to Morning Edition from NPR News. Our theme music was written by B.J. Leiderman, arranged by Jim Pew. I'm Steve Inskeep.

GREENE: And I'm David Greene. Transcript provided by NPR, Copyright NPR.