April 14, 2005, 9:48PM
Humans are hard-wired to trust one another. So says Dr. P. Read Montague, director of the Human Neuro-imaging Laboratory at Baylor College of Medicine. "You have to erode that trust. Con men have survived on that for thousands of years."
|
ADVERTISEMENT
|
A report by Montague in the April 1 issue of Science examines how people develop trust in business relationships.
Houston Chronicle reporter Lynn J. Cook recently spoke with Montague about his neuroeconomic study that marries brain imaging software with financial games that probe how people react to situations when money is at stake.
Q: What's the premise for this study?
A: Two subjects are simultaneously put into special magnetic
resonance imaging scanners at Baylor and the California Institute of
Technology. They play 10 rounds where one is the "investor" with $20
and the other is the "trustee." The investor can keep the money or
choose to invest it with the trustee, knowing it will triple in value.
Where it gets interesting is seeing
how the trustee divvies up the profit and how that encourages — or discourages — the investor from placing more money in
the next round.
If it's divided 50-50, the next time the investor might decide to invest more. If the trustee keeps it all, then in the next round the investor might not give up even $1.
Q: How did people behave?
A: First, there's some background you need to know. The test is
mathematically engineered so that cooperation yields the most money in
the game. There's a real financial incentive to trust. And the study
was anonymous. This was a critical feature. It's like doing a deal over
the Internet in your closet. Our scanner linked to Caltech's scanner,
and these people didn't meet. They were 1,500 miles apart and had a
strong assurance that they never would meet.
So, when you're not seeing me, reciprocity is the biggest predictor. It's a case of tit-for-tat behavior. If the trustee gives back a greater amount of money — or withholds it — these are strong signals. People's brains respond to that.
We saw brain signals about intention to increase or decrease trust on the next play. If I were monitoring your brain as the game proceeded, we could start to see the signals before the actual decision was made. We call that "intention to trust," and it's centered in the part of the brain called the caudate nucleus. We also saw signals that shift relative to time that correlate with the building of reputation. We think we're seeing the building of a model of the trustee inside the brain of the investor.
Q: Where does the study go next?
A: Right now we're doing another version of this trust exercise that hasn't been published yet. This time it's personal.
Two strangers meet in our waiting area. They're instructed on the test. They both get into our scanners, and they're paid according to how well they do. They know they'll see each other again. The results are completely different when it's personalized.
The best way to describe it is that in the personalized version, the subjects start to immediately operate on a set of assumptions about the other person, not just the financial data they are sending.
Next we're going to study subject behavior between Houston and Germany and China to see whether the styles of play and brain responses are different between different cultures. We think they will be. At least that's our hypothesis. Either answer is important. If people respond differently, it's important. And if they don't, that's equally important to know.
Q: If people operate on assumptions when they meet in
the study, that tells me they're craving as much information as they
can get, even if it's misinformation. Would you disagree?
A: Roll the clock back half a million years. These mechanisms
were forged back then, when it was very efficient to operate on
internal models and not collect good data on other people. This was
back to when survival was a big deal — back when we were starving all
the time and every efficiency counted. So even if the internal model is
bad, if it's good enough, then that's efficient. Moral systems are
similar. We don't have to sit around in a quandary about a lot of
situations. We know if a certain case comes up, we'll act a certain way.
Q: What got you interested in trust in the first place?
A: Trust is a gigantic concept. Think of how many novels are
written on trust and betrayal — they connote relationship
entanglements. It's as complicated as human interaction gets. But you
can strip away a lot of that and look at the central feature of a
transaction. I do something for you, and I do that even though it's not
risk-free. But I'm hoping you will, in some measure, reciprocate.
There are papers in economics that suggest GDP may even scale, to some degree, alongside the measure of trust in a society. Cultures that learn to interpret higher levels of trust can do more creative and complex transactions.
The vision of the Human Neuroimaging Lab is understanding social cognition. Studying trust is a baseline for that. It's essential to all interaction. There's no risk-free way to trust somebody, just like there's no risk-free investment.
