Study of Social Interactions Starts With a Test of Trust

n a finding that could help explain why a sucker never gets an even break, scientists are reporting today that they have succeeded in visualizing feelings of trust developing in a specific region of the brain.

In the study, pairs of anonymous subjects were strapped into magnetic resonance imaging scanners 1,500 miles apart. The participants played 10 consecutive rounds of a risk-taking game that involved balancing monetary profit and trust. While they played, the scanners, synchronized through the Internet, measured how the subjects' brains reacted.

	Advertisement

With the development of trusting feelings, increased blood flow occurred in the caudate nucleus, an area in the rear part of the brain that is involved in processing rewards. Over time, this increased blood flow appeared earlier as an expectation of trustworthiness was established.

The study's authors, from Baylor College of Medicine in Houston and the California Institute of Technology, say their work shows that, at some level, the process of building trust is as basic as obtaining food or other rewards. The caudate nucleus appears to play a central role in evaluating the fairness of another person's actions and in signaling the intention to trust that person. Future studies, they said, may prove useful for understanding autism, schizophrenia or other behavioral disorders where the ability to form internal models of other people may be impaired.

By allowing neuroscientists to measure how two brains act and interact, the novel M.R.I. technique, called hyperscanning, also opens avenues of research in a relatively new field, real-time brain imaging of human social interactions.

"Researchers have been stuck looking at one brain at a time," said Dr. Steven R. Quartz, a neuroscientist at Caltech and an author of the study, which is being published today in the journal Science. "This really begins a new chapter in looking at the neural basis of human social interaction."

Dr. Joy Hirsch, a professor at Columbia and director of the Functional Magnetic Resonance Imaging Research Center at the university, called the paper a "tour de force."

"It's rich with innovation, both from the experimental paradigm point of view and from the view of extending brain imaging into very complex social domains," Dr. Hirsch said.

In the game used in the study, one participant, called the investor, is given $20 and instructed that he may hold on to it or give some or all of it to his anonymous partner, called the trustee. The amount given to the trustee is then tripled, and the trustee must decide how much to return to the investor. Players can either trust each other, by increasing the amount they turn over in each round, for example, or betray each other by reducing the amount.

The real-time brain scans showed that as the players proceeded through several rounds, an "intention to trust" signal, signified by activity in the caudate nucleus, developed in the trustee. Initially, this signal came after it was revealed how much money the investor would give. But with succeeding rounds of the game, the signal developed earlier, eventually appearing even before the investor's decision was revealed.

"The trustee is acting on what we think is their internal model of the investor," said Dr. P. Read Montague, a professor of neuroscience at Baylor and an author of the paper.

Dr. Richard J. Davidson, a professor of psychology and psychiatry at the University of Wisconsin, who was not involved in the work, said the finding of an "unfolding" of trust over time was something predicted from other studies.

"The circuits being engaged are areas we know play an important role in reward," Dr. Davidson said. The study, he added, "shows how these brain regions come to be recruited in establishing a phenomenon as complex as trust."

Researchers are using imaging techniques, including functional or fast M.R.I. scans, which detect brain activity through changes in oxygen flow, to observe the brains of gamblers or alcoholics, say, or of people when they are afraid or anxious. The new technique allows scientists to observe social interactions from both sides in real time, to see how each person's actions affect brain activity in the other.

Dr. Montague spent three years developing software to compensate for lag times and other delays through the Internet so the scanners could be synchronized. This allows them to be far apart, which is critical to the experimental design.

Trust is a complex phenomenon, one that many scientists would think incapable of being studied. In order to do so, the social interaction under study must be stripped to a bare minimum.

In this case, by having the scanners 1,500 miles apart, the players in the risk-taking game were completely anonymous. The participants never saw each other, instead seeing simple bar charts and numbers indicating how much money they were receiving.

"As social interactions go, this was about as impoverished as you can get," Dr. Montague said. "Because of that, we were able to make all sorts of findings."

Dr. Montague said that in building trust, the brain drew on existing mental models of the other person.

"The thing to remember is that we have to conjure a kind of virtual model of what is going on that is very similar to each other, or we won't be able to understand each other," he said. So rather than building a model from scratch, he said, as trust builds up "you're probably augmenting an extremely rich model you come equipped with."

Subscribe Today: Home Delivery of The Times from $2.90/wk.