It might seem like a joke at first, that researchers could gain insight into international relations from studying MySpace, the once-popular online choice for social networking among teenagers. Yet researchers have done just that, and using the networking site’s information on a mass scale, they arrived at a number of conclusions, including one that suggests alliances between nations have similarities to the way we form friendships. Heading the research was psychology professor Peter DeScioli, whose work analyzed more than 11 million data points in MySpace friends’ rankings.
"If you think about friendship in terms of alliances, in the context of game theory and international relations, one of the main things about allies is that they are fundamentally jealous of each other," DeScioli says. "If Saudi Arabia is allies with the United States, it’s not just concerned about its relationship with the United States. It’s also concerned about the relationship that the United States has with other nations such as Iran."
Previously, DeScioli completed a Ph.D. from the University of Pennsylvania in psychology in 2008 and a postdoctoral fellowship in experimental economics at Chapman University before landing at Brandeis University. He is a past recipient of a research grant from the Wharton Risk Management and Decision Processes Center. Currently, he’s completing a post-doctoral fellowship in psychology and economics at Brandeis University.
An edited transcript of the conversation follows.
Arabic Knowledge at Wharton: What is the "alliance hypothesis" of friendship?
Peter DeScioli: We started this research because we don’t really have very good theories of friendship. The most common theory is that friendship is a trade relationship. In other words, I only do things for my friends if they do things for me. One problem with this theory is that it predicts that we should be good accountants in our friendships and we’re not. If someone happens to need a lot from their friend and at the same time they can’t repay the favor, that’s not a problem in friendships. We’re not very careful accountants. We’re not keeping track of what we do for close friends but we do keep track for acquaintances.
So we asked, ‘What sort of function does friendship serve?’ We started thinking about conflict resolution. In non-human animals, disputes are usually between two individuals at a time. But humans recruit allies. They attempt to recruit other humans to take their side in disputes. This is a very unusual strategic ability in the animal world. Once you have recruitment of allies in disputes, then you need to have allies you can count on. When forming alliances, you’re very concerned with the other alliances your friends have. If your friend has stronger alliances, or a stronger relationships with other people, then they will tend to side with those individuals against you if conflicts arise. As a result, the strength of a person’s relationships with others influences how you view your friendship with them. We’ve tested this theory in lab studies and the MySpace network.
Arabic Knowledge at Wharton: Can you tell me how this translates to country alliances?
DeScioli: Usually we do it the other way around: We use alliances between nations to help understand friendship. The alliance argument is easy to see on the international scale. Between nations, it is not uncommon to have jealousy in alliances. Nations are very concerned with their allies’ alliances with other nations. If my ally forms a treaty with my enemy, then I can’t count on them in disputes with that enemy. One of the foundations of my research is international relations.
Arabic Knowledge at Wharton: So if a country doesn’t show appropriate respect to a close ally by ranking them very high, the other country could act jealously and rank them lower?
DeScioli: They could rank them lower. They could threaten to rank them lower if they don’t receive enough respect. They could threaten to seek another alliance with someone else. Those are all things we see people doing interpersonally and of course, nations are going to do the same thing.
Arabic Knowledge at Wharton: That brings an example to mind. When the Chinese President Hu Jintao recently visited President Obama and the U.S., he received a five-gun salute and a state banquet dinner as his welcome. The last time President Hu visited the U.S., it was during George W. Bush’s presidency, and he was welcomed with a luncheon, which was seen as a bit of a snub. Does this example fit into to the "alliance" hypotheses?
DeScioli: Yes, countries want to see their power being acknowledged. Another example where jealousy is really clear is China’s attitude toward the Dalai Lama. As the Dalai Lama goes around trying to form alliances with different nations — obviously that is against China’s interests — China has made a series of threats to nations for meeting with the Dalai Lama. They have issued stern warnings threatening that ‘bilateral ties between our nations will be damaged if you meet with the Dalai Lama.’ They’ve done this to the U.S., Germany, and France. That’s a particular example of jealousy on the international stage.
In grade school, kids say, ‘I won’t be your friend anymore if you do such and such.’ But when you get older, you say, ‘This will damage our bilateral ties.’
Arabic Knowledge at Wharton: Does this sort of respect for friendships and alliances vary culturally from the Middle East to the West to Asia?
DeScioli: Yes, I would expect alliance behavior to vary cross-culturally. One of the main sources of variation is how important are the disputes you’re getting into and what are your options for resolving disputes. In the modern world, we have police forces. So if I get in an argument with someone, I don’t absolutely depend on my network of friends for my life and to be able to hold onto my property. For much of human history, this wasn’t the case. You had to depend on your alliances for your security. But there is still variation in policing institutions. Some police forces are very corrupt. You can call them all you want and they’re not going to do anything. And even when police protect your safety, for other types of disputes, like banks foreclosing on people’s homes, state enforcement is not perfect. Basically, I would expect that alliance concerns would be heightened whenever people depend more on their allies in disputes. You’ll see more or less alliance-building and jealousy depending on what are the disputes and what are the stakes.
Arabic Knowledge at Wharton: So if you’re living in a more corrupt country, the stakes are getting heightened. Are people relying on their network of friends rather than the police force or a security force?
DeScioli: Yes, that’s basically the prediction. Alliances would be more important than abiding by the laws, since the laws are not being enforced reliably and across the board. Who you know becomes a lot more important. Alliance building becomes much more important.
Arabic Knowledge at Wharton: With the unrest in the Arab world, protestors connected via Facebook and Twitter. These are alliances that are made between acquaintances and strangers. Even more remarkable are alliances made through Facebook fan pages. The alliances created were powerful enough to influence, and some cases, topple governments.
DeScioli: It makes sense that the Internet can facilitate these alliance formations, especially when you get to alliances with large numbers of people. In that case, one of the main problems for staging a revolution is that you have to coordinate your decisions. You can’t have a few people revolt on one day, a few others revolt the next day, and so on. You need to revolt all at the same time and social networks can be really useful for solving this problem.
Arabic Knowledge at Wharton: What about the case of groups of people who might sometimes be your allies, and sometimes they’re not?
DeScioli: First, whether someone decides to align with you depends on whom you’re disputing with. Someone you’re really good friends with on Day 1 could be your enemy on Day 2, if you get in a dispute with someone else. That’s a pretty straightforward scenario you see in international relations. All of a sudden, there are several countries taking sides with the U.S. against Libya, but some of those nations have recently been on the opposing side against them. That makes the alliance world very complicated because you might help someone on Day 1, but on Day 2 they’re your enemy.
But then, there’s another ambiguity worse than that. If there’s a dispute between two individuals, you might want to hedge your bets and make both individuals feel like you’re taking their side. That might be the best-case scenario for any one person. They want to make everyone feel like their best friend because those individuals will have an interest in supporting them.
Strategically, I want to go around telling everyone, ‘You’re my best friend,’ or ‘You’re my best friend. I’ll always side with you.’ But at some point, I’ll have to show my cards and actually side with someone. At that point, someone’s going to find out and decide, ‘No, you’re not really my best friend.’ But any time I can keep those true feelings ambiguous, I should try to keep them ambiguous. And I can try to court both sides. That way, whoever wins, I’ll end up being in the favor of the winner.
We have seen these tactics in the lab. We ask people to tell us their 10 closest friends and distribute 100 friend points to these friends. And we ask them to do it again, but this time, imagine all your friends will know how you distributed those points. And what they do is make the distribution more equal if they know the points will be public. They will make it less clear to everyone whose side they’ll take in a dispute. This is known as a ‘straddle strategy’ in international relations — you straddle both sides. And there seems to be evidence that people are using this tactic in close friendships as well.
Arabic Knowledg@Wharton: That strategy is useful in the business environment as well. People are working in corporations and they want to climb the corporate ladder. They would want a lot of people to be their allies and bring them up in promotions.
DeScioli: Yes, I think that’s right. If you were trying to cultivate business relationships within or between corporations, the same dynamics would apply. You would often want to conceal your true ranking of others, who you feel stronger towards, so everyone can maintain a positive illusion that you are a good friend of theirs.
One issue this also brings up is that the business world often combines different types of relationships, especially trade relationships with alliance relationships. And that can raise tensions. If one person views it as a trade relationship and another person views it as an alliance relationship, then they could be offended by each other. In a trade relationship, if you go to the market, and you trade with somebody, you would expect them to take the best deal they can find. If someone else offers them a higher price and they take it, you don’t take offense because that’s exactly what you expect them to do. That’s why they’re there. But in the alliance world, if someone makes a proposition to your best friend, and you say, ‘OK, I’ll be your best friend, see you later,’ to you, you do find that offensive. That’s a betrayal and you would be upset about it.
Say there’s a partnership between two companies, and if one of them finds a better price somewhere else, one company might treat it as an issue of loyalty and see it as offensive. That’s one area where there might be consequences.
Arabic Knowledge at Wharton: What led you to MySpace to conduct the research? You mined through 11 million data points? How did you do that?
DeScioli: Not by hand. We started collecting data in 2007. We worked with two computer scientists. They built programs to collect data from MySpace. When we went to analyze it, it was challenging because you can’t fit 11 million lines of data into standard programs like Excel. For all of our analyses, the computer scientists had to design algorithms for doing even simple things, like looking at averages, because there’s no way to get that much data into standard statistical packages.
Arabic Knowledge at Wharton: Are you continuing to do more research with that MySpace data?
DeScioli: We might do something else with that data set but we’re not working on that right now. One next step is to use simulations to look at alliance dynamics. In other words, to see which are better strategies and which are worse strategies. We’re using agent-based simulations. That just means we’re designing little computer people and giving them strategies to see what strategies are better.
The other issue I’m eager to address is that we haven’t been able to manipulate in a lab environment people’s perceptions about how their friends rank them. That obviously is really hard because why would they believe me, the experimenter, when they have longstanding relationships with their friends? In order to look at this issue, I want to try to create new friendships in the lab. They won’t have a long history so people won’t feel very confident about the stability of their relationship. Then I want to change people’s perceptions of where they think other people rank them. Basically, in a lab environment, I’ll create a world where people will have disputes and recruit their allies to take their sides in disputes. And then, have friend rankings to see whose sides they’ll take. I’ll be able to test in a more controlled way the predictions made by the alliance hypothesis.
Arabic Knowledge at Wharton: How are countries similar to MySpace friends?
DeScioli: I don’t usually think of it that way. When you see these strategies among nations, we just view them as good, reasonably smart strategies. When these dynamics happen in friendships, we often don’t see them happening because they’re working through our emotions, or if we do recognize these tactics, we often regard them as childish. So jealousy often looks crazy and childish and totally baseless. The way our friendship psychology works is to make us feel closer to people who rank fewer friends above us, because that will lead us to having stronger alliances. For the most part, this is all occurring unconsciously, except for a few exceptions in very strategic individuals. This is unlike nations that consciously pursue their strategies. So when we see interpersonal strategies, we don’t understand them because we don’t have an explicit representation in our heads that we are pursuing a strategic alliance the way that nations explicitly outline it. That’s why international relations can help us understand interpersonal relations.