Saturday, July 04, 2020

The Evolution of the Natural Desire for Friendship in Social Networks

                       A Short TED Talk by Nicholas Christakis on Friendship in Social Networks

         A Longer Version of the Lecture That Puts Social Networks Within Sociological Theory


My list of 20 natural human desires includes friendship.  As part of our evolved human nature, we generally desire friendship.  We seek social relationships of mutual affection and respect based on shared interests and cooperative endeavors with our friends.  Although we can be friends with our sexual partners and family members, our desire for friendship is a natural desire for a social bond of reciprocal attachment that differs from our natural desires for sexual matting or familial bonding.  We can have intense and enduring friendships with only a few people, because such friendships require shared experiences over a long period of life.

If the good is the desirable, then the 20 natural desires constitute a natural standard for judging the good society, which would include promoting the conditions for friendship.

I have written about the Darwinian evolution of the natural desire for friendship in previous posts.  I have argued that Aristotle's account of friendship (philia) is rooted in his biological science of social bonding among humans and other animals (herehere, and here).  For Aristotle, the friendship among philosophers is the highest form of friendship, and indeed the perfection of the moral and intellectual virtues.  This life of philosophic friendship could be lived in Athens, I have argued (here), because Athens was a relatively liberal commercial society that allowed philosophers to form voluntary associations devoted to the philosophic life.  Similarly, Adam Smith and David Hume could live a life of philosophic friendship because, I have argued (here and here), the modern commercial society of Scotland was liberal enough to allow such a life.

Critics of liberalism, like Patrick Deneen, have claimed that liberalism's radical individualism dissolves all the social connections of family life and friendship, and thus makes us desperately lonely.  But I have argued (here) that the modern liberal social order actually gives us the freedom to form social networks that satisfy our natural desires for social bonding and friendship.  Moreover, I have suggested (here and here) that Friedrich Hayek showed how liberal open societies allow us to live in two worlds of evolved social instincts--the small world of face-to-face personal relationships in families and groups of friends and the large world of impersonal market exchange.

Now, the work of Nicholas Christakis and his colleagues in studying the evolution of social networks of friends among humans and other animals can explain the biological science of all this.  Much of this research is surveyed in Christakis's recent book Blueprint: The Evolutionary Origins of a Good Society (Little, Brown Spark, 2019).  Christakis is a physician and sociologist, who teaches at Yale University, where he runs a "Human Nature Lab."  Unlike most sociologists, he does not scorn biological explanations of human social behavior.  On the contrary, he contends that there is a biological blueprint for human nature that shapes human social life.

At the core of all societies, Christakis claims, is what he calls a social suite of natural desires and capacities:

(1)  The capacity to have and recognize individual identity
(2)  Love for partners and offspring
(3)  Friendship
(4)  Social networks
(5)  Cooperation
(6)  Preference for one's own group (that is, "in-group bias")
(7)  Mild hierarchy (that is, relative egalitarianism)
(8)  Social learning and teaching

This social suite overlaps with my list of 20 natural desires and with much of Stephen Sanderson's list of 14 natural human preferences, except that Sanderson does not include friendship on his list.  I have written about Sanderson's list in a series of posts that begins here.  Like Christakis, Sanderson is one of those rare sociologists who embraces the biological science of human nature.

What is distinctive about Christakis's research is his adoption of "social network analysis."  (Wikipedia has a good article on this.)


In a social network diagram, each person in a population (a family, a club, an organization, a village, a school, or even a whole country) is indicated by a circle or a node in the diagram, and every connection between any two people (two friends, two relatives, two co-workers) is indicated by a line or an edge.  Connections among people are determined by asking people questions that are called a name generator.  Who are your closest friends?  Who do you prefer to spend time with?  Who can you trust to discuss your personal problems with?  To whom would you give some valuable gift?

Or researchers could map social connections by observing individuals and recording who is near whom for how long.  One could also use email or online social-network data to identify social bonds.

This is not limited to humans.  One can do a social network analysis of non-human animals to determine their patterns of friendship based on an association index of two animals, which is based on the amount of time they spend together.  Studying chimpanzees and bonobos, for example, one can see that they have social networks of friendship similar to that of human beings (Christakis 2019, 208-216).  One can also see that these primates are status seeking animals--like human beings--in that the more popular individuals are friends with one another, while the less popular individuals are also friends with one another; and the more popular ones are at the center of the network.

The popular individuals in positions of leadership in primate groups act as a kind of social police in that they mediate disputes to keep the peace in the group.  If high-ranking individuals are taken out of the group by human researchers, conflict and aggression rise, and the group collapses into chaos.  This shows how natural selection favors a desire for what Christakis calls "mild hierarchy," in which group leaders at the center of social networks lessen conflict and promote peaceful connections between high-ranking and low-ranking individuals.  This confirms Aristotle's observations of friendship and leadership among social animals.

Christakis and his colleagues have shown that there are similar networks of friendship in human groups, including bands of hunter-gatherers.  Studying the Hadza, a population of hunter-gatherers in Tanzania, they found that the networks of friendship among these foragers resembled the networks of people in modern developed societies in five or more traits (Coren Apicella, Frank Marlowe, James Fowler, and Nicholas Christakis, "Social Networks and Cooperation in Hunter-Gatherers," Nature 481 [26 January 2012]: 497-501.)

First, the likelihood of a social tie decreased with increased geographic distance and increased with increases in genetic relatedness.

Second, the networks showed reciprocity, in that if one person named another person as a friend, that second person was likely to name the first person as a friend.

Third, the networks showed transitivity, in that two of a person's friends were likely in turn to be friends with one another.

Fourth, there was degree assortativity, in that popular people tended to befriend other popular people.

Fifth, there was homophily, in that similar people tended to befriend one another, so that friends are like to resemble one another in many of their physical and social traits.

Here is a YouTube video with Coren Apicella explaining this research.

If the Hadza are a good proxy for our prehistoric hunter-gatherer ancestors, then this suggests that the desire for networks of friends is a natural instinct shaped by natural selection in our environment of evolutionary adaptation.  Of course, critics of this research will say that the Hadza have long had contacts with the modern world--including modern Western scientists!--and therefore they cannot give us any clear window into our prehistoric past.

And if hunter-gatherers really do show social networks of friendship similar to those in modern liberal societies, that suggests that critics of liberalism like Deneen are wrong in claiming that liberal social orders cannot satisfy the natural human need for social connection.  This might also suggest that Hayek was wrong in claiming that a modern open society must repress the ancient primitive instincts for social solidarity to which socialism appeals.  (I have written about this here.)

No comments: