John B. Gough – An Autobiography – 1845
Why Would I Even Ask?
When you think of social networks, you are probably thinking about Facebook, LinkedIn, Twitter, and so on. You go to the websites but what part of them is the network?
The social networks aren’t the software. They are the networks of connections that the software allows. People have been aware of social networks long before social networking websites. What were they talking about?
The short answer is, they are the networks of relationships we have with others. You may not even think about them but they are responsible for who you are. This blog entry gives you the background that will allow you to use them to your benefit.
The first book with “social network” in the title is shown in Figures 1&2.
Family and Social Network (Figures #1 and #2) was copyrighted and published in 1957. It is the first book to use “social network” in the title.
The book followed some families in England to understand the relationship between their marriages and their social networks, the networks of friends, family, and colleagues they had when they entered the marriage.
Today, we call strongly interconnected groups like that clusters. When two people interact with many people in common, their tie is a strong one. Because of the frequent interaction, they have common knowledge and norms. This kind of closeness can be comforting, but it can also constrain you. If you go against the group norms, “everyone” will know and, potentially, disapprove. This pulled individuals away from their marriage.
Bott found evidence for her hypothesis by interviewing a number of working-class people in England. She considered the clusters she found to be social networks.
“When many of the people a person knows interact with one another, that is when the person’s network is close-knit, the members of his network tend to reach consensus on norms and they exert consistent informal pressure on one another to conform to the norms, to keep in touch with one another, and, if need be, to help one another. If both husband and wife come to marriage with such close-knit networks, and if conditions are such that the previous pattern of relationships is continued, the marriage will be superimposed on these pre-existing relationships, and both spouses will continue to be drawn into activities with people outside their own elementary family (family of procreation). Each will get some emotional satisfaction from these external relationships and will likely demand correspondingly less of the spouse. Rigid segregation of roles will be possible because each spouse can get help from other people.”
One remarkable aspect of the study showed that the stronger the external bonds the more likely the couple was to be working-class. Today, we think that, probably, the relationship in the marriage was similar in other relationships making it difficult for them to make the connections necessary to find work and get information.
The cartoon in Figure is an ongoing strip called Andy Capp. It is the saga of a wastrel named Andy Capp whose wife supports him and has her network of friends, and he has his. This comic exaggerates the behavior for the sake of humor, but it gives a taste of some of the people she wrote about.
It is clear that people have been thinking about social networks for over 170 years, at least. Today, in psychology, sociology, anthropology, information systems, and other fields social networks is used is close to how Gough and Bott used it.
The Beginning of Social Network Analysis
Knowing there are social networks is one thing, but keeping track of the members and the interactions is another.
Once you decide that social networks are important, you will find that it is difficult to know who is interacting with whom. The number of connections and potential connections grows rapidly.
The first person (that we know of) that started to record and analyzed the relationships between people was Jacob Moreno. Moreno developed what he called “sociograms” which are drawings of the relationships between people. He developed the therapeutic technique called “psychodrama.” Psychodramas are based on the idea that an important part of our behavior is the result of our interactions with others. People would act out roles in a group situation to understand how their social roles influenced their thoughts and behavior.
Moreno was a complex and fascinating person who, as you can glean from the quote, had a high regard for himself. However, he is credited with being the first person that drew out these relationships and analyzed social networks.
Moreno realized that a person’s relationship with others in their immediate social environment had a lot to do with both how they thought of themselves and others thought of them. He used his theories to help make Sing Sing Prison, which was progressive at the time. He analyzed the relationships in a girls reform school where many of the girls were running away. He found that he could predict who would do it by the relationships of the girls.
Limits of Social Network Analysis
There are two main difficulties of Social Network Analysis of the type Moreno pioneered.
- It is hard to collect the data.
- Networks paths grow much faster than the members.
In Moreno’s study, there are 31 students. He had to interview each student to get the answer to the question, “who would you prefer to sit next to?” That could take quite a while. What happens if there are 100 students? 500? What is the limit? What if the questions become more dependent on memory, like “how many people do you interact with at your company?” If Fred does not say he interacts with Mary, but Mary says she interacts with John, the researcher needs to double check. It would take an impossible amount of time.
shows that the number of lines grows faster than the number of nodes.
You can see by looking at Figure that the top row is the sum of the bottom row. For those of you that like math equations, it is easy enough to get that the formula
(1 + n) / 2 * n
will allow you to calculate the number of there will be if every node is connected to every other one. Everyone knows each other. Since there are about 325 million people in the US, if everyone had a link, then, if everyone was connected, there would be about 58 quintillion connections. Even a million people, a small city, would have 500 trillion potential connections.
With everything connected, you would have a direct connection, but, obviously, it is not possible to maintain more than a minuscule portion of those contacts. A popular guess about the limit of the contacts we can maintain is 150, the Dunbar Number. If you have 150 direct connections that means that in a city of a million people there are 999850 people you do not have direct connections to. Since there are billions in the world there are billions of people you do not have a direct connection with.
However, even though we don’t have a direct connection to a person. We might know someone that knows her. She might know someone that knows someone and so on. With, potentially, sextillions of connections, how many people, on average does it take to get to anyone else?
The answer to this question and others about social networks came from a psychologist and a group of sociologists at Harvard in the 60’s.
The psychologist was Stanley Milgram.
Milgram is best known for his “obedience to authority” experiment.
In that experiment, Milgram set up a situation where one person was strapped to a chair with electrodes and another would be on the other side of a glass window. Milgram got psychology students to follow instructions and help with a learning experiment. The person in the chair was asked questions and if he got them wrong, the student was supposed to shock them with a device they were given. The device had a dial that would increase the voltage of the shock and, thus, the pain it caused.
Many of the students would keep turning up the voltage until the person in the chair was writhing and screaming with pain.
Fortunately, the person strapped in the chair was an actor and the shock device did nothing. The students didn’t know that. They thought they were really shocking the guy.
Milgram’s experiment has become famous, but he did another experiment that gave an answer to the question of how many people we need to go through to get to someone else. He called it a “small worlds study” because of the common comment we make when we are in a new place, we meet someone, and find we have an acquaintance in common.
Invariable, we exclaim, “it’s a small world.”
Milgram went to the most remote parts of the US (as far as he was concerned), Kansas and Missouri, and gave random people packets with the name of a stockbroker in Boston. They were asked to give the packet to someone they knew that might know him and if not, pass the packed on to someone who might (see Figure ).
How many people would it take? Some guessed thousands, though others thought it might be just a few. No one knew.
The result was that of the packets that made it the average distance was about five hops. In other words, we are six degrees of separation from each other. His experiment was criticized because the sample was so small, but subsequently, the experiment has been repeated using email and social networking sites and the results have been similar around 5 hops.
The idiom, “six degrees of separation” doesn’t come directly from this experiment, but rather a play by John Guere, Six Degrees of Separation, that premiered in 1990 (long before online “social networks”) and the subsequent movie of the same name in 1994. There is a scene in it that has the link, “every person is a door into another world” which I have taken as a personal motto.
Milgram’s study was a clue in how social space was organized. If connections were random, in other words, you are just as likely to be connected to anyone as anyone else, then we would not be so close. Just thinking about it tells you that that doesn’t make sense. You are far more likely to be connected to people in your environment that someone in another country, another profession, or another age group.
At nearly the same time as Milgram’s study, a group had formed at Harvard in the Department of Social Relations. The people in the group were drawn to the sociologist, Harrison White. They were finding ways to talk about social structure that could be measured.
White was a member of the faculty at the Department of Social Relations, an experiment in combining the social sciences into one program. Many participated in a groundbreaking class, Soc Rel 10 where Harrison White laid out his theories about networks. Most of the people in the class went on to make contributions to sociology.
One of White’s students was Mark Granovetter. He did a study of white-collar workers in the Northeast and found that people are more likely to get jobs from people they don’t know well than from those they do. Granovetter puts the reason succinctly, “your close friends know all the same things you do.” At least, as far as jobs are concerned. His paper, The Strength of Weak Ties, is foundational in the study of social networks. When you get connected to a different community you have the opportunity to learn new things.
Which is why every person is a door into another world.
The Sunbelt conference is the International Network of Social Network Analysis ( INSNA). “XXVIII” translates to the 38th occurrence, which means that it started in 1980. People have been experiencing, studying, and benefiting from social networks for a long time. Mark Granovetter, quips, “they started with Adam and Eve.”