Musing on Connective Tissue

Imagine 10 or 100 people describe you, and their descriptions become a word cloud. Now take your word cloud, and compare it to other people’s.

Would we find that the people you spend time with have similar word clouds — similar descriptors, similar nouns about them? Does your cloud sit close to the clouds of the people near you? If so, in what ways alike, and in what ways different? (That’s graphing friend groups or political links in n-dimensional space according to personal traits, as a word cluster of noun descriptors of the person: clustering by cloud similarity and then drawing lines between the points based on social or political ties.)

My bet: people befriend and connect based on proximity. But the friend network is different than the leader. The map of who we like is not the map of who leads us.

Maybe “Yalie” is a noun about you, and a bunch of the people around you are also Yalies. Maybe most of the people you spend time with are quite tall — about the same height. But what about groups with diversities of traits instead: An office might cluster around Yalies but diversify by role: one person described “HR,” one described “accounting,” one described “CEO.” Some friend groups are all CEOs; other circumstances produce diversity. How does that work in an office? In politics? 

I wonder what they’d learn about politics. Is the most successful politician of a group the person who overlaps most with the rest of the group? I wouldn’t think the leader is necessarily the person with the most words overlapping others (because that person might miss a key trait, e.g. the desire to lead). Maybe the person who is friends with everybody is the one with the most in common with everybody. Or maybe the leader’s lowest commonality is higher than anybody else’s lowest (i.e. “The leader is nobody’s stranger.”). It’s got to be more complicated than that, depending on the group structure and incentives and the outsized impact of specific traits like desire for power.

Why I’ve been thinking about this: in a group that I interacted with today, the leader shares the most traits with the other group members. Each bloc shares one of her dimensions. Thinking about the descriptors of these different people and how she ended up being president: she is the person all of these different groups can agree on.

But often the leader of a group is not the leader because they have much in common with everybody, but because they plan things. That’s an interesting difference. The leader of a friend group is not the one most overlapping with others but more likely the one who organizes the events. In her case, part of it is she makes the thing happen — the leadership thing. This differs from the second-in-command, who doesn’t organize people, just makes the wheels run. Doing isn’t organizing.

Does anybody map social groups this way? A thing I’d be curious about: how do the friend network and the leader differ? The mapping half is studied — sociologists call it homophily. The gap between who connects and who leads is the fresher question.

Games Played

The writers missed out by failing to name it, “The Devil Wears Prada 2: The Devil Still Wears Prada.”

Leave a comment