Greg Satel (29:57)
Remember the riot example from earlier? You remember how in one mix of thresholds, almost every everyone started throwing rocks, but in another mix of thresholds, it stopped at six people? Okay, so in the grand model of how cascades play out, those groups would each be considered a cluster, and the everyone is throwing rocks group would be considered a saturated cluster. The behavior Change in that group has saturated the entirety of every interaction in an entire nation like, say, Ireland. There are many clusters of human beings who share strong ties, people who know each other's stories, who help each other move, who exchange gifts on birthdays and holidays. Any large group of humans, like nations is made up of lots and lots of strongly connected clusters like this. And you may have heard of Granovetter's 1973 paper, the Strength of Weak Ties, one of the most widely cited and influential articles in all of social science. In it, he showed that weak ties, casual acquaintances, have an enormous impact on social networks, on the flow of information, on the flow of change and behavior. So a very large group of people like Ireland is made up of lots of clusters of people who know each other well but don't interact much with people outside those clusters. Weak ties form bridges between clusters of strong ties. Okay, we have everything in place now to make sense of this. If one person within a network of close friends and family and co workers and neighbors is an acquaintance of one other person in a different cluster of close friends and family and co workers and neighbors, that weak tie between those two people becomes the conduit by which information will flow from one closely connected network to the other. And here's the absolutely amazing and bonkers point we've been marching toward this entire time. When you have the right kind of weak ties between two clusters, if one cluster saturates, it can cause another cluster that doesn't have the right mix of thresholds and to saturate as well. Saturation is when everyone in the group follows suit, everyone conforms, or almost everyone. Right. That's a saturation moment within a cluster. So in this scenario, that weak tie adds a person from outside a cluster whose threshold, should it be met, this other person in another cluster, if their threshold is met, that will then influence the threshold of the person that. That they know in another cluster. Okay, let's use an example. The riot, once again. Okay, remember how the riot stopped at 6 people if there was no one whose number was 6? Like if you have a threshold of 7 and the person next to you has a threshold of 7, and there's a gap, there's a gap there. It's not getting crossed. You need that seventh person to join in. They don't join in, the cascade stops. Well, what if you got a text message right then from a person you met once at a conference who said they were getting into baking sourdough just like you were because they overheard you talking about how you just got into baking sourdough bread. And you exchanged numbers at the conference, but you never talked again. For some reason, today, that person is in another riot in another city, Started over the same grievance as your riot. And just now they share with you by text that they just threw rock. Well, there it is. You now have your threshold met. That's the seventh person you needed to throw a rock. It's just they're not in the group you're in, they're in another group. So now you throw a rock in your group, and the cascade crosses the gap and continues to build from there. And to make this even more amazing, imagine that over in this other riot, this acquaintance, this person, is a holdout. This is the kind of person that needs 80 other people to throw a rock to meet their threshold of conformity. But in that group, that's happened. And when they text you, even though their disposition, their threshold is very high, they serve the role that a person with a very low threshold would have served in your group. If you had that sort of person in your group, they become that person via the weak tie. So now your cluster will begin to saturate. And you can further imagine that someone in your group with a medium threshold, once your cluster has saturated, might text someone mid saturation in some other group, some third group, just to let them know about the riot that seems to be happening out of nowhere. And that one message will then meet the threshold of that acquaintance in this third cluster, Allowing their stalled out saturation to keep going as well. And all of these clusters will influence each other and their neighbors via weak ties. And a group like this, a group of groups, clusters of people with strong ties, some of whom are connected to other clusters of people via weak ties, form what they call a percolating local cluster. And this, as Sattell points out in his book, as many sociologists and psychologists and network scientists have demonstrated via both mathematical models and examining the real world, this is the dynamic, the natural phenomenon, the mass collective psychology at play in the sort of social change that will eventually spread rapidly across entire populations. Percolating local clusters, that's the key. Groups of people who are strongly connected can be weakly connected to other groups in a way that allows change to leap from cluster to cluster to cluster to cluster, as thresholds are met and those clusters saturate. And when you get clusters of clusters and clusters of superclusters, Everything scales into a massive system, wide s curve of status quo, shattering change or adoption or innovation, and then it levels off into a new normal until the Next cascade comes along, and mind you, this whole thing is constantly in flux. It's planet wide today, but not too long ago, these networks were bound by geography. But no matter their size, human social networks are constantly changing in dynamic, as events and information flows and affects thresholds themselves, and as people with different interests and goals and anxieties join and leave groups, making and losing friends, coming into and out of contact with ideas. And this is why most on purpose efforts to generate this kind of widespread change took place, because there were people in many different places continuously striking at the system until they made contact with one of these percolating local clusters. But the system is far too complex to predict most of the time. So that percolating cluster, it moves around, which means you have to keep striking, and you might strike your entire lifetime if so. Many of these movements, they pass down the hammer, a continuous, relentless striking at the status quo. That's the final piece of all this, at least in our examples. Cascades, whether they're naturally forming or they've been set in motion by directed, on purpose human action, they begin all the time in social networks. They get going for a while, and then, usually without much impact, they fizzle out. And if you'll forgive all the mixed metaphors here, let's move from hammers striking to forest fires. That's how sociologist Duncan Watts puts it. He says they're like forest fires caused by lightning strikes and tossed cigarettes, inciting moments that could be very big or very small. But these create continuous fires year round, and most of the time, those fires quickly die out. But every once in a while, when the conditions are just right, no matter the size of the initial spark, if it ignites a flame in just the right place at the right time, the resulting fire will persist and grow and spread for millions of acres. And that's due to the state of the network where that fire began. Now, when the cascade is part of a human interaction passing from human brain to brain, there's all sorts of factors that affect the likelihood of a movement persisting and growing and spreading. Like media representation, activism, events big and small. All of these things can either lower or raise people's existing thresholds, either all across a population or just in isolated clusters that may or may not have weak ties across them. So there are millions of variables that can alter the conditions of a network. But when those conditions are just right, one spark, one rock in just the right place at the right time, among the right mix of thresholds, can lead to a global cascade of change.