Different diseases spread differently. A simple observation, but with a level of complexity behind it that is worth understanding. It is not solely the infectiousness of a disease that affects its spread; the type of network the infection is dropped into plays a crucial role.
For many years, epidemiologists assumed a “mass action model” where the rate of infectiousness of a disease was a fixed number (basic reproduction number or R0), representing the expected number of secondary cases produced by a typical infected individual, early in an epidemic. For the flu, for example, it is roughly 2. Meaning one infection leads to two secondary infections, on average. Comparatively, for measles, it is 18.
In practice, the actual rate of spread is significantly impacted (and R0 effectively modified) by the host network: by both the density and the structure of that network.
At our recent network design symposium with the Santa Fe Institute (SFI), Lauren Ancel Meyers (Professor of Integrative Biology- University of Texas; Faculty- SFI) shared three simple network maps to illustrate how network structures influence the spread of epidemics. Although there are sophisticated mathematical descriptors of network structure, some simple explanations can easily demonstrate the range of impacts possible.
At the front of this article is a picture of three networks with the same number of nodes, but interconnected differently. The degree distribution (number of links each node has) varies across them, from only 2 on the left, to a range of 1 to 6 on the right. Two simple questions for you: 1) Which network will be the most susceptible to an epidemic? and 2) Which is likely to sustain the largest epidemic?
Are the answers obvious?
Networks with the greatest “degree distribution” (like the network to the right) are the most vulnerable to an epidemic. Variability implies vulnerability. But when an infection occurs, homogeneous networks (to the left) fare the worst.
Think of it this way: variability (or diversity) provides the greatest openness to the introduction of something new (whether virus or message), while homogeneity presents the greatest risk (or opportunity) for spread once established in the network. Beyond epidemiology, Meyers’ research has huge implications for the spread of ideas and information.
Take a Random Walk to Find “Super-Connectors”
Another network task in epidemiology is a desire to find so-called “super-connectors” because the likelihood of spread of a disease rises exponentially if “hubs” like these are infected in an outbreak (think of the highly connected nodes in the far right diagram above). It turns out super-connectors are not that hard to find. Randomly polling individuals in the network offers broad situational awareness and then asking to speak to one of their colleagues inexorably leads researchers towards super-connectors. Thus a random beginning can quickly lead to highly connected nodes. So, in any new network you enter, simple inquiries can uncover important understanding. And just as all rivers lead to the sea (or as Josh Baer joked, all tech in Austin leads to the Capital Factory), one can inevitably find “super-connectors.”
The infectious nature of an idea: Does it engage head and heart?
If you take Howard Gardner’s perspective, a cognitive leadership model recognizes that effective leaders “speak to narratives already present in their audiences’ minds.” For both organizational leaders and marketers alike, we not only need to understand the infectious nature of an idea and map the network, we also need to map the narratives already in existence in the network. In a future post, we’ll explore more on mapping narratives and the current state of a network’s mindset.
Thus, the capacity for an idea to spread depends on several factors: the content (infectiousness) of the idea itself (R0), the narratives already present in peoples’ minds, and the structure of the network in which the idea is released. This makes mapping networks and narratives key to orchestrating the spread of ideas, and yet, so few firms practice systematic network or narrative mapping. So, whether it’s epidemiology or the spreading of ideas, understanding the relevant network is crucial. Stopping the spread and accelerating the spread are simply two sides to the same network coin.
Stay tuned for more insights, and join us in conversation online using the hashtag #NetworksInAction