Slow ideas, complex ideas

Slow Ideas, Dr Atul Gawande in the New Yorker

Some ideas spread fast. How do you speed the ones that don’t? asks the New Yorker.

When anaesthesia was introduced in 1849, it soon swept across the world and was taken up as a worldwide innovation within seven years. When in 1867 Joseph Lister published a new method—the use of antiseptics—to eliminate germs and reduce sepsis and death, it took decades to catch on, and even today simple but effective actions like handwashing can still be perfunctory.

This is the story in a New Yorker article this week on innovation. Atul Gawande, a doctor working for less risky childbirth in Bangladesh, reflects on innovation and how you can speed it up, and concludes that one major factor is the effort involved by the user:

 “We yearn for frictionless, technological solutions.”

Ain’t that the truth. With anaesthesia the result was immediate and actually made life easier for the doctor; with antiseptics the doctor  had to follow lots of procedures, including spraying himself with carbolic acid in the early years, and the results would follow later.

His*  solution is conversations, talking to people—midwives and doctors—one by one to help them adopt better practices. It seems that once they actually see results, the practices become a regular part of what they do.

 “Diffusion is a social process through which people talking to people spread an innovation”

said Everett Rogers (quoted in this article), innovation scholar.

So far so good. The problem is when systems are complex. In the case in this article, we still have a linear relationship, it is simply a linear relationship with a timelag. You have to wait a few months but you still will see that X (your actions) led to Y (more surviving babies). This creates a nice positive feedback loop—more good practices, more surviving babies.

I am certainly not knocking this, as the innovation of going out to change national practice through one-to-one conversations is impressive. But what about in more complex systems? What about where, for example, the environment is concerned, where X (your actions) might lead to Z not Y, which eventually leads to A, B, C, finally arriving at Y (what you wanted) but not here and now, somewhere else in place and time? Not even here where you can see it. Just to make the plot a little thicker, why not add some contingencies? Maybe it will only lead to the A,B, C chain if another event, J, happens.

Do you see what I am saying? Not only is this tricky to research, it is tricky to be motivated, it is tricky to evaluate and it is a pig to sell to donors, who need to show their taxpayers, “We did this, it cost that, and the result is 1 million survived babies.”

Gregory Bateson, the great systems thinker observed :

 “Surely the mountain lion when he kills the deer is not acting to protect the grass from overgrazing.”

What we normally think when we see a situation of overgrazing is to reduce the number of deer. What happens when nature is in balance, is that the deer get reduced anyway as the landscape has co-evolved in a way that works. The deer, the lion, the grass and countless other variables keep each other in check, but that is not their purpose. Their purpose is another: the lion wants to be fed, not to take care of the grass. Preventing overgrazing is part of the inter-relationships of each one’s individual intentions.

The other example Bateson gives, very pertinent in our modern society of obesity and malnutrition, is that of when we eat. We do not wait till we are hungry (Hunger leads to Food in a direct relationship, like X leads to Y), but our eating is governed by social conventions of how we eat, and what, and where and with whom. We eat because it is dinner time, we eat meat and two veg (or rice and vegetables, or pasta, fish and vegetables, or noodles) because that is our tradition. The loss of these supporting conventions on our eating habits—where increasingly people graze all day, or eat purely to input X  to satisfy Y—is one factor behind our poor nutrition today.

So, yes, frictionless technological solutions are not going to make the social innovations we need. The importance of conversations cannot be underestimated—going out making change one person at a time conversation by conversation—is a great idea. But still more is needed when situations are complex, nonlinear and uncertain.

*When originally posted I said “Her solution…” mistakenly thinking that Atul was a woman’s name. Apologies.

Tags: agricultural research, complexity, environment