On Chesterton’s fences and dogs that don’t bark
A few weeks ago we were looking into a bug and while trying to find the root cause, we went down the rabbit hole of some seemingly random configurations and restrictions set in the past. Coincidentally, a few days later I read about Chesterton’s fence.
As detailed in Farnam Street blog, Chesterton’s Fence is a heuristic inspired by a quote from G. K. Chesterton’s book, The Thing.
There exists in such a case a certain institution or law; let us say, for the sake of simplicity, a fence or gate erected across a road. The more modern type of reformer goes gaily up to it and says, “I don’t see the use of this; let us clear it away.” To which the more intelligent type of reformer will do well to answer: “If you don’t see the use of it, I certainly won’t let you clear it away. Go away and think. Then, when you can come back and tell me that you do see the use of it, I may allow you to destroy it.”
Chesterton advices not to remove a fence until you know why it was put up in the first place. Not knowing the use of a fence doesn’t mean the fence is not useful. This is a heuristic for understanding context and previous decisions. This is especially relevant in product management as you might have to deal with products and practices that were set up several years before you joined a company.
An extension of this concept is the idea of second order thinking: understand the deeper consequences and longer term effects of an action in addition to the immediate ones. Let’s take for example the decision for running a price promotion. In the short term, this will likely have a positive first order effect to the topline as we manage to sell more products. However, in a longer timeframe, it might cause a second order effect of training customers to expect promotions and erode the brand. If we don’t think of second order effects, we might easily approve the promo, while the overall impact could be negative.
An even more fascinating mental model is the equivalent of Chesterton’s fence but for non-events: the Baskerville effect, inspired by Arthur Conan Doyle’s novel, The Hound of the Baskervilles. [spoiler alert] Sherlock Holmes solves the case by noticing the “curious incident of the dog that didn’t bark”. The clue that led him to deduct critical information about the case was based not on something that happened, but on something that didn’t happen. Holmes expected the dog to bark given that a stranger approached. Since the dog didn’t bark, whoever approached the dog was not a stranger.
Similar to how we can use the Chesterton’s fence (why was this fence built) before we proceed with the action to bring it down, we can also consider non-events (why didn’t the dog bark).
Many apps don’t allow for specific transactions that seem to be obvious offerings. You can’t purchase an ebook through the Amazon mobile app or a premium subscription through the Spotify app. Before concluding that this is a bug or trying to get credit for proposing this idea, one should think why the teams didn't develop such features given they are very obvious. Of course, the answer lies in Apple’s in-app purchases policy, requiring a 30% cut on transactions happening through iOS apps.
One important consideration in both cases is the probability that some act is intentional. Is it reasonable to assume that any past act is well thought, therefore there should be a reason the fence was built in the first place? The answer is probably yes. While there is some randomness in the world, people don’t usually build fences by chance. This doesn't mean that the fence should be there. The reasoning for the fence might be dated or wrong but it helps to understand it before removing it. Now is it reasonable to assume that a non action is also intentional? In this case we need to set a higher bar for our hypothesis. It is a helpful framework to think through the probability of something being intentional but in many cases the reasons for a non action might be more simple, like lack of resources.
Next time you consider a decision, think through Chesterton fences and dogs that don’t bark.
I used DALL-E for this article’s images and the AI is impressive. Highly recommend to try it out.