The main advantage of Tradition is time. I think the areas where the article is most compelling is when it focuses on drawing knowledge from traditional practices that have been done for long period of time in a stable society (the bit about "clam gardens" in particular).<p>A traditional practice is a sequence of small, imprecise experiments extended throughout a large period of time. It's risk averse and only tweaks things slowly, but it has the benefit of probably not breaking everything when you do it, because if it was going to have catastrophic consequences it probably wouldn't have persisted for this long without anyone noticing. A 'traditional' system of medicine probably doesn't have the underlying principles exactly right, and it gets stuck in local optima, but it usually falls into the 'ineffective' category when it goes wrong rather than 'insane side effects.' Admittedly, tradition has issues when the underlying system changes rapidly and practices that made sense in the past no longer make sense to do, and there's a catch-up time that has to happen.<p>Science has advantages of being able to more rigorously and skeptically re-evaluate assumptions and to tease out underlying causes and principles. But at the same time it's also prone to human frailties in how it's conducted -- see the replication crisis. Taleb might call it a lack of 'skin in the game,' where researchers are institutionally motivated to publish whatever they can that gets them a p-value below 0.05 and at the end of the day they're probably not changing their own personal habits or practices based on what their research says (because the strength and weakness of science is that it finds ways to detach the researcher from the research). The 'danger of a single study' comes when initial findings become loudly reported and the general population (or rather institutional powers) who want to be Modern and Cutting-Edge and moving toward the Future and Progress will take the initial findings as a stamp of approval.<p>The main issue is when we use science not just for 'what are the facts?' but 'how should we live?' and apply our initial findings universally. Because society doesn't want to wait to make a change, and our scientific processes usually do not have the advantage of time that tradition does, we start to embody the long-term experiment into the culture. And what's worse is that we get pressure to adopt it more widely than may be prudent. Why would a government official only promote a new idea in a single isolated population when they can reap the benefits of the new science by pushing the idea onto the whole country? When it's right, it has great reward, but when it's wrong the costs are great. And so we got the 'low-fat' craze that has led to great costs and suffering, forcing these generational oscillations to try to get people back on track to something with more scientific support.<p>Scott Alexander's book review of "Seeing Like a State" [0] points to this same attitude of hubris when it came to the 'modern rational scientific' thinking of the High Modernists in architecture:<p>>First, there can be no compromise with the existing infrastructure. It was designed by superstitious people who didn’t have architecture degrees, or at the very least got their architecture degrees in the past and so were insufficiently Modern. The more completely it is bulldozed to make way for the Glorious Future, the better.<p>>Second, human needs can be abstracted and calculated. A human needs X amount of food. A human needs X amount of water. A human needs X amount of light, and prefers to travel at X speed, and wants to live within X miles of the workplace. These needs are easily calculable by experiment, and a good city is the one built to satisfy these needs and ignore any competing frivolities.<p>>Third, the solution is the solution. It is universal. The rational design for Moscow is the same as the rational design for Paris is the same as the rational design for Chandigarh, India. As a corollary, all of these cities ought to look exactly the same. It is maybe permissible to adjust for obstacles like mountains or lakes. But only if you are on too short a budget to follow the rationally correct solution of leveling the mountain and draining the lake to make your city truly optimal.<p>>Fourth, all of the relevant rules should be explicitly determined by technocrats, then followed to the letter by their subordinates. Following these rules is better than trying to use your intuition, in the same way that using the laws of physics to calculate the heat from burning something is better than just trying to guess, or following an evidence-based clinical algorithm is better than just prescribing whatever you feel like.<p>>Fifth, there is nothing whatsoever to be gained or learned from the people involved (eg the city’s future citizens). You are a rational modern scientist with an architecture degree who has already calculated out the precise value for all relevant urban parameters. They are yokels who probably cannot even spell the word architecture, let alone usefully contribute to it. They probably make all of their decisions based on superstition or tradition or something, and their input should be ignored For Their Own Good.<p>The result was ugly rectangles that no one wanted to live in, at the cost of destroying sections of cities that had grown organically over time to solve their local particular problems.<p>The experiments of reality have to be conducted no matter what, and sometimes those experiments cannot be sped up. Not everything about social effects can be revealed in a 12-week study, and sometimes 50 years or more are needed to reveal the negative effects of a policy. Are there ways to effectively contain the high-variance experiments of science to small populations while keeping most people on the low-variance experiments of tradition?<p>[0] <a href="http://slatestarcodex.com/2017/03/16/book-review-seeing-like-a-state/" rel="nofollow">http://slatestarcodex.com/2017/03/16/book-review-seeing-like...</a>