Resilience 💪 & Meat 🥩
How to build systems that optimise & systems that survive
Summary
This week, an exploration of two kinds of resilience engineering and ecological. Also, a review of one author’s approach to answering the question of whether we should eat meat.
Engineering Resilience and Ecological Resilience
Resilience, the ability to withstand shocks, is much lauded. However, it is often not clearly defined. We can separate it into two kinds: engineering resilience and ecological resilience [1].
Engineering resilience means a system rapidly returns to a stable point, but only with respect to certain shocks. This is the resilience of a combustion engine. A burst of cold air can enter the intake, but this has no meaningful impact on the engine's operation. If you make a hole in the engine with a hammer, however, you may not have so much luck.
Ecological resilience means a system can survive a large variety of larger shocks, even if it never returns to the exact previous state. This is the resilience of an Ontario forest. Every year, fires burn parts of the forest to the ground but these clear patches offer new avenues for growth. Yet, no two growths are ever fully alike.
Another way to put it: engineering resilience prioritises maintaining efficiency of a specific function, while ecological resilience prioritises maintaining existence of function, even if it means that function is less precise.
Some types of system display both kinds of resilience. For example, a company may develop engineering resilience via specialisation. Cherry, a manufacturer of keyboard key switches, is considered the best in class for high-end mechanical keyboards. Cherry's superior technology means they consistently maintain a buying market among mechanical keyboard makers. However, if the market were to fundamentally shift next year to replace keyboards with touchscreens, these capabilities would do the company no good.
Alternatively, a company could have ecological resilience through variety. GE creates MRI machines alongside aircraft engines and, up until recently, appliances. Global airline usage may collapse and harm engine sales, but the overall business can remain profitable due to its other ventures. However, when there was fierce competition in the appliances market, the company was forced to eventually sell the division, it could not compete in that specific domain.
Generally, though not absolutely, there is a tradeoff between these two kinds of resilience. As a system specialises, a large shock is more likely to prevent its operation. Alternatively, as a system gains in variety, less energy can be exerted in any one area to operate in an efficient manner. When seeking resilience we should know what matters more: achieving the highest possible quality and consistency, or increasing our chance of survival.
[1] This distinction was first made by a set of authors in the book Resilience and the Behaviour of Large Scale Systems as part of a new approach to managing regional scale ecologies.
📘 Book Review: Should We Eat Meat?
Should We Eat Meat? is the first book I’ve read by famed Manitoban researcher Vaclav Smil, much beloved by Bill Gates. The book’s intent is to provide the data needed to answer its stated question. It is a compendium of facts and figures and each page is loaded with references, yet it lacks the holistic view needed for a satisfying answer.
Smil’s method is reductive: nutrition becomes the composition of protein, fat & carbohydrate with some micronutrients; environmental impact becomes land use, greenhouse gas emissions, and some observations on heavy metals. This approach is dangerous. The systems in question are too complex for closed models to accurately describe their operation. If we focus only on the ways these systems can have engineering resilience around specific components we will miss their ecological resilience. That said, knowing this, and duly discounting certain recommendations, the book is useful for covering what we have been able to glean from this approach. As an overview of the scientific literature for meat production it appears comprehensive.
Thanks to Casey Li for reading a draft of this newsletter.
