Stepping outside the modern habitus

Sociologists use the word habitus to describe ideas that are so ingrained in a society they influence discourse and agency in the people that grow up in it. Until relatively recently in historical terms the idea of the gentlemanly elite - aristocrats - was part of this habitus. For a very long period of human history most believed this class of people were born to lead (and that this was unquestionable). Today this idea has almost been written out of society and replaced with the idea of the virtuous worker and the bourgeois fat cat. The best example of this is William Wallace (Mel Gibson in Braveheart) and King Edward I. In the film William Wallace is a scrappy farmer, but from what we know of the real history he was actually a member of the Scottish nobility himself. It seems strange today but for the era it was such a commonly held belief that the nobility were born to lead, likely it would have been difficult for Wallace to have seeded a rebellion at all had he not been “high born”. There are many historical stories of people preferring to be led by the nobility believing they had some kind of inbuilt gift for leadership. In the feudal system God placed those people in command because God decided which souls would become heirs, and for a long time that was that.

Just as those people were influenced by how things were then, we are influenced by how things are today. One of the most pervasive ideas of today is how we think about managers and managing. Modern managers are orchestrators of people, responsible for efficiency. Usually middle managers are given specific objectives by their managers and so on, until the person at the top is indirectly in control of the actions of all below them.

It seems like it was not always like this however. Time was not at a premium in the medieval and ancient world because more often than not, people didn’t actually know exactly what time it was. What’s more, people didn’t need to know exactly what time it was. In the pre-modern era, it didn’t matter how fast you milked all the cows because of course you stop when you’re done. Once you’ve ploughed all the fields, you’ve ploughed all the fields, that’s all there is. The industrial revolution changed that mindset. When introducing machinery suddenly we needed the introduction of time across the population because the machine workers needed to all start together. This idea of orchestration of people represents the first part of modern management. The second part of modern management is maximising efficiency - the art of combining people and machines in a way to maximise profit. This concept was no less profound than the rise of a shared, absolute clock because efficiency is the basis for the idea of continuous improvement, year over year profit and a slightly perfidious justification for compound interest (which was deemed horrible usury for most of pre-industrial Europe). From these two axiomatic pivots it seems the role of leaders begins to change trajectory profoundly too, away from hereditary controllers of outcome (for example, “half of the grain belongs to me”) towards orchestrators and programmers of people (“I’ll pay you to plough my field exactly as I describe”).

Henry Ford and Frederick Taylor were early champions (and benefactors) of systemizing and relentlessly seeking efficiency gains. Ford of course popularised the assembly line but before then Taylor introduced the idea of standardising the actions people take through what were essentially time and motion studies - investigations into exactly how people interacted with machines (which was a base to edit those interactions). This is a clever way to make processes increasingly repeatable - necessary for a production line to exist and they represented a revolution in leadership that Taylor called scientific management. If you could program people’s actions entirely you could test empirically which ways of working were faster, starting the world’s current fervour for efficiency. It’s hard to understate how influential these ideas became: they are the foundation of globalism, consumerism and the workplace as we know it. What’s more they work really well (albeit distastefully) in the situations they were created for.

It seems however that the world is changing and the ideas rooted in scientific management have been so historically useful we have started to incorrectly apply them to the dissimilar problems we face today.

Scientific management solves an optimization problem - what’s the cheapest/fastest way to make a thing. Factories solve this problem. Now we have algorithms and computers, most of these optimization problems are solvable with computers instead of people. These optimization problems are of the “complicated” type, like owning a car factory. There may be new models but the goal to optimise - the process of making a car - never changes much. There are plenty of problems in which the aim is not that concrete however. For example, a tech startup is trying to find product market fit and in doing so they produce totally different features every day. It doesn’t make sense to release the same feature many times like you would build many of the same car model. There’s not a production process there to optimise. This type of problem, where we see non-repeating, non predictable tasks is a hallmark of a complex system.

Humans have built organisations operating in complex systems for considerably longer than we have been optimising complicated ones like factories. It could be the answer of how to organise industry in the future is embedded in how we approached complexity with large groups of people in the past, for example how we organised soldiers, sailors, ships and alike.

What’s more we’re used to avoiding thinking traps in this “military schema”. For example, it would be strange to put a number on how efficient an individual soldier is. We might have a set of synchronisation plays (standard operating procedures) but they can change very, very quickly. We know what soldiers do influences what the enemy does, and the enemy influences what they do (the recursive loop common in complex systems). We have plenty of prior examples that common scientific management tropes have led to disaster in a military context. In late world war 2 German tanks were so valuable that people were reluctant to use them - something which was exploited greatly by the allies by building cheap and barely sufficient tanks like the Sherman and the T34. The making of a tank is a complicated task ripe for optimization but deliberately trying to out-produce the enemy with lower quality tanks, knowing that your enemy is reluctant to “waste” their materiel on them is a play for a complex one. In this “military sphere” it’s obvious to us that this was the right play yet in other schemas, complex tasks like the long term future of NASA we insist on scientific management style predicates like Faster, Better and Cheaper.

Organisations that thrive in complex systems like war do not favour efficiency over all. In the total wars of the 20th century industrial output was essential to attrition but this is successful because it maximises the amount of battles you can engage in and how much pressure you can apply to a specific area. What organisations in complex environments chase is decision making throughput - the ability to correct behaviour to match your environment as fast as possible. This is what seems to have broken down in WW1.

NATO currently practises a strategic doctrine known as manoeuvre warfare. It gears the entire organisation towards decision making throughput (known as tempo). NATO sets out to achieve this through a decentralised leadership structure. This is not hierarchy-less - on the contrary - it’s that each layer has autonomy (known as Command) at a specific scale, and they do not need permission from their bosses to perform most actions. Bosses also do not typically plan actions for their subordinates, they share their own plan and have subordinates fill in their own implementation. This by action is not dissimilar to feudalism, albeit feudalism had a different goal. In the feudal period(s) delegating command was essential because communication was slow (as was the environment). NATO is interested in a similar system because the environment moves very quickly, even if communications are also fairly fast.

Scientific management greatly slows decision making throughput because it forces process changes to be made not by those doing the job (with the most information) but their bosses through a rulebook. This is lossy because it introduces politics, missing information and compromises. Planned economies like the USSR were scientifically managed and it seems obvious today how the bureaucracy there created a market that was slow to react. This is quite the opposite to manoeuvre warfare’s concept of tempo.

It shouldn’t be hard to draw a connection here to agile/XP in tech. Both are organisational systems designed to improve decision making throughput. Their purpose is to make sure we’re consistently performing relevant activities rather than efficiently performing the wrong activities.

It seems the problem is that our habitus has such a high gravity even after creating systems like agile we gradually are pulled back to scientific management. For example standups trend towards giving the boss a status update so that they can tune the day’s actions. With slack requiring permission for things is very easy and plans coming entirely top down is common. This is quite different to Fred Brooks’ “small team of surgeons” and continuously takes us closer to Taylor’s motion studies.

We love the Ford model T story because it made cars cheap but by the 1920s the Ford model T was an old fashioned joke that would have bankrupted Ford had it continued production. We all seem to spend a huge amount of time focussed on efficiently working, often knowing we’re working on the wrong things. The confusing question to me is, why? Why are we stuck chasing efficiency and clock watching in situations we know that to not be particularly useful?

Interesting associated reads: