Epistemic status: Speculative. Almost certainly not an idea that is original to me, but it’s hard to google for this without being swamped in people who don’t actually understand what evolution is.
Evolution is not only real, it’s everywhere. The set of criteria required for an evolutionary selection process is so mild that it’s almost weird to find things where it isn’t a factor (though it may not be the dominant one).
In order to be subject to evolution, roughly all you need is the following:
- Heritability – the ability to reproduce and create new things that are in some way “based” on the original and share its traits.
- Mutation – a certain degree of random changes to traits as you reproduce.
- Selection – reproductive success must be influenced by inherited traits, usually due to some sort of interaction with the environment (e.g. “survival of the fittest”, but also things like traits that make you more able to attract mates)
This absolutely includes evolution in the classical, biological, sense. A while back Dawkins proposed memes (in the original sense that just means “ideas that are subject to selection pressure” rather than pictures of cats with words on them. There’s also a fairly significant evolutionary process in how language changes), but really it’s just about everything. Programming languages evolve, natural languages evolve, genres of literature evolve, etc.
Often there is some mix of evolution and design going on (evolution is not design!) but the evolution is still there, the design is just a major part of the inheritance and mutation processes.
Note that evolve doesn’t mean “improve”, it just means “adapt to selection pressures”, and often that adaptation again isn’t one that we’d particularly think of as a good thing. As Dawkins (again. Sorry. In my defence this was all from before his evolution transformation into his current profile as chief new atheist ranter) pointed out in The Selfish Gene, evolution of our genetics doesn’t select in any way for our well-being, it just does whatever makes the numbers go up (which is why you get things like genes that work great when you have one of them and kill you when you have two. Statistically it’s a great deal!)
One thing I’ve been noticing for a while is that company cultures are absolutely in the camp of things that are subject to evolutionary pressures. A company culture:
- Reproduces as people leave it and join or form other companies, or as people outside the company are exposed to ideas about how the company works.
- Mutates, as people change their opinions over time through their natural growth and change and through exposure to other companies.
- Is selected for, because company culture is a large part of determining company success.
I’ve previously thought of this as a mostly neutral thing with some interesting implications. e.g. it probably tends to mean that over time we evolve companies which are more or less rational economic actors even if they’re not made out of rational economic actors. Not that I necessarily think that’s a good thing, but it’s at least a useful thing for reasoning about company behaviour.
But I realised this morning a rather unfortunate problem: Although having highly profitable companies which treat their employees great and make economically rational decisions is one perfectly viable reproductive strategy in this model (where in that case reproduction tends to come more from people trying to emulate you than from people leaving, although there tends to be enough turnover in even great companies that it’s both), there is another extremely adaptive model.
That model is to basically just be incredibly dysfunctional and haemorrhage employees as a result.
This might not be good for the survival of the company, but it’s not the company that we’re considering here, it’s the population of company cultures. If the company goes under and all its employees go off and form five different companies that look exactly the same as their parent company, that’s evolutionarily speaking a great success.
The problem is that people tend to reproduce the cultures they’re used to. Even when they realise that those cultures are bad, they will typically only pick up on a few of the ways in which they were bad, and will tend to just carry the rest of them along without ever really noticing.
As a result, dysfunctional company cultures are fairly actively selected for, and will tend to thrive even as they destroy the companies they live in.
This is perhaps analogous to r/K selection theory – some animals reproduce by having a small number of offspring that they invest a lot of effort in to ensure they thrive (K-selection, roughly analogous to well functioning company cultures that reproduce slowly but tend to promote the survival of the companies that have them) and those that produce many offspring, most of whom will die but enough of whom will survive to adulthood to continue on to another generation (r-selection, roughly analogous to dysfunctional company cultures that destroy their companies but are passed on when the employees leave for other companies). The analogy isn’t perfect because the mechanism of “gene” transmission is so different, but it’s at least suggestive.
Which one is more adaptive varies from niche to niche, and there are plenty of niches, so it’s unlikely that either type of culture or organism is ever going to out-compete the other and replace it, but anecdotally it certainly seems like we have enough niches for the dysfunctional company cultures to thrive in right now.
So that’s the problem. What’s the solution?
Good question. Unfortunately I’m not sure of the answer. Culture is hard.
But I do think that whatever the solution is, it’s going to have to work with the evolutionary dynamic, not against it – evolution is going to happen regardless of whether we want it to or not, but we can try to shape it so that it selects for things we want.
So here are some ideas for how we can try to guide each of the three aspects of that evolution:
- To control how dysfunctional cultures reproduce, we can can reduce how much hiring we do along friendship and network lines – when a company fails and people leave, they tend to bring their friends at that company along with them. This is a perfectly natural thing to do, but tends to perpetuate the culture they are used to. (That being said, I don’t think it’s a good idea to actively avoid hiring people from dysfunctional companies – that will just send them to companies that will happily reproduce their culture. Also it punishes individuals unfairly for things that are not their fault. Just… maybe treat their opinions on how things should be done with a grain of salt.)
- To control how culture mutates, we can be more actively involved in designing our cultures for the traits we want – culture isn’t just a thing that happens, it’s a thing we create. If we let it emerge purely organically, we will tend to reproduce what has come before. If we design it, and in particular design it to avoid things we know are dysfunctional, the mutation will more actively push us in the directions we want.
- To control how dysfunction is selected for, we as individuals can more actively avoid working for dysfunctional companies – investigate the culture before we join, and leave sooner rather than later once we realise there are major problems we can’t fix. Where possible, if we know a company has actively bad culture we should warn people off, though whisper networks and doing this in public have problems that I don’t entirely know how to solve.
I don’t necessarily think those ideas are either necessary or sufficient, but I do think they will probably help. Evolution is hard to manage, and tends to do things we don’t expect, but a certain amount of selective breeding is possible, and I think we need to start thinking about how to do that for the cultures we create.
Of course, when we’re doing that, we need to be careful about going too far in the other direction.