Flying Squirrels: A Survival Guide for New Engineering Managers (Multi-Part Series, Part 9)

This is part 9 of a multi-part series. If you missed part 8, it's in the link above!

Going Native

To continue the discussion about Dunbar's Number and the ways in which it places constraints on our human platform, we need to talk a bit about its relationship to another very human construct: the tribe.

We believe, and generally accept, that Dunbar's Number effectively defined the maximum tribe size for early humans. As with many things that are the product of evolution, the causal relationship is difficult to assert, so the issue of whether maximum tribe size is a result of the physiological constraint that Dunbar's Number represents or whether Dunbar's Number was encoded in our biology as the result of human tribe size repeatedly finding success in the 150-person range is kind of unimportant. When you ask whether the chicken or the egg came first in an evolutionary context, the answer is usually "yes", accompanied by a smug look that says "stop asking stupid questions."

Humans are tribal, and the fact that we've evolved this thing called civilization hasn't really changed that. We've generally accepted the idea that we shouldn't just go around killing each other based on tribal identity alone, but our desire to form and belong to tribes is still an ingrained part of our humanity. You've probably even heard someone speak about "finding their tribe" as a way of expressing the desire to bond with a community of people that they can relate to, feel safe around, and feel understood within. Something that's interesting to me in that statement is that if you could go back thousands of years to the times before cities and nations, you probably wouldn't ever hear it. Imagine that conversation:

"I just need to get out there and find my tribe."

"Your tribe is here. Why would it be somewhere else? This is your tribe."

"I just don't feel understood here. I'm not the same as everyone else. No one gets me."

"No one is the same as everyone else. You have enough context for every member of this tribe that generalizing everyone and applying stereotypes is a pretty huge expression of selfishness since scale hasn't forced you to take cognitive shortcuts when it comes to people. But hey, I mean if that's how you feel, you can always leave."

"See that's a problem. I'll probably get killed, eaten, or made a slave pretty quickly on my own."

"That would be problematic. Also we do understand you. You're a little weird, but that's cool. We still help feed you, clothe you, provide shelter for you, and protect you. You're pretty good at hunting, so that's worth a lot and we're willing to put up with your weirdness even when it finds expression as insecurity and self-centeredness like right now."

"Awww, you're just saying that."

"No, seriously. There's only like a hundred and fifty of us. Everyone knows you. A lot of people understand you. You're unique just like everyone else. So it would be really cool if we could fast forward to the part of this exchange where you realize that people do love (or at least tolerate) you and then you stop pouting and commit to being a team player, because we're low on meat and those animals out there don't just hop onto our spears without a little coaxing. Capiche?"

The idea of needing to "find your tribe" is a relatively modern problem (when viewed in evolutionary timeframes), because once upon a time our tribes were largely hermetic. To illustrate this, let's visualize the tribe as a graph. In a simple, self-contained tribe, you could explore that graph from any vertex and find relationships to the same 150 people. Put another way, if you visualized each individual's tribe relationships and their intersectionality with every other tribe member's relationships using a Venn diagram, you'd have exactly one circle -- total overlap. Sure, connection strength would vary among members, but the vertices in the graph are all the same. Now try doing that today.

My loving partner of ten years is the strongest single connection in my Dunbar's graph. If you were to create a Venn diagram to illustrate our overlapping tribes, we probably have something like fifty people in common. Now think about your own tribal circles. Compare your circle to one of your co-workers. Even if you choose someone that you have a high-strength connection to, your overlap isn't likely to expand much beyond the boundaries of the workplace and the team members you have in common.

Where the concept of tribe evolved as a communal construct, it has now become a thing unique to the individual.

To complicate matters further, each individual also participates in more than one tribe. If you are like me, you probably have a work tribe. If you work for a large company, there are likely multiple sub-tribes within that company and you may belong to more than one. You probably also have a family tribe. If you're lucky, you also have at least one circle, or tribe, of friends that exists outside of these other tribes. All of these tribes vie for priority in your life, and all of these tribes are made of members who are similarly timeshared out to other tribes.

The tribe was once a communal, singular well of shared context, and it is now a multiplexed property attached to you and you alone. Still, our desire to belong to and feel a sense of community within the tribal spheres that we are a part of is an overriding factor in all facets of life and the decisions we make. And that feeling, among other things, is an expression of desire -- of a preference for making decisions in environments where we feel like we have high levels of shared context. When it comes to making important decisions, high context equals high confidence.

There are many factors that can play into the success or failure of a small startup company, but Dunbar's Number is absolutely a part of their capacity to move fast and get so much done when compared with the later stages of the company. Think about it -- at the beginning, the tribe is small and the stakes are high. The size of the group makes it easy to maintain fully-shared context for each group member. The fact that the stakes are high sends an unconscious signal to Dunbar's service to place each member into a high-quality connection layer. The amount of shared context for that setting and the individuals within it provides lubrication for every decision and action. The signal is clear: "Go go go! Food, shelter, and safety are at stake. Stick close to the tribe and help each other out or we could all end up alone and hungry, out in the cold."

Bring it on Home

Let's finally tie this back now to the statement about the fragility of the networking layer with something that vaguely resembles a proof based on the things we’ve discussed so far.

1. Humans today operate in multiple,? context-poor environments

• because modern tribes mostly consist of scattered pockets of loosely-connected members, and
• members each belong to multiple tribes.

2. This totally sucks

• because speedy, accurate decision-making (agility and adaptability) is what keeps tribes safe,
• but speedy, accurate decision-making requires large amounts of context.

3. So decision-makers in context-poor environments under time constraints must either

• sacrifice accuracy and rely on luck (bad idea), or
• become adept at quickly acquiring and transferring context within time constraints.

Successful tribes of modern humans get really good at moving context because our lack of high-context environments places a functional requirement on us to acquire and transfer context with both frequency and velocity. As humans, we call this communicating. As independent nodes in a distributed application... yeah, ok we'd also call that communicating. So if you've spent time working with either humans or distributed applications, you know that the networking layer is prone to all kinds of problems. Dropped packets, stampeding herds, and version mismatch are just the beginning. We haven't taught computers how to screw up communication yet in all the ways that we humans have managed to do it, but I have faith we'll get there.

We'll talk later in the book about ways to work with this problem, but for now I just want to clearly define the constraint: the networking layer is pretty fragile, and in a distributed application where instructions and their accompanying truckloads of context are constantly being thrown around, that's kind of a big deal.

Don’t Lose Hope

No, seriously. I mean that. Yes, leadership is a hard gig. Yes, humans, both at the individual level and at the aggregate are a huge pain in the ass, and yes, learning how to effectively work with them to get consistent results is probably the hardest thing a person can learn to do. I will also freely admit that the journey will change you, for the better or, if you let it, for the worse. It can be humbling, exciting, boring, terrifying, and exhilarating, and it can go from one of those to its polar opposite faster than we can deal with sometimes.

It is all of those things.

But please, stay with me.

Everything we've just talked about in this part of the book is the What. It defines, in some way, the job and the path that has been set before you.

If that's made you anxious, if that's made you doubt, if that's prompted you to draft an email to your boss asking if you can go back to being an engineer -- please don't hit send yet, because what's next is important.

I want to tell you Why.

...

[End of Part 1: What]

(Flying Squirrels continues in part 10, below)