I’m currently reading Leadership is language by L. David Marquet, the same author of Turn the Ship Around. In this book, Marquet describes the different ways people communicate and how it influences the outcomes we get. This should not be any news really, but the book explains in much more detail how this happens. Especially when there is a power dynamic between the participants of a conversation or meeting.

The main theme across the book is the sinking of a ship called El Faro caused by a hurricane. The National Transportation Safety Board (NTSB) found the ship and recovered it’s voice recorder. It contained the last 25 hours of conversation on the bridge. The sad story is that the people on board were aware of the dangerous situation but the environment created by the company and captain did not allow for them to speak up. Seeing what they were getting into, the subordinates were not capable of convincing the captain to change route. The ship sank and all 35 people on board died.

In the domain of tech the consequences usually are not that drastic on an individual level. Hopefully no one gets killed because of a bad running project. Exceptions like writing software for medical equipment persist of course.

On an organizational level though, the consequences might just be that: life or death. Stafford Beer teaches in the Viable System Model that the main purpose of an organization is to survive. Thus the environment inside an organization influences survival and we need to be mindful of it. Does it work in the organization’s favour? Does it unwillingly work against us?

I want to describe how the language we use sometimes results in the exact opposite of what we aim for. Instead of the expected outcomes, it created side-effects that hindered the organization of moving faster or even prevented collaboration between teams.

“Own your development” Link to heading

There is hardly any on-boarding material handed out by HR that does not include this statement. Obviously this is right: when you don’t work on your development, who will? Promotions and pay rises are not given automatically based on the number of years people spend in a company. Working for the state might include a different point of view.

This statement alone is not enough though. There is also an organizational responsibility to it. If only you own your development, the smallest important work unit in the organization is the individual, not the team. People do not have a direct incentive to cooperate. Apart from the “we are all in this together” mantra. But when there is no cooporation within the team, everyone is on their own.

Consider the situation of senior engineers in a team. They will only get a promotion to something like a principal/staff engineer when they demonstrated they can run projects or initiatives across multiple teams. If the Engineering Manager’s task is not to find good projects for them when they are ready, everyone will fight about the projects that appear. Whether they are ready for the next step or not. If those projects don’t appear, they might find projects from other teams. Or make projects larger than they need to be, just because they want to advance their career.

Looking at this from a systems perspective a balancing feedback loop is missing. There is a reenforcing feedback loop (you own your career), which initially provides the intended results. But at some point the system becomes instable, because the process requires a controlling hand. Everyone only owns their career and suddenly it is not about the company anymore but careers. The Engineering Managers task is to be part of this balancing feedback loop: is the engineer capable of working on the next level. If yes, then go search for a project that would help them demonstrate they are ready for the step.

“Even better if” Link to heading

This was the main part of a Post Mortem or Incident Review template I noticed. It was the main section in the “what happened” part. No real root cause analysis like the 5-why’s. This technique helps to get to the core of a problem that happened. I have used this extensively to dig as deep as possible to find out what happened and how we could prevent it from happening again. My personal most memorable one was were we found a bug in the AWS network driver for a certain instance type.

But don’t get me wrong, the idea behind it is obviously right: there rarely is this one root cause for an incident. Most of the time an incident happened because a number of factors were at play.

But:

And it’s true. If the broader organization is not good at post-incident handling and improving the systems, not doing 5-why’s is an invitation for getting sloppy. And even better if states this plain and simple: whishful thinking.

“I’m open to feedback” Link to heading

I’ve heard this from a senior leader in an organization. It is probably meant as encouragement for feedback, but it does result in the opposite. Feedback is a two-way street and if you want to get feedback, especially as a leader, the first step is to provide feedback.

If the leader needs to tell their subordinates, that they “are open to feedback” this might show the opposite. It seems like they do not receive regular feedback and maybe need to tell their subordinates to give them feedback. Either way: this statement alone made from a boss is a the prime example that there probably is not a feedback culture.

“Strong opinions, loosely held” Link to heading

Michael Natkin writes in Strong ideas loosely held might be the worst idea in tech, “The loudest, most bombastic engineer states their case with certainty, and that shuts down discussion.” In the presence of a power dynamic I would go even further though:

If the top executive in the company has “strong opinions” it effectively destroys creativity. Everyone will always only want to do things that fit the opinions of that executive. But what if their reality does not match the reality on the ground? What if the people actually implementing the things the exec ask them to have a lot more insights? What if they could find a much better solution, if the leader would allow them to think for themselves? The outcome could actually be the same or close enough, but they way to it might be simpler or adjusted more to the situation engineers have in this environment.

And what does this say about the person stating it?

You have to accept that we actually need others to make us smart. And if we surround ourselves with people who confirm our existing thoughts, we get stupid. Quite frankly, stupid. (Jonathan Haidt in The Knowledge Project)

There is a simple hack that can give a way out of this: state the strong opinions in probabilities (“I’m 90% sure we should not use MongoDB”). This has at least two benefits: 1/ it opens the possibility of disagreement. People have the option of saying something contrary. Probably still only those with a lot of courage, but at least some can speak up.

And 2/, as Annie Duke writes in Thinking in Bets: Making Smarter Decisions When You Don’t Have All the Facts: it does enable better decision making. Especially when one does not have all the facts.

Conclusion Link to heading

Language is important in any social endeavour, private or business. Its influence goes a lot further and dependeing on what language a leader chooses, they get a different result and a different culture.

Be mindful of it and choose words wisely.

I have linked a number of books, blog posts, and podcasts on this topic, they are well worth the time invested.