Earlier this summer I was at a conference, watching a talk from a speaker about refactoring and rewriting parts of a codebase. I'm sort of assuming that if you're on this blog you understand what that means, but if you don't: imagine re-building a chunk of your house (your hallway, say) while continuing to use the building as your home. It's a complicated task, and I mean why bother refitting the carpet, just because it's threadbare in places and the jagged hooks at the door thresholds have started to come through?
Throughout this particular conference talk, as with several others I saw, I found myself thinking: "is this true?".
A speaker would say something like "this was a success, but the team reported being burned out after four months" and you think "is something a success if, after sixteen weeks, most people involved tell you they overextended and probably cannot continue working at that pace? Is that not something we should be attuned to as leaders? Should we breeze past that trade-off and call this a success?".
I think sometimes the background technical complexity associated with our profession, that of building and maintaining software, creates a sort of "correctness by association". People see that you work in a complicated field, and maybe assume that you've taken as much care over your ideas and language as would be required in their imagined idea of your profession. I do this too. I imagine nuclear safety experts have thorough meal plans, and never sleep through their alarm clock.
Dear reader, I have a Ph.D. and about a decade's experience building software and once upon a time, in the not too distant past, I cut fully into my hand while cutting some paper. I routinely forget most of what I am told in a twenty-four hour period (even it's important) unless I remember to write it down. We're all a little bit like this, or at least the vast majority of us are, or at least I am.
So long as I'm not out here giving talks about the best way to slice paper and not your own flesh, this is fine.
I think we can all agree that it's important to say things that are true. Or at least true-enough.
If I can hear something and ask "is that true?" in direct response to that statement, that statement might not be quite true enough. I'm not talking about second-order effects or a step in a chain of thought, I'm talking about "a minute ago you said X, and now you are saying Y, and I don't think X and Y can both be true at the same time".
If you're talking about an idea, or making decisions based on an idea, the core premise needs to withstand a certain amount of interrogation. The burden of proof shouldn't crush any pragmatic efforts, but someone at some point should ask "is this true?"
I personally arrive at tested ideas by writing, doing research for that writing, and by using the drafts for that writing as the basis for conversations with other professionals. Recently at work, I've found relatively dense, technical, and researched prose to be the best tool for convincing groups of people into a broad consensus: in order to deliver the next six-months of features, we are going to have to take a more event-native approach - here's what that looks like and what we should be careful of.
There are a lot of questions that fall out of that: why do it? Is it worth it? What do we do first? What should I learn?
These are natural follow up questions. They let me know that people are listening and engaging, and they start to give a specific shape to the implications of the things that came before it. But to give one of my favourite quotes:
There is surely nothing quite so useless as doing with great efficiency what should to be done at all
(A quote which, like so much, has its roots in the systemised formal education system.)
When the implications of making a poor choice are so high, i.e. when they lead you to do things which should not be done at all, your ideas need to stand up to at least one question.
This probably means that you need to be interrogating your work as you go: how does this relate to my specific context? What are the alternatives? Where has this been done before, and how did it turn out?
Because the even greater risk is that nobody asks you if what you're saying is true.
See other articles