There’s a saying I picked up from Gardener’s Question Time (ikr) that if something’s gone wrong in the garden, the issue’s probably underground.
Above the surface, things always look clearer. Leaves go pale. Stems wilt. Fruit sets and then drops. We call it sunburn, or overwatering, or disease. But more often than not, we’re diagnosing shadows. What’s really happening is out of sight.
So we dig.
We check the roots. Because nine times out of ten, that’s where the system failed quietly, weeks before symptoms appeared. And once you’ve seen enough root failures, you realise it’s the same pattern you’ve seen in servers, relationships, or institutions: fragility forming in the layer no one’s watching.
That’s what this post is about.
Failure is Slow, Then Sudden
When a plant collapses from the top down, it looks like it’s happened all at once. One morning everything is fine; by evening the leaves have dropped.
But it’s never instant.
What you’re seeing is the surface layer reacting to failure that started days or weeks earlier. A fungal infection. Root rot. A pest you never spotted. Or a simple mismatch—too much water, too little air, and the roots suffocate before they can signal distress.
This is what most post-mortems in tech get wrong: they analyse the moment of impact. But roots die quietly. They fail in isolation. And by the time the system goes down, the damage is embedded.
Healthy Roots Are Invisible
In any system, garden or digital, the healthiest components are the ones no one notices. Good root systems hold quietly. They don’t ask for praise. They don’t throw alerts. They simply function, absorbing what they need, supporting what they must.
The trouble is, that invisibility leads to neglect.
In infrastructure, it’s the legacy code no one wants to touch. In a team, it’s the person who never misses a deadline and never asks for help. In a garden, it’s the soil you forgot to test because everything looked fine.
We’re trained to prioritise what’s visible. And that’s where most preventable failures begin.
Blame Is a Surface-Level Response
In systems engineering, root cause analysis is meant to be a calm, structured process. But in reality, it’s often about blame. Who missed the sign? Who approved the deploy? Who let the plant sit in waterlogged soil?
But roots don’t lie. They don’t care who was on call. They just reflect the environment they were given.
You want the real answers? Dig.
What conditions allowed this failure to take hold? What signals were ignored—not because of laziness or incompetence, but because they were designed out of the monitoring process?
That’s the heart of root cause analysis: not “what failed”—but “what was never seen”.
Theory into Practice
In my own systems, garden, software, life; I’ve started building in more observation. Not dashboards. Not alerts. Just space.
- Dig before planting.
- I talk to the people I work with, even when things seem fine.
- I log changes when I make them, even if the outcome is unclear.
- I listen to the roots, not just the leaves.
Because the problem is rarely the thing that broke. It’s the thing that we didn’t know was breaking.
And the only way to know is to stop assuming the surface has the answers.
Soil and Software and Self
A plant that fails from the root isn’t doomed. It can be repotted. It can be fed, pruned, given new ground. But it takes time. Recovery is never instant, and it never starts at the top. It starts below. Quietly. Patiently.
That’s true for people, too.
If something feels off, if a part of your life, your project, your team, or your garden starts to wilt; don’t just treat the symptoms.
Dig deeper.
And don’t stop until you hit roots.
Part of my Garden as System series.