Compli/cat/ed
Complicated
Overview
- Something complicated has many parts, dependencies, or steps, so it can feel difficult or expensive to work with—but the underlying situation is still knowable and analyzable in principle.
- With enough expertise, time, and decomposition, you can often trace causes, predict outcomes, and apply a right answer or a good procedure (even if discovering it was painful).
Context
- The word is casual English (“this deployment is complicated”) and also technical in sensemaking frameworks (for example the Cynefin framework), where complicated names the domain of known unknowns resolved by analysis and good practice, contrasted with Complex situations where cause and effect emerge and are clearer only after the fact.
- Complicated work can still be hard; the distinction is not “easy vs hard” but decomposability and predictability of the system’s behavior under study.
Key principles
- Decomposition — Subproblems can be separated, owned, and reasoned about with expert judgment and established methods.
- Legibility — With enough mapping, the system’s structure and interaction rules can be made explicit (diagrams, contracts, runbooks).
- Optimizable — There is often a meaningful sense of better vs worse designs once constraints and goals are stated.
- Complicated — Relationships are intricate but often stable enough that experts can forecast what will happen when you change X; surprises tend to come from scale, detail, or missing information, not from novel emergence every time you intervene.
- Complex — Many interacting forces produce emergent behavior; the same intervention can land differently depending on history and context; useful knowledge is often built through probing, feedback, and safe-to-fail experiments, not a single exhaustive blueprint.
- Practical heuristic — If the hard part is doing all the work carefully, it may be complicated. If the hard part is that the situation keeps teaching you something new when you touch it, it may be complex.
Examples
- Complicated — A large but well-specified migration with many moving parts; tuning a known performance model with many knobs; a dense legal agreement where experts can still say what clauses imply.
- Not automatically complex — “Many components” alone does not imply Complex; a complicated assembly line can still be mostly predictable if interactions are constrained and monitored.