The problem. After the toggle switch and repressilator, building anything bigger was miserable: every part behaved differently in every new context, and each project started from scratch. Endy’s claim is that this is not a biology problem but an engineering one, biology lacked the abstractions that let other fields compose complexity.

The idea. Three borrowed principles. Standardization, define parts (promoters, RBSs, coding sequences) with measured, documented behavior so they can be shared and reused (the BioBrick/registry vision). Decoupling, separate design from fabrication, so you can reason about what to build without also solving how to synthesize it. Abstraction, a hierarchy (DNA → parts → devices → systems) where each layer hides the mess below it, so a systems designer needn’t think in base pairs.

Why it matters. This is the conceptual scaffold the whole field’s tooling hangs on, and the analogy to software is exact: datasheets, interfaces, layered abstraction, separation of concerns. It’s also the clearest articulation of why reproducibility is an engineering value, not a virtue, standard, characterized, composable units are what make work cumulative. The same instinct drives containerized, versioned analysis pipelines: decouple the definition of an analysis from the machine that runs it, and hide the plumbing behind a clean interface.

Verdict. Twenty years on, the diagnosis is right and the cure is partial. Standardization keeps foundering on biological context-dependence, a part’s behavior still leaks across its boundaries in a way a resistor’s doesn’t. But as a statement of what “engineering biology” should mean, it hasn’t been improved on. Read it as the philosophy under everything else in this list.