Alan Perlis, 1982

AI-generated editorial portrait of Alan Perlis in an early-1980s computer science office

“Simplicity does not precede complexity, but follows it.”

Alan Perlis published this sentence as epigram 31 in “Epigrams on Programming,” a playful and serious collection in the September 1982 issue of SIGPLAN Notices. Perlis, the first recipient of the ACM Turing Award, used compact paradoxes to expose habits programmers rarely examine. This one reverses an attractive story about design: that good systems begin with a perfectly simple idea and become complicated only through mistakes. In practice, the simplicity people trust often arrives after difficult exploration. Designers encounter exceptions, competing goals, and awkward interactions, then discover which distinctions matter and which can be removed without hiding something essential.

Early simplicity can be ignorance wearing a clean diagram. A new interface looks elegant when it has not yet met unreliable networks, unusual data, adversarial input, accessibility needs, or users who combine features in unexpected ways. Complexity supplies evidence. It reveals where two concepts that seemed identical must remain separate and where ten special cases are expressions of one deeper rule. The later simplification is therefore an achievement of compression. It preserves the behavior that experience proved necessary while replacing accidental detail with a smaller model. That is why a short, mature API may contain more thought than a long first version, even though it exposes fewer controls.

Perlis was not arguing that every project should wander through disorder or that complexity is automatically profound. Teams can learn from earlier systems, constrain scope, and avoid known traps. His point is about sequence: durable simplicity must answer the world it simplifies. Premature cleanup can merely push hard decisions into undocumented assumptions, where they return as bugs. The useful discipline is iterative. Build enough to discover the real shape of the problem, name the recurring concepts, and then refactor around them. Tests help distinguish essential behavior from historical clutter, while clear boundaries let developers remove an internal complication without changing the promise made to users.

The epigram is especially valuable in an era of frameworks, cloud platforms, and artificial-intelligence tools that can produce polished beginnings almost instantly. A generated prototype may appear coherent before anyone has tested failure recovery, privacy boundaries, maintenance costs, or operational scale. Perlis’s sequence warns against confusing the absence of visible detail with understanding. The path to a system that feels obvious may include experiments, rejected abstractions, and revisions that teach the team what “obvious” should mean. Simplicity deserves trust when it has survived complexity, not when complexity has merely been postponed. The clean result is the residue of learning: fewer concepts, better chosen, carrying more of the truth.

Perlis included the sentence as epigram 31 in his 1982 collection of concise observations about programming practice.

The collection used paradox and humor to challenge assumptions about languages, design, learning, and the daily work of programmers.

Reliable simplicity is a compression of lessons learned from real constraints, exceptions, failures, and competing goals.

The idea does not praise needless complication; it warns that an untested clean design may only hide the complexity it has not encountered yet.

Fast prototypes and generated code can look finished before teams understand their failure, privacy, maintenance, or scaling requirements.

Iteration, testing, and refactoring turn that early surface simplicity into a smaller model that has earned the right to appear obvious.

Explore more "Quotes of The Day"

Discover more notable quotes from influential voices across politics, science, business, technology, sports, and culture. Each quote offers insight into how ideas, beliefs, and decisions shape the world around us.