Grace Hopper was a mathematician, computer programmer and U.S. Navy rear admiral. She led pioneering compiler work, helped develop word-oriented programming through Flow-Matic, and contributed to the standards work that shaped COBOL. Her own interviews and lectures ground the lessons below. — CHM Oral History (1980).

Part 1: Challenging the Status Quo
- Question inherited procedures: Hopper treated the claim that a computer shop had always done something one way as a warning to reconsider the procedure in light of new tools. — NSA Lecture, Part One.
- Use a counterclockwise clock to challenge assumptions: Hopper kept a counterclockwise clock to show that an apparently fixed convention could have worked differently. — NSA Lecture, Part One.
- Automate work you have already solved: When programmers repeatedly copied routines and adjusted addresses, Hopper asked why the computer could not do those mechanical tasks; that insight led to her compiler work. — CHM Oral History (1980).
- Challenge assumptions about what computers can do: Hopper recalled that her compiler idea took years to gain acceptance even after she demonstrated that machines could translate higher-level instructions. — Vassar Interview.
- Expect resistance to new technical ideas: Hopper urged younger colleagues not to accept the first refusal: some decision-makers needed to see a proposal return before taking it seriously. — NSA Lecture, Part One.
Part 2: Managing Things, Leading People
- Manage things and lead people: Hopper distinguished administering resources from caring for and guiding the people doing the work. She applied the distinction beyond the military. — NSA Lecture, Part Two.
- Do not let management displace leadership: Hopper warned that narrow attention to current financial or reporting measures could crowd out longer-term leadership and care for a team. — NSA Lecture, Part Two.
- Make loyalty reciprocal: Hopper described leadership as loyalty up and down: keep superiors informed while taking care of the crew. — NSA Lecture, Part Two.
- Learn the organization you must work within: In a direct television interview, Hopper joked that she could not derive the Navy’s organization logically and had to memorize it. — Letterman Interview.
- Rest when duty allows: Hopper recalled a practical boot-camp lesson: if ordered to wait with nothing to do, sleep because the next chance is uncertain. — Letterman Interview.
- Turn data into useful decisions: Hopper argued that data was the raw material and information the output of computing; system design should prioritize the information people need to act. — NSA Lecture, Part One.
- Give young colleagues room to solve problems: Hopper told interviewers that she assigned young staff a goal, backed them, and often stepped away so they could work out a new solution together. — CHM Oral History (1981).
Part 3: The Architecture of Computing
- Remember the Mark I’s scale: Hopper described the Mark I as roughly 51 feet long, eight feet high and eight feet deep, with 72 words of storage and three additions per second. — NSA Lecture, Part One.
- Notice how Hopper spoke of the Mark I: Hopper referred to the Mark I as “she” while explaining the machine’s dimensions and capabilities; the old bracketed “computer was originally she” line was not a standalone quotation. — NSA Lecture, Part One.
- See the computer as a tool for thought: Hopper contrasted the Mark I with earlier tools: in her telling, computing extended a person’s mental rather than physical work. — NSA Lecture, Part One.
- Think in systems, not isolated machines: Hopper warned against choosing a computer first and then adding programs; organizations should understand the whole flow of information before selecting hardware. — NSA Lecture, Part One.
- Describe relationships across a whole organization: Hopper argued that effective systems design begins with the information moving among parts of a company and the interfaces between its subsystems. — CHM Oral History (1980).
- Remember that the field had no playbook: Asked how she knew about computers when assigned to the Mark I, Hopper replied that she had to learn because it was her first encounter with one. — Letterman Interview.
- Treat early computers as a Model T stage: Hopper used the Model T analogy to argue that the computing industry was entering a wider phase of adoption and needed to plan for its second-order effects. — NSA Lecture, Part One.
- Prioritize information, not mere volume: Hopper pressed organizations to compare the value, timing and decision impact of different information instead of giving priority to the loudest requester. — NSA Lecture, Part One.
- Value information separately from hardware: Hopper asked how organizations valued the information they processed and argued that the data and resulting information deserved more attention than the tools alone. — NSA Lecture, Part One.
Part 4: Language, Compilers, and COBOL
- Make programs portable across machines: Hopper championed standard languages and compiler validation so programs could move among machines rather than stay bound to one vendor. — CHM Oral History (1980).
- Write for the people who use the language: Hopper argued that data processors were accustomed to words, not mathematical symbols, and should be able to express work in English-like programming statements. — Vassar Interview.
- Design languages for distinct users: Hopper distinguished symbol-oriented mathematical work from word-oriented data processing and favored languages fitted to particular application areas. — CHM Oral History (1980).
- Let machines assemble reusable routines: Hopper’s compiler work automated the copying and relocation of tested subroutines so programmers did not have to rebuild the same components by hand. — CHM Oral History (1980).
- Show a working compiler before asking for belief: Hopper recalled that her compiler idea initially met disbelief and took roughly two years to be accepted; the demonstration mattered more than asserting that it was possible. — Vassar Interview.
- Work within early memory and control constraints: Hopper explains how her one-pass compiler dealt with branches to program locations that had not yet been assembled; the legacy tape-pass claim was not established. — CHM Oral History (1980).
- Test language implementations against a standard: Hopper’s Navy work included compiler testing to make standard-language programs portable rather than vendor-specific. — NSA Lecture, Part Two.
- Keep modules and interfaces explicit: Hopper described independent compiler modules with documented interfaces so a change in one component would not silently break another. — NSA Lecture, Part Two.
- Validate compilers for portability: Hopper supported a testing service for COBOL and FORTRAN compilers; compatible language claims needed to be checked in practice. — NSA Lecture, Part Two.
Part 5: Bureaucracy and Execution
- Act on good ideas without endless permission loops: Hopper described doing useful work and apologizing if necessary when a hierarchy would otherwise stall it; this was her context, not a universal rule to ignore safeguards. — NSA Lecture, Part One.
- Go back after an initial refusal: Hopper told younger officers and civilians that a first no need not end a worthwhile proposal; she sometimes tried again or found another route. — NSA Lecture, Part One.
- Value service beyond formal awards: A U.S. Navy retrospective reports that Hopper regarded serving in the Navy as her highest honor, despite the many other awards she received. — Navy CHIPS Retrospective.
- Return when the work still matters: Hopper described her 1966 reserve retirement and recall to active duty in 1967, after which she continued work on standards and technology. — Vassar Interview.
- Enter service despite an unusual age: Hopper said she entered the Navy at 37 during World War II, motivated by the shared war effort. — Letterman Interview.
- Acknowledge family influence without reducing the decision: Hopper said her admiral grandfather helped draw her toward the Navy, while the wartime need provided the immediate reason to serve. — Vassar Interview.
- Expect a practical transition after military service: After retiring, Hopper joked to Letterman that changing from flat naval shoes to civilian footwear was unexpectedly awkward. — Letterman Interview.
- Remember the collective wartime purpose: Hopper recalled entering the Navy when, in her view, the country was collectively focused on the World War II effort. — Letterman Interview.
Part 6: Systems Thinking and The Model T
- Add another computer rather than only enlarging one: Hopper used the oxen analogy: when one could not move a load, people added a second; she applied that thinking to computer power. — NSA Lecture, Part One.
- Build systems of cooperating computers: Hopper argued that increasing power could come from multiple coordinated computers, not just ever-larger single machines. — NSA Lecture, Part One.
- Start with total information flow: Before acquiring hardware, Hopper wanted organizations to map how information moved through their work and then select machines to support it. — NSA Lecture, Part One.
- Notice tradeoffs in operational decisions: In a television interview, Hopper recalled that a Pentagon–Navy Yard boat service had ended because the Carter administration judged its energy use too high. — Letterman Interview.
- See computing as an emerging industry: In her 1982 lecture, Hopper compared computers with the early Model T era and predicted a much larger computing industry ahead. This remains a historical forecast, not a current market estimate. — NSA Lecture, Part One.
- Share the search for knowledge across a team: Hopper described assigning staff different technical reading areas and circulating the useful articles so the group could track more than any one person could. — CHM Oral History (1981).
- Use machines to handle data while people judge information: Hopper said data processing should produce useful information; she did not equate more hardware or processing with better human decisions. — NSA Lecture, Part One.
Part 7: Debugging, Latency, and the Physical World
- Make speed serve better answers: Hopper explained that demand for better answers faster drove computer performance, while physical limits still mattered. — Letterman Interview.
- Make a nanosecond tangible: Hopper asked engineers to cut a wire representing the maximum distance light could travel in a nanosecond, helping audiences visualize latency. — NSA Lecture, Part One.
- Use physical distance to explain latency: Hopper gave about 11.8 inches as the distance light travels in one nanosecond; the wire was a teaching model, not a claim that electricity in wire reaches light speed. — NSA Lecture, Part One.
- Show the size of a microsecond: Hopper used a roughly 984-foot wire to make a microsecond concrete and show programmers that small delays accumulate. — NSA Lecture, Part One.
- Explain the next unit before optimizing for it: Hopper described the picosecond as one-thousandth of a nanosecond while discussing the physical limits of faster computation. — NSA Lecture, Part One.
- Correct the moth story’s date and participants: The preserved Mark II logbook records a moth found in a relay in 1947. It was a team incident, not proof that Hopper herself discovered or taped it. — Smithsonian Bug Logbook.
- Do not claim Hopper invented “bug”: The Smithsonian notes that engineers used “bug” before electronic computers; Hopper and the Mark II team helped popularize the computer usage after the moth incident. — Smithsonian Bug Logbook.
- Read the original logbook wording: The 1947 logbook labels the taped moth the first actual case of a bug being found, a literal joke about an already-used engineering term. — Smithsonian Bug Logbook.
Part 8: The Future, Youth, and Education
- Invest in young people as a legacy: Hopper said training and encouraging younger colleagues mattered to her at least as much as building the compiler. — Vassar Interview.
- Back people when they try: Hopper said she urged young colleagues to try ambitious work, supported them, and checked in later so they did not stop taking chances. — Vassar Interview.
- Look ahead instead of defending old methods: Hopper said change is slow and urged people to revisit established procedures as computing capabilities changed. — Vassar Interview.
- Protect equal access to education: Hopper credited her father with insisting that his daughters have the same educational opportunity as his son, which let her and her sister attend Vassar. — Vassar Interview.
- Notice how teaching skills emerge: Vassar’s archival history records that Hopper tutored fellow students in mathematics and physics and discovered an aptitude for teaching. — Vassar Archival History.
- Support curiosity with tools and confidence: Hopper’s own account ties curiosity to learning how things work and ties mentoring to giving younger colleagues the chance to try; the legacy exact aphorism is not independently authenticated. — Vassar Interview.
- Listen to technically current younger people: Hopper argued that senior officials who could not keep pace with rapid technical change needed to listen to junior colleagues. — NSA Lecture, Part One.
- Carry new technical work back to the classroom: Vassar’s archival account of Hopper’s 1946 campus talk says a long Mark I calculation was to be given to its library, making early computing work available to her alma mater. — Vassar Archival History.