#148 1952 · Grace Hopper / Remington Rand · Computing / software
Hopper had a working compiler that let programmers reuse code instead of hand-writing binary every time — and then spent years just getting anyone to believe a computer could do that at all.
问题
a genuine technical solution exists and works, but the field's own consensus about what the technology is even capable of prevents anyone from adopting it
背景
In the early 1950s, writing any new program for the UNIVAC computer meant hand-coding the same common subroutines in raw machine-level binary from scratch every single time, because the prevailing consensus among computer scientists and engineers was that computers were fundamentally arithmetic machines — capable of calculation, not of following or translating higher-level symbolic instructions on their own. This assumption wasn't a stated policy anyone had to defend; it was simply taken as an obvious fact about what a computer could and couldn't do.
Grace Hopper, working at Remington Rand's UNIVAC division, believed the field's assumption was wrong and built a working translator, the A-0 System, that let programmers reference reusable, previously written subroutines by a short code rather than retyping the full binary each time — technically demonstrating that a computer could, in fact, translate and execute higher-level instructions.
换别人会怎么做
The natural approach after building a working technical solution would be to let the demonstration speak for itself — show that the compiler works, publish the results, and expect colleagues to adopt it once its functionality is proven, treating adoption as a natural consequence of a working proof.
他们看到了什么
Hopper saw that technical proof alone couldn't overcome an assumption the field didn't even know it was making — colleagues weren't rejecting the compiler because they'd evaluated it and found it lacking, they were rejecting it because their entire mental model of what a computer was, an arithmetic machine incapable of translating higher-level instructions, made the compiler's actual function nearly impossible for them to correctly perceive. The fix wasn't a better demonstration of the technology, it was sustained, repeated persuasion aimed at changing the underlying belief itself, since the resistance was categorical, not evidentiary.
那一手
Hopper completed a functioning A-0 compiler in 1952 that measurably reduced redundant hand-coding, but rather than immediate adoption, she faced years of active skepticism from colleagues and management who, in her own words, 'told me computers could only do arithmetic' — she spent the following years running repeated demonstrations, presentations and direct persuasion of skeptical managers and programmers to get the tool actually used.
为什么管用
Presenting a genuinely working A-0 compiler to colleagues who held a prior, unstated belief that computers could only do arithmetic showed that a single demonstration doesn't overturn a categorical assumption — people could watch a working compiler and still not update their broader model of what computers were capable of in general. Hopper responded not by building a more convincing technical demonstration but by running the same demonstration repeatedly, across many audiences and over years, giving skeptics multiple opportunities to have their assumption challenged until belief gradually shifted. Once enough people's underlying model of computing capability updated, the compiler concept became acceptable not because the technology had changed, but because the field's assumptions about what such technology could do finally caught up to what had already been technically proven years earlier.
值了多少
Hopper's sustained persuasion effort eventually succeeded, and the compiler concept she pioneered led directly to her subsequent work on FLOW-MATIC and the eventual development of COBOL, one of the most widely used programming languages in business computing history — establishing compilers and higher-level programming languages as standard practice rather than a rejected curiosity.
什么时候会失灵
The mechanism depends on the resistance genuinely being a categorical assumption rather than a legitimate technical objection — if skeptics have real, specific concerns about a solution's actual limitations, repeated demonstration of the same working proof wouldn't address those concerns and could look like stubbornness rather than persuasion. It also requires the innovator to have the standing, patience and institutional access to keep demonstrating over years, which not every innovator with a working solution has the position or career runway to sustain. And the technique specifically addresses adoption blocked by belief, not adoption blocked by cost, infrastructure or competing priorities — a genuinely believed-in but still unaffordable solution requires a different kind of persuasion that repeated technical demonstration alone wouldn't resolve.
后来呢
Grace Hopper's compiler work is now recognized as foundational to all modern software development, and her own account of the resistance she faced — technical proof existing well before institutional belief caught up — is frequently cited in the history of computing as the template case for how genuinely field-changing technical innovations often require as much sustained persuasion work as the original invention itself.
资料来源
- [1]A-0 SystemWikipedia, 2026en.wikipedia.org
- [2]Milestone-Proposal: Grace Hopper's Compiler and Programming Language Work, 1952-1959IEEE Milestones Wiki, 2021ieeemilestones.ethw.org
- [3]The Woman Who Taught Computers EnglishMissed History, 2024missedhistory.com