Their Names Were Left Off the Paper

In 1932, a 27-year-old Hungarian mathematician named Rózsa Péter stood before the International Congress of Mathematicians in Zürich and presented a paper on recursive functions. What she laid out that day—a rigorous, systematic account of functions that call themselves, building complexity from simple self-referential rules—was the mathematical foundation on which all modern computing would eventually rest. She was not yet famous. She would never become famous, in the way that word gets applied to scientists.

She had arrived at mathematics sideways, enrolling at Pázmány Péter University in Budapest to study chemistry before the mathematics pulled her in. By her late twenties she was working on the foundations of logic independently, devising her own proofs for results that contemporaries who would become far more celebrated were also pursuing. Her 1932 paper was not a detour or a preliminary sketch. It was the opening move of a four-decade research program: a systematic study of recursion, establishing hierarchies of recursive functions, clarifying precisely what it means for a process to be computable at all.

Without her work, there's no loop. There's no function calling another function. There's no programming language with any expressive power. Recursion is why a computer can do anything requiring nesting, repetition, or self-reference. That covers most of what computers do.

In 1951, she published Rekursive Funktionen—the first book on modern mathematical logic ever published by a woman. It remains a foundational text. In 1961 came Playing with Infinity, a popular mathematics book translated into a dozen languages, still in print. She was by all accounts a devoted and inspired teacher at the Budapest Teacher's College, where students called her “Aunt Rózsa.” She wrote and translated poetry. She was a complete person of unusual warmth and intellectual force.

In 1939, Hungary's anti-Jewish laws stripped her of her teaching position. She continued working. She continued publishing. She and her mother survived the war in hiding in Budapest. She returned to teaching in 1945 and was appointed to her alma mater in 1955. She died in February 1977, one day before her seventy-second birthday.

Her name does not appear in most popular histories of computing. The standard genealogy runs from Turing's theoretical machines to von Neumann's architecture to the first programmable computers. Rózsa Péter's recursive functions—the abstract machinery that makes Turing's formalism work—appear as an anonymous foundation. Everyone builds on it. Nobody says her name.

In 1900, a girl named Hadassah Itzkowitz was born in a small Jewish village in Ukraine, roughly midway between the towns of Nemirow and Tulcin. Thirteen years later, her family brought her to the United States. Six years after that, she won a scholarship to Cornell University. She studied mathematics. She would eventually do something almost nobody alive at the time had done: design a programming language.

At the National Bureau of Standards in Washington, working through the early 1950s, Ida Rhodes designed the C-10 language for the Census Bureau's UNIVAC I. This was the first operational stored-program computer in the United States, and someone had to tell it what to do in a structured, repeatable way. Rhodes was that person. C-10 was not a high-level language in the modern sense—the concepts needed to build one were still being invented—but it was a systematic framework for instructing the machine. It worked. The Census Bureau ran it. The Social Security Administration ran programs she designed on the same platform.

Think about what that last sentence means. The Social Security Administration in the early 1950s was tracking the retirement contributions of tens of millions of Americans. A Jewish woman who had immigrated from Ukraine at age thirteen wrote the program that processed that data.

She received an Exceptional Service Gold Medal from the Department of Commerce in 1949. She was cited as a UNIVAC I Pioneer at the 1981 National Computer Conference in Chicago. She died in 1986, at 86, having outlived many of the machines she programmed.

The history of programming languages, as it's typically taught, begins with FORTRAN in 1957, moves through COBOL to C and beyond. Ida Rhodes appears in specialist literature and not much else. She doesn't feature in the canonical popular accounts. The C-10 system is rarely mentioned in the standard narrative of how humans learned to talk to machines. She is a woman who was there at the beginning, doing foundational work, and is not in the story.

In the summer of 1955, Mary Tsingou sat at a terminal connected to the MANIAC computer at Los Alamos National Laboratory and watched the simulation she had written run for the first time. She was 26 years old. She was one of the first programmers on a machine built to calculate thermonuclear yields. What she saw on that terminal screen had never been seen before, by anyone, anywhere.

The experiment had been Enrico Fermi's idea: take a mathematical model of a one-dimensional crystal—a chain of masses connected by springs—add a small nonlinear term to the interactions, and run it forward in time. Classical statistical mechanics predicted that energy introduced into one mode of oscillation would gradually spread into all other modes, the system settling eventually into equilibrium. This was equipartition. It was expected.

Tsingou had translated this physical model into an algorithm the MANIAC could execute. She wrote the code. She ran the computation. And then she watched something happen that nobody had predicted: the energy spread into other modes, yes—and then it flowed back. Almost perfectly. The system returned to something very close to its starting state, as if it remembered where it had been. Fermi, standing nearby, reportedly said it was one of the most interesting results he had ever seen in physics.

What Tsingou had witnessed became the seed of chaos theory, soliton physics, and a fundamental reconsideration of how nonlinear systems behave. The result has been continuously studied and extended for seventy years. Fermi died of cancer five months later, in November 1955, before he could publish formally. The 1955 report circulated as a preprint for decades and became one of the most cited documents in theoretical physics.

The problem it described was called the Fermi-Pasta-Ulam problem. For fifty-three years.

In 2008, a paper in Physics Today made the case that this name was historically wrong. The computation was Tsingou's. The algorithm was Tsingou's. The discovery was, in an essential sense, made by watching the output she had coded run on a machine she had programmed. The problem was renamed. It is now officially the Fermi-Pasta-Ulam-Tsingou problem.

Mary Tsingou-Menzel, born in Milwaukee in 1928, is still alive at the time of this writing. She continued working at Los Alamos until 1991. She saw her name restored before she turned eighty. Fifty-three years is a long time to wait for your name on your own discovery. It is not the longest wait in this article.

In 1972, Karen Spärck Jones published a paper in the Journal of Documentation containing a deceptively simple idea. Words that appear in many documents are less useful for identifying any particular document than words that appear in very few. A term's value for information retrieval should be weighted inversely by how common it is across the entire collection. She called this inverse document frequency.

Combined with term frequency—how often a word appears in a specific document—it produces a score known as TF-IDF. The idea is elegant, mathematically clean, and demonstrably effective. It became the standard method in information retrieval. It underpins every search engine that existed before the neural search era. Google's early architecture incorporated TF-IDF weighting. Every keyword search you ran in the 1990s, 2000s, and much of the 2010s was using, in some form, what Karen Spärck Jones worked out in a paper fifty-two years ago.

Spärck Jones was born in Huddersfield, England, in 1935. She taught herself programming. She joined the Cambridge Language Research Unit in 1957 and spent the next fifty years at Cambridge, working on natural language processing and information retrieval when both fields were barely named. In 1964, she published a foundational paper on semantic classification. The 1972 IDF paper followed. Her body of work over five decades shaped two disciplines.

She was an advocate for women in computing with a combination of dry wit and genuine urgency. Her slogan: “Computing is too important to be left to men.” She used it in talks and interviews for decades. It scanned as provocation. It was also a diagnosis.

She died in April 2007. The Karen Spärck Jones Award is presented annually for outstanding research in natural language processing and information retrieval. Her 1972 paper had over 4,500 recorded citations at the time of her death. That number has continued to grow.

What she built lives inside every phone in your pocket. There is no Wikipedia without her work. No Google. No search bar. She is not in any school textbook.

The pattern in these four lives is not subtle. A woman does the work. The work is recorded. Her name does not attach to it. Decades pass. Sometimes someone notices the gap.

None of these are bitter stories, exactly. Péter was loved by her students. Rhodes was honoured in her eighties. Tsingou lived to see her name restored. Spärck Jones spent fifty years at Cambridge doing work she cared about. They were, by most measures, scientists who had careers and lives. The erasure didn't stop them from working. It just stopped other people from seeing what they had done.

That's what makes it consequential. Names are not vanity. A named discovery is a path. Students can follow a path they can see. Funding and recognition flow toward visible work. A field whose history excludes women tells women, implicitly and continually, that they are guests. The telling doesn't have to be intentional to have effects.

Rózsa Péter built the logical foundation for all computation. Ida Rhodes designed one of the first programming languages. Mary Tsingou ran the first computer simulation in history. Karen Spärck Jones invented the mathematics inside every search engine. These are not small contributions. They are not peripheral achievements. They are the kind of foundational, durable, structurally indispensable work that earns statues and institutions and names on buildings.

Instead: a footnote. A gold medal. A name restored fifty-three years late. An annual award that most people in the field have never heard of.

Computing is too important to be left to men. Karen Spärck Jones was right. She was also, given everything above, far too generous with the people who had been running the story for so long without acknowledging who wrote it.