For weeks this study carried a headline finding: across every class-10 science book in the corpus — eleven volumes, roughly 530,000 words of biology, chemistry and physics — not one woman scientist is named. It was clean, it was damning, and it had worked its way into the synthesis, the gender analysis and the list of things every board has in common.
It never reached a reader, because the site was not yet public. It was three weeks from being published as fact.
Then someone who had studied in Punjab in the early 1990s read the draft and asked a simple question: isn’t Madam Curie mentioned anywhere? That’s weird, it used to be.
He was right and the finding was wrong.
What is actually there
Marie Curie appears in Khyber Pakhtunkhwa’s Physics 10, with her dates, on page 244. She appears in Punjab’s Physics 10 on page 174, in a sentence that runs: “Henri Becquerel discovered radioactivity in 1896, and later, Marie Curie and Ernest Rutherford studied it further.” She appears in Khyber Pakhtunkhwa’s Chemistry 9 on page 40, inside a nicely-turned puzzle: “Cm is the chemical symbol for curium, named after the famous scientist Madam Curie. Why wasn’t the symbol C, Cu or Cr used instead?”
And there is a fourth name we had not found at all. On page 252 of Khyber Pakhtunkhwa’s Physics 10:
Nuclear fission was discovered in December 17, 1938 by Otto Hahn and his assistant Fritz Strassmann, and explained theoretically in January 1939 [by] Lise Meitner and her nephew Otto Robert Frisch.
On the next page there is a portrait: Figure 18.12 — Lise Meitner.
That is a woman credited with the theoretical explanation of nuclear fission, in her own right, in a Pakistani government physics textbook, with her face printed beside it.
How the error happened, which matters more than the error
Most occurrences of the string “curie” in this corpus are not the person. They are the curie, a unit of radioactivity, and the Curie temperature, the point at which a magnet stops being a magnet. Balochistan’s General Science 6 uses the second one three times and never mentions her. A count that does not separate the woman from the unit will produce nonsense in either direction.
But that is an explanation, not an excuse, and here is the part that should have caught it. The project’s own records already contained the contradiction. The synthesis said zero named women scientists. The gender analysis had located the Chemistry 9 mention and written the disagreement down, explicitly, as unresolved. Our editorial guideline says that where two of our own documents disagree we must resolve it or publish both readings and mark it unresolved.
Instead the wrong side was repeated. The contradiction sat in the files, correctly flagged, and got quoted past.
The corrected finding is more interesting than the false one
Here is what the recount actually shows.
| Board | Named women scientists in science books |
|---|---|
| Khyber Pakhtunkhwa | Marie Curie and Lise Meitner (Physics 10), Madam Curie (Chemistry 9) |
| Punjab | Marie Curie (Physics 10), once, in a list |
| Sindh | none |
| Balochistan | none |
Four mentions, in three books, from two of the five boards. None below class 9. The Federal Board, added later, names none in any of its six science volumes.
Now set that against the gender scores. On this project’s measures of women’s presence and agency across its whole corpus, Khyber Pakhtunkhwa is the worst board of the four. It scores 4.31 on a scale where 5 means women are largely absent or their roles prescriptively bounded. It is worse than Punjab, worse than Sindh, worse than Balochistan.
And it is the only board that names two women scientists, the only one that credits a woman with a major discovery in her own right rather than as somebody’s co-worker, and the only one that prints a woman scientist’s portrait.
What that tells you about aggregate numbers
An average over a whole board cannot see this. It reports Khyber Pakhtunkhwa as the most gender-conservative curriculum in the study, which on the aggregate evidence it is. It cannot tell you that a girl in Peshawar opening her physics book in class 10 meets Lise Meitner, and a girl in Karachi opening hers does not.
That is why this site publishes every book’s record separately, with the evidence attached, rather than only the board means. The means are real. They are also, on their own, capable of concealing the single most encouraging thing in the corpus.
What survives
The silence is still real, and it should not be softened by a good correction. Women scientists are named four times in 530,000 words. Two of four boards name none at all. No board names one below class 9. Rosalind Franklin has zero occurrences anywhere in the corpus while Watson and Crick are credited for DNA in three separate books.
“Vanishingly rare” is the accurate description. “Absent” was not, and we should not have said it.
This one was caught before publication, so there is nothing in the corrections log about it — that log records changes to claims readers have actually relied on, and no reader ever saw this one. From launch onward every correction goes there, dated, whether or not it suits the argument.
What the episode is worth keeping for is the shape of it. The error survived a scoring pass, a synthesis and a summary list. It was not caught by the method, by the checks, or by rereading. It was caught by one person with no access to the data, working purely from what he remembered being taught thirty years ago, who found a claim implausible and said so.
If you find another, that is the most useful thing you can do with this site.