Truncated text of tweet from MrPitBull, Mar 11, 2026:
She kept finding women in laboratory photographs from the 1800s. Then she read the published papers—and every single woman had vanished. Someone had erased them from history.
Yale University, 1969.
Margaret Rossiter was a graduate student studying the history of science. She was one of very few women in her program.
Every Friday afternoon, students and faculty gathered for beers and informal conversation. One week, Margaret asked a simple question: "Were there ever any women scientists?"
The faculty answered firmly: No.
Someone mentioned Marie Curie. The group dismissed it—her husband Pierre really deserved the credit.
Margaret didn't argue. But she also didn't believe them.
So she started looking.
She found a reference book called "American Men of Science"—essentially a Who's Who of scientific achievement. Despite the title, she was shocked to discover it contained entries about women. Botanists trained at Wellesley. Geologists from Vermont.
There were names. There were credentials. There were careers.
The professors had been wrong.
But Margaret's discovery was just the beginning. Because as she dug deeper into archives across the country, she found something far more disturbing.
Photograph after photograph showed women standing at laboratory benches, working with equipment, listed on research teams.
But when she read the published papers, the award citations, the official histories—those same women had disappeared. Their names were missing. Their contributions erased.
It wasn't random. It was systematic.
Women who designed experiments watched male colleagues publish results without giving them credit. Women whose discoveries were assigned to supervisors. Women listed in acknowledgments instead of as authors. Women passed over for awards that went to male collaborators who contributed far less.
Margaret realized she was witnessing a pattern that stretched across centuries.
Women had always been present in science. The record had simply pushed them aside.
She needed a name for what she was documenting.
In the early 1990s, she found it in the work of Matilda Joslyn Gage—a 19th-century suffragist who had written about this exact phenomenon in 1870.
In 1993, Margaret published a paper formally naming it: The Matilda Effect.
The term captured something that had been hidden in plain sight for generations. Once you knew the term, you saw it everywhere.
Her dissertation became a lifelong mission.
For more than 30 years, Margaret researched and wrote her landmark three-volume series: Women Scientists in America. She examined letters, institutional policies, individual careers. She gathered undeniable evidence that women in science had been consistently under-credited and structurally excluded.
Her work faced resistance. Many dismissed women's history as political rather than academic. Others insisted she was exaggerating.
Margaret didn't argue emotionally. She presented data. Documented cases. Patterns repeated across decades and institutions.
Eventually, the evidence became undeniable.
Her research helped restore recognition to scientists who had been erased:
Rosalind Franklin, whose X-ray work revealed DNA's structure—credit went to Watson and Crick.
Lise Meitner, who explained nuclear fission—omitted from the Nobel Prize.
Nettie Stevens, who discovered sex chromosomes—received little credit.
Cecilia Payne-Gaposchkin, who discovered stars are made of hydrogen—initially dismissed.
And countless others whose names had nearly vanished.
Margaret changed the narrative. Science was no longer just the story of solitary male geniuses. It became a story of collaboration that included women who had been written out.
The Matilda Effect became standard terminology. Scholars used it to examine how credit is assigned, how authors are listed, who receives awards, who gets left out.
This is an important concept, but the piece is written by AI.
There are a number of tells, but this is an excellent example to talk about em-dashes, which people often either take as permanent AI tells or run the other way and say "humans use em-dashes and that's why AI does, too! they're not tells!" Both are kind of right and both are kind of wrong.
What you'll see if you look closely at this text is that it ONLY uses em-dashes. Every time it needs to put in some kind of break or set off some text, it goes for the em-dash. There are no phrases in parentheses. There are commas, but only in places where the absolute rule is to use a comma (like in a series, for instance). There is one colon, again placed where the absolute rule is to use at (at the top of a list). Whenever there's an option, where a human writer would be actively making a choice about what punctuation to use, the AI defaults to an em-dash.
On top of that, look at the content. The AI bot people are obsessed with feminism, ironically. I suspect it's because very basic feminist narratives about women pushing back against barriers or doing something heroic are popular and gets shared widely. So, first of all, you should be on your guard when you see a "what this woman did CHANGED HISTORY!" kind of piece. (I wonder if the twitter/tumblr trend of BUCKLE UP history posts has affected the AI ...) And then you should check out the specific claims.
She kept finding women in laboratory photographs from the 1800s. Then she read the published papers—and every single woman had vanished. Someone had erased them from history.
I can't find this anywhere else. The paper "The Matthew Matilda Effect in Science" doesn't talk about photos! The Wikipedia page doesn't talk about photos! This Smithsonian article doesn't talk about photos! Her piece on her career in Writing and Revising the Disciplines (2002) (good read) DOES mention photos, in that she got the Mount Holyoke archivist to send her a few from the 1880s showing women doing scientific work as a nice illustration that "epitomized" what she was already aware of.
Rossiter started with textual primary sources that documented women as named individuals contributing to scientific discoveries. The idea of her being confused by photos is a hallucination.
Despite the title, she was shocked to discover it contained entries about women. Botanists trained at Wellesley. Geologists from Vermont.
There's definitely something to be said about the framing of this bit as shocking!!! but since I'm talking about facts and sources, it's clear to me that the AI recognized the botany-Wellesley connection from the paper but could not parse that the reference was to a female botanist who taught at Wellesley. There is also nothing in the paper about Vermont geologists, so I have no idea where the AI got that; I would suspect it's another hallucination attempting to create a pattern from the first reference.
But Margaret's discovery was just the beginning. Because as she dug deeper into archives across the country, she found something far more disturbing. Photograph after photograph showed women standing at laboratory benches, working with equipment, listed on research teams. But when she read the published papers, the award citations, the official histories—those same women had disappeared. Their names were missing. Their contributions erased.
Again, back to the mysterious photographs. But the rest of this text is an issue as well: what Rossiter describes in the paper is not a complete absence of these women in any official documentation, but that these women were amply documented and known to be working within the scientific community and yet did not receive public credit or awards. It's not a complete smothering out, but a sort of complacent back-burnering, which is too nuanced for the AI to be able to handle when told to "write a post about the Matilda effect that will get engagement on social media". She didn't prove that discoveries attributed to male authorship actually had women involved and only she knew their names: she collected many stories that people already knew of overlooked/underplayed female scientists and put them together to say, "This is a pattern and we should have a name for it." Some of her examples were even recent enough (1970s-80s) that she was able to point to a feminist backlash.
And again ironically, the AI itself engages in the Matilda Effect by presenting this whole thing as utter silence -> Rossiter gets curious -> the case is blown open. Rossiter actually refers to the work of other female historians and social scientists! In fact, she started this line of research after noticing the female biographies in American Men of Science when her housemate, Cynthia Thompson, recommended that she keep track of them.
Her research helped restore recognition to scientists who had been erased: Rosalind Franklin, whose X-ray work revealed DNA's structure—credit went to Watson and Crick. Lise Meitner, who explained nuclear fission—omitted from the Nobel Prize. Nettie Stevens, who discovered sex chromosomes—received little credit. Cecilia Payne-Gaposchkin, who discovered stars are made of hydrogen—initially dismissed. And countless others whose names had nearly vanished.
Rossiter did not claim to be rediscovering these women. She refers to Franklin and Meitner as having been famously denied credit, in fact! Meitner specifically is "one of the best-known examples of the phenomenon". Stevens she uses as one in the list of examples in the paper, and Payne-Gaposchkin actually just gets a reference at the end that's doesn't even tell you the specific field of scientific study. (To be fair, there may be more about them in her other publications.) This was not about Badass Historian of Science Tells the Establishment What's What. Everybody knew about the concept of female scientists being publicly ignored as collaborators by 1993 — and women's history as a field had been around for 15-20 years. She was not working in a vacuum where nobody else thought that it was important to study these topics until she forced them to see the light.
Please, please, everyone, be on the lookout for bad feminist history written by AI. If you're not with me on the tells and hallucinations here, then at least be on the lookout for bad "feminist" history regardless of the source. If it sounds like it's sensationalizing, it probably is.
Some important updates to a Xitter post I shared uncritically because I'm definitely one of the people who doesn't auto-judge AI based on word choices, punctuation, or text layout (as some other debunkers have.)
Aside: I now judge that Xitter account to be an AI factory based on the sheer volume of "stories" posted every single day on a mind-boggling array of topics. This was the first post of theirs I saw and the reason I followed them.
It's disappointing to me because I might never have heard of Margaret Rossiter without this post, but now I've only gotten the disneyfied account from "mrpitbull": flashy, entertaining, and only loosely based on legend that might be related to facts.
Thank you for correcting that.
