Meteorologist, oceanographer, Python solvem probler, experimenter, Raspberry Pi tinkerer, geek dad, manager, space enthusiast, ham radio lover, science columnist, ADHDer, synthetic intellect curious, linux admin cosplayer.

DRAFT CORRECTIONS - Post 19: The Women They Wrote Out of the Photo

Status: COMPLETE

SUMMARY

Post 19 is largely well-researched. All biographical details for the six ENIAC programmers, Adele Goldstine, Margaret Hamilton, Ellen Fetter, Mary Tsingou, and Arianna Rosenbluth are CONFIRMED accurate against Wikipedia, MacTutor, ENIAC Programmers Project, and Lost Women of Science sources.

One section requires significant correction (Klara von Neumann’s role in the 1950 weather forecast). One minor date claim needs nuance (ENIAC unveiling). No corrections needed for any other women profiled.

CRITICAL CORRECTION 1: Klara von Neumann and the 1950 Weather Forecast

Issue

The post states (lines 145-148):

In 1950, Jule Charney, Agnar Fjortoft, and John von Neumann ran the first computer-generated weather forecast on ENIAC – the founding moment of numerical weather prediction… Klara coded it. … She worked for 32 days on the project, managing 100 000 punch cards. A 2021 Smithsonian correction later clarified that her primary contribution was writing and checking the code and training the meteorologists to program the ENIAC – she was not merely a punch-card operator, as earlier accounts had implied. She was the programmer.

Problems

  1. “Klara coded it” is factually unsupported. Historian Thomas Haigh (the author of ENIAC in Action and the Tom and Maria piece cited in footnote 2) explicitly debunked this claim in the Lost Women of Science bonus episode “The Weather Myth.” The 1950 paper’s acknowledgments thank her for “instruction in the technique of coding for the ENIAC and for checking the final code” — not for writing the code.
  2. “32 days” + “100 000 punch cards” belong to the 1948 Monte Carlo work, not the 1950 weather forecast. The 32-day stretch was the April 1948 ENIAC Monte Carlo run at Aberdeen for nuclear weapons simulations. Klara was central to that work.
  3. The 2021 Smithsonian correction went the OPPOSITE direction from how the post frames it. Smithsonian Magazine’s editor’s note explicitly retracted the punch-card claim: “the original piece misstated that Klara was in charge of hand-punching and managing the 100,000 punchcards” — “in fact she wasn’t present for this part of the experiment.” The correction downgraded, not upgraded, her role.
  4. The actual computational operators in 1950 were ballistics lab staff (likely including Hedi Selberg and Margaret Smagorinsky).

Suggested rewrite

REPLACE the paragraph beginning “In 1950, Jule Charney…” (lines 147-149) with:

In 1950, Jule Charney, Agnar Fjortoft, and John von Neumann ran the [first computer-generated weather forecast](/weather/hpc/history/2026/03/29/The-man-who-tamed-the-equations.html) on ENIAC -- the founding moment of numerical weather prediction, the event that this entire blog series has been building toward. By then, Klara was the world's leading expert on ENIAC coding. The published paper thanked "Mrs. K. von Neumann for instruction in the technique of coding for the ENIAC and for checking the final code." Historians have debated the exact extent of her role on the day; what is certain is that no one outside Los Alamos understood ENIAC programming as deeply as she did, and the meteorologists learned to write code for the machine in a paradigm she had helped invent. She trained them, and she reviewed the code that made the first numerical forecast possible.

For her own programming work, look earlier. In April 1948, a team including John and Klara von Neumann and Nick Metropolis spent 32 days straight at Aberdeen running the first computerized Monte Carlo simulations on ENIAC. Klara had diagrammed and coded the Monte Carlo algorithm herself. These were not only the first Monte Carlo calculations by computer -- they were among the first code written in the modern stored-program paradigm ever to be executed. Her team discovered they had been "testing the conveniency of the H-bomb."[^2] The simulations were for thermonuclear weapon feasibility studies.

In the published Charney-Fjortoft-von Neumann paper, she appears only in the acknowledgments. "Mrs. K. von Neumann." Not even her first name.

This rewrite:

  • Keeps the emotional center (uncredited expertise; “Mrs. K. von Neumann”) which is fully accurate.
  • Moves the “32 days” detail to where it belongs (1948 Monte Carlo).
  • Removes the unsupported “Klara coded it” claim for the 1950 forecast.
  • Removes the “100,000 punch cards” claim (which Smithsonian retracted).
  • Removes the misleading reference to the 2021 Smithsonian correction supposedly elevating her — the correction did the opposite.

MINOR CORRECTION 2: ENIAC unveiling date (Feb 14 vs Feb 15, 1946)

Issue

The post says (line 16):

It is February 15, 1946. The U.S. Army is unveiling ENIAC – the Electronic Numerical Integrator and Computer – to the press and the public at the Moore School of Engineering in Philadelphia.

Nuance

Multiple histories distinguish:

  • Feb 14, 1946 (evening): ENIAC press demonstration / announcement to public
  • Feb 15, 1946: formal dedication at University of Pennsylvania

Both dates are commonly cited as “the unveiling.” This is defensible but readers familiar with the precise history may notice.

Optional adjustment

If desired, change “It is February 15, 1946” to “It is February 14-15, 1946” or “It is the evening of February 14, 1946” (the actual press demo with the celebratory dinner). The “dinner” referenced later in the post was held the evening of Feb 14, when the press demonstration occurred, so this would also fix the dinner timing.

Lower priority — the current wording is widely seen in popular histories.

NOTE: Trajectory speed (20 seconds vs 40 hours)

The post states (line 24):

the program demonstrated… wired the machine to perform a ballistics trajectory calculation in 20 seconds – a problem that took a human computer 40 hours to solve by hand.

Wikipedia ENIAC article says “15 seconds” vs “several weeks.” Other sources give variations.

However, the “20 seconds / 40 hours” specific pairing is the Kleiman ENIAC Programmers Project / Proving Ground (2022) version, which is the primary source cited in the post’s footnote [^1]. Since the post is following Kleiman’s account, this is consistent with the cited source. NO CHANGE recommended.

VERIFIED — all other biographical claims

All other facts in the post were checked and are accurate:

  • The six ENIAC programmers (all birth/death dates, places, education, contributions)
  • Adele Goldstine (1920-1964, including children born 1952 and 1959)
  • Klara Dan von Neumann (1911-1963, including 15 lbs wet sand, BAC 0.18%, fourth husband Carl Eckart)
  • Mary Tsingou (b. 1928; first computer graphics with Pasta on MANIAC oscilloscope; 1953 FPUT; 2008 Dauxois renaming; 1972 Tuck collaboration)
  • Margaret Hamilton (Earlham 1958, MIT 1959, hired Fetter 1961, SAGE then Apollo, Presidential Medal Nov 22 2016)
  • Ellen Fetter (Mount Holyoke 1961, LGP-30 replacement to Hamilton, 1963 paper acknowledgment, married John Gille)
  • Arianna Rosenbluth (Houston 1927, Harvard PhD 1949 fifth woman, Van Vleck advisor, fencing championships, Olympics 1944/1948, married Marshall 1951, died COVID Dec 28 2020)
  • Kathy Kleiman as Harvard undergraduate, “Refrigerator Ladies” framing

Wayback Machine posting

NOT posted (per instructions, skip if blocked or if not necessary for the core verification work).

Status: READY FOR REVIEW

Main action item: rewrite the Klara/1950-weather-forecast paragraph as described above. Optional: tweak ENIAC unveiling to “evening of February 14-15, 1946.” Everything else is accurate and well-sourced.