POPL 2022—Et Tu, Brute?
Some things just cannot wait. Men must stand up now for women’s equality— Rick Goings
Kathleen Booth, Cicely Popplewell, Grace Hopper, and Jean Sammet were four pioneers of the field of Programming Languages. They are credited for (co-)creating the first assembly language, programming manual, compiler, and widest-use language (COBOL), the last alongside Mary Hawes and Gertrude Tierney. Popplewell’s name heads the proceedings of the 1962 IFIP conference, while Sammet wrote a defining textbook of the field, Programming Languages: History and Fundamentals, in 1969.
Today Ken and I wonder why the 56-person strong programming committee for the 2022 Principles of Programming Languages conference has only six women.
If you were asked to name a principle of programming languages, there’s a good chance you’d think of the Liskov substitution principle, which is named for Barbara Liskov. Of all fields of computing, we would hope that this would be the leader toward gender balance. It is true that across the board in computer science, we are struggling to get over 20% female representation. But we should be doing better than 
The Full Committee
Here is the full POPL committee. We have starred the six women—whom we salute.
- Aditya V. Thakur
- Alex Kavvos
- Alvin Cheung
- *Amal Ahmed
- Andrew Kennedy
- Anthony W. Lin
- Atsushi Igarashi
- Bas Spitters
- Benjamin Lucien Kaminski
- C.-H. Luke Ong
- *Caterina Urban
- Christos Dimoulas
- Chung-chieh Shan
- David Broman
- Dimitrios Vytiniotis
- Fan Long
- Filip Sieczkowski
- Francois Pottier
- Garrett Morris
- *Hongjin Liang
- Hongseok Yang
- James Wilcox
- Jan Hoffmann
- John Wickerson
- Justin Hsu
- Ken McMillan
- Kihong Heo
- Kohei Suenaga
- *Koko Muroya
- Leonidas Lampropoulos
- Matthew Hague
- Matthew Parkinson
- Michael Carbin
- Michael D. Adams
- Michael Emmi
- Neel Krishnaswami
- Nicolas Wu
- *Niki Vazou
- Ori Lahav
- P. Madhusudan
- Pavel Panchekha
- Philipp Ruemmer
- Pierre Clairambault
- Qirun Zhang
- Radu Grigore
- Ralf Hinze
- Ranjit Jhala
- Robert Atkey
- Ronald Garcia
- Ronghui Gu
- Sam Staton
- *Sandrine Blazy
- Shin-ya Katsumata
- Tachio Terauchi
- Wei-Ngan Chin
- Woosuk Lee
I Cannot Believe It
I have always enjoyed the POPL conference. I recall being there ages ago—I was there in 1980 on a paper, “Theoretical and Empirical Studies on Using Program Mutation to Test the Functional Correctness of Programs,” with Timothy Budd, Richard DeMillo, and Frederick Sayward.
I had planned to do a long post on POPL 2022. This POPL is planned to be a real in-person conference for this coming 2022—not virtual. But I was upset to say the least by the above ratio.
Ken and I were similarly stopped the week before Halloween. We had intended a post saluting those who aggregate theory of computing blogs and including a list of over 50 active blogs. We stopped it because, as noted at the end of the post that Ken wrote instead, one has to stretch to reach even 4 mathematics/theory blogs by women.
Counts of the most recent FOCS and STOC program committees give slightly better percentages. Slightly. For STOC 2022 we count 7 women out of 46, giving This is
This is just a notch better. FOCS 2021 (meeting in February 2022) manages to improve both the numerator and the denominator:
That’s double POPL, and over 20%. But it is still single-digits on a fairly large committee.
Planting a Field
We’ve led off this post with pioneers of generations back. What about 20–30 years back? That seems to be the time frame alluded to in this statement by Valerie King—whom both of us have known since the late 1980s—in the first-year report to NSF of the inaugural TCS Women meeting in 2018:
“I chaired the STOC program committee last year [2017], and while there were plenty of women on the committee, what a disappointment it was [in 2017] to see rooms full of men with only a sprinkling of woman in sight. I remember a time when there were more women at FOCS and STOC and It’s hard not to wonder what’s happened, and to want to fix this problem. I think it can lead to a downward spiral, the fewer the women, the more women who do come fill uncomfortable and the fewer that come next time.
Thanks to efforts of Barna [Saha], Sofya [Raskhodnikova] and Virginia [Vassilevska Williams], STOC 2018 seemed different. As I looked around the conference rooms, I actually saw women and it felt good.”
So we wondered if we could take a snapshot of POPL 20 years ago. I (Ken writing this part) counted out the “Language People” list that is linked from Wikipedia’s own shorter list of programming language researchers. The longer, former list was compiled circa 2000 by Mark Leone, whom I knew as a student in a theory course I taught once in Cornell’s summer session, and who now works at NVIDIA.
I count 319 names. I noticed Susan Horwitz, for whom we wrote a memorial in 2014. I did not check whether others are still living, as I viewed this as a generation-ago snapshot. The list is not perfect—it CLU-lessly omits Liskov—but it serves the purpose. I included Horwitz in my quick count of women:
That makes a ratio of 
Open Problems
Even with Ken’s little experiment, I cannot understand how the low ratio could be possible with POPL. Here are the goals for POPL. Gender equity is in the list, but is seventh. Hmmm.
[Fixed count of POPL committee; “wide-use”->”widest-use” in intro, FOCS 2022 8/36 –> FOCS 2021 8/35]
You are somewhat overstating some of the achievements here:
Kathleen Booth wrote the first assembler for the particular computer installed at her University.
Cicely Popplewell probably deserves to be better known for doing a lot more than writing a manual.
Grace Hopper was the real deal, but Cobol was one of the first (people can argue about whether Fortran came first).
Jean Sammet wrote a not particularly good book in a new area, which has become the default citation in the field.
If we are signalling our virtue, what about all the non-Americans who were doing interesting things at the dawn of electronic computing?
Some of your points understood: that’s why the intro has “(co-)”, meant to distinguish the scope of FORTRAN and COBOL, and made sure to put in something more about Popplewell (British like the still-living Booth).
The authors of this post have graduated 18 PhD students according to
https://mathgenealogy.org/id.php?id=69524
https://mathgenealogy.org/id.php?id=172360
How many of them are women? I can only count 2/18, or 11.1%, which is exactly the same percentage as the POPL PC.
Dear Sam G:
You are correct on percent. But I have done that over 40 years. So it says that the percent has not changed for a long time…
dick
The actual PC count is 56 not 54, the following are two people each.
Thanks—fixed.
Thanks. Funny—I fixed the decimal and thought I’d done that.
The problem is that when you try to improve these ratios as much as possible, it means that the (too few) women in CS will have a big chunk of their research time taken out to be representative in PCs, thesis commitee, etc. It is a big disservice to them, and can make them feel their main job is not to do research, but to appear on various committees for quotas. At least the women I talked to are tired of being constantly asked this kind of things, and must say no to a majority of them, if they want to have time for research. So unfortunately, a solution to this will have to take a long time, and address the problem at every step of the career (probably starting at early childhood), we cannot just magically fix it in academia if the unbalance has only been created earlier.
Agreed. Indeed, a remark of something like “we can’t have (the) 17 women serving on the PC every year, else they could never submit to the conference” was left on the cutting-room floor (along with saying Liskov was the first female CS PhD and references to some of our earlier posts on promoting gender equity—the mindspace of the title includes Dick and me among the “honourable men” of the Julius Caesar allusion).
Could it simply be that this field isn’t equally attractive to both genders, and the necessary talent to advance to the highest levels in it is also not even distributed between men and women?
A gender disparity- or any other significant gap between the overall population’s demographics and that of a specific subset (like CS professors)- is not ipso facto a problem, and the null hypothesis should NOT be that any such disparity is the product of some kind of malign action or institutional flaw. In other words, you need to prove that there is actually discrimination against women or something truly broken about the talent pipeline for this field before any “diversity and inclusion” changes are made. Because in practice letting in the “equity” zealots has NEVER improved any academic field. Just ask your colleagues in the humanities how that worked out. Such a ‘cure’ is far worse than most diseases, let alone a situation where its far from clear you’re actually sick.
As for why there might be an organic decline in female computer scientists, there’s perfectly reasonable hypotheses for that that require no malevolence or failure from anybody. Namely, after the 1970s modern computing became a major engineering field just like civil or electrical engineering. And now, 2 generations later, its demographics are similar to many of those other fields of engineering. If there is any real problem here, it is the gender disparity that appears to be common throughout engineering and the hard sciences, not just CS.
There is a lot of research and experiments indicating the presence of strong bias against women, unrelated to their skill, see e.g. https://www.pnas.org/content/109/41/16474 together with its bibliograhy and citing articles.
Cribbing. I bet the amount of African Americans is less than 5% and the amount of Africans is less than the digits on a cows hoover in absolute number. I bet it is systemic bias within the community and less morale among these groups and a negative feedback. It it like we know 90% of phds have no use at all even within the community let alone to the world. Why not give equal opportunity to everyone to waste their time with earnest so some random minority with real talent has a chance to really succeed.