Not an advocate, but isn't a core idea of feminism, that claims, that men think technically, in terms of power, while women think about social stuff, child-care, etc., are just roles and made up?
How can than be claimed under the banner of feminism, that something isn't feminine (e.g. not social, to much technical) enough? Isn't that the "misogynous" kind of thinking, that feminism is supposed to fight?
(Edit: Anyone cares to elaborate why this was flagged?)
> Let’s consider a definition in a simple grammar: digit : 0..9. [...] In Arabic, digits are not 0-9 but 0..9.
Aren't those just different glyphs? This can be even seen by the fact that HN converts them both to the same symbols. The mapping from characters to glyphs isn't defined by a programming language, it is defined by a font.
> HTML, which is executable however, is not seen as a programming language in regular PL discourse
HTML is executable?
> UML is not executable, so could easily be disregarded as a programming language, but is not
Who claims, that UML is a programming language? Is there a program that is written in "UML"? It is more a language in the same sense as a "design language" of a building.
Had a chat with Claude over this paper and it did ultimately help me understand it better. So in case it helps others, I'm going to share the bits I think are important from that. (Just wanted to disclose my use of AI)
To start with, this is a paper on programming language research first and programming languages themselves second (by way of how that research influences them).
I think some readers might be put off by the feminist framing and miss that these are ideas many on HN have expressed before.
For example, the paper makes the point that PL research tends towards mathematical formalism over human experience. This sounds like an odd thing to say at first, but consider the well established meme on HN: "What is a monad? Monads are monoids in the category of endofunctors" as though that were a helpful answer. That meme is saying the same thing. That focusing exclusively on the mathematically pure concept is actually an obstacle to understanding which limits the usefulness in practice of PLs, independent of a PL's capabilities.
It's an interesting framing to be sure, and perhaps the observed values and behaviors of the PL research community is deeply influenced by a masculine framing. However, I do wonder how much of it at least originated from the domain itself.
The hardware limitations of early computing made strict formalism essential - unless you got your formalism correct, your intent couldn't be mapped to execution on silicon. Mathematical rigor wasn't just preferred, it was often the only path to creating anything that actually worked. Even today those formalisms remain the fundamental building blocks that enable the higher levels of expressiveness we enjoy.
Fields of research, science in general really, have momentum. Strict formalism and a willingness to deal with, or even value, high complexity was what it once took to make meaningful headway in the field of programming languages. Today, perhaps, that is less true, but those that succeeded in the past will continue to shape the field for some time.
I do think this is ultimately a problem for the field. There are aspects of PL research and design that that certainly do require formalism and are inherently complex in a way that requires a certain approach. But that doesn't mean the field needs to remain forever closed to discussions and research around how the decisions that are made impact the world at large and the people in it.
> For example, western PL designers might disregard cultural aspects of non-Western languages, like non-English numerals
There is a reason why even in native Arabic¹ or Thai² communication Western "Arabic" numerals are used ... deliberately.
Also I didn't find any suggestions or criticism actually regarding the _design_ of programming languages (from a feminist perspective); after all the title suggests that this is what the paper is about. There is only a single irrelevant reference to Python - I mean, practically thinking: what should be changed about Python to make it more "feminist"?
Java, JavaScript, Rust, C++ aren't mentioned even once ... wouldn't it make sense to make the point clear by providing concrete examples of what is not satisfying about the status quo? Or is it maybe that there is no point to the question to begin with?
1: https://www.aljazeera.net/ebusiness/2025/6/5/countries-large...
Stopped reading as soon as I saw a positive reference to "Glaciers, gender and science", one of the most batshit insane things I have ever read.
"For as long as I1 can remember, I have been in love with pro- gramming. As a pre-teen, I excitedly taught myself program- ming with BASIC books from the local library, and as a late teenager, I joined my high school’s computer club, and my high school capstone project was a rudimentary machine learning algorithm. In university, I chose compiler courses as electives, and my final MSc project included the develop- ment of a small programming language for SAT solving. It seemed a logical next step, when given the opportunity, to pursue a PhD in the direction of programming languages (PL). I am, for all intents and purposes, a very normal pro- gramming languages person, with one notable exception: I am female."
but then:
"Feminism Can Help Question Values and Priorities of Programming Languages. As I was trying to under- stand my mismatch with the PL community, I started to see that the current state of the art in PL research and de- sign prioritizes machines over people, there is a focus on programming languages that are easy to read for the ma- chine, rather than easy for people. The current standards of evaluation in the PL community are set from a masculine perspective as well, valuing formalism and formal methods over user studies, quantitative over qualitative work, and the examining of technical aspects over context and people"
So they loved something as a kid growing up but now think it was dehumanizing ("prioritizing machines over people")?
"Feminism in PL introduces interesting and exciting challenges that will require the deliberate work of examining different perspectives; seeking out users from a variety of different backgrounds and identities, and carefully investigating our programming languages and systems from diverse perspec- tives."
I don't know that you can rely on feminism to drive inclusion of diverse perspectives at large. There are many backgrounds and identities that I do not expect feminism to bring to the fore for consideration and inclusion in PL design. I might be wrong on this, so would love a concrete example that draws that throughline.
> Error messages are another PL design feature shaped by peoples’ standpoint, by their lived experiences and values. For many programming language designers, error messages are an afterthought. Other aspects of design will likely over- power the careful crafting of error messages, such as an op- timized compiler or an elegant syntax. Many language de- signers, explicitly or implicitly, take the stance that ‘users will just have to deal with error messages when they oc- cur’. Error messages in many languages explicitly do not include the user in their phrasing, saying things like ‘miss- ing bracket’ or ‘unknown variable’, which leaves the user out of the equation. For a person with the lived experience of being told they don’t belong in programming and being afraid of failure in the programming world (a common expe- rience for women in computing), unclear or unhelpful error messages are more likely to have negative impacts. More readable error messages are not just helpful for those with less prior knowledge, clear error messages are a better ex- perience and will help all users—even professional develop- ers. A recent study on error messages in Elm showed that their well-phrased error messages were the most-named positive experience of professionals working with the language [citation].
The author must consider the audience is either stupid or childish. An intelligent adult deserves respect and should not be talked about in this way, implying that they are emotionally unequipped to handle a straightforward computer error message.
“ Our goal is to illustrate opportunities for more inclusive practices that will introduce greater diversity to the design of programming languages and the demographic makeup of the programming language community.”
Why would I care at all who wrote a programming language, and if the people writing them are diverse?
I care about the language features and nothing else.
Instead of writing that more programming languages should have a female perspective why doesn't the author simply write a programming language?
Edit - apparently I can't read (bad habit of skimming long articles and reading the body of the more interesting headers). She did make one - designed to teach children Python. I still think it would be more interesting to write a general purpose language with the values in the essay.
Guys, the author presents an overall reasonable argument and I think it's more useful to engage with it in good faith than going "so it's all my fault just because I'm a man?" - no one's implying that.
At its simplest, the point is that much of programming language design is done with a masculine perspective that values technical excellence and very little feminine perspective that focuses more on social impact. Most, including myself, have a knee-jerk reaction to dismiss this argument since at first glance it appears to trade off something known useful for something that's usually little else than a buzzword, but upon further reflection the argument is sensible.
The theme of forsaking technological perfectionism in favor of reaching whatever end goal you have set is widely circulated on this forum and generally agreed with. Those of us that work as software engineers know that impact of your work is always valued more than the implementation or technical details. It's thus reasonable that when building programming languages, the needs and experience of the users should be considered. Not override everything else, but be factored into the equation.
I know if I were to write a programming language I'd probably focus on pushing the boundaries of what's technologically possible, because I find it fun and interesting. But I would have to agree that even if I did succeed in doing so, the actual impact of my work would probably be lower than that of Hedy - the author's language. Hedy is not novel technologically, but the fact that it makes it meaningfully easier to learn programming for significant numbers of people is real, undeniable impact.
Lastly, I want to note that the author's argument for underrepresentation of women in PL cannot be reduced to "those nasty men are keeping us out". Humans are tribal and any group of humans is bound to form complex social structures. Those are going to affect different people in different ways, linked paper investigates the effect on those structures on specifically women because the topic is close to the author. Whether you care about low numbers of women in PL design or not, the dynamics that have led to that being the case are worth investigating and are quite interesting on their own.