My Response to Coraline Ehmke's Attack on Meritocracy, Nine Years Later

This is a response to Ehmke's article "The Dehumanizing Myth of the Meritocracy". While the article technically isn't directed to any individual, the social impact of similar thoughts is obviously not to be ignored. I shall condense my opinion as much as possible.

Update(2024.7.31): I thought of something while writing a reply to a reply to this blog post (said reply) (archived). I now will summarize my point into the following list:

 

As we can see, the topic of the article is obviously of the ethics of software development process; note that we are not discussing the ethics of software per se which, despite what some people might say, is a very different topic. When we discuss the ethics of software development process, we must admit that software quality is one of the core virtues and not to be lightly ignored or to be regarded as less important than other desired aspects; and when we ned to discuss about software quality, the first thing we must discuss is that how much of the criteria for a piece of software to be considered to be good is related to the objective a-priori truth of the world (e.g. claims of security proven through mathematics) and how much of the criteria is related to how people treat each other; my opinion, is that the latter half is of insignificant amount, if not straight-up non-existent.

One of the biggest lies people like Coraline Ehmke has lied to the modern software industry is that purely-societal problems are somehow something that can and should be addressed and dealt with by software developers within the range of their own individual software projects in no other way than integrating purely-societal power dynamics into their development process. Considering what Ehmke and their peers-in-spirit did and are doing to the current open-source software scene, it's quite ironic when Ehmke asked:

…why is it that we can proudly refuse to use software created by corporations whose often aggressive business policies we disagree with, but continue to adopt software written by sexists, racists, homophobes, transphobes? What makes these people immune?

One can always refuse to use softwares that are developed by the people one disagrees with as long as one is willing to accept the implications and consequences, and people have been making forks of softwares due to different opinions on technical decisions for as long as this industry has existed. One could try one's best not to out-right claim this was (one of) the pre-move Ehmke and their peers-in-spirit made to justify their pursuit in power (because of what use is walking down the same path of accusing people like how they did, which we all have seen have lead to so much disarray?), but it is hard to prevent people from thinking that way when the things they did effectively lead to the same result. Yet Ehmke went on and said:

It’s the idea that our humanity doesn’t matter as much as the quality of our intellectual output.

One could obviously bring out the argument that software development is far from being everything about one's existence and any normal person would have agreed, but that argument does not stand well on its own, so I would like to point out an observation here: when someone is only having interactions with you that are strictly within the context of software development, all the judgements he make is strictly limited to the space that all the possible behaviours within the space of software development can imply; but since all interactions are strictly limited to the space of software development as well, this would not be a problem. This is the real truth behind the old-school notion of meritocracy, unlike the grim scene Ehmke and their peers-in-spirit like to imply where every single one would be asked of their gender or other identity and then be discriminated with it.

One must realize that the "pure meritocracy" in the context of software development, the one principle people have been upholding since the very beginning, is not that "we judge everything about you by how good you can be at developing software" but rather that "it's only because we are developing software that we judge everything about you by how good you can be at developing software". Any argument that deliberately ignore the first part is at best invalid and at worst malicious and full of ulterior motives, no matter what words one may use to dress up both the one who's making the argument and the argument itself.

 

I suppose, that due to our discussion space being constantly bombarded with moral judgements and all from these individuals since the very beginning, people have already been forgetting that being a part of the "marginalized people" group never directly implies you (and any aspects about you) are superior in any other way than the people that's not in the group. Ehmke addressed about how the contribution from "bad actors" are held highly while other contributions from "good actors" and "just as important as other contributions" (important in Ehmke's book, of course) are not treated the same:

It seems that ignoring the worst of our heroes is easy, but should the opposite also hold true? Should we ignore the positive, community-oriented contributions of others as quickly as we dismiss some people’s negative attributes? Are the contributions of bad actors really superior to those who bring humane, non-code contributions to our corner of the world?

To do so underlines the dehumanizing effect of purely meritocratic thinking. One product of this sort of thinking is the division of our skills into “hard” and “soft” categories. Almost every technical conference draws a line based on this false dichotomy. Not surprisingly, marginalized people deliver most of the “soft” talks, while cisgender, white, heterosexual men deliver the more technical, “hard” talks. Value judgements are ascribed to each category of presentation: technical talks are recognized as immediately relevant, while talks addressing the humanity of our profession are relegated to interesting thought exercises or simply as useful breaks to be taken between the “real” talks.

Obviously the first question one should ask after reading this is that how much this "humane", "positive" and "community-oriented" contribution (I'll spare them by not surrounding the word "contribution" with double-quotes) is actually needed, and the second question should be how is deciding that non-technical talks are not as relevant as technical talks an issue. Any technical development could be a solution to a problem the people who would attend these technical conference would be having, of course they are more likely to be considered immediately relevant, it's only logical for it to be that way. Of course, one could also make the cynical suggestion that the reason why Ehmke argued for "non-technical contributions" is that they themselves were not good enough to make enough actual technical contributions to put them into higher positions within the community, but that's near-conspiracy level which I would like not to discuss further here.

Back

Last update: 2024.7.31