Most of these languages are neither dead nor dying. They are just niche.

I just started learning APL this year. It is still actively developed and no you don't need a special keyboard. Plus we have unicode now which makes the special symbols a non-issue these days.

Same with most of the other languages. There is Pharo and Squeak for Smalltalk, the pascal community has free pascal. Sure those communities might not be huge but they are not in acute danger of vanishing any time soon.

And COBOL is still carrying the economy.

Now some languages like ALGOL might actually be dying, sure.

Natural languages are considered dead when they lose their last native speaker. Similarly when the last person being able to use a programming languages dies, we can consider that language dead.

Which means that we have lots of languages that died at childbirth but once a programing language has managed to go over a certain popularity threshold, it is very hard to kill.

Languages don't need to win any popularity contests to be alive. Language maximalism in the sense that you need to be one of the most popular languages or you are considered a failure and dead is just silly.

> Natural languages are considered dead when they lose their last native speaker. Similarly when the last person being able to use a programming languages dies, we can consider that language dead.

I don't think this is the right comparison. A native speaker would be more like someone who learned the language as their first or maybe second language, rather than someone who can use it at all. And by that metric, these languages are pretty much dead/dying, since they mostly have no new learners who aren't into PL history.

They do have have new learners, that was my whole point. Whether they learn out of historical interest, to become better programmer in general, for a job, for research, or because they need it for a specific project does not matter. (And yes all those reasons apply.)

Take a look at companies using APL: https://github.com/interregna/arraylanguage-companies

Or look how many people use it to solve Advent of Code.

As for learning them as a first language, if we applied that criteria then most programming languages would be born absolutely dead and stay there. I don't think anyone ever learned Elm or Purescript as their first language, are they dead?

I really don't get why people make such weird claims, declaring healthy and obviously alive communities to be dead. Again, things don't need to popular to be alive.