> - Doing academic research on "non-employable" languages decreases the impact of your research and your own employability
That is not a metric for choice of language for academic research at all. The academic research itself is a job, paid with public money, and you better not hinder it with your career plans. Most computer science research is not on languages itself, so you need a language expressive enough to allow you study whatever computer science concept you need to study; it's mostly C and Haskell around me, but I can't imagine my friends going into embedded C development - the C they write is very different, as are the practices.
I was referring to academic research in software engineering or PL related areas.
I think your view of academic research is nice but perhaps a bit naive. Academic research in most countries is filled with pressure from private money to achieve valuable outputs to feed the private sector.
> Academic research in most countries is filled with pressure from private money to achieve valuable outputs to feed the private sector.
This is not my experience working in PL research, personally. Mostly government grants here (US) that are either wholly research-oriented ("Build a cool thing because it improves the research community") or government-oriented ("Develop this technique because it will provide us with better (cybernetic) defenses). Admittedly I'm pretty new, but the funding does not seem generally to come from or be focused on developments for the private sector.
I imagine it would be different in compilers, though, for example. That's certainly an area that the private sector has a more vested interest in, due to direct applicability.
In US universities in machine learning fields, a lot of labs work with private companies or do interships during summer at private companies and the grad students eventually graduate and get a job at the private company. Software engineering research might be similar but not sure.
That is not a metric for choice of language for academic research at all. The academic research itself is a job, paid with public money, and you better not hinder it with your career plans. Most computer science research is not on languages itself, so you need a language expressive enough to allow you study whatever computer science concept you need to study; it's mostly C and Haskell around me, but I can't imagine my friends going into embedded C development - the C they write is very different, as are the practices.