> deep tech work in physics

And how long did they have funding for to continue doing this work? Is there some part of "it's expensive to do deep work" that you disagree with?

> Would you say every programming language is fungible? If so, then why are there so many? How many should there be?

This is a complete non-sequitur.

If I want to do numerical computing (to solve a problem that needs it), which languages should I consider? Which should I not? Is any language appropriate?

Which languages do you reach for most often (whatever the reason for doing so)?

I still have no clue what you're getting at but the answer is 100% determined by the community you're a part of - the same PDE can be solved by different communities using Fortran/C++/Matlab/Julia/Python.

You're conveniently not addressing the more salient, central point: deep work costs lots of money.

Edit: but no one is solving PDEs using Haskell lol.

My point actually is that different languages suit different purposes with no statement about the cost to fund anything. While there may be many different paths to achieving a result—in this case numerical computing with any of "Fortran/C++/Matlab/Julia/Python"—certain languages are more suited to certain problem domains than others. You, yourself, stated the most popular languages for numerical computing. Most of those would not be suitable for general web development.

Thus languages are neither fungible nor does it seem from the languages you just mentioned (which most would say the same) that there is "zero causal relationship" between a language and solving problems in the domain of interest. In this particular hypothetical, someone would notice a desire to solve deep problems with lots of funding that use numerical computing and could thus look for the languages you mentioned as an initial filter. Such a person would probably not be looking at a javascript role, for instance, given these aims.

Thank you.

> nor does it seem from the languages you just mentioned

Lololol what exactly are the unifying themes of

Fortran, C++, Matlab, Julia, Python

They're all pairwise as different as can be.

Man it's hilarious how delusional some people are to insist on their flawed point despite the loudest clearest contradictions.