This paper is awesome, and so unexpectedly so given its obscure title. A more descriptive summary might be, “how hackers and scientists contribute differently to the advancement of software, and why the tendency of academics to shut engineering out of its work is leading to some strange results.”
The basic premise is that programming language research, before the 90’s, was about discovering the designs of programming systems – that is, systems you could actually use, like new languages. But now, it is becoming about programming language design as an abstract “field in itself”. Thus the move toward the abstract (type systems and such) rather than the practical (modularity and reuse). It is written by Richard Gabriel, who is perhaps best known for writing that famous “Worse is Better” paper.
I think about this paper probably once every two weeks; the incommensurability aspect ala Kuhn creeps into many of our different social systems in the computer field. A particularly glaring example is if you review the Journal of Software: Practice and Experience from the late 60s into the 90s. At some tipping point in the early 90s, late 80s, things suddenly spiral into the realm of esoterica and gobblygook and - last I checked - it had never gotten back to actual practice.