My point was that this article is specifically using the term engineering in the sense of computer science. Using engineering as a whole is misleading because as I have demonstrated, software engineering had a historically larger amount of women which faced a decline post-80s. It's burying the lede to make a point about something as a whole rather than the specifics of the article.
Also, your point about those language features being harder is also, objectively wrong. All of those features you mentioned existed before the 80s and were used at the time. If anything software engineering has gotten easier to understand over time, not harder thanks to more robust tooling, debugging and support. Back then you didn't have access to easy resources and had to do everything yourself, so you would have to provide a better explanation than 'things are harder now than they were before'.
Also, your point about those language features being harder is also, objectively wrong. All of those features you mentioned existed before the 80s and were used at the time.
Existed, yes. Widely used? That was a period of transition!
The period from the early 1980s to the late 1980s coincided with a host of changes. Including the rise of personal PCs and Unix, less VMS and mainframes, a decline in COBOL, a rise in C and Pascal, etc, etc, etc.
In 1980 it was reasonable to take introductory programming courses in COBOL (the most commonly used language in business) or FORTRAN (the most commonly used language in science). At the time neither language supported recursion, closures, or object oriented programming. Pointers existed, but were not as heavily used as they were later. That cohort graduated 37.1% women in 1984.
If you were taking your introductory programming course in 1986 on a personal computer in C or Pascal, you would have dealt with both recursion and pointers heavily. That was about when I took that course. This cohort graduated 29.9% women.
Object oriented programming didn't break into the mainstream until the 1990s. (The original Code Complete written in the 1980s doesn't even mention it.) It existed, but was not widely adopted "by industry".
Closures began to spread out from Scheme but remained a fairly niche concept until people got serious about them in JavaScript in the 2000s.
So the timeline is not evidence against my suspicion.
Of course so many other things were in flux that it isn't evidence for either.
>software engineering had a historically larger amount of women which faced a decline post-80s.
How much resemblance does software engineering pre-80's and post-80's have with each other? Furthermore, how much generalizations can one make when the sample size is small? Presumably when computer programming first became a distinct activity, the pool of players were significantly smaller than it currently is.
Also, your point about those language features being harder is also, objectively wrong. All of those features you mentioned existed before the 80s and were used at the time. If anything software engineering has gotten easier to understand over time, not harder thanks to more robust tooling, debugging and support. Back then you didn't have access to easy resources and had to do everything yourself, so you would have to provide a better explanation than 'things are harder now than they were before'.