The examples the author cites of having to write a great deal of code that one does not understand is the motivation for Racket's language levels [0]. However,

The examples the author cites of having to write a great deal of code that one does not understand is the motivation for Racket's language levels [0]. However, it is not just a matter of Java having a syntax that imposes a lot of cognitive load on novices. It is about being able to talk basic programming language features independently from advanced languages features. This is independent of both syntactic and semantic complexity. Python's clean syntax, for example, masks some really complex semantics. Neither Java and Python are poor programming languages, but their overriding design goals are to be languages that proficient programmers can be very productive in. The pervasive mistake in CS education seems to be that these languages are therefore great choices for teaching programming to non-programmers. They are almost certainly not. I really hope that the discussion of languages for CS1 can move beyond this. There are many languages whose overriding design goal is to be phenomenal for introductory computing education. They are designed by computer science educators, and are shaped foremost by the observed successes and difficulties of actual students that are learning to program with them. They provide answers for educators beyond just a syntax and semantics; frequently, they come battle-tested textbooks and teaching materials. [0] https://docs.racket-lang.org/drracket/htdp-langs.html https://docs.racket-lang.org/drracket/htdp-langs.html