Despite the articles several flaws — not the least of which is the fact that article seems to be vaguely insulting to its intended audience — it brings up a very interesting point: programming languages and human languages are (surprise!) very similar! They both, at the core, follow strictly logical rule sets made up of syntax and semiotics. Rule sets (grammar) for different languages have points of similarity which can be useful for using one language as the basis of learning another.
Most of the rest is learning the words.
Note that I said at the core languages are similar. Colloquialism can be equated to non-standards-compliance (or genetic drift), and can usually be reverse-engineered to find the original rules from which they sprang.
As an example of how this kind of thing works, try reverse-engineering some “English” words in order to learn a little Latin, Greek, or Anglo-Saxon: economy ecology psychiatry psychology absent abate neologism physiology physiognomy
Perhaps I will explore the interesting interplay between -logos and -nomos and what they imply about the way the words we use have drifted from their original meanings.