Lambda the Ultimate

But I can't locate the specific thread of which you are referring. These matters seem to be mostly opinions. I do know that as I learn more programming languages, it becomes easier to learn more programming languages.

Off Topic: So in the interest of Science, I make the motion that we do a collaboration with the Psyc departments that study the various regions of the brain. Let's find out once and for all whether people who use Dynamic PL's think differently than those who use Static PL's. Should be easy enough to determine which part of the brain they are exercising when doing a programming task in their particular language of choice.

If sample size presents a financial barrier, then I vote we strap Oleg into a chair and ask him a series of questions on Scheme and then Haskell. Even if we don't find a difference in the thought patterns between the two languages/paradigms, at least we'll have a bunch of interesting answers for scheme and haskell. :-)

if polylingual programmers who are polylingual in human languages can learn programming languages quicker.

I will guess that a natural language polyglot has no advantage when first learning to program, because natural languages are of a fundamentally different nature computer languages. A natural-polyglot who is also a computer-polyglot may have a small advantage.

If one knows (say) C++, Perl, and Java reasonably well, one is unlikely to have much difficulty with any imperative language. However, if you are unable to let go of that mindset, learning Haskell can be more difficult for you than somebody who has no programming experience, but has a modicum of mathematical sophistication.

I suspect the same thing is true of natural languages. If you know French, Latin, and German, you shouldn't have much difficultly learning most any European language. The mindset is largely the same, and there are many similarites in phenomes and vocabulary. Learning to speak Chinese or Hindi would be a great deal more difficult. Could knowning european languages possibly be a detriment?

Then you have languages that are radically different. African Languages comprising of clicks are well known. There is an endangered language in India whose 18 phonetic sounds are all different trills of the tongue. Learning to listen to these languages is a challenge for the most skilled linguists, and I imagine learning to speak them is nearly impossible for a non-native speaker.

They're (mostly) unambiguous, they come with tools that tell you when you are talking nonsense, and the number of lexemes in even the most complex programming languages is seldom over 100. Plus, the compiler/REPL will wait patiently for you to finish your sentence.

Natural languages, OTOH, have vocabularies consisting of thousands of words; and that's for basic literacy. They are inherently ambiguous, and they frequently depend on a lifetime's worth of acquired cultural context. They are almost always irregular. There are often two forms to learn (spoken and written); fluency in one implies nothing about fluency in the other. And when one utters something in a foreign language, one may encounter native speakers who are either impatient or downright hostile. Some written languages use scripts which are mutually incompatible (English vs Arabic vs Chinese); and many use tones which are difficult for non-native tongues to pronounce and non-native ears to distinguish.

Natural languages do have one advantage: much greater redundancy; it is frequently possible to make gross errors in speech or writing and still be understood correctly. Errors in programming languages usually either fail to compile (if you're lucky) or produce incorrect programs (if you're not).