BTW I just realized the XKCD article I posted may suggest I’m replying to stupid comments in a long way. I don’t mean that and instead was intending to poke fun at myself for my ridiculously long comment =) All in good fun.

I meant that PLs can’t mirror NLs in the evolution in that machines, because they do not have social understanding, cannot evolve PL’s in the same way that NL’s evolve. While the code itself may change, the compiler or interpreter only knows one way to parse the code, hence the language does not evolve.

If I as a programmer could decide that I wanted to write a new programming construct that I could then describe to the compiler, and then the compiler could remember that and use that construct in the future with out me having to make a change to the core language, then perhaps PLs would be just like NLs, for better or worse.

As it is, you are correct: the interpretation of the code, by a humane, is a social construct, as it is language interpretation by two or more humans. However, again, for the machine, there is and can only be one way to interpret the code, until it is deliberately updated by a human.

So that is the nuance behind what I was saying, although I didn’t make that very clear in the first post, and hopefully it is more clear now.

Short summary: Natural Language is very fluid; programming languages, due to several factors, are much more rigid.

If this is not what you were trying to say then please, by all means, correct me. Otherwise, here’s why I disagree:

Language is bidirectional. There’s a dichotomy between the ease of expression and the ease of interpretation.

While using NL, humans are constantly tweaking the language to keep equilibrium between expression and interpretation as the greater social interpretation of the language gradually shifts. This is a form of natural language evolution. It’s also paradoxical but we won’t go there =).

Programming languages, at first glance, may seem unidirectional because its used for humans to express ideas to machines. But this is not true because humans are trying to interpret the ideas they are expressing, even while expressing them (reading code as it’s written). Also, it’s common for other humans to read code written by another or even come back and read their own code after forgetting about it. This is clearly interpretation and therefore PLs are also bidirectional, but in a different form.

In both NLs and PLs, humans are doing the expression and interpretation. But with PLs, there is also a third party: machines. Since machines are involved, PLs evolve incrementally because humans have to write new interpreters (here’s another place humans must interpret the PL) for the machines as they find better language tweaks to keep close to the I/O equilibrium.

PLs and NLs mirror each other because they both evolve based the user society. Q.E.D.

http://www.xkcd.com/406/

Perhaps I rambled a bit, but I think my meaning is clear.

Nice article.