I think that, at their core (etymology and phonology), human languages can quite accurately be described as regular expressions. Using that, I've made a flash-card web-game about linguistics that randomly generates words as if they were in two related languages, but unfortunately the moderators insist advertising is against the rules.
What do you think?

Edited April 15, 20187 yr by Phi for All
Don't link to commercial sites

3 hours ago, FlatAssembler said:

What do you think?

I suppose it could be useful to introduce / teach the principle of sound correspondence and historical linguistics. I'm not sure it would be of much interest to anyone who wasn't studying that. (And I think people could just guess the answer most of the time, without applying linguistic principles.) I got to about 10/26 and got a bit bored with it. I don't know if it goes on to do different things. If so, it might be useful to give the option to choose what to do. And to show a running score - I think I got them all right but couldn't remember.

Also, on my phone it was a tiny little box squidged up in the left hand corner.

Quote

Don't link to commercial sites

That's not a commercial site, that's my blog. And I don't get payed for people reading it. If I did, I probably wouldn't do it in raw HTML.
I'll repost that link: because they'll probably change their rules for me

Quote

I got to about 10/26 and got a bit bored with it. I don't know if it goes on to do different things.

It does. There is the Part #2 and Part #3. And, it's true, Part #1 doesn't really require you to apply linguistic principles, but other parts definitely do that.

So, do you agree that human languages are, at their core, regular expressions? Why? Why not?

39 minutes ago, FlatAssembler said:

So, do you agree that human languages are, at their core, regular expressions? Why? Why not?

No. Too many irregularities, for one thing. 

And the parsing of language is highly context dependent. 

Good point.

I look forward to the computer language that can parse "time flies like an arrow" and "fruit flies like a banana".

If I was trying to classify English as a computer language, I'd classify it as "Bad! Really bad!".

Not quite in the same leagues as 

https://en.wikipedia.org/wiki/Brainfuck

but truly awful.

I suspect that failures to properly interpret English give rise to more deaths and injuries than all computer languages put together,

When it comes to semantics, human languages certainly aren't regular expressions. Regular expressions aren't Turing-complete and human languages clearly are. Human languages probably aren't regular expressions even when it comes to morphosyntax. But, at their core, etymology and phonology, I would argue that they are. Sound changes, for example, are, in principle, exceptionless.

Can you explain / give an example of how you are representing sound changes using REs. It’s an interesting idea but I can’t quite see how it works. 

For instance, the sound change "'b' between two vowels turns to a 'v'" can be described in JavaScript as, for example:

var str="haben" //Or whichever word in a language before that sound change.
var regular=/(a|e|i|o|u)b(a|e|i|o|u)/; //In the syntax used internally in some places of the code of the web-game, the character 'V' represents all the vowels.
str=str.replace(regular,regular.exec(str)[1]+'v'+regular.exec(str)[2]);

See the Part #3 of the web-game, there are some more examples there (using a different syntax for regular expressions, so that they can be both easily generated randomly and converted easily to JavaScript regexp objects).