Haskell: there are too many experimental pure lazy functional languages.

Scala: why should you have to choose between functional and object-oriented programming?

F#: OCaml has no libraries.

Erlang: C doesn't scale.

Arc: Common Lisp is a kludge, and Scheme has no libraries.

This seems to address the reasons the languages were created more than the reasons people continue to use them today. The former is insightful, but the latter might be more useful.

>...and Scheme has no libraries.

That's a strange reason to write a brand-new language that has even fewer libraries. If that's the problem with Scheme. I hope PG and JAR had better reasons than that.

jar is working on Arc?

No, but jar was involved with T, which had the same reason according to pg.

This is interesting because it's a good way to look at the domain of language not in its power or application but its audience.

Haskell: Because programming isn't enough like math

Mozart: Because Prolog is pretty foreign

Lisp: Because parens are really underappreciated ; )

My favorite is "Lisp: Turing Machines are an awkward way to describe computation", especially given that McCarthy originally only intended Lisp to be used as a theoretical device, and not for actual programming! (so goes the story, anyway)

I wonder if that is true or a myth? I wouldn't be surprised if it was true.

Perl: C isn't as good with text.

PHP: I'd rather write my own Perl.

Visual Basic: Crap, we desperately need a way to inflate the number of apps available for Windows.

The same is true for most companies.

PHP: Doesn't [insert other language] have that function? Let's add it, too!

Original JavaScript: HTML can't validate forms

Modern JavaScript: we are the best language committee in the world

Perl wasn't the motivation behind creating Python.

will there ever be one language to rule them all?

maybe Arc 2...

Hopefully the version digits of Arc will converge to e.