"Lisp is worth learning for the profound enlightenment experience you will have when you finally get it; that experience will make you a better programmer for the rest of your days, even if you never actually use Lisp itself a lot."

Rhetoric about Lisp is often couched in the language of the arduous journey of a novice monk to Nirvana. (Perhaps Lisp advocates are the bodhisattvas in this narrative.)

I don't think this sort of quasi-religious approach is a good way of evaluating programming languages. I would prefer a more scientific, even quantitative approach.

Reviewing the packages available on common-lisp.net, one notices that most of them, in fact the vast majority of them have not thrived. Quite a few have been sitting idle in permanent alpha stage since I graduated high school.

When it comes to claims about Lisp's superior expressive power (as claimed here for example: http://www.paulgraham.com/power.html), quantitative data are also lacking. The only hard metrics on the subject that I can find are from the Computer Language Benchmarks Game. Incidentally, these measure the length of code in gzip bytes, which seems like a more fundamental consideration than LOC.

How do they compare Lisp to comparable languages in this regard? Not favorably. Here I'm using SBCL, as the default referent of Lisp is usually Common Lisp, but you'll find broadly similar results with other Lisp implementations, like Racket.

SBCL vs Python 3:

http://shootout.alioth.debian.org/u32/benchmark.php?test=all&lang=sbcl&lang2=python3

SBCL vs Ruby:

http://shootout.alioth.debian.org/u32/benchmark.php?test=all&lang=sbcl&lang2=yarv

SBCL vs OCaml:

http://shootout.alioth.debian.org/u32/benchmark.php?test=all&lang=sbcl&lang2=ocaml

And, indeed, SBCL vs Java:

http://shootout.alioth.debian.org/u32/benchmark.php?test=all&lang=sbcl&lang2=java

Here, Lisp appears to be a teensy bit more verbose on average than Java, which brought unto the world such marvels as the anonymous inner class.

Are there any more objective (as opposed to anecdotal) measures for the superior efficacy of Lisp?