We've heard over and over again that perl is like line noise - especially by people who don't program perl and are not used to seeing its code.
I think it's time to take a scientific approach to this problem, and figure out how "line-noisy" perl is. I believe that by comparing character frequencies in perl files versus C/java/plain English/etc we can ascertain once and for all which language looks more like English, which one more like line noise, etc.
The only problem I see with this is, just what exactly is line noise? Well, for perfect line noise, I would expect any character to be as likely as any other at any given time; so I think that by using that definition, we can easily compare.
I'm not going to do it right now because I'm awfully lazy, but I will soon. I'm more than ready to hear comments and criticisms on my approach, and on which code to use.