Peter Seibel of Practical Common Lisp fame gave a talk to some group or other about syntax and whether it matters. His answer is (of course) that it does and it doesn't. You can watch videos of the talk if you'd like. I found it very interesting.
Obviously I'm on a Lisp kick lately. I tend to get really excited about new languages I learn. My first infatuation was Perl. Then Ruby, which is a better Perl than Perl in many ways. Then Lisp, which may be a better Ruby than Ruby in certain ways that matter.
I heard said once that a sign of intelligence is being able to entertain an idea without accepting it. I think I like to take it one step further. I think a sign of intelligence is being willing to try on an idea like a pair of shoes, and walk around in it for a while to see how it feels.
Clearly some ideas (like some shoes) are so filthy and nasty that I won't go near them. But other ideas seem bad to me only because of my habits and prejudices. It's a fundamental fact of the universe that I'm wrong in some of my beliefs (probably many of them) at the moment. Well not really a fundamental fact, more like a statistical likelihood so great that it may as well be a fact. The problem is that everything I believe seems right to me (otherwise I wouldn't believe it).
In spite of this problem, the only way to survive in the world is for me to keep acting as though what I believe is right, until it's shown to be otherwise. Luckily this attitude acts as a good test of whether I'm right or not, as long as I'm always open to noticing the things that point out that I'm wrong.
But on top of that, I like to throw some random mutation into the mix. I like to take things I strongly believe and test them by trying something very different. I like to put my tried-and-true Vim aside and try Emacs for a while. I like to turn away from my comfortable Ruby and give Lisp a go. This is inevitably painful. But it's a nice kind of painful.
Sometimes it works out and I change my mind, which is great. Sometimes I don't change my mind, which is also great. I was very heavy into C++ for a while; now not so much. But at least now I have good reasons not to like it. Same with Java. At this point I'm not sure I'm ever going back there, but who knows. I may. Any theory worth testing is worth re-testing.
But Lisp is a nice pair of shoes to try on. It's sufficiently different from ALGOLish languages that it tests many beliefs many people probably take for granted. Like that syntax matters.