Anurag Mendhekar has an interesting post about why he uses Lisp and thinks you should too. He uses several different Lisps—Scheme, Common Lisp, and Racket—but boils down their essence to, “An s-expression based, dynamically typed, mostly functional, call-by-value λ-calculus-based language.”
Most of his post is an explication of each of those characteristics, including their advantages and disadvantages. Mendhekar takes a pragmatic view on each of them. While he believes they are generally the right thing, he understands that sometimes you want to violate them. He illustrates this most clearly, I think, in his discussion of statically typed versus dynamically typed languages.
Mendhekar is an advocate of dynamically typed languages and contra conventional wisdom makes a good case for them. That basically boils down to static typing seldom really helping make better programs but always getting in the way by enforcing rules that are so general that they often don’t make sense. The power of Lisp is that although it’s generally dynamically typed, you can make it statically typed when you need to. Take a look at Mendhekar’s post for a more nuanced explanation.
He also discusses how s-expressions eliminate many of the syntax related annoyances of “conventional” languages. There is no associativity or precedence rules to worry about. Everything is either a list or an atom: that’s all there is.
As many of you know, I spent most of my career hacking C but once I discovered Lisp, I didn’t want to program in anything else. If you’d like to explore Lisp, I think Scheme is the best place to start. Some folks recommend Clojure but Mendhekar avoids it for reasons he explains in his post. In any event, take a look at Mendhekar’s post for a good discussion of why you should be using Lisp.