>>1
Lisp is an almost interesting experiment which has grown up to a language. It offers somewhat difficult to read syntax, thus rendering it a terrible language in the general sense -- since a programming language is our best achievement so far in interfacing human reasoning with computational power.
But it is still an interesting application of mathematical concepts and theories into computer interfaces. By being math-based, Lisp is coherent and marginally simple to implement. Naturally, complexity can quickly grow into unmanageable levels in Lisp, as it can grow in any language. Lisp has some applications in some fields, and, as always, such applications
should be limited to such fields.
Lisp is all about lambda-calculus, albeit most Lisp dialects have dispensed mathematical purism on behalf of 'more useful' features. There's nothing really bright on L-calculus, neither on Lisp; instead, it's just 'another way' of thinking things. If you're a whorfist, you'll know that different languages yields different thinking processes, and Lisp has a lot to add in this aspect to the average imperative-oriented programmer.
Obviously, Lisp is not a silver bullet. There are tasks much better left apart of it. For example, no one should implement webservers or device drivers in Lisp. Well, none
should but some do. As always, the fact that some brain-crippled people do this or that doesn't justify the application. The world is very big and there are
horrid lots of grotesquely stupid people out there: just thrive hard for not being one.
Lisp is notable for being original and rather powerful in expressive power (whatever that means). It compiles and interprets rather fast, and is particularly prone to unique optimization techniques.
That said, see Lisp as some specific tool to some specialized, precise job, instead of the generic power drill or screwdriver. It cannot be used on every task, but if used correctly and in the exact problem, it yields much less effort, much prettier code, and much better quality results.