Please explain some of Paul Graham's points on Lisp

I always tend to think of LISP as following a Von Neumann architecture where there is no difference between code and data and non-LISP derived languages as Harvard architecture where there are distinct differences between code and data.

To say that symbols are strings that happened to be interned, while true, doesn't, I think, capture their essence. I think the best way to understand symbols is that they are literals of an infinite enumeration type. Symbols are <i>names</i> that refer to things.

While there's an implied question here, the link has plenty of adequate responses.<p>If you want to learn Lisp, I recommend Common Lisp, or else a Scheme implementation with lots of practical extensions (e.g. Chicken or PLT). _Land of Lisp_ is a good intro. _SICP_ is also excellent, but hell-bent on teaching you long-term deep theoretical wisdom rather than useful-today tricks. But either way: <i>Do the exercises.</i> Try writing your daily experimental code in it. Think in it. Make it work. Live in it.