The Beauty of Lisp

• Programming languages are about turning your ideas into a form a computer can understand.
  
  • The ideas must be mathematically defined, but beyond that they can be arbitrary (Turing equivalence).
• Most good ideas need to be reused.
• Programming involves defining ideas and reducing them to "words", and then referencing those words whenever the idea needs to be reused.
  
  • In this way, programs can be built in layers, one on top of the other, where each layer is abstracted away to make room for the next one.
    
    • A good language has semantics that let you easily label common patterns of ideas (such as functions, data types, and object classes).
    • A compiler turns source code from that language into machine code.
• Some ideas can't be reduced down to a single word in certain languages, because they don't follow the patterns the language designers anticipated, even though the ideas themselves are mathematically defined.
• As a result, the programmer has to "preprocess" his ideas into source code — a mechanical task but one necessitated by the lack of expressiveness of the language. In essence, the programmer is doing the work of the compiler.
• Therefore, what's needed is the ability to define arbitrary patterns of ideas, not just arbitrary ideas.
  
  • The only way to do this, is to be able to define the patterns in the programming language itself — to write programs which can manipulate programs.
    
    • That is what Lisp is.