> For my expander I actually implemented all let forms, for instance, as macros that reduced to lambda applications. > Now other lisps might have more forms th

> For my expander I actually implemented all let forms, for instance, as macros that reduced to lambda applications. > Now other lisps might have more forms that don't reduce quite nicely, I'm not sure, I still strongly believe in my original point that lisp practically has no syntax This isn't an argument against Lisp having syntax: it does have those operators, and each one defines its own syntax. The fact that they can be written entirely in the language itself doesn't mean Lisp doesn't have syntax, it means that Lisp has extensible syntax: you can add more syntax whenever you like, and in some dialects, you can remove or replace the pre-defined syntax forms. Syntactic sugar is still syntax, too: in C, arr[i] is just sugar for *(arr + i), but I don't know anyone who would claim that array indexing isn't part of C's syntax.