There are alternative syntaxes which are more minimal, for example if you require using parens around abstractions then you are correct that a lambda is not needed (e.g. `(x.y x)z(a.a a)` is easily parsed). There are also syntaxes which don't even use names! De Bruijn indices allow you to define an abstraction in terms of how many levels "up" the param came from, à la `λ λ λ 3 1 (2 1)` which is the S combinator. LC as it was published wasn't necessarily trying to go for minimal syntax as a goal though, presumably Church et al felt the lambda helped distinguish abstractions from other terms, so we get more syntax than strictly necessary - but the same goes for most programming languages, e.g. why do we have to write `function foo (x) { ... }` in JS when they could have made it `fun foo` or `fn foo`? One assumes for readability. 🙂