> There will inevitably be some computations which can be more compactly or efficiently expressed in the untyped concatenative calculus than in the untyped lamb

> There will inevitably be some computations which can be more compactly or efficiently expressed in the untyped concatenative calculus than in the untyped lambda calculus, and visa versa. What would be an example of a computation more compactly expressed in the untyped concatenative calculus than in the binary lambda calculus [1] (where compactness can be directly measured as size in bits) ? I suppose that your programs can also be represented in binary with 3 bits encoding the 6 intrinsics plus the two symbols '[' and ']', and the end of a program could be encoded by an unmatched ']' at the end. [1] https://tromp.github.io/cl/Binary_lambda_calculus.html https://tromp.github.io/cl/Binary_lambda_calculus.html