"John D. Burger" <john(at)mitre(dot)org> writes:
> Arguably a tad inconsistent. I don't know what the OP's rationale was,
> but I can imagine that allowing the syntax on INSERT might simplify the
> generation of SQL in certain circumstances. There's also the Principle
> of Least Surprise, etc.
There's also something called the SQL standard, which forbids both of
those syntaxes.
regards, tom lane