> As described in the SRFI, the define-XXX forms introduce new bindings
> (locations, keywords, etc.). Contrary to those, alias doesn't. This is
> one reason why it is called `alias` and not `define-alias`. Note that
> this SRFI in principle allows an implementation to alias an unbound
> identifier (it is left unspecified), in which case not even a "new
> identifier is introduced" (whatever this means).

I'm convinced by your reasoning. `alias` is the right name after all.