I actually can't see how that works, at least out of the languages that I know, none of them fit the dot-syntax. The only ones that do (again, in the ones I know), are the lisp variants, but you need to flip the sexpr around. (I suspect the reason why lisp is "flipped" could be because programs are action centric whereas utterances are subject centric (this is a very risky statement to make, but interesting to think about.))
Anyhow, with this in mind, translating the above is absolutely trivial:
((action kick) (object boy) (object ball)) for prefix, or
((object boy) (object ball) (action kick)) for postfix
Boy.ball.kick() isn't an example from any particular language, but I was thinking something along the lines of "boy.ball" (the boy's ball) ".kick" (is kicked). That's different from boy.kick(ball), though.
In Forth, which is a postfix and stack-based language, "boy ball kick" would work by pushing boy and ball on the stack, and kick would pop two values, the kicker and kickee.
I just saw this comment and pedantically, I still think boy.ball is unworkable in this context (concerning SOV order and such).
The above expression illustrates a verb executed on a subject and object which are two distinct entities. This is your Forth example (which I agreed with earlier).
In the dot syntax though, you use it as an indicator of possessiveness, in which case the subject is reduced to "ball", modified by "boy". Furthermore "is kicked" is a passive form, which means it has an implied subject, which is the boy. Therefore the completed expression cannot be written as-is; you'd have to insert the subject (boy) into the expression and delineate it as a separate entity.
boy.kick(boy.ball) would convey this meaning. boy.ball.kicked() could also convey it, but in this case, the innards of kicked() must make reference to boy.
