% Here the trick is that the first two rules imply minimally p <-> r.
% So not(p) and not(r) are always together true or false,
% and thus not(s) follows.
% This is a difference to D-WFS, which would not yield not(s).

p v q.
q v r.
s v t <- not(p).
t <- not(r).

$v.