Modellings for belief change: Prioritization and entrenchment*
Hans Rott · Theoria · 1992
However, it is not easy to formulate analogues of the Levi identity for "pick" and "bunch revisions" in the sense explained below. 4Andre Fuhrmann [5] was probably the first to study pick and bunch contractions.He called them choice and meet contractions.For danger of confusion with maxichoice and (full, partial) meet contraction, we introduce new names. 5We note some special cases for -0 e K [theory!] and maximality) and (ii) 0->y/ 6 K' (Proof: since K' \r 0, we get A" ^(0-1//)-0, so K U { -+yf} tf 0, so 0--y/€ K\ by 0->y eK [theory!] and maximality).However, in general A±[0,y/| (ZAT1 (0 V y/) even if A" is a theory; consider K -Cn({p,q}), for which K ±-[p,q] = {Cn(p V q), Cn(p~q)} t but K 1 (p V q) = {Cn(p~q)}.And HL (0 V y/) C HI [0,y] breaks down for non-theories //; just consider H = {p,q,p V q}, for which HI (p V q) -{0 J and HI \p,q\ = {{pvq}}. Presumably because of the problemswith the recovery postulate, NebeEs attention switches from contractions in [20] to revisions in [21].