Anon05/09/26, 15:10No.16973047
You're asking this on /sci/, so presumably you are too autistic to parse logic in plain language. Very well.Let [math]D={D1,D2}[/math] be the two Doors (one leads to freedom, the other to death).
Let [math]G={GT,GL}[/math] be the two Guards (one always tells the Truth, one always Lies)
Let [math]f:D{0,1}[/math] be the freedom function:
[eqn]f(D_i) = \begin{cases} 1 & \text{if } D_i \text{ leads to freedom} \\ 0 & \text{if } D_i \text{ leads to death} \end{cases}[/eqn]
Define a truth-functional operator for each guard. For any yes/no proposition [math]P[/math]:
[math]answer(GT,P)=[[P]][/math]
[math]answer(GL,P)=¬[[P]][/math]
where [math][[P]]∈{0,1}[/math] denotes the actual truth value of [math]P[/math].We want to find a question [math]Q[/math] such that
[math]answer(GT,Q)=answer(GL,Q) [/math]We achieve this by composing the two truth functions. We ask a nested question that forces the lie to be applied twice.
Consider the question "If I asked the other guard whether [math]D1[/math] leads to freedom, would he say yes?"
[math]P≡(f(D1)=1)[/math]
[math]Q(Gi)=answer(Gj,P)[/math] where [math]j \neq i[/math]
the response is:
[math]R(Gi)=answer(Gi,Q(Gi))=answer(Gi,answer(Gj,P))[/math]Case 1:
[math]R(GT)=answer(GT,answer(GL,P)) [/math]
[math]R(GT)=answer(GL,P)=¬[[P]] [/math]
Case 2:
[math]R(GL)=answer(GL,answer(GT,P)) [/math]
[math]R(GL)=¬answer(GT,P)=¬[[P]] [/math]Thus:
[eqn]\boxed{R(G_T) = R(G_L) = \lnot [[P]]}[/eqn]
The composition answer always involves exactly one truthful evaluation and one inversion, regardless of order: Since the composition [math]id∘¬=¬∘id=¬[/math], the guard you happen to ask is irrelevant.