Aussagenlogik
Transcription
Aussagenlogik
Aussagenlogik Schnelldurchlauf Michael Leuschel Softwaretechnik und Programmiersprachen 05/11/2012 Lecture 3 Teil 1: Sprache (Syntax) 05/11/2012 Bestandteile n Atomare Aussagen (atomic propositions) n Entweder wahr oder falsch (Wahrheitswert, truth value auf Englisch; true oder false) n n zwei-wertig Wahrheitswert einer komplexen Aussage: n bestimmt von Wahrheitswerten der Bestandteile it rains and it is cold 05/11/2012 it rains (compositionality) it is cold Alphabet der Aussagenlogik n Connectives n n n n n n Atomare Aussagen n n Und (Konjunktion): ∧ Oder (Disjunktion): ∨ Wenn (Implikation): → Not (Negation): ¬ Äquivalenz ⟺, Exklusives Oder, . . . Bezeichner: p q ( ) Klammerung n 05/11/2012 Klammern rains ... Well Formed Formulas (wff’s) n n Atomic propositions are wff’s If α and β are wff’s then so are: n n n n n n (¬ α) (α ∧ β) (α ∨ β) (α → β) No other expressions are wff’s Note: α, β : meta-variables (outside of propositional logic) 05/11/2012 Precedence 1. Negation ¬ binds highest 2. Then come ∧ and ∨ 3. Finally comes → Instead of n (((¬ p) ∧ (¬ q)) → ((¬ p) ∨ (¬ q))) we can write n n 05/11/2012 (¬ p ∧ ¬ q) → (¬ p ∨ ¬ q) ¬p∧¬q → ¬p∨¬q Teil 2: Semantik der Aussagenlogik Interpretations, Models, Logical Consequence 05/11/2012 Wahrheitswert einer Formel n n Zuweisung: wff ↝ Wahrheitswert Junktoren: n n n n feste Bedeutung Bsp: true ∧ false ergibt false Wahrheitstabellen (truth tables) Atomare Aussagen: n n 05/11/2012 egal (wahr oder falsch) (we study form, not content !) Wahrheitstabellen α β ¬α α∧ β α∨ β α→ β α⇔ β true true false true true true true true false false false true false false false true true false true true false false false true false false true true 05/11/2012 Interpretations and Models n An interpretation v is a mapping from atomic propositions to {true,false} Example: v(p) = true, v(q) = false n Extension to wff’s by using truth tables Example: v(p ∨ q) = ? n 1. Determine v for subparts: n n v(p) = true, v(q) = false 2. Apply truth table: true ∨ false = true n v(p ∨ q) = true For v’(p) = false, v’(q) = false: v’(p ∨ q) = false n Interpretation v is a model for α iff v(α) = true 05/11/2012 Exercises: n n v(p) = true,v(q) = false v’(p) = false,v’(q) = false n n n n n n n 05/11/2012 v(p ∧ (q ∨ (¬p) )) = v(p → (p → q )) = v’(p → (p → q )) = v(p ∨ (¬p)) = v’(p ∨ (¬p)) = v(p ∧ (¬p)) = v’(p ∧ (¬p)) = Exercises: Solutions n n v(p) = true,v(q) = false v’(p) = false,v’(q) = false n n v(p ∧ (q ∨ (¬p) )) false n n v(p) v(q ∨ (¬p) ) true n v(p ∧ (q ∨ (¬p) )) = false v(p → (p → q )) = false v’(p → (p → q )) = true v(p ∨ (¬p)) = true v’(p ∨ (¬p)) = true false v(q) false 05/11/2012 v(¬p) false n n v(p) true v(p ∧ (¬p)) = false v’(p ∧ (¬p)) = false Satisfaction, Tautology, Contradiction n A wff α is called satisfiable (erfüllbar) iff it has a model n unsatisfiable or a contradiction (unerfüllbar, Widerspruch) iff it has no model n a tautology iff all interpretations are models n p ∧ (¬p) p ∧ (q ∨ (¬p) ) p ∨ (¬p) 05/11/2012 Consequence and Equivalence n Two wff’s α and β are logically equivalent, denoted by α ≡ β, iff they have the same models (i.e., same truth table) (p ∧ q) ≡ (q ∧ p) (p → q) ≡ (¬p ∨ q) n A wff α is called a logical consequence of β, denoted by β ⇒ α, iff every model of β is also a model of α p ⇒ (q ∨ p) 05/11/2012 often instead of ⇒ Consequence for Sets n An interpretation v is a model for a set S of wff’s iff it is a model for every element of S v(p)=true, n v is a model of {(p ∨ q), (¬p → ¬q)} A set A of wff’s is called a logical consequence of another set B, denoted by B ⇒ A, iff every model of B is also a model of Α {p} ⇒ {(p ∨ q), (¬p → ¬q)} 05/11/2012 Beispiel Ja {p → q , ¬q} ⇒ {¬p} ? auch Modell von {¬p} v(p)=true, v(q)= true v(p)=true, v(q)=false v(p)= false, v(q)= true 4 mögliche Interpretationen 05/11/2012 v(p)= false, v(q)= false einziges Modell von {p → q , ¬q} Ein anderes Puzzle Ein Zettel sagt die Wahrheit der andere nicht. ? In dieser Zelle befindet sich eine Prinzessin und ein Tiger in der anderen Zelle 05/11/2012 ? ? ? In einer der Zellen befindet sich eine Prinzessin und in der anderen ein Tiger Beware: Logic vs Language n n Commutativity: p ∧ q ≡ q ∧ p But: He got scared and he killed the intruder. n He killed the intruder and he got scared. (causal and temporal meaning attached to and) n n Implication If 1 = 2 then I am Winnie the Poh. 05/11/2012 What (definitely) to know for the exam: n Language of Propositional Logic n n wff’s Semantics of Propositional Logic n n n 05/11/2012 Truth tables, Interpretations, Models, Satisfiability,Tautology, Contradiction Logical Equivalence and Consequence