Propositional Equivalences What are the roots of x 3 4 x ? Well, in high-school algebra you learned how to deal with such numeric formulas: x 3 4 x = x x 2 4 factor out x = x x 2 x 2 The identity a 2 b 2 a b a b with a being x, and b being 2.

This last expression happens to be useful since it is in a form which lets us read off the roots 0, +2, -2. The rules of algebra tell us that these three different formulas are all equivalent. In fact, our very definition of two formulas being equivalent is that for any value of x the two formulas return the same value. We are distinguishing between syntax (the expression itself, as data), and semantics (what the expression means). Usually, when presented with syntax, one is supposed to bypass that and focus on its meaning (e.g., reading a textbook). However, in logic and post-modern literature alike, we are actually studying the interplay between syntax and semantics. The general gist is that in each step, you rewrite subparts of your formula according to certain rules (“replacing equals with equals”). Well, we can use a similar set of rules about rewriting formulas with equivalent ones, to answer the questions of whether two formulas are equal, or whether a formula is a tautology. George Boole was the first to realize that and are just values in the way that numbers are, and he first codified the rules for manipulating them; thus Boolean algebra is named in his honor. The term “algebra” comes from the values , and operators ∧, ∨ having some very specific properties similar to those of numbers with ×, +.

George Boole (1815-1864) Again, each individual step consists of rewriting a formula according to certain rules. So, just what are the rules for manipulating Boolean values? We'll start with an example. 1 a b 2≡ b Dominance of over ∧ 3≡ b Commutativity of ∨ 4≡b Identity element for ∨ is 5≡bIdentity element for ∧ is

Thus we have a series of equivalent formulas, with each step justified by citing a propositional equivalence. By and large, the equivalences are rather mundane. A couple are surprisingly handy; take a moment to consider DeMorgan's laws. φ ψ φ ψ φ ψ φ ψ

(Try φ being “Leprechauns are green”, and ψ being “Morgana Le Fay likes gold”. Do these laws make sense, for each of the four possible truth assignments?) Augustus DeMorgan was also an important figure in the formalization of logic.

Augustus DeMorgan (1806-1871) Here is another example. For a statement φψ, the contrapositive of that formula is ψ φ . We can show that a formula is equivalent to its contrapositive: Contrapositive 1 φ ψ 2≡ φ ψ Definition of ⇒ 3≡ ψ φ Commutativity of ∨ 4≡ ψ φ Double Complementation 5≡ ψ φ Definition of ⇒

Don't confuse the contrapositive of a statement with the converse of a formula: The converse of φψ is the formula ψφ; in general a formula is not equivalent to its converse! This next example is actually a proof of one of the laws from the given list, using (only) others from the list. Absorption of ∨ 1 φ ψ ψ 2≡ φ ψ ψ Identity of ∧ 3≡ ψ φ ψ Commutativity of ∨ 4≡ ψ φ Distributivity of ∧ over ∨ 5≡ψ Dominance of ∨ 6≡ψIdentity of ∧

Show that the “Absorption of ∧” equivalence holds, given the other equivalences. I.e., show a b b b . 1 a b b 2≡ a b b Identity of ∨ 3≡ b a b Commutativity of ∨ 4≡ b a Distributivity of ∨ over ∧ 5≡b Dominance of ∧ 6≡bIdentity of ∨

Compared to proofs using truth tables, Boolean algebra gives us much shorter proofs. But, determining which equivalence to use in the next step of a proof can be difficult. In this case, compare the solution for this exercise to the previous absorption proof. These two proofs have a special dual relationship described in the next section. Show that the modus ponens rule, a a b b always holds. I.e., show that it is a tautology, and thus equivalent to . 1 a a b b 2≡ a a b b Definition of ⇒ 3≡ a a a b b Distributivity of ∨ over ∧ 4≡ a b b Complement 5≡ a b b Commutativity of ∨ 6≡ a b b Identity of ∨ 7≡ a b b Definition of ⇒ 8≡ a b b DeMorgan's law 9≡ a b b Associativity of ∨ 10≡ a b b Commutativity of ∨ 11≡ a Complement 12≡Dominance of ∨

So, what would it mean to use Boolean algebra as reasoning for WaterWorld? That is, if you wanted to show that G-safe was true, how would you do that using Boolean algebra? As with truth-tables, we would take the conjunction of all the WaterWorld domain axioms (call it ρ), and the board's observed state (ψ). We would then want to show that asserting G-safe was already equivalent to the rules-and-observed-state: ρ ψ ρ ψ G-safe .

Duals (optional) Duals: a symmetry between ∧, ∨ mediated by ¬. Looking at the provided propositional equivalences, you should notice a strong similarity between those for ∨ and those for ∧. Take any equivalence, swap ∨s and ∧s, swap s and s, and you'll have another equivalence! For instance, there are two flavors of DeMorgan's law, which are just duals of each other: φ ψ φ ψ φ ψ φ ψ

In terms of circuit diagrams, we can change each AND gate to an OR gate and add negation-bubbles to each gate's inputs and outputs. The principle of duality asserts that this operation yields an equivalent circuit. The idea of duality is more general than this. For example, polyhedra have a natural dual of interchanging the role of vertices and faces.