Question

Using boolean algebra (not a specific language implementation), can we evaluate 1 ^ 1 + 1 (or 1 XOR 1 OR 1) unambiguously?

I can derive two evaluations:

 [I]:  (1 ^ 1) + 1 = 0 + 1 = 1
[II]:  1 ^ (1 + 1) = 1 ^ 1 = 0

Perhaps there’s some stated order of operations, or of a left-to-right evaluation? Or is this not defined in Boolean algebra?

Answer 1

We can use the rules of boolean algebra to attempt to evaluate the expression 1 XOR 1 OR 1.

Now:

• XOR is derived from OR such that A XOR B = (¬A AND B) OR (¬B AND A);
• Associativity tells us that A OR (B OR C) = (A OR B) OR C;
• Associativity also tells us that A AND (B AND C) = (A AND B) AND C

So, taking either of the possible interpretations of evaluation order:

  (1 XOR 1) OR 1                       1 XOR (1 OR 1)

Even though we have no left-to-right “evaluation order” defined, these rules are all we need to show that the two possible interpretations are not equivalent:

= (¬1 AND 1) OR (¬1 AND 1) OR 1      = (¬1 AND (1 OR 1)) OR (¬(1 OR 1) AND 1)
= (0 AND 1) OR (0 AND 1) OR 1        = (0 AND 1) OR (0 AND 1)
= 0 OR 0 OR 1                        = 0 OR 0
= 1                                  = 0

Unless I’m forgetting some crucially pertinent axiom, I’ve confirmed that you need more context to evaluate the given expression.

^{(And, of course, examining the expression A XOR B OR C ∀A,B,C is of course substantially more tricky! But if the expression is ambiguous for just one value of all three inputs, then why bother checking for any others?)}

This context is usually provided in language-specific evaluation-order rules. C-type languages give XOR a low precedence (something Richie came to dislike); by contrast, mathematical convention dictates a left-to-right associativity where no other axiom can be applied and an ambiguity is otherwise present.

So, basically, since we’re ignoring language-specific rules, you’d probably go with [I].