Keltham takes a quick look at the nametag of whoever that was. Why the Chelians collectively aced this problem but not the predicate-logic one... presumably it's just down to more actual practice with algebra?
"Precisely. If we substitute in 1 for x and y, and evaluate the left-hand sides and right-hand sides of each equation, we get the following assertions:"
[1] |
(1 = 1) |
x = 1 |
(premise) |
[2] |
(1 = 1) |
y = 1 |
(premise) |
[3] |
(1 = 1) |
1 = 1 |
(id. 1) |
[4] |
(1 = 1) |
x = y |
(subst lh [1] ; subst rh [2]) |
[5] |
(1 = 1) |
x*x = y*x |
(mult. x) |
[6] |
(0 = 0) |
x*x - y*y = y*x - y*y |
(sub. (y*y)) |
[7] |
(0 = 0) |
(x + y)*(x - y) = y*(x - y) |
(diff-squares lh. x, y ; factor rh. y) |
[8] |
(2 = 1) |
x + y = y |
(cancel. *(x - y)) |
[9] |
(2 = 1) |
2 = 1 |
(conclusion) |
"The tactics of algebra - like being allowed to add 3 to both sides of an equation - are meant to preserve truth, not create it from scratch. If an equation starts out true, a tactic in algebra should not produce a false equation from that true equation."
"This way of thinking holds even if the elements of the equation refer to things in the outside world. Let x be the number of people sitting in the brown chair, 2 as it happens, and let y be the number of people sitting in the red chair, currently 3. It is then an unnecessary truth, not a necessary truth, that x + 1 = y, as I have defined those terms to refer to the outside world. In our world, x + 1 = y evaluates to 3=3, which happens to be true; but if you cast an illusion showing two people sitting in the brown chair and two people sitting in the red chair, the equation in that world would evaluate to 3 = 2, which is false. And if I said x + 10 = y, that would be an unnecessary falsehood; in our world it evaluates to the false statement 12 = 3."
"Now apply the rules of algebra, add 2 to both sides, and transform the first equation x + 1 = y to the new equation x + 3 = y + 2. In our world, this evaluates to 5 = 5, which is again true. If we apply the same tactic to x + 10 = y, it yields x + 12 = y + 2, which evaluates to 14=5, again false."
"We term a step of inference valid when it is truth-preserving; when it transforms true statements into only other true statements. It doesn't have to preserve falsehood; multiplying both sides of an equation by zero will produce truth even where it didn't previously exist."
"What makes the tactic of adding 2 to both sides of an equation, allowed in math, is not that some Watcher or representative from Governance told you it was allowed." This part got hammered into Keltham and his agemates a lot as a kid, so it was probably determined to be important in practice to emphasize?? "What makes it an allowed step is that, if you have two weights balanced on either side of a scales, and you add two identical rocks to both the left side and the right side, the scales will still balance after that."
"If you look back at the original flawed proof that 2=1, it goes from a true statement in step [7], to a false statement in step [8]. Then between [7] and [8] we must have applied some operation of inference which is not 'valid', which has the ability to take in a true statement and spit out a false statement. This tactic was canceling the multiplication by (x - y) from both sides, which is to say, dividing both sides by x - y. Dividing both sides of an equation by 2 is valid; if you have a scales in balance, and remove half the weight from each sides of a scale, it will still be in balance. Here, we see that division by 0 is not valid, because it can produce falsehood from truth. What makes division by 0 unlawful is not that your Watcher told you not to do it while doing algebra; it is that division by 0 is not generally truth-preserving. We can find some equations that will still be true after dividing both sides by a term equal to 0, but it is not a safe step in general."
"Sorry if that part about Watchers seems overly obvious, by the way. It's just that apparently human brains by default try to reuse the part of ourselves that learns from adults not to steal cookies outside of mealtimes or we'll get slapped on the wrist, in order to relate to the rules for manipulating necessary truths that existed outside the start of Time. And these are actually quite different topics; like, rules change sometimes, when Legislators vote on them, but algebra doesn't. So you want to be explicitly aware of the difference, and not go bugging adults to let you divide by zero just this once."