## 28 October 2009

### Irrational

A cute math puzzle that I saw twice in the past week.

There are three parts. The first two are supposed to be very easy.

1. Are there two irrational numbers a and b such that a+b is rational?
2. Are there two irrational numbers a and b such that ab is rational?
3. Are there two irrational numbers a and b such that ab is rational?

Sources: The Princeton Companion and Richard Lipton's blog, quoting a talk by Manny Blum.

Anonymous said...

1. Let a = sqrt(2) and b = -sqrt(2)
2. Let a = b = sqrt(2)
3. If sqrt(2)^sqrt(2) is rational, let a = b = sqrt(2). Otherwise, let a = sqrt(2)^sqrt(2) and let b = sqrt(2). a^b = (sqrt(2)^sqrt(2))^sqrt(2) = sqrt(2)^(sqrt(2)*sqrt(2)) = sqrt(2)^2 = 2.

rgrig said...

That's exactly the proof in The Companion!

Anonymous said...

#3 is the simplest non-constructive proof I know, other than the various ways of saying LEM.

rgrig said...

other than the various ways of saying LEM.

Are you saying that the proof of, say, x=~~x starting from x\/~x is non-constructive? I remember writing down natural deduction style proofs for various equivalent formulations of the excluded middle, but I find the constructive / non-constructive distinction to be rather meaningless for such proofs.

Perhaps I'm misunderstanding.

rgrig said...

I realized I misread the puzzle in Lipton's post: It wants a=b, so the proof above doesn't work. The answer is supposedly still `yes'.

If we replace sqrt(2) by q in the argument above and force a=b then we get: "either (1) q^q is rational or (2) q^(q^(q+1)) is". So it is sufficient to find an irrational number q such that q^(q^(q+1)) is rational. If we set q=sqrt(p) where p is some prime then we see that it's sufficient to find some prime number such that sqrt(p)^(sqrt(p)+1) is even. But none of these two tasks seems easier to me :(

Perhaps a different approach is needed?

rgrig said...

Oops. For sqrt(p)^(sqrt(p)+1) to be even we need sqrt(p) to be odd, which is clearly not possible.

rgrig said...

Oops. I think I'll better wait until I wake up today.

Anonymous said...

re:LEM:

I guess it depends on which proof one thinks one is doing. Certainly, proofs of (forall a . a \/ ~ a) -> (forall a . ~~a -> a) are constructive. But proofs of (forall a . ~~a -> a) are (always?) non-constructive.