4

It's not like we can say that, because $\mathbb{N}$ is a subset of $\mathbb{R}$, the restriction of a bijection between $\mathbb{R}$ and $\mathbb{R}^2$ is going also to be a bijection between $\mathbb{N}$ and $\mathbb{N}^2$ (although such a bijection may exist). There are other ways to prove that $\mathbb{N} \cong \mathbb{N}^2$, but is it possible to show this using the fact that $\mathbb{R} \cong \mathbb{R}^2$?

Henno Brandsma
  • 242,131
  • 1
    Can you conclude ${a,b}$ is in bijection with ${a,b} \times {a,b } = {(a, a), (a,b), (b,a), (b,b)}$? –  Nov 09 '19 at 13:05

1 Answers1

3

No, you cannot show $\mathbb{N}^2 \simeq \mathbb{N}$ from $\mathbb{R}^2 \simeq \mathbb{R}$, as there is no reason to think the witnessing bijection for the former equivalence sends integer pairs to integers.

You can deduce $\mathbb{R}^2 \simeq \mathbb{R}$ from $\mathbb{N}^2 \simeq \mathbb{N}$ if you know that $\mathbb{R} \simeq 2^{\mathbb{N}}$:

$$\mathbb{R} \simeq 2^{\mathbb{N}} \simeq 2^{\mathbb{N}^2} \simeq 2^{\mathbb{N}} \times 2^{\mathbb{N}} \simeq \mathbb{R}^2$$

using the obvious bijections. But not the other way around. But there are even explicit formulas for $\mathbb{N}^2 \simeq \mathbb{N}$, see here, e.g., so there is no need for arguments from a harder to show fact.

Henno Brandsma
  • 242,131