Uncountability of the set of functions from the set of natural numbers to itself

45 Views Asked by At

I'm new to this website so forgive me for any formatting mistakes or other errors you encounter. I would appreciate it if you pointed them out. For anyone interested, the problem is from Algebra by Roger Godement.

I'm struggling with an exercise that aims to furnish a proof that the set $\mathbb{N^N}$ of functions from $\mathbb{N}$ to $\mathbb{N}$ is uncountable. The problem starts by assuming the existence of a sequence of functions $\left(f_n\right)_{n\in\mathbb{N}}$ from $\mathbb{N}$ to $\mathbb{N}$, and defining a mapping $f:\mathbb{N}\rightarrow\mathbb{N}$ by $f(n)=f_n(n)+1$ for all $n\in\mathbb{N}$.

The first part asks you to show that there is no number $p\in\mathbb{N}$ such that $f=f_p$, which is easy enough. For if $f=f_p$ for some $p$, then by definition we would have $f_p(p)+1=f(p)=f_p(p)$. But since any cardinal $x$ that satisfies $x=x+1$ is infinite, this implies $f_p(p)$ is infinite, which contradicts the fact that $f_p(p)\in\mathbb{N}$. Hence for all $p\in\mathbb{N}$ we have $f_p\neq f$.

From this, you are supposed to deduce the uncountability of $\mathbb{N^N}$, which one can proceed with by using the lack of surjectivity of $f$ to show that $f$ is not a bijection, as in the problem before this. In particular, since $f(p)\in\mathbb{N}$, we find that there exists $y=f_p(p)\in\mathbb{N}$ such that $\forall p\in\mathbb{N}$, $y\neq f(p)$, which is to say that $f$ is not surjective, hence not bijective (In the preceding discussion you can exchange $f$ and $f_p$; either way I haven't been able to figure it out).

Where this gets fuzzy for me is in trying to use this to show that any function from $\mathbb{N}$ to $\mathbb{N}$ cannot be bijective. I can't find a way to generalize the result to an arbitrary function from $\mathbb{N}$ to $\mathbb{N}$. I've tried two approaches so far.

The first is to show that for any function $h:\mathbb{N}\rightarrow\mathbb{N}$ we can find another function $g:\mathbb{N}\rightarrow\mathbb{N}$ such that $h=g\circ f$. A theorem from a previous chapter guarantees the existence of $g$ provided that if $m$,$n\in\mathbb{N}$, then $f(m)=f(n)\implies h(m)=h(n)$. And since $f$ isn't bijective, neither would $h$ be. But given that $h$ is supposed to be arbitrary, I'm not sure how to prove this implication.

The second approach is to find some way of showing that the sequence $\left(f_p\right)_{n\in\mathbb{N}}$ covers $\mathbb{N^N}$, in which case we could make use of the fact that $f_p$ isn't bijective. However, since no conditions are set on any of the $f_n$, I don't see any way of proving this.

Once you show that a bijection does not exist between $\mathbb{N}$ and $\mathbb{N^N}$, the result follows. Thanks in advance!