I know this question has been asked before, but this is (hopefully) not a duplicate. Dummit and Foote's Algebra asks us to show $$|(\mathbb{Z}/n\mathbb{Z})^{\times}| = \varphi(n)$$ using the following definitions, and nothing else:
$(\mathbb{Z}/n\mathbb{Z})^{\times} := \{\bar a \in \mathbb{Z}/n\mathbb{Z} : \exists \bar c \in \mathbb{Z}/n\mathbb{Z}, \bar a \bar c = \bar 1\}$
Here, $\bar a$ denotes the equivalence class of $a$ in $\mathbb{Z}/n\mathbb{Z}$, namely one of $\bar 0,\bar 1,...,\overline{n-1}$. $\bar a\bar c$ refers to the usual multiplication, i.e. we take any elements from the equivalence classes of $\bar a$ and $\bar c$, multiply them, and then reduce the result $\bmod n$. We have shown this works in an earlier proposition, i.e. modular multiplication is well-defined. So, $(\mathbb{Z}/n\mathbb{Z})^{\times}$ is just the collection of residue classes of $\mathbb{Z}/n\mathbb{Z}$ which have a multiplicative inverse in $\mathbb{Z}/n\mathbb{Z}$.
$\varphi(n)$ is Euler's totient function, as is popularly known. It is not defined using $|(\mathbb{Z}/n\mathbb{Z})^{\times}| = \varphi(n)$ in this text. Rather it is defined as $\varphi(n) = |\{a\in\mathbb{N}: a\leq n, (a,n)=1 \}|$, i.e. number of positive integers less than and co-prime with $n$. $(a,n) = 1$ stands for $\operatorname{g.c.d}(a,n) = 1$.
We have not yet shown that
$(\mathbb{Z}/n\mathbb{Z})^{\times} = \{\bar a \in \mathbb{Z}/n\mathbb{Z} : (a,n) = 1\}$
so please do not use this in your proof. Knowing this, it is trivial, but in a later exercise, we are supposed to prove this using $|(\mathbb{Z}/n\mathbb{Z})^{\times}| = \varphi(n)$ - so let's avoid circular arguments.
Could someone please point me in the right direction? Thanks!