My book says that for elements $\alpha$ in $\mathbb F_p$, where $p$ is prime, it holds that $$ \alpha^{p^n}=\alpha, $$ because of Fermat's little theorem, which says that $$ a^p=a\mod p. $$ Of course it's clear that $\alpha^p=\alpha$, but I don't see why this would hold for arbitrary powers of $p$. We have $$ \alpha^{p^n}=\alpha^n, $$ but this isn't necessarily equal to $\alpha$, is it?

For context: my book uses this argument, to show that the roots of $X^{p^n}-X$ in $\overline{F_p}$ form a field of $p^n$ elements, and they argue that $F_p$ is a subset of the set of roots of $X^{p^n}-X$.

  • 7
    $\begingroup$ $\alpha^{p^2} \equiv \alpha^{pp} \equiv (\alpha^p)^p \equiv \alpha^p \equiv \alpha \pmod{p}$. $\endgroup$
    – Hubble
    Mar 29, 2019 at 2:52

2 Answers 2


Fixed points stay fixed under iteration, by an obvious induction:

if $\,f(x) = x\,$ then $\, \color{#c00}{f^{\large n}(x) = x}\,\Rightarrow\, f^{\large n+1}(x) = f(\color{#c00}{f^{\large n}(x)}) = f(\color{#c00}x) = x$

OP is the special case $\, f(x) := x^{\large p}\,$ so $\,f^{\large n}(x) = x^{\large p^{\Large n}}\pmod {p}$


Observe that, in any ring $R$, for $\alpha \in R$ and $m \in \Bbb N$ we have

$\alpha^{m^2} = (\alpha^m)^m; \tag 1$

$\alpha^{m^3} = (\alpha^{m^2})^m; \tag 2$

and in general, for $n \in \Bbb N$,

$\alpha^{m^n} = \alpha^{m^{n - 1}m} = (\alpha^{m^{n - 1}})^m. \tag 3$

Now suppose that $R$ is such that

$\alpha^m = \alpha; \tag 4$

then from (1) and (2),

$\alpha^{m^2} = (\alpha^m)^m = \alpha^m = \alpha, \tag 5$

$\alpha^{m^3} = (\alpha^{m^2})^m = \alpha^m = \alpha; \tag 6$

thus if

$\alpha^{m^j} = \alpha, \tag 7$


$\alpha^{m^{j + 1}} = \alpha^{m^jm} = (\alpha^{m^j})^m = \alpha^m = \alpha, \tag 7$

and by induction it may be concluded that

$\alpha^{m^n} = \alpha \tag 8$

holds for all $n \in \Bbb N$.

Now taking

$m = p \in \Bbb P, \tag 9$


$R = \Bbb F_p \tag{10}$

we have

$\alpha^p = \alpha; \tag{11}$

thus it follows that

$\alpha^{p^n} = \alpha, \; \forall n \in \Bbb N. \tag{12}$

  • 1
    $\begingroup$ Simpler: fixed points stay fixed on iteration - see my answer. $\endgroup$ Mar 30, 2019 at 2:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.