I just want to see if this is a misprint or am I making some really silly mistake. the question says let $α$ be a root of $f(x) = x^4 − 2$ over $\text{GF}(5)$.

Obviously what one should do in such cases is take it s.t. $α^4=2$.

However my difficulty arises through the fact that this doesn't seem to be true for any element of $\text{GF}(5):=\{0,1,2,3,4\}$.

what am I doing wrong, because honestly I doubt its a misprint, and suspect it's much more likely that because I'm very new to galois fields that I'm doing something incorrectly.

  • $\begingroup$ (a) ‘Obviously α 625 = α’. Explain in a very few sentences why this is so! (b) Calculate $α^5$,$ α^{25}$ ,$ α^{125}$ . (c) What is the degree of α over GF(5)? (d) Is f(x) irreducible over GF(5)? (e) Write down all roots of f(x) in GF(5)(α). $\endgroup$
    – excalibirr
    Mar 27 '18 at 1:58
  • $\begingroup$ these are the related questions, I don't want answers to these I leave them here purely for context. $\endgroup$
    – excalibirr
    Mar 27 '18 at 1:59
  • 4
    $\begingroup$ $\alpha$ won't lie in $GF(5)$, bit rather in an extension field of $GF(5)$. $\endgroup$ Mar 27 '18 at 2:01
  • $\begingroup$ @Lord Shark the Unknown ah yes , that makes much more sense. So Really what the question wants is to take for granted the fact that $\alpha^4=2$ but rather just to use this property in the rest of the question ? $\endgroup$
    – excalibirr
    Mar 27 '18 at 2:03
  • 2
    $\begingroup$ Further to @LordSharktheUnknown 's comment, there isn't anything in the question which suggests that $\alpha$ ought to lie in $GF(5)$. The phrasing "let $\alpha$ be a root of $f(x) = x^4 - 2$ over $GF(5)$" is telling you to interpret $f(x)$ as an element of $GF(5)[x]$ $\endgroup$
    – abc
    Mar 27 '18 at 2:05

I think there are some important things to remark concerning what you wrote above:

To a). Your calculation is absolutely correct, but the fact that the question as you posted it in your comment speaks about explaining an "obvious" fact in "a few sentences", I would suspect that it aimed more for something like the following (I'm not entirely sure though, as the ordering of the given subtasks seems to disagree with my interpretation):

Suppose we already know that $f(x)=x^4-2$ is irreducible over $\mathrm{GF}(5)$ (part d)). This would tell us that the degree of $\alpha$ over $\mathrm{GF}(5)$ is $4$ (part c)), so $K:=\mathrm{GF}(5)(\alpha)$ is a field with $5^4=625$ Elements. Its multiplicative group $K^\times=K\setminus\{0\}$ has order $\lvert K^\times \rvert=624$. So $a^{624}=1$ for any $a\in K^\times$, in particular $$a^{625}=a\cdot a^{624}=a\cdot 1=a,\quad\text{for any }a\in K^\times$$

Remark: Obviously, this equality holds for $a=0$ as well, so that every $a\in K$ is a root of $x^{625}-x\in\mathrm{GF}(5)[x]$. Considering that a polynomial of degree $n$ over a field can have only a maximum of $n$ roots (in some algebraic closure), one thus gets that every finite field $L$ of characteristic $p$ is a splitting field of $x^{\lvert L\rvert}-x$ over $\mathrm{GF}(p)$.

To c). Here I am not sure what the first line of your solution is used for? An later on, I think it would have been more precise to explicitly give some $\mathrm{GF}(5)$-Basis of $\mathrm{GF}(5)(\alpha)$, you might for example prove that $\{1,\alpha,\alpha^2,\alpha^3\}$ is such a basis (your argument already shows it is generating, so you only had to use the irreducibility of $f$ to prove linear independence), hence $$[\mathrm{GF}(5)(\alpha):\mathrm{GF}(5)]=\dim_{\mathrm{GF}(5)}(\mathrm{GF}(5)(\alpha))=4.$$

To d). How exactly did you end up with only checking there is no $a\in \mathrm{GF}(5)$ with $a^2=2$ to ensure that $f$ has no irreducible factor of degree $2$? You were definitely right about checking for roots to exclude factors of degree $1$, but the point of excluding factors of degree $2$ could need some supplementation, I think. Basically, there are two natural ways to do this (unfortunately both are a bit lengthy):

  1. As $f$ is monic, you may without loss of generality assume both degree $2$ factors monic. So you could multiply out the assumption $$f(x)=(x^2+ax+b)(x^2+cx+d)$$ compare the coefficients and try to solve for $a,b,c,d\in \mathrm{GF}(5)$, getting a contradiction.

  2. As $\mathrm{GF}(5)$ consists of only five elements, there are just $5^2=25$ monic polynomial of degree $2$ in $\mathrm{GF}(5)[x]$. By testing for roots in $\mathrm{GF}(5)$ you might find and list all irreducible ones, and then, for any of these, test if it divides $f$.

To e). If I'm not mistaking what you wrote, this is an important misconception, I think! You could of course label $\alpha$ as $\sqrt[4]{2}$, which makes perfect sense, but the use of $i$ suggests you are thinking about complex numbers. However, the question happens in characteristic $5$! The procedure is similar, though. You already know one root of $f$ in $K=\mathrm{GF}(5)(\alpha)$, which is $\alpha$ (by definition). Now, what are the other roots? This is where the roots of unity come in. By definition, a $n$-th root of unity over a field $k$ (in some fixed algebraic closure of $k$) is nothing but a root of the polynomial $x^n-1\in k[x]$, that is, the roots of unity are the elements $\zeta $ with $\zeta ^n=1$. Now, if $\alpha$ is a root of $f$, that is $\alpha^4=2$, and $\zeta$ is a fourth root of unity over $\mathrm{GF}(5)$, then $$(\zeta\alpha)^4=\zeta^4\alpha^4=\alpha^4=2,$$ so $\zeta\alpha$ is a root of $f$ as well, for any fourth root of unity $\zeta$. Now $x^4-1\in\mathrm{GF}(5)[x]$ is separable (as the characteristic $5$ does not divide $4$), so there are four different 4th roots of unity over $\mathrm{GF}(5)$, call them $1,\zeta,\eta,\tau$ (actually, the group of roots of unity is cyclic as a finite subgroup of a multiplicative group of a field, so we might as well write them as $1,\zeta,\zeta^2,\zeta^3$), and the four elements $\alpha,\zeta\alpha,\eta\alpha,\tau\alpha$ are all roots of $f$. As $f$ is of degree $4$, these are of course all roots. Now, what are the fourth roots of unity over $\mathrm{GF}(5)$? As the multiplicative group $\mathrm{GF}(5)^\times$ has order $4$, we immediately have $a^4=1$ for all $a\in \mathrm{GF}(5)^\times$ - that means, the $4$-th roots of unity are nothing but $1,2,3,4\in\mathrm{GF}(5)$, and the roots of $f$ in $\mathrm{GF}(5)(\alpha)$ are thus $$\alpha,2\alpha,3\alpha,4\alpha.$$


So here's what i ended up doing.

a) $\alpha^4=2$


b)$\alpha^5=\alpha\times\alpha^4=2\alpha$, $\alpha^{25}=\alpha^5\alpha^5=2\alpha\times2\alpha=4\alpha^2$, $\alpha^{125}=\alpha^{25}\alpha^{5}=2\alpha4\alpha^2=8\alpha^3=3\alpha^3$

c)$x^4=2$ over GF(5)

$x^4-2$ is irreducible over GF(5)

so $GF(5)(\alpha)\cong GF(5)[x]/<x^4-2>$

and $GF(5)(\alpha):=\{a+bx+cx^2+dx^3|a,b,c,d \in \Bbb Q \}$

so $\alpha$ has degree 4 over GF(5).

d) there is no $a \in GF(5)$ s.t. $ a^2=2$ so it cant be factored into quadratics and further no elements of GF(5) are solutions to $x^4-2=0$ so it is irreducible


So the roots are $+-\alpha,+-i\alpha$ where $\alpha=\sqrt[4]{2}$.

Is this all okay or am I doing anything wrong ?

  • $\begingroup$ Where did you prove that $x^4-2$ is irreducible over $GF(5)$? $\endgroup$
    – bof
    Mar 27 '18 at 5:15
  • $\begingroup$ Why does $\alpha$ have degree4 $3$ over $GF(5)$? Where did the $3$ come from? $\endgroup$
    – bof
    Mar 27 '18 at 5:17
  • $\begingroup$ @bof I wrote 3 by mistake, it should be of degree 4 because basis for the extension has 4 elements. I've edited it now. $\endgroup$
    – excalibirr
    Mar 27 '18 at 15:45
  • $\begingroup$ @bof and also what method should one use to show irreducibility over a finite field ? $\endgroup$
    – excalibirr
    Mar 27 '18 at 15:46

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.