Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

The following corollary is quite confusing to me.

Let $N$ be a normal subgroup of a group, $G$ and let $\bar{G}$ denote the set of cosets of $N$ in $G$. Let $\pi: G \rightarrow \bar{G}$ be the canonical homomorphism. Let $a_{1},...a_{k}$ be elements of $G$ such that the product $a_{1}...a_{k}$ is in $N$. Then $\bar{a_{1}}...\bar{a_{k}} = 1$.

The proof is that the product $p=a_{1}...a_{k}$ is in $N$, so $\pi(p) = \bar{p} = \bar{1}$. Since $\pi$ is a homomorphism, so $\bar{a_{1}}...\bar{a_{k}} = \bar{p}$.

I'm wondering why it equals 1.

NB: a bar over something indicates an element from the set of cosets... I think (ref. page 66, Algebra by Artin).

Thanks in advance for your help.

share|cite|improve this question
$\bar a = \bar b$ iff $aN = bN$ iff $b^{-1}a \in N$. Apply this with $\bar p = \bar 1$. – Jack Schmidt Aug 31 '12 at 18:51
Not sure why this result needs the title of corollary actually. The identity element of the group $G/N$ is the coset $N = N1,$ where $1$ denotes the identity element of $G.$ – Geoff Robinson Aug 31 '12 at 18:53
@JackSchmidt Is there someplace I would be able to see a proof for that? @ GeoffRobinson The title of corollary was not chosen by me (it was from a textbook) – achacttn Aug 31 '12 at 19:03
@achattn: My comment was not directed against you- I just meant that I did no see why the book felt the need to state the result as a corollary, when it looks like a tautological statement to me. – Geoff Robinson Aug 31 '12 at 19:07
@GeoffRobinson Ah ok. – achacttn Aug 31 '12 at 19:09
up vote 3 down vote accepted

Recall that the cosets of $N$ form a partition of $G$: for any $a,b\in G$, either $aN=bN$, or $aN\cap bN=\varnothing$.

Now let $p\in N$. $N$ is a subgroup of $G$, so $p^{-1}\in N$, and therefore $1=pp^{-1}\in pN\cap N$. Thus, $pN$ and $N=1N$ are not disjoint, so they must be equal: $pN=1N$, or, in the bar notation, $\bar p=\bar 1$.

share|cite|improve this answer
When $aN = bN$, does it mean that it can be $a.n_{1}$ and $b.n_{2}$ where $n_{1}$ is not necessarily equal to $n_{2}$ (and both are in $N$). Thanks Edit again: Ah thanks, I understand now. I'm new to algebra so it is difficult to constantly be holding all the theorems and propositions etc at the same time. – achacttn Aug 31 '12 at 19:05
@achacttn: $aN=bN$ means that $\{an:n\in N\}=\{bn:n\in N\}$. If $c\in aN=bN$ and $a\ne b$, then yes, there will be distinct $n_1,n_2\in N$ such that $c=an_1=bn_2$. – Brian M. Scott Aug 31 '12 at 19:15

(Presumably redundant response to questions in comments; Brian's answer already contains this)

$\bar a$ is by definition $aN$, the coset of $N$ containing $a$.

If $\bar a = \bar b$, then by definition $aN = bN$.

Guido's lemma is that $aN = bN$ if and only if $b^{-1}aN = N$ if and only if $b^{-1}a \in N$.

While simple, this is extremely useful.

Proof: If $aN = bN$ then every element of $aN$ is contained in $bN$, so in particular, $a\cdot 1 \in aN$ is also in $bN$, so there is some $n \in N$ such that $a \cdot 1 = b \cdot n$. Dividing on the left by $b$ gives $b^{-1} a = n \in N$ and $b^{-1} a N = n N = N$. Conversely if $b^{-1}a \in N$, then there is some $n \in N$ such that $b^{-1} \cdot a = n$, and then $a \cdot 1 = b \cdot n$ so that $b$ is in both $aN$ and $bN$. Since cosets are either equal or disjoint, $aN = bN$. $\square$

share|cite|improve this answer
Thanks very much for that. – achacttn Aug 31 '12 at 19:30

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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