Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Help finding homomorphism

15 views
Skip to first unread message

Michael F. Stemper

unread,
Mar 21, 2022, 5:10:41 PM3/21/22
to
I am trying to find the homomorphism from S_3 to the subgroup
{(),(abc),(acb)}. Since it's normal in S_3, there must be
such a homomorphism. Call it "f".

Since o((abc))=o((acb))=3, we either have f((abc)) = (abc) or
f((abc)) = (acb).

What about the other elements? They're all transpositions, which
are involutions. Presumably, they must all map to involutions in
the image. But, there's only one involution in the image -- the
identity permutation.

This doesn't work, since if all of the transpositions were
in the kernel of the homomorphism (call it "f"), we would
have f((abc)) = f((ab)(bc)) = f((ab))f((bc)) = ()(), showing
that f maps all of S_3 to the identity.

There is obviously something very wrong with my understanding.
Would somebody please set me straight?


P.S. I do *not* want somebody to hand me the homomorphism. I just
want to have my mistake pointed out. If I don't work this out for
myself, I won't be any more knowledgeable than I am now.
--
Michael F. Stemper
There's no "me" in "team". There's no "us" in "team", either.

Arturo Magidin

unread,
Mar 24, 2022, 12:56:41 PM3/24/22
to
On Monday, March 21, 2022 at 4:10:41 PM UTC-5, Michael F. Stemper wrote:
> I am trying to find the homomorphism from S_3 to the subgroup
> {(),(abc),(acb)}. Since it's normal in S_3, there must be
> such a homomorphism. Call it "f".

This assertion is false. What we know is that if G is a group and N is a normal subgroup, then there is a homomorphism from G to some group with KERNEL equal to N. there is no guarantee that there is a homomorphism to N.

The only normal subgroups of S_3 are the trivial group, S_3, and the subgroup { (), (abc), (acb)} that you describe. That means that all images of S_3 have order 1 (trivial map), 2 (if the kernel is this group) or 6. The only possible map from S_3 to a group of order 3 is, therefore, the trivial map.

> P.S. I do *not* want somebody to hand me the homomorphism. I just
> want to have my mistake pointed out.

Your mistake is the belief that such a map exists. The only homomorphism S_3 -> { (), (abc), (acb)} is the trivial map, sending everything to (). You have misunderstood the connection between normal subgroups and homomorphisms.

--
Arturo Magidin

Michael F. Stemper

unread,
Mar 25, 2022, 12:13:11 PM3/25/22
to
Hey! Long time no see, and thanks for the reply.

A few hours of quality time with Dummit & Foote, Isaacs, and Pinter showed
me exactly what you said, although I kept trying to force it into my
mistaken understanding.

Ironically, I was doing this to try and improve my understanding
of quotient groups. I was right in needing to do that, but got off
on the wrong foot entirely.

Thanks again.

--
Michael F. Stemper
Psalm 94:3-6
0 new messages