Group factorisation theorem

From Maths
Jump to: navigation, search
(Unknown grade)
This page is a stub
This page is a stub, so it contains little or minimal information and is on a to-do list for being expanded.The message provided is:
This page seems very long for what was (at most) half a page of A4, but otherwise seems good!
  • This page is waiting for a final review, then this notice will be removed.

Statement

Diagram
Let G and H be groups, and consider any NG [Note 1] and π:GG/N the canonical projection of the quotient group, let φ:GH be any group homomorphism, then[1]:
  • If NKer(φ) then φ factors uniquely through π to yield ˉφ:G/NH given by ˉφ:[g]φ(g) [Note 2] and ˉφ is a group homomorphism.

Additionally we have φ=¯φπ (or in other terms, the diagram on the right commutes)

Proof

The bulk of this proof involves invoking the function factorisation theorem:

Let f:XY and w:XW be functions.
  • If x,yX[[w(x)=w(y)][f(x)=f(y)]] then f "factors (uniquely, if w is surjective) through w" via:
    • ˉf:WY given by ˉf:uf(w1(u)) (which is "well defined")
  • Which has the property that f=ˉfw (the diagram on the right commutes)

Applying the function factorisation theorem

Before we can apply this theorem we need to check that the objects in play are eligible (satisfy the requirements to factor) for the theorem. We set up as follows:

  • Let G, H be groups, let NG and let φ:GH be a group homomorphism with NKer(φ), then:
    • a,bX[[π(a)=π(b)][φ(a)=φ(b)]]
  • where π:GG/N is the canonical projection of the quotient group

Proof

Let a,bX be given

  1. If π(a)π(b) then it doesn't matter what φ(a) and φ(b) are, as for logical implication if the LHS is false, it doesn't matter what the RHS for the implication to be considered true.
  2. If π(a)=π(b) then we require φ(a)=φ(b) for the implication to be true.
    • By hypothesis we have π(a)=π(b) so:
      1. π(a)[π(b)]1=e (here e will denote the identity of whatever group makes sense in the context, in this case G/N and may not be the strongest logical relation that can be used)
        1. π(a)π(b1)=e
        2. π(ab1)=e
      • ab1Ker(π) But! The kernel of the canonical projection of a quotient group is the normal subgroup of the quotient
      • ab1N
      • ab1Ker(φ) (by hypothesis, as NKer(φ) using the implies-subset relation)
      • φ(ab1)=e (e is the identity of H this time)
      • φ(a)[φ(b)]1=e
      • φ(a)=φ(b)
    • We have shown π(a)=π(b)φ(a)=φ(b) as required.

Result

We have shown that the factor (function) theorem is eligible for application here, thus we apply it and yield:

  • A function ˉφ:G/NH given by ˉφ:[u]φ(u)[Note 3][Note 4]
  • By the surjectivity of the canonical project of the quotient group (π:GG/N) we know from the factor theorem that the induced function, ˉφ:G/NH is the unique function that satisfies:
    • φ=ˉφπ (there is no other ˉφ with this property that is distinct from ˉφ)

Showing ˉφ is a group homomorphism

We now have a function:

  • ˉφ:G/NH given by ˉφ:[u]φ(u). We claim this is a group homomorphism. To show this we must show that:
    • a,bG/N[ˉφ(a,b)=ˉφ(a)ˉφ(b)]

We shall write [u]G/N and choose u[u] to represent the classes in what follows, however as has been mentioned in the notes several times, this isn't a problem.

  • Let [u],[v]G/N be given; as usual [uv]=[u][v] where is the binary operation of the G/N group.
    • Then ˉφ([uv])=φ(uv)=φ(u)φ(v) as φ is a group homomorphism.
    • Note that ˉφ([u])ˉφ([v])=φ(u)φ(v)
      • So φ(u)φ(v)=ˉφ([u])ˉφ([v])=ˉφ([uv])
  • We have shown ˉφ([u])ˉφ([v])=ˉφ([uv]) as required.

Thus ˉφ is a grop homomorphism from G/N to H

See also

Notes

  1. Jump up The notation AB means A is a normal subgroup of the group B.
  2. Jump up This may look strange as obviously you're thinking "what if we took a different representative h[g] with hg, then we'd have φ(h) instead of φ(g)!", these are actually the same, see Factor (function) for more details, I shall explain this here.
    • Technically we have this: ˉφ:uφ(π1(u)) for the definition of ˉφ
      • Note though that if g,hπ1(u) that:
        • π(g)=π(h)=u and by hypothesis we have [π(x)=π(y)][φ(x)=φ(y)]
          • Thus φ(g)=φ(h)
      • So whichever representative of [g] we use φ(h) for h[g] is the same.
    • This is actually all dealt with as a part of factor (function) not this theorem. However it is worth illustrating.
  3. Jump up This is just a nicer way of writing:
    • ˉφ:gφ(π1(g))
  4. Jump up Please note, again, this may be an abuse of notation, but the result is well defined, as for a,bπ1(x) we see φ(a)=φ(b) by hypothesis.

References

  1. Jump up Abstract Algebra - Pierre Antoine Grillet