We can restate the theorem, and get some fascinating results.
The theorem states that there exists an isomorphism I:G/Ker(φ) → Im(φ). Now, Im(φ) is a subgroup of H, so we can make function i:G/Ker(φ) → H out of I with all the same values, just a different domain.
i inherits lots of properties from I, just doesn’t go the parts of H that are not in Im(φ). Or, the property we lose is surjectivity.
An injective homomorphism is just saying “take a smaller group, and put it exactly the same in a larger group”
So, we can really restate this theorem as “if φ:G→H is a group homomorphism, then you can do the same thing as φ by quotienting, and then injecting the quotient into H.”
Let’s try it in mathier terms: Let’s say φ:G→H is a group homomorphism, and q:G→G/Ker(φ) is the quotient map, then there exists an injective homomorphism i:G/Ker(φ)→H such that i∘q=φ
A solid line means “assumed (like φ) or already proven from assumption (like q),” and a dashed line means “the theorem states that this exists (like i)“
If you see the “φ=i∘q” version, you might realize that this theorem is stating: “For any group homomorphism φ, you can factor φ into a part that crunches it down, q, and a part that puts the crunched down part into the target group, i.”
Here’s a homomorphism from the Klein 4 group, which we denote V, toanother copy of V.
Here is it factored into a quotient and an injection.