1
$\begingroup$

I am studying Algebraic Operads with the book Algebraic Operads, by Jean-Louis Loday and Bruno Vallette. The authors present the following definitions of coaugmented coalgebra and reduced coproduct: enter image description here

My first problem is: Who is $1$? Is it an element that we can choose?

My first idea of choice was $1=u(1_{\mathbb{K}})$. Since $u$ is a coalgebra morphism, this gives us that $\Delta(1) = 1\otimes 1$ and maybe I could use this for to solve my last question: How I prove that $\bar{\Delta}(x) \in \bar{C}\otimes \bar{C}$?

Thanks! ;)

$\endgroup$
2
$\begingroup$

Yes, $1 = u(1_{\mathbb{K}})$. You have a short exact sequence $$0 \to \ker(\epsilon) \to C \xrightarrow{\epsilon} \mathbb{K} \to 0$$ and it is split by the coaugmentation $u : \mathbb{K} \to C$, so by the splitting lemma, $C \cong \ker(\epsilon) \oplus \mathbb{K}$, and $\mathbb{K}$ is identified as a submodule of $C$ as the image of $u$.

You simply need to compute $(\epsilon \otimes \operatorname{id})(\bar{\Delta}(x))$ and $(\operatorname{id} \otimes \epsilon)(\bar{\Delta}(x))$ to check that $\bar{\Delta(x)}$ is in $\bar{C} \otimes \bar{C}$. But for example $$(\epsilon \otimes \operatorname{id})(\bar{\Delta}(x)) = (\epsilon \otimes \operatorname{id})(\Delta(x)) - \epsilon(x) \otimes 1 - \epsilon(1) \otimes x;$$ since $x \in \bar{C} = \ker(\epsilon)$, $\epsilon(x) \otimes 1 = 0$; since $\epsilon \circ u = \operatorname{id}$, $\epsilon(1) = 1_K$; and by the counit relations, $(\epsilon \otimes \operatorname{id})(\Delta(x)) = x$. So you're left with $x - 0 - x = 0$, and so $\bar{\Delta(x)} \in \bar{C} \otimes C$. Similarly $(\operatorname{id} \otimes \epsilon)(\bar{\Delta}(x)) = 0$ so $\bar{\Delta}(x) \in \bar{C} \otimes \bar{C}$.

$\endgroup$
2
  • 1
    $\begingroup$ @Pryscilla Regarding your suggested edit, you misunderstood what I wrote: by applying the counit only on the first factor one shows that only the first factor is in $\bar C$. $\endgroup$ Dec 12 '15 at 16:47
  • $\begingroup$ Ok! I understand now. You are right. Thanks! ;) $\endgroup$ Feb 12 '16 at 11:42
1
$\begingroup$

You are correct about the identity $1=u(1_{\mathbb{K}})$. For the second part, this should follow from the distributivity of tensor product over direct sum and the fact that, as an element of $\overline{C}$, $x$ has no nonzero part in $\mathbb{K}1$.

$\endgroup$

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.