Remark 1.2.1. Taking some of the input identifications to be identities gives compatibility with identifications in each individual functor.
It will sometimes be convenient to abbreviate the conditions from Axiom A.7 as follows:
\[
\id \ast \alpha = \alpha = \alpha \ast \id \qquadtext{ and } (\gamma \ast \beta) \ast \alpha = \gamma \ast (\beta \ast \alpha).
\]
More generally, if we are given identifications \(\alpha\colon f = g\) and \(\alpha' \colon f' = g'\) of functors \(C \to D\), and we happen to be given identifications \(f = f'\) and \(g = g'\), we will write \(\alpha = \alpha'\) to indicate the existence of a commutative square of identifications as follows:
From Section 1.2, Coherences.