## Not signed in

Want to take part in these discussions? Sign in if you have an account, or apply for one below

## Site Tag Cloud

Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.

• CommentRowNumber1.
• CommentAuthorHarry Gindi
• CommentTimeNov 8th 2017
• (edited Nov 8th 2017)

In the case of strict ω-categories, there is an ostensible generalization of the comma construction, where, given maps $f:A \to B \leftarrow C:g$ we let

$f\downarrow_n g$ denote the limit of the cospan

$A^{D^{n-1}} \times_{B^{\partial D^{n-1}}} C^{D^{n-1}} \to B^{D^{n-1}} \times_{B^{\partial D^{n-1}}} B^{D^{n-1}} \leftarrow B^{D^n}$.

The objects of this strict ω-category are pairs of $n-1$-cells $a\in A$, $c \in C$ whose images under $f$ resp. $g$ are parallel, together with an $n$-cell $\gamma: f(a) \to g(c)$, and the arrows are obvious arrows forming commuting diagrams.

I think this construction might be interesting since we have a natural equipment of pairs of projections $\pi_{s,t}: f \downarrow_n g \to f \downarrow_{n-1} g$ for each $n$.

In particular, this should be interesting if we want to define cartesian fibrations of ω-categories globally, when, given an isofibration $p:E\to B$

we ask for the existence of a left adjoint right inverse to the induced maps $E^{D^n} \to p\downarrow_n B$ for each $n\in \mathbb{N}$, and moreover we require that these are compatible with the projections (some kind of Beck-Chevalley condition, I think, and what this should mean intuitively, is that every cartesian $n$-cell lies between cartesian $n-1$-cells).

Is this kind of idea covered anywhere?

• CommentRowNumber2.
• CommentAuthorHarry Gindi
• CommentTimeNov 12th 2017
• (edited Nov 12th 2017)

I found something in Gray’s old papers through Mike’s old notes on 3-toposes, and the correct construction uses the (op)lax Gray cotensor but is related.