Quasi-unital \(\infty\)-categories (Q498739)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Quasi-unital \(\infty\)-categories
scientific article

    Statements

    Quasi-unital \(\infty\)-categories (English)
    0 references
    0 references
    29 September 2015
    0 references
    One model for the theory of \((\infty, 1)\)-categories are complete Segal spaces, as introduced by Rezk. These are certain simplicial spaces \(X_\bullet\), where we think of the points of \(X_n\) as chains of \(n\) composable morphisms and of paths in \(X_n\) as invertible morphisms between such chains. In particular, they satisfy the Segal condition that \(X_{n+m}\) is the homotopy fiber product \(X_n\times^h_{X_0}X_m\). In the spirit of \(\infty\)-categories, compositions are only defined up to contractible choice. In contrast, units are given by the degeneracies in the simplicial structure and are therefore strictly defined. This can be a technical problem, e.g. in the theory of cobordism \(\infty\)-categories, where every cobordism diffeomorphic to \(M\times [0,1]\) should be equally good as an identity. Following a remark of Lurie in his cobordism hypothesis paper, the author sets up a theory of quasi-unital complete semi-Segal spaces that solves this issue. In more detail: Following Lurie, the author defines a \textit{semi-Segal space} as a semisimplicial space \(X_\bullet\) satisfying the Segal condition -- semisimplicial means here that we index only over the injective maps in \(\Delta\) so that we do not have degeneracies. Denoting by \(Ho(\mathcal{S})\) the homotopy category of simplicial sets, we can associate with a semi-Segal space a non-unital \(Ho(\mathcal{S})\)-enriched category \(Ho(X_\bullet)\), its \textit{homotopy category}. We call \(X_\bullet\) \textit{quasi-unital} if \(Ho(X_\bullet)\) admits the structure of a \textit{unital} \(Ho(\mathcal (S))\)-enriched category. Define \(X_1^{inv} \subset X_1\) as the subspace of \textit{invertible edges}, i.e. those edges such that pre- and postcomposition defines isomorphisms of mapping spaces in \(Ho(X_\bullet)\). Then we define a \textit{complete semi-Segal space} as a quasi-unital semi-Segal space such that the \(d_i: X_1^{inv} \to X_0\) are weak equivalences for \(i=0,1\) -- this is analogous to the completeness condition of Rezk. Consider now the \(\infty\)-category \(CsS\) associated with the following category with weak equivalences: Objects are complete semi-Segal spaces, morphisms are those morphisms of semi-Segal spaces that preserve invertible edges and weak equivalences are levelwise weak equivalences. The main theorem of the paper under review is that this \(\infty\)-category is equivalent to that of complete Segal spaces in the sense of Rezk. More precisely, the author constructs a model structure on marked semisimplicial sets whose bifibrant objects are exactly the Reedy fibrant complete semi-Segal spaces. He then constructs a Quillen equivalence to the Rezk model structure on simplicial spaces, where the bifibrant objects are exactly the Reedy fibrant complete Segal spaces. Thus complete semi-Segal spaces are a valid model for \((\infty,1)\)-categories. One can argue that the quasi-unitality condition is more natural from the \(\infty\)-point of view than the strict unitality of complete Segal spaces.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    Segal space
    0 references
    strictification
    0 references
    semisimplicial
    0 references
    0 references
    0 references