Stable homotopy hypothesis in the Tamsamani model (Q2145907)

The authors formulate and prove the stable homotopy hypothesis in all finite dimensions with respect to a recursively defined notion of weak \(n\)-groupoid due to \textit{Z. Tamsamani} [\(K\)-Theory 16, No. 1, 51--99 (1999; Zbl 0934.18008)]. The original homotopy hypothesis (attributed to Grothendieck and named as such by Baez) states that the standard homotopy theory \(\mathcal{S}\) of spaces is \textit{the} homotopy theory of weak \(\infty\)-groupoids. That is to say, first, that the homotopy theory \(\mathcal{S}_{[0,n]}\) of \(n\)-truncated spaces is equivalent to the well-established models of weak \(n\)-groupoids for small integers \(n\), and, second, that any other description of a homotopy theory \(\infty\)-\(\mathrm{Gpd}\) of weak \(\infty\)-groupoids equipped with a sequence of \(n\)-truncations \(\infty\text{-}\mathrm{Gpd}\rightarrow n\text{-}\mathrm{Gpd}\) comes together with a fundamental \(\infty\)-groupoid functor \(\Pi_{\infty}\colon\mathcal{S}\rightarrow\infty\text{-}\mathrm{Gpd}\) which induces an equivalence \(\Pi_{n}\colon\mathcal{S}_{[0,n]}\rightarrow n\text{-}\mathrm{Gpd}\) of homotopy theories on all truncation levels \(n\) for \(n\leq\infty\). For instance, in the case \(\infty\)-\(\mathrm{Gpd}\) is defined as the homotopy theory \(\mathcal{S}\) of spaces itself, the first part of the homotopy hypothesis holds by fact that the homotopy theory of \(n\)-types for \(n\leq 3\) does recover the homotopy theory of the common algebraic models of \(n\)-groupoids; the second part is trivial. In the case \(\infty\)-\(\mathrm{Gpd}\) is defined as Quillen's homotopy theory for Kan complexes, the second part of the homotopy hypothesis is a theorem due to Quillen himself [(1967; Zbl 0168.20903)]. In the case \(n\)-\(\mathrm{Gpd}\) is defined in terms of Tamsamani's \(n\)-groupoids for all \(n<\infty\), it is a theorem due to \textit{Z. Tamsamani} [\(K\)-Theory 16, No. 1, 51--99 (1999; Zbl 0934.18008)] -- although of course it does not capture \(n=\infty\) itself. The according definitions and results concering Tamsamani \(n\)-groupoids are recalled in Section 2 of the given paper. Based on this homotopy hypothesis, variations of the theme arise considering \(n\)-groupoids equipped with various additional structure. A central such variation is the stable homotopy hypothesis, which states that the stable homotopy theory of (\(n\)-truncated) spectra is \textit{the} homotopy theory of symmetric monoidal weak \(n\)-groupoids for \(n\leq\infty\). The stable homotopy theory of (\(n\)-truncated) spectra is summarized in Section 3 of the paper; a notion of symmetric monoidal weak \(n\)-groupoids for \(n<\infty\) is defined in Section 4 of the paper in terms of \textit{Picard-Tamsamani \(n\)-categories}. A Picard-Tamsamani \(n\)-category is essentially (Lemma 4.8) a ``very special'' \(\Gamma\)-object in Tamsamani \(n\)-groupoids. The according stable homotopy hypothesis (Theorem 4.17) states that the fundamental \(\infty\)-groupoid functor associated to the homotopy theory of Tamsamani n-groupoids induces equivalences between the homotopy theories of Picard-Tamsamani \(n\)-categories and the stable homotopy theories of \(n\)-truncated spectra for \(n<\infty\). The proof is based on Tamsamani's version of the original homotopy hypothesis, together with a result of \textit{A. K. Bousfield} and \textit{E. M. Friedlander} [Lect. Notes Math. 658, 80--130 (1978; Zbl 0405.55021)] who show that the stable homotopy theory of (\(n\)-truncated) spectra is equivalent to the homotopy theory of very special (\(n\)-truncated) \(\Gamma\)-spaces. In Sections 5 and 6, the authors show that their general framework recovers a proof of a classic formulation of the stable homotopy hypothesis in the case \(n=1\). Namely, that Segal's K-theory functor from Picard categories (in other places called \textit{Picard groupoids}, i.e.\ symmetric monoidal groupoids with every object invertible with respect to the monoidal structure) to stable 1-types induces an equivalence of homotopy theories. Indeed, in Section 5 it is shown that the K-theory functor is equivalent to a 2-categorical nerve construction which factors through Picard-Tamsamani 1-categories. In Section 6 it is shown that this corestricted functor is part of an equivalence of associated homotopy theories whose inverse is induced by a left biadjoint of the given nerve.