Categorical homotopy theory (Q2575129)

In the mid 1970s Grothendieck wrote a series of letters to Larry Breen outlining a theory of stacks of homotopy types. These stacks can be thought of as the homotopy coherent analogues of sheaves and his ideas amounted to a conjecture that the Galois-Poincaré correspondence between \(\pi_1X\)-sets and covering spaces over \(X\), should have an analogue in higher dimensions. In about 1983, Grothendieck started a correspondence with Ronnie Brown and the reviewer in which these earlier ideas, and much more, were developed in a 650 page manuscript. A key idea was to find `test categories' and `modelisers', these were to give categories of models for homotopy types, generalising simplicial sets in a way so as to gain extra geometric intuition on what information was encoded in a homotopy type. This correspondence, which will shortly be available in Latexed form, contained quite a detailed, but informal, discussion of these modelisers, but with few proofs given in detail. These ideas were influential `behind the scenes' for 15 years, but it was \textit{D.-C. Cisinski}'s thesis [Université de Paris 7 (2002)] that explored the fine detail and pushed the theory considerably further. The paper under review provides an excellent exposition of these ideas in the context of \(\mathcal{A}\)-sheaves, where \(\mathcal{A}\) is an arbitrary test category in the sense of Grothendieck. The point of the exposition is not just to make available the ideas and methods of Cisinski's interpretation and extension of Grothendieck, but to apply them to cubical sets and thence, for possible future development, to motivic homotopy. (This latter was another of the ideas developed initially by Grothendieck in that 650 page correspondence \textit{cum} manuscript of 1983.)