32 |
32 |
33 \medskip |
33 \medskip |
34 |
34 |
35 The axioms for an $n$-category are spread throughout this section. |
35 The axioms for an $n$-category are spread throughout this section. |
36 Collecting these together, an $n$-category is a gadget satisfying Axioms \ref{axiom:morphisms}, |
36 Collecting these together, an $n$-category is a gadget satisfying Axioms \ref{axiom:morphisms}, |
37 \ref{nca-boundary}, \ref{axiom:composition}, \ref{nca-assoc}, \ref{axiom:product}, \ref{axiom:extended-isotopies} and \ref{axiom:vcones}. |
37 \ref{nca-boundary}, \ref{axiom:composition}, \ref{nca-assoc}, \ref{axiom:product}, \ref{axiom:extended-isotopies} and \ref{axiom:splittings}. |
38 For an enriched $n$-category we add Axiom \ref{axiom:enriched}. |
38 For an enriched $n$-category we add Axiom \ref{axiom:enriched}. |
39 For an $A_\infty$ $n$-category, we replace |
39 For an $A_\infty$ $n$-category, we replace |
40 Axiom \ref{axiom:extended-isotopies} with Axiom \ref{axiom:families}. |
40 Axiom \ref{axiom:extended-isotopies} with Axiom \ref{axiom:families}. |
41 |
41 |
42 Strictly speaking, before we can state the axioms for $k$-morphisms we need all the axioms |
42 Strictly speaking, before we can state the axioms for $k$-morphisms we need all the axioms |
670 In addition, collar maps act trivially on $\cC(X)$. |
670 In addition, collar maps act trivially on $\cC(X)$. |
671 \end{axiom} |
671 \end{axiom} |
672 |
672 |
673 \medskip |
673 \medskip |
674 |
674 |
675 We need one additional axiom, in order to constrain the poset of decompositions of a given morphism. |
675 We need one additional axiom. |
676 We will soon want to take colimits (and homotopy colimits) indexed by such posets, and we want to require |
676 It says, roughly, that given a $k$-ball $X$, $k<n$, and $c\in \cC(X)$, there exist sufficiently many splittings of $c$. |
677 that these colimits are in some sense locally acyclic. |
677 We use this axiom in the proofs of \ref{lem:d-a-acyclic}, \ref{lem:colim-injective} \nn{...}. |
678 Before stating the axiom we need a few preliminary definitions. |
678 All of the examples of (disk-like) $n$-categories we consider in this paper satisfy the axiom, but |
|
679 nevertheless we feel that it is too strong. |
|
680 In the future we would like to see this provisional version of the axiom replaced by something less restrictive. |
|
681 |
|
682 We give two alternate versions of he axiom, one better suited for smooth examples, and one better suited to PL examples. |
|
683 |
|
684 \begin{axiom}[Splittings] |
|
685 \label{axiom:splittings} |
|
686 Let $c\in \cC_k(X)$, with $0\le k < n$. |
|
687 Let $X = \cup_i X_i$ be a splitting of $X$. |
|
688 \begin{itemize} |
|
689 \item (Version 1) Then there exists an embedded cell complex $S_c \sub X$, called the string locus of $c$, |
|
690 such that if the splitting $\{X_i\}$ is transverse to $S_c$ then $c$ splits along $\{X_i\}$. |
|
691 \item (Version 2) Consider the set of homeomorphisms $g:X\to X$ such that $c$ splits along $\{g(X_i)\}$. |
|
692 Then this subset of $\Homeo(X)$ is open and dense. |
|
693 \end{itemize} |
|
694 \nn{same something about extension from boundary} |
|
695 \end{axiom} |
|
696 |
|
697 We note some consequences of Axiom \ref{axiom:splittings}. |
|
698 |
|
699 First, some preliminary definitions. |
679 If $P$ is a poset let $P\times I$ denote the product poset, where $I = \{0, 1\}$ with ordering $0\le 1$. |
700 If $P$ is a poset let $P\times I$ denote the product poset, where $I = \{0, 1\}$ with ordering $0\le 1$. |
680 Let $\Cone(P)$ denote $P$ adjoined an additional object $v$ (the vertex of the cone) with $p\le v$ for all objects $p$ of $P$. |
701 Let $\Cone(P)$ denote $P$ adjoined an additional object $v$ (the vertex of the cone) with $p\le v$ for all objects $p$ of $P$. |
681 Finally, let $\vcone(P)$ denote $P\times I \cup \Cone(P)$, where we identify $P\times \{0\}$ with the base of the cone. |
702 Finally, let $\vcone(P)$ denote $P\times I \cup \Cone(P)$, where we identify $P\times \{0\}$ with the base of the cone. |
682 We call $P\times \{1\}$ the base of $\vcone(P)$. |
703 We call $P\times \{1\}$ the base of $\vcone(P)$. |
683 (See Figure \ref{vcone-fig}.) |
704 (See Figure \ref{vcone-fig}.) |
784 \caption{(a) $P$, (b) $P\times I$, (c) $\Cone(P)$, (d) $\vcone(P)$} |
805 \caption{(a) $P$, (b) $P\times I$, (c) $\Cone(P)$, (d) $\vcone(P)$} |
785 \label{vcone-fig} |
806 \label{vcone-fig} |
786 \end{figure} |
807 \end{figure} |
787 |
808 |
788 |
809 |
|
810 \begin{lem} |
|
811 \label{lemma:vcones} |
|
812 Let $c\in \cC_k(X)$, with $0\le k < n$, and |
|
813 let $P$ be a finite poset of splittings of $c$. |
|
814 Then we can embed $\vcone(P)$ into the splittings of $c$, with $P$ corresponding to the base of $\vcone(P)$. |
|
815 Furthermore, if $q$ is any decomposition of $X$, then we can take the vertex of $\vcone(P)$ to be $q$ up to a small perturbation. |
|
816 \end{lem} |
|
817 |
|
818 \begin{proof} |
|
819 After a small perturbation, we may assume that $q$ is simultaneously transverse to all the splittings in $P$, and |
|
820 (by Axiom \ref{axiom:splittings}) that $c$ splits along $q$. |
|
821 We can now choose, for each splitting $p$ in $P$, a common refinement $p'$ of $p$ and $q$. |
|
822 This constitutes the middle part of $\vcone(P)$. |
|
823 \end{proof} |
|
824 |
|
825 |
|
826 \noop{ %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
827 |
|
828 We need one additional axiom, in order to constrain the poset of decompositions of a given morphism. |
|
829 We will soon want to take colimits (and homotopy colimits) indexed by such posets, and we want to require |
|
830 that these colimits are in some sense locally acyclic. |
|
831 Before stating the axiom we need a few preliminary definitions. |
|
832 If $P$ is a poset let $P\times I$ denote the product poset, where $I = \{0, 1\}$ with ordering $0\le 1$. |
|
833 Let $\Cone(P)$ denote $P$ adjoined an additional object $v$ (the vertex of the cone) with $p\le v$ for all objects $p$ of $P$. |
|
834 Finally, let $\vcone(P)$ denote $P\times I \cup \Cone(P)$, where we identify $P\times \{0\}$ with the base of the cone. |
|
835 We call $P\times \{1\}$ the base of $\vcone(P)$. |
|
836 (See Figure \ref{vcone-fig}.) |
|
837 \begin{figure}[t] |
|
838 \centering |
|
839 \begin{tikzpicture} |
|
840 [kw node/.style={circle,fill=orange!70}, |
|
841 kw arrow/.style={-latex, very thick, blue!70, shorten >=.06cm, shorten <=.06cm}, |
|
842 kw label/.style={cca}, |
|
843 ] |
|
844 |
|
845 \definecolor{cca}{rgb}{.1,.4,.3}; |
|
846 |
|
847 \node at (0,0) { |
|
848 \begin{tikzpicture} |
|
849 \draw |
|
850 (0,0) node[kw node](p1){} |
|
851 (1,.5) node[kw node](p2){} |
|
852 (2,0) node[kw node](p3){}; |
|
853 |
|
854 \draw[kw arrow] (p1) -- (p3); |
|
855 \draw[kw arrow] (p2) -- (p3); |
|
856 \draw[kw arrow] (p1) -- (p2); |
|
857 |
|
858 \draw[kw label] (1,-.6) node{(a)}; |
|
859 \end{tikzpicture} |
|
860 }; |
|
861 |
|
862 \node at (7,0) { |
|
863 \begin{tikzpicture} |
|
864 \draw |
|
865 (0,0) node[kw node](p1){} |
|
866 ++(0,2.5) node[kw node](q1){} |
|
867 (1,.5) node[kw node](p2){} |
|
868 ++(0,2.5) node[kw node](q2){} |
|
869 (2,0) node[kw node](p3){} |
|
870 ++(0,2.5) node[kw node](q3){} |
|
871 ; |
|
872 |
|
873 \draw[kw arrow] (p1) -- (p3); |
|
874 \draw[kw arrow] (p2) -- (p3); |
|
875 \draw[kw arrow] (p1) -- (p2); |
|
876 \draw[kw arrow] (q1) -- (q3); |
|
877 \draw[kw arrow] (q2) -- (q3); |
|
878 \draw[kw arrow] (q1) -- (q2); |
|
879 \draw[kw arrow] (p1) -- (q1); |
|
880 \draw[kw arrow] (p2) -- (q2); |
|
881 \draw[kw arrow] (p3) -- (q3); |
|
882 |
|
883 \draw[kw label] (1,-.6) node{(b)}; |
|
884 \end{tikzpicture} |
|
885 }; |
|
886 |
|
887 \node at (0,-5) { |
|
888 \begin{tikzpicture} |
|
889 \draw |
|
890 (0,0) node[kw node](p1){} |
|
891 (1,.5) node[kw node](p2){} |
|
892 ++(0,2.5) node[kw node](v){} |
|
893 (2,0) node[kw node](p3){} |
|
894 ; |
|
895 |
|
896 \draw[kw arrow] (p1) -- (p3); |
|
897 \draw[kw arrow] (p2) -- (p3); |
|
898 \draw[kw arrow] (p1) -- (p2); |
|
899 \draw[kw arrow] (p1) -- (v); |
|
900 \draw[kw arrow] (p2) -- (v); |
|
901 \draw[kw arrow] (p3) -- (v); |
|
902 |
|
903 \draw[kw label] (1,-.6) node{(c)}; |
|
904 \end{tikzpicture} |
|
905 }; |
|
906 |
|
907 \node at (7,-5) { |
|
908 \begin{tikzpicture} |
|
909 \draw |
|
910 (0,0) node[kw node](p1){} |
|
911 ++(-2,2.5) node[kw node](q1){} |
|
912 (1,.5) node[kw node](p2){} |
|
913 ++(-2,2.5) node[kw node](q2){} |
|
914 ++(4,0) node[kw node](v){} |
|
915 (2,0) node[kw node](p3){} |
|
916 ++(-2,2.5) node[kw node](q3){} |
|
917 ; |
|
918 |
|
919 \draw[kw arrow] (p1) -- (p3); |
|
920 \draw[kw arrow] (p2) -- (p3); |
|
921 \draw[kw arrow] (p1) -- (p2); |
|
922 \draw[kw arrow] (p1) -- (v); |
|
923 \draw[kw arrow] (p2) -- (v); |
|
924 \draw[kw arrow] (p3) -- (v); |
|
925 \draw[kw arrow] (q1) -- (q3); |
|
926 \draw[kw arrow] (q2) -- (q3); |
|
927 \draw[kw arrow] (q1) -- (q2); |
|
928 \draw[kw arrow] (p1) -- (q1); |
|
929 \draw[kw arrow] (p2) -- (q2); |
|
930 \draw[kw arrow] (p3) -- (q3); |
|
931 |
|
932 \draw[kw label] (1,-.6) node{(d)}; |
|
933 \end{tikzpicture} |
|
934 }; |
|
935 |
|
936 \end{tikzpicture} |
|
937 \caption{(a) $P$, (b) $P\times I$, (c) $\Cone(P)$, (d) $\vcone(P)$} |
|
938 \label{vcone-fig} |
|
939 \end{figure} |
|
940 |
|
941 |
789 \begin{axiom}[Splittings] |
942 \begin{axiom}[Splittings] |
790 \label{axiom:vcones} |
943 \label{axiom:vcones} |
791 Let $c\in \cC_k(X)$, with $0\le k < n$, and |
944 Let $c\in \cC_k(X)$, with $0\le k < n$, and |
792 let $P$ be a finite poset of splittings of $c$. |
945 let $P$ be a finite poset of splittings of $c$. |
793 Then we can embed $\vcone(P)$ into the splittings of $c$, with $P$ corresponding to the base of $\vcone(P)$. |
946 Then we can embed $\vcone(P)$ into the splittings of $c$, with $P$ corresponding to the base of $\vcone(P)$. |
1462 We want to define $y_1\bullet y_2 \in \cl\cC(W)$. |
1616 We want to define $y_1\bullet y_2 \in \cl\cC(W)$. |
1463 Choose a permissible decomposition $\du_a X_{ia} \to W_i$ and elements |
1617 Choose a permissible decomposition $\du_a X_{ia} \to W_i$ and elements |
1464 $y_{ia} \in \cC(X_{ia})$ representing $y_i$. |
1618 $y_{ia} \in \cC(X_{ia})$ representing $y_i$. |
1465 It might not be the case that $\du_{ia} X_{ia} \to W$ is a permissible decomposition of $W$, |
1619 It might not be the case that $\du_{ia} X_{ia} \to W$ is a permissible decomposition of $W$, |
1466 since intersections of the pieces with $\bd W$ might not be well-behaved. |
1620 since intersections of the pieces with $\bd W$ might not be well-behaved. |
1467 However, using the fact that $\bd y_i$ splits along $\bd Y$ and applying Axiom \ref{axiom:vcones}, |
1621 However, using the fact that $\bd y_i$ splits along $\bd Y$ and applying Axiom \ref{axiom:splittings}, |
1468 we can choose the decomposition $\du_{a} X_{ia}$ so that its restriction to $\bd W_i$ is a refinement |
1622 we can choose the decomposition $\du_{a} X_{ia}$ so that its restriction to $\bd W_i$ is a refinement |
1469 of the splitting along $\bd Y$, and this implies that the combined decomposition $\du_{ia} X_{ia}$ |
1623 of the splitting along $\bd Y$, and this implies that the combined decomposition $\du_{ia} X_{ia}$ |
1470 is permissible. |
1624 is permissible. |
1471 We can now define the gluing $y_1\bullet y_2$ in the obvious way, and a further application of Axiom \ref{axiom:vcones} |
1625 We can now define the gluing $y_1\bullet y_2$ in the obvious way, and a further application of Axiom \ref{axiom:splittings} |
1472 shows that this is independent of the choices of representatives of $y_i$. |
1626 shows that this is independent of the choices of representatives of $y_i$. |
1473 |
1627 |
1474 |
1628 |
1475 \medskip |
1629 \medskip |
1476 |
1630 |