pnas/pnas.tex
changeset 584 7fc1a7ff9667
parent 583 08e80022a881
child 585 e2996d7b4e6c
child 590 6de8871d5786
equal deleted inserted replaced
583:08e80022a881 584:7fc1a7ff9667
   194 
   194 
   195 There are five basic ingredients of an $n$-category definition:
   195 There are five basic ingredients of an $n$-category definition:
   196 $k$-morphisms (for $0\le k \le n$), domain and range, composition,
   196 $k$-morphisms (for $0\le k \le n$), domain and range, composition,
   197 identity morphisms, and special behavior in dimension $n$ (e.g. enrichment
   197 identity morphisms, and special behavior in dimension $n$ (e.g. enrichment
   198 in some auxiliary category, or strict associativity instead of weak associativity).
   198 in some auxiliary category, or strict associativity instead of weak associativity).
   199 We will treat each of these it turn.
   199 We will treat each of these in turn.
   200 
   200 
   201 To motivate our morphism axiom, consider the venerable notion of the Moore loop space
   201 To motivate our morphism axiom, consider the venerable notion of the Moore loop space
   202 \nn{need citation}.
   202 \nn{need citation -- \S 2.2 of Adams' ``Infinite Loop Spaces''?}.
   203 In the standard definition of a loop space, loops are always parameterized by the unit interval $I = [0,1]$,
   203 In the standard definition of a loop space, loops are always parameterized by the unit interval $I = [0,1]$,
   204 so composition of loops requires a reparameterization $I\cup I \cong I$, and this leads to a proliferation
   204 so composition of loops requires a reparameterization $I\cup I \cong I$, and this leads to a proliferation
   205 of higher associativity relations.
   205 of higher associativity relations.
   206 While this proliferation is manageable for 1-categories (and indeed leads to an elegant theory
   206 While this proliferation is manageable for 1-categories (and indeed leads to an elegant theory
   207 of Stasheff polyhedra and $A_\infty$ categories), it becomes undesirably complex for higher categories.
   207 of Stasheff polyhedra and $A_\infty$ categories), it becomes undesirably complex for higher categories.
   221 For each $0 \le k \le n$, we have a functor $\cC_k$ from 
   221 For each $0 \le k \le n$, we have a functor $\cC_k$ from 
   222 the category of $k$-balls and 
   222 the category of $k$-balls and 
   223 homeomorphisms to the category of sets and bijections.
   223 homeomorphisms to the category of sets and bijections.
   224 \end{axiom}
   224 \end{axiom}
   225 
   225 
   226 Note that the functoriality in the above axiom allows us to operate via
   226 Note that the functoriality in the above axiom allows us to operate via \nn{fragment?}
   227 
   227 
   228 Next we consider domains and ranges of $k$-morphisms.
   228 Next we consider domains and ranges of $k$-morphisms.
   229 Because we assume strong duality, it doesn't make much sense to subdivide the boundary of a morphism
   229 Because we assume strong duality, it doesn't make much sense to subdivide the boundary of a morphism
   230 into domain and range --- the duality operations can convert domain to range and vice-versa.
   230 into domain and range --- the duality operations can convert domain to range and vice-versa.
   231 Instead, we will use a unified domain/range, which we will call a ``boundary".
   231 Instead, we will use a unified domain/range, which we will call a ``boundary".
   281 (For $k=n$ in the plain (non-$A_\infty$) case, see below.)
   281 (For $k=n$ in the plain (non-$A_\infty$) case, see below.)
   282 \end{axiom}
   282 \end{axiom}
   283 
   283 
   284 \begin{axiom}[Strict associativity] \label{nca-assoc}
   284 \begin{axiom}[Strict associativity] \label{nca-assoc}
   285 The composition (gluing) maps above are strictly associative.
   285 The composition (gluing) maps above are strictly associative.
   286 Given any splitting of a ball $B$ into smaller balls
   286 Given any decomposition of a ball $B$ into smaller balls
   287 $$\bigsqcup B_i \to B,$$ 
   287 $$\bigsqcup B_i \to B,$$ 
   288 any sequence of gluings (in the sense of Definition \ref{defn:gluing-decomposition}, where all the intermediate steps are also disjoint unions of balls) yields the same result.
   288 any sequence of gluings (where all the intermediate steps are also disjoint unions of balls) yields the same result.
   289 \end{axiom}
   289 \end{axiom}
   290 \begin{axiom}[Product (identity) morphisms]
   290 \begin{axiom}[Product (identity) morphisms]
   291 \label{axiom:product}
   291 \label{axiom:product}
   292 For each pinched product $\pi:E\to X$, with $X$ a $k$-ball and $E$ a $k{+}m$-ball ($m\ge 1$),
   292 For each pinched product $\pi:E\to X$, with $X$ a $k$-ball and $E$ a $k{+}m$-ball ($m\ge 1$),
   293 there is a map $\pi^*:\cC(X)\to \cC(E)$.
   293 there is a map $\pi^*:\cC(X)\to \cC(E)$.
   372 \subsubsection{Decompositions of manifolds}
   372 \subsubsection{Decompositions of manifolds}
   373 
   373 
   374 \nn{KW: I'm inclined to suppress all discussion of the subtleties of decompositions.
   374 \nn{KW: I'm inclined to suppress all discussion of the subtleties of decompositions.
   375 Maybe just a single remark that we are omitting some details which appear in our
   375 Maybe just a single remark that we are omitting some details which appear in our
   376 longer paper.}
   376 longer paper.}
       
   377 \nn{SM: for now I disagree: the space expense is pretty minor, and it always us to be "in principle" complete. Let's see how we go for length.}
   377 
   378 
   378 A \emph{ball decomposition} of $W$ is a 
   379 A \emph{ball decomposition} of $W$ is a 
   379 sequence of gluings $M_0\to M_1\to\cdots\to M_m = W$ such that $M_0$ is a disjoint union of balls
   380 sequence of gluings $M_0\to M_1\to\cdots\to M_m = W$ such that $M_0$ is a disjoint union of balls
   380 $\du_a X_a$ and each $M_i$ is a manifold.
   381 $\du_a X_a$ and each $M_i$ is a manifold.
   381 If $X_a$ is some component of $M_0$, its image in $W$ need not be a ball; $\bd X_a$ may have been glued to itself.
   382 If $X_a$ is some component of $M_0$, its image in $W$ need not be a ball; $\bd X_a$ may have been glued to itself.
   454 For $k=2$, we have a two types of generators; they each consists of a field $f$ on $W$, and two balls $B_1$ and $B_2$. In the first case, the balls are disjoint, and $f$ restricted to either of the $B_i$ is a null field. In the second case, the balls are properly nested, say $B_1 \subset B_2$, and $f$ restricted to $B_1$ is null. Note that this implies that $f$ restricted to $B_2$ is also null, by the associativity of the gluing operation. This ensures that the differential is well-defined.
   455 For $k=2$, we have a two types of generators; they each consists of a field $f$ on $W$, and two balls $B_1$ and $B_2$. In the first case, the balls are disjoint, and $f$ restricted to either of the $B_i$ is a null field. In the second case, the balls are properly nested, say $B_1 \subset B_2$, and $f$ restricted to $B_1$ is null. Note that this implies that $f$ restricted to $B_2$ is also null, by the associativity of the gluing operation. This ensures that the differential is well-defined.
   455 
   456 
   456 \section{Properties of the blob complex}
   457 \section{Properties of the blob complex}
   457 \subsection{Formal properties}
   458 \subsection{Formal properties}
   458 \label{sec:properties}
   459 \label{sec:properties}
   459 The blob complex enjoys the following list of formal properties.
   460 The blob complex enjoys the following list of formal properties. The first three properties are immediate from the definitions.
   460 
       
   461 The proofs of the first three properties are immediate from the definitions.
       
   462 
   461 
   463 \begin{property}[Functoriality]
   462 \begin{property}[Functoriality]
   464 \label{property:functoriality}%
   463 \label{property:functoriality}%
   465 The blob complex is functorial with respect to homeomorphisms.
   464 The blob complex is functorial with respect to homeomorphisms.
   466 That is, 
   465 That is, 
   514 obtained by adding an outer $(k{+}1)$-st blob consisting of all $B^n$.
   513 obtained by adding an outer $(k{+}1)$-st blob consisting of all $B^n$.
   515 For $k=0$ we choose a splitting $s: H_0(\bc_*(B^n)) \to \bc_0(B^n)$ and send 
   514 For $k=0$ we choose a splitting $s: H_0(\bc_*(B^n)) \to \bc_0(B^n)$ and send 
   516 $x\in \bc_0(B^n)$ to $x - s([x])$, where $[x]$ denotes the image of $x$ in $H_0(\bc_*(B^n))$.
   515 $x\in \bc_0(B^n)$ to $x - s([x])$, where $[x]$ denotes the image of $x$ in $H_0(\bc_*(B^n))$.
   517 \end{proof}
   516 \end{proof}
   518 
   517 
   519 \nn{Properties \ref{property:functoriality} will be immediate from the definition given in
       
   520 \S \ref{sec:blob-definition}, and we'll recall it at the appropriate point there.
       
   521 Properties \ref{property:disjoint-union}, \ref{property:gluing-map} and 
       
   522 \ref{property:contractibility} are established in \S \ref{sec:basic-properties}.}
       
   523 
   518 
   524 \subsection{Specializations}
   519 \subsection{Specializations}
   525 \label{sec:specializations}
   520 \label{sec:specializations}
   526 
   521 
   527 The blob complex has two important special cases.
   522 The blob complex has two important special cases.