author | Scott Morrison <scott@tqft.net> |
Wed, 23 Jun 2010 10:54:29 -0400 | |
changeset 388 | eec4b1f9cfc2 |
parent 368 | eb7a1ea85179 |
child 385 | b1da2a454ee7 |
permissions | -rw-r--r-- |
70 | 1 |
%!TEX root = ../blob1.tex |
2 |
||
236 | 3 |
\section{Action of \texorpdfstring{$\CH{X}$}{$C_*(Homeo(M))$}} |
100
c5a43be00ed4
No new content, just rearranging (and procrastinating)
kevin@6e1638ff-ae45-0410-89bd-df963105f760
parents:
92
diff
changeset
|
4 |
\label{sec:evaluation} |
c5a43be00ed4
No new content, just rearranging (and procrastinating)
kevin@6e1638ff-ae45-0410-89bd-df963105f760
parents:
92
diff
changeset
|
5 |
|
246 | 6 |
\nn{should comment at the start about any assumptions about smooth, PL etc.} |
7 |
||
236 | 8 |
Let $CH_*(X, Y)$ denote $C_*(\Homeo(X \to Y))$, the singular chain complex of |
9 |
the space of homeomorphisms |
|
10 |
between the $n$-manifolds $X$ and $Y$ (extending a fixed homeomorphism $\bd X \to \bd Y$). |
|
249 | 11 |
We also will use the abbreviated notation $CH_*(X) \deq CH_*(X, X)$. |
12 |
(For convenience, we will permit the singular cells generating $CH_*(X, Y)$ to be more general |
|
70 | 13 |
than simplices --- they can be based on any linear polyhedron. |
249 | 14 |
\nn{be more restrictive here? does more need to be said?}) |
70 | 15 |
|
236 | 16 |
\begin{prop} \label{CHprop} |
70 | 17 |
For $n$-manifolds $X$ and $Y$ there is a chain map |
18 |
\eq{ |
|
244
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
19 |
e_{XY} : CH_*(X, Y) \otimes \bc_*(X) \to \bc_*(Y) |
70 | 20 |
} |
244
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
21 |
such that |
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
22 |
\begin{enumerate} |
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
23 |
\item on $CH_0(X, Y) \otimes \bc_*(X)$ it agrees with the obvious action of |
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
24 |
$\Homeo(X, Y)$ on $\bc_*(X)$ (Proposition (\ref{diff0prop})), and |
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
25 |
\item for any compatible splittings $X\to X\sgl$ and $Y\to Y\sgl$, |
70 | 26 |
the following diagram commutes up to homotopy |
27 |
\eq{ \xymatrix{ |
|
244
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
28 |
CH_*(X\sgl, Y\sgl) \otimes \bc_*(X\sgl) \ar[r]^(.7){e_{X\sgl Y\sgl}} \ar[d]^{\gl \otimes \gl} & \bc_*(Y\sgl) \ar[d]_{\gl} \\ |
236 | 29 |
CH_*(X, Y) \otimes \bc_*(X) |
244
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
30 |
\ar@/_4ex/[r]_{e_{XY}} & |
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
31 |
\bc_*(Y) |
70 | 32 |
} } |
244
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
33 |
\end{enumerate} |
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
34 |
Up to (iterated) homotopy, there is a unique family $\{e_{XY}\}$ of chain maps |
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
35 |
satisfying the above two conditions. |
70 | 36 |
\end{prop} |
37 |
||
345 | 38 |
Before giving the proof, we state the essential technical tool of Lemma \ref{extension_lemma}, |
39 |
and then give an outline of the method of proof. |
|
303
2252c53bd449
minor changes in a few places
Scott Morrison <scott@tqft.net>
parents:
256
diff
changeset
|
40 |
|
70 | 41 |
Without loss of generality, we will assume $X = Y$. |
42 |
||
43 |
\medskip |
|
44 |
||
244
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
45 |
Let $f: P \times X \to X$ be a family of homeomorphisms (e.g. a generator of $CH_*(X)$) |
cf01e213044a
start working on "evaluation map" section
Kevin Walker <kevin@canyon23.net>
parents:
236
diff
changeset
|
46 |
and let $S \sub X$. |
70 | 47 |
We say that {\it $f$ is supported on $S$} if $f(p, x) = f(q, x)$ for all |
345 | 48 |
$x \notin S$ and $p, q \in P$. Equivalently, $f$ is supported on $S$ if |
49 |
there is a family of homeomorphisms $f' : P \times S \to S$ and a `background' |
|
236 | 50 |
homeomorphism $f_0 : X \to X$ so that |
245
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
51 |
\begin{align*} |
70 | 52 |
f(p,s) & = f_0(f'(p,s)) \;\;\;\; \mbox{for}\; (p, s) \in P\times S \\ |
53 |
\intertext{and} |
|
54 |
f(p,x) & = f_0(x) \;\;\;\; \mbox{for}\; (p, x) \in {P \times (X \setmin S)}. |
|
245
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
55 |
\end{align*} |
70 | 56 |
Note that if $f$ is supported on $S$ then it is also supported on any $R \sup S$. |
245
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
57 |
(So when we talk about ``the" support of a family, there is some ambiguity, |
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
58 |
but this ambiguity will not matter to us.) |
70 | 59 |
|
60 |
Let $\cU = \{U_\alpha\}$ be an open cover of $X$. |
|
236 | 61 |
A $k$-parameter family of homeomorphisms $f: P \times X \to X$ is |
245
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
62 |
{\it adapted to $\cU$} |
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
63 |
\nn{or `weakly adapted'; need to decide on terminology} |
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
64 |
if the support of $f$ is contained in the union |
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
65 |
of at most $k$ of the $U_\alpha$'s. |
70 | 66 |
|
67 |
\begin{lemma} \label{extension_lemma} |
|
236 | 68 |
Let $x \in CH_k(X)$ be a singular chain such that $\bd x$ is adapted to $\cU$. |
69 |
Then $x$ is homotopic (rel boundary) to some $x' \in CH_k(X)$ which is adapted to $\cU$. |
|
70 | 70 |
Furthermore, one can choose the homotopy so that its support is equal to the support of $x$. |
71 |
\end{lemma} |
|
72 |
||
245
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
73 |
The proof will be given in Appendix \ref{sec:localising}. |
70 | 74 |
|
75 |
\medskip |
|
76 |
||
236 | 77 |
Before diving into the details, we outline our strategy for the proof of Proposition \ref{CHprop}. |
70 | 78 |
|
73 | 79 |
%Suppose for the moment that evaluation maps with the advertised properties exist. |
236 | 80 |
Let $p$ be a singular cell in $CH_k(X)$ and $b$ be a blob diagram in $\bc_*(X)$. |
246 | 81 |
We say that $p\ot b$ is {\it localizable} if there exists $V \sub X$ such that |
82 |
\begin{itemize} |
|
70 | 83 |
\item $V$ is homeomorphic to a disjoint union of balls, and |
84 |
\item $\supp(p) \cup \supp(b) \sub V$. |
|
246 | 85 |
\end{itemize} |
245
7537032ad5a0
more evmap.tex; also testing using hg from office computer; also
Kevin Walker <kevin@canyon23.net>
parents:
244
diff
changeset
|
86 |
(Recall that $\supp(b)$ is defined to be the union of the blobs of the diagram $b$.) |
246 | 87 |
|
88 |
Assuming that $p\ot b$ is localizable as above, |
|
89 |
let $W = X \setmin V$, $W' = p(W)$ and $V' = X\setmin W'$. |
|
70 | 90 |
We then have a factorization |
91 |
\[ |
|
92 |
p = \gl(q, r), |
|
93 |
\] |
|
236 | 94 |
where $q \in CH_k(V, V')$ and $r \in CH_0(W, W')$. |
73 | 95 |
We can also factorize $b = \gl(b_V, b_W)$, where $b_V\in \bc_*(V)$ and $b_W\in\bc_0(W)$. |
70 | 96 |
According to the commutative diagram of the proposition, we must have |
97 |
\[ |
|
73 | 98 |
e_X(p\otimes b) = e_X(\gl(q\otimes b_V, r\otimes b_W)) = |
99 |
gl(e_{VV'}(q\otimes b_V), e_{WW'}(r\otimes b_W)) . |
|
100 |
\] |
|
236 | 101 |
Since $r$ is a plain, 0-parameter family of homeomorphisms, we must have |
73 | 102 |
\[ |
103 |
e_{WW'}(r\otimes b_W) = r(b_W), |
|
70 | 104 |
\] |
236 | 105 |
where $r(b_W)$ denotes the obvious action of homeomorphisms on blob diagrams (in |
73 | 106 |
this case a 0-blob diagram). |
107 |
Since $V'$ is a disjoint union of balls, $\bc_*(V')$ is acyclic in degrees $>0$ |
|
303
2252c53bd449
minor changes in a few places
Scott Morrison <scott@tqft.net>
parents:
256
diff
changeset
|
108 |
(by Properties \ref{property:disjoint-union} and \ref{property:contractibility}). |
73 | 109 |
Assuming inductively that we have already defined $e_{VV'}(\bd(q\otimes b_V))$, |
83 | 110 |
there is, up to (iterated) homotopy, a unique choice for $e_{VV'}(q\otimes b_V)$ |
73 | 111 |
such that |
112 |
\[ |
|
113 |
\bd(e_{VV'}(q\otimes b_V)) = e_{VV'}(\bd(q\otimes b_V)) . |
|
114 |
\] |
|
115 |
||
116 |
Thus the conditions of the proposition determine (up to homotopy) the evaluation |
|
246 | 117 |
map for localizable generators $p\otimes b$. |
73 | 118 |
On the other hand, Lemma \ref{extension_lemma} allows us to homotope |
246 | 119 |
arbitrary generators to sums of localizable generators. |
73 | 120 |
This (roughly) establishes the uniqueness part of the proposition. |
121 |
To show existence, we must show that the various choices involved in constructing |
|
122 |
evaluation maps in this way affect the final answer only by a homotopy. |
|
123 |
||
246 | 124 |
Now for a little more detail. |
125 |
(But we're still just motivating the full, gory details, which will follow.) |
|
126 |
Choose a metric on $X$, and let $\cU_\gamma$ be the open cover of by balls of radius $\gamma$. |
|
127 |
By Lemma \ref{extension_lemma} we can restrict our attention to $k$-parameter families |
|
128 |
$p$ of homeomorphisms such that $\supp(p)$ is contained in the union of $k$ $\gamma$-balls. |
|
129 |
For fixed blob diagram $b$ and fixed $k$, it's not hard to show that for $\gamma$ small enough |
|
130 |
$p\ot b$ must be localizable. |
|
131 |
On the other hand, for fixed $k$ and $\gamma$ there exist $p$ and $b$ such that $p\ot b$ is not localizable, |
|
132 |
and for fixed $\gamma$ and $b$ there exist non-localizable $p\ot b$ for sufficiently large $k$. |
|
133 |
Thus we will need to take an appropriate limit as $\gamma$ approaches zero. |
|
83 | 134 |
|
246 | 135 |
The construction of $e_X$, as outlined above, depends on various choices, one of which |
136 |
is the choice, for each localizable generator $p\ot b$, |
|
137 |
of disjoint balls $V$ containing $\supp(p)\cup\supp(b)$. |
|
138 |
Let $V'$ be another disjoint union of balls containing $\supp(p)\cup\supp(b)$, |
|
139 |
and assume that there exists yet another disjoint union of balls $W$ with $W$ containing |
|
140 |
$V\cup V'$. |
|
141 |
Then we can use $W$ to construct a homotopy between the two versions of $e_X$ |
|
142 |
associated to $V$ and $V'$. |
|
143 |
If we impose no constraints on $V$ and $V'$ then such a $W$ need not exist. |
|
144 |
Thus we will insist below that $V$ (and $V'$) be contained in small metric neighborhoods |
|
145 |
of $\supp(p)\cup\supp(b)$. |
|
146 |
Because we want not mere homotopy uniqueness but iterated homotopy uniqueness, |
|
147 |
we will similarly require that $W$ be contained in a slightly larger metric neighborhood of |
|
148 |
$\supp(p)\cup\supp(b)$, and so on. |
|
149 |
||
83 | 150 |
|
151 |
\medskip |
|
152 |
||
303
2252c53bd449
minor changes in a few places
Scott Morrison <scott@tqft.net>
parents:
256
diff
changeset
|
153 |
\begin{proof}[Proof of Proposition \ref{CHprop}.] |
83 | 154 |
Notation: Let $|b| = \supp(b)$, $|p| = \supp(p)$. |
155 |
||
156 |
Choose a metric on $X$. |
|
157 |
Choose a monotone decreasing sequence of positive real numbers $\ep_i$ converging to zero |
|
158 |
(e.g.\ $\ep_i = 2^{-i}$). |
|
159 |
Choose another sequence of positive real numbers $\delta_i$ such that $\delta_i/\ep_i$ |
|
160 |
converges monotonically to zero (e.g.\ $\delta_i = \ep_i^2$). |
|
88 | 161 |
Let $\phi_l$ be an increasing sequence of positive numbers |
89 | 162 |
satisfying the inequalities of Lemma \ref{xx2phi}. |
236 | 163 |
Given a generator $p\otimes b$ of $CH_*(X)\otimes \bc_*(X)$ and non-negative integers $i$ and $l$ |
83 | 164 |
define |
165 |
\[ |
|
88 | 166 |
N_{i,l}(p\ot b) \deq \Nbd_{l\ep_i}(|b|) \cup \Nbd_{\phi_l\delta_i}(|p|). |
83 | 167 |
\] |
247 | 168 |
In other words, for each $i$ |
169 |
we use the metric to choose nested neighborhoods of $|b|\cup |p|$ (parameterized |
|
88 | 170 |
by $l$), with $\ep_i$ controlling the size of the buffers around $|b|$ and $\delta_i$ controlling |
171 |
the size of the buffers around $|p|$. |
|
83 | 172 |
|
236 | 173 |
Next we define subcomplexes $G_*^{i,m} \sub CH_*(X)\otimes \bc_*(X)$. |
174 |
Let $p\ot b$ be a generator of $CH_*(X)\otimes \bc_*(X)$ and let $k = \deg(p\ot b) |
|
83 | 175 |
= \deg(p) + \deg(b)$. |
176 |
$p\ot b$ is (by definition) in $G_*^{i,m}$ if either (a) $\deg(p) = 0$ or (b) |
|
84 | 177 |
there exist codimension-zero submanifolds $V_0,\ldots,V_m \sub X$ such that each $V_j$ |
83 | 178 |
is homeomorphic to a disjoint union of balls and |
179 |
\[ |
|
84 | 180 |
N_{i,k}(p\ot b) \subeq V_0 \subeq N_{i,k+1}(p\ot b) |
181 |
\subeq V_1 \subeq \cdots \subeq V_m \subeq N_{i,k+m+1}(p\ot b) . |
|
83 | 182 |
\] |
183 |
Further, we require (inductively) that $\bd(p\ot b) \in G_*^{i,m}$. |
|
184 |
We also require that $b$ is splitable (transverse) along the boundary of each $V_l$. |
|
185 |
||
186 |
Note that $G_*^{i,m+1} \subeq G_*^{i,m}$. |
|
73 | 187 |
|
83 | 188 |
As sketched above and explained in detail below, |
189 |
$G_*^{i,m}$ is a subcomplex where it is easy to define |
|
190 |
the evaluation map. |
|
84 | 191 |
The parameter $m$ controls the number of iterated homotopies we are able to construct |
87 | 192 |
(see Lemma \ref{m_order_hty}). |
83 | 193 |
The larger $i$ is (i.e.\ the smaller $\ep_i$ is), the better $G_*^{i,m}$ approximates all of |
236 | 194 |
$CH_*(X)\ot \bc_*(X)$ (see Lemma \ref{Gim_approx}). |
83 | 195 |
|
249 | 196 |
Next we define a chain map (dependent on some choices) $e_{i,m}: G_*^{i,m} \to \bc_*(X)$. |
197 |
(When the domain is clear from context we will drop the subscripts and write |
|
198 |
simply $e: G_*^{i,m} \to \bc_*(X)$). |
|
83 | 199 |
Let $p\ot b \in G_*^{i,m}$. |
200 |
If $\deg(p) = 0$, define |
|
201 |
\[ |
|
202 |
e(p\ot b) = p(b) , |
|
203 |
\] |
|
236 | 204 |
where $p(b)$ denotes the obvious action of the homeomorphism(s) $p$ on the blob diagram $b$. |
83 | 205 |
For general $p\ot b$ ($\deg(p) \ge 1$) assume inductively that we have already defined |
206 |
$e(p'\ot b')$ when $\deg(p') + \deg(b') < k = \deg(p) + \deg(b)$. |
|
84 | 207 |
Choose $V = V_0$ as above so that |
83 | 208 |
\[ |
84 | 209 |
N_{i,k}(p\ot b) \subeq V \subeq N_{i,k+1}(p\ot b) . |
83 | 210 |
\] |
84 | 211 |
Let $\bd(p\ot b) = \sum_j p_j\ot b_j$, and let $V^j$ be the choice of neighborhood |
83 | 212 |
of $|p_j|\cup |b_j|$ made at the preceding stage of the induction. |
213 |
For all $j$, |
|
214 |
\[ |
|
88 | 215 |
V^j \subeq N_{i,k}(p_j\ot b_j) \subeq N_{i,k}(p\ot b) \subeq V . |
83 | 216 |
\] |
217 |
(The second inclusion uses the facts that $|p_j| \subeq |p|$ and $|b_j| \subeq |b|$.) |
|
218 |
We therefore have splittings |
|
219 |
\[ |
|
247 | 220 |
p = p'\bullet p'' , \;\; b = b'\bullet b'' , \;\; e(\bd(p\ot b)) = f'\bullet f'' , |
83 | 221 |
\] |
236 | 222 |
where $p' \in CH_*(V)$, $p'' \in CH_*(X\setmin V)$, |
84 | 223 |
$b' \in \bc_*(V)$, $b'' \in \bc_*(X\setmin V)$, |
86 | 224 |
$f' \in \bc_*(p(V))$, and $f'' \in \bc_*(p(X\setmin V))$. |
236 | 225 |
(Note that since the family of homeomorphisms $p$ is constant (independent of parameters) |
86 | 226 |
near $\bd V$, the expressions $p(V) \sub X$ and $p(X\setmin V) \sub X$ are |
83 | 227 |
unambiguous.) |
86 | 228 |
We have $\deg(p'') = 0$ and, inductively, $f'' = p''(b'')$. |
229 |
%We also have that $\deg(b'') = 0 = \deg(p'')$. |
|
84 | 230 |
Choose $x' \in \bc_*(p(V))$ such that $\bd x' = f'$. |
248 | 231 |
This is possible by \ref{bcontract}, \ref{disjunion} and the fact that isotopic fields |
232 |
differ by a local relation \nn{give reference?}. |
|
83 | 233 |
Finally, define |
234 |
\[ |
|
235 |
e(p\ot b) \deq x' \bullet p''(b'') . |
|
236 |
\] |
|
73 | 237 |
|
84 | 238 |
Note that above we are essentially using the method of acyclic models. |
239 |
For each generator $p\ot b$ we specify the acyclic (in positive degrees) |
|
240 |
target complex $\bc_*(p(V)) \bullet p''(b'')$. |
|
241 |
||
242 |
The definition of $e: G_*^{i,m} \to \bc_*(X)$ depends on two sets of choices: |
|
243 |
The choice of neighborhoods $V$ and the choice of inverse boundaries $x'$. |
|
88 | 244 |
The next lemma shows that up to (iterated) homotopy $e$ is independent |
84 | 245 |
of these choices. |
88 | 246 |
(Note that independence of choices of $x'$ (for fixed choices of $V$) |
247 |
is a standard result in the method of acyclic models.) |
|
84 | 248 |
|
88 | 249 |
%\begin{lemma} |
250 |
%Let $\tilde{e} : G_*^{i,m} \to \bc_*(X)$ be a chain map constructed like $e$ above, but with |
|
251 |
%different choices of $x'$ at each step. |
|
252 |
%(Same choice of $V$ at each step.) |
|
253 |
%Then $e$ and $\tilde{e}$ are homotopic via a homotopy in $\bc_*(p(V)) \bullet p''(b'')$. |
|
254 |
%Any two choices of such a first-order homotopy are second-order homotopic, and so on, |
|
255 |
%to arbitrary order. |
|
256 |
%\end{lemma} |
|
84 | 257 |
|
88 | 258 |
%\begin{proof} |
259 |
%This is a standard result in the method of acyclic models. |
|
260 |
%\nn{should we say more here?} |
|
261 |
%\nn{maybe this lemma should be subsumed into the next lemma. probably it should.} |
|
262 |
%\end{proof} |
|
84 | 263 |
|
87 | 264 |
\begin{lemma} \label{m_order_hty} |
84 | 265 |
Let $\tilde{e} : G_*^{i,m} \to \bc_*(X)$ be a chain map constructed like $e$ above, but with |
266 |
different choices of $V$ (and hence also different choices of $x'$) at each step. |
|
267 |
If $m \ge 1$ then $e$ and $\tilde{e}$ are homotopic. |
|
268 |
If $m \ge 2$ then any two choices of this first-order homotopy are second-order homotopic. |
|
269 |
And so on. |
|
270 |
In other words, $e : G_*^{i,m} \to \bc_*(X)$ is well-defined up to $m$-th order homotopy. |
|
271 |
\end{lemma} |
|
272 |
||
273 |
\begin{proof} |
|
274 |
We construct $h: G_*^{i,m} \to \bc_*(X)$ such that $\bd h + h\bd = e - \tilde{e}$. |
|
275 |
$e$ and $\tilde{e}$ coincide on bidegrees $(0, j)$, so define $h$ |
|
276 |
to be zero there. |
|
277 |
Assume inductively that $h$ has been defined for degrees less than $k$. |
|
278 |
Let $p\ot b$ be a generator of degree $k$. |
|
279 |
Choose $V_1$ as in the definition of $G_*^{i,m}$ so that |
|
280 |
\[ |
|
281 |
N_{i,k+1}(p\ot b) \subeq V_1 \subeq N_{i,k+2}(p\ot b) . |
|
282 |
\] |
|
283 |
There are splittings |
|
284 |
\[ |
|
285 |
p = p'_1\bullet p''_1 , \;\; b = b'_1\bullet b''_1 , |
|
286 |
\;\; e(p\ot b) - \tilde{e}(p\ot b) - h(\bd(p\ot b)) = f'_1\bullet f''_1 , |
|
287 |
\] |
|
236 | 288 |
where $p'_1 \in CH_*(V_1)$, $p''_1 \in CH_*(X\setmin V_1)$, |
84 | 289 |
$b'_1 \in \bc_*(V_1)$, $b''_1 \in \bc_*(X\setmin V_1)$, |
290 |
$f'_1 \in \bc_*(p(V_1))$, and $f''_1 \in \bc_*(p(X\setmin V_1))$. |
|
88 | 291 |
Inductively, $\bd f'_1 = 0$ and $f_1'' = p_1''(b_1'')$. |
84 | 292 |
Choose $x'_1 \in \bc_*(p(V_1))$ so that $\bd x'_1 = f'_1$. |
293 |
Define |
|
294 |
\[ |
|
295 |
h(p\ot b) \deq x'_1 \bullet p''_1(b''_1) . |
|
296 |
\] |
|
297 |
This completes the construction of the first-order homotopy when $m \ge 1$. |
|
298 |
||
299 |
The $j$-th order homotopy is constructed similarly, with $V_j$ replacing $V_1$ above. |
|
300 |
\end{proof} |
|
301 |
||
302 |
Note that on $G_*^{i,m+1} \subeq G_*^{i,m}$, we have defined two maps, |
|
249 | 303 |
$e_{i,m}$ and $e_{i,m+1}$. |
304 |
An easy variation on the above lemma shows that |
|
305 |
the restrictions of $e_{i,m}$ and $e_{i,m+1}$ to $G_*^{i,m+1}$ are $m$-th |
|
84 | 306 |
order homotopic. |
307 |
||
236 | 308 |
Next we show how to homotope chains in $CH_*(X)\ot \bc_*(X)$ to one of the |
85 | 309 |
$G_*^{i,m}$. |
310 |
Choose a monotone decreasing sequence of real numbers $\gamma_j$ converging to zero. |
|
311 |
Let $\cU_j$ denote the open cover of $X$ by balls of radius $\gamma_j$. |
|
345 | 312 |
Let $h_j: CH_*(X)\to CH_*(X)$ be a chain map homotopic to the identity whose image is |
313 |
spanned by families of homeomorphisms with support compatible with $\cU_j$, |
|
314 |
as described in Lemma \ref{extension_lemma}. |
|
86 | 315 |
Recall that $h_j$ and also the homotopy connecting it to the identity do not increase |
85 | 316 |
supports. |
317 |
Define |
|
318 |
\[ |
|
319 |
g_j \deq h_j\circ h_{j-1} \circ \cdots \circ h_1 . |
|
320 |
\] |
|
321 |
The next lemma says that for all generators $p\ot b$ we can choose $j$ large enough so that |
|
322 |
$g_j(p)\ot b$ lies in $G_*^{i,m}$, for arbitrary $m$ and sufficiently large $i$ |
|
247 | 323 |
(depending on $b$, $\deg(p)$ and $m$). |
324 |
%(Note: Don't confuse this $n$ with the top dimension $n$ used elsewhere in this paper.) |
|
85 | 325 |
|
87 | 326 |
\begin{lemma} \label{Gim_approx} |
236 | 327 |
Fix a blob diagram $b$, a homotopy order $m$ and a degree $n$ for $CH_*(X)$. |
85 | 328 |
Then there exists a constant $k_{bmn}$ such that for all $i \ge k_{bmn}$ |
255 | 329 |
there exists another constant $j_{ibmn}$ such that for all $j \ge j_{ibmn}$ and all $p\in CH_n(X)$ |
85 | 330 |
we have $g_j(p)\ot b \in G_*^{i,m}$. |
331 |
\end{lemma} |
|
332 |
||
255 | 333 |
For convenience we also define $k_{bmp} = k_{bmn}$ |
334 |
and $j_{ibmp} = j_{ibmn}$ where $n=\deg(p)$. |
|
254 | 335 |
Note that we may assume that |
336 |
\[ |
|
337 |
k_{bmp} \ge k_{alq} |
|
338 |
\] |
|
339 |
for all $l\ge m$ and all $q\ot a$ which appear in the boundary of $p\ot b$. |
|
255 | 340 |
Additionally, we may assume that |
341 |
\[ |
|
342 |
j_{ibmp} \ge j_{ialq} |
|
343 |
\] |
|
344 |
for all $l\ge m$ and all $q\ot a$ which appear in the boundary of $p\ot b$. |
|
345 |
||
254 | 346 |
|
85 | 347 |
\begin{proof} |
348 |
Let $c$ be a subset of the blobs of $b$. |
|
248 | 349 |
There exists $\lambda > 0$ such that $\Nbd_u(c)$ is homeomorphic to $|c|$ for all $u < \lambda$ |
85 | 350 |
and all such $c$. |
86 | 351 |
(Here we are using a piecewise smoothness assumption for $\bd c$, and also |
90 | 352 |
the fact that $\bd c$ is collared. |
353 |
We need to consider all such $c$ because all generators appearing in |
|
247 | 354 |
iterated boundaries of $p\ot b$ must be in $G_*^{i,m}$.) |
85 | 355 |
|
86 | 356 |
Let $r = \deg(b)$ and |
357 |
\[ |
|
90 | 358 |
t = r+n+m+1 = \deg(p\ot b) + m + 1. |
86 | 359 |
\] |
85 | 360 |
|
361 |
Choose $k = k_{bmn}$ such that |
|
362 |
\[ |
|
248 | 363 |
t\ep_k < \lambda |
85 | 364 |
\] |
365 |
and |
|
366 |
\[ |
|
90 | 367 |
n\cdot (2 (\phi_t + 1) \delta_k) < \ep_k . |
85 | 368 |
\] |
369 |
Let $i \ge k_{bmn}$. |
|
370 |
Choose $j = j_i$ so that |
|
371 |
\[ |
|
90 | 372 |
\gamma_j < \delta_i |
373 |
\] |
|
374 |
and also so that $\phi_t \gamma_j$ is less than the constant $\rho(M)$ of Lemma \ref{xxzz11}. |
|
375 |
||
236 | 376 |
Let $j \ge j_i$ and $p\in CH_n(X)$. |
90 | 377 |
Let $q$ be a generator appearing in $g_j(p)$. |
378 |
Note that $|q|$ is contained in a union of $n$ elements of the cover $\cU_j$, |
|
379 |
which implies that $|q|$ is contained in a union of $n$ metric balls of radius $\delta_i$. |
|
380 |
We must show that $q\ot b \in G_*^{i,m}$, which means finding neighborhoods |
|
381 |
$V_0,\ldots,V_m \sub X$ of $|q|\cup |b|$ such that each $V_j$ |
|
382 |
is homeomorphic to a disjoint union of balls and |
|
383 |
\[ |
|
384 |
N_{i,n}(q\ot b) \subeq V_0 \subeq N_{i,n+1}(q\ot b) |
|
385 |
\subeq V_1 \subeq \cdots \subeq V_m \subeq N_{i,t}(q\ot b) . |
|
386 |
\] |
|
248 | 387 |
Recall that |
388 |
\[ |
|
389 |
N_{i,a}(q\ot b) \deq \Nbd_{a\ep_i}(|b|) \cup \Nbd_{\phi_a\delta_i}(|q|). |
|
390 |
\] |
|
90 | 391 |
By repeated applications of Lemma \ref{xx2phi} we can find neighborhoods $U_0,\ldots,U_m$ |
392 |
of $|q|$, each homeomorphic to a disjoint union of balls, with |
|
393 |
\[ |
|
394 |
\Nbd_{\phi_{n+l} \delta_i}(|q|) \subeq U_l \subeq \Nbd_{\phi_{n+l+1} \delta_i}(|q|) . |
|
85 | 395 |
\] |
248 | 396 |
The inequalities above guarantee that |
397 |
for each $0\le l\le m$ we can find $u_l$ with |
|
90 | 398 |
\[ |
399 |
(n+l)\ep_i \le u_l \le (n+l+1)\ep_i |
|
400 |
\] |
|
401 |
such that each component of $U_l$ is either disjoint from $\Nbd_{u_l}(|b|)$ or contained in |
|
402 |
$\Nbd_{u_l}(|b|)$. |
|
403 |
This is because there are at most $n$ components of $U_l$, and each component |
|
404 |
has radius $\le (\phi_t + 1) \delta_i$. |
|
405 |
It follows that |
|
406 |
\[ |
|
407 |
V_l \deq \Nbd_{u_l}(|b|) \cup U_l |
|
408 |
\] |
|
409 |
is homeomorphic to a disjoint union of balls and satisfies |
|
410 |
\[ |
|
411 |
N_{i,n+l}(q\ot b) \subeq V_l \subeq N_{i,n+l+1}(q\ot b) . |
|
412 |
\] |
|
86 | 413 |
|
90 | 414 |
The same argument shows that each generator involved in iterated boundaries of $q\ot b$ |
415 |
is in $G_*^{i,m}$. |
|
86 | 416 |
\end{proof} |
417 |
||
418 |
In the next few lemmas we have made no effort to optimize the various bounds. |
|
419 |
(The bounds are, however, optimal in the sense of minimizing the amount of work |
|
420 |
we do. Equivalently, they are the first bounds we thought of.) |
|
421 |
||
422 |
We say that a subset $S$ of a metric space has radius $\le r$ if $S$ is contained in |
|
423 |
some metric ball of radius $r$. |
|
424 |
||
425 |
\begin{lemma} |
|
426 |
Let $S \sub \ebb^n$ (Euclidean $n$-space) have radius $\le r$. |
|
427 |
Then $\Nbd_a(S)$ is homeomorphic to a ball for $a \ge 2r$. |
|
428 |
\end{lemma} |
|
429 |
||
430 |
\begin{proof} \label{xxyy2} |
|
431 |
Let $S$ be contained in $B_r(y)$, $y \in \ebb^n$. |
|
89 | 432 |
Note that if $a \ge 2r$ then $\Nbd_a(S) \sup B_r(y)$. |
433 |
Let $z\in \Nbd_a(S) \setmin B_r(y)$. |
|
434 |
Consider the triangle |
|
435 |
\nn{give figure?} with vertices $z$, $y$ and $s$ with $s\in S$. |
|
436 |
The length of the edge $yz$ is greater than $r$ which is greater |
|
437 |
than the length of the edge $ys$. |
|
438 |
It follows that the angle at $z$ is less than $\pi/2$ (less than $\pi/3$, in fact), |
|
439 |
which means that points on the edge $yz$ near $z$ are closer to $s$ than $z$ is, |
|
440 |
which implies that these points are also in $\Nbd_a(S)$. |
|
441 |
Hence $\Nbd_a(S)$ is star-shaped with respect to $y$. |
|
442 |
\end{proof} |
|
443 |
||
444 |
If we replace $\ebb^n$ above with an arbitrary compact Riemannian manifold $M$, |
|
445 |
the same result holds, so long as $a$ is not too large: |
|
249 | 446 |
\nn{what about PL? TOP?} |
89 | 447 |
|
448 |
\begin{lemma} \label{xxzz11} |
|
449 |
Let $M$ be a compact Riemannian manifold. |
|
450 |
Then there is a constant $\rho(M)$ such that for all |
|
451 |
subsets $S\sub M$ of radius $\le r$ and all $a$ such that $2r \le a \le \rho(M)$, |
|
452 |
$\Nbd_a(S)$ is homeomorphic to a ball. |
|
453 |
\end{lemma} |
|
454 |
||
455 |
\begin{proof} |
|
456 |
Choose $\rho = \rho(M)$ such that $3\rho/2$ is less than the radius of injectivity of $M$, |
|
457 |
and also so that for any point $y\in M$ the geodesic coordinates of radius $3\rho/2$ around |
|
458 |
$y$ distort angles by only a small amount. |
|
459 |
Now the argument of the previous lemma works. |
|
85 | 460 |
\end{proof} |
461 |
||
462 |
||
89 | 463 |
|
464 |
\begin{lemma} \label{xx2phi} |
|
465 |
Let $S \sub M$ be contained in a union (not necessarily disjoint) |
|
86 | 466 |
of $k$ metric balls of radius $r$. |
89 | 467 |
Let $\phi_1, \phi_2, \ldots$ be an increasing sequence of real numbers satisfying |
468 |
$\phi_1 \ge 2$ and $\phi_{i+1} \ge \phi_i(2\phi_i + 2) + \phi_i$. |
|
469 |
For convenience, let $\phi_0 = 0$. |
|
248 | 470 |
Assume also that $\phi_k r \le \rho(M)$, |
471 |
where $\rho(M)$ is as in Lemma \ref{xxzz11}. |
|
89 | 472 |
Then there exists a neighborhood $U$ of $S$, |
473 |
homeomorphic to a disjoint union of balls, such that |
|
86 | 474 |
\[ |
89 | 475 |
\Nbd_{\phi_{k-1} r}(S) \subeq U \subeq \Nbd_{\phi_k r}(S) . |
86 | 476 |
\] |
477 |
\end{lemma} |
|
478 |
||
479 |
\begin{proof} |
|
89 | 480 |
For $k=1$ this follows from Lemma \ref{xxzz11}. |
481 |
Assume inductively that it holds for $k-1$. |
|
86 | 482 |
Partition $S$ into $k$ disjoint subsets $S_1,\ldots,S_k$, each of radius $\le r$. |
89 | 483 |
By Lemma \ref{xxzz11}, each $\Nbd_{\phi_{k-1} r}(S_i)$ is homeomorphic to a ball. |
484 |
If these balls are disjoint, let $U$ be their union. |
|
485 |
Otherwise, assume WLOG that $S_{k-1}$ and $S_k$ are distance less than $2\phi_{k-1}r$ apart. |
|
486 |
Let $R_i = \Nbd_{\phi_{k-1} r}(S_i)$ for $i = 1,\ldots,k-2$ |
|
487 |
and $R_{k-1} = \Nbd_{\phi_{k-1} r}(S_{k-1})\cup \Nbd_{\phi_{k-1} r}(S_k)$. |
|
488 |
Each $R_i$ is contained in a metric ball of radius $r' \deq (2\phi_{k-1}+2)r$. |
|
91 | 489 |
Note that the defining inequality of the $\phi_i$ guarantees that |
490 |
\[ |
|
491 |
\phi_{k-1}r' = \phi_{k-1}(2\phi_{k-1}+2)r \le \phi_k r \le \rho(M) . |
|
492 |
\] |
|
89 | 493 |
By induction, there is a neighborhood $U$ of $R \deq \bigcup_i R_i$, |
494 |
homeomorphic to a disjoint union |
|
495 |
of balls, and such that |
|
86 | 496 |
\[ |
89 | 497 |
U \subeq \Nbd_{\phi_{k-1}r'}(R) = \Nbd_{t}(S) \subeq \Nbd_{\phi_k r}(S) , |
86 | 498 |
\] |
89 | 499 |
where $t = \phi_{k-1}(2\phi_{k-1}+2)r + \phi_{k-1} r$. |
86 | 500 |
\end{proof} |
501 |
||
70 | 502 |
\medskip |
503 |
||
254 | 504 |
Let $R_*$ be the chain complex with a generating 0-chain for each non-negative |
505 |
integer and a generating 1-chain connecting each adjacent pair $(j, j+1)$. |
|
358
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
506 |
(So $R_*$ is a simplicial version of the non-negative reals.) |
254 | 507 |
Denote the 0-chains by $j$ (for $j$ a non-negative integer) and the 1-chain connecting $j$ and $j+1$ |
508 |
by $\iota_j$. |
|
509 |
Define a map (homotopy equivalence) |
|
250 | 510 |
\[ |
254 | 511 |
\sigma: R_*\ot CH_*(X, X) \otimes \bc_*(X) \to CH_*(X, X)\ot \bc_*(X) |
250 | 512 |
\] |
254 | 513 |
as follows. |
514 |
On $R_0\ot CH_*(X, X) \otimes \bc_*(X)$ we define |
|
515 |
\[ |
|
516 |
\sigma(j\ot p\ot b) = g_j(p)\ot b . |
|
517 |
\] |
|
255 | 518 |
On $R_1\ot CH_*(X, X) \otimes \bc_*(X)$ we define |
519 |
\[ |
|
520 |
\sigma(\iota_j\ot p\ot b) = f_j(p)\ot b , |
|
521 |
\] |
|
522 |
where $f_j$ is the homotopy from $g_j$ to $g_{j+1}$. |
|
86 | 523 |
|
254 | 524 |
Next we specify subcomplexes $G^m_* \sub R_*\ot CH_*(X, X) \otimes \bc_*(X)$ on which we will eventually |
525 |
define a version of the action map $e_X$. |
|
255 | 526 |
A generator $j\ot p\ot b$ is defined to be in $G^m_*$ if $j\ge j_{kbmp}$, where |
254 | 527 |
$k = k_{bmp}$ is the constant from Lemma \ref{Gim_approx}. |
255 | 528 |
Similarly $\iota_j\ot p\ot b$ is in $G^m_*$ if $j\ge j_{kbmp}$. |
254 | 529 |
The inequality following Lemma \ref{Gim_approx} guarantees that $G^m_*$ is indeed a subcomplex |
530 |
and that $G^m_* \sup G^{m+1}_*$. |
|
250 | 531 |
|
254 | 532 |
It is easy to see that each $G^m_*$ is homotopy equivalent (via the inclusion map) |
533 |
to $R_*\ot CH_*(X, X) \otimes \bc_*(X)$ |
|
534 |
and hence to $CH_*(X, X) \otimes \bc_*(X)$, and furthermore that the homotopies are well-defined |
|
535 |
up to a contractible set of choices. |
|
250 | 536 |
|
254 | 537 |
Next we define a map |
538 |
\[ |
|
539 |
e_m : G^m_* \to \bc_*(X) . |
|
540 |
\] |
|
255 | 541 |
Let $p\ot b$ be a generator of $G^m_*$. |
542 |
Each $g_j(p)\ot b$ or $f_j(p)\ot b$ is a linear combination of generators $q\ot c$, |
|
543 |
where $\supp(q)\cup\supp(c)$ is contained in a disjoint union of balls satisfying |
|
544 |
various conditions specified above. |
|
545 |
As in the construction of the maps $e_{i,m}$ above, |
|
546 |
it suffices to specify for each such $q\ot c$ a disjoint union of balls |
|
547 |
$V_{qc} \sup \supp(q)\cup\supp(c)$, such that $V_{qc} \sup V_{q'c'}$ |
|
548 |
whenever $q'\ot c'$ appears in the boundary of $q\ot c$. |
|
549 |
||
550 |
Let $q\ot c$ be a summand of $g_j(p)\ot b$, as above. |
|
551 |
Let $i$ be maximal such that $j\ge j_{ibmp}$ |
|
552 |
(notation as in Lemma \ref{Gim_approx}). |
|
553 |
Then $q\ot c \in G^{i,m}_*$ and we choose $V_{qc} \sup \supp(q)\cup\supp(c)$ |
|
554 |
such that |
|
555 |
\[ |
|
556 |
N_{i,d}(q\ot c) \subeq V_{qc} \subeq N_{i,d+1}(q\ot c) , |
|
557 |
\] |
|
558 |
where $d = \deg(q\ot c)$. |
|
559 |
Let $\tilde q = f_j(q)$. |
|
560 |
The summands of $f_j(p)\ot b$ have the form $\tilde q \ot c$, |
|
561 |
where $q\ot c$ is a summand of $g_j(p)\ot b$. |
|
562 |
Since the homotopy $f_j$ does not increase supports, we also have that |
|
563 |
\[ |
|
564 |
V_{qc} \sup \supp(\tilde q) \cup \supp(c) . |
|
565 |
\] |
|
566 |
So we define $V_{\tilde qc} = V_{qc}$. |
|
567 |
||
568 |
It is now easy to check that we have $V_{qc} \sup V_{q'c'}$ |
|
569 |
whenever $q'\ot c'$ appears in the boundary of $q\ot c$. |
|
570 |
As in the construction of the maps $e_{i,m}$ above, |
|
571 |
this allows us to construct a map |
|
572 |
\[ |
|
573 |
e_m : G^m_* \to \bc_*(X) |
|
574 |
\] |
|
575 |
which is well-defined up to homotopy. |
|
576 |
As in the proof of Lemma \ref{m_order_hty}, we can show that the map is well-defined up |
|
577 |
to $m$-th order homotopy. |
|
578 |
Put another way, we have specified an $m$-connected subcomplex of the complex of |
|
579 |
all maps $G^m_* \to \bc_*(X)$. |
|
580 |
On $G^{m+1}_* \sub G^m_*$ we have defined two maps, $e_m$ and $e_{m+1}$. |
|
581 |
One can similarly (to the proof of Lemma \ref{m_order_hty}) show that |
|
582 |
these two maps agree up to $m$-th order homotopy. |
|
583 |
More precisely, one can show that the subcomplex of maps containing the various |
|
584 |
$e_{m+1}$ candidates is contained in the corresponding subcomplex for $e_m$. |
|
585 |
\nn{now should remark that we have not, in fact, produced a contractible set of maps, |
|
586 |
but we have come very close} |
|
256
2a5d54f51808
small test on new computer
Kevin Walker <kevin@canyon23.net>
parents:
255
diff
changeset
|
587 |
\nn{better: change statement of thm} |
253
3816f6ce80a8
evmap; about to delete a few paragraphs, but committing just so there's
Kevin Walker <kevin@canyon23.net>
parents:
251
diff
changeset
|
588 |
|
358
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
589 |
\medskip |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
590 |
|
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
591 |
Next we show that the action maps are compatible with gluing. |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
592 |
Let $G^m_*$ and $\ol{G}^m_*$ be the complexes, as above, used for defining |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
593 |
the action maps $e_{X\sgl}$ and $e_X$. |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
594 |
The gluing map $X\sgl\to X$ induces a map |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
595 |
\[ |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
596 |
\gl: R_*\ot CH_*(X\sgl, X \sgl) \otimes \bc_*(X \sgl) \to R_*\ot CH_*(X, X) \otimes \bc_*(X) , |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
597 |
\] |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
598 |
and it is easy to see that $\gl(G^m_*)\sub \ol{G}^m_*$. |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
599 |
From this it follows that the diagram in the statement of Proposition \ref{CHprop} commutes. |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
600 |
|
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
601 |
\medskip |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
602 |
|
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
603 |
Finally we show that the action maps defined above are independent of |
8589275ac65b
CH_* action -- gluing compatibility
Kevin Walker <kevin@canyon23.net>
parents:
357
diff
changeset
|
604 |
the choice of metric (up to iterated homotopy). |
359
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
605 |
The arguments are very similar to ones given above, so we only sketch them. |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
606 |
Let $g$ and $g'$ be two metrics on $X$, and let $e$ and $e'$ be the corresponding |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
607 |
actions $CH_*(X, X) \ot \bc_*(X)\to\bc_*(X)$. |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
608 |
We must show that $e$ and $e'$ are homotopic. |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
609 |
As outlined in the discussion preceding this proof, |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
610 |
this follows from the facts that both $e$ and $e'$ are compatible |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
611 |
with gluing and that $\bc_*(B^n)$ is contractible. |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
612 |
As above, we define a subcomplex $F_*\sub CH_*(X, X) \ot \bc_*(X)$ generated |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
613 |
by $p\ot b$ such that $|p|\cup|b|$ is contained in a disjoint union of balls. |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
614 |
Using acyclic models, we can construct a homotopy from $e$ to $e'$ on $F_*$. |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
615 |
We now observe that $CH_*(X, X) \ot \bc_*(X)$ retracts to $F_*$. |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
616 |
Similar arguments show that this homotopy from $e$ to $e'$ is well-defined |
6224e50c9311
metric independence for homeo action (proof done now)
Kevin Walker <kevin@canyon23.net>
parents:
358
diff
changeset
|
617 |
up to second order homotopy, and so on. |
84 | 618 |
\end{proof} |
619 |
||
620 |
||
368
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
621 |
\noop{ |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
622 |
|
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
623 |
\nn{this should perhaps be a numbered remark, so we can cite it more easily} |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
624 |
|
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
625 |
\begin{rem} |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
626 |
For the proof of xxxx below we will need the following observation on the action constructed above. |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
627 |
Let $b$ be a blob diagram and $p:P\times X\to X$ be a family of homeomorphisms. |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
628 |
Then we may choose $e$ such that $e(p\ot b)$ is a sum of generators, each |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
629 |
of which has support arbitrarily close to $p(t,|b|)$ for some $t\in P$. |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
630 |
This follows from the fact that the |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
631 |
\nn{not correct, since there could also be small balls far from $|b|$} |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
632 |
\end{rem} |
eb7a1ea85179
aborted attempt at remark for small blobs lemma
Kevin Walker <kevin@canyon23.net>
parents:
359
diff
changeset
|
633 |
} |
84 | 634 |
|
357
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
635 |
\begin{prop} |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
636 |
The $CH_*(X, Y)$ actions defined above are associative. |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
637 |
That is, the following diagram commutes up to homotopy: |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
638 |
\[ \xymatrix{ |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
639 |
& CH_*(Y, Z) \ot \bc_*(Y) \ar[dr]^{e_{YZ}} & \\ |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
640 |
CH_*(X, Y) \ot CH_*(Y, Z) \ot \bc_*(X) \ar[ur]^{e_{XY}\ot\id} \ar[dr]_{\mu\ot\id} & & \bc_*(Z) \\ |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
641 |
& CH_*(X, Z) \ot \bc_*(X) \ar[ur]_{e_{XZ}} & |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
642 |
} \] |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
643 |
Here $\mu:CH_*(X, Y) \ot CH_*(Y, Z)\to CH_*(X, Z)$ is the map induced by composition |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
644 |
of homeomorphisms. |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
645 |
\end{prop} |
70 | 646 |
|
357
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
647 |
\begin{proof} |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
648 |
The strategy of the proof is similar to that of Proposition \ref{CHprop}. |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
649 |
We will identify a subcomplex |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
650 |
\[ |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
651 |
G_* \sub CH_*(X, Y) \ot CH_*(Y, Z) \ot \bc_*(X) |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
652 |
\] |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
653 |
where it is easy to see that the two sides of the diagram are homotopic, then |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
654 |
show that there is a deformation retraction of $CH_*(X, Y) \ot CH_*(Y, Z) \ot \bc_*(X)$ into $G_*$. |
70 | 655 |
|
357
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
656 |
Let $p\ot q\ot b$ be a generator of $CH_*(X, Y) \ot CH_*(Y, Z) \ot \bc_*(X)$. |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
657 |
By definition, $p\ot q\ot b\in G_*$ if there is a disjoint union of balls in $X$ which |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
658 |
contains $|p| \cup p\inv(|q|) \cup |b|$. |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
659 |
(If $p:P\times X\to Y$, then $p\inv(|q|)$ means the union over all $x\in P$ of |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
660 |
$p(x, \cdot)\inv(|q|)$.) |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
661 |
|
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
662 |
As in the proof of Proposition \ref{CHprop}, we can construct a homotopy |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
663 |
between the upper and lower maps restricted to $G_*$. |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
664 |
This uses the facts that the maps agree on $CH_0(X, Y) \ot CH_0(Y, Z) \ot \bc_*(X)$, |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
665 |
that they are compatible with gluing, and the contractibility of $\bc_*(X)$. |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
666 |
|
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
667 |
We can now apply Lemma \ref{extension_lemma_c}, using a series of increasingly fine covers, |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
668 |
to construct a deformation retraction of $CH_*(X, Y) \ot CH_*(Y, Z) \ot \bc_*(X)$ into $G_*$. |
bbd55b6e9650
associativity for CH_* action
Kevin Walker <kevin@canyon23.net>
parents:
345
diff
changeset
|
669 |
\end{proof} |