Independent scholarship · Open accessThe monthly digest

Indexed text attachment

differential-topology-in-higher-categories.tex

Source for Differential Topology in Higher Categories: Mapping Objects, Derived Intersections, Spaces of Choices, and Cobordism

View paper · Download original file

Lines 211–310 of 3851. Source text is displayed literally.

211\noindent\begin{tikzpicture}[x=1cm,y=1cm]212  \draw[thick,fwd] (-0.2,0) .. controls (1.2,1.5) and (2.4,-1.5) .. (3.8,0.2);213  \draw[thick,bwd] (-0.2,0.25) -- (3.8,-0.1);214  \node[dec] at (0.07,0.23) {};215  \node[dec] at (1.83,0.07) {};216  \node[dec] at (3.63,-0.08) {};217  \node[lab] at (4.6,0.05) {$=$};218  \node[lab] (P) at (5.6,0.7) {$M\times_N P$};219  \node[lab] (M) at (5.6,-0.7) {$M$};220  \node[lab] (Q) at (7.8,0.7) {$P$};221  \node[lab] (N) at (7.8,-0.7) {$N$};222  \draw[fw,black] (P) -- (M);223  \draw[fw,black] (P) -- (Q);224  \draw[fw,black] (M) -- (N);225  \draw[fw,black] (Q) -- (N);226  \draw[thick] (6.0,0.25) -- (6.25,0.25) -- (6.25,0.45);227\end{tikzpicture}\par228\vspace{0.5in}229\noindent{\large \bookauthor}\par230\vspace{4pt}231\noindent{\small 2026}232\end{titlepage}233 234\thispagestyle{empty}235\vspace*{\fill}236{\small237\noindent\textbf{About this text.}238This monograph is a higher-categorical treatment of the territory mapped by Morris~W.~Hirsch's239\emph{Differential Topology} (Graduate Texts in Mathematics~33, Springer, 1976): manifolds and maps, function240spaces, transversality, vector bundles and tubular neighbourhoods, degrees and intersection numbers, Morse241theory, cobordism, isotopy, and the classification of surfaces. It is an independent work. It reproduces none242of that book's text or proofs verbatim; it is not affiliated with or endorsed by its publisher.243\Cref{app:concordance} gives a concordance so that the two can be read side by side.244 245\medskip246\noindent\textbf{Status of results.}247Every numbered statement carries a tag: \textsf{classical} (standard differential topology, attributed),248\textsf{literature} (established in the categorical, homotopical or derived-geometric literature, attributed),249\textsf{reformulation} (a known result restated in a different language, with the restatement proved or250sketched here), \textsf{observation} (an elementary remark we have not found stated in this form), or251\textsf{programme} (a conjecture or a direction, not a result). No cited theorem is claimed as new.252 253\medskip254\noindent\textbf{Authorship and assistance.}255The text was drafted and typeset with a large language model (Anthropic's Claude) under the direction of256the submitting author, who is responsible for it. Generative AI assisted with drafting, literature recall,257mathematical consistency checks, computation, and typesetting. The numerical examples and a number of finite258structural claims are recomputed by the accompanying script \texttt{verify-differential-topology.py} (\cref{app:computations}).259Proofs and bibliographic details have not been independently refereed; corrections are welcome.260 261\medskip262\noindent Open access. Rights remain with the author.263}264\clearpage265 266\tableofcontents267 268\chapter{Preface}269 270Hirsch's \emph{Differential Topology} \cite{Hir76} is a book about what can be done to a smooth map by moving it a271little. Its method is uniform across nine chapters. One topologizes the set $C^r(M,N)$ of maps; one shows272that the maps with a desirable property (immersions, embeddings, maps transverse to a submanifold, Morse273functions) form an open and dense, or at least residual, subset; one defines an invariant using a map in274that subset; and one shows that the invariant does not depend on which map was used, because any two such275maps are joined by a path that is itself generic one dimension up. Degree, intersection number, Euler276characteristic, the cobordism class of a preimage, and the diffeomorphism type of a surface are all obtained277in this way.278 279Each step of that method has a counterpart in the language of higher categories, and the counterparts are280by now theorems in the literature and no longer analogies. The set of maps is the object of global points of281a mapping object in a topos of smooth sets, whose homotopy type is its \emph{shape}; the classification282theorems of the subject (for vector bundles, for immersions, for manifolds up to cobordism) compute such283shapes. Transversality is the condition under which a fibre product of manifolds exists and is correct; the284category of manifolds has a universal completion in which every fibre product is correct, the285$\infty$-category of derived manifolds, and the genericity theorems say that every object of the completion286can be deformed back into the original category. Independence of choices is the statement that a space of287choices (of tubular neighbourhoods, collars, metrics, embeddings in $\R^\infty$) is contractible, which is288what makes a construction functorial. Invariants are traces. Decompositions by Morse functions are289presentations by generators and relations of cobordism categories.290 291This book follows Hirsch's order and states each of his main theorems in that language, with attribution. It292is not a replacement for his book, and it does not reprove his analytic theorems. The inverse function293theorem, partitions of unity, the Morse--Sard theorem, and the existence and uniqueness of solutions of294ordinary differential equations are used as stated. What changes is the form of the conclusions, and the last295chapter says plainly how much that change is worth. In brief: of the three volumes in this series so far,296this is the one in which the higher-categorical statements carry the most real content, because the297subject is where a good part of higher category theory came from; and it is also the one in which the298gap between a formal statement and its proof is widest, because the proofs are analysis.299 300A companion volume, \emph{Parametrized Neural Networks and Their Representations in Differential Topology},301applies the results of this book to smooth families of maps given by neural network architectures. It was302written to be read after this one and cites it by chapter.303 304\chapter{Introduction}305 306\section*{What is assumed}307 308The reader should know the definition of a smooth manifold and of a smooth map, the inverse function309theorem, and the language of categories, functors, limits and adjunctions. Sheaves on a site, simplicial310sets, and the idea of an $\infty$-category as a category with spaces of morphisms are used from