Equivalence of germs (of mappings and sets) over 𝕜 \Bbbk vs that over 𝕂 \mathbbm{K}

Dmitry Kerner · 2025

Read Equivalence of germs (of mappings and sets) over 𝕜 \Bbbk vs that over 𝕂 \mathbbm{K} by Dmitry Kerner (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-10922.txt.

Introduction

Dmitry Kerner, a solo author, wrote a math paper in 2025 on when two maps count as the same. In singularity theory, maps are sorted by smooth changes of coordinates. A puzzle arises when the field of numbers grows: could two maps match over a big field but not over a small one?

Kerner proves they cannot, in a broad setting. The result covers maps between spaces with rough points, in formal, analytic, and Nash forms. It holds for the classic sorting rules of the field and for any flat extension of rings. One payoff: if a family of maps is trivial over the big field, it is trivial over the small one too. A like result holds for spaces themselves, via the zero sets of maps. The paper also writes down theory for maps with singular targets.

Catalog record

CREATORDmitry Kerner
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE103 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-10922

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Equivalence of germs (of mappings and sets) over 𝕜 \Bbbk vs that over 𝕂 \mathbbm{K} Thanks: I was supported by the Israel Science Foundation, grants No. 1910/18 and 1405/22 Dmitry Kerner Abstract . Consider ℝ \mathbb{R} -analytic mapping-germs, ( ℝ n , o ) → f , f ~ ( ℝ m , o ) . (\mathbb{R}^{n},o)\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\to}\limits^{f,{\tilde{f}}}}$}\ \!(\mathbb{R}^{m},o). They can be equivalent (by coordinate changes) ℂ \mathbb{C} -analytically, but not ℝ \mathbb{R} -analytically. However, if the transformation of ℂ \mathbb{C} -equivalence is I ​ d Id modulo higher order terms, then it implies the ℝ \mathbb{R} -equivalence. On the other hand, starting from ℂ \mathbb{C} -analytic map-germs ( ℂ n , o ) → f , f ~ ( ℂ m , o ) , (\mathbb{C}^{n},o)\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\to}\limits^{f,{\tilde{f}}}}$}\ \!(\mathbb{C}^{m},o), and taking any field extension ℂ ↪ 𝕂 , \mathbb{C}\hookrightarrow\mathbbm{K}, one has: if f ∼ 𝕂 f ~ f\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\sim}\limits^{\mathbbm{K}}}$}\ \!{\tilde{f}} then f ∼ ℂ f ~ . f\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\sim}\limits^{\mathbb{C}}}$}\ \!{\tilde{f}}. These (quite useful) properties seem to be not well known. We prove slightly stronger properties in a more general form: • for Maps ⁡ ( X , Y ) , {\rm Maps}\left(X,Y\right), where X , Y X,Y are (formal/analytic/ 𝕜 \Bbbk -Nash) germs of spaces, with arbitrary singularities, over a base ring 𝕜 ; \Bbbk; • for the classical groups of (right/left-right/contact) equivalence of Singularity Theory; • for faithfully-flat extensions of rings 𝕜 ↪ 𝕂 . \Bbbk\hookrightarrow\mathbbm{K}. In particular, for arbitrary extensions of fields. The case “ 𝕜 \Bbbk is a ring” is important for the study of deformations/unfoldings. E.g. it implies the statement for field extensions: if a family of 𝕜 \Bbbk -maps { f t } \{f_{t}\} is 𝕂 \mathbbm{K} -trivial then it is also 𝕜 \Bbbk -trivial. Similar statements for germs of spaces (“isomorphism over 𝕂 \mathbbm{K} vs isomorphism over 𝕜 \Bbbk ”) follow by the standard reduction “Two maps are contact equivalent if and only if their zero sets are ambient isomorphic”. This study involves the contact equivalence of maps with singular targets, which seems to be not well-established. We write down the relevant part of this theory. 2020 Mathematics Subject Classification Primary 14B05 Secondary 13J05, 13J07, 13J15, 32A05 1. Introduction 1.1. Consider mapping-germs {\rm Maps}\left((\Bbbk^{n},o),\!(\Bbbk^{m},o)\right). Here \Bbbk is an arbitrary field, (\Bbbk^{n},o) is the (algebraic, resp. analytic, formal) germ of the affine space. See § 2 for notations and conventions. These mapping-germs are traditionally studied up to the right ( \mathscr{R} ), left-right ( {\mathscr{L}}\mathscr{R} or \mathscr{A} ), and contact ( {\mathscr{K}}\! ) equivalences. The case of analytic maps, {\rm Maps}\!\left((\mathbb{R}^{n},o),\!(\mathbb{R}^{m},o)\right)\! and {\rm Maps}\!\left((\mathbb{C}^{n},o),\!(\mathbb{C}^{m},o)\right)\!, was intensively studied through 20’th century, see [ Mond. Nuñ.-Bal. ] . Many results of \mathscr{R},{\mathscr{K}}\! -equivalences were extended (in numerous works) from \mathbb{R},\mathbb{C} to arbitrary fields, see Chapter 3 of the second edition of [ Gr.Lo.Sh ] . The extension of {\mathscr{L}}\mathscr{R} -equivalence to arbitrary fields was done in [ Kerner.24 ] . Below \mathscr{G} is one of the groups {\mathscr{L}},\mathscr{R},{\mathscr{L}}\mathscr{R}, {\mathscr{C}},{\mathscr{K}}\!, see § 3.1 . The classical approach is: first to study the complex-analytic situation, then to deduce results in the real-analytic case. Hence the old question is to relate the real and complex equivalence of real map-germs. Suppose f,{\tilde{f}}\!\in\!{\rm Maps}\!\left((\mathbb{R}^{n},o),\!(\mathbb{R}^{m},o)\right)\! are \mathscr{G}_{\mathbb{C}} -equivalent, i.e. their complexifications are equivalent as elements of {\rm Maps}\!\left((\mathbb{C}^{n},o),\!(\mathbb{C}^{m},o)\right)\!. This does not imply their \mathscr{G}_{\mathbb{R}} -equivalence. The simplest example is the Morse critical point, \sum(\pm)x^{2}_{i} . Indeed, the real right equivalence, \mathscr{R}_{\mathbb{R}}, preserves the signature of this quadratic form, (\lambda_{+},\lambda_{0},\lambda_{-}). (Thus e.g. \sum x^{2}_{i}\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\not\sim}\limits^{\mathscr{R}_{\mathbb{R}}}}$}\ \!-\sum x^{2}_{i}. ) The real contact equivalence, {\mathscr{K}}\!_{\mathbb{R}}, can swap the signs of all the eigenvalues, (\lambda_{+},\lambda_{0},\lambda_{-})\leftrightarrow(\lambda_{-},\lambda_{0},\lambda_{+}). But the difference |\lambda_{+}-\lambda_{-}| is still preserved. However one might guess: “The real and complex equivalence of real mapping-germs can differ only at low orders.” Namely, if {\tilde{f}}=g_{\mathbb{C}}f, where the element g_{\mathbb{C}}\!\in\!\mathscr{G}_{\mathbb{C}} is identity modulo higher order terms, then f\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\sim}\limits^{\mathscr{G}_{\mathbb{R}}}}$}\ \!{\tilde{f}}. Even if one studies {\rm Maps}\!\left((\mathbb{C}^{n},o),\!(\mathbb{C}^{m},o)\right)\!, occasionally one has to extend the base-field. For example, some proofs go via deformation arguments, i.e. additional parameters are introduced, \mathbb{C}\hookrightarrow\mathbb{C}(t)=:\mathbbm{K}. Then, having established the \mathscr{G}_{\mathbbm{K}} -equivalence, one would like to deduce the \mathscr{G}_{\mathbb{C}} -equivalence. The natural guess is: “This should work, because \mathbb{C} is algebraically closed”. 1.2. The results ​We convert these guesses into stronger statements in more general situation: • For {\rm Maps}\left(X,Y\right) whose source X\!\subseteq\!(\Bbbk^{n},o) and target Y\!\subseteq\!(\Bbbk^{m},o) are (formal or analytic or algebraic) germs over a ring \Bbbk, and with arbitrary singularities. • For a large class of ring extensions \Bbbk\!\subset\!\mathbbm{K}. (In particular for arbitrary field extensions.) The precise statement is in Theorem 5.1 , for faithfully flat ring extensions. Now we give just the conclusion for (arbitrary) field extensions \Bbbk\!\subset\!\mathbbm{K}. 1. (For char(\Bbbk)=0 ) If {\tilde{f}}\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\sim}\limits^{\mathscr{G}_{\mathbbm{K}}}}$}\ \!f by an element whose linear part is unipotent, then {\tilde{f}}\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\sim}\limits^{\mathscr{G}}}$}\ \!f. 2. Suppose the base-field is algebraically closed, of any characteristic. If {\tilde{f}}\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\sim}\limits^{\mathscr{G}_{\mathbbm{K}}}}$}\ \!f then {\tilde{f}}\!\ \!\lower 1.29167pt\hbox{$\mathrel{\mathop{\sim}\limits^{\mathscr{G}}}$}\ \!\!f. Part 1 is sharp, as we show by explicit examples in § 5.1.1 . While part 2 is “naturally expected”, part 1 seems “a bit unexpected”. (At least for some people in Singularities.) Below we list some corollaries of Theorem 5.1 and other results. • Working over base rings (not just base fields) is important for deformations and unfoldings. E.g. it implies, Corollary 5.6 , for char(\Bbbk)=0 : Let \Bbbk\!\subset\!\mathbbm{K} be any extension of fields. If a family f_{t} is \mathscr{G}_{\mathbbm{K}} -trivial, then it is also \mathscr{G} -trivial. (Roughly: the \Bbbk -part of the \mathscr{G}_{\mathbbm{K}} -constant stratum locally coincides with the \mathscr{G}_{\Bbbk} -constant stratum.) This seems to be new even in the classical case of \mathbb{R}\!\subset\!\mathbb{C}. Example 5.7 . f\in{\rm Maps}\left(X,Y\right) is stable as a \Bbbk -map if and only if it is stable as a \mathbbm{K} -map. • The “ \Bbbk vs \mathbbm{K} question” is finitely determined. Namely, if two maps f,{\tilde{f}}\!\in\!{\rm Maps}\left(X,Y\right) are \mathscr{G}_{\mathbbm{K}} -equivalent, then their \mathscr{G} -equivalence can be verified at a finite jet. (Here f,{\tilde{f}} are not assumed finitely determined.) More precisely, Corollary 5.9 reads in char=0 : 1. Every map f\!\in\!{\rm Maps}\left(X,Y\right) satisfies: \mathscr{G}_{\mathbbm{K}}f\cap(\{f\}+{\mathfrak{m}}^{N}\cdot R^{\oplus m}_{X})\!\subset\!\mathscr{G}f for some N\!\in\!\mathbb{N}. 2. Moreover, this number N can be chosen uniformly for the whole \mathscr{G}_{\mathbbm{K}} -orbit of f . The smallest such N is of course crucial for various applications, but the proof is based on Artin’s \beta -function, for which not many bounds are known. • Though the \mathscr{G}_{\mathbbm{K}} -equivalence does not imply the \mathscr{G} -equivalence, we have at least the finiteness of \mathscr{G} -orbits. Corollary 5.10 . The real part of the complex orbit, \mathscr{G}_{\mathbb{C}}f\cap{\rm Maps}\left(X,Y\right), splits into a finite number of real orbits, \amalg\mathscr{G}_{\mathbb{R}}f_{\alpha}. In fact Corollary 5.10 is for finite extensions of “fields of type F ”, e.g. \mathbb{R} or p -adic fields. • The classical group-actions on spaces of map-germs, \mathscr{G}\circlearrowright{\rm Maps}\left(X,\!(\Bbbk^{m},o)\right), for \mathscr{G}={\mathscr{L}},\mathscr{R}, {\mathscr{L}}\mathscr{R},{\mathscr{C}},{\mathscr{K}}\!, are well-studied. Maps with singular targets, {\rm Maps}\left(X,Y\right), were studied only in rather particular cases, e.g. [ Mond-Montaldi ] , [ Nuñ.-Ball. Tomazella ] . We could not find the general references for the actions \mathscr{G}\circlearrowright{\rm Maps}\left(X,Y\right) with singular targets. The relevant basic notions are set-up in § 3 . Giving the full exposition of this topic “ \mathscr{G}\circlearrowright{\rm Maps}\left(X,Y\right) ” would (at least) double the paper. Therefore

Read the complete text