Hodge Theory of p p -adic analytic varieties: a survey

Pierre Colmez, Wiesława Nizioł · 2026

Read Hodge Theory of p p -adic analytic varieties: a survey by Pierre Colmez, Wiesława Nizioł (2026) 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-2601-16557.txt.

Introduction

Pierre Colmez and Wiesława Nizioł survey p-adic Hodge theory for analytic forms. Their 2026 preprint sums a field born with Tate in 1967. Tate set a claim of a Hodge like split for p-adic sums. Those sums live on firm analytic forms in the p-adic world. Fontaine then honed that claim in key ways through his work.

The survey traces how the claim grew to a full theory. It maps main proofs, tools and open ends in one place. Such maps let new teams enter a deep field with speed. They also show which claims still lack full proof. The work is a guide note, not a new proof.

Catalog record

CREATORPierre Colmez, Wiesława Nizioł
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE107 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-16557

Browse the full library · About the Tower of Babel

Full text

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

Hodge Theory of p p -adic analytic varieties: a survey Thanks: P.C. and W.N.’s research was supported in part by the Simons Foundation. Pierre Colmez, Wiesława Nizioł Abstract . Hodge Theory of p p -adic analytic varieties was initiated by Tate in his 1967 paper on p p -divisible groups, where he conjectured the existence of a Hodge-like decomposition for the p p -adic étale cohomology of proper analytic varieties. Tate’s conjecture was refined by Fontaine who gave the theory its definite shape. A lot of work has been done for algebraic varieties and a number of proofs of Fontaine’s conjectures have been obtained between years 1985 and 2011. But the study of Hodge Theory of p p -adic analytic varieties started really only in 2011 with Scholze’s proof of Tate’s conjecture using perfectoid methods. Methods that opened the way to an avalanche of results. In this paper, we survey our results and conjectures (comparison theorems and their geometrization, dualities, etc.), focusing on the case of nonproper analytic varieties, where a number of new phenomena occur. We also describe the new objects that appeared along the way. 1. Introduction 1.1. Complex periods 1.1.1. {\mathcal{C}}^{\infty} -manifolds If M is a compact, connected, orientable {\mathcal{C}}^{\infty} -manifold of dimension d , the classical theorem of de Rham states that integrating a closed differential form along a cycle without boundary induces a pairing H^{i}_{{\rm dR},{\mathcal{C}}^{\infty}}(M)\times H_{i}(M,{\bf Z})\to{\bf R} between de Rham cohomology and singular homology, which identifies H^{i}_{{\rm dR},{\mathcal{C}}^{\infty}}(M) with {\rm Hom}(H_{i}(M,{\bf Z}),{\bf R}) . This can be rephrased using Betti cohomology H^{i}_{\rm B}(M,{\bf Q})\simeq{\rm Hom}(H_{i}(M,{\bf Z}),{\bf Q}) by saying that we have a functorial period isomorphism H^{i}_{{\rm dR},{\mathcal{C}}^{\infty}}(M)\simeq{\bf R}\otimes_{\bf Q}H^{i}_{\rm B}(M,{\bf Q}), (1.1) a statements that can be proved, following Weil (see [ 75 ] , Weil 18/01/47, Weil 02/02/47), by means of the {\mathcal{C}}^{\infty} -Poincaré Lemma. Moreover H^{d}_{{\rm dR},{\mathcal{C}}^{\infty}}(M)\simeq{\mathrm{R}} and H^{d}_{\rm B}(M)\simeq{\bf Q} , and the resulting cup-products H^{i}_{{\rm dR},{\mathcal{C}}^{\infty}}(M)\otimes H^{d-i}_{{\rm dR},{\mathcal{C}}^{\infty}}(M)\to{\mathrm{R}} and H^{i}_{\rm B}(M)\otimes H^{d-i}_{\rm B}(M)\to{\bf Q} are perfect dualities (Poincaré dualities). The isomorphism ( 1.1 ) extends to manifolds which are countable at infinity but, since the cohomology groups can now be infinite dimensional, one needs to take a completed tensor product. In this noncompact situation we have at our disposal other cohomology theories, using classes with compact support, and Poincaré duality is now a perfect duality between cohomology and cohomology with compact support. 1.1.2. Complex algebraic and analytic varieties Now, if M=X({\bf C}) is the {\mathcal{C}}^{\infty} -manifold defined by the {\bf C} -points of a proper smooth algebraic variety X defined over a subfield K of {\bf C} , the de Rham cohomology of M can be recovered [ 54 ] from the cohomology of the algebraic de Rham complex {\rm R}\Gamma_{\rm dR}(X):={\mathrm{R}}\Gamma(X,{\mathcal{O}}\to\Omega^{1}\to\Omega^{2}\to\cdots) of X : we have functorial isomorphisms {\bf C}\otimes_{K}H^{i}_{\rm dR}(X)\simeq{\bf C}\otimes_{\bf R}H^{i}_{{\rm dR},{\mathcal{C}}^{\infty}}(X({\bf C})). Combined with the theorem of de Rham from ( 1.1 ), this gives functorial isomorphisms {\bf C}\otimes_{K}H^{i}_{\rm dR}(X)\simeq{\bf C}\otimes_{\bf Q}H^{i}_{\rm B}(X({\bf C}),{\bf Q}). (1.2) The coefficients of the matrix of this isomorphism in bases over K and {\bf Q} respectively are called “periods”; for example 2\pi i is a period of {\bf P}^{1} (for H^{2} ) or of {\bf G}_{m} (for H^{1} , but {\bf G}_{m} is not proper), and \frac{\Gamma(1/4)\Gamma(1/2)}{\Gamma(3/4)} is a period of the elliptic curve y^{2}=x^{3}-x (for H^{1} ). See [ 60 ] for general conjectures about these numbers. The algebraic de Rham cohomology is endowed with a natural filtration (the Hodge filtration) given by the filtration of the de Rham complex: {\rm Fil}^{i}{\rm R}\Gamma_{\rm dR}(X):={\rm R}\Gamma(X,0\to\cdots\to 0\to\Omega^{i}\to\Omega^{i+1}\to\cdots) Hodge theory provides a canonical splitting of this filtration over {\bf C} by means of harmonic forms of type (p,q) , which gives a functorial (Hodge) decomposition {\bf C}\otimes_{\bf Q}H^{i}_{\rm B}(X({\bf C}),{\bf Q})\simeq\oplus_{p+q=i}{\bf C}\otimes_{K}H^{p}(X,\Omega^{q}) (1.3) This decomposition does not, in general, come from a decomposition over K . Most of the above extends to compact complex analytic varieties (taking K={\bf C} ), but the existence of the Hodge decomposition can fail if the variety is not Kähler. 1.2. p -adic periods Now, let p be a prime and let {\bf C}_{p} be the completion of an algebraic closure \overline{\bf Q}_{p} of {\bf Q}_{p} for the p -adic valuation, and {\bf Q}_{p}^{\rm nr}\subset\overline{\bf Q}_{p} be the maximal unramified extension of {\bf Q}_{p} . Let K\subset{\bf C}_{p} be a finite extension of {\bf Q}_{p} , and let G_{K}={\rm Aut}_{\rm cont}({\bf C}_{p}/K)={\rm Gal}(\overline{\bf Q}_{p}/K) be the absolute Galois group of K . The Hodge theory for varieties over K was initiated by Tate in his incredibly influential paper [ 71 ] and given a definitive form by Fontaine (see [ 47 , 48 , 49 ] ) who introduced various rings {\bf B}_{{\rm cris}} , {\bf B}_{{\rm st}} , {\bf B}_{{\rm dR}} , in which the periods of algebraic varieties over K should live. The situation is more complicated here than over {\bf C} : The results over {\bf C} sketched above show that the periods of varieties defined over subfields of {\bf C} live in {\bf C} . But in the p -adic case, as Tate showed, these periods do not, in general, live in {\bf C}_{p} (for example, the p -adic avatar of 2\pi i does not live in {\bf C}_{p} – this is a rephrasing of Tate’s result [ 71 ] that 1 1 1 If r\in{\bf Z} , and V is a G_{K} -module, we let V(r) be its r -th Tate twist: i.e., the module V with action of \sigma\in G_{K} multiplied by \chi^{r}(\sigma) , where \chi:G_{K}\to{\bf Z}_{p}^{\boldsymbol{*}} is the cyclotomic character. H^{0}(G_{K},{\bf C}_{p}(1))=0 ). If X is a smooth and proper algebraic variety defined over K , there is no direct analog of H^{i}_{\rm B}(X({\bf C}),{\bf Q}) ; what plays its role in the p -adic world is the étale cohomology H^{i}_{\operatorname{\acute{e}t}}(X_{{\bf C}_{p}},{\bf Q}_{p}) . Indeed, if X is defined over a subfield F of K that embeds into {\bf C} , and if we choose an embeding into {\bf C} of the algebraic closure \overline{F} of F in {\bf C}_{p} , we have natural isomorphisms (Artin [ 7 ] for the last one) H^{i}_{\operatorname{\acute{e}t}}(X_{{\bf C}_{p}},{\bf Q}_{p})\simeq H^{i}_{\operatorname{\acute{e}t}}(X_{\overline{F}},{\bf Q}_{p})\simeq H^{i}_{\operatorname{\acute{e}t}}(X_{\bf C},{\bf Q}_{p})\simeq{\bf Q}_{p}\otimes_{\bf Q}H^{i}_{\rm B}(X({\bf C}),{\bf Q}) Fontaine’s rings {\bf B}_{{\rm cris}}\subset{\bf B}_{{\rm st}}\subset{\bf B}_{{\rm dR}} are topological rings endowed with compatible continuous actions of G_{K} , and containing a p -adic avatar t of 2\pi i (Fontaine’s 2\pi i ) on which \sigma\in G_{K} acts by \sigma(t)=\chi(\sigma)t . Moreover {\bf B}_{{\rm dR}}^{G_{K}}=K and {\bf B}_{{\rm cris}}^{G_{K}}={\bf B}_{{\rm st}}^{G_{K}}=K\cap{\bf Q}_{p}^{\rm nr} , and: \bullet {\bf B}_{{\rm dR}}={\bf B}_{{\rm dR}}^{+}[\frac{1}{t}] , where {\bf B}_{{\rm dR}}^{+} is complete for the t -adic topology which is stronger that the natural topology (for which \overline{K} is dense [ 26 ] ), and we have {\bf B}_{{\rm dR}}^{+}/t={\bf C}_{p} . The filtration {\bf B}_{{\rm dR}}^{i}:=t^{i}{\bf B}_{{\rm dR}}^{+} is stable by G_{K} . \bullet {\bf B}_{{\rm st}}={\bf B}_{{\rm cris}}[\log\tilde{p}] is endowed with a Frobenius \varphi satisfying \varphi(t)=pt , \varphi(\log\tilde{p})=p\,\log\tilde{p} , and a {\bf B}_{{\rm cris}} -derivation N with N(\log\tilde{p})=-1 (hence N\varphi=p\varphi N ); \varphi and N commute with G_{K} . \bullet {\bf Q}_{p} can be recovered inside {\bf B}_{{\rm st}} using the above structure: we have the fundamental exact sequence 0\to{\bf Q}_{p}\to{\bf B}_{{\rm st}}^{N=0,\varphi=1}\to{\bf

Read the complete text