Logarithmic Form¶

A complex meromorphic differential form whose form and exterior derivative have controlled first-order behavior along a specified reduced divisor.

Version: v1 · 2026-10-07 · History
Domain-specific #: 13934
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomain: Complex Geometry → Mathematics

Core Idea¶

A logarithmic form is a complex differential form whose behavior at a specified divisor is controlled in two linked ways. Let \(X\) be a smooth complex analytic space and \(D\) a reduced divisor. Locally write \(D=\{f=0\}\), with \(f\) a reduced equation. A meromorphic \(p\)-form \(\omega\), holomorphic on \(X\setminus D\), is logarithmic along \(D\) when both \(f\omega\) and \(f\,d\omega\) extend holomorphically across \(D\). A regular holomorphic form qualifies too; a nonzero pole or residue is not required.^[1]^[2]

Deligne describes the normal-crossings case through local holomorphic forms and generators \(dz_i/z_i\). Novikov and Yakovenko give the paired extension criterion for an analytic hypersurface and an explicit three-line divisor that is not normal crossings at its triple intersection. The second case shows why the definition cannot be reduced to Deligne's free local normal-crossings basis.^[1]^[2]

Structural Signature¶

Sig role-phrases:

Smooth complex ambient space and reduced divisor: specify \(X\), \(D\), and a local reduced equation \(f\). The divisor is the boundary relative to which logarithmic behavior is tested; its branches may meet singularly.
Individual meromorphic \(p\)-form: identify \(\omega\) on \(X\setminus D\), with meromorphic continuation across \(D\). It is a form, not the whole logarithmic complex or the pair \((X,D)\).
Form extension: require \(f\omega\) to extend holomorphically. This limits the form's pole order but does not alone settle the classification.
Derivative extension: also require \(f\,d\omega\) to extend holomorphically. This prevents a first-order appearance of \(\omega\) from hiding a worse pole in its exterior derivative.^[1]^[2]

The recognition test is local: choose a reduced equation for the divisor and check both products. Replacing \(f\) by another reduced local equation multiplies it by a holomorphic unit, so the two extension tests do not depend on that choice. The conditions concern the allowed maximum singularity; they do not demand an actual pole.^[2]

What It Is Not¶

A logarithmic form is not every meromorphic form with a first-order-looking coefficient. The exterior derivative has its own extension test. For example, \(dx/x^2\) along \(x=0\) fails even the first test: \(x(dx/x^2)=dx/x\) is still polar. A meromorphic form that passes \(f\omega\) but fails \(f\,d\omega\) also lies outside this class.^[2]

Nor does one individual form entail a mixed Hodge structure, a particular cohomology calculation, a holomorphic residue at a singular divisor, or a free basis of \(dz_i/z_i\) at every singular point. Deligne's local generator description assumes normal crossings. The logarithmic de Rham complex is assembled from the sheaves of such forms; its downstream constructions are not constitutive attributes of each member.^[1]^[2]

Scope of Application¶

The identity belongs to complex analytic geometry and its use of meromorphic differential forms along divisors. The present source set supports reduced analytic hypersurfaces in a smooth complex ambient space. Normal-crossings divisors admit Deligne's particularly transparent local generators; singular divisors require the general paired extension test and may behave differently at intersections.^[1]^[2]

An algebraic logarithmic Kähler differential was separately requested in the frozen screening context and redirected toward this term. Its universal-derivation/module scope has not been shown to coincide with the individual analytic form identity here. It remains a held candidate, not an alias. Arithmetic log schemes and p-adic variants are likewise outside what these two inspected sources establish.

Clarity¶

Ask two questions, in order. First, which divisor is specified and which reduced local equation defines it? Second, do both \(f\omega\) and \(f\,d\omega\) extend? Merely seeing \(dx/x\) notation answers neither question for an arbitrary singular boundary. For a smooth branch \(D=\{z=0\}\), \(dz/z\) passes immediately: \(z(dz/z)=dz\) and \(d(dz/z)=0\).^[1]

At a triple intersection, a residue may be singular even though the form is logarithmic. The three-line example below meets both extension tests but Novikov and Yakovenko display its singular residues. Thus “logarithmic” names a controlled differential condition, not a guarantee that every residue is holomorphic.^[2]

Manages Complexity¶

The paired test compresses a large set of local meromorphic expressions into a checkable class. It separates forms suitable for logarithmic differential calculus from arbitrary polar forms without requiring an analyst to enumerate all coordinate expressions. On a normal-crossings divisor, Deligne's generators simplify calculation further. At a singular crossing, the general test remains valid while the normal-crossings basis cannot be carried over automatically.^[1]^[2]

That economy depends on retaining the divisor and derivative condition. If the divisor is omitted, “simple pole” has no specified boundary. If only \(f\omega\) is checked, differentiating may reveal an unallowed singularity. If the normal-crossings basis is asserted at a triple point, a special-case calculation has silently become a false general rule.

Abstract Reasoning¶

The classification is local but controls further operations. On each chart, identify \(D\) by a reduced \(f\), write the candidate meromorphic \(p\)-form, and check the two holomorphic extensions. The exterior derivative \(d\omega\) remains in the logarithmic class under the source's general calculus; Novikov and Yakovenko also state closure under wedge. These are properties of the class, not extra tasks every individual form must visibly perform.^[2]

A counterexample may fail either test. A proof of a positive should show both, even when one is zero. The three-line calculation illustrates this discipline: multiplying the form by \(f\) is easy, but explicitly differentiating prevents a claim based on surface pole appearance alone.

Knowledge Transfer¶

The paired criterion transfers between smooth and singular reduced divisors inside the inspected analytic setting. What changes is the local geometry: one smooth branch supports the \(dz/z\) generator; three concurrent branches cannot be treated as normal crossings at the origin, although the example form still passes the criterion.^[1]^[2]

The abstract idea of controlling an object and its derivative near a boundary may suggest analogies elsewhere. The name “logarithmic form,” the exterior derivative, holomorphic extension, and divisor equation are mathematical cargo; these sources do not justify promoting this entry to a substrate-independent Prime or identifying it with the held Kähler-module redirect.

Examples¶

Smooth branch: punctured disc¶

Take a complex disc \(X\) with coordinate \(z\), let \(D=\{z=0\}\), and set \(\omega=dz/z\) on the punctured disc. This is a direct local specialization of Deligne's normal-crossings generator, not a separately named example in his article.^[1]

Mapped back: ambient and divisor = disc and its origin, \(f=z\); individual form = \(dz/z\); form extension = \(f\omega=dz\); derivative extension = \(f\,d\omega=0\). A genuine simple pole appears, but nonzero pole is not a universal requirement.

Singular crossing: three concurrent lines¶

In \(\mathbb C^2\) with coordinates \(x,y\), let \(D=\{xy(y-x)=0\}\) and \(\omega=(dx/x-dy/y)/(y-x)\). Novikov and Yakovenko's Example 3.6 gives this form for the same three lines, using an equivalent sign convention for their reduced equation. The three branches meet at the origin, so this is not a normal-crossings divisor there.^[2]

Mapped back: ambient and divisor = \(\mathbb C^2\) and the three lines, \(f=xy(y-x)\); individual form = the displayed \(\omega\); form extension = \(f\omega=y\,dx-x\,dy\); derivative extension = \(f\,d\omega=dx\wedge dy\). The last two equalities are direct algebraic checks of the published example. Its singular residues at the triple point warn against importing a holomorphic-residue rule from the normal-crossings case.^[2]

Structural Tensions¶

There is no sourced all-instance conflict of objectives to optimize. The requirement that both a form and its derivative have controlled poles is a paired definition, not a tradeoff. At a singular divisor, local calculation becomes harder, but that difficulty does not change the two constitutive tests.

Structural–Framed Character¶

The mathematical recognition test travels between analytic charts and examples: identify the complex form and divisor, then check the two holomorphic extensions. Its vocabulary does not travel without complex differential forms, reduced divisors and exterior differentiation; calling an unrelated boundary behavior “logarithmic” would import a mathematical classification that has not been established. Choosing which \(X\) and \(D\) to study is a human research decision, while passing \(f\omega\) and \(f\,d\omega\) is determined by the stated local data. No institution, moral judgment or evaluation of usefulness makes an instance a member. Recognition checks the paired condition; importing Deligne's normal-crossings free basis or a holomorphic-residue guarantee into a singular point would assert a different, stronger claim. Its character: predominantly structural within complex geometry, with framing limited to the choice of ambient space, divisor and research purpose rather than the truth of membership.

Structural Core vs. Domain Accent¶

The live Complex Differential Form parent supplies the genus: a complex-valued exterior-form section on \(X\setminus D\), with any type claim governed by its complex structure. The named child adds a reduced divisor \(D\) and the two holomorphic extension tests for \(f\omega\) and \(f\,d\omega\). Those tests, rather than a particular pole or residue, are the structural core. Deligne's \(dz_i/z_i\) local generators belong to the normal-crossings presentation; the singular residues in Novikov and Yakovenko's three-line example belong to that case. Neither is an all-instance requirement. The residue filtration and mixed Hodge consequences are downstream constructions. Removing complex forms, divisors, exterior differentiation and holomorphic extension leaves only an analogy about boundary control. The named Logarithmic Form therefore does not meet a new substrate-independent Prime bar; its portable complex-form skeleton is already represented by the live parent.^[1]^[2]

This entry is a kind of Complex differential form.

A logarithmic form is, in every case, a kind of Complex Differential Form: on \(X\setminus D\), \(\omega\) is a holomorphic complex-valued \(p\)-form, and the extra paired extension conditions narrow that genus. The bidegree discussion under Complex Differential Form is conditional, not a requirement to decompose each instance explicitly. Differential Form is a broad neighboring vocabulary, but its full orientation/integration signature does not establish it as a second, independent broader abstraction here. Logarithmic Pair names an ambient-with-boundary carrier, whereas this entry names a form on that carrier.

Relationships to Other Abstractions¶

Current abstraction Logarithmic Form Domain-specific

Parents (1) — more general patterns this builds on

Logarithmic Form is a kind of Complex differential form Domain-specific

A logarithmic form is a complex differential form on the complement with added divisor-relative extension conditions.

Hierarchy path (1) — routes to 1 parentless root

Logarithmic Form → Complex differential form → Decomposition

Neighborhood in Abstraction Space¶

Logarithmic Form sits in a sparse region of the domain-specific corpus (77^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Sheaves & Birational Geometry (22 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-08

Not to Be Confused With¶

Arbitrary meromorphic form: it may lack one or both required extensions.
Normal-crossings generator: \(dz_i/z_i\) is a useful special local presentation, not a basis theorem at an arbitrary singular divisor.^[1]
A holomorphic residue: residues can be singular at a triple intersection.^[2]
Logarithmic de Rham complex: a graded differential complex formed from the sheaves, not one member form.^[1]^[2]
Logarithmic Kähler differential: the held algebraic derivation-module redirect has not been accepted as an alias.

References¶

[1] Pierre Deligne, Théorie de Hodge II, Publications Mathématiques de l'IHÉS 40 (1971), pp. 5–57; §3.1.2–3, printed p. 31/PDF p. 28 (one-based), local normal-crossings generators and form-plus-derivative criterion; §3.1.5, printed p. 32/PDF p. 29, residue structure. Full original scan inspected on the Institute for Advanced Study host. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l

[2] Dmitry Novikov and Sergei Yakovenko, Lectures on Meromorphic Flat Connections, author-hosted preprint arXiv:math.CA/0212334 v1 (24 December 2002), §3.2 printed/PDF p. 15, paired \(f\omega\), \(f\,d\omega\) criterion; §3.3 printed/PDF p. 17, differential and wedge closure; Example 3.6 printed/PDF p. 18, three-line form and singular residues; Theorem 3.8 printed pp. 18–19, normal-crossings local representation. Full original preprint inspected. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p