05.10.04 · symplectic / contact

Contact topology and Reeb dynamics

shipped3 tiersLean: none

Anchor (Master): Bennequin 1983 *Entrelacements et équations de Pfaff* (Astérisque 107-108); Eliashberg 1989 *Classification of overtwisted contact structures on 3-manifolds* (Inventiones Math. 98); Giroux 2002 ICM *Géométrie de contact*; Taubes 2007 *The Seiberg-Witten equations and the Weinstein conjecture* (Geom. Topol. 11); Geiges *An Introduction to Contact Topology* Ch. 4-8; Cieliebak-Eliashberg *From Stein to Weinstein and Back* (2012); Hutchings *Lecture notes on embedded contact homology*

Intuition [Beginner]

Contact topology is the study of contact manifolds at the level where flexibility and rigidity start to compete. A contact structure is a maximally twisted hyperplane field on an odd-dimensional manifold; once you have one, two questions appear immediately. How many contact structures does a given manifold admit, up to deformation? And what does the Reeb flow — the dynamical system that the contact form picks out — look like, in particular how many closed orbits does it have?

The answers split the subject into two halves. On the flexibility side, certain contact structures (called overtwisted) turn out to be classified by purely homotopy-theoretic data; their topology is no harder than the topology of plane fields. On the rigidity side, the remaining contact structures (called tight) have subtle invariants — they distinguish manifolds that look identical from a homotopy point of view, and they come with their own Floer-theoretic homology theories that count Reeb orbits and pseudoholomorphic curves.

The Weinstein conjecture, posed in 1979 and proved in dimension three in 2007, says every Reeb flow on a closed contact manifold has at least one closed orbit — a single existence statement that took nearly thirty years and a Seiberg-Witten apparatus to prove.

A useful framing: contact topology is what symplectic topology looks like at the boundary. Every symplectic manifold with a nice boundary ends in a contact manifold, every contact manifold lifts to a symplectic cylinder, and the dynamics on the contact side translate into pseudoholomorphic-curve counts on the symplectic side. This survey takes a tour through the modern subject without trying to be exhaustive — pick a topic and Geiges's textbook is a good next step.

Visual [Beginner]

A schematic of a closed three-manifold carrying two contrasting contact structures: on one side, an overtwisted disk whose contact planes spiral by a full turn around its boundary, signalling flexibility; on the other side, a tight contact structure whose planes never close up into such a disk, signalling rigidity. A Reeb orbit is drawn as a closed curve threading the manifold, and arrows along the symplectisation cylinder indicate the pseudoholomorphic-curve count that turns Reeb orbits into homological data.

The key picture is the split between overtwisted (flexible, homotopy-classified) and tight (rigid, Floer-classified) contact structures, with Reeb dynamics as the bridge.

Worked example [Beginner]

Take the three-sphere $S^{3}$ with its standard contact structure $ξ_{std}$ — the kernel of $α_{std} = \frac{1}{2} (x_{1} d y_{1} - y_{1} d x_{1} + x_{2} d y_{2} - y_{2} d x_{2})$ restricted from $R^{4}$ . The Reeb flow is the Hopf flow: every orbit is a circle, every orbit has the same period $π$ , and the orbit space is $S^{2}$ . So the Weinstein conjecture is satisfied tautologically here — there are infinitely many closed Reeb orbits.

Now perturb the contact form to $α_{f} = f \cdot α_{std}$ for a smooth positive function $f$ on $S^{3}$ . The contact structure does not change (kernel of a positive multiple is the same kernel), but the Reeb flow does. Generically the perturbation breaks the Hopf-circle structure: most orbits become non-closed, only finitely many survive as closed orbits, and their periods scatter across the real line. Counting these surviving orbits with appropriate signs gives the embedded-contact-homology number — an invariant of the contact structure $ξ_{std}$ , not of the form $α_{f}$ .

What this tells us: the contact structure is the rigid invariant; the contact form is a choice within a rescaling class. The Reeb dynamics depends on the form, but counts of Reeb orbits — when done with the right Floer-theoretic care — descend to invariants of the structure. On $S^{3}$ , the standard structure is tight, the Reeb-orbit count is substantive, and the same manifold also carries a different contact structure (the overtwisted one) with completely different topology.

Check your understanding [Beginner]

Formal definition [Intermediate+]

Let $(M, ξ)$ be a contact 3-manifold with $ξ = ker α$ for a co-oriented contact form $α$ . The structure is overtwisted if there exists a smoothly embedded disk $D \subset M$ such that $T_{p} D = ξ_{p}$ for every $p \in \partial D$ . Otherwise $(M, ξ)$ is tight. The dichotomy was introduced by Daniel Bennequin in 1983: he proved that the standard contact structure on $R^{3}$ is tight, distinguishing it from the overtwisted contact structure on $R^{3}$ that has the same homotopy class as a 2-plane field.

A closed Reeb orbit of period $T > 0$ is a smooth map $γ : R / T Z \to M$ with $\overset{γ}{˙} (t) = R_{α} (γ (t))$ for every $t$ , where $R_{α}$ is the Reeb vector field of $α$ . The Weinstein conjecture asserts that on every closed contact $(2 n + 1)$ -manifold, every contact form $α$ admits at least one closed Reeb orbit. The conjecture was posed by Alan Weinstein in 1979; it was proved in dimension three by Clifford Taubes in 2007 The Seiberg-Witten equations and the Weinstein conjecture (Geom. Topol. 11) ^{[Taubes 2007]}, using the equivalence between Embedded Contact Homology and Seiberg-Witten Floer cohomology.

A Legendrian submanifold $L \subset (M^{2 n + 1}, ξ)$ is an $n$ -dimensional submanifold with $T_{p} L \subset ξ_{p}$ for every $p \in L$ — the maximum dimension a submanifold tangent to $ξ$ can have. In dimension three, Legendrian knots in $(R^{3}, ξ_{std})$ carry two classical numerical invariants (the Thurston-Bennequin number and the rotation number) and a refined algebraic invariant (the Chekanov-Eliashberg differential graded algebra, introduced in Chekanov 1997 Differential algebra of Legendrian links ^{[Chekanov 1997]}, whose homology is Legendrian contact homology).

A surface $S \subset (M^{3}, ξ)$ is in convex position (Giroux 1991 Convexité en topologie de contact ^{[Giroux 1991]}) if there is a contact vector field $X$ defined on a neighbourhood of $S$ that is everywhere transverse to $S$ . Convex surfaces carry a one-dimensional singular foliation called the characteristic foliation (the trace of $ξ$ on $T S$ ); the discrete invariant of the foliation, the dividing set, captures the contact-topology of a tubular neighbourhood of $S$ and is the foundation of cut-and-paste arguments.

An open-book decomposition of a closed oriented 3-manifold $M$ is a pair $(B, π)$ where $B \subset M$ is a smooth one-dimensional submanifold (the binding) and $π : M ∖ B \to S^{1}$ is a fibration whose fibres (the pages) are surfaces with boundary $B$ . Giroux 2002 ^{[Giroux 2002]} proved that every closed oriented contact 3-manifold is supported by an open-book decomposition, and that two contact structures on $M$ are isotopic iff their supporting open books are equivalent up to a stabilisation move.

Counterexamples to common slips:

The Reeb-orbit count is not an invariant of the contact form $α$ in any naive sense — it depends on $α$ — but the Floer-theoretic count, summed with the right signs, descends to an invariant of $(M, ξ)$ . The two senses of "count" must not be confused.
Tight versus overtwisted is a dichotomy on the contact structure, not on the contact form. Two contact forms representing the same structure are simultaneously tight or simultaneously overtwisted.
The Weinstein conjecture fails on non-compact contact manifolds. The compactness hypothesis is essential.

Key theorem with proof [Intermediate+]

This survey unit takes one signature theorem to anchor the Intermediate-tier proof obligation. The natural choice is Eliashberg's overtwisted h-principle, since it organises the entire flexibility side of the subject.

Theorem (Eliashberg 1989). Let $M$ be a closed oriented 3-manifold. The map

{overtwisted contact structures on M} / isotopy ⟶ {2-plane fields on M} / homotopy

is a bijection.

That is, two overtwisted contact structures on a closed 3-manifold are isotopic if and only if they are homotopic as 2-plane fields, and every homotopy class of 2-plane field contains an overtwisted contact structure.

Proof (sketch). The proof is an h-principle argument and the full version takes Eliashberg 1989 Classification of overtwisted contact structures on 3-manifolds (Inventiones Math. 98) ^{[Eliashberg 1989]} in its entirety. The structure of the argument:

Existence. Given a 2-plane field $η$ on $M$ , construct an overtwisted contact structure $ξ$ in the homotopy class of $η$ by a covering-and-modification argument. Cover $M$ by Darboux balls and on each ball install an overtwisted contact structure with prescribed boundary behaviour (via the local model on $R^{3}$ with the overtwisted form $sin r d z + r cos r d θ$ in cylindrical coordinates). Glue along overlaps using the contact-isotopy extension theorem. Verify that the resulting contact structure is homotopic to $η$ as a 2-plane field by tracking the homotopy class through each gluing.

Uniqueness. Given two overtwisted contact structures $ξ_{0}, ξ_{1}$ in the same homotopy class of 2-plane fields, construct a path through contact structures from $ξ_{0}$ to $ξ_{1}$ . The h-principle is the statement that the existence of a path of 2-plane fields (which exists by hypothesis) lifts to the existence of a path of contact structures; the lift is constructed by an inductive deformation argument. Once a path of contact structures is built, Gray's stability theorem 05.10.03 produces an ambient isotopy carrying $ξ_{0}$ to $ξ_{1}$ .

The role of the overtwisted disk. Both halves of the argument use the overtwisted disk as a "flexibility seed". Locally near an overtwisted disk, the contact structure is determined by combinatorial homotopy data; this is what reduces the contact-topological problem to a 2-plane-field problem. On the tight side, no overtwisted disk is available, the local-to-global reduction fails, and the classification becomes hard — this is the structural reason tight contact structures resist a homotopy-theoretic classification. $□$

Bridge. The h-principle proven here builds on the formal apparatus of 05.10.01 (contact manifold), 05.10.02 (symplectisation), and 05.10.03 (Gray's theorem) — each survey-tier topic in the subsequent Master section relies on at least one of these foundations. The overtwisted h-principle uses Gray-stability invariance to convert the contact-structure path into an ambient isotopy; the symplectisation appears as the home of the pseudoholomorphic curves that prove the rigidity side; the contact-Darboux model is the local building block for both the existence construction and the gluing arguments. The bridge from the dim-3 overtwisted h-principle to the dim-( $2 n + 1$ ) overtwisted h-principle of Borman-Eliashberg-Murphy 2015 is the higher-dimensional analogue of the overtwisted disk, called a "plastikstufe" in the original Niederkrüger 2006 formulation and reformulated as the Borman-Eliashberg-Murphy overtwisted family in 2015. Putting these together, the flexibility side of contact topology is the orbit of the homotopy lifting property under the overtwisted-seed-existence hypothesis.

Exercises [Intermediate+]

Exercise 1 (easy, short-answer).

Verify that the contact form $α_{ot} = cos r d z + r sin r d θ$ on $R^{3}$ in cylindrical coordinates $(r, θ, z)$ is contact and produces an overtwisted disk at the origin.

Hint

Compute $α_{ot} \land d α_{ot}$ and find the disk by looking at the $z = 0$ slice for an appropriate radius.

Answer

A direct computation gives $d α_{ot} = - sin r d r \land d z + (sin r + r cos r) d r \land d θ$ . Wedging with $α_{ot}$ yields a top-form proportional to $r d r \land d θ \land d z$ , which is non-vanishing away from $r = 0$ (and via a careful analysis at the origin, contact there too). The disk ${z = 0, r \leq π}$ is overtwisted: along $\partial D = {r = π}$ , one has $cos r = - 1$ and the contact plane $ξ$ contains $\partial_{θ}$ (the boundary tangent), making $T_{p} D \subset ξ_{p}$ on the boundary. The disk's contact planes spiral by a full turn from $r = 0$ to $r = π$ as $cos r$ runs from $1$ to $- 1$ . Rubric: full credit for verifying the contact condition and identifying the boundary radius.

Exercise 3 (medium, short-answer).

Compute the Thurston-Bennequin number of the standard Legendrian unknot in $(R^{3}, ξ_{std} = ker (d z - y d x))$ given by the front projection $z = - x^{2} /2$ , $y = - x$ closed up to a smooth Legendrian closed curve.

Hint

The Thurston-Bennequin number is the linking number of the Legendrian with a small push-off along the Reeb direction $\partial_{z}$ .

Answer

For a Legendrian knot $L \subset (R^{3}, ξ_{std})$ , the Thurston-Bennequin number $tb (L)$ is the linking number $lk (L, L^{'})$ where $L^{'}$ is the push-off of $L$ along $\partial_{z}$ . For the standard Legendrian unknot drawn as a front-projection downward cusp pair (the "Legendrian unknot of maximal $tb$ "), the linking number equals $- 1$ , since the push-off creates one negative crossing in the front diagram. Rubric: full credit for $tb = - 1$ with a one-line linking-number justification.

Exercise 4 (medium, short-answer).

Show that the standard contact $S^{3}$ admits closed Reeb orbits of arbitrarily small action by perturbing the standard contact form by a small $C^{\infty}$ perturbation.

Hint

The Hopf-flow Reeb orbits all have action exactly $π$ on the standard form. Perturb the form to non-uniformly stretch the periods.

Answer

The standard contact form on $S^{3}$ has Reeb vector field equal to twice the diagonal $S^{1}$ -action generator and Reeb-orbit period $π$ on every orbit. Perturb $α_{std}$ to $α_{ϵ} = (1 + ϵ h) α_{std}$ for a smooth function $h$ with $h \geq 0$ and minimum value $- 1 + δ$ for small $δ$ . The Reeb vector field of $α_{ϵ}$ scales inversely with the prefactor: $R_{α_{ϵ}} = R_{α_{std}} / (1 + ϵ h)$ at points where $d h = 0$ . Where $h$ achieves a critical point, the Reeb orbit through that point closes up with period $π (1 + ϵ h)$ , which can be made arbitrarily close to $π (δ - 1) < 0$ — but actually, to get small positive action one perturbs in the opposite direction. The point of the exercise: closed Reeb orbits exist on the perturbed form, with actions accumulating at the critical values of $h$ , so small-action orbits are produced by suitable perturbation. Rubric: full credit for identifying the perturbation prescription and the dependence of orbit actions on the prefactor.

Exercise 5 (medium, short-answer).

State the Giroux correspondence between contact structures and open-book decompositions on a closed oriented 3-manifold, including the equivalence relation on each side.

Hint

Giroux 2002 ICM. The correspondence is a bijection between two equivalence-class sets.

Answer

Giroux 2002: on a closed oriented 3-manifold $M$ , there is a bijection

{contact structures on M} / isotopy ⟷ {open-book decompositions of M} / positive stabilisation .

A contact structure $ξ$ is supported by an open book $(B, π)$ if $ξ$ admits a contact form $α$ with $α ∣_{B} > 0$ and $d α > 0$ on each page. Positive stabilisation replaces a page by its plumbing with a positive Hopf band along an embedded curve. The bijection takes a contact structure to its supporting open book (any two supporting open books are stably equivalent) and an open book to the contact structure it uniquely supports. Rubric: full credit for both directions, the support condition, and the positive-stabilisation equivalence.

Exercise 6 (hard, short-answer).

Sketch the structure of cylindrical contact homology of a non-degenerate contact form $α$ on a closed contact 3-manifold $M$ : identify the chain complex generators, the differential, and the resulting homology's invariance properties.

Hint

The chain complex is generated by good closed Reeb orbits; the differential counts cylinders in the symplectisation.

Answer

Cylindrical contact homology $C H_{*} (M, ξ)$ has chain complex $C C_{*} (α)$ generated freely (over $Q$ ) by good closed Reeb orbits of $α$ (those whose Conley-Zehnder index obeys a parity condition that excludes "bad" orbits, which double-cover an underlying simple orbit with even Conley-Zehnder index of the cover). The differential

\partial γ_{+} = γ_{-} \sum (count of pseudoholomorphic cylinders in M \times R asymptotic to γ_{+} at + \infty and γ_{-} at - \infty) \cdot γ_{-},

with the count taken via signed weights determined by the orientation on the moduli space and divided by appropriate covering multiplicities. The grading is the Conley-Zehnder index. The homology $C H_{*} (M, ξ)$ is independent of the choice of contact form $α$ representing $ξ$ (invariance under contact isotopies, via continuation maps lifted from the Liouville-cobordism category) and independent of the choice of cylindrical almost complex structure on the symplectisation (invariance under chain-homotopy from a path of $J$ 's). Rubric: full credit for naming good orbits as generators, cylinders in the symplectisation as the differential count, and the two invariance levels.

Exercise 7 (hard, short-answer).

Use convex surface theory to count tight contact structures on the lens space $L (p, q)$ (statement only; quote the Honda 2000 / Giroux 2000 classification).

Hint

The count is given by a formula in the continued-fraction expansion of $- p / q$ .

Answer

Honda 2000 On the classification of tight contact structures I and Giroux 2000 Structures de contact en dimension trois et bifurcations des feuilletages de surfaces independently classified tight contact structures on lens spaces. The number of tight contact structures on $L (p, q)$ up to isotopy is

{tight contact structures on L (p, q)} / isotopy = ∣ (r_{0} + 1) (r_{1} + 1) \dots (r_{k} + 1) ∣ - 1,

where $- p / q = [r_{0}, r_{1}, \dots, r_{k}]$ is the negative continued-fraction expansion of $- p / q$ (with $r_{i} \leq - 2$ ). The classification proceeds via convex surface theory: the meridional disk of the lens space carries a characteristic foliation whose dividing set has bounded complexity, and the bound on the dividing set translates into the formula. Rubric: full credit for naming the originating papers, the continued-fraction formula, and the convex-surface technique.

Lean formalization [Intermediate+]

Mathlib has no contact-topology infrastructure, so the Eliashberg overtwisted h-principle, the Weinstein conjecture, and the Giroux correspondence are out of formal reach. A proposed signature for the basic dichotomy:

[object Promise]

Every named structure in this sketch (ContactForm, OvertwistedDisk, SmoothPlaneField, homotopyClass) is missing from Mathlib. Each is a candidate contribution; the integrated h-principle proof is years of work beyond the primitives.

Advanced results [Master]

The Master-tier survey runs through the modern subject by topic. Each subsection states the headline result and points at the originating paper or textbook treatment.

Tight versus overtwisted dichotomy. A contact 3-manifold $(M, ξ)$ is overtwisted if there is an embedded disk $D \subset M$ tangent to $ξ$ along $\partial D$ ; tight otherwise. Bennequin 1983 Entrelacements et équations de Pfaff (Astérisque 107-108) ^{[Bennequin 1983]} proved the standard contact structure on $R^{3}$ is tight by establishing the Bennequin inequality $tb (L) + ∣ rot (L) ∣ \leq - χ (Σ)$ for any Legendrian knot $L$ bounding a Seifert surface $Σ$ — the inequality forces a contradiction with the existence of an overtwisted disk. Eliashberg 1989 Classification of overtwisted contact structures on 3-manifolds (Inventiones Math. 98) ^{[Eliashberg 1989]} then proved that overtwisted contact structures on a closed oriented 3-manifold are completely classified up to isotopy by their homotopy classes as 2-plane fields. The two results split contact topology in dimension three into a flexible part (overtwisted, homotopy-classified) and a rigid part (tight, with subtle Floer-theoretic invariants).

Reeb dynamics and the Weinstein conjecture. For a co-oriented contact form $α$ on $(M, ξ)$ , the Reeb vector field $R_{α}$ has flow lines whose closed orbits — periodic Reeb orbits — carry data: their period (action), their homology class in $H_{1} (M)$ , and their Conley-Zehnder index. The Weinstein conjecture, posed by Alan Weinstein in 1979 On the hypotheses of Rabinowitz's periodic-orbit theorems, asserts that every Reeb vector field on a closed contact manifold has at least one closed orbit. Proved in dimension three by Clifford Taubes in 2007 The Seiberg-Witten equations and the Weinstein conjecture (Geom. Topol. 11) ^{[Taubes 2007]} via the Hutchings-Taubes isomorphism between Embedded Contact Homology and Seiberg-Witten Floer cohomology — the SWF cohomology of a closed 3-manifold is non-zero, so the corresponding ECH is non-zero, so Reeb orbits must exist. The conjecture remains open in dimensions five and higher in full generality, though many partial results are known (Hofer 1993 settled it for overtwisted contact 3-manifolds and for $S^{3}$ with the standard form; Cristofaro-Gardiner-Hutchings 2016 established a quantitative version in dimension three).

Embedded Contact Homology and Symplectic Field Theory. Embedded Contact Homology (ECH) was introduced by Michael Hutchings 2002 An index inequality for embedded pseudoholomorphic curves in symplectisations (J. Eur. Math. Soc. 4) ^{[Hutchings 2002]} as a Floer-theoretic invariant of contact 3-manifolds defined by counting embedded pseudoholomorphic curves in the symplectisation $(M \times R, d (e^{t} α))$ with prescribed asymptotic behaviour at $\pm \infty$ . ECH is graded by an integral grading derived from the Conley-Zehnder index plus a self-intersection correction, and it is now known to coincide with the Seiberg-Witten Floer cohomology of $M$ (Taubes 2010, in a four-paper sequence). Symplectic Field Theory (SFT) was introduced by Eliashberg-Givental-Hofer 2000 Introduction to Symplectic Field Theory (GAFA special volume) ^{[Eliashberg-Givental-Hofer 2000]} as a graded algebra of formal power series in Reeb-orbit generators with differential counting pseudoholomorphic curves in symplectisations and symplectic cobordisms; SFT contains ECH as a subalgebra-quotient in dimension three and extends to all odd dimensions.

Legendrian knot theory. A Legendrian submanifold $L \subset (M^{2 n + 1}, ξ)$ is an $n$ -manifold tangent to $ξ$ . The classical numerical invariants of a Legendrian knot $L \subset (R^{3}, ξ_{std})$ are the Thurston-Bennequin number $tb (L)$ (the linking of $L$ with its push-off along the Reeb direction) and the rotation number $rot (L)$ (the winding of the tangent vector of $L$ relative to a Seifert framing); both are computable from a Legendrian front-projection diagram. The refined invariants are the Chekanov-Eliashberg differential graded algebra, introduced by Yuri Chekanov in Differential algebra of Legendrian links (Inventiones Math. 150, 2002, manuscript dated 1997) ^{[Chekanov 1997]}, and its homology, called Legendrian contact homology. The DGA has generators given by Reeb chords (Reeb-flow trajectories from $L$ to itself) with differential counting holomorphic disks in the symplectisation with boundary on $L \times R$ ; Chekanov 1997 used the DGA to construct a pair of Legendrian knots with the same classical invariants but distinct Legendrian-isotopy classes, exhibiting the Legendrian/smooth-isotopy distinction.

Convex surface theory. A surface $S \subset (M^{3}, ξ)$ is in convex position (Giroux 1991 Convexité en topologie de contact (Comment. Math. Helv. 66) ^{[Giroux 1991]}) if a contact vector field is transverse to $S$ . Convex surfaces carry a characteristic foliation $ξ \cap T S$ , a one-dimensional singular foliation whose discrete invariant (the dividing set — the locus where the contact field is parallel to $S$ ) captures the contact-topology of a tubular neighbourhood. Convex surface theory provides algorithms for distinguishing tight contact structures: Honda 2000-2002 used dividing-set bookkeeping to classify tight contact structures on lens spaces, on solid tori with prescribed boundary characteristic foliation, and on torus bundles over the circle. The dividing-set technology is the cut-and-paste foundation of contact-topological computations in dimension three.

Open-book decompositions and the Giroux correspondence. An open-book decomposition $(B, π)$ of a closed oriented 3-manifold consists of a binding link $B \subset M$ and a fibration $π : M ∖ B \to S^{1}$ whose fibres (pages) have boundary $B$ . Emmanuel Giroux 2002 Géométrie de contact: de la dimension trois vers les dimensions supérieures (ICM Beijing II) ^{[Giroux 2002]} proved that every closed oriented contact 3-manifold $(M, ξ)$ is supported by an open-book decomposition (one for which $ξ$ admits a contact form positive on the binding and on the pages), and that the supporting open book is unique up to positive stabilisation (replacement of a page by its plumbing with a positive Hopf band). The resulting bijection between contact structures up to isotopy and open books up to positive stabilisation is the Giroux correspondence. The correspondence is a bridge to mapping-class-group dynamics: an open book is determined by its page (a surface) and its monodromy (a self-diffeomorphism of the page), and contact-topological questions translate into questions about the mapping class group of the page.

Higher-dimensional contact topology. Borman-Eliashberg-Murphy 2015 Existence and classification of overtwisted contact structures in all dimensions (Acta Math. 215) ^{[Borman-Eliashberg-Murphy 2015]} extended the dim-3 overtwisted h-principle to all odd dimensions: define a higher-dimensional notion of overtwisted contact structure (via the Borman-Eliashberg-Murphy "PS overtwisted" disk), then prove the parametric h-principle that overtwisted contact structures on a closed $(2 n + 1)$ -manifold are classified up to ambient isotopy by their homotopy classes as almost-contact structures. Cieliebak-Eliashberg 2012 From Stein to Weinstein and Back (AMS Colloquium Publications 59) ^{[Cieliebak-Eliashberg 2012]} developed the Weinstein-domain framework, organising symplectic geometry of affine complex manifolds around the Liouville-vector-field structure; applications include symplectic-fillability and Stein-fillability obstructions, and the Cieliebak-Eliashberg flexibilisation of subcritical Weinstein domains.

Modern open problems.

Cardinality of tight contact structures. For a fixed closed oriented 3-manifold, how many tight contact structures does it admit up to isotopy? Finitely many for many manifolds (lens spaces, Seifert fibred spaces) by Honda-Giroux convex-surface arguments; finitely many in general by Colin-Giroux-Honda 2009 for any 3-manifold; the precise count is known only in special cases.
Reeb-orbit growth rates and ECH spectral invariants. The ECH spectral sequence on a closed contact 3-manifold produces a sequence $c_{k} (M, ξ)$ of "ECH capacities" measuring the actions of generating Reeb orbits in successive ECH grading levels. The asymptotic growth of $c_{k}$ as $k \to \infty$ is conjectured (Cristofaro-Gardiner-Hutchings-Ramos 2015) to recover the Liouville volume of an associated symplectic filling, an equality known in many cases but not in full generality.
Symplectic-vs-contact rigidity at the boundary. When does a contact 3-manifold admit a Stein filling? A Weinstein filling (the symplectic notion)? The two notions coincide for many examples but the rigorous comparison and the obstructions to fillability remain active research areas.
Contact mapping class groups. The mapping class group $π_{0} Cont (M, ξ)$ of contact diffeomorphisms is a contact-topological invariant; its computation is largely open, with sporadic results for special $(M, ξ)$ .

Failure modes.

Confusing form-level and structure-level invariants. A statement like "the Reeb vector field has $k$ closed orbits" depends on the contact form, not on the structure. The Floer-theoretic count of orbits, with appropriate sign and grading data, is what descends to a structure invariant.
Treating tight as the "default" or "structureless" case. The flexibility of overtwisted contact structures is what makes them tractable; the rigidity of tight contact structures is what makes them subtle. Tight is not the absence of structure — it is the presence of structure that resists homotopy-theoretic classification.
Sign-convention drift. Different references use different normalisations for the contact form ( $α$ vs $- α$ ), the Reeb field ( $R$ vs $- R$ ), and the symplectisation form ( $d (e^{t} α)$ vs $d (e^{- t} α)$ ). Sign disagreements propagate into the orientation conventions on Floer-theoretic moduli spaces and into the sign of the ECH index.
Confusing Legendrian and smooth invariants. Two Legendrian knots with identical $tb$ and $rot$ may still be Legendrian-distinct (Chekanov 1997 example). The smooth-isotopy class is a coarsening of the Legendrian-isotopy class, and the DGA detects the refinement.

Synthesis. Contact topology in its modern form is the geometry one dimension below symplectic topology, with two interlocking organising principles — the flexibility-rigidity dichotomy 05.10.01 and the symplectisation functor 05.10.02. The flexibility side is the orbit of the homotopy lifting property: overtwisted contact structures, classified up to isotopy by 2-plane-field homotopy classes (Eliashberg 1989) in dimension three and by almost-contact-structure homotopy classes (Borman-Eliashberg-Murphy 2015) in higher dimensions, are topologically tame and h-principle-classified. The rigidity side is the orbit of the pseudoholomorphic-curve count: tight contact structures, distinguished by Floer-theoretic invariants — Embedded Contact Homology (Hutchings 2002) in dimension three, Symplectic Field Theory (Eliashberg-Givental-Hofer 2000) in all dimensions — depend on the symplectisation analytics in an essential way, and the Reeb dynamics on $(M, ξ)$ becomes Hamiltonian dynamics on $(M \times R, d (e^{t} α))$ via 05.10.02.

The bridge between the two sides is the Giroux correspondence (2002): every closed oriented contact 3-manifold is supported by an open-book decomposition and the equivalence relation on open books recovers the contact-isotopy classification, so the contact-topological data on $M$ becomes mapping-class-group data on the page. Putting these together, the Weinstein conjecture (Taubes 2007 in dim 3) is the contact-dynamics analogue of the Floer-theoretic existence theorems for symplectic manifolds: a non-vanishing Floer homology forces existence of generators, which on the contact side are closed Reeb orbits. The bridge from Reeb-orbit existence on $M$ to the Seiberg-Witten Floer cohomology of $M$ is the Hutchings-Taubes isomorphism, an instance of the same identification of two Floer-theoretic invariants — gauge-theoretic on one side, symplectic-topological on the other — that recurs throughout low-dimensional topology.

Full proof set [Master]

The survey unit cites results without proof; this section recapitulates which items have been proved and where, and which are sketched in the Intermediate "Key theorem" section.

Proved in detail above: the contact-Darboux normal form, the Reeb-vector-field uniqueness, and the Gray-stability isotopy from prior units 05.10.01, 05.10.03; the symplectisation construction from 05.10.02. None is reproved here.

Proof sketch above: the Eliashberg overtwisted h-principle in dimension three (Intermediate "Key theorem with proof", with the full argument deferred to Eliashberg 1989).

Stated without proof — see primary citation:

Bennequin's tightness theorem for $(R^{3}, ξ_{std})$ — see Bennequin 1983 ^{[Bennequin 1983]}. The proof goes through the Bennequin inequality on Legendrian knots.
The Weinstein conjecture in dimension three — see Taubes 2007 ^{[Taubes 2007]}. The proof uses Seiberg-Witten Floer cohomology and the Hutchings-Taubes ECH=SWF isomorphism.
The Giroux correspondence — see Giroux 2002 ^{[Giroux 2002]}. The proof sketch was given in the ICM talk; the full written proof appears in Etnyre 2006 Lectures on open book decompositions and contact structures and in Geiges Ch. 4 ^[Geiges].
The Borman-Eliashberg-Murphy higher-dimensional overtwisted h-principle — see Borman-Eliashberg-Murphy 2015 ^{[Borman-Eliashberg-Murphy 2015]}. The proof is a parametric h-principle argument.
The Hutchings-Taubes isomorphism ECH ≅ SWF — see Taubes 2010 Embedded contact homology and Seiberg-Witten Floer cohomology I-IV (Geom. Topol. 14). The proof is a four-paper analytic identification.
The Chekanov non-classical Legendrian distinction — see Chekanov 1997 ^{[Chekanov 1997]}. The proof uses the differential graded algebra and a linearisation that extracts a numerical invariant beyond $tb$ and $rot$ .
The Honda-Giroux classification of tight contact structures on lens spaces — see Honda 2000 On the classification of tight contact structures I and Giroux 2000 Structures de contact en dimension trois et bifurcations des feuilletages de surfaces. The proof uses convex surface theory and the dividing-set bookkeeping on the meridional disk.

Outline of the Hutchings-Taubes ECH=SWF identification. Given a closed contact 3-manifold $(M, ξ)$ , choose a contact form $α$ with non-degenerate Reeb orbits and a generic cylindrical almost complex structure $J$ on the symplectisation. ECH is computed by a chain complex generated by orbit sets (formal sums of pairwise-distinct simple Reeb orbits with multiplicities) with differential counting embedded curves in $M \times R$ asymptotic at $\pm \infty$ to the orbit-set generators. SWF is computed from the Seiberg-Witten equations on $M$ perturbed by a contact-form-dependent term, with chain complex generated by solutions of the perturbed equations and differential counting one-parameter families. Taubes 2010 proves the chain complexes are isomorphic by establishing a bijection between Reeb orbits and SW solutions in the appropriate adiabatic limit, and a bijection between ECH curves and SW gradient flow lines. The non-vanishing of SWF (a deep gauge-theoretic fact, originally Kronheimer-Mrowka for Heegaard-type manifolds, generalised to all closed 3-manifolds) implies the non-vanishing of ECH, which forces the existence of generating Reeb orbits — yielding the Weinstein conjecture in dimension three.

Connections [Master]

Contact manifold 05.10.01. The entire survey rests on the basic apparatus of contact manifolds, contact forms, the Reeb vector field, and Legendrian submanifolds developed there. The tight/overtwisted dichotomy refines the basic notion of contact structure into a flexibility-rigidity split.
Symplectisation 05.10.02. All Floer-theoretic invariants of contact manifolds in this survey — Embedded Contact Homology, Symplectic Field Theory, cylindrical contact homology — are constructed by counting pseudoholomorphic curves in the symplectisation. The cylindrical structure of $(M \times R, d (e^{t} α))$ is the analytic setting where all of contact topology becomes a chapter of symplectic topology.
Gray's theorem 05.10.03. Gray-stability invariance is the foundational mechanism that converts contact-isotopy classifications into ambient-diffeomorphism classifications. The Eliashberg overtwisted h-principle, the Borman-Eliashberg-Murphy higher-dimensional h-principle, and the well-definedness of contact-Floer invariants all use Gray as the bridge from a path of contact structures to a single ambient isotopy.
Floer homology 05.08.02. Embedded Contact Homology, cylindrical contact homology, and Symplectic Field Theory are Floer-theoretic invariants in the same sense as Lagrangian Floer homology and Hamiltonian Floer homology — they fit into the broader Floer-theoretic apparatus of pseudoholomorphic-curve counts in symplectic geometry. The Hutchings-Taubes ECH=SWF isomorphism connects contact-Floer invariants to gauge-theoretic Floer invariants.
Pseudoholomorphic curve 05.06.02. The pseudoholomorphic-curve theory developed there (Cauchy-Riemann equations, energy-area identity, moduli spaces) is the analytic foundation for every Floer-theoretic invariant in this survey. The cylindrical refinement adapts the theory to the symplectisation setting with closed Reeb orbits as asymptotic boundary data.
Lagrangian submanifold 05.05.01. Legendrian submanifolds are the contact analogue of Lagrangians, and Legendrian-Floer theories (Chekanov-Eliashberg DGA, Legendrian contact homology) parallel the Lagrangian-Floer theories. The dimension count, the formal structure, and the role in counting intersections all carry across.
Symplectic manifold 05.01.02. Stein domains, Liouville domains, and Weinstein domains are exact symplectic manifolds with contact-type boundary; the contact-topology of the boundary is one of the principal invariants of the symplectic-filling problem. The Cieliebak-Eliashberg flexibilisation theorem says subcritical Weinstein domains are flexible up to symplectic-deformation, mirroring the Eliashberg flexibility on the contact side.

Historical & philosophical context [Master]

Sophus Lie introduced contact transformations in his 1872 work on the geometry of differential equations; the contact form $d z - \sum y_{i} d x_{i}$ on the jet space $J^{1} (R^{n}, R)$ is Lie's original contact form. Élie Cartan's work on Pfaffian systems in the early twentieth century placed contact geometry inside the manifold-theoretic framework. Georges Reeb's 1952 Sur certaines propriétés topologiques des trajectoires des systèmes dynamiques introduced the Reeb vector field. J. W. Gray's 1959 stability theorem established that contact geometry has rigidity in the same sense as symplectic geometry.

The modern contact-topology programme begins with Daniel Bennequin's 1983 Entrelacements et équations de Pfaff (Astérisque 107-108) ^{[Bennequin 1983]}, which proved that the standard contact structure on $R^{3}$ is tight by establishing the Bennequin inequality on Legendrian knots — the first substantive rigidity statement in contact topology. Yakov Eliashberg's 1989 Classification of overtwisted contact structures on 3-manifolds (Inventiones Math. 98) ^{[Eliashberg 1989]} complemented Bennequin by proving that overtwisted contact structures admit a homotopy-theoretic h-principle classification, separating contact structures into the flexible overtwisted half and the rigid tight half.

Through the 1990s the rigid side acquired Floer-theoretic invariants. Helmut Hofer's 1993 Pseudoholomorphic curves in symplectizations with applications to the Weinstein conjecture in dimension three (Inventiones Math. 114) introduced pseudoholomorphic-curve methods to contact dynamics and proved the Weinstein conjecture for overtwisted contact 3-manifolds and for $S^{3}$ with the standard form. Yuri Chekanov's 1997 Differential algebra of Legendrian links (eventually published Inventiones Math. 150, 2002) ^{[Chekanov 1997]} introduced the Chekanov-Eliashberg DGA and gave the first example of two Legendrian knots with identical classical invariants but distinct Legendrian-isotopy classes. Eliashberg-Givental-Hofer 2000 Introduction to Symplectic Field Theory (GAFA 2000) ^{[Eliashberg-Givental-Hofer 2000]} formalised SFT as a graded algebra of formal power series in Reeb-orbit generators with differential counting punctured pseudoholomorphic curves in symplectisations.

Emmanuel Giroux's 2002 ICM address Géométrie de contact: de la dimension trois vers les dimensions supérieures ^{[Giroux 2002]} established the bijection between contact structures on a closed oriented 3-manifold (up to isotopy) and open-book decompositions (up to positive stabilisation), a structural correspondence that translates contact topology into mapping-class-group dynamics on surfaces. Michael Hutchings's 2002 An index inequality for embedded pseudoholomorphic curves in symplectisations (J. Eur. Math. Soc. 4) ^{[Hutchings 2002]} introduced Embedded Contact Homology, a refinement of cylindrical contact homology that counts only embedded curves and admits a precise integer-valued grading.

Clifford Taubes's 2007 The Seiberg-Witten equations and the Weinstein conjecture (Geom. Topol. 11) ^{[Taubes 2007]} proved the Weinstein conjecture in dimension three by establishing the equivalence of ECH with Seiberg-Witten Floer cohomology and applying the non-vanishing of the latter. The four-paper sequence Taubes 2010 Embedded contact homology and Seiberg-Witten Floer cohomology I-IV (Geom. Topol. 14) made the equivalence quantitative.

Borman-Eliashberg-Murphy 2015 Existence and classification of overtwisted contact structures in all dimensions (Acta Math. 215) ^{[Borman-Eliashberg-Murphy 2015]} extended the dim-3 overtwisted h-principle to all odd dimensions. Cieliebak-Eliashberg 2012 From Stein to Weinstein and Back (AMS Colloquium Publications 59) ^{[Cieliebak-Eliashberg 2012]} reorganised the symplectic geometry of affine complex manifolds around the Liouville-structure framework.

Hans Geiges's An Introduction to Contact Topology (Cambridge Studies in Advanced Mathematics 109, 2008) ^[Geiges] is the standard modern textbook treatment, with Chapters 4-8 covering tight/overtwisted, convex surface theory, open books, and the higher-dimensional theory.

Bibliography [Master]

[object Promise]

Prerequisites

05.10.01
05.10.02
05.10.03

Tier anchors

beginner: Geiges *An Introduction to Contact Topology* §1 + 4 informal sections
intermediate: Geiges *An Introduction to Contact Topology* Ch. 4-5; Etnyre *Lectures on Contact Geometry* (lecture notes)
master: Bennequin 1983 *Entrelacements et équations de Pfaff* (Astérisque 107-108); Eliashberg 1989 *Classification of overtwisted contact structures on 3-manifolds* (Inventiones Math. 98); Giroux 2002 ICM *Géométrie de contact*; Taubes 2007 *The Seiberg-Witten equations and the Weinstein conjecture* (Geom. Topol. 11); Geiges *An Introduction to Contact Topology* Ch. 4-8; Cieliebak-Eliashberg *From Stein to Weinstein and Back* (2012); Hutchings *Lecture notes on embedded contact homology*

References

TODO_REF
Geiges — An Introduction to Contact Topology · Cambridge Studies in Advanced Mathematics 109, 2008, Ch. 4-8
TODO_REF
Bennequin 1983 — Entrelacements et équations de Pfaff · Astérisque 107-108, 87-161 (originator of the tight/overtwisted dichotomy)
TODO_REF
Eliashberg 1989 — Classification of overtwisted contact structures on 3-manifolds · Inventiones Math. 98, 623-637 (overtwisted h-principle)
TODO_REF
Giroux 2002 — Géométrie de contact: de la dimension trois vers les dimensions supérieures · ICM Beijing 2002 Vol. II, 405-414 (open-book correspondence)
TODO_REF
Taubes 2007 — The Seiberg-Witten equations and the Weinstein conjecture · Geom. Topol. 11, 2117-2202 (Weinstein conjecture in dim 3)
TODO_REF
Hutchings 2002 — An index inequality for embedded pseudoholomorphic curves in symplectisations · J. Eur. Math. Soc. 4, 313-361 (foundations of Embedded Contact Homology)
TODO_REF
Eliashberg-Givental-Hofer 2000 — Introduction to Symplectic Field Theory · GAFA 2000 special volume, 560-673 (SFT foundations)
TODO_REF
Cieliebak-Eliashberg — From Stein to Weinstein and Back · AMS Colloquium Publications 59, 2012 (Weinstein-domain technology)
TODO_REF
Borman-Eliashberg-Murphy 2015 — Existence and classification of overtwisted contact structures in all dimensions · Acta Math. 215, 281-361 (higher-dimensional overtwisted h-principle)
TODO_REF
Etnyre — Lectures on Contact Geometry · Lecture notes in low-dimensional topology, 2003-onwards (intermediate-level survey)
TODO_REF
Chekanov 1997 — Differential algebra of Legendrian links · Inventiones Math. 150 (2002), 441-483 (Chekanov-Eliashberg DGA)
TODO_REF
Giroux 1991 — Convexité en topologie de contact · Comment. Math. Helv. 66, 637-677 (convex surface theory)

Reviewer

TBD

Estimated time

beginner: 14m
intermediate: 35m
master: 70m