--- title: 'Legendrian Self-Shrinkers: Rigidity & Classification' url: https://www.emergentmind.com/topics/legendrian-self-shrinkers type: topic --- # Legendrian Self-Shrinkers: Rigidity & Classification Legendrian self-shrinkers are self-similar shrinking solutions of Legendrian mean curvature flow in contact or Sasakian geometry. Their defining feature is that the shrinker equation is not the ordinary Euclidean condition \(H+\frac12X^\perp=0\), because preservation of the Legendrian constraint requires a correction by the Legendrian angle or phase and the Reeb field. In current formulations this leads to equations such as \(H-\theta\xi=-\frac12X^\perp\), \(H+\theta\xi=\alpha F^\perp\) with \(\alpha<0\), or \(H-2\alpha T=-F^\perp\), depending on the normalization and ambient conventions. Recent work has established a Bernstein-type rigidity theorem for entire smooth graphical Legendrian self-shrinkers in the standard contact Euclidean space, explicit low-dimensional families in \(\mathbb R^3\) and \(\mathbb R^5\), and a curvature-pinched rigidity theorem singling out the generalized Legendrian Clifford torus and the Harvey–Lawson special Lagrangian cone [2508.07900] [2508.15279] [2312.15996]. ## 1. Contact and Sasakian geometric setting In the Sasakian formulation, a Legendrian submanifold is a maximally isotropic \(n\)-submanifold \(F:\Sigma^n\to M^{2n+1}\) satisfying \(F^*\eta=0\), equivalently \(\eta|_{T\Sigma}=0\). For a Sasakian manifold \((M^{2n+1},\eta,T,g,\Phi)\), the standard splitting along a Legendrian \(\Sigma\) is \[ TM|_\Sigma=T\Sigma\oplus \Phi T\Sigma\oplus \mathbb RT, \] so the normal bundle has a distinguished \(\Phi T\Sigma\)-part together with the Reeb direction. In dimension \(2\), this becomes \(N\Sigma=\Phi(T\Sigma)\oplus \mathbb RT\), a decomposition that is used directly in the shrinker equation and in blow-up analysis [2312.15996]. The principal Euclidean model is the standard contact Euclidean \((2n+1)\)-space \[ (\mathbb R^{2n+1},\eta,\Phi,\xi,g), \] with coordinates \((x_1,\dots,x_n,y_1,\dots,y_n,z)\) and \[ \eta=\frac12\,dz-\frac14\sum_{i=1}^n (y_i\,dx_i-x_i\,dy_i),\qquad \xi=2\frac{\partial}{\partial z}, \] \[ g=\frac14\sum_{i=1}^n(dx_i^2+dy_i^2)+\eta\otimes\eta. \] In this model the papers use an adapted orthonormal frame \(E_i,E_{n+i},\xi\), the explicit Levi-Civita identities for \(\bar\nabla\), and the almost contact tensor \(\Phi\) satisfying \(\Phi(E_i)=E_{n+i}\) and \(\Phi(\xi)=0\) [2508.07900]. A structural bridge to symplectic geometry is furnished by the Lagrangian projection \[ \Pi:(\mathbb R^{2n+1},\Phi,\xi,\eta,g)\to (\mathbb C^n,J,\omega_{\mathrm{can}}), \] obtained by forgetting the \(z\)-coordinate. Since \(\Pi^*(\omega_{\mathrm{can}})=\frac12 d\eta\) and \(\Pi_*(\xi)=0\), a Legendrian immersion projects to a Lagrangian immersion. This projection is central in the low-dimensional rigidity theory, where Legendrian shrinkers are reduced to Lagrangian shrinkers in \(\mathbb C^2\) [2508.15279]. ## 2. Flow equations, normalizations, and the graphical formulation Legendrian self-shrinkers are tied to Legendrian mean curvature flow rather than to ordinary MCF. In the graphical Euclidean setting of the rigidity theorem, the modified Legendrian flow is \[ \frac{\partial X_t}{\partial t}=H-\theta\,\xi, \] because the usual mean curvature flow does not preserve the Legendrian condition. A self-similar shrinking solution satisfies \[ \Sigma_t=\sqrt{-t}\,\Sigma_{-1},\qquad t<0, \] and the corresponding elliptic equation is \[ H-\theta\xi=-\frac12 X^\perp. \] For an entire Legendrian graph over \(\mathbb R^n\), the immersion is parametrized by a global smooth potential \(u:\mathbb R^n\to\mathbb R\) as \[ X(x)=\Bigl(x,Du(x),u(x)-\frac12 x\cdot Du(x)\Bigr), \] with tangent vectors \[ e_i=\frac12\bigl(E_i-u_{ki}E_{n+k}\bigr), \] induced metric \[ g_{ij}=\frac14\bigl(\delta_{ij}+u_{ki}u_{kj}\bigr), \] and Legendrian phase \[ \theta=\sum_{i=1}^n \arctan \lambda_i, \] where \(\lambda_i\) are the eigenvalues of \(D^2u\). The shrinker equation reduces to the scalar PDE \[ \sum_{i=1}^n \arctan\lambda_i(x) = \frac14\left(u(x)-\frac12 x\cdot Du(x)\right), \] and “entire smooth” means precisely that the Legendrian submanifold is a global smooth graph over all of \(\mathbb R^n\) [2508.07900]. A frequent source of confusion is that the literature uses different sign and scaling conventions. In the Sasakian blow-up framework, the flow is written \[ \partial_t F_t = H-2\alpha T, \] with \(H=-\nabla^k\alpha\,U_k\), and self-shrinkers are defined by \[ H-2\alpha T=-F^\perp. \] In the later classification and rigidity paper on \(\mathbb R^3\) and \(\mathbb R^5\), the convention is instead \[ H+\theta\xi=\alpha F^\perp,\qquad \alpha<0 \] for shrinkers, and the main rigidity theorem is stated in the normalization \[ H+\theta\xi=-F^\perp. \] These formulas should therefore be read as exact paper-specific normalizations rather than as a single universal convention [2312.15996] [2508.15279]. ## 3. Weighted structure and optimal volume growth The decisive analytic innovation in the graphical theory is a weighted-volume framework adapted to the Legendrian vertical direction. Following the self-shrinker technology of Colding–Minicozzi, the standard drifted Laplacian is \[ \mathcal L=\Delta-\frac12\langle X,\nabla(\cdot)\rangle = e^{\frac{|X|^2}{4}}\operatorname{div}\!\left(e^{-\frac{|X|^2}{4}}\nabla(\cdot)\right), \] with Gaussian weight \(e^{-|X|^2/4}d\mu\). However, for Legendrian graphs this is not the correct proper function for the volume argument. Introducing \[ V:=u-\frac12 x\cdot Du,\qquad f:=\frac14\left(|X|^2-\frac14V^2\right), \] the paper shows that the relevant weighted measure is \(e^{-f}d\mu\) rather than \(e^{-|X|^2/4}d\mu\) [2508.07900]. The geometric input is the decomposition of the position vector \[ X=\frac12 x_iE_i-\frac12 u_iE_{n+i}+\frac12 V\xi, \] hence \[ \langle X,\xi\rangle=\frac12 V. \] Using the shrinker equation, one obtains the fundamental identity \[ V=4\theta. \] The modified shrinker potential then satisfies \[ \nabla f=\frac12 X^T,\qquad \Delta f=\frac n2+\frac12\langle X,H\rangle, \] and with \[ \Delta_f:=\Delta-\langle \nabla f,\nabla\cdot\rangle \] one gets \[ \Delta_f f+f=\frac n2,\qquad |\nabla f|^2\le f. \] A further direct computation gives \[ f=\frac1{16}(x_i^2+u_i^2)\ge0, \] so \(f\) is nonnegative and proper [2508.07900]. These identities allow the use of Cheng–Zhou’s theorem for shrinker-type weighted manifolds. If \[ D_r:=\{p\in\Sigma^n\mid 2\sqrt f\le r\}, \] then an entire graphical Legendrian self-shrinker satisfies \[ \int_{\Sigma^n} e^{-f}\,d\mu<\infty, \qquad \operatorname{Vol}(D_r)\le C_1r^n,\quad r\ge1. \] The paper calls this estimate optimal volume growth, because the exponent \(n\) matches the natural Euclidean growth rate for an \(n\)-dimensional entire graph. In parallel, the Legendrian phase is shown to satisfy the weighted harmonic equation \[ \mathcal L\theta=0, \] which becomes the key rigidity input [2508.07900]. ## 4. Bernstein-type rigidity for entire smooth graphical shrinkers The main rigidity theorem states that every entire smooth solution of \[ \sum_{i=1}^n \arctan \lambda_i(x) = \frac14\left(u(x)-\frac12 x\cdot Du(x)\right) \quad\text{in }\mathbb R^n \] is quadratic: \[ u(x)=u(0)+\frac12\langle D^2u(0)x,x\rangle. \] Equivalently, every entire smooth Legendrian self-shrinking graph in the standard contact Euclidean space is globally classified by a quadratic potential, so there are no nontrivial entire smooth graphical Legendrian self-shrinkers [2508.07900]. The proof is an integral rigidity argument of Bernstein type. Since \[ \mathcal L\theta=0, \] one chooses a cutoff \(\eta\in C^\infty(\Sigma)\) with \(\eta\equiv1\) on \(D_r\), \(\eta\equiv0\) outside \(D_{2r}\), and \(|\nabla\eta|\le C_2/r\), multiplies the weighted harmonic equation by \(\theta\eta^2 e^{-|X|^2/4}\), and integrates by parts. Using Cauchy–Schwarz and Young’s inequality yields \[ \int_\Sigma |\nabla\theta|^2\eta^2 e^{-\frac{|X|^2}{4}} \le 4\int_\Sigma \theta^2|\nabla\eta|^2 e^{-\frac{|X|^2}{4}}. \] Because \[ |\theta|\le \frac{n\pi}{2}, \] and because the polynomial volume bound is coupled with Gaussian decay, the right-hand side tends to \(0\) as \(r\to\infty\). The conclusion is \[ \nabla\theta\equiv0, \] so the phase is constant [2508.07900]. Once \(\theta\) is constant, the scalar shrinker equation reduces to a homogeneity constraint on \(u\): \[ \theta=\frac14\left(u-\frac12 x\cdot Du\right). \] Differentiating twice shows that each second derivative \(u_{ij}\) is homogeneous of degree \(0\). Smoothness at the origin then forces \(D^2u\) to be constant, and \(u\) is quadratic. This theorem is presented as a Bernstein-type rigidity result for entire smooth Legendrian self-shrinkers and as the first general rigidity theorem of this type in the Legendrian contact setting [2508.07900]. ## 5. Low-dimensional models, partial classification, and the generalized Clifford torus A separate line of work treats low-dimensional Legendrian self-shrinkers in \(\mathbb R^3\) and \(\mathbb R^5\). In dimension \(1\), the paper gives a Legendrian analogue of the Abresch–Langer family. For an immersed curve \[ \tilde{\gamma}(t)=\left(y(t)\cos\theta(t)+x(t)\sin\theta(t),\, y(t)\sin\theta(t)-x(t)\cos\theta(t),\,-\frac{B}{2}\theta\right), \] with \(x(t)=\kappa(t)\) and \[ \frac{d}{dt}x(t)=x(t)y(t),\qquad \frac{d}{dt}y(t)=B-x^2(t),\qquad B>0, \] one has \[ \kappa \hat N+B\theta(t)\xi=-4\tilde{\gamma}^\perp. \] When \(B=1\), this gives a self-shrinker for the Legendre curve shortening flow \[ \tilde{\gamma}_t=\kappa \hat N+\lambda(t)\xi. \] The construction is explicit and is presented as the natural Legendrian self-shrinking counterpart of the classical Abresch–Langer description [2508.15279]. In dimension \(2\), the classification is partial and assumes harmonic Legendrian angle with respect to the transverse metric \(g^T\). Under that hypothesis, an immersed Legendrian self-shrinker \(F:M^2\to(\mathbb R^5,\Phi,\xi,\eta,g)\) is embedded and locally congruent to one of four explicit models: a cylinder-type model \(\tilde C\), a torus-type model \(\tilde T\), and two one-parameter families \(\tilde\Upsilon_\gamma\) and \(\tilde\Psi_\nu\). All of these satisfy \[ H+\theta\xi=8aF^\perp,\qquad a<0. \] The paper also proves that there is no Legendrian self-shrinker in \((\mathbb R^5,\Phi,\xi,\eta,g)\) with spherical topology [2508.15279]. The main low-dimensional rigidity theorem is a Legendrian analogue of the Li–Wang characterization of the Clifford torus. If \[ F:\Sigma^2\to(\mathbb R^5,\Phi,\xi,\eta,g) \] is an orientable Legendrian self-shrinker satisfying \[ H+\theta\xi=-F^\perp, \] if the associated Legendrian immersion \[ \bar F(\Sigma)\subset \mathbb R^4\times \mathbb S^1 \] is compact, and if \[ \|A\|_g^2\le2, \] then equality holds, \[ \|A\|_g^2=2, \] and \[ F(t,s)=(2e^{it},2e^{is},-2(t+s)), \qquad \bar F(t,s)=(e^{it},e^{is},e^{-i(t+s)}). \] The compact lift \(\bar F\) is a flat minimal generalized Legendrian Clifford torus in \(\mathbb S^5\), and its cone is the Harvey–Lawson special Lagrangian cone in \(\mathbb C^3\). The proof proceeds by projecting \(F\) to a compact orientable Lagrangian self-shrinker in \(\mathbb C^2\), applying Lagrangian rigidity at the threshold \(2\), and then lifting the resulting Clifford torus back to the Legendrian setting [2508.15279]. ## 6. Singularity models, Lagrangian analogies, and current scope Legendrian self-shrinkers also enter the singularity theory of Legendrian mean curvature flow. In a 5-dimensional Sasaki-Einstein manifold, the flow \[ \partial_t F_t=H-2\alpha T \] preserves the Legendrian condition. After isometric embedding into Euclidean space and applying parabolic blow-up at a type I singularity, one obtains a smooth ancient limit satisfying \[ H-2\alpha T=\frac{1}{2s}F^\perp,\qquad -\infty\delta>0, \] type I singularities are excluded, and compact oriented Legendrian surfaces with \(*\omega>0\) have long-time smooth Legendrian mean curvature flow. In this framework, self-shrinkers are singularity models exactly in the Huisken sense, but with the Legendrian correction term built into the velocity [2312.15996]. An important scholarly distinction is that the same Sasakian-flow paper only announced, rather than proved in full detail, a future rigidity and classification theory for \(2\)-dimensional Legendrian self-shrinkers in \(\mathbb R^5\), including partial classification under harmonic Legendrian angle and the case of constant \[ S=|A|^2. \] It also announced a reconstruction of the Harvey–Lawson special Lagrangian cone from a Legendrian self-shrinker. Those results were not stated there as complete theorems with full proofs. This later became a point requiring careful bibliographic separation between proved blow-up results and subsequently established rigidity/classification statements [2312.15996]. The modern Legendrian theory is explicitly modeled on the Lagrangian self-shrinker literature. The graphical rigidity theorem in the contact Euclidean space is compared with rigidity results for Lagrangian self-shrinking graphs due to Chau–Chen–Yuan and Ding–Xin, while the \(\mathbb R^5\) curvature-pinched rigidity theorem is described as an analogue of Li–Wang. On the Lagrangian side, a complete classification is available for \(2\)-dimensional complete Lagrangian self-shrinkers in \(\mathbb R^4\) with constant squared norm of the second fundamental form: \[ \mathbb R^2,\qquad S^1(1)\times\mathbb R,\qquad S^1(1)\times S^1(1). \] The case \(S=2\) forces the surface into \(S^3(\sqrt2)\) as a minimal flat torus, hence the Clifford torus. This provides the exact rigidity template imported into the Legendrian setting by projection and lift, and it clarifies why the generalized Legendrian Clifford torus and the Harvey–Lawson cone recur as rigid models [1802.02396]. A plausible implication is that Legendrian self-shrinker theory is now organized around three interacting paradigms: global graphical rigidity in \(\mathbb R^{2n+1}\), low-dimensional explicit classification in \(\mathbb R^3\) and \(\mathbb R^5\), and singularity-model analysis in Sasakian manifolds. Across all three, the distinctive contact-geometric content is the nontrivial Reeb contribution, which modifies both the flow equation and the weighted analytic structure, while the decisive classification mechanisms remain closely coupled to Lagrangian projection, phase harmonicity, and Clifford-type rigidity [2508.07900] [2508.15279]. Source: https://www.emergentmind.com/topics/legendrian-self-shrinkers