diff --git a/Surface_modeling/doc/Surface_modeling/Concepts/DeformationClosestRotationTraits_3.h b/Surface_modeling/doc/Surface_modeling/Concepts/DeformationClosestRotationTraits_3.h
index 4f19787dc60..3c0f7993663 100644
--- a/Surface_modeling/doc/Surface_modeling/Concepts/DeformationClosestRotationTraits_3.h
+++ b/Surface_modeling/doc/Surface_modeling/Concepts/DeformationClosestRotationTraits_3.h
@@ -4,7 +4,7 @@
 
 @brief Concept describing the set of requirements for computing a close rotation to a 3x3 matrix together with basic computations used in the class `CGAL::Deform_mesh`.
 The fact that some basic operations are hidden behind a function is to allow to benefit from optimizations like expression template from libraries used
-to implement a model of this concept.
+to implement models of this concept.
 
 \cgalHasModel `CGAL::Deformation_Eigen_closest_rotation_traits_3`
 
@@ -52,9 +52,9 @@ public:
   /// Returns `i`th coefficient of a vector
   double vector_coeff(const Vector& v, int i);
     
-  /// Computes a close rotation to `m` and places it into `R`
-  /// \note The algorithm expects the closest rotation in Frobenius norm, however not returning the closest rotation does not lead to a crash or malfunction. 
-  ///  For example, returning directly identity matrix without regarding `m`, will result in naive Laplacian deformation.
+  /// Computes a rotation matrix close to `m` and places it into `R`
+  /// \note It is expecting to provide the closest rotation in Frobenius norm, however not returning the closest rotation does not lead to a crash or non-convergence.
+  ///  For example in the context of the deformation, always returning the identity matrix independently of `m` will result in a naive Laplacian deformation.
   void compute_close_rotation(const Matrix& m, Matrix& R);
   
 /// @}