Skip to main content

Welcome to a Grasp lesson

Create a free account to keep studying and create your own study path.

Lesson 11

Symmetry Groups of Hamiltonians and Lagrangians

Recommended resources

Welcome to the first lesson in our module on Symmetry and Conservation Laws. In our previous module, we saw how dimensional analysis provides a powerful set of "rules of grammar" for physical laws. Today, we begin our exploration of a much deeper source of constraints and insight: symmetry. Emmy Noether herself described it as a study of "those properties of a system of equations which remain invariant under a group of transformations." We will unpack this statement and learn how to identify the symmetry group of a given physical system, a foundational skill in theoretical physics.

This lesson addresses the first step in that process: given a Hamiltonian or Lagrangian that describes a physical system, how do we systematically identify all the transformations that leave it unchanged? The set of these transformations forms a mathematical structure known as a group, and our goal is to learn how to determine this "symmetry group." We will explore continuous and discrete symmetries, those acting on spacetime and those acting on the fields themselves, and see how the abstract language of group theory provides a precise and powerful framework for describing them.

1. From Physical Invariance to Mathematical Groups

At its core, a symmetry is a transformation you can perform on a system that leaves its fundamental properties—and thus its equations of motion—unchanged. A sphere can be rotated by any angle about its center, and it looks the same. This invariance under rotation is a symmetry. In theoretical physics, we formalize this by saying a transformation is a symmetry if it leaves the Lagrangian (or Hamiltonian) of the system invariant.

The collection of all such symmetry transformations for a given system is not just a random set; it has a rich mathematical structure. Let's watch a segment from a lecture by Tobias Osborne that builds the definition of a group from the physical requirements that a set of symmetry operations must satisfy.

Symmetries and quantum mechanics: basics, Wigner's theorem, linear representations of groups

This video introduces the concept of symmetries as physical operations and then formally defines a mathematical group.

Please watch from the introduction to symmetry operations until the formal definition of a group is complete. Focus on the three key consistency conditions he proposes for any set of symmetry operations: Doing nothing: The identity operation must be a symmetry. Composition: Performing one symmetry followed by another must result in an overall transformation that is also a symmetry (closure). Inversion: If you can perform a symmetry operation, you must also be able to "undo" it. Notice how these physical requirements map directly onto the mathematical axioms of a group: the existence of an identity element, closure under the group operation, the existence of an inverse for every element, and associativity.

The video establishes that the set of symmetries of a physical system forms a group. Our task is now to figure out, for a given Lagrangian or Hamiltonian, precisely what that group is.


2. Identifying Symmetries of a Lagrangian

The fundamental principle for identifying a symmetry group is straightforward:

A transformation of the fields and/or spacetime coordinates is a symmetry if the Lagrangian, when expressed in terms of the new variables, takes the same functional form as the original Lagrangian (up to a possible total derivative, which doesn't affect the equations of motion).

Let's work through some concrete examples from a set of QFT notes by David Morrissey. These examples will illustrate the process and help us start classifying different types of symmetries.

[PDF] P528 Notes #2: Symmetries 1 Symmetries in QFT - David Morrissey

These notes provide clear, hands-on examples of identifying symmetries in Lagrangians.

Please read the introductory section, focusing on the three examples provided. Example 1: For the Lagrangian L = \frac{1}{2}(\partial\phi)^2 - \frac{1}{2}m^2\phi^2 - \frac{\lambda}{4!}\phi^4, verify by substitution that the transformation \phi(x) \to \phi'(x) = -\phi(x) leaves L unchanged. This is a discrete symmetry. Since applying the transformation twice returns the identity ((-\phi) \to -(-\phi) = \phi), this two-element group is called \mathbb{Z}_2. Example 3: For the Lagrangian L = |\partial\phi|^2 - M^2|\phi|^2 + \dots, consider the transformation on the complex field \phi given by \phi \to e^{i\alpha Q_\phi}\phi. This is a continuous symmetry because it holds for any real value of the parameter \alpha. The set of these phase rotations forms the group U(1), the group of 1 \times 1 unitary matrices.

These examples highlight a crucial first distinction:

  • Discrete Symmetries: Consist of a finite number of distinct transformations (e.g., reflections, permutations). The groups are finite, like (cyclic groups) or (permutation groups).
  • Continuous Symmetries: Depend on one or more continuous parameters (e.g., rotation angle, phase). The groups are Lie groups, like SO(3) (rotations in 3D) or U(1).

3. A Classification of Symmetries

To build a more complete picture, we can further classify symmetries based on what they act on (spacetime vs. fields) and how they act (globally vs. locally). A short lecture from MIT OpenCourseWare provides a concise overview.

Lecture 2: Symmetries and Conservation Laws

This video discusses different types of symmetries in the context of a simple scalar field theory.

Please watch the following segments: First, see the formal definition of a symmetry as a transformation leaving the action invariant (definition of symmetry). Next, observe examples of spacetime symmetries, where the transformation acts on the coordinates x^\mu: translations and Lorentz transformations. The group of all translations and Lorentz transformations is the Poincaré group, the fundamental spacetime symmetry of special relativity. Finally, see the distinction between global and local (gauge) symmetries (global vs. local). A global symmetry has a transformation parameter that is constant everywhere, while a local symmetry's parameter can vary with spacetime position.

This leads to a more refined classification:

Spacetime Symmetries (act on coordinates ) Internal Symmetries (act on fields )
Global Translations, Rotations (Poincaré group) U(1) phase rotation, SU(3) flavor symmetry
Local (Gauge) General Coordinate Invariance (General Relativity) U(1) gauge symmetry of QED, SU(2)U(1) of Electroweak theory

Gauge symmetries are not "true" symmetries in the sense of relating distinct physical states. Instead, they represent a redundancy in our mathematical description. The physics is invariant under these transformations because the different mathematical expressions they relate all describe the same physical state. Understanding how to build Lagrangians with local gauge invariance is a cornerstone of modern physics, as it dictates the nature of fundamental interactions.

The QED Lagrangian is the canonical example.

[PDF] P528 Notes #2: Symmetries 1 Symmetries in QFT - David Morrissey

This section from Morrissey's notes demonstrates how the QED Lagrangian is constructed to be invariant under a local U(1) transformation.

Read Section 2: Gauge Invariance and QED. Pay close attention to how the simple derivative \partial_\mu fails to be invariant under a local transformation \psi_i \to e^{iQ_i\alpha(x)}\psi_i. The problem is that the derivative compares the field at two different points, x and x+\delta x, which transform with different phases \alpha(x) and \alpha(x+\delta x). The solution is to introduce a gauge field A_\mu that also transforms, and to replace the ordinary derivative with the covariant derivative D_\mu = \partial_\mu + ieQ_iA_\mu. This new object is designed to transform "covariantly," ensuring the kinetic term as a whole remains invariant. This principle, that promoting a global symmetry to a local one requires introducing a gauge field, is the foundation of the Standard Model.


4. The Formal Structure: Lie Groups and Lie Algebras

For continuous symmetries, the corresponding groups are Lie groups—groups that are also smooth manifolds. This means we can study their properties by considering infinitesimal transformations near the identity. An infinitesimal transformation from a Lie group can be written as:

The matrices are called the generators of the group. They form a vector space called the Lie algebra, and their commutation relations, , define the local structure of the group. The constants are the structure constants of the algebra.

For a physicist, this is incredibly powerful. Instead of dealing with the infinite number of elements in a continuous group, we can often work entirely with the finite set of generators and their algebra.

Symmetries in physical systems are described by Lie groups, which are generated by Lie algebras. This diagram schematically illustrates how different symmetry groups, such as the U(1) of electromagnetism or the SU(N) groups of the Standard Model, represent different kinds of transformations on a system's state space.

The document from David Morrissey provides a concise summary of the key ideas.

[PDF] P528 Notes #2: Symmetries 1 Symmetries in QFT - David Morrissey

This section provides a formal but physically-motivated introduction to the language of groups, representations, and Lie algebras.

Read Section 3: Symmetries and Groups. This is a compact review, so focus on grasping the hierarchy of concepts: Group: The abstract set of symmetry operations. Representation: A concrete realization of the group as a set of matrices acting on a vector space (e.g., the Hilbert space of quantum states). Lie Group: A continuous group. You can study it via its infinitesimal transformations. Generators/Lie Algebra: The generators (t_a) capture the essence of infinitesimal transformations. Their commutation relations define the Lie algebra, which in turn determines the local structure of the group. Generators and Lie algebra. Familiarize yourself with the "classical groups" listed, as they appear constantly in physics: U(1), SU(N), and SO(N).


5. Symmetry in Condensed Matter: Point Groups

Your background includes work on crystal structures and bubble rafts. These systems often exhibit discrete symmetries. For example, the Hamiltonian describing the interactions of atoms in a crystal lattice is not invariant under any rotation, but only under a finite set of rotations (e.g., by 90, 180, 270 degrees for a cubic lattice) that map the lattice back onto itself.

These finite subgroups of the full rotation group SO(3) are called point groups. Identifying the correct point group is essential for classifying electronic orbitals, vibrational modes (phonons), and magnetic structures in a crystal.

Various possible magnetic orderings in a crystal structure. Each distinct pattern of spins (arrows) is symmetric under a different subgroup of the full rotation and translation group of the lattice. For example, a simple ferromagnetic state (all spins aligned, as in (a)) has a higher symmetry than a complex antiferromagnetic or helical state (like (h) or (n)).

The classification of these groups is a beautiful piece of mathematics. For our purposes, the key idea is that the requirement of mapping a lattice onto itself severely restricts the possible rotational symmetries.

[PDF] Symmetries in Physics, Lecture Notes

These lecture notes by Niklas Beisert offer a deep and elegant treatment of symmetry in physics. We will look at the section on point groups.

Please read the first part of Section 4.1 Point Groups, specifically the subsection "Discrete Subgroups of SO(3)". The derivation uses a clever counting argument on the fixed points of the rotation axes to prove that only a few families of finite rotation groups are possible. You don't need to absorb the entire proof, but appreciate the logic: the discrete nature of the group leads to strong constraints. The possible groups are the cyclic groups C_n, the dihedral groups D_n, and the symmetry groups of the Platonic solids (tetrahedron, cube/octahedron, icosahedron/dodecahedron).


Conclusion

In this lesson, we established a procedure for identifying the symmetry group of a physical system. The core method is to test which transformations leave the system's Lagrangian or Hamiltonian invariant.

Key Takeaways:

  • A symmetry is a transformation that leaves the action (or Hamiltonian) invariant.
  • The set of all such symmetries for a system forms a mathematical group.
  • Symmetries can be classified as discrete or continuous, spacetime or internal, and global or local (gauge).
  • Continuous symmetries are described by Lie groups, which can be analyzed through the commutation relations of their infinitesimal generators (the Lie algebra).
  • The procedure is to propose transformations (rotations, phase shifts, etc.), apply them to the Lagrangian/Hamiltonian, and check for invariance. The collection of all valid transformations is the symmetry group.

Identifying the symmetry group is the crucial first step. Once we know what the symmetries are, we can unlock their profound consequences. In our next lesson, we will explore one of the most beautiful results in all of physics: Noether's Theorem, which reveals an intimate connection between every continuous symmetry and a conserved physical quantity.