Lorentz Group Representations

In physics, symmetry is not merely an aesthetic quality but the very language in which the laws of nature are written. Among the most fundamental of these are the symmetries of spacetime, described by the Lorentz group. Every rotation and every change in velocity must leave the underlying principles of physics unchanged. This raises a profound question: what are all the possible types of objects—the fundamental particles and fields—that can exist in a universe governed by these rules? The answer lies in the representation theory of the Lorentz group, a mathematical framework that provides a complete catalog of nature's building blocks. This article serves as a guide to this essential theory, addressing the challenge of classifying all possible relativistic objects.

The journey begins in the first chapter, ​​Principles and Mechanisms​​, where we will uncover the stunning mathematical simplicity hidden within the Lorentz group. We will explore how its seemingly complicated structure elegantly splits into two simpler, familiar parts, leading to a universal classification scheme known as the (jA, jB) system. This system allows us to identify and understand the properties of foundational entities like spinors and vectors. In the second chapter, ​​Applications and Interdisciplinary Connections​​, we will see this abstract framework in action. We will learn how it is used to define a particle's identity through its spin, to construct the rules governing their interactions in quantum field theory, and even to bridge the gap between particle physics and Einstein's theory of gravity, revealing a deep and unexpected unity in the physical world.

Imagine you are trying to understand an extraordinarily complex machine. You see gears turning, levers moving, and you want to know the rules. What are the fundamental principles governing its operation? This is precisely the position we are in when we study the symmetries of spacetime. The Lorentz transformations—rotations and boosts—are the "motions" of this machine, and the objects they act upon, the fields and particles that fill our universe, are its "parts." Our goal is to find a complete catalog of these parts and understand the rules they follow.

At first glance, the rules look messy. The "gears" of our machine are the six ​​generators​​ of the Lorentz group: three for rotations, which we'll call $J_1, J_2, J_3$, and three for boosts, $K_1, K_2, K_3$. They obey a specific set of commutation relations, which are the mathematical expression of how these operations intertwine. For example, performing a rotation and a boost in a different order doesn't give you the same result. A key relation that captures this is $[J_i, K_j] = i \epsilon_{ijk} K_k$, which tells us that the commutator of a rotation and a boost gives us another boost. While boosts among themselves give rotations: $[K_i, K_j] = -i \epsilon_{ijk} J_k$. This all seems a bit complicated and tangled.

But here, a wonderful mathematical "trick" reveals an astonishing simplicity, a trick so beautiful it feels like we're uncovering a deep secret of nature. Instead of working with the $J_i$s and $K_i$s, let's define two new sets of generators:

Why the imaginary number $i$? It seems like a strange move, but it's the key that unlocks everything. If you work out the commutation relations for these new generators, the tangled mess miraculously unravels. You find that all the $A$ generators commute with all the $B$ generators, $[A_i, B_j] = 0$. And within their own sets, they obey:

Look at that! These are just the familiar commutation relations for the generators of ordinary rotations, the algebra of angular momentum known as $\mathfrak{su}(2)$. The complex-looking Lorentz algebra, $\mathfrak{so}(1,3)$, has decomposed into two completely independent, identical copies of the simple rotation algebra: $\mathfrak{su}(2) \oplus \mathfrak{su}(2)$. It's as if our complicated machine was, all along, just two simple, independent gyroscopes spinning.

This simplification is not just mathematically elegant; it is tremendously powerful. It gives us a universal system to classify every possible kind of object that can exist in a relativistic theory. Since the two algebras are independent, any ​​irreducible representation​​—which corresponds to a fundamental type of particle or field—can be labeled by a pair of numbers, $(j_A, j_B)$. These numbers are simply the "spin" quantum numbers for each of the two $\mathfrak{su}(2)$ algebras, which can be integers or half-integers ($0, 1/2, 1, 3/2, \dots$).

This pair of numbers is like a universal parts number. If you tell me the $(j_A, j_B)$ of a field, I can tell you everything about how it must transform under any rotation or boost. The number of components it has, for example, is simply $(2j_A+1)(2j_B+1)$. This is the complete classification scheme for the building blocks of any relativistic theory.

But what do these numbers mean physically? What is the connection to the familiar concept of ​​spin​​? The ordinary spin of a particle tells us how it behaves under simple spatial rotations. In our new language, the rotation generators are $\vec{J} = \vec{A} + \vec{B}$. This is exactly the formula for adding two independent sources of angular momentum in quantum mechanics. So, an object of type $(j_A, j_B)$ will, when you only look at its rotational properties, appear to have a spin $J$ that can take any value between $|j_A - j_B|$ and $j_A + j_B$, in integer steps. The "highest spin content" of the particle is $j_A+j_B$.

With this classification scheme, we can take inventory of the fundamental entities in our theories.

The simplest non-trivial objects are those where one of the labels is zero. These are the ​​Weyl spinors​​:

• The $(\frac{1}{2}, 0)$ representation describes a ​​left-handed spinor​​. It has $(2 \cdot \frac{1}{2} + 1)(2 \cdot 0 + 1) = 2$ components.
• The $(0, \frac{1}{2})$ representation describes a ​​right-handed spinor​​. It also has 2 components.

These spinors are the true fundamental building blocks of matter. Particles like electrons and quarks are described by combinations of them. The generators for these representations are built from the famous Pauli matrices, $\vec{\sigma}$. For the $(\frac{1}{2}, 0)$ representation, we have $\vec{J} = \frac{1}{2}\vec{\sigma}$ and $\vec{K} = -\frac{i}{2}\vec{\sigma}$. Notice the crucial $i$ in the boost generator. This has a profound physical consequence. When you boost a left-handed spinor, its transformation involves a factor of $\exp(-\zeta/2)$, while a right-handed spinor's transformation involves $\exp(+\zeta/2)$, where $\zeta$ is the rapidity of the boost. This means a boost affects left- and right-handed particles differently, scaling their components by different amounts. This isn't just a mathematical curiosity; it's a measurable physical effect essential to understanding particle physics.

What about a familiar object like a ​​four-vector​​, such as the spacetime position $x^\mu=(ct, x, y, z)$? It has four components. In our scheme, what $(j_A, j_B)$ gives dimension $(2j_A+1)(2j_B+1)=4$? The simplest answer is $(\frac{1}{2}, \frac{1}{2})$. This is a beautiful revelation! A vector, which we often think of as fundamental, is actually a composite object in this deeper sense, built from combining the A-spinor and B-spinor parts. Its spin content ranges from $J=|\frac{1}{2}-\frac{1}{2}|=0$ to $J=\frac{1}{2}+\frac{1}{2}=1$, which is why a four-vector field (like the photon) is a "spin-1" particle.

We can even describe more complicated objects, like the electromagnetic field tensor $F^{\mu\nu}$. This antisymmetric tensor has 6 independent components. It turns out that this representation is reducible. It splits neatly into two irreducible parts: a ​​self-dual​​ part, which corresponds to the $(1,0)$ representation (3 components), and an ​​anti-self-dual​​ part, corresponding to the $(0,1)$ representation (3 components). So, the richness of electromagnetism is captured by these two fundamental representations of the Lorentz group.

Just as we can combine atoms to make molecules, we can combine these field representations to describe more complex systems or interactions. This is done via the ​​tensor product​​, and the rule is delightfully simple: you just use the standard quantum mechanical rules for adding angular momentum on each of the labels independently. For example, combining a $(1,0)$ object with a $(\frac{1}{2}, \frac{1}{2})$ object results in a mix of two new irreducible objects: $(\frac{1}{2}, \frac{1}{2})$ and $(\frac{3}{2}, \frac{1}{2})$. This mathematical machinery is the basis for constructing interactions in quantum field theory.

The Lorentz group we've discussed so far only includes "proper" transformations that can be reached continuously from doing nothing (rotations and boosts). What about discrete transformations, like looking at the world in a mirror? This is the ​​parity transformation​​, $P$, which sends $\vec{x} \to -\vec{x}$.

Parity acts on our generators in a very interesting way. It leaves the rotation generators alone ($P \vec{J} P^{-1} = \vec{J}$) but flips the sign of the boost generators ($P \vec{K} P^{-1} = -\vec{K}$). What happens to our beautiful $\vec{A}$ and $\vec{B}$ generators?

And similarly, $P \vec{B} P^{-1} = \vec{A}$. Parity swaps the two $\mathfrak{su}(2)$ algebras! This means it swaps the labels in our classification scheme: an object of type $(j_A, j_B)$ is transformed by parity into an object of type $(j_B, j_A)$.

This has profound consequences. A left-handed Weyl spinor $(\frac{1}{2}, 0)$ is turned into a right-handed one $(0, \frac{1}{2})$. They are mirror images of each other. For a long time, it was assumed that the laws of physics should be the same in the mirror world—that they should respect parity symmetry. The discovery of the weak nuclear force proved this wrong. Nature, at a fundamental level, can tell the difference between left and right.

Representations where the labels are the same, like the vector $(\frac{1}{2}, \frac{1}{2})$, are mapped to themselves under parity. This means they can be states of definite parity. In contrast, a Weyl spinor cannot. A Dirac spinor, which describes massive electrons, is constructed by combining a left-handed and a right-handed spinor, $\Psi = \begin{pmatrix} \psi_L \\ \psi_R \end{pmatrix}$. This larger object, belonging to the $(\frac{1}{2}, 0) \oplus (0, \frac{1}{2})$ representation, can have a definite parity because the parity operator just swaps its upper and lower components.

So far, we have built a beautiful, finite-dimensional zoo of fields labeled by $(j_A, j_B)$. These are the fields that appear in our Lagrangians. But there's a crucial distinction to be made. When we move to quantum mechanics, the physical states of particles (like an electron with a specific momentum) must live in a Hilbert space, and for probabilities to be conserved, transformations like boosts and rotations must be represented by ​​unitary operators​​.

Here we hit a surprising and deep fact: for a ​​non-compact​​ group like the Lorentz group, all non-trivial unitary irreducible representations are ​​infinite-dimensional​​. The beautiful, finite-dimensional $(j_A, j_B)$ representations we've been discussing are not unitary! (The exception is the trivial $(0,0)$ representation.)

This means that the tools and theorems we use for compact groups like the rotation group $SO(3)$, where all unitary representations are finite-dimensional, do not directly apply. For instance, the famous Wigner-Eckart theorem, which simplifies the calculation of matrix elements in quantum mechanics, must be substantially modified to handle the continuous spectra and infinite dimensions of the Lorentz group's unitary representations. This distinction between the finite-dimensional fields used to build theories and the infinite-dimensional state spaces where particles live is one of the subtle and beautiful complexities of relativistic quantum theory, opening the door to a much richer mathematical world.

After our deep dive into the principles and mechanisms of Lorentz group representations, you might be wondering, "What is all this beautiful mathematics for?" The answer, which is a source of constant delight and astonishment for physicists, is that it is nothing less than the language nature uses to write its own rules. To understand the world of elementary particles, to build theories of their interactions, and even to grapple with the very fabric of spacetime in gravity, we must become fluent in this language. This chapter is a journey through these applications, revealing how the abstract algebra of symmetries gives rise to the concrete, observable universe.

The first and most direct application of Lorentz group representations is in a task of cosmic importance: classifying the fundamental particles. Just as a biologist classifies species, a physicist classifies particles. Einstein's famous equation, $E = mc^2$, tells us that one fundamental property of a particle is its mass, which corresponds to the eigenvalue of the Casimir operator $P^2$ of the Poincaré group. The second great label is ​​spin​​. What we call spin is, at its heart, a label specifying which representation of the Lorentz group a particle's quantum field transforms under.

The finite-dimensional representations are labeled by a pair of half-integers, $(j_L, j_R)$. For a massive particle, we can always go to its rest frame and ask how it behaves under ordinary rotations, the $SO(3)$ subgroup of the Lorentz group. The answer is that a single Lorentz representation $(j_L, j_R)$ decomposes into a collection of rotational representations, with spins $j$ given by the familiar rules of adding angular momentum:

$j \in \{|j_L - j_R|, |j_L - j_R| + 1, \dots, j_L + j_R\}$

Consider a field that transforms in the $(1, 1)$ representation. In the rest frame of a massive particle described by this field, we would find not one, but three species of spin: a spin-2 part, a spin-1 part, and a spin-0 part. This mathematical fact has profound physical consequences. It means that to describe a single massive particle of, say, spin-2, one cannot simply use a $(1,1)$ field; one must impose additional constraints to project out the unwanted spin-0 and spin-1 components. Similarly, a more complex object like the Rarita-Schwinger field, which transforms under $(1, 1/2) \oplus (1/2, 1)$, contains both spin-3/2 and spin-1/2 components, and is used in theories of particles like the delta baryon.

For massless particles, something truly wonderful happens. A massless particle travels at the speed of light, so we can never go to its rest frame. The concept of "spin" is replaced by ​​helicity​​: the projection of its angular momentum onto its direction of motion. The constraints on massless states, first elucidated in Wigner's monumental work on the classification of particles, are severe. The allowed helicities are sharply restricted by the representation.. Let's revisit our particle in the $(1,1)$ representation. If it is massive, it contains spins 0, 1, and 2. But if it is massless, the physical states that propagate are found to have only helicities $h = \pm 2$. This isn't just a mathematical curiosity; it's a description of the ​​graviton​​, the quantum of the gravitational field!. The same mathematical object behaves completely differently depending on whether it is massive or massless.

So what, then, is spin? The clearest definition comes from the Pauli-Lubanski pseudovector, $W_\mu$. The operator $W^2 = W_\mu W^\mu$ is, like $P^2$, a Casimir invariant of the Poincaré group, meaning its eigenvalue is a fixed number for any given irreducible representation. This eigenvalue is precisely $-m^2 s(s+1)$, where $s$ is the particle's spin. For a scalar field, which by definition transforms trivially under the Lorentz group (it's in the $(0,0)$ representation), the generators of Lorentz transformations have no effect on its internal state. Consequently, the Pauli-Lubanski operator must yield zero when acting on a single-particle state of a scalar field. This means its spin is fundamentally zero. This is a purely geometric fact about how the field transforms, and it holds true regardless of any other properties, such as whether it hypothetically follows Fermi-Dirac statistics in violation of the spin-statistics theorem.

Knowing what particles exist is only half the story. The other half is understanding how they interact. Here too, the Lorentz group is the ultimate arbiter. The principle is simple: the laws of physics must be the same for all inertial observers. In the language of quantum field theory, this means that the Lagrangian density, the function that encodes all the dynamics and interactions, must be a Lorentz scalar—it must transform in the trivial $(0,0)$ representation.

An interaction term typically involves a product of several fields at the same spacetime point. When fields are multiplied, their Lorentz representations combine via a "tensor product." The crucial question is whether this resulting combination contains a scalar $(0,0)$ component. If it doesn't, a direct interaction of that form is forbidden by the laws of special relativity.

This provides an incredibly powerful set of "selection rules." For example, one can ask if a direct, local interaction is possible between three hypothetical particles transforming in the $(3/2, 1/2)$, $(1, 3/2)$, and $(1, 0)$ representations. A careful analysis of the number of spinor indices shows that it's impossible to contract them all to form an invariant. The sum of left-handed indices and the sum of right-handed indices are not both even, which is a prerequisite for forming a scalar. Thus, such an interaction vertex is forbidden. Symmetry alone, without any knowledge of the detailed dynamics, rules out a possible interaction.

When interactions are allowed, the "recipe" for combining the representations is given by the Lorentz Clebsch-Gordan coefficients. These are the precise mathematical factors that tell us how to project the tensor product of several fields onto a specific final representation. For instance, constructing an interaction between a four-vector field (like the photon, in the $(1/2, 1/2)$ representation) and a Weyl spinor (like a neutrino, in the $(1/2, 0)$ representation) requires these coefficients to build a consistent, covariant term in the Lagrangian. This is the intricate mathematical grammar that underpins the Standard Model of Particle Physics.

Perhaps the most profound interdisciplinary connection arises when we try to take our particles, so beautifully described by Lorentz representations, and place them in the curved spacetime of Einstein's General Relativity. A dramatic inconsistency immediately appears.

Spinors, by their very definition, are objects that know how to transform under the Lorentz group, $SO(1,3)$. However, the principle of general covariance at the heart of GR deals with arbitrary coordinate transformations, whose Jacobians form a much larger group, the general linear group $GL(4, \mathbb{R})$. A spinor simply does not have a rule for transforming under this larger group. There is a fundamental mismatch between the language of particle physics and the language of gravity.

The resolution to this puzzle is one of the most elegant ideas in theoretical physics. We introduce a new mathematical tool: the ​​tetrad​​ field (or vierbein). You can think of the tetrad as a set of four "flagpoles" planted at every point in curved spacetime. These flagpoles define a small patch of flat Minkowski space—a local Lorentz frame—at that point. The tetrad, then, acts as a dictionary or a Rosetta Stone, translating between the "curved" world indices of the spacetime manifold and the "flat" Lorentz indices of the local tangent space where the spinor lives.

But this introduces a new subtlety. The local Lorentz frame at one point in spacetime is, in general, rotated relative to the frame at a neighboring point. To meaningfully compare a spinor at point A to one at point B—a procedure essential for defining a derivative—we need to account for this rotation. This requires another new field: the ​​spin connection​​. The spin connection acts as a gauge field for the local Lorentz symmetry. Its role is to ensure that the derivative of a spinor transforms covariantly, "parallel transporting" the spinor from one frame to the next in a consistent way.

In this beautiful synthesis, the need to describe spinor particles in a gravitational field forces us to recast General Relativity as a gauge theory of local Lorentz symmetry. This reveals a deep and unexpected unity between the symmetries of special relativity, the dynamics of gravity, and the gauge principles that govern the fundamental forces of nature.

As magnificent as the Lorentz group is, it is not the end of the story. In certain physical regimes—at ultra-high energies where particle masses become negligible, or in systems at a critical point of a phase transition—an even larger symmetry can emerge: the ​​conformal group​​, $SO(4,2)$. This 15-parameter group contains the Lorentz group and translations, but also includes transformations of scale (dilatations) and special conformal transformations.

A single irreducible representation of this larger conformal group is a vast, infinite-dimensional space. When we view this single object from the limited perspective of its Lorentz subgroup, what we see is an entire infinite tower of familiar Lorentz representations. For example, the representation corresponding to a free, massless scalar field in a conformal field theory can be decomposed level by level. At level zero is the primary scalar field itself, a $(0,0)$ Lorentz representation. Its first set of descendants transforms as a four-vector, the $(1/2, 1/2)$ representation. The next level contains a symmetric traceless tensor, the $(1,1)$ representation, and so on. A single, highly symmetric conformal representation contains an infinite spectrum of Lorentz representations, $(j_n, j_n) = (n/2, n/2)$ for all $n \geq 0$.

This shows a stunning hierarchy of structure. The objects we study in our world may simply be the "low-energy" shadows or components of much larger, more symmetric structures that are only fully revealed in more extreme conditions.

From labeling particles to dictating their interactions, and from shaping the theory of gravity to hinting at grander symmetries beyond, the representation theory of the Lorentz group is an indispensable tool. It is a testament to the idea that the deepest secrets of the universe are written in the language of symmetry, a language we are privileged to learn and, with it, to appreciate the profound unity and beauty of the physical world.