Prerequisites: Part I: Lie Groups and Lie Algebras, Part II: The Lorentz Group, familiarity with the Pauli matrices, and basic quantum mechanics (spin-1/2 systems).
At the end of Part II we had six generators, three rotations and three boosts , satisfying the Lorentz algebra,
We also had one explicit representation, the 4-dimensional vector representation, where the generators are matrices acting on 4-vectors. The algebra itself has infinitely many representations, though, and the physically most important ones are smaller than the vector one. To find them we need a trick.
Decomposition of the Lorentz algebra
The complexification trick
The three commutation relations above mix the ‘s and ‘s in an awkward way, so the goal is to find linear combinations that decouple. We define
and compute their commutators using the Lorentz algebra (this is a straightforward exercise, and honestly you should work it out yourself at least once),
This is a really nice result. The generators form an algebra on their own, the generators form a separate algebra, and the two copies commute with each other. So the Lorentz algebra has decomposed,
A subtlety you should not skip: this decomposition holds for the complexified Lorentz algebra. We had to multiply by to define , so we have extended the algebra over the complex numbers. The real Lorentz algebra is not the same as as a real Lie algebra. It is isomorphic to , the complexification of . The distinction matters when we talk about unitarity, but for classifying representations the complexified version is exactly what we need.
Classification by
Since there are two independent algebras, every finite-dimensional representation of the Lorentz algebra is labelled by two half-integers,
where labels the representation of and that of . Each follows the same rules as angular momentum in quantum mechanics, and the dimension of the representation is
This is why the decomposition is useful. Instead of fighting with the full Lorentz algebra (where the boosts make everything non-compact and painful), we have reduced the classification to something we already know, two copies of angular momentum.
From the definitions we can recover the physical generators,
The Casimir operators of the two ‘s are and , with eigenvalues and respectively.
The representations
Let us now go through the most important representations.
: scalar
Dimension . Both and , so and and nothing transforms. This is the trivial representation we already met in Part II.
: left-handed Weyl spinor
Dimension . The generators act in the spin- representation (i.e., as the Pauli matrices divided by 2), while ,
and the physical generators are
The rotation generators are Hermitian, as they should be, since rotations are generated by Hermitian operators in quantum mechanics.
The boost generators are anti-Hermitian, and this is a consequence of non-compactness. The Lorentz group is not compact, so its finite-dimensional representations cannot be unitary. Concretely, boosting a left-handed Weyl spinor does not preserve its norm. This makes sense physically, since Lorentz boosts are not symmetries of any positive-definite inner product on spinor space.
An object transforming in the representation is called a left-handed Weyl spinor, written or (with a two-component undotted index ). Under a Lorentz transformation,
Note the structure of the exponent. The rotation term comes with a factor of , so it acts as a unitary, phase-like rotation, while the boost term comes with a real , so it rescales the spinor and is not unitary.
: right-handed Weyl spinor
Dimension . Now and acts as spin-,
with physical generators
The rotation generators are the same as for the left-handed spinor (both are spin- under rotations), but the boost generators have the opposite sign. This is the whole difference between left- and right-handed spinors, they rotate the same way but boost differently.
A right-handed Weyl spinor is written or (with a dotted index), and under a Lorentz transformation
Compared with the left-handed case, the rotation piece is the same and the boost piece has flipped sign. A parity transformation () reverses the boost direction but not the rotation, so parity exchanges , i.e. left-handed right-handed.
: vector
Dimension . This 4-dimensional representation is in fact equivalent to the vector representation we built with explicit matrices in Part II, i.e. the four-vector transforms in the representation.
The connection can be made explicit with the map , where . Under a Lorentz transformation, the undotted index transforms under and the dotted index under , and the product is .
Higher representations
The pattern continues. The and representations are 3-dimensional each and correspond to self-dual and anti-self-dual antisymmetric tensors. The representation is 9-dimensional and corresponds to symmetric traceless tensors. The representation for general describes higher-spin objects. In practice, most of particle physics lives in the representations listed above.
Spinors in non-relativistic quantum mechanics
Before getting to fields, it is worth pausing to connect all this to something you already know.
Spinors in non-relativistic QM
In non-relativistic quantum mechanics, spin- particles are described by two-component objects that transform under rotations as
This is a representation of the rotation group , with no mention of boosts, because non-relativistic QM has only rotational symmetry and no Lorentz symmetry. The two-component spinor is a representation of and nothing more.
Spinors in relativistic QM
Once we move to special relativity, we need representations of the full Lorentz group and not only of its rotation subgroup. Saying how a two-component spinor rotates no longer fixes it, we also have to say how it boosts, and this is where the classification becomes essential.
A left-handed Weyl spinor and a right-handed Weyl spinor are both spin- under rotations, but they transform differently under boosts. Non-relativistic QM does not see the difference because it has no boosts. Relativistic QM has to, and that is why we need the full Lorentz representation theory.
Field representations
So far we have discussed how objects at a single point transform under the Lorentz group. A field is a function on spacetime, and under a Lorentz transformation two things happen.
- The argument transforms, (the field is evaluated at the transformed point).
- The field components mix, (according to the representation).
So a field in the representation transforms as
The in the argument (instead of ) makes sure the transformation is a proper group homomorphism. That is the general framework, and we now apply it to the specific representations.
Scalar fields:
A scalar field has no indices and transforms as
The value of the field does not change, it is just “moved” to the new location. The Higgs field of the Standard Model is a (complex) scalar field, and the Klein-Gordon equation describes a free scalar field.
Weyl fields: and
A left-handed Weyl field has two components and transforms as
and a right-handed Weyl field transforms in the same way with the opposite sign of the boost.
Weyl fields describe massless fermions (in the standard treatment). The equation of motion of a left-handed Weyl field is
where . This is a two-component equation for a two-component field, about as minimal as it gets.
Why are Weyl fields associated with massless particles? A Dirac mass term has the form , which needs both a left-handed and a right-handed field. is not Lorentz invariant (it is the time component of a four-vector), so it cannot be a mass term. A single Weyl field can form one Lorentz scalar with itself, , which is the Majorana mass term discussed below. That term carries two units of any charge of , so it is forbidden whenever carries a conserved charge (electric charge, or lepton number if you impose it). So a Weyl fermion with a conserved charge and no right-handed partner has to be massless.
Before the discovery of neutrino oscillations (which imply neutrino masses), neutrinos were thought to be described by a single left-handed Weyl field. The situation is now more subtle, but Weyl fields are still the basic building blocks.
Dirac fields:
To describe a massive fermion we need both chiralities. The Dirac field is a reducible representation of the Lorentz group,
i.e. a four-component object built from a left-handed Weyl spinor (top two components) and a right-handed one (bottom two). It transforms as the direct sum , which is reducible because the two Weyl spinors do not mix under Lorentz transformations.
What couples them is the mass term. The Dirac equation,
mixes and through the mass . In the massless limit it decouples into two independent Weyl equations, and the left- and right-handed components propagate independently.
The gamma matrices are matrices that connect the two Weyl representations. In the chiral (Weyl) basis they are
where and . The off-diagonal structure is precisely what couples to .
The electron, the muon and the quarks, i.e. all the massive fermions of the Standard Model, are described by Dirac fields.
Majorana fields
A Majorana field is a Dirac field with one extra constraint, namely that the particle is its own antiparticle. Formally this is the condition
where is the charge conjugate and is the charge conjugation matrix (its explicit form depends on the gamma matrix convention).
In terms of Weyl components, the Majorana condition sets . So a Majorana field has only two independent degrees of freedom instead of four, since the right-handed component is fixed by the left-handed one (or vice versa). In short,
- a Dirac field has four independent components, with and unrelated,
- a Majorana field has two independent components, with the charge conjugate of .
Majorana fields can have a mass (unlike a single charged Weyl field), but the mass term looks different. It is a Majorana mass , which violates lepton number by two units. This is why Majorana masses for neutrinos, if they exist, would matter a lot. They would imply lepton number violation and could be connected to the matter-antimatter asymmetry of the universe.
Whether neutrinos are Dirac or Majorana particles is one of the major open questions in particle physics, and neutrinoless double beta decay experiments are designed specifically to answer it.
Vector fields:
A vector field has four components and transforms as
which is the representation from Part II, now promoted to a field. The photon field of electrodynamics and the bosons of the weak interaction are vector fields.
One subtlety. A massive vector field has three physical degrees of freedom (the three polarizations), not four. The timelike component is not an independent propagating degree of freedom, and it is removed by the constraint equations (or, in the language of gauge theory, by gauge fixing). For a massless vector field like the photon only two polarizations are physical (the two transverse modes), and gauge invariance is essential for consistency.
The representations we need
| Representation | Dimension | Object | Example | |
|---|---|---|---|---|
| Scalar | 1 | Higgs boson | ||
| Left-handed Weyl | 2 | Left-handed neutrino | ||
| Right-handed Weyl | 2 | Right-handed electron | ||
| Vector | 4 | Photon | ||
| Dirac | 4 | Electron | ||
| Self-dual tensor | 3 | — | ||
| Anti-self-dual tensor | 3 | — |
Every field of the Standard Model transforms in one of the representations in this table. This is because the Standard Model is built to be Lorentz invariant, and these are the building blocks from which Lorentz-invariant Lagrangians are constructed.
Connection to what comes next
We have now classified how fields transform under the Lorentz group. The Lorentz group is not the full symmetry of spacetime, though, since it is missing translations. The full spacetime symmetry group is the Poincaré group, Lorentz transformations plus translations.
In the next post we study the Poincaré group and its representation theory. The central result is Wigner’s classification, which says that every irreducible unitary representation of the Poincaré group (i.e. every type of elementary particle) is labelled by exactly two numbers,
- the mass (with ),
- the spin (for massive particles) or the helicity (for massless particles).
The representation theory will explain why massless particles have only two helicity states while massive spin- particles have polarizations, why there is no smooth limit for certain representations, and why the little group ( for massive particles, for massless ones) determines the internal structure in each case.
Next post: Part IV: The Poincaré Group and the Classification of Particles
This post is based on my own self-study notes that I created in 2022 in order to get a deeper understanding of all of this.