Fermionic fields change sign under a rotation by $2\pi$. However, in $SO^+\left(1,3\right)$ a rotation by $2\pi$ is the identity. For any representation $R$ of $SO\left(1,3\right)$ then we have $$R\left(\Lambda_1\right)R\left(\Lambda_2\right)=R\left(\Lambda_1\Lambda_2\right)$$ taking $\Lambda_1=\Lambda_2=$rotation by $2\pi=$ the identity, we have $$R\left(1\right)R\left(1\right)=R\left(1\right)$$ giving $R\left(1\right)=1\neq-1$. Hence, these fermionic fields do not transform under a representation of $SO^+\left(1,3\right)$. Of course, fermions do transform under a representation of the double cover of $SO^+\left(1,3\right)$, $SL\left(2,\mathbb{C}\right)$, but by the above argument these are not representations of the Lorentz group.

I feel like my professors talk as if this is not the case, calling them all representations of the Lorentz group. I just wanted to check that this is correct, and they are not really representations of the Lorentz group?

  • 1
    $\begingroup$ Define "so-called representations." Do you mean projective representations? $\endgroup$ – Qmechanic Feb 6 '19 at 14:36

Before answering the question, I'll review the definition of "representation" so that the answer will be clear. If $G$ and $H$ are groups, then a homomorphism from $G$ to $H$ is a map that assigns each element $g\in G$ to some element $\sigma(g)\in H$ such that the basic group operations are respected, such as $\sigma(g_1)\sigma(g_2)=\sigma(g_1 g_2)$ and $\sigma(1)=1$ where $1$ is the identity element in $G$ or $H$, respectively.

A matrix representation of a group $G$ is a homomorphism $\sigma:G\to H$ where $H$ is a group of matrices (usually over $\mathbb{C}$) using matrix multiplication as the multiplication rule for the group. This answer will focus on matrix representations, as implied by the OP.

By this definition, the spin-$1/2$ "representation" of the rotation group $SO(3)$ is not a representation, for precisely the reason noted in the OP. It is a legitimate reprentation of the covering group $SU(2)$, again as noted in the OP. Here I'm using just the rotation group $SO(3)$ as an example instead of the Lorentz group $SO(1,3)$, but the idea is the same.

If $F$ is a covering group of $G$, then a representation of $F$ can be regarded as a projective representation of $G$. Intuitively, if $F$ reduces to $G$ when certain differences between elements of $F$ are ignored (technically, when $G$ is regarded as a quotient of $F$), then a representation of $F$ gives a "representation" (in scare quotes) of $G$ if certain differences bewteen representation-matrices are ignored. The word "projective" refers to ignoring these factors (like factors of $-1$ in the OP's example).

On the other hand, if $G$ is a Lie group, then $G$ has an associated Lie algebra, and we can also talk about representations of the Lie algebra. Since $SO(3)$ and $SU(2)$ have the same Lie algebra (just like $SO(1,3)$ and $SL(2,\mathbb{C})$ have the same Lie algebra), a representation of the Lie algebra of $SU(2)$ is also a legitimate representation of the Lie algebra of $SO(3)$.

So the OP's suspicion is correct: people often use the language carelessly (or at least in non-standard ways). People might say "representation of the group $G$" when they really mean "projective reprentation of the group $G$" or "representation of a covering group of $G$" or "representation of the Lie algebra corresponding to $G$."


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.