1

Suppose we have two 4-components Dirac spinors, that is two non commuting objects, $\psi$ and $\chi$. We know that:

$ \bar{\psi} \chi= - \chi \bar{\psi} $

$ \bar{\psi} = \psi^{\dagger} \gamma_0 $

$\dagger=T*$

The question is the following. What is $( \bar{\psi} \chi)^{T } $ ? For $T$ I mean taking the transpose, $\dagger$ is the adjoint, $*$ the complex conjugate.

Explicitly, is $ (\bar{\psi} \chi )^{T }=\chi^{T }\bar{\psi} ^{T } $ or $ (\bar{\psi} \chi )^{T }=- \chi^{T }\bar{\psi} ^{T } $ ?

And what can we say about $( \bar{\psi} \chi)^{\dagger} $? Is $ (\bar{\psi} \chi )^{\dagger}=\chi^{\dagger}\bar{\psi} ^{\dagger} $ or $ (\bar{\psi} \chi )^{\dagger}=- \chi^{\dagger}\bar{\psi} ^{\dagger} $ ?

Please, make the explanation clear, not only giving the correct answer.

Valac
  • 2,893
Rexcirus
  • 4,953

2 Answers2

0

There're a few things that should be clarified. First of all, are you talking about classical spinors?

If they are classical spinors, then they take values in Grassmann-odd numbers, and so its transpose should be

\begin{align} (\bar{\psi}\chi)^{T}&=(\psi^{\dagger}\gamma^{0}\chi)^{T} \\ &=\left[(\psi^{\ast})^{T}\gamma^{0}\chi\right]^{T} \\ &=-\bar{\psi}\chi, \end{align}

where in the last step, one has

\begin{align} \bar{\psi}\chi=(\psi^{\ast})^{T}\gamma^{0}\chi &=\sum_{i=0}^{3}\sum_{j=0}^{3}(\psi^{\ast})_{i}(\gamma^{0})_{ij}(\chi)_{j} \\ &=-\sum_{i=0}^{3}\sum_{j=0}^{3}(\chi)_{j}(\psi^{\ast})_{i}(\gamma^{0})_{ij} \\ &=-\sum_{i=0}^{3}\sum_{j=0}^{3}(\chi)_{i}(\psi^{\ast})_{j}(\gamma^{0})_{ji} \\ &=-\sum_{i=0}^{3}\sum_{j=0}^{3}(\chi)_{i}\left[(\gamma^{0})^{T}\right]_{ij}(\psi^{\ast})_{j} \\ &=-\chi^{T}(\gamma^{0})^{T}\psi^{\ast}=-\chi^{T}\bar{\psi}^{T}\equiv-(\bar{\psi}\chi)^{T}. \end{align}

Now, the problem has to do with the involution of the Grassmann algebra $\Lambda_{\infty}$. To be specific, suppose you have a bunch of anti-commuting elements $\theta_{i}$, i.e. $$\theta_{i}\theta_{j}+\theta_{j}\theta_{i}=0.$$ Then, each element (also known as the supernumber) in $\Lambda_{\infty}$ takes the form $$z=z_{B}+\sum_{k=1}^{\infty}\frac{1}{k!}C_{i_{1}\dots i_{k}}\,\theta_{i_{1}}\cdots\theta_{i_{k}},$$ where $z_{B}\in\mathbb{C}$ is its body part, and $C_{i_{1}\dots i_{k}}\in\mathbb{C}$ is the coefficient of its soul part. The involution (complex conjugation) $\ast$ in $\Lambda_{\infty}$ is defined as an action satisfying the following rules:
$1.\quad(\theta_{i})^{\ast}=\theta_{i}$.
$2.\quad(\alpha z)^{\ast}=\bar{\alpha}z^{\ast}$, where $\alpha\in\mathbb{C}$.
$3.\quad(z+w)^{\ast}=z^{\ast}+w^{\ast}$.
$4.\quad(zw)^{\ast}=w^{\ast}z^{\ast}$.
$5.\quad(z^{\ast})^{\ast}=z$.

Back to your question, the classical Dirac spinors should take values in the infinite dimensional $\Lambda_{\infty}$. So you would expect

\begin{align} (\bar{\psi}\chi)^{\dagger}&=\left[(\bar{\psi}\chi)^{T}\right]^{\ast}=\left[\chi^{T}\bar{\psi}^{T}\right]^{\ast} \\ &=\left(\sum_{i=0}^{3}\sum_{j=0}^{3}(\chi)_{i}\left[(\gamma^{0})^{T}\right]_{ij}(\psi^{\ast})_{j}\right)^{\ast} \\ &=\sum_{i=0}^{3}\sum_{j=0}^{3}\left[(\chi)_{i}\left[(\gamma^{0})^{T}\right]_{ij}(\psi^{\ast})_{j}\right]^{\ast} \\ &=\sum_{i=0}^{3}\sum_{j=0}^{3}(\psi)_{j}\left[(\gamma^{0})^{\dagger}\right]_{ij}(\chi^{\ast})_{i} \\ &=\sum_{i=0}^{3}\sum_{j=0}^{3}(\psi)_{j}(\gamma^{0})_{ij}(\chi^{\ast})_{i} \\ &=-\sum_{i=0}^{3}\sum_{j=0}^{3}(\chi^{\ast})_{i}(\gamma^{0})_{ij}(\psi)_{j}\equiv-\chi^{\dagger}\gamma^{0}\psi=-\bar{\chi}\psi. \end{align}

Notice that in the last step one also has $\bar{\psi}^{\dagger}=(\psi^{\dagger}\gamma^{0})^{\dagger}=\gamma^{0}\psi$, and so you should obtain

$$(\bar{\psi}\chi)^{\dagger}=-\chi^{\dagger}\bar{\psi}^{\dagger}.$$

Valac
  • 2,893
-1

EDIT: The correct answer is at: https://physics.stackexchange.com/a/458455/50207

Leaving the (wrong) answer below as it can be instructive.


I've clarified my doubt. The expression

$ (\bar{\psi} \chi )^{T }=\chi^{T }\bar{\psi} ^{T } $

is simply the definition of trasposition (the same argument for the adjoint), independently on the nature of the objects in the parenthesis. What is subtle is that, the following is true:

$ \bar{\psi} \chi =-\chi^{T }\bar{\psi} ^{T } =-(\bar{\psi} \chi )^{T }$

The minus in the first step arise only due to the anticommuting nature of the variables.

Rexcirus
  • 4,953
  • I had the same question and found an answer here. Just I can not understand the final line of your answer. You say that the minus sign appears in the first step just due to anticommutating nature of variable but why they are transposed too? If they are just commutated you must have the same expression as the first expression of your question. In this way your are saying that \xi^T=\xi – Vahid Sep 13 '16 at 21:59
  • This is not correct. What you call 'the definition of transposition' is actually a corollary. It automatically comes with a minus sign for Grassmann-variables. See https://physics.stackexchange.com/questions/458451/transpose-of-fermion-bilinears. – GaloisFan Dec 09 '19 at 07:54
  • @GaloisFan, please answer the question correctly – Michael Levy Dec 14 '21 at 00:28
  • I think for Grassman valued matrix field $(AB)^T=-B^T A^T$ see https://physics.stackexchange.com/a/701245/310987 so you made a mistake – Марина Marina S Mar 30 '22 at 19:01