In the book of Freedman & van Proeyen on Supergravity a table (3.2) can be found which shows for dimensions from 2-11 the number of components of Majorana spinors.

For instance in 4 dimensions we have 4 components for a spin-1/2 Majorana spinor (represented by a bispinor), whereas in 8 dimensions there are 16 components.

Actually, in 4 dimensions all components have a physical meaning, 2 correspond to the spin 1/2 and another 2 to the particle anti-particle symmetry. The latter two get kind of superfluous once the reality condition for a Majorana spinor is imposed.

However, in 8 dimensions there are 16 components whereas only 2 degrees of freedom for the spin 1/2 are needed. What is the physical meaning of the other components ? Are they also "superfluous" due to some conditions as the reality condition ? Or do they get "lost" in the higher-dimensional space ?


In order to make my point clearer let's assume a Majorana spinor in Weyl-respresentation

$$\Psi_M = \left( \begin{array}{c} \xi \\ \xi^\dagger \end{array}\right)$$

(Other representations can be found by multiplying $\Psi_M$ with an appropiate matrix, for instance one which makes all components real.)

where $\xi$ (if considered in the rest frame of the concerned particle and the axes (s,t,u,v,w,x,y,z) of the coordinate system are appropiately chosen) is an eigenmode of the z-component of the spin (or helicity) operator -- apparently there is no further condition:

$$ S_z \left( \begin{array}{c} \xi_1 \\ \xi_2 \\ \xi_3 \\ \xi_4 \\ \xi_5 \\ \xi_6\\ \xi_7 \\ \xi_8 \end{array}\right) = \pm \frac{1}{2} \left( \begin{array}{c} \xi_1 \\ \xi_2 \\ \xi_3 \\ \xi_4 \\ \xi_5 \\ \xi_6\\ \xi_7 \\ \xi_8 \end{array}\right)$$

Why in 8 dimensions do I need an 8-component long spinor for fulfilling such a simple eigenmode equation whereas in 4D spacetime I only need a 2-component long spinor ? May be there is a misconception in my thinking, so I would appreciate if such a misconception could be cleared up.

  • $\begingroup$ Doesn't a spin state have 4 extra degrees of freedom in 8 dimensions? $\endgroup$ Feb 8, 2022 at 10:44
  • $\begingroup$ @Felicia The question is not about the spin states. Moreover, the number of components of a (Majorana)-spinor in $n$-dimensional space is not equal to $n$. $\endgroup$ Feb 8, 2022 at 12:26
  • $\begingroup$ @FredericThomas I don't mean the number of components (which is 8), but the number of degrees of extra freedom. So in 4D space up or down (thanks to 3 space and 1 time dimension), and for every extra space dimension an up and down degree extra. $\endgroup$ Feb 8, 2022 at 12:37
  • $\begingroup$ related/possible dup? physics.stackexchange.com/q/381752/84967 $\endgroup$ Feb 8, 2022 at 17:44

2 Answers 2


There is no such a thing as "the z-component of spin $S_z$" in higher dimensions. There is no z-axis for general $d$, rather there are multiple "$z$-axes" all of which we need to diagonalize simultaneously, so to speak.

Of course I cannot give you a crash course on representation theory of simple groups here (the book on CFT by di Francesco et all does a good job, it has a physics-oriented chapter on Lie groups, see chapter 13). But I can give you a flavor of some of the ingredients that are important for your question.

Consider the Lie algebra $g=\mathfrak{so}(d)$. It has some commutation relations, which are probably familiar to you. There is a nice basis of the algebra, known as the Cartan-Weyl basis, which takes the form $$ [H^i,H^j]=0,\quad [H^i,E^\alpha]=\alpha^i E^\alpha,\qquad [E^\alpha,E^\beta]\propto E^{\alpha+\beta} $$ Here, the set $H^i$ (where $i=1,2,\dots,\text{rank}(g)$) is a set of commuting operators, which replace $S^z$ for general $d$. Note that the rank of $SO(3)$ is 1, hence a single $S^z$, but the rank of $SO(d)$ is $\lfloor d/2\rfloor$ for higher $d$, so you need multiple $S^z$ that you have to diagonalize.

It just so happens that the spinor representation (i.e., the representation that you get when you simultaneously diagonalize all the $H^i$), has dimension $\sim 2^d$. Roughly speaking, this representation has $H^i=0$ for all $i<\text{rank}(g)$, and $H^i=1/2$ for $i=\text{rank}(g)$. So this plays the role of a higher dimensional fermion, this is the smallest representation that has half-integral eigenvalues. It is a fact of life that this representation requires $\sim 2^d$ components, no smaller vector spaces is able to diagonalize all the require operators.

(Note: of course, sometimes you can half the size of the representation by imposing suitable reducibility conditions, such as Majorana. This depends on $d\mod8$, and the reference in the OP does a good job at explaining when and why can you impose such conditions.)

  • $\begingroup$ That looks interesting. I will in any case go through it. $\endgroup$ Feb 9, 2022 at 18:01

For a spin $j=1/2$ spinor field, the number of DOF is not just 2 as OP writes in the title. Instead

$$\text{Off-shell DOF = #(components)},$$


$$\text{On-shell DOF = #(helicity states) = #(components)/2}. $$

A Dirac spinor has $2^{[D/2]}$ complex components in $D$ spacetime dimensions, while a Majorana spinor has $2^{[D/2]}$ real components, cf. e.g. my related Phys.SE answer here.

See also e.g. this related Phys.SE post.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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