432 H. Flores / Nuclear Physics B 925 (2017) 431–454
however, that Penrose formulation is invariant under conformal symmetries while Fronsdal for-
mulation is not [10].
At first glance, it may seem strange that tw
o descriptions of the same theory have different
symmetries. The present paper aims to solve this conflict. After a brief review of the two formu-
lations, we construct a first order action for Penrose theory and show that both theories describe
the same classical phase space via a canonical transformation. Using the canonical map, we can
push forw
ard the conformal transformations of Penrose theory to a set of non-local conformal
symmetries in Fronsdal description.
Plan of this paper We or
ganize our presentation as follows. Section 2 is a brief review, where
we explain the two approaches for free massless higher-spin theories.
In section 3 we write an action for Penrose higher
-spin theory. To our knowledge, such action
for general higher-spins has never appeared before in the literature. First-order formulations,
however, were used by Fradkin and Vasiliev in [14] for AdS space, where they were also ex-
tended to interactions. More recently, Kirill Krasnov described full self-dual gra
vity in [2] using
an action that resembles ours; but, in our case, this action is defined over complex field configura-
tions, and it describes off-shell a doubled set of the higher-spin modes. In phase space, however,
there is a well-defined notion of reality, and it is where we obtain a single cop
y of the spectrum.
At this point, we look at some e
xamples, so the spins 1, 3/2 and 2 cases are discussed in
detail, each of which highlights a particular feature of our construction outlining our strategy for
dealing with general spins. The spin s case is done in section 4; our construction is a particular
instance of the prescription gi
ven in [13], where a set of equations of motion and a presymplectic
structure are shown to lift to a well-defined Lagrangian.
With this map, we can in
vestigate conformal invariance. In section 5 we show that Penrose
action does have conformal symmetry for every spin s. Therefore one is able to push forward
these transformations to the Fronsdal case. For spins lower than 2, these new transformations
agree with usual conformal change of coordinates. The first non-tri
vial case is linearized gravity.
We write explicitly the resulting transformation, where one is able to see the difference from
standard Lie derivatives.
On notation Our con
ventions follow those of [8]; we are concerned with 4-dimensional Min-
skowski space; so, through out the paper, the various indices will always be running over fixed
intervals. Small Latin letters, for example, are spacetime indices running from 0to 3, so that
A
m
is a spacetime covector. Capital Latin letters, in turn, are spinor indices in Va n der Warden
notation, that is, dotted and undotted running from 0to 1. In particular, a Dirac spinor is a two
component Weyl and anti-Weyl spinor written like
=
ψ
A
χ
˙
A
(1.1)
for some chiral spinor ψ
A
and anti-chiral χ
˙
A
.
Such notation is designed so that there is a correspondence between spacetime and spinor
indices where, for instance, m will correspond to the pair M
˙
M. The explicit realization is given
by the Pauli matrices with index structure σ
m
M
˙
M
, where
σ
0
=−1 and σ = (σ
1
,σ
2
,σ
3
).
The epsilon symbol satisfies
AB
BC
= δ
C
A
for undotted and dotted indices. This enables one
to raise the indices of σ
m
M
˙
M
to obtain
评论0
最新资源