Unfolded Formulation of 4d Yang–Mills Theory

Nikita Misuna

Tamm Department of Theoretical Physics, Lebedev Physical Institute,

Leninsky prospekt 53, 119991, Moscow, Russia
arXiv:2408.13212v1 [hep-th] 23 Aug 2024

[email protected]

Abstract
In this note we construct a formulation of 4d pure Yang–Mills theory in the unfolded
language of higher-spin gravity. It represents a first-order formulation possessing manifest
diffeomorphism- and gauge-invariance. We make use of an unfolding method that was
recently put forward and used to unfold scalar electrodynamics. We also discuss features
of various unfolding maps defined by the unfolded equations.

1
Introduction
Symmetry considerations are among the most important guiding principles in constructing
theories of fundamental interactions. Higher-spin gravity, representing a theory of interacting
massless fields of all spins, sets an example. To a large extent, this theory is determined by an
infinite-dimensional higher-spin gauge symmetry [1]. For a partial review of the literature on
the subject see [2].
To keep the higher-spin symmetry under control is of crucial importance. Originally, higher-
spin gravity was formulated in the form of Vasiliev equations [3, 4] which are classical equations
of motion written in the so-called unfolded form. Unfolded equations are first-order differen-
tial equations on unfolded fields being exterior forms, which possess manifest gauge symmetry.
Thus, the unfolded dynamics approach [3–8] represents a classical first-order formalism pro-
viding manifest diffeomorphism- and gauge-invariance. It also allows for an effective control
over degrees of freedom, since unfolded fields (whose spectrum is infinite in dynamical field
theories) parameterize all d.o.f. of a theory under consideration. In particular, this makes the
unfolded dynamics approach a promising tool for the study of AdS/CF T -correspondence and
other dualities [9–11]. A way to quantize classical unfolded theories was put forward in [12].
Apart from higher-spin gravity models [3, 4, 13, 14], not that many examples of unfolded
formulations of nonlinear theories are available [12, 15, 16], which limits the ability to explore
the full potential of the approach. The reason behind this was a lack of a practical algorithm
of unfolding, which was especially critical for nonlinear theories. Therefore, linear models were
mainly studied [17–21]. In [16], a novel unfolding method was put forward, which allowed one
to construct an unfolded formulation of 4d scalar electrodynamics. In brief, it consists in the
initial postulation of the form of an unfolded master-field and the subsequent derivation of
unfolded equations as identities for this master-field.
In this note, we apply the method of [16], slightly improving it, in order to construct an
unfolded formulation of 4d on-shell pure Yang–Mills theory.
The note is organized as follows. First, we present basic definitions of the unfolded dynamics
approach. Next, we construct unfolded Yang–Mills equations by fixing the form of the unfolded
master-fields and using certain operator relations that we derive. Then we discuss the properties
of the resulting unfolded system and comment on the unfolding maps it defines. In conclusion,
we provide a brief overview of possible directions for further research.

Unfolded dynamics approach

Within the unfolded dynamics approach [3–8], a field theory is formulated in terms of first-order
equations
dW A (x) + GA (W ) = 0 (1)
on unfolded fields W A (x), which are exterior forms on a space-time manifold M d . Here A
stands for all indices of an unfolded field, d is the exterior differential on M d and GA (W ) is
built from exterior products of unfolded fields (the wedge symbol is omitted throughout the
paper). For every W A there is one and only one unfolded equation (1).
There is a consistency condition for the system (1)

δGA
GB ≡ 0, (2)
δW B

2
which follows from d2 ≡ 0 and restricts possible GA . Considering this, unfolded equations (1)
possess a manifest gauge symmetry1

δGA
δW A = dεA (x) − εB , (3)
δW B
so that every (n > 0)-form field W A gives rise to a gauge symmetry with a (n − 1)-form
parameter εA (x). 0-form fields are transformed only under gauge symmetries of 1-form fields
via the second term in (3).
In dynamical field theories, a spectrum of unfolded fields is infinite, since they parameterize
all d.o.f. Usually, the space of unfolded fields possesses some grading bounded from below, so
that equations (1) express (perhaps, in a very complicated nonlinear way) higher-grade fields in
terms of derivatives of the lower-grade ones. In addition, (1) may implicitly impose dynamical
constraints on the lowest-grade fields, in which case the system is said to be on-shell.
Ultimately, an unfolded system (1) describes a theory of lowest-grade fields (also referred
to as primaries), possibly subjected to some differential constraints (e.o.m. or whatever), while
higher-grade fields form (infinite) towers of their covariant differential descendants. Impor-
tantly, an unfolded formulation is manifestly diffeomorphism- and gauge-invariant and provides
the control over d.o.f. of the theory. All this makes unfolded dynamics approach a powerful
tool for studying higher-spin gravity. But besides higher spins, these remarkable features can
be useful for more standard field theories as well. Here our goal is to reformulate 4d pure
Yang–Mills theory in the unfolded form (1), making use of the unfolding method proposed in
[16].

Yang–Mills fields and auxiliary spinors

In sl(2, C)-spinor notations, 4d Yang–Mills equations together with Bianchi identities are

Dβ α̇ F β α = 0, Dαβ̇ F̄ β̇ α̇ = 0, (4)

where the (anti-)self-dual components of the strength tensor are

∂ ∂
Fαα := Aα β̇ − i[Aαβ̇ , Aα β̇ ], F̄α̇α̇ := β α̇
Aβ α̇ − i[Aβ α̇ , Aβ α̇ ], (5)
∂xαβ̇ ∂x
and the covariant derivative is
∂
Dαα̇ := − i[Aαα̇ , •], (6)
∂xαα̇
so that
[Dαα̇ , Dβ β̇ ] = −iǫαβ F̄α̇β̇ − iǫα̇β̇ Fαβ . (7)
Antisymmetric spinor metric
   
0 1 αβ α̇β̇ 0 1
ǫαβ = ǫα̇β̇ = , ǫ =ǫ = (8)
−1 0 −1 0
1
Strictly speaking, this is true provided that consistency holds regardless of the dimension of M d , which is
the case for all known examples [8].

3
moves spinor indices up/down according to

vα = ǫβα v β , v α = ǫαβ vβ , v̄α̇ = ǫβ̇ α̇ v̄ β̇ , v̄ α̇ = ǫα̇β̇ v̄β̇ . (9)

Indices of multispinors, denoted with the same letter, are either contracted or symmetrized,
depending on their relative positions (the same holds for dotted indices)

Tαα := T(α1 α2 ) , Tα α := ǫαβ Tαβ . (10)

In order to unfold Yang–Mills theory, one has to introduce, on top of primaries Fαα and
F̄α̇α̇ , the infinite towers of all their differential on-shell descendants. This can be conveniently
performed by means of auxiliary commuting spinors2 Y = (y α , ȳ α̇), then the whole towers get
packed into unfolded Yang–Mills master-fields, which we postulate to be of the form

F (Y |x) = eDyȳ Fαα (x)y α y αe−Dyȳ , F̄ (Y |x) = eDyȳ F̄α̇α̇ (x)ȳ α̇ ȳ α̇ e−Dyȳ , (11)

where right-acting derivatives are contracted with spinors as

Dy ȳ := y α ȳ α̇Dαα̇ . (12)

Unfolded master-fields (11) contain primary Yang-Mills tensors as (anti-)holomorphic in Y

components
Fαα (x)y α y α = F (Y |x)|ȳ=0 , F̄α̇α̇ (x)ȳ α̇ ȳ α̇ = F̄ (Y |x)|y=0 , (13)
together with an infinite sequence of their fully symmetrized traceless covariant derivatives
of all orders. This constitutes a set of all independent covariant descendants of the primary
Yang–Mills tensor, since antisymmetrizations and contractions are determined by (4) and (7).
By construction, unfolded master-fields (11) inherit the adjoint representation of the gauge
algebra of primary tensors.
The aforementioned grading on the space of unfolded fields can be introduced in terms of
the spinor Euler operators
N := y α ∂α , N̄ := ȳ α̇ ∂¯α̇ , (14)
where ∂α and ∂¯α̇ are y α - and ȳ α̇-derivatives. From their definitions (11), master-fields obey

(N − N̄)F = 2F, (N − N̄ )F̄ = −2F̄ , (15)

which corresponds to helicities ±1 (in higher-spin gravity, strength tensors with |N −N̄ |F = 2sF
describe spin-s massless fields) and, further,

Dy ȳF = N̄ F, Dy ȳF̄ = N F̄ . (16)

Then a grading operator on the space of unfolded fields can be defined as

1
G := (N + N̄ − 2). (17)
2
2
In 4d Vasiliev equations [3, 4], these spinors become generators of an associative higher-spin gauge algebra
through a certain non-commutative star product defined on them.

4
It decomposes the space of master-fields F into a direct sum of its eigenspaces (the same applies
to F̄ )
X∞
F (Y |x) = F (n) (Y |x), GF (n) = nF (n) , (18)
n=0

so that the primary has 0-grade, while n-grade unfolded fields represent its covariant derivatives
of n-th order, as follows from (11).
To formulate the unfolded equations (1), one needs to express the derivatives of the unfolded
fields in algebraic terms. In our particular case, the task is to express Dαβ̇ F for F (11) in terms
of Y and ∂/∂Y acting on F and F̄ (and the same for Dαβ̇ F̄ ). Before we start processing (11),
let us work out some general operator relations, which simplify the analysis implied by the
method of [16]. Combining a commutator formula
Z 1
D̂
[Â, e ] = dtetD̂ [Â, D̂]e−tD̂ eD̂ (19)
0

with an Euler-operator representation of a homotopy integral

Z 1
1
dttk F (tz) = ∂ F (z), (20)
0 z ∂z + 1 + k

one obtains
1 D̂ 1 −D̂
[Â, eD̂ ] = ( e [Â, D̂]e−D̂ )eD̂ = eD̂ ( e [Â, D̂]eD̂ ), (21)
ND̂ ND̂
where the inverse Euler operator is understood as (for F (0) = 0)
Z 1
1 1
F (D̂) := dt F (tD̂). (22)
ND̂ 0 t
Applying this to an operator of the form

B̂ = eD̂ Ĉe−D̂ (23)

yields
1 D̂
[Â, B̂] = [( e [Â, D̂]e−D̂ ), B̂] + eD̂ [Â, Ĉ]e−D̂ . (24)
ND̂
Next, considering the case Ĉ = D̂, from (24) one gets
1
eD̂ [Â, D̂]e−D̂ = [Â, D̂] − (eD̂ [[Â, D̂], D̂]e−D̂ ), (25)
ND̂ − 1

so that (24) can be equivalently rewritten as

1
[Â, B̂] = [( eD̂ [D̂, [Â, D̂]]e−D̂ ), B̂] + [[Â, D̂], B̂] + eD̂ [Â, Ĉ]e−D̂ . (26)
ND̂ (ND̂ − 1)

For convenience, we denote an "unfolding map" for an arbitrary function Cα(n),β̇(m) (Y |x) as

≪ Cα(n),β̇(m) (Y |x) ≫:= eDyȳ Cα(n),β̇(m) (Y |x)e−Dyȳ , (27)

5
so that, in particular,

F (Y |x) =≪ Fαα (x)y α y α ≫, F̄ (Y |x) =≪ F̄α̇α̇ (x)ȳ α̇ ȳ α̇ ≫ . (28)

Then for an arbitrary C(Y |x) one has from (26)

1
≪ ∂µ C ≫= (∂µ − Dµα̇ ȳ α̇) ≪ C ≫ +iyµ [ F̄ , ≪ C ≫], (29)
(N + 1)(N + 2)
1
≪ ∂¯µ̇ C ≫= (∂¯µ̇ − Dαµ̇ y α ) ≪ C ≫ +iȳµ̇ [ F, ≪ C ≫]. (30)
(N̄ + 1)(N̄ + 2)
From here on, all derivatives inside the angle brackets always act only on an expression within
brackets and never differentiate unfolding exponents of (27). The square brackets stand for the
commutator in the gauge Lie algebra and the Euler operators of (26) are expressed in terms
of spinor Euler operators (14). Master-fields F and F̄ arise in (29)-(30) through (7) without
assuming Yang–Mills equations (4).
Now let us get down to solving the problem. Direct application of (24) to Dµµ̇ F (Y |x) gives
1
Dµµ̇ F = [ ≪ [Dµµ̇ , Dy ȳ] ≫, F ]+ ≪ Dµµ̇ Fαα y α y α ≫ . (31)
N
The task is to get rid of all covariant derivatives and angle brackets on the r.h.s. by re-expressing
∂
them in terms of unfolded master-fields F and F̄ acted on by Y ’s and ∂Y ’s. To accomplish
this, we have the following tools at our disposal: Yang–Mills equations (4) together with (7),
relations (29) and (30), Schouten identities for spinors and Jacobi identity of the gauge Lie
algebra.
Applying (7) to the first term on the r.h.s. of (31) and (4) (plus Schouten identities) to the
second one, one has
i 1 i 1 1
Dµµ̇ F = ȳµ̇ [ ≪ ∂µ Fαα y α y α ≫, F ]+ yµ [ ≪ ∂¯µ̇ F̄α̇α̇ ȳ α̇ȳ α̇ ≫, F ]+ ≪ ∂µ ∂¯µ̇ Dy ȳFαα y α y α ≫ .
2 N 2 N̄ 3
(32)
First, we process the first term on the r.h.s. (the second term will be resolved by conjuga-
tion). Applying (29) to the problematic factor yields
1
≪ ∂µ Fαα y α y α ≫= ∂µ F − Dµα̇ ȳ α̇ F + iyµ [ F̄ , F ]. (33)
(N + 1)(N + 2)
So, one needs to process Dµα̇ ȳ α̇ F . Contracting (32) with ȳ µ̇ gives
1 1
Dµµ̇ ȳ µ̇ F = iyµ [ F̄ , F ] + ≪ ∂µ Dy ȳFαα y α y α ≫ . (34)
N̄ − 1 3
On the other hand, from (16) and (29) one finds
1
≪ ∂µ Dy ȳFαα y α y α ≫= ∂µ N̄ F − Dµα̇ ȳ α̇N̄ F + iyµ [ F̄ , N̄ F ]. (35)
(N + 1)(N + 2)
Combining (34) and (35), one obtains after rearranging the Euler-operator ratios
N̄ 2 1 1
Dµα̇ ȳ α̇ F = ∂µ F + iyµ [ F̄ , F ] + iyµ [ F̄ , F ]. (36)
N +1 N +2 N +2 (N + 1)(N + 2)

6
Thus one finds for (33)
2 1
≪ ∂µ Fαα y α y α ≫= (∂µ F − iyµ [ F̄ , F ]). (37)
N +1 N +2
Conjugation gives
2 1
≪ ∂¯µ̇ F̄α̇α̇ ȳ α̇ȳ α̇ ≫= (∂¯µ̇ F̄ − iȳµ̇ [ F, F̄ ]). (38)
N̄ + 1 N̄ + 1
These bring first two terms on the r.h.s. of (32) to the admissible form.
Now we process the last term in (32). By virtue of (29) one has

≪ ∂µ ∂¯µ̇ Dy ȳFαα y α y α ≫= (∂µ − Dµα̇ ȳ α̇ ) ≪ ∂¯µ̇ Dy ȳFαα y αy α ≫ +

1
+iyµ [ F̄ , ≪ ∂¯µ̇ Dy ȳFαα y α y α ≫]. (39)
(N + 1)(N + 2)
Since obviously
≪ ∂¯µ̇ Fαα y α y α ≫= 0, (40)
one has from (30)
1
Dαµ̇ y α F = ∂¯µ̇ F + iȳµ̇ [ F, F ]. (41)
(N̄ + 1)(N̄ + 2)
Using this and contracting (32) with y µ yields
1
≪ ∂¯µ̇ Dy ȳFαα y αy α ≫= ∂¯µ̇ F − iȳµ̇ [ F, F ], (42)
(N̄ + 2)
which together with (36) turns (39) to
1 1
≪ ∂µ ∂¯µ̇ Dy ȳFαα y α y α ≫= ∂µ ∂¯µ̇ F − iȳµ̇ ∂µ [ F, F ] + i[ F̄ , yµ ∂¯µ̇ F ] +
N̄ + 2 (N + 1)(N + 2)
1 1 N̄ + 1 ¯
+yµ ȳµ̇ [ F̄ , [ F, F ]] + Dµµ̇ F − ∂µ ∂µ̇ F −
(N + 1)(N + 2) N̄ + 2 N +1
1 1 1
−2iyµ ∂¯µ̇ [ F̄ , F ] − iyµ ∂¯µ̇ [ F̄ , F ] +
N +2 N +2 (N + 1)(N + 2)
1 N̄ 2 1 1
+iȳµ̇ [ ( ∂µ F + iyµ [ F̄ , F ] + iyµ [ F̄ , F ]), F ] +
N̄ + 1 N + 1 N +2 N +2 (N + 1)(N + 2)
1 N̄ 2 1 1
+iȳµ̇ [ F, ( ∂µ F + iyµ [ F̄ , F ] + iyµ [ F̄ , F ])]. (43)
N N +1 N +2 N +2 (N + 1)(N + 2)
Substituting (37), (38) and (43) into (32) and combining like terms using the gauge algebra
Jacobi identity, the final result is
1 1 1 1 1
Dµµ̇ F = ∂µ ∂¯µ̇ F + iN[ ȳµ̇ ∂µ F, F ] − iyµ ∂¯µ̇ [ F̄ , F ] +
N +1 N(N + 1) N N +2 N +2
i 1 N +3 1
+[ yµ ∂¯µ̇ F̄ , F ] + yµ ȳµ̇ [ [ F̄ , F ], F ] +
(N + 1)(N + 2) 2 (N + 1)(N + 2) N + 2
3 1 1 1 1 1
+ yµ ȳµ̇ [ [ F, F̄ ], F ] + yµ ȳµ̇ [ [ F̄ , F ], F ]. (44)
2 (N + 1)(N + 2) N N +2 N +2 N

7
Poincaré symmetry and diffeomorphism-invariance
The manifestly Poincaré-invariant relation (44) is written in Cartesian coordinates, together
with (6) and (11), since they involve ∂x∂µµ̇ inside of Dµµ̇ . The unfolded dynamics approach
requires manifest coordinate-independence, which is ensured by formulating equations in terms
of exterior forms. To attain this, we switch to the fiber space picture: we claim that F (Y |x) and
F̄ (Y |x) are now 0-forms on the Minkowski base manifold with some local coordinates xn , while
spinor variables {y α , ȳ α̇} are coordinates in the fiber. This requires appropriate generalization
of Dµµ̇ in (44).
According to the ideology of unfolding, global Poincaré symmetry of a relativistic theory
should arise in terms of the unique general formula (3). This is achieved by introducing a
non-dynamical 1-form Ω(x), which takes values in Lie algebra iso(1, 3)

Ω = eαβ̇ Pαβ̇ + ω αα Mαα + ω̄ α̇α̇ M̄α̇α̇ , (45)

with Pαα̇ , Mαα and M̄α̇α̇ being generators of translations and rotations of R1,3 , and eαβ̇ and
ω αα (ω̄ α̇α̇ ) being 1-forms of a vierbein and a Lorentz connection.
Ω is subjected to the flatness condition (square brackets stand for the iso(1, 3)-commutator)
1
dΩ + [Ω, Ω] = 0, (46)
2
so that the corresponding gauge symmetry (3) is
δΩ = dε(x) + [Ω, ε] (47)
and describes an infinite-dimensional freedom in switching between all possible local coordi-
nates on R1,3 . This boils down to 10-dimensional global Poincaré symmetry after fixing some
particular solution Ω0 and restricting to those residual ε(x) which leave it invariant
dε0 + [Ω0 , ε0 ] = 0. (48)
The simplest non-degenerate global solution to (46) is provided by Cartesian coordinates

em αβ̇ = (σ̄m )β̇α , ωm αα = 0, ω̄m α̇α̇ = 0, (49)

with global symmetries (48) parameterized by x-independent ξ αβ̇ , ξ αα and ξ¯α̇α̇

εα0 β̇ = ξ αβ̇ + ξ α γ (σ̄m )β̇γ xm + ξ¯β̇ γ̇ (σ̄m )γ̇α xm , εαα αα

0 = ξ , ε̄α̇0 α̇ = ξ¯α̇α̇ . (50)
Analogously, in order to realize the Yang–Mills gauge symmetry via (3), one introduces a
1-form A(x), with Aαα̇ (x) being its expansion in the vierbein
A(x) = eαα̇ Aαα̇ . (51)
Now an appropriate coordinate-independent generalization of Dµµ̇ is a 1-form operator D, sup-
plemented by Lorentz-connection terms rotating fiber coordinates Y ,
D := d + ω αα yα ∂α + ω̄ α̇α̇ ȳα̇ ∂¯α̇ − i[A, •]. (52)
In Cartesian coordinates (49), this indeed boils down to
D = dxµµ̇ Dµµ̇ . (53)

8
Unfolded Yang–Mills equations and unfolding maps
Now we are ready to write down an unfolded system for Yang–Mills theory. Contracting (5)
and (44) with the vierbeins yields
1 1
dA + [A, A] = eα β̇ eαβ̇ ∂α ∂α F |ȳ=0 + eβ α̇ eβ α̇ ∂¯α̇ ∂¯α̇ F̄ |y=0 , (54)
4 4
1 ¯ + iN[ 1 1 1 1
DF = e∂ ∂F e∂ ȳF, F ] − iey ∂¯ [ F̄ , F ] +
N +1 N(N + 1) N N +2 N +2
i 1 N +3 1
+[ ey ∂¯F̄ , F ] + ey ȳ[ [ F̄ , F ], F ] +
(N + 1)(N + 2) 2 (N + 1)(N + 2) N + 2
3 1 1 1 1 1
+ ey ȳ[ [ F, F̄ ], F ] + ey ȳ[ [ F̄ , F ], F ] (55)
2 (N + 1)(N + 2) N N +2 N +2 N
plus a conjugate equation for F̄ resulting from exchanging barred and unbarred objects in (55).
Here
e∂ ∂¯ := eαβ̇ ∂α ∂¯β̇ , e∂ ȳ := eαβ̇ ∂α ȳβ̇ , ey ∂¯ := eαβ̇ yα ∂¯β̇ ey ȳ := eαβ̇ yα ȳβ̇ . (56)
A full spectrum of unfolded fields consists of a 1-form Ω describing Minkowski background,
a 1-form of the gauge potential A and the 0-forms of master-fields F (Y |x) and F̄ (Y |x) encoding
the Yang–Mills tensor together with an infinite tower of its covariant derivatives. Corresponding
unfolded equations are (46), (54) and (55) plus a conjugate for F̄ . The formulation is manifestly
diffeomorphism-invariant. 1-forms Ω and A give rise to two manifest symmetries in accordance
with (3): the global (after fixing Ω) Poincaré one and the local Yang–Mills one. The Yang–Mills
symmetry is realized as
δA(x) = Dε(x), δF (Y |x) = i[ε(x), F (Y |x)], δ F̄ (Y |x) = i[ε(x), F̄ (Y |x)]. (57)
By construction, the unfolded system is consistent, provided unfolded 0-form fields obey the
helicity constraint (15). At the same time, a direct check of (2) seems hardly executable due
to the complexity of the equations. One of the possible forms of solution to the system is (11)
with primary fields subjected to (4). Let us quickly derive this directly from (55).
Master-fields F and F̄ are assumed to be analytical in Y and obey (15). For the sake of
brevity, here and below we consider only an anti-selfdual component F , but everything applies
to F̄ as well. Acting on (55) with y β ȳ β̇ δeδβ β̇ produces (16), whose solution, accounting for (15),
is (11). On the other hand, acting on (55) with 12 ∂µ ∂ β δeδβ µ̇ and putting ȳ α̇ = 0 yields, accounting
for (15) again,
Dβ µ̇ F β µ = 0. (58)
Thus, the unfolded system presented above indeed provides a consistent manifestly diffeomorphism-
and gauge-invariant first-order formulation of 4d Yang-Mills theory. Unfolded master-fields
F (Y |x) and F̄ (Y |x) encode all on-shell d.o.f. of the Yang-Mills tensor as expansions in auxil-
iary spinors Y .
The system allows for two obvious reductions. The first one is the anti-selfdual case
F̄ (Y |x) = 0 (or the selfdual F (Y |X) = 0), then only first two terms on the r.h.s. of (55)
survive. In different (tensor) terms, this was presented in [22]. The second one is the abelian
case with all commutators vanishing, then only the first term on the r.h.s. of (55) remains.
One can think of the equation (55) as defining an unfolding map from x-space to Y -space
Fαα (x)|on−shell → F (Y |x) → F (Y ) := F (Y |x = 0). (59)

9
Then (11) explicitly realizes the first arrow in (59). The field F (Y ) carries precisely the same
information as on-shell Fαα does. In a sense, spinors Y = {y α; ȳ α̇ } effectively replace space-time
coordinates xαα̇ for on-shell configurations, hence being conjugate to spinor-helicity variables
that resolve light-like momenta pαα̇ = π α π̄ α̇ . This is the way the unfolded system imposes e.o.m.
on primary fields: via (59), it maps 4d space-time fields onto an effectively 3d hypersurface (in
the sense that y αȳ α̇ is a light-like vector). Note, however, that the role of auxiliary spinors is
much more important and sophisticated. As follows directly from (11), they in fact equalize
and mix translational and spin degrees of freedom. So they should not be thought of simply as
coordinates on some null hypersurface in 4d Minkowski space.
To get a better idea of the unfolding map (59), it is instructive to consider the abelian case,
where the unfolding exponent in (11) boils down to a space-time translation so that (59) can
be constructed explicitly. Then a plane-wave solution to (54), (55) in Cartesian coordinates is
β β̇ β β̇ +y β ȳ β̇ )
Aαα̇ (x) = πα µ̄α̇ eiπβ π̄β̇ x + c.c., F (Y |x) = iπ̄α̇ µ̄α̇ (πα y α )2 eiπβ π̄β̇ (x (60)

with µ̄α̇ being an arbitrary reference spinor, defined up to a gauge transformation µ̄α̇ → µ̄α̇ +
const · π̄α̇ . Putting x = 0, one has
β π̄ β̇
F (Y ) = iπ̄α̇ µ̄α̇ (πα y α )2 eiπβ y β̇ ȳ , (61)

which represents a plane-wave Maxwell tensor formulated purely in Y -terms.

In fact, one can start directly from (61) and then make use of (55) in order to fully recover
x-dependence. It means that although we have derived the unfolded equation (55) starting
from postulating (11), for (55) the expression (11) per se is nothing more than just one of many
possible forms of solution. In particular, instead of (59) one can think of (55) as defining a
Y -to-x map (a similar interpretation has been proposed in [23])

F (Y ) → F (Y |x) → Fαα (x)|on−shell . (62)

In this picture, F (Y ) is completely unconstrained aside from the helicity condition

(N − N̄)F (Y ) = 2F (Y ) ⇒ F (Y ) = Fαα (y ȳ)y αy α . (63)

This distinguishes the unfolded formulation constructed here from that of [8], where the fiber
structure is tantamount to the base one. In the abelian case, the map (62) is obviously realized
as
1
F (Y |x) = exp( xβ β̇ ∂β ∂¯β̇ )Fαα (y ȳ)y αy α . (64)
N +1
It generates a solution to (55), and hence implicitly to Maxwell equations, for arbitrary Fαα (y ȳ).
This signals that relativistic dynamics in fact can be realized in terms of Y without any reference
to a space-time. Corresponding action principle for an arbitrary-mass integer-spin field was
constructed in [24] by means of supplementing the set of Y with a Lorentz-invariant proper-
time coordinate serving as an evolution parameter.
The unfolding map (62) hints that the unfolded dynamics approach may provide a new way
to study the problem of integrability. In particular, the question is whether it is possible to
extend (64) to the non-abelian case. If so, an appropriate unfolding map will generate solutions
to Yang–Mills equations in Minkowski space.

10
 Another interesting problem is to relate the unfolded dynamics approach to twistor theory
[25, 26]. Potentially, twistors may arise from treating an unfolded system as defining an un-
folding map to x-space from some complex plane in (Y |x)-space, different from x = 0 of (62)
and associated with the incidence relation. Then this complex plane should be identified with
the twistor space, and the corresponding unfolding map with the Penrose transform.

Conclusion
In this note, we constructed an unfolded formulation of 4d pure Yang–Mills theory making
use of the unfolding method proposed in [16] and improved here by preliminary derivation of
general relations for unfolded functions.
A natural generalization of our result would be the inclusion of charged matter. It is straight-
forward and should not cause any problems, as shown by the example of scalar electrodynamics
[16].
Another possible direction of further research is to include supersymmetry, as well as to man-
ifest conformal symmetry. To this end one needs to introduce the corresponding gauge 1-forms
of (super)conformal gravity in addition to the Poincaré connection and to deform appropriately
the unfolded equations. In its turn, this requires further non-trivial modifications of the un-
folding method of [16]. In particular, it would be interesting to unfold N = 4 super-Yang–Mills
theory in order to apply unfolding tools to the problems of AdS/CF T and amplitudes.
Quantization of the unfolded Yang–Mills system can be performed along the lines of [12] but
with necessary modifications in order to include ghosts. The first step is to build an off-shell
extension of the unfolded system constructed here, which is equivalent to coupling it to external
currents [27, 28].
Finally, there are two vague but potentially promising topics: the study of integrability
and the derivation of twistors. Both of them seem to be related to the investigation of various
unfolding maps defined by the unfolded Yang–Mills equations.

Acknowledgments
The author is grateful to A.V. Korybut, D.S. Ponomarev and M.A. Vasiliev for valuable com-
ments and remarks.

References
[1] E.S. Fradkin, M.A. Vasiliev, Annals Phys. 177 (1987) 63.

[2] X. Bekaert, N. Boulanger, A. Campoleoni, M. Chiodaroli, D. Francia, M. Grigoriev, E.

Sezgin, E. Skvortsov, Snowmass White Paper: Higher Spin Gravity and Higher Spin Sym-
metry, [arXiv:2205.01567].

[3] M.A. Vasiliev, Phys.Lett.B 243 (1990) 378-382.

[4] M.A. Vasiliev, Phys.Lett.B 285 (1992) 225-234.

[5] M.A. Vasiliev, Annals Phys. 190 (1989) 59-106.

11
 [6] M.A. Vasiliev, Class.Quant.Grav. 11 (1994) 649-664.

[7] O.V. Shaynkman, M.A. Vasiliev, Theor.Math.Phys. 123 (2000) 683-700, Teor.Mat.Fiz.
123 (2000) 323-344 [hep-th/0003123].

[8] M.A. Vasiliev, Int.J.Geom.Meth.Mod.Phys. 3 (2006) 37-80 [hep-th/0504090].

[9] M.A. Vasiliev, J.Phys.A 46 (2013) 214013 [arXiv:1203.5554].

[10] F. Diaz, C. Iazeolla, P. Sundell, Fractional Spins, Unfolding, and Holography: I. Parent
field equations for dual higher-spin gravity reductions, [arXiv:2403.02283].

[11] F. Diaz, C. Iazeolla, P. Sundell, Fractional Spins, Unfolding, and Holography: II. 4D
Higher Spin Gravity and 3D Conformal Dual, [arXiv:2403.02301].

[12] N. Misuna, JHEP 12 (2023) 119 [arXiv:2208.04306].

[13] V.E. Didenko, JHEP 10 (2022) 191 [arXiv:2209.01966].

[14] V.E. Didenko, A.V. Korybut, Phys.Rev.D 108 (2023) 8, 086031 [arXiv:2304.08850].

[15] E. Joung, M. Kim, Y. Kim, JHEP 12 (2021) 092 [arXiv:2108.05535].

[16] N. Misuna, Scalar Electrodynamics and Higgs Mechanism in the Unfolded Dynamics Ap-
proach, [arXiv:2402.14164].

[17] D.S. Ponomarev, M.A. Vasiliev, JHEP 1201 (2012) 152 [arXiv:1012.2903].

[18] N.G. Misuna, M.A. Vasiliev, JHEP 05 (2014) 140 [arXiv:1301.2230].

[19] M.V. Khabarov, Yu.M. Zinoviev, Nucl.Phys.B 953 (2020) 114959 [arXiv:2001.07903].

[20] I.L. Buchbinder, T.V. Snegirev, Yu.M. Zinoviev, JHEP 08 (2016) 075 [arXiv:1606.02475].

[21] N.G. Misuna, Phys.Lett.B 840 (2023) 137845 [arXiv:2201.01674].

[22] E. Skvortsov, R. Van Dongen, JHEP 08 (2022) 083 [arXiv:2204.09313].

[23] M.A. Vasiliev, Lect.Notes Phys. 892 (2015) 227-264 [arXiv:1404.1948].

[24] N.G. Misuna, Spinors, Proper Time and Higher-Spin Fields, [arXiv:2301.02207].

[25] M.G. Eastwood, R. Penrose, R.O. Wells, Commun.Math.Phys. 78 (1981) 305–351.

[26] N.M.J. Woodhouse, Class.Quant.Grav. 2 (1985) 257-291.

[27] N. Misuna, Phys.Lett.B 798 (2019) 134956 [arXiv:1905.06925].

[28] N.G. Misuna, JHEP 12 (2021) 172 [arXiv:2012.06570].

12