ELSEVIER

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Nuclear Physics B 890 (2015) 279-301

www.elsevier.com/locate/nuclphysb

Higher spins in hyper-superspace

Ioannis Florakisab, Dmitri Sorokinc, Mirian Tsulaiad*

a Department of Physics, CERN — Theory Division, CH-1211 Geneva 23, Switzerland b Max-Planck-Institut für Physik, Werner-Heisenberg-Institut, 80805 München, Germany c INFN, Sezione di Padova, via F. Marzolo 8, 35131 Padova, Italy d Faculty of Education, Science, Technology and Mathematics, University of Canberra, Bruce, ACT 2617, Australia

Received 15 September 2014; received in revised form 13 November 2014; accepted 21 November 2014

Available online 26 November 2014

Editor: Stephan Stieberger

Abstract

We extend the results of arXiv:1401.1645 on the generalized conformai Sp(2n)-structure of infinite multiplets of higher-spin fields, formulated in spaces with extra tensorial directions (hyperspaces), to the description of OSp( 1|2n)-invariant infinite-dimensional higher-spin supermultiplets formulated in terms of scalar superfields on flat hyper-superspaces and on OSp(1ln) supergroup manifolds. We find generalized su-perconformal transformations relating the superfields and their equations of motion in flat hyper-superspace with those on the OSp(1ln) supermanifold. We then use these transformations to relate the two-, three-and four-point correlation functions of the scalar superfields on flat hyperspace, derived by requiring the OSp(1|2n) invariance of the correlators, to correlation functions on the OSp(1ln) group manifold. As a byproduct, for the simplest particular case of a conventional N = 1, D = 3 superconformal theory of scalar superfields, we also derive correlation functions of component fields of the scalar supermultiplet including those of auxiliary fields.

© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/). Funded by SCOAP3.

* Corresponding author.

E-mail addresses: ioannis.florakis@cern.ch (I. Florakis), dmitri.sorokin@pd.infn.it (D. Sorokin), mirian.tsulaia@canberra.edu.au (M. Tsulaia).

http://dx.doi.org/10.1016/j.nuclphysb.2014.11.017

0550-3213/© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/). Funded by SCOAP3.

1. Introduction

In [1] we have studied some aspects of the description of infinite sets of integer and halfinteger massless higher-spin fields in flat and anti-de-Sitter (AdS) spaces in terms of scalar and spinor 'hyperfields' propagating in hyperspaces. In addition to a conventional space-time as a subspace, hyperspaces are endowed with extra tensorial coordinates encoding the spin degrees of freedom of conventional space-time fields. This formulation, which was originally put forward by Fronsdal as an alternative to the Kaluza-Klein theory [2], has been extensively developed by several authors [3-22].

The theories on tensorially extended (super)spaces, which we will henceforth refer to as hyper-(super)spaces, offer many interesting and challenging problems regarding higher-spin fields, one of them being the further development and study of generalized (super)conformal theories on these spaces. This motivated our recent work [1] in which, using generalized conformal transformations, we established an explicit relation between the equations of motion of hyperfields on flat hyperspace and on Sp(n) group-manifolds, the latter being tensorial generalizations1 of AdS spaces. This relation was then employed in order to explicitly derive the Sp(2n)-invariant two-, three- and four-point correlation functions for fields on Sp(n) group manifolds, from the known Sp(2n)-invariant correlation functions on flat hyperspaces, thus generalizing the results obtained in [5,10,23].

In this paper we further extend the results of [1] to the description of supersymmetric systems of higher-spin fields in hyper-superspaces, which were previously studied, e.g. in [3,4,6-8,11,14, 20,24]. In particular, by means of a generalized superconformal transformation, we establish an explicit relation between the superfield equations of motion [11] on flat hyper-superspace and on an OSp(1\n) supergroup manifold. Furthermore, the explicit solution of the generalized superconformal Ward identities allows us to derive the OSp(1 \2n)-invariant two-, three- and four-point superfield correlation functions on flat hyper-superspace and, consequently, using the generalized superconformal transformations, we obtain the corresponding correlation functions on the OSp(1\n) group manifolds. Our results, therefore, generalize the superfield description and computation of superfield correlators in conventional superconformal field theories, considered, e.g. in [25-28], to superconformal higher-spin theories. A byproduct of our analysis is the derivation of correlation functions involving the component fields of the scalar supermultiplet, including the auxiliary fields, for the simple special case of a three-dimensional N = 1 superconformal theory of scalar superfields.

As in the case of the N = 1, D = 3 superconformal theory, the fact that 3- and 4-point correlation functions are non-zero for hyperfields of an anomalous conformal weight may indicate the existence of interacting conformal higher-spin fields which involve higher orders of their field strengths.

It should be noted that in the literature [29-49] various supersymmetric higher-spin systems have been considered in either irreducible or reducible representations of the Poincare and AdS groups (see,e.g. [50,51] for a discussion of reducible higher-spin multiplets in the "metric-like" approach). As we will see, the systems of integer and half-integer higher-spin fields considered in [3,4,6-8,11,14,20] and in this paper form irreducible infinite-dimensional supermultiplets of space-time supersymmetry. These supersymmetric higher-spin systems are therefore different

1 Here Sp(n) stands for the real non-compact form Sp(n, R) of Sp(n, C), where R will be omitted for brevity.

from finite-dimensional higher-spin supermultiplets considered in [29-49]. We will provide the algebraic reasoning for this in Section 2.4.

The paper is organized as follows. Section 2 begins with a review of some basic known results about hyper-superspaces. We describe in detail the generalized superconformal algebra, the realization of the generalized superconformal group OSp(1\2n) on hyper-superspace and the precise connection between generalized and conventional conformal weights for scalar superfields and their components in various dimensions. Finally, we demonstrate how an infinite-dimensional N = 1 supersymmetry multiplet is formed by the component fields of the hyper-superfield in the case of four-dimensional flat space-time.

In Section 3 we provide a description of the geometric structure of OSp(1\n) manifolds. These manifolds exhibit the property of generalized superconformal flatness (or GL-flatness) observed earlier in [7,8], which is similar to the superconformal flatness property of certain conventional AdS superspaces and superspheres [52-56]. We then consider the relation between the OSp(1 \2n)-invariant field equations for scalar superfields on flat hyper-superspace and those on the OSp(1 \n) group manifold derived in [11]. We show that, similarly to the non-supersymmetric case [1], the supersymmetric field equations on flat hyper-superspace and on OSp(1\n) group manifolds are related to each other via a generalized superconformal transformation of the scalar hyper-superfield and its derivatives.

In Section 4, as a preparation for the computation of correlation functions on flat hyper-superspace and on OSp(1\n) supergroup manifolds, we consider the simplest example of an OSp(1\4)-invariant superconformal theory of a conventional N = 1, D = 3 massless scalar su-perfield. Even though higher-spin fields are absent in this case, it is a simple setup in which one can illustrate the salient features of our approach. To this end, we present the OSp(1 \4)-invariant two-, three- and four-point correlation functions of scalar superfields, as well as the correlators of the component fields of the scalar supermultiplet, including those of auxiliary fields.

Finally, in Section 5 we use the requirement of OSp(1\2n) invariance to derive the expressions for two-, three- and four-point correlation functions of the scalar hyper-superfields. Again, in a complete analogy with the non-supersymmetric systems [1], the correlation functions on flat hyper-superspaces and OSp(1\n) supergroup manifolds are related via generalized superconformal Weyl rescaling. Thus, our basic result is that the GL-flatness is a key property of Sp(n) and OSp(1\n) manifolds that renders them amenable to the same type of analysis as for the case of flat hyper (super) spaces.

We conclude with a discussion on open problems and perspectives for further development of the hyperspace formulation of higher-spin fields.

2. Scalar superfields in flat hyper-superspace, equations of motion and correlators

2.1. Flat hyper-superspace and its symmetries

The flat hyper-superspace (see, e.g. [3,4,11]) is parametrized by "("2+1) bosonic matrix coordinates Xvv = XvV and n real Grassmann-odd 'spinor' coordinates (v = 1, ..., n). We call 'spinors', since they are indeed so from the perspective of conventional space-time, which is a subspace of hyperspace.

For instance, when n = 4, we can decompose the ten bosonic coordinates Xvv using the Majorana (real) representation of the gamma-matrices of a D = 4 space-time as follows

X^v = XvV = 1 xm(ymrv + 1 y^iymn)^, v,v = 1, 2, 3,4, m,n = 0, 1, 2, 3, (2.1)

where (ym)»v = (Ym)v» = C^(ym)rv, (ymn)»v = (ymn)v» = C^(ymn)rv, with CT = -C being the charge conjugation matrix and the gamma-matrices (ym)»v satisfy the Clifford algebra {Ym, Yn} = 2nmn. The space-time metric signature is chosen to be mostly plus (-, +, ■ ••, +).

The four coordinates xm parametrize the conventional flat space-time which is extended to flat hyperspace by adding six extra dimensions, parametrized by ymn = -ynm. This bosonic hyperspace is then further extended to the hyper-superspace by adding four Grassmann-odd directions parametrized by 0», which transform in the spinor representation of the D = 4 Lorentz group 50(1, 3).

The supersymmetry variation of the coordinates

80» = €», 8X»v = -ie(»0v), (2.2)

leaves invariant the Volkov-Akulov-type one-form

n»v = dX»v + i0(»d0v). (2.3)

The round brackets denote symmetrization of indices with the standard normalization

Y(»i...»k) = 1 (y»i...»k + all permutations of indices). (2.4)

The supersymmetry transformations form a generalized super-translation algebra

{Q»,Qv } = 2P»v, [Q»,Pvp ]=0, [P»v,PPx] = 0, (2.5)

with P»v generating translations along X»v. Namely, 8X»v = iapXPpx ■ X»v = a»v, with a»v being constant parameters.

The realization of P»v and Q» as differential operators is given by

d v d P»v = -l dX»v =-ld»v^ Q» =d» -10 d» = do», (2.6)

where, by definition,

B»vXpk = 8»p8^). (2.7)

Furthermore, in the case n = 4, D = 4, the partial derivative associated with (2.1) takes the form

d _ I (Ym) d . 1 (ymn) d (2 8)

»v 2 »v dxm + 2 '»v^ymn. ^ ' '

The algebra (2.5) is invariant under rigid GL(n) transformations

Q» = g»vQv, P» v = g»PgvlPpx, (2.9)

generated by

L»v = -2i^Xvp + l-0v0p)dp» - i0vQ», (2.10) which act on P v and Q as

[P» v,Lkp ] = -i (8pPvk + 8pP»x), [Q»,Lvp ] = -i8pQv, (2.11) and close into the gl(n) algebra

[Lv»,Lkp] = i(8»Lvp - 8pvLx»). (2.12)

The algebra (2.5), (2.11) and (2.12) is the hyperspace counterpart of the conventional super-Poincare algebra enlarged by dilatations. That this is so can be most easily seen by considering, e.g. n = 2 (i.e. x = 1, 2), in which case this algebra is recognized as the D = 3 super-Poincare algebra with Liv - 2VlxLpp = Mm(Ym)iv generating the SL(2, R) ~ SO(1, 2) Lorentz rotations (note that m = 0, 1, 2) and D = 1 Lpp being the dilatation generator. Note that the factor 2 in the definition of the dilatation generator is required in order to have the canonical scaling of the momentum generator Pxv with weight 1 and the supercharge Qx with weight 1, as follows from Eq. (2.11).

This algebra may be further extended to the OSp(1\2n) algebra, generating generalized su-perconformal transformations of the flat hyper-superspace, by adding the additional set of super-symmetry generators

sx = ^xxv + l-exev^Qv, (2.13)

together with the generalized conformal boosts

Kxv = i(xiP + l-exe^(Xvk + l-eve^jdPk - w(isv). (2.14)

The generators Sx and Kxv form a superalgebra similar to (2.5)

{Sx,Sv} = -2Kxv, [Sx,KvP ] = 0, [Kiv,KPk] = 0, (2.15)

while the non-zero (anti)commutators of Sx and Kxv with Qx, Pxv and Lxv read

{Qx,Sv} = -Liv, [Sx,Pvp ] = iS'ivQp), [Qi,Kvp ] = -isivSp),

[Sx,LvP ] = iSx;SP. (2.16)

2.2. Generalized superconformal algebra OSp(1\2n)

We now collect together all the non-zero (anti)commutation relations among the generators of the OSp(1\2n) algebra

{Qi,Q}} = 2Piv, [Q i,Pvp 1 = 0, [Piv,Ppi] = 0,

{Sx,Sv} = -2Kxv, [Sx,KvP ] = 0, [Kxv,KPk] = 0,

{Qx,Sv} = -Liv, [Sx,Pvp ] = iSiQp), [Qi,Kvp ] = -iS(vSp),

[Piv,Lxp] = -i{SPpPvk + spPi,), [Qi,Lvp ] = -HQ

[Sx,LvP] = iSi[SP, [LvX,LkP] = i[S'lLvP - SpLkx),

[Kxv,LkP ] = i[SxKvP + SvxKiP),

[Piv,KkP ] = 4 [SpLvk + SPLxk + SkLvP + SkLxP). (2.17)

Let us note that in the case n = 4, in which the physical space-time is four-dimensional (see Eq. (2.1)) the generalized superconformal group OSp(1\8) contains the D = 4 conformal symmetry group SO(2, 4) ~ SU(2, 2) as a subgroup, but not the superconformal group SU(2, 2\1). The reason being that, although OSp(1\8) and SU(2, 2\ 1) contain the same number of (eight) generators, the anticommutators of the former close on the generators of the whole Sp(8), while

those of the latter only close on an U(2, 2) subgroup of Sp(8), and the same supersymmetry generators cannot satisfy the different anti-commutation relations simultaneously. In fact, the minimal 0Sp-supergroup containing SU(2,211 ) as a subgroup is 0Sp(2|8).

2.3. Scalar superfields and their OSp(1\2n)-invariant equations of motion

Let us now consider a superfield ®(X, 0) transforming as a scalar under the super-translations given in Eq. (2.6)

8® — -(eaQa + ia»vP^v )®. (2.18)

To construct equations of motion for ®(X, 0) which are invariant under (2.18) and comprise the equations of motion of an infinite tower of integer and half-integer higher-spin fields with respect to conventional space-time, we introduce the spinorial covariant derivatives

DIX — + i0vdv{D^,Dv} — 2id»v, (2.19)

which (anti)commute with QM and Piv.

The ®-superfield equations then take the form [11]

DiDv]® — 0, (2.20)

where the brackets denote the anti-symmetrization of indices with unit overall strength similarly to (2.4). As was shown in [11], these superfield equations imply that all components of ®(X, 0) except for the first and the second one in the 0M-expansion of ®(X, 0) should vanish

®(X, 0) — b(X) + i0IfiAX) + i0»0vA^v + ■■■, (2.21)

(i.e. AM1.. ..vk — 0 for k > 1) while the scalar and spinor fields b(X) and f^(X) satisfy the equations first derived in [4]

(d^vdpx - dipdvx)b(X) — 0, (2.22)

divfp(X) - dipfv(X) — 0. (2.23)

For n — 4, 8 and 16 these equations encode the Bianchi identity and equations of motion for the curvatures of infinite towers of conformally invariant, massless higher-spin fields in 4-, 6- and 10-dimensional flat space-time, respectively (see [4,12]).

The superfield equations (2.20) are invariant under the generalized superconformal OSp(1\2n) symmetry, provided that ®(X, 0) transforms as a scalar superfield with the 'canonical' generalized scaling weight 5, i.e.

8® — -(^Qi + HS + iaivPiv + ikivKiv + ig/Lv1)®

- 2 (gi1 - kiv(XIiv + 2-0^0^ + (2.24)

where the factor 1 in the second line is the generalized conformal weight and e1, aIiv, kiv and g^v are the rigid parameters of the OSp(1 \2n) transformations.

Scalar superfields with anomalous generalized conformal dimension A transform under OSp(1\2n) as

8® = -( e^Q^ + ZS + ia,ÂV + ik^vK'ÂV +

- A^g^ - + 20M0V) + Z^®. (2.25)

It is instructive to demonstrate how the generalized conformai dimension A, which is defined to be the same for all values of n in OSp(1\2n), is related to the conventional conformai weight of scalar superfields in various space-time dimensions. As we have already mentioned in Section 2.1, the dilatation operator should be identified with D = 2 Li1. Therefore, considering a GL(n) transformation (2.25) with parameter g^V

8® = -igiVLVi®,

the part of the transformation corresponding to the dilatation reads

8d$ = -l-giiLVV$ = --gi^D® = -igD®, (2.26)

where g = ng^i is the genuine dilatation parameter. From (2.25) it then follows that the conventional conformal weight AD of the scalar superfield is related to the generalized one A via

ad = - A. (2.27)

In the n = 2 case corresponding to the N = 1, D = 3 scalar superfield theory the two conformal dimensions coincide, whereas in the case n = 4 describing conformal higher-spin fields in D = 4 one finds A4 = 2A. Relation (2.27) indeed provides the correct conformal dimensions of scalar superfields (and consequently of their components) in the corresponding space-time dimensions. For instance, when A = in D = 3 one finds 1 as the canonical conformal dimension of the scalar superfield, while in the cases D = 4 and D = 6, n = 8 it is found to be equal to one and two, respectively. For convenience, we shall henceforth associate the scaling properties of the fields to the universal D- and n-independent generalized conformal weight A.

2.4. Infinite-dimensional higher-spin representation of N = 1, D = 4 supersymmetry

Using the example of n = 4 in D = 4 we will now show that in four space-time dimensions, the fields of integer and half-integer spin s = 0, 2, 1, ..., œ encoded in b(X) and f^(X) form an irreducible infinite-dimensional supermultiplet with respect to the supersymmetry transformations generated by the generalized super-Poincaré algebra (2.5)-(2.8). The hyperfields b(X) and f^(X), satisfying (2.23), transform under the supertranslations (2.18) as follows

8b(X) = -ieifi(X), 8fi(X) = -eVdVib(X). (2.28)

The D = 4 higher-spin field curvatures are contained in b(X) and f^(X) as the components of the series expansion in the powers of the tensorial coordinates ymn of the flat hyperspace (2.1)

b{xl,ymn)

= 4(x) + ym1n1 Fmini (x) + ym1n1 ym2n2

Rm\n\,m2n (x) ^ nm\m2 dn\

ym1n1 .. .ym*ns\R (r)J____I

y y \_Rm\n\,...,msns\x ) + J,

fp (xl,ymn) = Cp»f^ = fp(x) + ymini

Kn (x) - 19m, (Yni f)p

Em i n i n _

ymini •••y s-2 s-1 [Rmin1,...,m_,n, (x) + -]. (2.29)

~2 s-2

Remember that in (2.29), Cp» = -C»p is the charge conjugation matrix used to raise spinor indices, $(x) and fp(x) are a D = 4 scalar and a spinor field, respectively, Fmini (x) is the Maxwell field strength, Rminim2n2 (x) is the curvature tensor of linearized gravity, Rfmini (x) is the Rarita-Schwinger field strength and other terms in the series stand for generalized Riemann curvatures of spin-s fields2 that also contain contributions of derivatives of the fields of lower spin denoted by dots, as in the case of the Rarita-Schwinger and gravity fields (see [i2] for further details).

The fact that the higher-spin fields should form an infinite-dimensional representation of the generalized N = i, D = 4 supersymmetry (2.5) is prompted by the observation that the spectrum of bosonic fields contains a single real scalar field $(x), which alone cannot have a fermionic superpartner, while each field with i > 0 has two helicities ±s. Indeed, from (2.28) we obtain an infinite entangled chain of supersymmetry transformations for the D = 4 fields

i —t

SR^mn(x) = 2d[m(Yn]Sf )^ - 2^Y^^pFm

84(x) = -ic»f,Ax), = --(V{yrnidm$ + vVmnFmn)

SFmn = -ie'H R^mn(x) d[m(Yn}f )ii,

- e Yp," yRpq,mn(x) - 2dqVp[mdn]$(x)J, (2.30)

and so on.

The algebraic reason behind the appearance of the infinite-dimensional supermultiplet of the D = 4 higher-spin fields is related to the following fact. In the n = 4, D = 4 case the superalgebra (2.5) takes the following form

[Q^Qv } = {Ym)VPm + {Ymn)^vZmn, (2.31)

where Pm is the momentum along the four-dimensional space-time and Zmn = -Znm are the tensorial charges associated with the momenta along the extra coordinates ymn. On the other hand, the conventional N = 1, D = 4 super-Poincaré algebra is

[Q^Qv } = (ym)/ÂVPm. (2.32)

Though the both algebras have the same number of the supercharges QM, their anti-commutator closes on different sets of bosonic generators. So the super-Poincaré algebra (2.32) is not a subalgebra of (2.31). Hence the representations of (2.31) do not split into (finite-dimensional) representations of the standard super-Poincaré algebra. In this sense the supersymmetric higherspin systems under consideration differ from the most of supersymmetric models of finite-dimensional super-Poincaré or AdS higher-spin supermultiplets considered in the literature (see, e.g. [29-49]).

2 The pairs of the indices separated by the commas are antisymmetrized.

It will be of interest to study which higher-spin superalgebra, associated with the enveloping algebra of osp(1\2n), underlies the super-hyperspace system under consideration. In particular, one should understand whether and how this superalgebra can be embedded into the higherspin superalgebra hu(1, 1 \2n) considered in [34] and, in the context of hyperspace constructions, in [4]. For instance, in the D = 4 case the superalgebra hu(1,1 \8) contains osp(2\8) as a finite-dimensional subalgebra [4], the latter contains the D = 4 superconformal algebra su(1, 1 \4) and, hence, the usual N = 1, D = 4 super-Poincare algebra as sub-superalgebras, thus allowing for an hu(1, 1\ 8)-invariant higher-spin system to split into the conventional finite-dimensional N = 1, D = 4 supermultiplets. As we have argued above (see also the comment in the end of Section 2.2), this is not so for the osp(1 \8)-invariant higher-spin model under consideration. In this respect let us also note that, as has been pointed out, e.g. in [57], although higher-spin superalge-bras exist in any space-time dimension D they admit usual finite-dimensional sub-superalgebras only in space-times of lower dimensions3 such as D = 3, 4, 5 and 7. In other words, higher-spin supersymmetry does not necessarily imply conventional supersymmetry.

3. Scalar superfields on OSp(1\n) group manifolds and their equations of motion

3.1. Geometric structure of the OSp(1\n) group manifolds

The geometric structure of the OSp(1\n) group manifolds in the form we shall review below and use extensively in this paper for the description of higher-spin fields in the associated AdS spaces has been discussed in [3,7,8,11,24]. The OSp(1\n) superalgebra is formed by n anti-commuting supercharges Qa and n(n2+1) generators Maß = Mßa of Sp(n)

{Qa, Qß} = 2Maß, [Qa,Mßy ] = y Ca(ß Qy),

[Maß,MyS ] = -J (Cy(aMß)S + Cs(aMß)y), (3.1)

where Caß = -Cßa is the Sp(n) invariant symplectic metric and Z is a parameter of inverse dimension of length related to the AdS radius via r = 2/Z (see also [1]). The OSp(1\n) algebra (3.1) is recognized as a subalgebra of (2.17) with the identifications

Qa = ^Qa + 4Sa^J, Maß = Paß - ^TKaß - 4L(aß), (3.2)

where Sa = SßCßa, Laß = LayCvß and Kaß = KySCyaCgß.

The OSp(1\n) manifold is parametrized by the coordinates (X^v, 0^) and its geometry is described by the Cartan forms

Q = O-1dO(X, 0) = -iQaßMaß + iEaQa, (3.3)

where O(X, 0) is an OSp(1 \n) supergroup element. The Cartan forms satisfy the Maurer-Cartan equations associated with the OSp(1\n) superalgebra (3.1)

dQaß + %Qay A Qyß = -iEa A Eß, dEa + Zey A Qya = 0, (3.4)

with the external differential acting from the right.

3 The case of D = 6 still has to be analyzed. We thank Mikhail Vasiliev for comments on this issue.

The Maurer-Cartan equations (3.4) are then solved by the following forms

Qa? = dx^G^G/X) + -(0aD0? + 0?D0a) = n»vQ^QV?(X, 0), (3.5)

Ea = P(02)D0a - 0aDP(02), (3.6)

where 0 is related to 0 through

0a = 0?G-iaP-i(02), 02 = 00 P 2(0 2) = i + -0 2, (3.7)

while the covariant derivative

V0a = d0a + ^0Pmpa(X),

contains the Cartan form of the Sp(n) group manifold

MaP(X) = dX»VG^a(X)GV^(X), (3.9)

Çap(X,0) = G/(X) - l8 (0a - 2GaY 0Y )0P, G= &ap + 4 Xj. (3.10) Note also the relations

0a g J = 0pp (02), 0a = 0P g~lap (02), (3.11) and the fact that the inverse matrix of (3.10) is given by

Q~^(X, 0) = G-^(X) - |(0sG-ai)(0sG-1f>)P-2(02)

= G-^(X) - %a0? = % + UxJ - l-00). (3.12)

8 a a 4\ a 2

The form of the bosonic Cartan form (3.5) prompts us that the latter is related to the superinvariant form (2.3) in flat hyper superspace via the GL(n) transformation with matrix element (3.i0). This property was revealed in [7] and called GL-flatness of the OSp(i\n) supermanifold. It will allow us to generalize the results of [i] and relate the scalar superfield ®(X, 0) and its field equation (2.20) in flat superspace to a scalar superfield and its equation of motion on the supergroup manifold OSp(i\n).

3.2. Scalar superfield on OSp(i \n) and its OSp(i \2n) invariant equation of motion The scalar superfield equation on OSp(i\n) takes the form [ii]

V[aV?]- jCa^$OSp(X,0) = 0, (3.i3)

where the Grassmann-odd covariant derivatives Va and their bosonic counterparts Va? satisfy the OSp(i\n) superalgebra similar to (3.i), namely

{Va, V?} = 2lVaP (3.i4)

[Vy, Va?] = 2CY(aV?), (3.i5)

[Va?, Vys ] = 2 (Ca(Y VS)? + C?(y VS)a)- (3.i6)

A somewhat tedious but straightforward algebra then shows that the superfield @OSp(X, 0) satisfying (3.13) is related to the superfield @(X, 0) satisfying the flat superspace equation (2.20) by the super-Weyl transformation

®osp(i\n)(X, 0) = (det G)-2 flt(X, 0) = (det G)-1P (02)fl (X,0), (3.17)

while the OSp(1\n) covariant derivatives are obtained from the flat superspace ones by the following GL ('generalized superconformal') transformations

Va = G-^D^,

Vap = G-MG-1v(d^v + 2iD(^ ln((det G)1P-1(02))Dv]). (3.18)

Substituting (2.21) into (3.17) and using the definition (3.7), together with the fact that on the mass shell all higher components in (2.21) vanish, we find

$OSp(n)(X, 0)

= (det G)-2 b(X) + 0a(det G)-1 (X)f,AX) + O (0 2,b(X)), (3.19)

where the first two terms are the fields

B(X) = (det G)-2 b(X), Fa(X) = (det G)-1 G-^(X)f^(X) (3.20)

propagating on the Sp(n) group manifold, and O(02, b(x)) stands for higher order terms in 02 which only depend on b(X). The fields (3.20) satisfy the equations of motion

- o (Cay VpS - Cap VyS + CpsVay - CysVap + 2Cpy Vas)B ^ 2

(Vap VyS -Vay Vps)B

ol (CayCps - CapCYs + 2CpyCas)B = 0, (3.21)

VafiFy - VayFp + i(CyaFp - CfiaFy + 2CyPFa) = 0 (3.22)

discussed in detail in [1]. Note that in (3.21) and (3.22) the covariant derivatives are restricted to the bosonic group manifold Sp(n), i.e. Vap = (X)G-v (X)dMV.

Since the flat superspace field equation is invariant under the generalized superconformal OSp(1 \2n) transformations (2.24), the above relation leads us to conclude that also the OSp(1 \n) superspace equations (3.13) are invariant under the OSp(1\2n) transformations, under which the superfield @OSp(X, 0) varies as

S^OSp = - e»Q„ + iS + ia^v + ik^ KMV + igCv^&OSp

- - v(Vv + 20^ + i^O^0OSp. (3.23)

21 ■'¡j.vt-- i 2

= -Dv = -i( d^v + 8G(aP) ), (3.24)

Qm = Qm - l^0^P(0)- (3.25)

Using the relations

Qp0a = p-1(0 Gpa + l^0p0a + ^Gpa0a0a + ( y) 20 (3.26)

(Qp0a )0a = p (0 2)( Gpa + lj-0p0^0a, (3.27)

dap0Y = 4 0(aGP)s(^l + j0s0yy (3.28) ..Y — llp (0 2)(0 -2G..P0-)C„ Y

Dp GaY = J P{0 2) (0a - 2GaP0p) GpY (3.29)

dafi 0yS = 7 GY (a G p) , (3.30)

5 ^ r> r> 5 ' = 4GY(aGP) ,

Qa G»v = -%P (0 2)0v G»a, (3.31)

l-ip(0 ^ 4

one may check that the operators (3.24) and (3.25) obey the flat hyperspace supersymmetry algebra

[V»v, Vpa]=0, {Q», Qv} = -2V»v, [V»v, Qp]=0. (3.32)

The other generators of the OSp(l\2n) are

s» = -( x»v + l-e»ev)Qv, L»v = -2i(xvp + l-evep D» -ievq», (3.33)

2 ) " » V 2

K»v = i^x»p + l-e»e^ (XvÀ + l-eve^jVpk - ie(»sv). (3.34)

Taking into account the commutation relations (3.32) we see that the operators Q», S », V»v, L»v, K»v obey the same OSp(1\2n) algebra (2.17) as the operators Q», S», P»v, L»v and K»v.

4. Correlation functions in N = 1, D = 3 superconformal models

Before considering correlation functions for superfields in hyper superspaces, it is instructive to discuss in detail analogous structures arising in the superconformal theory of a real scalar superfield in a conventional N = 1, D = 3 superspace. The reason being that this model is the simplest example (with n = 1) of the OSp(1\2n) invariant systems considered above. The physical content of this system is a real scalar and a D = 3 Majorana spinor field whereas the massless higher-spin fields are absent.

The superconformally invariant two- and three-point correlation functions of the N = 1, D = 3 model have been constructed in [26] with the use of a slightly different notation. Below we shall discuss properties of the two- and three-point functions for the D = 3 scalar superfield and its components using a formalism which straightforwardly generalizes to higher-dimensional hyperspaces.

Let us use the spinor-tensor representation for the description of the three-dimensional spacetime coordinates

хав = хва = xm(ymfe, (4.1)

where а, в = 1, 2 are D = 3 spinorial indices and m = 0, 1, 2 is the vectorial one. Since (4.1) provides a representation of the symmetric 2 x 2 matrices хав, no extra coordinates, like ymn, are present and, hence, no higher-spin fields. The inverse matrix of (4.1), х-

хавх-1 = SY, (4.2)

takes the simple form

-l 1 n 1

хав = - xm^x (yn)aP = - X2 хав- (43)

x х-m, х

We may now consider a real scalar superfield in D = 3

Ф(х, в) = ф(х) + 1ва/а(х) + вава¥(х), (4.4)

with ф(х) being a physical scalar, fa(x) a physical fermion and F(x) an auxiliary field. If (4.4) satisfies the free equation of motion (2.20), which in the D = 3 case reduces to

DaDаФ = 0, (4.5)

the auxiliary field F(x) vanishes, the scalar field ф(х) satisfies the massless Klein-Gordon equation and fa(x) satisfies the massless Dirac equation.

Let us consider a superconformal transformation of (4.4). The Poincare supersymmetry transformations read

, в) = '"(-¿а- ^¿фУ^ в) = <а 2афх в) (4.б)

and imply the supersymmetry transformations of the component fields

Sф(x) = ¡еа/а(х), (4.7)

Sfa(x) = -2i'aF(x) - 'вдавФ(х), (4.8)

SF(x) = 1 д"в fe (х), (4.9)

where we have made use of the identity

вавв = 1 Сав (вуву). (4.10)

Moreover, under conformal supersymmetry, Ф(х, в ) transforms as

SФ(x,в) = ^а(хав + ^вав^дрФ(х,в) - i(Zaea)АФ(х,в), (4.11)

where A is the conformal weight of the superfield. The superconformal transformations of the component fields are given by

Sф(x) = iZaxaefp(x), (4.12)

Sfa(x) = -2iZpxeaF(x)+ ^хвудуаф(х) + &Аф(х), (4.13)

SF(x) = 1 ZaxaedeYfY(x) - 2zJ2 - A\fa(x). (4.14)

The conformai weights of ф, fa and F are A, A + 1 and A + 1, respectively.

It should be noted that the field equation (4.5) is superconformally invariant if the superfield Ф(х, в) has the canonical conformal weight A = 1.

4.1. Two-point functions

The form of correlation functions in superconformal theories is drastically restricted by the requirement of their superconformal invariance.

The two-point correlation function of the superfield Ф(х, в ) with conformal weight A is obtained by first solving the superconformal Ward identities which involve Q- and S-supersym-metry transformations. The invariance under bosonic translations, rotations, conformal boosts and dilations then follows as a consequence of the properties of the superconformal algebra. The Q- and S-supersymmetry Ward identities are

д v д д v д V ,

— - ieï-Tï + ття - ТЦ7 [Ф(Х1 ,в1)Ф(х2,в2) = 0, (4.15)

дв' 1 дх' дв' 2 дх2

xfv + 1-в'вА(^~ - + ( xfv + 1-в'вЛ(^- -

1 2 1 V\двv 1 дх\р ) V 2 2 2 2 J \дв% 2 дх1Р

• [Ф(х1,в1)Ф(х2,в2)) + iAZf(ef + в£)(Ф(х1 ,в1)Ф(х2,в2)) = 0. The solution to these equations takes the form

(Ф(хьв1)Ф(х2,в2)) = C2 (det |Z12|)-A, (4.16) where c2 is an arbitrary normalization constant and

zJ = xfV - хfV - -в?в) - -evef, (4.17) - j 2 1 j 2 j

is invariant under Q-supersymmetry. As usual, for the two-point function to be non-vanishing, the conformal weights of the two superfields should be equal.

Expanding the expression on the right hand side of (4.16) in powers of в, we obtain

(det |Z121) A = (det |х121)-A --дав(det ^''^в^в^

- 2ду&дар(det ^п^в^в^в^в^. (4.18)

Using the identities

даР (det |х |)-A = -Ax-1 det х^, (4.19)

дарду,(det ^|)-A = ^х^х- + 1 х^х- + 1 (det |хD~A, (4.20)

one may rewrite the expression (4.18) as

(det |z12|)-A = (det х^уЧ 1 - iAХm^вaв2 - (2A -1)A-1rв2в2). (4.21)

V х12 4 х12 '

Thus, from Eqs. (4.18) or (4.21), one may immediately read off the expressions for the correlation functions of the component fields of the superfield (4.4):

(0(*i)<M*2)) = C2(det|xi2|) 2, (4.22)

{fa(xi)fp(x2)) =-ic2dafi (det |xi2|)-2, {^(xi)fa(x2)) = 0, (4.23)

{F(xi)$(x2)) = 0, [F(xi)fa(x2)) = 0, (4.24)

(F(xi)F(x2)) =-C2dapdap (det |x. (4.25)

Let us note that when the superfield ®(x, 0) has the canonical conformal dimension A = 1, due to the identity

1 1 d d 1 CaYC^9>et |x12D-1 = --nmn^ixn (det |x12Q-2 (4.26)

the last term in (4.18) is proportional to the 5-function if one moves to the Euclidean signature. Then one has for the two-point function for the auxiliary field

{F(x1)F(x2)) =-nC2S(3)(x1 - x2). (4.27)

Note that the correlation functions of the auxiliary field F with the physical fields and with itself (for xm = x2m) vanish.

On the other hand, if the conformal weight of the superfield (4.4) is anomalous, i.e. A = 2, the correlators of the auxiliary field with the physical ones still vanish (in agreement with the fact that their conformal weights are different), but the (FF) correlator is

, , (2A - 1)A 1 ( )

(F(xOF(x2)) =-C2 (-^^(det |xx21)

(2A - 1)A ( ^^ 1

= -C2(-(det ^ A ■ (4.28)

This situation may correspond to an interacting quantum N = 1 superconformal field theory [58], where the auxiliary field is non-zero, and fields acquire anomalous dimensions due to quantum corrections.

4.2. Three-point functions

We now consider three-point functions involving three real scalar superfields carrying scaling dimensions Ai (i = 1,2, 3). Solving the supeconformal Ward identities for Q- and S-supersymmetry transformations we find

{&A1 (x1,01)&A2 (x2,02)®A'3 (x3,03))

= C3 (det ^n^-1 (det |z23 |)~k2 (det |z31 |)-k3, (4.29)

k1 = - (A1 + A2 - A3), k2 = - (A2 + A3 - A1), k3 = - (A3 + A1 - A2).

2 2 2 (4.30)

Using the expansion (4.21), one obtains the three-point functions of the component fields of (xi, 01), (x2, 02) and (x3,03), whose labels of scaling dimension we skip for simplicity

(0(*i)0(*2)0M = c3(det|xi20 M(det^D 42(det\x3i\) h, (4.31)

[fa(xi)fp(X2)^(X3)]

= -iC3 ^^ (det \xi2\)-kl (det |x23\)-"2 (det |x3i \)--k3 x12

= -ic3kix'mi(Ym)ap (det \xi2\)-kl-1(det \x23\)-k2 (det \x3i\)~ks, (4.32)

fa(xi)F (x2)fp(x3))

= C32xkT (Ym)J(Yn)&p (xî2)(x23)(det \xi2\)-kl (det \x23\)~"2 (det \x3i\)-k3

2x12x23

= C3kl2k2(ym)as(yn)sp(x^feXdet\xi2\)-kl-1(det\x23\)-k2-1(det\x3i\)-k3. (4.33)

(F(xi)F(x2)t(x3)] = -c3dmdm((det\xi20 )(det\x23 0 "^(det\x3i\) (4.34)

The remaining three-point functions containing an odd number of fermions, as well as the correlator (F$$), vanish. Note that, dimensional arguments would allow for a non-zero (F$$) correlator, but supersymmetry forces it to vanish. The correlator (F(x\)F(x2)F(x3)) is zero as well, since it is proportional to (KmynKi)xm!xn3x|1 = 2iemnpxm!xn3x|1 = 0.

Moreover, from the above expressions we see that superconformal symmetry does not fix the values of the scaling dimensions Ai (4.30) entering the right hand side of (4.29). This indicates that quantum operators may acquire anomalous dimensions and the quantum N = 1, D = 3 superconformal theory of scalar superfields can be non-trivial, in agreement, e.g. with the results of [58].

If the value of A were restricted by superconformal symmetry to its canonical value and no anomalous dimensions were allowed (for all the operators which are not protected by super-symmetry) one would conclude that the conformal fixed point is that of the free theory. This is the case, for instance, for the N = 1, D = 4 Wess-Zumino model in which the chirality of N = 1 matter multiplets and their three-point functions restricts the scaling dimensions of the chiral scalar supermultiplets to be canonical. This implies that in the conformal fixed point the coupling constant is zero, i.e. the theory is free [59,60].

5. Correlation functions in OSp(1\2n)-invariant models

Following the example of the N = 1, D = 3 superconformally invariant model of the previous section, we now proceed to compute correlation functions on hyper superspace for generic OSp(1\2n) invariant models. Again, it is sufficient to require the invariance of the correlation functions under Q- and S-supersymmetry transformations. The invariance under the generalized translations, rotations and conformal transformations will then be guaranteed by the form of the OSp(1\2n) superalgebra. As we will see, the form of the super-correlators will be exactly the same as in the D = 3 case with only difference that the superinvariant intervals (4.17) are now n x n matrices.

5.1. Two-point functions

Let us denote the two-point correlation function by

W(Z!,Z2) = (0(X1,01)0(X2,02)j. (5.1)

The invariance under Q-supersymmetry requires

d v d d v d \ —Ü - 0-öU + —Ü - i0Vv-äv )W(Z1,Z2) = 0, (5.2)

d0f 1 dxfv d0f 2 dxfv J 2

which implies

{<P(X1,01)$(X2,02)) = W (det \Zx2\), (5.3)

Zf2v = Xfv - Xfv - l-0f0l - l-0v0ff (5.4)

is the interval between two points in hyper-superspace which is invariant under the rigid super-symmetry transformations (2.2).

We next impose invariance of the correlator under the S-supersymmetry transformation

Xfv + -0f0Mi^- - i0p+ (Xfv + ^Ail- - i0p d

1 1 o 1 1 /1 a/iv 1 avvP I 1 \ 2 1 o 2 2 l\ aQv

1 01 M d0v 1 dX\p) V2 202 2) Vd02v 2 dx2vp

x W(det\Zx2\) + i^2of + lo^W(det\Zx2\) = 0, (5.5)

which is solved by

_ 1 _ 1

W(det\Zx2\) = c2(detZd) 2 ^ {^(X1,01)^(X2,02)) = c^det\Zn\) 2. (5.6)

The two-point function (5.6) reproduces the correlators of the component bosonic and fermionic hyperfields b(X) and fM(X) after the expansion of the former in powers of the Grassmann coordinates 0^0?. Since on the mass shell the superfield (2.21) has only two non-zero components, all terms in the 0 -expansion of the two-point function (5.6), starting from the ones quadratic in 01(M02v should vanish. This is indeed the case, as a consequence of the field equations.

To see this, let us recall that in the separated points the two-point function of the bosonic hyperfield of weight \ satisfies the free field equation. Therefore for Xlp = X^ one has4

df vdpa - d^pPdl)[b(XDh(X2)) = df vd^ - dfpdl)(det\x 1 2O"2 = 0. (5.7)

Similarly, for Xaß = Xa2ß the fermionic two-point function satisfies the free field equation for the fermionic hyperfield. Written in terms of the superfields, these equations are encoded in the superfield equation

(DfD 1 - DlDf){0(X 1,0 1)0(X2,02))

= (DfD 1 - D¿Df)(det \Z 12\)-2 = 0 (for Z 12 = 0). (5.8)

4 When the two points coincide, one can define an analog of the Dirac delta function in the tensorial spaces, see [5] for the relevant discussion.

Expanding the two-point function (det |Zi2|) 2 in powers of the Grassmann theta-variables

(det IZ12I) 1 = (det IX12I) 1 - idaß(det IX12I) 22e(aeß

-1 dySBaß (det IX12I)- 2 e^e^e^e^ + ■■■, (5.9)

one may see that terms in the expansion starting from (O^O^)2 vanish due to the free field equation (5.7). From Eqs. (5.6), (5.9) and from the explicit form of the superfield (2.21), one may immediately reproduce the correlation functions for the component fields [10]

(b(X1)b(X2)) = c2(det |X12i) 2,

[f»(Xl)fv(X2)) = (X12)-1(det IX12I)"1 • (5.10)

Notice also that, contrary to the non-supersymmetric case, where the two-point functions for bosonic and fermionic hyperfields contain an independent normalization constant each, in the supersymmetric case the number of independent constants is reduced to one.

The two-point functions on the OSp(1In) manifold may now be obtained from (5.6) via the rescaling (3.17), which relates the superfields in flat superspace and on the OSp(1In) group manifold

{<P0Sp(X1,S1)$0Sp(X2,02))

= (det G(Xx))" 2 P (0f)(det G(X2))~ 2 P (&l)[<P(Xl,0l)$(X2,02)). (5.11)

Finally, as in the D = 3 case, one may derive the superconformally invariant two-point function for superfields carrying an arbitrary generalized conformal weight A, which on flat hyper superspace has the form

1 (Xu0x )®A2 (X2,02)) = C2(det |Z12i)"A, A1 = A2 = A. (5.12)

In principle, in order to obtain the OSp(1In) correlator, as in the case A = 5, one may apply to

(5.12) a Weyl rescaling similar to (5.11). However, when A = 2 the superfields no longer satisfy

the quadratic equations (2.20) and (3.13), because the latter equations are superconformally in-1 -1 2 variant only for A = j. Thus, fixing the power of (det G(X)) 2 P(02) in the Weyl transform of

quantities carrying anomalous dimensions remains an interesting open problem.

5.2. Three-point functions

The three-point functions for the superfields with arbitrary generalized conformal dimensions Ai (i = 1, 2, 3)

W(Z!,Z2,Z3) = (@(X\,0O$(X2,02)$(X3,03)), (5.13)

may be computed in a way similar to the two-point functions using the superconformal Ward identities. The invariance under g-supersymmetry implies that they depend on the superinvariant intervals Zij, i.e.

[^(Xl,0l)$(X2,02)$(X3,0)) = W(Z12, Z23,Z31), (5.14)

Z'.v = X/v - X/v - -{0/0vj + 0V0/), ij = 1 , 2, 3. (5. 1 5)

J J 2 J J

Invariance under S-supersymmetry then fixes the form of the function W to be

{&(X 1,0 1)$(X2,02)$(X3,03)]

= c3 (det Z 12)-2(A 1 +A2-A3)(det Z23)-2 (A2+A3-A1)(det Z31)-2(A3+A1-A2>. (5. 1 6)

Let us note that the three-point function is not annihilated by the operator entering the free equations of motion (2.20) for generic values of the generalized conformal dimensions, including the case in which the values of all the generalized conformal dimensions are canonical

(DlDl - D1vDl){^(X1,01), &(X2,02), &(X2,02))

= c3(DlDl - d;d£)((det|Zx2|) 4(det IZ23O 4 (detIZ31I) ?) = 0.

_ 1 1 1

// V V /}

The component analysis of the superfield three-point correlation function (5.16) proceeds in the same way as in the N = 1, D = 3 case of Section 4.2. The difference lies, however, in the presence of many more auxiliary fields.

Again, the three-point functions on the supergroup manifold OSp(1In) can be obtained via the Weyl rescaling (3.17), as in the case of the two-point functions, Eq. (5.11).

5.3. Four-point functions

Finally, let us consider, first in flat hyper superspace, the correlation function of four real scalar superfields with arbitrary generalized conformal dimensions, Ai (with i = 1,2, 3, 4)

W(Zx,Z2,Z3) = [<P(X1,01)$(X2,02)$(X3,03 )$(X4,04)). (5.17)

Invariance under g-supersymmetry again implies that the correlation function depends only on

nt intervals Z

theory we find

the superinvariant intervals Z// (5.15). Following the analogy with conventional conformal field

W(X1 ,X2,X3,X4) = C4 ft (det ,Z.. {)kj W M), (5.18)

-J,i<J ( 1 -J I)

with W being an arbitrary function of the cross-ratios

z = det( ), z' = det( \ (5.19) VIZ1311Z241), V1Z23I1Z141/,

JZ23IIZ14I/, subject to the crossing symmetry constraints

: ?t )=-(?•?

W = *( 11) = z ,±). (5'20)

Furthermore, the kij's are constrained by invariance of the four-point function under the S-supersymmetry to satisfy

J^k-J = A-. (5.21)

Similarly to the case of two- and three-point functions, the four-point function of the scalar superfields on OSp(1In) can be obtained from (5.18) via the Weyl rescaling (3.17).

6. Conclusion and outlook

A detailed study of the OSp(1I2n)-invariant generalized superconformal theories is still an interesting open problem, which is important for better understanding the properties of con-formally invariant higher-spin field theories (see, e.g. [61-69] for recent progress in studying conformal higher-spin fields). Our results are a further step in this direction. Following the program outlined in [1], we have extended the results on the structure of Sp(2n)-invariant field equations to supersymmetric higher-spin systems. We constructed generalized superconformal transformations relating the field equation on flat hyper-superspace and on OSp(1In) supergroup manifolds, which correspond to a generalization of supersymmetric AdS spaces. We computed the two-, three- and four-point functions of real hyper-superfields both on flat and on OSp(1In) supergroup manifolds and, as a simple illustration of our approach, applied this technique to the example of N = 1, D = 3 superconformal theory of scalar superfields.

It is important to further study possible interactions (which might be associated with nontrivial three- and four-point correlation functions) in this type of models. Since a Lagrangian description of OSp(1\2n) invariant field equations is still not known even in the free case, one can approach the problem using non-Lagrangian methods similar to those in Conformal Field Theories (see for example [70]). Following these methods one can try to introduce OSp(1 \2n) invariant vertexes and compute explicit expressions for anomalous dimensions for generalized conformal weights. Recall that according to the results of Section 5 the Ward identities for three- and four-point functions do not necessarily require the values of the generalized conformal weights to be canonical, therefore one may expect interesting outcomes of this study.

The question of the existence of anomalous values for generalized conformal dimensions can be related to the question of a possible breaking of OSp(1 \2n) symmetry down to a corresponding AdSD (super)symmetry. In this respect one can also note that the hyperspace formulation considered in this paper does not involve higher-spin gauge field potentials, but only their field strengths. So far higher-spin potentials have been introduced only in an unfolded extension of the hyperspace formulation of D = 4 higher-spin fields in such a way that the resulting equations are invariant under SU(2,2) and O(3, 3) subgroups of the original Sp(8) symmetry, motivating to speculate on their origin due to a mechanism of spontaneous breaking of higher-spin and Sp(8) symmetries [13]. Further study in this direction may help in searching for interacting systems of fields on hyper-(super)spaces and their possible connection to Vasiliev's interacting higher-spin gauge theories.

It would be also of interest to consider in detail the implication of our results in the framework of higher-spin AdS/CFT correspondence. The origin of higher-spin holographic duality can be traced back [4] to the work of Flato and Fronsdal [71] who showed that the tensor product of single-particle states of a 3D massless conformal scalar and spinor fields (singletons) produces the tower of all single-particle representations of 4D massless fields whose spectrum matches that of 4D higher-spin gauge theories. The hyperspace formulation provides an explicit field theoretical realization of the Flato-Fronsdal theorem in which higher-spin fields are also "packed" in a single scalar and spinor fields, though propagating in hyperspace. The relevance of the hyperspace formulation to holography has been pointed out in [4,72]. In this interpretation, holographically dual theories share the same unfolded formulation in extended spaces which contains twistor-like (or oscillator) variables and each of these theories corresponds to a different reduction, or "visualization", of the same "master" theory. For instance, the higher-spin field equations in either ordinary space-time or hyperspace can be obtained from the same set of unfolded equations [4,6-8]. Depending on the number of twistorial coordinates of the unfolded

formulation, one can obtain hyperfields of different ranks which can be fundamental fields, bifundamental fields (currents) etc. [9]. A connection between these fields in different dimensions can be established via embedding of lower-dimensional hyperspaces into higher-dimensional ones [19]. Thus, one can conclude that the hyperspace formulation provides an extra and potentially powerful tool for studying higher-spin AdS/CFT correspondence.

A detailed study of the higher-spin content of field equations on higher-dimensional curved hyper-superspaces, as well as their underlying higher-spin superalgebras containing OSp(1In), is yet another interesting issue. We hope to address these problems in future work.

Acknowledgements

We are grateful to I. Bandos, N. Berkovits, S. Kuzenko, I. Samsonov, M. Vasiliev and P. West for fruitful discussions. The work of D.S. was partially supported by the Padova University Project CPDA119349, the INFN Special Initiative ST&FI and by the Russian Science Foundation grant 14-42-00047 in association with Lebedev Physical Institute. D.S. would also like to acknowledge the warm hospitality extended to him at the Faculty of Education, Science, Technology and Mathematics, University of Canberra, during an intermediate stage of this work. M.T. would like to thank the Department of Physics, the University of Auckland, where part of this work has been performed, for its kind hospitality. The work of M.T. has been supported in part by an Australian Research Council grant DP120101340. M.T. would also like to acknowledge grant 31/89 of the Shota Rustaveli National Science Foundation.

References

[1] I. Florakis, D. Sorokin, M. Tsulaia, Higher spins in hyperspace, J. High Energy Phys. 1407 (2014) 105, arXiv: 1401.1645 [hep-th].

[2] C. Fronsdal, Massless particles, orthosymplectic symmetry and another type of Kaluza-Klein theory, UCLA-85-TEP-10, 1985. Published in: C. Fronsdal (Ed.), Essays on Supersymmetry, in: Mathematical Physics Studies, Reidel, Dordrecht, Netherlands, 1986, pp. 163-265.

[3] I.A. Bandos, J. Lukierski, D.P. Sorokin, Superparticle models with tensorial central charges, Phys. Rev. D 61 (2000) 045002, arXiv:hep-th/9904109.

[4] M.A. Vasiliev, Conformal higher spin symmetries of 4D massless supermultiplets and osp(L, 2M) invariant equations in generalized (super)space, Phys. Rev. D 66 (2002) 066006, arXiv:hep-th/0106149.

[5] M. Vasiliev, Relativity, causality, locality, quantization and duality in the S(p)(2M) invariant generalized spacetime, arXiv:hep-th/0111119.

[6] V.E. Didenko, M.A. Vasiliev, Free field dynamics in the generalized AdS (super)space, J. Math. Phys. 45 (2004) 197-215, arXiv:hep-th/0301054.

[7] M. Plyushchay, D. Sorokin, M. Tsulaia, Higher spins from tensorial charges and OSp(N \2n) symmetry, J. High Energy Phys. 0304 (2003) 013, arXiv:hep-th/0301067.

[8] M. Plyushchay, D. Sorokin, M. Tsulaia, GL flatness of OSp(1I2n) and higher spin field theory from dynamics in tensorial spaces, arXiv:hep-th/0310297.

[9] O.A. Gelfond, M.A. Vasiliev, Higher rank conformal fields in the Sp(2M) symmetric generalized space-time, Theor. Math. Phys. 145 (2005) 1400-1424, arXiv:hep-th/0304020.

[10] M. Vasiliev, V. Zaikin, On Sp(2M) invariant Green functions, Phys. Lett. B 587 (2004) 225-229, arXiv:hep-th/ 0312244.

[11] I. Bandos, P. Pasti, D. Sorokin, M. Tonin, Superfield theories in tensorial superspaces and the dynamics of higher spin fields, J. High Energy Phys. 0411 (2004) 023, arXiv:hep-th/0407180.

[12] I. Bandos, X. Bekaert, J.A. de Azcarraga, D. Sorokin, M. Tsulaia, Dynamics of higher spin fields and tensorial space, J. High Energy Phys. 0505 (2005) 031, arXiv:hep-th/0501113.

[13] M. Vasiliev, On conformal, SL(4, R) and Sp(8, R) symmetries of 4d massless fields, Nucl. Phys. B 793 (2008) 469-526, arXiv:0707.1085 [hep-th].

[14] E. Ivanov, Nonlinear realizations in tensorial superspaces and higher spins, arXiv:hep-th/0703056.

[15] P.C. West, E(11) and higher spin theories, Phys. Lett. B 650 (2007) 197-202, arXiv:hep-th/0701026.

[16] O.A. Gelfond, M.A. Vasiliev, Higher spin fields in Siegel space, currents and theta functions, J. High Energy Phys. 0903 (2009) 125, arXiv:0801.2191 [hep-th].

[17] O. Gelfond, M. Vasiliev, Sp(8) invariant higher spin theory, twistors and geometric BRST formulation of unfolded field equations, J. High Energy Phys. 0912 (2009) 021, arXiv:0901.2176 [hep-th].

[18] O. Gelfond, M. Vasiliev, Unfolding versus BRST and currents in Sp(2M) invariant higher-spin theory, arXiv: 1001.2585 [hep-th].

[19] O. Gelfond, M. Vasiliev, Unfolded equations for current interactions of 4d massless fields as a free system in mixed dimensions, arXiv:1012.3143 [hep-th].

[20] I.A. Bandos, J.A. de Azcarraga, C. Meliveo, Extended supersymmetry in massless conformal higher spin theory, Nucl. Phys. B 853 (2011) 760-776, arXiv:1106.5199 [hep-th].

[21] S. Fedoruk, J. Lukierski, New spinorial particle model in tensorial space-time and interacting higher spin fields, J. High Energy Phys. 1302 (2013) 128, arXiv:1210.1506 [hep-th].

[22] O. Gelfond, M. Vasiliev, Higher-rank fields and currents, arXiv:1312.6673 [hep-th].

[23] V. Didenko, E. Skvortsov, Exact higher-spin symmetry in CFT: all correlators in unbroken Vasiliev theory, arXiv: 1210.7963 [hep-th].

[24] I.A. Bandos, J. Lukierski, C. Preitschopf, D.P. Sorokin, OSp supergroup manifolds, superparticles and supertwistors, Phys. Rev. D 61 (2000) 065009, arXiv:hep-th/9907113.

[25] J.-H. Park, N = 1 superconformal symmetry in four-dimensions, Int. J. Mod. Phys. A 13 (1998) 1743-1772, arXiv: hep-th/9703191.

[26] J.-H. Park, Superconformal symmetry in three-dimensions, J. Math. Phys. 41 (2000) 7129-7161, arXiv:hep-th/ 9910199.

[27] S.M. Kuzenko, S. Theisen, Correlation functions of conserved currents in N = 2 superconformal theory, Class. Quantum Gravity 17 (2000) 665-696, arXiv:hep-th/9907107.

[28] F. Dolan, H. Osborn, Conformal four point functions and the operator product expansion, Nucl. Phys. B 599 (2001) 459-496, arXiv:hep-th/0011040.

[29] T. Curtright, Massless field supermultiplets with arbitrary spin, Phys. Lett. B 85 (1979) 219.

[30] M.A. Vasiliev, 'Gauge' form of description of massless fields with arbitrary spin, Yad. Fiz. 32 (1980) 855-861 (in Russian).

[31] M.P. Bellon, S. Ouvry, D = 4 supersymmetry for gauge fields of any spin, Phys. Lett. B 187 (1987) 93.

[32] E.S. Fradkin, M.A. Vasiliev, Superalgebra of higher spins and auxiliary fields, Int. J. Mod. Phys. A 3 (1988) 2983.

[33] E. Bergshoeff, A. Salam, E. Sezgin, Y. Tanii, Singletons, higher spin massless states and the supermembrane, Phys. Lett. B 205 (1988) 237.

[34] S.E. Konstein, M.A. Vasiliev, Extended higher spin superalgebras and their massless representations, Nucl. Phys. B 331 (1990) 475-499.

[35] S. Kuzenko, A. Sibiryakov, V. Postnikov, Massless gauge superfields of higher half integer superspins, JETP Lett. 57 (1993) 534-538.

[36] S. Kuzenko, A. Sibiryakov, Massless gauge superfields of higher integer superspins, JETP Lett. 57 (1993) 539-542.

[37] S. Kuzenko, A. Sibiryakov, Free massless higher superspin superfields on the anti-de Sitter superspace, Phys. At. Nucl. 57 (1994) 1257-1267, arXiv:1112.4612 [hep-th].

[38] I. Buchbinder, S. Kuzenko, A. Sibiryakov, Quantization of higher spin superfields in the anti-de Sitter superspace, Phys. Lett. B 352 (1995) 29-36, arXiv:hep-th/9502148.

[39] J. Gates, S. James, S.M. Kuzenko, A.G. Sibiryakov, N = 2 supersymmetry of higher superspin massless theories, Phys. Lett. B 412 (1997) 59-68, arXiv:hep-th/9609141.

[40] J. Gates, S. James, S.M. Kuzenko, A.G. Sibiryakov, Towards a unified theory of massless superfields of all superspins, Phys. Lett. B 394 (1997) 343-353, arXiv:hep-th/9611193.

[41] E. Sezgin, P. Sundell, Higher spin N = 8 supergravity, J. High Energy Phys. 9811 (1998) 016, arXiv:hep-th/ 9805125.

[42] K.B. Alkalaev, M.A. Vasiliev, N = 1 supersymmetric theory of higher spin gauge fields in AdS(5) at the cubic level, Nucl. Phys. B 655 (2003) 57-92, arXiv:hep-th/0206068.

[43] J. Engquist, E. Sezgin, P. Sundell, Superspace formulation of 4-D higher spin gauge theory, Nucl. Phys. B 664 (2003) 439-456, arXiv:hep-th/0211113.

[44] E. Sezgin, P. Sundell, Supersymmetric higher spin theories, J. Phys. A 46 (2013) 214022, arXiv:1208.6019 [hep-th].

[45] Y. Zinoviev, Massive N = 1 supermultiplets with arbitrary superspins, Nucl. Phys. B 785 (2007) 98-114, arXiv: 0704.1535 [hep-th].

[46] A. Fotopoulos, M. Tsulaia, Gauge invariant Lagrangians for free and interacting higher spin fields. A review of the BRST formulation, Int. J. Mod. Phys. A 24 (2009) 1-60, arXiv:0805.1346 [hep-th].

[47] J. Gates, S. James, K. Koutrolikos, On 4D, N = 1 massless gauge superfields of arbitrary superhelicity, J. High Energy Phys. 1406 (2014) 098, arXiv:1310.7385 [hep-th].

[48] S.J.J. Gates, K. Koutrolikos, On 4D, N = 1 massless gauge superfields of higher superspin: half-odd-integer case, arXiv:1310.7386 [hep-th].

[49] C. Candu, C. Peng, C. Vollenweider, Extended supersymmetry in AdS3 higher spin theories, arXiv:1408.5144 [hep-th].

[50] D. Francia, A. Sagnotti, On the geometry of higher-spin gauge fields, Class. Quantum Gravity 20 (2003) S473-S486, arXiv:hep-th/0212185.

[51] A. Campoleoni, D. Francia, Maxwell-like Lagrangians for higher spins, J. High Energy Phys. 1303 (2013) 168, arXiv:1206.5877 [hep-th].

[52] I.A. Bandos, E. Ivanov, J. Lukierski, D. Sorokin, On the superconformal flatness of AdS superspaces, J. High Energy Phys. 0206 (2002) 040, arXiv:hep-th/0205104.

[53] S.M. Kuzenko, G. Tartaglino-Mazzucchelli, Three-dimensional N = 2 (AdS) supergravity and associated supercur-rents, J. High Energy Phys. 1112 (2011) 052, arXiv:1109.0496 [hep-th].

[54] S.M. Kuzenko, U. Lindstrom, G. Tartaglino-Mazzucchelli, Three-dimensional (p, q) AdS superspaces and matter couplings, J. High Energy Phys. 1208 (2012) 024, arXiv:1205.4622 [hep-th].

[55] I. Samsonov, D. Sorokin, Superfield theories on S3 and their localization, J. High Energy Phys. 1404 (2014) 102, arXiv:1401.7952 [hep-th].

[56] S.M. Kuzenko, D. Sorokin, Superconformal structures on the three-sphere, J. High Energy Phys. 1410 (2014), arXiv:1406.7090 [hep-th].

[57] M. Vasiliev, Higher spin superalgebras in any dimension and their representations, J. High Energy Phys. 0412 (2004) 046, arXiv:hep-th/0404124.

[58] F. Synatschke, J. Braun, A. Wipf, N = 1 Wess-Zumino model in d = 3 at zero and finite temperature, Phys. Rev. D 81 (2010) 125001, arXiv:1001.2399 [hep-th].

[59] S. Ferrara, J. Iliopoulos, B. Zumino, Supergauge invariance and the Gell-Mann-Low eigenvalue, Nucl. Phys. B 77 (1974) 413.

[60] B. Conlong, P.C. West, Anomalous dimensions of fields in a supersymmetric quantum field theory at a renormal-ization group fixed point, J. Phys. A 26 (1993) 3325-3332.

[61] A. Gover, A. Shaukat, A. Waldron, Tractors, mass and Weyl invariance, Nucl. Phys. B 812 (2009) 424-455, arXiv: 0810.2867 [hep-th].

[62] R. Metsaev, Gauge invariant two-point vertices of shadow fields, AdS/CFT, and conformal fields, Phys. Rev. D 81 (2010) 106002, arXiv:0907.4678 [hep-th].

[63] R. Metsaev, CFT adapted approach to massless fermionic fields, AdS/CFT, and fermionic conformal fields, arXiv:1311.7350 [hep-th].

[64] R. Metsaev, Arbitrary spin conformal fields in (A)dS, arXiv:1404.3712 [hep-th].

[65] M.S. Costa, J. Penedones, D. Poland, S. Rychkov, Spinning conformal correlators, J. High Energy Phys. 1111 (2011) 071, arXiv:1107.3554 [hep-th].

[66] J. Maldacena, A. Zhiboedov, Constraining conformal field theories with a higher spin symmetry, arXiv:1112.1016 [hep-th].

[67] Y.S. Stanev, Correlation functions of conserved currents in four dimensional conformal field theory, arXiv:1206. 5639 [hep-th].

[68] S. Giombi, I.R. Klebanov, S.S. Pufu, B.R. Safdi, G. Tarnopolsky, AdS description of induced higher-spin gauge theory, J. High Energy Phys. 1310 (2013) 016, arXiv:1306.5242 [hep-th].

[69] T. Nutma, M. Taronna, On conformal higher spin wave operators, J. High Energy Phys. 1406 (2014) 066, arXiv: 1404.7452 [hep-th].

[70] A. Petkou, Conserved currents, consistency relations and operator product expansions in the conformally invariant O(N) vector model, Ann. Phys. 249 (1996) 180-221, arXiv:hep-th/9410093.

[71] M. Flato, C. Fronsdal, One massless particle equals two Dirac singletons: elementary particles in a curved space. 6, Lett. Math. Phys. 2 (1978) 421-426.

[72] M.A. Vasiliev, Holography, unfolding and higher-spin theory, J. Phys. A 46 (2013) 214013, arXiv:1203.5554 [hep-th].