Volume 105B, number 2,3 PHYSICS LETTERS 1 October 1981
THE EQUIVALENCE OF Sp(2N) AND SO(-2N) GAUGE THEORIES
R.L. MKRTCHYAN
Landau Institute for Theoretical Physics, Academy of Sciences, Moscow, USSR
Received 5 February 1981
It is proved that aU gauge invariant quantities in the Sp (2N) quantum gauge theory transform into those of SO (2N) after
changing N--* -N, at fixed h = g2N. Matter multiplets transform into multiptets with transposed Young diagrams with nf
nf(-1) d, where nf is the number of flavors, and d the number of squares in the Young diagram.
In this letter we prove a simple relation between
gauge theories, based on symplectic and orthogonal
groups, in gauge invariant sectors. The following ex-
act statement is claimed.
Consider a gauge theory with the gauge group
Sp(2N) and matter in multiplets with the Young dia-
grams Mi, i = 1,2 ..... and with number of flavors n i.
Then all gauge invariant expectation values in this the-
ory coincide with the corresponding quantities in the
SO(2N) gauge theory with the matter in transposed
multiplets M/T , and with the number of flavors
n i (- 1 ) di (d i is the number of squares in the Young
diagram Mi), after changing N~ -N at fixed ~ = g2N,
where g2 is the coupling constant, common for both
theories.
Notes:
(i) As is well-known from the N expansion [1 ] the
relevant coupling constant in nonabellian gauge theories
is ~ = g2 N, its positiveness corresponds, for example,
to the property of asymptotic freedom (see below).
(ii) One may avoid the reversing of the sign of n i in
the case of odd d i by introducing a subsidiary index
(i.e. multiply all flavors), the trace of which is equal to
1 or -1.
(iii) We shall see later that actually our statement
holds for quantities invariant under global gauge trans-
formations only.
(iv) Naturally, there are ambiguities in changing the
sign of N in quantities which are defined only for in-
teger positive N. But in perturbation theory we give a
174
simple and natural definition: the group weight of each
diagram may be represented by a polynomial of N(see
below) and in this form they make sense for allN.
(v) All representations of the group Sp(2N) are la-
beled by Young diagrams, but SO(2N) has (spinor)
representations, which do not correspond to any
Young diagrams. So, more exactly one may speak
about the embedding of the Sp(2N) theory in the
SO (2N) one.
(vi) The group weight of each diagram may be ex-
pressed in terms of eigenvalues of the invariant
(Casimir) operators on irreducible representations,
hence this equivalence has to correspond to some rela-
tions between the eigenvalues of the Casimir operators
for the groups Sp(2N) and SO(2N). Actually, we
proved that these eigenvalues for Sp(2N) transform
to the eigenvalues for SO(2N) on a transposed repre-
sentation under the change ofN ~ -N, but we do not
continue this line here.
(vii) Some relations between the groups Sp (2N)
and SO(-2N) were known previously [2] in a purely
group-theoretical framework, without any reference
to gauge theories.
In what follows we check the equivalence on some
examples from perturbation theory, and describe gen-
eral diagrammatic proofs. Then we carry out the most
simple approach to the phenomena considered - the
reformulation of the gauge invariant sector of the
gauge theory in loop space, developed recently in the
works of Makeenko and Migdal [3]. In the end of this
letter we briefly discuss the case of matter in the defin-
ing and adjoint representations, leaving aside the general
case.
0 031-9163/81/0000-0000/$ 02.75 © 1981 North-Holland