Professional Documents
Culture Documents
1, 150--158 (1965)
Introduction
I n recent time much attention has been paid to the problem of
extending relativistieally the S U (6) supermultiplet theory which has been
proposed independently b y m a n y authors [1--3]. Such an extension
should have the property t h a t the resulting invariance group contains
the Poincar6 group ~ as a subgroup and S U (6) as a "little group". The
latter assumption is motivated b y the fact t h a t S U(6) is supposed to
describe the spin as well as the S U(3) internal degrees of freedom of the
elementary system.
One such relativistic extension** is tile "inhomogeneous SL(6, C)"
group [hereafter denoted I SL (6, C)]. I t is built in complete analogy with
ISL(2, C) which is isomorphic to the covering group of the Poincar6
group ~ . The group ISL(2, C) contains as a subgroup SL(2, C) -- the
group of 2 × 2 complex matrices with determinant 1 - - which is iso-
morphic to the covering group of the proper Lorentz group. I t contains
also an invariant Abelian subgroup, the translation group, which is the
additive group of 2 × 2 tIermitian matrices. Correspondingly the
ISL(6, C) group contains as a subgroup SL(6, C), the group of 6 × 6
complex matrices with determinant 1, which constitutes the "homo-
geneous p a r t " of 1SL(6, C). I t also contains as an invariant Abelian
subgroup the 36 dimensional additive group of 6 × 6 Hermitian matrices.
* On leave from Vniversitg de MarseiIle, Institut de Physique Th~orique.
** The ISL(6, C) group has been proposed independently by B. SAKITX[1],
L. l-~mm~L(private communication), T. FtrLTO-~and J. W]~ss [4], and H. BAcnY [5].
On a Class of Generalized Poincar6 Groups: Inhomogeneous SL(n, C) 151
where ~b and fib are real coefficients and E~ is the n × n matrix all
elements of which are zero except one; this element equals 1 and belongs
* See, for instance, [6]. This paper al~o contains references to other papers.
152 H. BAe~Yand A. KntLB]~G:
to the a th row and the bth column. The matrices E~ and iE~ generate
GL(n, C). The SL(n, C) group is generated by
E~ = E~ - - %-
For S V (n), D (p, q) reduces to (p} ® {q}, the usual direct product of the
irreducible representations {p} and (q} of S U (n).
In the special case of SL(2, C) the numbers p and q are usually
replaced by the spin eigenvalues ] and )" with p = 2] ÷ 1 and q = 2]' ÷ 1.
The vector space V2 is the well-known two-dimensional Weyt spinor space
corresponding to the representation D 1 ° ~- D (2, 1). By taking the direct
sum D 1 0 ~ D 1 we get the four-dimensional Dirac spinor space.
0y
I t ~411 appear useful to generalize the Dirae representation to the
SL (n, C) case. It is built from the sum of the two representations D (n, 1)
and D(1, n). From Eq. (6) and definition (8) the generators of SL(n, C)
can be written in such a reducible representation in the form of 2n × 2n
matrices
(1t)
P~
L~
where L~ are for instance those of Eq. (4). The P~'s generate obviously
an invariant Abelian subgroup and they belong to the (~} ® {n} represen-
tation of S U (n) and to the D (~, n) representation of SL (n, C). It can
also readily be verified that the matrices Mi, Ni and P~ generate
I SL (n, C). Obviously we could have chosen the representation D (n, ~)
instead of D (~, n) for the translation operators P~, a choice which is
actually suggested by Eq. (1).
The 2n-dimensional representation t h a t we have just defined suggests
the following choice of generators for I GL (n, C). The generators of the
* This was shown for n = 2 and n = 6 in [9].
154 H. BACRr and A. KIHLB:ERG:
homogeneous group are E~, defined as before, and E~ where the dots mean
t h a t we have to find the only non-vanishing d e m e n t 1 in the bth row and
a th column of the right-lower n × n matrix. Those of the translation group
can be denoted Pai, whose non-vanishing element belongs to the a t~
column and bth row of the left-lower n × n matrix. We get the following
commutation rules for I G L (n, C)
[E ab, E~] ~ E ~ - - (5~E~ (13a)
[ g < =0
[E~, Pc~] = - - ~, Pb3 (13d)
[Pa~, P ~ ] = 0 . (13f)
If we are interested in the subgroup, I S L ( n , C), we have to suppose
that ~" E a and ~ E~ are zero, but the commutation relations (13) are
still valid.
I n S L (n, C) there exists an invariant antisymmetric form of n th order
which is the well-known Levi-Civita-Ricci tensor e a ~... f. Moreover, given
a set of n vectors belonging to the D (~, n) representation (X1)ar~, (Xz)b;,...
(Xn)t; we can define the following symmetric form
Mal= - -
i (Ei < ~+ "1+ "
2
M~,a_ 1 {~1 2 i ½
-- ~- ~*~1 - - E 2 + E i - - E:z)
• / 1 2 i
M 01__
- ~ ~ ~E2~- E l - E ; 2 - E ~ )
11 2 iE~) (18)
M ° ~= ~ (E 2 - - E 1 + E{
Mo.a_
-- ~i {~'q
~,l -- E~--
" E ii + E~)
W ° = W l i + W22
W~ = W12+ W2i
(21)
w 2 = i ( w ~ 2 - w2i)
W 8 = W1 i - - W2.2.
W h a t invariants can be built using the two vectors Pab and idab ? Since
they must be invariants with respect to SL(2, C) they must be scalars
for this group. The only scalar one can form from two vectors Aab and
Be~ is A (Aab, Bc~), according to Section I. One easily verifies t h a t
A (W~, W ~ . . . . . WgD
involving, respectively, n P's, (n - - 1) P ' s and one W, . . . and so on up to
n W's. Due to the antisymmctry of the invariant tensor ~a b . . . f and to the
constraints ,~ E a = O, ~ E h = 0 one finds that the second invariant in-
tt
References
[1] SAKITA,B.: Phys. Rev. 136, B 1756 (1964).
[2] Gi~sn~, F., and L. A. R2~DICATI:Phys. Rev. Letters 13, 173 (1964).
[3] B~cRY, H., and J. Nv¥~s: Phys. Letters 12, 156 and 13, 359 (1964).
[4] FvLTo~, T., and J. W~ss: Phys. Letters 14, 57 (1965).
[5] BAcR~~, H.: A possible enlarged Poincar~ group (unpublished).
[6] B ~ , hr. A. B., and A. P ~ s : Lorentz inva.rianee and the interpretation of S U {6)
theory (to be published in Phys. l~ev.).
[7] R~)tIL, W. : A relativistic generalization of the S U(6) symmetry group, CEI~N
peprint (TH. 505), and Baryons and mesons in a theory which combines
relativistic invarianee with S U(6) symmetry, CERN preprint (TH. 514).
[8] BAcR~, It., and J. NvYTs: Remarks on an enlarged Poincar6 group, to be
published in the Nuovo Cimento.
[9] BAcaY, H.: Some classical groups of space-time transformations and theh"
S U (6) generalization, CERN preprint (TH. 526), to be published in Annales
de l'Institut Henri Poincar~.
[10] R A c ~ , G. : Group theory and spectroscopy, Lecture given at the Institute for
Advanced Study, Princeton (1951).