Acta Polytechnica


doi:10.14311/AP.2013.53.0395
Acta Polytechnica 53(5):395–398, 2013 © Czech Technical University in Prague, 2013

available online at http://ojs.cvut.cz/ojs/index.php/ap

THE SHMUSHKEVICH METHOD FOR HIGHER SYMMETRY
GROUPS OF INTERACTING PARTICLES

Mark Bodnera, Goce Chadzitaskosb, Jiří Paterac,a,
Agnieszka Tereszkiewiczd,c,∗

a MIND Research Institute, 111 Academy Drive, Irvine, CA 92617, USA
b Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University in Prague, Břehová 7,

CZ-11519 Praha 1, Czech Republic
c Centre de Recherches Mathématiques, Université de Montréal, C. P. 6128, succ. Centre-ville, Montréal,

H3C 3J7, Québec, Canada
d Institute of Mathematics, University of Bialystok, Akademicka 2, PL-15-267 Bialystok, Poland
∗ corresponding author: a.tereszkiewicz@uwb.edu.pl

Abstract. About 60 years ago, I. Shmushkevich presented a simple ingenious method for computing
the relative probabilities of channels involving the same interacting multiplets of particles, without
the need to compute the Clebsch-Gordan coefficients. The basic idea of Shmushkevich is “isotopic
non-polarization” of the states before the interaction and after it. Hence his underlying Lie group was
SU (2). We extend this idea to any simple Lie group. This paper determines the relative probabilities of
various channels of scattering and decay processes following from the invariance of the interactions with
respect to a compact simple a Lie group. Aiming at the probabilities rather than at the Clebsch-Gordan
coefficients makes the task easier, and simultaneous consideration of all possible channels for given
multiplets involved in the process, makes the task possible. The probability of states with multiplicities
greater than 1 is averaged over. Examples with symmetry groups O(5), F (4), and E(8) are shown.

Keywords: isospin, particle collisions, Lie group representation.

Submitted: 1 April 2013. Accepted: 7 May 2013.

1. Introduction
The method of Shmushkevich [1, 2] was conceived
as a simpler alternative to computing the relative
probabilities of various channels of scattering and
decay processes under strict isospin invariance (SU (2)
invariance). The traditional alternative method for
calculating the probabilities of the same channels is to
calculate first all pertinent Clebsch-Gordan coefficients
(CGC) for the channel.

The most remarkable feature of the Shmushkevich
method is the complete avoidance of the need to calcu-
late the Clebsch-Gordan coefficients. The underlying
idea is to consider isotopically unpolarized states be-
fore and after the interaction, assuming that each
possible state came with equal probability.
The simplicity of the idea has attracted the atten-

tion of many physicists [3–6].
Technically, the two methods differ in their ob-

jective: Shmushkevich’s method calculates just the
probabilities. The conventional alternative method
calculates the CGC, their squares, and then provides
the probabilities. Neither of the tasks is easy for
higher ranks of the representations.

From the point of view of symmetries the two meth-
ods differ by the symmetry group that they exploit:
Shmushkevich uses just the Weyl group of the Lie
group, while CGCs are built using the symmetry group
of Demazure-Tits, see [7, 8].

The difficulty of generalizing of Shmushkevich’s
method to higher rank groups lies in the frequent
occurrence of multiple states with the same quantum
numbers, equivalently labeled by the same weights
of irreducible representations [9], as well as the sheer
number of channels that need to be written down.

It is likely that practical exploitation of Shmushke-
vich’s idea for higher groups and possibly representa-
tions of much higher dimensions, will not proceed by
spelling out the large number of channels for each case
and counting the number of occurrences of each state
in all the channels. Instead, one would start from
one known channel and use the symmetry group, the
Weyl group in this case, to produce other channels
with the same probability. This is a routine oper-
ation which, however does not produce all possible
channels. This will relate only the states which are
situated on the same Weyl group orbit. This may
be all that one needs as long as only the channels
defined by individual orbits are studied. However, if
the equal probability of all the states of the Lie group
is to be involved, the link between different orbits
present in the same representation has to be imposed
independently. For the probabilities, a natural link
is provided by the requirement that the probabilities
add up to one. If an orbit is present in irreducible
representation more than once, say m times, we count
them as equally probable m channels. In this paper

395

http://dx.doi.org/10.14311/AP.2013.53.0395
http://ojs.cvut.cz/ojs/index.php/ap


M. Bodner, G. Chadzitaskos, J. Patera, A. Tereszkiewicz Acta Polytechnica

C2 Decomposition into
irreducible representations

with multiplicities

Multiplicity Orbit size

(20) × (10) (30) (11) (10)
10 × 4 [30] 1 4

[11] [11] 2 8
2[10] 2[10] [10] 5 4
Decomposition into orbits

with multiplicities
[30] 2[11] 5[10] 40

Table 1. Decomposition of the product in the C2 example. Decomposition is given in the weight system of irreducible
representations (the first line) and in terms of orbits (bottom line). The dimensions of the representations and the
sizes of the orbits are shown together with the multiplicities.

we provide an illustration of this approach, for the
symmetry groups O(5), F (4), and E(8). The process
that consider is the simplest, where three multiplets
are interacting, more precisely the interaction of two
particles yields a third one. Our aim is to show how
to average over different particles/states which carry
the same Lie group representation labels.

We write the highest weights of the representations
in round brackets, and the dominant weights of the
Weyl group orbits in square brackets.

In the examples, we show that there are many states
which have the group labels (weights) identical al-
though they label different particle states. In order
to avoid the almost impossible task of distinguishing
these states, we add them up and count their total
probability.

2. Symmetry group O(5)
Consider the example where the underlying symmetry
group is the Weyl reflection group of the Lie group
O(5), or equivalently of the Lie algebra C2. We label
the representations by their unique highest weight
(relative to the basis of the fundamental weights).
The product of representations of dimensions 10 and
4 decomposes as follows,

(20) × (10) = (30) + (11) + (10),
10 × 4 = 20 + 16 + 4 = 40,

where the second line shows the dimensions of repre-
sentations, see [10]. Labeling the Weyl group orbits
by their unique dominant weights, the product of the
weight systems decomposes into the Weyl group orbits
as follows,

(20) × (10) = [30] + 2[11] + 5[10],
10 × 4 = 4 + 2 · 8 + 5 · 4 = 40,

where the integers in front of the square brackets are
the multiplicities of the occurrence of the respective
orbits in the decomposition.

If only the product of the Weyl group orbits were to
be considered, the decomposition would be simpler:

[20] × [10] = [30] + [11] + [10],
4 × 4 = 4 + 8 + 4 = 16.

There are 40 states in the product. If equal probability
is assumed, each of the channels comes with the prob-
ability 140 . Consequently, we have the probabilities:
• 4 states from [30], each present once: 1/40;
• 8 states from [11], each present twice: 1/20;
• 4 states from [10], each present 5 ×: 1/8.
The results of the example are summarized in Table 1.

3. Symmetry group F (4)
Consider decomposition of the product of representa-
tions in terms of their weight systems,

(0001) × (0001) = (0002) + (0010) + (1000)
+ (0001) + (0000),

26 × 26 = 324 + 273 + 52 + 26 + 1 = 676.

Decomposition of the same product in terms of Weyl
group orbits,

(0001) × (0001) = [0002] + 2[0010] + 6[1000]
+ 12[0001] + 28[0000]

and the corresponding equality of the dimensions

26 × 26 = 24 + 2 · 96 + 6 · 24 + 12 · 24 + 28 · 1.

Suppose that we want to decompose only the product
of the orbits of the highest weights

[0001] × [0001] = [0002] + 2[0010] + 6[1000]
+ 8[0001] + 24[0000],

24 × 24 = 24 + 2 · 96 + 6 · 24 + 8 · 24 + 24 · 1

If equal probability of the 676 states is assumed we
have the following probabilities of the channels:

396



vol. 53 no. 5/2013 The Shmushkevich Method for Higher Symmetry Groups

F (4) Decomposition into irreducible
representations with multiplicities

Multiplicity Orbit size

(0001) × (0001) = (0002) (0010) (1000) (0001) (0000)
26 × 26 [0002] 1 24

[0010] [0010] 2 96
3[1000] 2[1000] [1000] 6 24
5[0001] 5[0001] [0001] [0001] 12 24

12[0000] 9[0000] 4[0000] 2[0000] [0000] 28 1

Decomposition into orbits with multiplicities
[0002] 2[0010] 6[1000] 12[0001] 28[0000] 676 = 262

Table 2. Decomposition of the product in the F (4) example. The decomposition is given in the weight system of
irreducible representations (the first line) and in terms of orbits (bottom line). The dimensions of the representations
and the sizes of the orbits are shown together with the multiplicities.

• 24 states from [0002], each present once: 1/676;
• 96 states from [0010], each present twice: 2/676;
• 24 states from [1000], each present 6 ×: 6/676;
• 24 states from [0001], each present 12 ×: 12/676;
• 1 state from [0000], each present 28 ×: 28/676.
The results of the example are summarized in Table 2.

4. Symmetry group E(8)
Consider decomposition of the product of the repre-
sentations in terms of their weight systems(

0
1000000

)
×

(
0

1000000
)

=
(

0
2000000

)
+

(
0

0100000
)

+
(

0
0000001

)
+

(
0

1000000
)

+
(

0
0000000

)
with the respective dimensions

248 × 248 = 27000 + 30380 + 3875 + 248 + 1
= 61504.

We write the components of E(8) weights as they
would be attached to the corresponding Dynkin dia-
gram.

The same product decomposed into the sum of the
Weyl group orbits has very different multiplicities,(

0
1000000

)
×

(
0

1000000
)

=
[

0
2000000

]
+ 2

[
0

0100000
]

+ 14
[

0
0000001

]
+ 72

[
0

1000000
]

+ 304
[

0
0000000

]
,

and the equality of the dimensions in the decomposed
product,

248 × 248 = 240 + 2 · 6720 + 14 · 2160
+ 72 · 240 + 304 · 1 = 61504.

If only the product of the orbits is to be calculated,
the result is much simpler,[

0
1000000

]
×

[
0

1000000
]

=
[

0
2000000

]
+ 2

[
0

0100000
]

+ 14
[

0
0000001

]
+ 56

[
0

1000000
]

+ 240
[

0
0000000

]
,

and the corresponding orbit sizes with the appropriate
multiplicities are

240 × 240 = 240 + 2 · 6720
+ 14 · 2160 + 56 · 240 + 240 · 1 = 57600.

If equal probability of the 61504 states is assumed,
we have the following probabilities of the channels:
• 240 states from

[
0

2000000
]
,

each present once: 1/61504;
• 6720 states from

[
0

0100000
]
,

each present twice: 2/61504;
• 2160 states from

[
0

0000001
]
,

each present 14 times: 14/61504;
• 240 states from

[
0

1000000
]
,

each present 72 times: 72/61504;
• 1 state from

[
0

0000000
]
,

each present 304 times: 304/61504.
The results of the example are summarized in Table 3.

Acknowledgements
The authors are grateful for support from Natural Sci-
ences and Engineering Research of Canada, to the MIND
Research Institute of Irvine, California, and to MITACS.
A.T. expresses her gratitude to the Centre de Recherches
Mathématiques, Université de Montréal for the hospitality
during her postdoctoral fellowship. G.C. wishes to express
thanks for support from the Ministry of Education of the
Czech Republic (project MSM6840770039).

References
[1] I.M. Shmushkevich, On deduction of relations between

sections that arise from the hypothesis of isotopic
invariance, Dokl. Akad. Nauk SSSR 103 (1955)
235–238 (in Russian)

[2] N. Dushin, I.M. Shmushkevich, On the Relations
Between Cross Sections which Result from the
Hypothesis of Isotopic Invariance, Dokl. Akad. Nauk
SSSR 106 (1956) 801-805; translated in Soviet Phys.
Dokl. 1 (1956) 94–98

397



M. Bodner, G. Chadzitaskos, J. Patera, A. Tereszkiewicz Acta Polytechnica

E(8) Decomposition into irreducible
representations with multiplicities

Multi-
plicity

Orbit
size(

0
1000000

)
×

(
0

1000000
) (

0
2000000

) (
0

0100000
) (

0
0000001

) (
0

1000000
) (

0
0000000

)
248 × 248

[
0

2000000
]

1 240[
0

0100000
] [

0
0100000

]
2 6720

6
[

0
0000001

]
7
[

0
0000001

] [
0

0000001
]

14 2160
29

[
0

1000000
]

35
[

0
1000000

]
7
[

0
1000000

] [
0

1000000
]

72 240
120

[
0

0000000
]

140
[

0
0000000

]
35

[
0

0000000
]

8
[

0
1000000

] [
0

0000000
]

304 1

Decomposition into orbits with multiplicities[
0

2000000
]

2
[

0
0100000

]
14

[
0

0000001
]

72
[

0
1000000

]
304

[
0

0000000
]

61504 = 2482

Table 3. Decomposition of the product in the E(8) example. The decomposition is given in the weight system of
irreducible representations (the first line) and in terms of orbits (bottom line). The dimensions of the representations
and the sizes of the orbits are shown together with the multiplicities.

[3] G. Pinsky, A.J. Macfarlane, E.C.G. Sudarshan,
Shmushkevich’s Method for a Charge-Independent
Theory, Phys. Rev. 140, (1965), B1045–B1053

[4] C.G. Wohl, Isospin relations by counting, Amer. J.
Phys. 50 , (1982), 748-753

[5] P. Roman, The Theory of Elementary Particles,
North-Holland Pub. Co. (1960)

[6] R. Marshak, E. Sudarshan, Introduction to
Elementary Particles Physics, New York, Interscience
Publishers, 1961

[7] R. V. Moody, J. Patera, General charge conjugation
operators in simple Lie groups, J. Math. Phys., 25
(1984) 2838–2847

[8] L. Michel, J. Patera, R. T. Sharp, Demazure-Tits

subgroup of a simple Lie group, J. Math. Phys., 29
(1988) 777–796

[9] P. Ramond, Group Theory. A Physicist’s Survey,
Cambridge Univ. Press, 2010

[10] W. G. McKay, J. Patera, D. W. Rand, Tables of
Representations of Simple Lie Algebras (Exceptional
Simple Lie Algebras : Vol. 1), Université de Montréal,
Centre de Recherches Mathématiques, 1990

[11] R. V. Moody, J. Patera, Fast recursion formula for
weight multiplicities, Bull. Amer. Math. Soc., 7 (1982)
237–242

[12] M. R. Bremner, R. V. Moody, J. Patera, Tables of
dominant weight multiplicities for representations of
simple Lie algebras, Marcel Dekker, New York 1985, 340
pages, ISBN: 0-8247-7270-9

398


	Acta Polytechnica 53(5):395–398, 2013
	1 Introduction
	2 Symmetry group O(5)
	3 Symmetry group F(4)
	4 Symmetry group E(8)
	Acknowledgements
	References