Abstract
We study when a tensor product of irreducible representations of the symmetric group \(S_n\) contains all irreducibles as subrepresentations—we say such a tensor product covers \(\textsf{Irrep}(S_n)\). Our results show that this behavior is typical. We first give a general sufficient criterion for tensor products to have this property, which holds asymptotically almost surely for constant-sized collections of (Plancherel or uniformly) random irreducibles. We also consider the minimal tensor power of a single fixed irreducible representation needed to cover \(\textsf{Irrep}(S_n)\). Here a simple lower bound comes from considering dimensions, and we show it is always tight up to a universal constant factor as was recently conjectured by Liebeck, Shalev, and Tiep.
Similar content being viewed by others
Notes
We show the latter from the former in Appendix A, simplifying [16] which used two separate arguments.
References
Bessenrodt, C., Behns, C.: On the Durfee size of Kronecker products of characters of the symmetric group and its double covers. J. Algebra 280(1), 132–144 (2004)
Baik, J., Deift, P., Johansson, K.: On the distribution of the length of the longest increasing subsequence of random permutations. J. Am. Math. Soc. 12(4), 1119–1178 (1999)
Bürgisser, P., Ikenmeyer, C.: The complexity of computing Kronecker coefficients. In: 20th Annual International Conference on Formal Power Series and Algebraic Combinatorics (FPSAC 2008), Discrete Math. Theor. Comput. Sci. Proc., AJ, pp. 357–368. Assoc. Discrete Math. Theor. Comput. Sci., Nancy (2008)
Borodin, A., Okounkov, A., Olshanski, G.: Asymptotics of Plancherel measures for symmetric groups. J. Am. Math. Soc. 13(3), 481–515 (2000)
Christandl, M., Harrow, A.W., Mitchison, G.: Nonzero kronecker coefficients and what they tell us about spectra. Commun. Math. Phys. 270(3), 575–585 (2007)
Christandl, M., Mitchison, G.: The spectra of quantum states and the Kronecker coefficients of the symmetric group. Commun. Math. Phys. 261(3), 789–797 (2006)
Dvir, Y.: On the Kronecker product of \(S_n\) characters. J. Algebra 154(1), 125–140 (1993)
Goh, W.M.Y., Schmutz, E.: The number of distinct part sizes in a random integer partition. J. Comb. Theory Ser. A 69(1), 149–158 (1995)
Harman, N., Ryba, C.: A tensor-cube version of the Saxl conjecture (2022). arXiv preprint arXiv:2206.13769
Heide, G., Saxl, J., Tiep, P.H., Zalesski, A.E.: Conjugacy action, induced representations and the Steinberg square for simple groups of Lie type. Proc. Lond. Math. Soc. 106(4), 908–930 (2013)
Ikenmeyer, C.: The Saxl conjecture and the dominance order. Discrete Math. 338(11), 1970–1975 (2015)
Ikenmeyer, C., Mulmuley, K.D., Walter, M.: On vanishing of Kronecker coefficients. Comput. Complex. 26(4), 949–992 (2017)
Klyachko, A.: Quantum marginal problem and representations of the symmetric group (2004). arXiv preprint arXiv:quant-ph/0409113
Li, X.: Saxl conjecture for triple hooks. Discrete Math. 344(6), 112340 (2021)
Logan, B.F., Shepp, L.A.: A variational problem for random Young tableaux. Adv. Math. 26(2), 206–222 (1977)
Luo, S., Sellke, M.: The Saxl conjecture for fourth powers via the semigroup property. J. Algebraic Comb. 45(1), 33–80 (2017)
Liebeck, M., Shalev, A., Tiep, P.H.: On the diameters of McKay graphs for finite simple groups. Isr. J. Math. 241, 449–464 (2020)
Liebeck, M., Shalev, A., Tiep, P.: McKay graphs for alternating and classical groups. Trans. Am. Math. Soc. 374(08), 5651–5676 (2021)
Mulmuley, K.D., Sohoni, M.: Geometric complexity theory II: towards explicit obstructions for embeddings among class varieties. SIAM J. Comput. 38(3), 1175–1206 (2008)
Pak, I., Panova, G., Vallejo, E.: Kronecker products, characters, partitions, and the tensor square conjectures. Adv. Math. 288, 702–731 (2016). https://doi.org/10.1016/j.aim.2015.11.002
Rosas, M.H.: The Kronecker product of Schur functions indexed by two-row shapes or hook shapes. J. Algebraic Comb. 14(2), 153–173 (2001)
Vershik, A.M.: Statistical mechanics of combinatorial partitions, and their limit shapes. Funct. Anal. Appl. 30(2), 90–105 (1996). https://doi.org/10.1007/BF02509449
Veršik, A.M., Kerov, S.V.: Asymptotic behavior of the Plancherel measure of the symmetric group and the limit form of Young tableaux. Dokl. Akad. Nauk SSSR 233(6), 1024–1027 (1977)
Wilf, H.S.: Three problems in combinatorial asymptotics. J Comb Theory Ser A 35(2), 199–207 (1983)
Acknowledgements
The author gratefully acknowledges support of NSF and Stanford Graduate Fellowships. I thank Daniel Bump, Pavel Etingof, Xiaoyue Gong, Sammy Luo, Alex Malcolm, Chris Ryba and the anonymous referee for helpful comments, corrections, discussions, and references. I thank Greta Panova for bringing [17] to my attention after the initial posting of this paper.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A: Alternate proof of fourth power Saxl theorem
Appendix A: Alternate proof of fourth power Saxl theorem
Here we give an alternate proof of Theorem 2.10 that \(\varvec{\varrho }_r^{\otimes 4}\) covers \(\textsf{Irrep}(S_n)\) for r sufficiently large (which is Theorem 1.4 of [16]) based on another main result from [16]. The implication is immediate from a lemma on the representation theory of an arbitrary finite group G which we suspect to be known but have not been able to locate in the literature.
Definition A.1
Let G be a finite group and let \(M_G\) be the Plancherel probability measure on \(\textsf{Irrep}(G)\) which assigns an irreducible representation \(\varvec{\lambda }\) a probability \(M_G(\varvec{\lambda })=\frac{\dim (\varvec{\lambda })^2}{|G|}\). For an arbitrary finite-dimensional G-representation V, let \(M_G(V)\) denote the Plancherel measure of the set of distinct irreducible subrepresentations of V.
Theorem 1.6 of [16] states that \(\varvec{\varrho }_r^{\otimes 2}\) contains Plancherel-asymptotically-almost-all of \({\mathcal {Y}}_n\) for \(n=\left( {\begin{array}{c}r+1\\ 2\end{array}}\right) \), i.e. \(\lim _{r\rightarrow \infty } M_{S_n}(\varvec{\varrho }_r^{\otimes 2})=1\). Therefore Theorem 2.10 follows immediately from the lemma below. We note that proving Theorem 1.6 of [16] relies on the deep work of [4], so the proof of Theorem 2.10 given in [16] is more elementary than the present proof. Nonetheless we find the connection enlightening.
Lemma A.2
Suppose \(M_G(V)+M_G(W)>1\). Then \(V\otimes W\) covers \(\textsf{Irrep}(G)\).
Proof
The conclusion is equivalent to the statement that \(\langle \chi ^V\chi ^W,\chi ^{\varvec{\lambda }^*}\rangle >0\) for any irreducible representation \(\varvec{\lambda }\), where \((\cdot )^{*}\) denotes the dual representation. As \(\langle \chi ^V\chi ^W,\chi ^{\varvec{\lambda }^*}\rangle =\langle \chi ^W,\chi ^{V^*}\chi ^{\varvec{\lambda }^*}\rangle \), this is equivalent to showing the tensor product \(V^*\otimes \varvec{\lambda }^*\) shares some irreducible subrepresentation with W. We will prove that \(M_G(V\otimes \varvec{\lambda })\ge M_G(V)\). This implies \(M_G(V^*\otimes \varvec{\lambda }^*)+M_G(W)=M_G(V\otimes \varvec{\lambda })+M_G(W)>1\) so that they share a subrepresentation by the pigeonhole principle.
To see this, we work in the standard inner product space \(L^2(G)\) and recall that irreducible characters \(\chi ^{\varvec{\mu }}\) for \(\varvec{\mu }\in \textsf{Irrep}(G)\) are orthonormal. We identify representations with their characters. Consider for any representation U the best \(L^2\) approximation to the regular representation \({{\textbf {Reg}}}\) of G lying in the linear space
From the point of view of irreducible representations it is clear that the best approximation \({\tilde{U}}\) is obtained by projection via \(a_{\varvec{\mu }_U}=\dim (\varvec{\mu }_U)\), and the \(L^2\) error of this approximation \({\tilde{U}}\) is therefore \(|{{\textbf {Reg}}}-\tilde{U}|=\sqrt{|G|\cdot (1-M_G(U))}\).
We form \({\tilde{V}}\) and multiply its character by \(\frac{\chi ^{\varvec{\lambda }}}{\dim \varvec{\lambda }}\), and by abuse of notation treat this as a tensor product of fractional representations. The key point is that \({\tilde{V}}\otimes \frac{\varvec{\lambda }}{\dim \varvec{\lambda }}\) has the same character value at the identity element of G, and a smaller character value (in absolute value) at all other elements. Since \({{\textbf {Reg}}}\) has character value 0 at all non-identity elements, computing the distances using the character basis implies
Moreover \({\tilde{V}}\otimes \frac{\varvec{\lambda }}{\dim \varvec{\lambda }}\) is in the \({\mathbb {R}}\)-span of the irreducible subrepresentations of \(V\otimes \varvec{\lambda }\). Since the function \(\sqrt{1-M_G(U)}\) is decreasing in \(M_G(U)\), the fact that by using subrepresentations of \(V\otimes \varvec{\lambda }\) we weakly improved upon the best \(L^2\) approximation to \({{\textbf {Reg}}}\) using subrepresentations of V implies \(M_G(V\otimes \varvec{\lambda })\ge M_G(V)\) as desired. \(\square \)
Note that Lemma A.2 becomes completely false if Plancherel measure is replaced by uniform measure. For instance, the group of invertible affine transformations of \({\mathbb {F}}_p\) has \(p-1\) irreducible representations of dimension 1 and one of dimension \(p-1\). If V, W each contain exactly the 1-dimensional irreducible representations then \(V\otimes W\) still consists of only 1-dimensional irreducibles, hence does not cover \(\textsf{Irrep}(G)\).
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Sellke, M. Covering \(\textsf{Irrep}(S_n)\) With tensor products and powers. Math. Ann. 388, 831–865 (2024). https://doi.org/10.1007/s00208-022-02532-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00208-022-02532-3