## Publications of George McNinch

## Summary list

## Detailed list

*Nilpotent elements and maximal rank subgroups of a reductive group over a local field*, 2019.**Citation:**Preprint.**URL(s):**[e-print]**Abstract:**Let \(K\) be the field of fractions of a complete discrete valuation ring \(A\) with perfect residue field \(k\) of characteristic \(p\), and let \(G\) be a connected and reductive algebraic group over \(K\) which splits over an unramified extension of \(K\).

Suppose that \(P\) is a parahoric group scheme over \(A\) with generic fiber \(P_K = G\). A nilpotent section \(\mathcal{X} \in \operatorname{Lie}(P)\) is

*balanced*if the fibers \(C_K\) and \(C_k\) are smooth group schemes of the same dimension, where \(C=C_P(\mathcal{X})\) is the scheme theoretic centralizer of \(\mathcal{X}\) for the adjoint action of \(P.\) The identity component of the centralizer \(C_P(\mathcal{X})\) of a balanced nilpotent section is*smooth*over \(A\). If \(X_0\) is a nilpotent element in the Lie algebra of the reductive quotient of the special fiber \(P_k\), we give conditions for the existence and conjugacy of balanced nilpotent sections \(\mathcal{X}\) of \(\operatorname{Lie}(P)\) with \(X_0 = \mathcal{X}_k\).The construction of balanced sections given here provides useful qualitative information about the parametrization of \(G(K)\)-orbits on the nilpotent elements of \(\operatorname{Lie}(G)\) described in [DeBacker 2002].

*Reductive subgroup schemes of a parahoric group scheme*, 2018.**Citation:**To appear in*Transformation Groups*.**Abstract:**Let \(K\) be the field of fractions of a complete discrete valuation ring \(A\) with residue field \(k\), and let \(G\) be a connected reductive algebraic group over \(K\). Suppose \(P\) is a parahoric group scheme attached to \(G\). In particular, \(P\) is a smooth affine \(A\)-group scheme having generic fiber \(P_K = G\); the group scheme \(P\) is in general not reductive over \(A\).

If \(G\) splits over an unramified extension of \(K\), we find in this paper a closed and reductive A-subgroup scheme \(M ⊂ P\) for which the special fiber \(M_k\) is a Levi factor of \(P_k\). Moreover, we show that the generic fiber \(M_K\) is a subgroup of G which is geometrically of type \(C(μ)\) -- i.e. after a separable field extension, \(M_K\) is the identity component \(M_K = C_G^o(φ)\) of the centralizer of the image of a homomorphism \(φ:μ_n → H\), where \(μ_n\) is the group scheme of \(n\)-th roots of unity for some \(n \gt 2\). For a connected and split reductive group \(H\) over any field \(F\), the paper describes those subgroups of \(H\) which are of type \(C(μ)\).

*Central subalgebras of the centralizer of a nilpotent element*, 2016. With Donna M. Testerman (EPFL).**Citation:***Proceedings of the American Mathematical Society*144(6), 2016, pp. 2383--2397.**Abstract:**Let \(G\) be a connected, semisimple algebraic group over a field \(k\) whose characteristic is

*very good*for \(G\). In a canonical manner, one associates to a nilpotent element \(X \in \operatorname{Lie}(G)\) a parabolic subgroup \(P\) -- in characteristic zero, \(P\) may be described using an \(\mathfrak{sl}(2)\) triple containing \(X\); in general, \(P\) is the “instability parabolic” for \(X\) as in geometric invariant theory. In this setting, we are concerned with the center \(Z(C)\) of the centralizer \(C\) of \(X\) in \(G\). Choose a Levi factor \(L\) of \(P\), and write \(d\) for the dimension of the center \(Z(L)\). Finally, assume that the nilpotent element \(X\) is*even*. In this case, we can*deform*\(\operatorname{Lie}(L)\) to \(\operatorname{Lie}(C)\), and our deformation produces a \(d\) dimensional subalgebra of \(\operatorname{Lie}(Z(C))\). Since \(Z(C)\) is a smooth group scheme, it follows that \(\dim Z(C) \ge d = \dim Z(L)\). In fact, Lawther and Testerman have proved that \(\dim Z(C) = \dim Z(L)\). Despite only yielding a partial result, the interest in the method found in the present work is that it avoids the extensive case checking carried out by Lawther-Testerman in the memoir [LT 11].*Levi factors of the special fiber of a parahoric group scheme and tame ramification*, 2014.**Citation:***Algebras and Representation Theory*17(2), 2014, pp. 469--479.**Abstract:**Let \(A\) be a Henselian discrete valuation ring with fractions \(K\) and with

*perfect*residue field \(k\) of characteristic \(p\gt 0\). Let \(G\) be a connected and reductive algebraic group over \(K\), and let \(P\) be a parahoric group scheme over \(A\) with generic fiber \(P_K = G\). The special fiber \(P_k\) is a linear algebraic group over \(k\).If \(G\) splits over an unramified extension of \(K\), we proved in some previous work that the special fiber \(P_k\) has a Levi factor, and that any two Levi factors of \(P_k\) are geometrically conjugate. In the present paper, we extend a portion of this result. Following a suggestion of Gopal Prasad, we prove that if \(G\) splits over a

*tamely ramified*extension of \(K\), then the*geometric*special fiber \(P_{\overline{k}}\) has a Levi factor.*Linearity for actions on vector groups*, 2014.**Citation:***Journal of Algebra*397, 2014, pp. 666--688.**Abstract:**Let \(k\) be an arbitrary field, let \(G\) be a (smooth) linear algebraic group over \(k\), and let \(U\) be a vector group over \(k\) on which \(G\) acts by automorphisms of algebraic groups. The action of \(G\) on \(U\) is said to be

*linear*if there is a G equivariant isomorphism of algebraic groups \(U \simeq \operatorname{Lie}(U)\).Suppose that \(G\) is connected and that the unipotent radical of \(G\) is defined over \(k\). If the \(G\) module \(\operatorname{Lie}(U)\) is simple, we show that the action of \(G\) on \(U\) is linear. If \(G\) acts by automorphisms on a connected, split unipotent group \(U\), we deduce that \(U\) has a filtration by \(G\) invariant closed subgroups for which the successive factors are vector groups with a linear action of \(G\). This verifies for such \(G\) an assumption made in earlier work of the author on the existence of Levi factors.

On the other hand, for any field \(k\) of positive characteristic we show that if the category of representations of \(G\) is not semisimple, there is an action of \(G\) on a suitable vector group \(U\) which is not linear.

*On the descent of Levi factors*, 2013.**Citation:***Archiv der Mathematik*100(1), 2013, pp. 7--24.**Abstract:**Let \(G\) be a linear algebraic group over a field \(k\) of characteristic \(p\gt0\), and suppose that the unipotent radical \(R\) of \(G\) is defined and split over \(k\). By a Levi factor of \(G\), one means a closed subgroup \(M\) which is a complement to \(R\) in \(G\). In this paper, we give two results related to the descent of Levi factors.

First, suppose \(\ell\) is a finite Galois extension of \(k\) for which the extension degree \([\ell:k]\) is relatively prime to \(p\). If \(G_\ell\) has a Levi decomposition, we show that \(G\) has a Levi decomposition. Second, suppose that there is a \(G\)-equivariant isomorphism of algebraic groups \(R \simeq \operatorname{Lie}(R)\) -- i.e. \(R\) is a vector group with a linear action of the reductive quotient \(G/R\). If \(G_L\) has a Levi decomposition for a separable closure \(L=k_{\operatorname{sep}}\) of \(k\), then \(G\) has a Levi decomposition.

Finally, we give an example of a disconnected, abelian, linear algebraic group \(G\) for which \(G_L\) has a Levi decomposition over a separable closure \(L\), but \(G\) itself has no Levi decomposition.

*Some good-filtration subgroups of simple algebraic groups*, 2013. With Chuck Hague (McKeogh Co.).**Citation:***Journal of Pure and Applied Algebra*217(12), 2013, pp. 2400--2413.**Abstract:**Let \(G\) be a connected and reductive algebraic group over an algebraically closed field of characteristic \(p\gt 0\). An interesting class of representations of \(G\) consists of those \(G\)-modules having a

*good filtration*-- i.e. a filtration whose layers are the*standard*highest weight modules obtained as the space of global sections of \(G\) linearized line bundles on the flag variety of \(G\). Let \(H \subset G\) be a connected and reductive subgroup of \(G\). One says that \((G,H)\) is a*Donkin pair*, or that \(H\) is a*good filtration subgroup*of \(G\), if whenever the \(G\)-module \(V\) has a good filtration, the \(H\)-module \(\operatorname{res}^G_H V\) has a good filtration.In this paper, we show when \(G\) is a “classical group” that the

*optimal*\(\operatorname{SL}(2)\) subgroups of \(G\) are good filtration subgroups. We also consider the cases of subsystem subgroups in all types and determine some primes for which they are good filtration subgroups.*Levi decompositions of a linear algebraic group*, 2010.**Citation:***Transformation Groups*15(4), 2010, pp. 937--964.**Abstract:**If \(G\) is a connected linear algebraic group over the field \(k\), a Levi factor of \(G\) is a reductive complement to the unipotent radical of \(G\). If \(k\) has positive characteristic, \(G\) may have no Levi factor, or \(G\) may have Levi factors which are not geometrically conjugate. We give in this paper some

*sufficient*conditions for the existence and the conjugacy of Levi factors of \(G\).Let \(A\) be a Henselian discrete valuation ring with fractions \(K\) and with

*perfect*residue field \(k\) of characteristic \(p \gt 0\). Let \(G\) be a connected and reductive algebraic group over \(K\). Bruhat and Tits have associated to \(G\) certain smooth \(A\) group schemes \(P\) whose generic fibers \(P_k\) coincide with \(G\); these are known as*parahoric group schemes*. The special fiber \(P_k\) of a parahoric group scheme is a linear algebraic group over \(k\). If \(G\) splits over an unramified extension of \(K\), we show that \(P_k\) has a Levi factor, and that any two Levi factors of \(P_k\) are geometrically conjugate.*Nilpotent centralizers and Springer isomorphisms*, 2009. With Donna M. Testerman (EPFL).**Citation:***Journal of Pure and Applied Algebra*213(7), 2009, pp. 1346--1363.**Abstract:**Let \(G\) be a semisimple algebraic group over a field \(K\) whose characteristic is

*very good*for \(G\), and let σ be any \(G\) equivariant isomorphism from the nilpotent variety to the unipotent variety; the map σ is known as a Springer isomorphism. Let \(y \in G(K)\), let \(Y ∈ \operatorname{Lie}(G)(K)\), and write \(C_y = C_G(y)\) and \(C_Y= C_G(Y)\) for the centralizers. We show that the center of \(C_y\) and the center of \(C_Y\) are smooth group schemes over \(K\). The existence of a Springer isomorphism is used to treat the crucial cases where \(y\) is unipotent and where \(Y\) is nilpotent.Now suppose \(G\) to be quasisplit, and write \(C\) for the centralizer of a rational

*regular*nilpotent element. We obtain a description of the normalizer \(N_G(C)\) of \(C\), and we show that the automorphism of \(\operatorname{Lie}(C)\) determined by the differential of σ at zero is a scalar multiple of the identity; these results verify observations of J-P. Serre.*Nilpotent subalgebras of semisimple Lie algebras*, 2009. With Paul Levy (Lancaster Univ) and Donna M. Testerman (EPFL).**Citation:***Comptes Rendus Mathématique, Académie des Sciences, Paris*347(9-10), 2009, pp. 477--482.**Abstract:**Let \(\operatorname{Lie}(G)\) be the Lie algebra of a semisimple linear algebraic group. Under mild conditions on the characteristic of the underlying field, one can show that any subalgebra of \(\operatorname{Lie}(G)\) consisting of nilpotent elements is contained in some Borel subalgebra. In this Note, we provide examples for each semisimple group \(G\) and for each of the torsion primes for \(G\) of nil subalgebras not lying in any Borel subalgebra of \(\operatorname{Lie}(G)\).

*The centralizer of a nilpotent section*, 2008.**Citation:***Nagoya Mathematical Journal*190, 2008, pp. 129--181.**URL(s):**[Project-Euclid:euclid.nmj/1214229081][arXiv:math.RT/0605626][e-print][MR:2423832][BibTeX][AMSRefs]**Abstract:**Let \(F\) be an algebraically closed field and let \(G\) be a semisimple \(F\) algebraic group for which the characteristic of \(F\) is

*very good*. If \(X ∈ \operatorname{Lie}(G) = \operatorname{Lie}(G)(F)\) is a nilpotent element in the Lie algebra of \(G\), and if \(C\) is the centralizer in \(G\) of \(X\), we show that (i) the root datum of a Levi factor of \(C\), and (ii) the component group \(C/C^o\) both depend only on the Bala-Carter label of \(X\); i.e. both are independent of very good characteristic. The result in case (ii) depends on the known case when \(G\) is (simple and) of adjoint type.The proofs are achieved by studying the centralizer \(C\) of a nilpotent section \(X\) in the Lie algebra of a suitable semisimple group scheme over a Noetherian, normal, local ring \(A\). When the centralizer of \(X\) is equidimensional on \(\operatorname{Spec}(A)\), a crucial result is that locally in the étale topology there is a smooth \(A\) subgroup scheme \(L\) of \(C\) such that \(L_t\) is a Levi factor of \(C_t\) for each \(t ∈ \operatorname{Spec}(A)\).

*Completely reducible Lie subalgebras*, 2007.**Citation:***Transformation Groups*12(1), 2007, pp. 127--135.**URL(s):**[DOI:10.1007/s00031-005-1130-5][arXiv:math.RT/0509590][e-print][MR:2308032][BibTeX][AMSRefs]**Abstract:**Let \(G\) be a connected and reductive group over the algebraically closed field \(K\). J-P. Serre has introduced the notion of a G completely reducible subgroup \(H ⊂ G\). In this note, we give a notion of \(G\) complete reducibility -- \(G\)-cr for short -- for Lie subalgebras of \(\operatorname{Lie}(G)\), and we show that if the closed subgroup \(H ⊂ G\) is \(G\)-cr, then \(\operatorname{Lie}(H)\) is \(G\)-cr as well.

*Completely reducible SL(2) homomorphisms*, 2007. With Donna M. Testerman (EPFL).**Citation:***Transactions of the American Mathematical Society*359(9), 2007, pp. 4489--4510 (electronic).**URL(s):**[DOI:10.1090/S0002-9947-07-04289-4][arXiv:math.RT/0510377][e-print][MR:2309195][BibTeX][AMSRefs]**Abstract:**Let \(K\) be any field, and let \(G\) be a semisimple group over \(K\). Suppose the characteristic of \(K\) is positive and is very good for \(G\). We describe all group scheme homomorphisms \(ϕ:SL(2) → G\) whose image is geometrically \(G\) completely reducible -- or \(G\)-cr -- in the sense of Serre; the description resembles that of irreducible modules given by Steinberg's tensor product theorem. In case \(K\) is algebraically closed and \(G\) is simple, the result proved here was previously obtained by Liebeck and Seitz using different methods. A recent result shows the Lie algebra of the image of ϕ to be geometrically \(G\)-cr; this plays an important role in our proof.

*On the centralizer of the sum of commuting nilpotent elements*, 2006.**Citation:***Journal of Pure and Applied Algebra*206(1-2), 2006, pp. 123--140.**URL(s):**[DOI:10.1016/j.jpaa.2005.04.016][arXiv:math.RT/0412283][e-print][MR:2220085][BibTeX][AMSRefs]**Abstract:**Let \(X\) and \(Y\) be commuting nilpotent \(K\) endomorphisms of a vector space \(V\), where \(K\) is a field of characteristic \(p ≥ 0\). If \(F=K(t)\) is the field of rational functions on the projective line \(\mathbf{P}^1_K\), consider the \(K(t)\) endomorphism \(A=X+tY\) of \(V\). If \(p=0\), or if \(A^{p-1}=0\), we show here that \(X\) and \(Y\) are tangent to the unipotent radical of the centralizer of \(A\) in \(\operatorname{GL}(V)\). For all geometric points \((a:b)\) of a suitable open subset of \(\mathbf{P}^1\), it follows that \(X\) and \(Y\) are tangent to the unipotent radical of the centralizer of \(aX + bY\). This answers a question of J. Pevtsova.

*Optimal SL(2) homomorphisms*, 2005.**Citation:***Commentarii Mathematici Helvetici*80(2), 2005, pp. 391--426.**Abstract:**Let \(G\) be a semisimple group over an algebraically closed field of

*very good*characteristic for \(G\). In the context of geometric invariant theory, G. Kempf has associated optimal cocharacters of \(G\) to an unstable vector in a linear \(G\)-representation. If the nilpotent element \(X ∈ \operatorname{Lie}(G)\) lies in the image of the differential of a homomorphism \(\operatorname{SL}(2) → G\), we say that homomorphism is optimal for \(X\), or simply optimal, provided that its restriction to a suitable torus of \(\operatorname{SL}(2)\) is optimal for X in Kempf's sense.We show here that any two \(\operatorname{SL}(2)\) homomorphisms which are optimal for \(X\) are conjugate under the connected centralizer of \(X\). This implies, for example, that there is a unique conjugacy class of

*principal homomorphisms*for \(G\). We show that the image of an optimal \(\operatorname{SL}(2)\) homomorphism is a*completely reducible*subgroup of \(G\); this is a notion defined recently by J-P. Serre. Finally, if \(G\) is defined over the (arbitrary) subfield \(K\) of \(k\), and if \(X ∈ \operatorname{Lie}(G)(K)\) is a \(K\)-rational nilpotent element with \(X^{[p]}=0\), we show that there is an optimal homomorphism for \(X\) which is defined over \(K\).*Nilpotent orbits over ground fields of good characteristic*, 2004.**Citation:***Mathematische Annalen*329(1), 2004, pp. 49--85.**URL(s):**[DOI:10.1007/s00208-004-0510-9][arXiv:math.RT/0209151][e-print][MR:2052869][BibTeX][AMSRefs]**Abstract:**Let \(X\) be an \(F\)-rational nilpotent element in the Lie algebra of a connected and reductive group \(G\) defined over the ground field \(F\). Suppose that the Lie algebra has a non-degenerate invariant bilinear form. We show that the unipotent radical of the centralizer of X is \(F\)-split. This property has several consequences. When \(F\) is complete with respect to a discrete valuation with either finite or algebraically closed residue field, we deduce a uniform proof that \(G(F)\) has finitely many nilpotent orbits in \(\operatorname{Lie}(G)(F)\). When the residue field is finite, we obtain a proof that nilpotent orbital integrals converge. Under some further (fairly mild) assumptions on \(G\), we prove convergence for arbitrary orbital integrals on the Lie algebra and on the group. The convergence of orbital integrals in the case where \(F\) has characteristic 0 was obtained by Deligne and Ranga Rao (1972).

*Adjoint Jordan Blocks*, 2003.**Citation:**Unpublished manuscript.**Abstract:**Let \(G\) be a quasisimple algebraic group over an algebraically closed field of characteristic \(p>0\). We suppose that \(p\) is

*very good*for G; since \(p\) is good, there is a bijection between the nilpotent orbits in the Lie algebra and the unipotent classes in \(G\). If the nilpotent \(X ∈ \operatorname{Lie}(G)\) and the unipotent \(u ∈ G\) correspond under this bijection, and if \(u\) has order \(p\), we show that the partitions of \(\operatorname{ad}(X)\) and \(\operatorname{ad}(u)\) are the same. When G is classical or of type \(G_2\), we prove this result with no assumption on the order of \(u\). In the cases where \(u\) has order \(p\), the result is achieved through an application of results of Seitz concerning good \(A_1\) subgroups of \(G\). For classical groups, the techniques are more elementary, and they lead also to a new proof of the following result of Fossum: the structure constants of the representation ring of a 1-dimensional formal group law \(F\) are independent of \(F\).*Component groups of unipotent centralizers in good characteristic*, 2003. With Eric Sommers (UMass Amherst).**Citation:***Journal of Algebra*260(1), 2003, pp. 323--337.**URL(s):**[DOI:10.1016/S0021-8693(02)00661-0][arXiv:math.RT/0204275][e-print][MR:1976698][BibTeX][AMSRefs]**Abstract:**Let \(G\) be a connected, reductive group over an algebraically closed field of good characteristic. For \(u ∈ G\) unipotent, we describe the conjugacy classes in the component group \(A(u)\) of the centralizer of \(u\). Our results extend work of the second author done for simple, adjoint \(G\) over the complex numbers.

When \(G\) is simple and adjoint, the previous work of the second author makes our description combinatorial and explicit; moreover, it turns out that knowledge of the conjugacy classes suffices to determine the group structure of \(A(u)\). Thus we obtain the result, previously known through case-checking, that the structure of the component group \(A(u)\) is independent of good characteristic.

*Faithful representations of SL(2) over truncated Witt vectors*, 2003.**Citation:***Journal of Algebra*265(2), 2003, pp. 606--618.**URL(s):**[DOI:10.1016/S0021-8693(03)00269-2][arXiv:math.RT/0109107][e-print][MR:1987019][BibTeX][AMSRefs]**Abstract:**Let \(Γ_2\) be the six dimensional linear algebraic \(k\)-group \(\operatorname{SL}(2)(W_2)\), where \(W_2\) is the ring of Witt vectors of length two over the algebraically closed field \(k\) of characteristic \(p>2\). Then the minimal dimension of a faithful rational k-representation of \(Γ_2\) is \(p+3\).

*Sub-principal homomorphisms in positive characteristic*, 2003.**Citation:***Mathematische Zeitschrift*244(2), 2003, pp. 433--455.**Abstract:**Let \(G\) be a reductive group over an algebraically closed field of characteristic \(p\), and let \(u ∈ G\) be a unipotent element of order \(p\). Suppose that \(p\) is a good prime for \(G\). We show in this paper that there is a homomorphism \(ϕ:\operatorname{SL}(2)_k → G\) whose image contains \(u\). This result was first obtained by D. Testerman (J. Algebra, 1995) using case considerations for each type of simple group (and using, in some cases, computer calculations with explicit representatives for the unipotent orbits).

The proof we give is free of case considerations (except in its dependence on the Bala-Carter theorem). Our construction of ϕ generalizes the construction of a principal homomorphism made by J.-P. Serre in (Invent. Math. 1996); in particular, ϕ is obtained by reduction modulo \(\mathfrak{P}\) from a homomorphism of group schemes over a valuation ring \(A\) in a number field. This permits us to show moreover that the weight spaces of a maximal torus of \(ϕ(\operatorname{SL}(2)))\) on \(\operatorname{Lie}(G)\) are “the same as in characteristic 0”; the existence of a ϕ with this property was previously obtained, again using case considerations, by Lawther and Testerman (Memoirs AMS, 1999) and has been applied in some recent work of G. Seitz (Invent. Math. 2000).

*Abelian unipotent subgroups of reductive groups*, 2002.**Citation:***Journal of Pure and Applied Algebra*167(2-3), 2002, pp. 269--300.**URL(s):**[DOI:10.1016/S0022-4049(01)00038-X][arXiv:math.RT/0007056][e-print][MR:1874545][BibTeX][AMSRefs]**Abstract:**Let \(G\) be a connected reductive group defined over an algebraically closed field \(k\) of characteristic \(p > 0\). The purpose of this paper is two-fold. First, when \(p\) is a good prime, we give a new proof of the “order formula” of D. Testerman for unipotent elements in \(G\); moreover, we show that the same formula determines the \(p\)-nilpotence degree of the corresponding nilpotent elements in the Lie algebra \(\operatorname{Lie}(G)\) of \(G\).

Second, if \(G\) is semisimple and \(p\) is sufficiently large, we show that \(G\) always has a faithful representation \((ρ,V)\) with the property that the exponential of \(dρ(X)\) lies in \(ρ(G)\) for each \(p\)-nilpotent \(X ∈ \operatorname{Lie}(G)\). This property permits a simplification of the description given by Suslin, Friedlander, and Bendel of the (even) cohomology ring for the Frobenius kernels \(G_d\), \(d ≥ 2\). The previous authors already observed that the natural representation of a classical group has the above property (with no restriction on \(p\)). Our methods apply to any Chevalley group and hence give the result also for quasisimple groups with “exceptional type” root systems. The methods give explicit sufficient conditions on \(p\); for an adjoint semisimple G with Coxeter number \(h\), the condition \(p > 2h-2\) is always good enough.

*The second cohomology of small irreducible modules for simple algebraic groups*, 2002.**Citation:***Pacific Journal of Mathematics*204(2), 2002, pp. 459--472.**Abstract:**Let \(G\) be a connected, simply connected, quasisimple algebraic group over an algebraically closed field of characteristic \(p>0\), and let \(V\) be a rational \(G\)-module such that \(\dim V ≤ p\). According to a result of Jantzen, \(V\) is completely reducible, and \(H^1(G,V)=0\). In this paper we show that \(H^2(G,V) = 0\) unless some composition factor of \(V\) is a non-trivial Frobenius twist of the adjoint representation of \(G\).

*Filtrations and positive characteristic Howe duality*, 2000.**Citation:***Mathematische Zeitschrift*235(4), 2000, pp. 651--685.*Semisimplicity of exterior powers of semisimple representations of groups*, 2000.**Citation:***Journal of Algebra*225(2), 2000, pp. 646--666.*Semisimple modules for finite groups of Lie type*, 1999.**Citation:***Journal of the London Mathematical Society*60(3), 1999, pp. 771--792.*Dimensional criteria for semisimplicity of representations*, 1998.**Citation:***Proceedings of the London Mathematical Society*76(1), 1998, pp. 95--149.*Semisimplicity in positive characteristic*, 1998.**Citation:***Algebraic groups and their representations, NATO Adv. Sci. Inst. Ser. C Math. Phys. Sci. 517, 1998, pp. 43--52*.